Safety evaluation of certain food additives and contaminants, supplement 1: non-dioxin-like polychlorinated biphenyls, prepared by the eightieth meeting of the Joint FAO/WHO Expert Committee on Food Additives (JECFA)

WHO FOOD ADDITIVES SERIES: 71-S1 Prepared by the eightieth meeting of the Joint FAO/WHO Expert Committee on Food Additives (JECFA)

Safety evaluation of certain food additives and contaminants Supplement 1:

Non-dioxin-like polychlorinated biphenyls

WHO FOOD ADDITIVES SERIES: 71-S1 Prepared by the eightieth meeting of the Joint FAO/WHO Expert Committee on Food Additives (JECFA)

Safety evaluation of certain food additives and contaminants Supplement 1:

Non-dioxin-like polychlorinated biphenyls

World Health Organization, Geneva, 2016

WHO Library Cataloguing-in-Publication Data Safety evaluation of certain food additives and contaminants, supplement 1: non-dioxin-like polychlorinated biphenyls / prepared by the eightieth meeting of the Joint FAO/WHO Expert Committee on Food Additives (JECFA). (WHO food additives series ; 71-S1) 1.Food Additives - toxicity. 2.Food Contamination. 3.Risk Assessment. I.Joint FAO/WHO Expert Committee on Food Additives. Meeting (80th : 2015 : Rome, Italy). II.World Health Organization. III.Series. ISBN 978 92 4 166171 3 ISSN 0300-0923 (NLM classification: WA 712)

© World Health Organization 2016 All rights reserved. Publications of the World Health Organization are available on the WHO website (www.who.int) or can be purchased from WHO Press, World Health Organization, 20 Avenue Appia, 1211 Geneva 27, Switzerland (tel.: +41 22 791 3264; fax: +41 22 791 4857; email: bookorders@who.int). Requests for permission to reproduce or translate WHO publications – whether for sale or for non-commercial distribution – should be addressed to WHO Press through the WHO website (www.who.int/about/licensing/copyright_form/en/index.html). The designations employed and the presentation of the material in this publication do not imply the expression of any opinion whatsoever on the part of the World Health Organization concerning the legal status of any country, territory, city or area or of its authorities, or concerning the delimitation of its frontiers or boundaries. Dotted and dashed lines on maps represent approximate border lines for which there may not yet be full agreement. The mention of specific companies or of certain manufacturers’ products does not imply that they are endorsed or recommended by the World Health Organization in preference to others of a similar nature that are not mentioned. Errors and omissions excepted, the names of proprietary products are distinguished by initial capital letters All reasonable precautions have been taken by the World Health Organization to verify the information contained in this publication. However, the published material is being distributed without warranty of any kind, either expressed or implied. The responsibility for the interpretation and use of the material lies with the reader. In no event shall the World Health Organization be liable for damages arising from its use. Design: Rania Spatha (www.raniaspatha.com)

Contents Preface Non-dioxin-like polychlorinated biphenyls 1. Explanation 2. Biological data 3. Analytical methods 4. Sampling protocols 5. Effects of processing 6. Prevention and control 7. Levels and patterns of contamination of food commodities 8. Dietary exposure assessment 9. Modelling of body burden from dietary exposure 10. Dose–response analysis 11. Comments 12. Evaluation 13. References Appendices v 1 6 11 169 180 182 185 188 200 243 254 258 277 279 336

Annex 1 Reports and other documents resulting from previous meetings of the Joint FAO/WHO Expert Committee on Food Additives 403 415

Annex 2 Abbreviations used in the monographs

Annex 3 Participants in the eightieth meeting of the Joint FAO/WHO Expert Committee on Food Additives 420

iii

Preface The monograph contained in this volume was prepared at the eightieth meeting of the Joint Food and Agriculture Organization of the United Nations (FAO)/World Health Organization (WHO) Expert Committee on Food Additives (JECFA), which met at FAO headquarters in Rome, Italy, on 16–25 June 2015. This monograph summarizes the data on one contaminant group reviewed by the Committee. Monographs on seven food additive groups discussed at the meeting have been published in WHO Food Additives Series 71, and a monograph on a second contaminant group will be published as a separate supplement in WHO Food Additives Series 71. The eightieth report of JECFA has been published by WHO as WHO Technical Report No. 995. Reports and other documents resulting from previous meetings of JECFA are listed in Annex 1. The participants in the meeting are listed in Annex 3 of the present publication. JECFA serves as a scientific advisory body to FAO, WHO, their Member States and the Codex Alimentarius Commission, primarily through the Codex Committee on Food Additives, the Codex Committee on Contaminants in Food and the Codex Committee on Residues of Veterinary Drugs in Foods, regarding the safety of food additives, residues of veterinary drugs, naturally occurring toxicants and contaminants in food. Committees accomplish this task by preparing reports of their meetings and publishing specifications or residue monographs and dietary exposure and toxicological monographs, such as that contained in this volume, on substances that they have considered. The monograph contained in this volume is based on a working paper that was prepared by JECFA experts. A special acknowledgement is given at the beginning of the monograph to those who prepared this working paper. The monograph was edited by M. Sheffer, Ottawa, Canada. The designations employed and the presentation of the material in this publication do not imply the expression of any opinion whatsoever on the part of the organizations participating in WHO concerning the legal status of any country, territory, city or area or its authorities, or concerning the delimitation of its frontiers or boundaries. The mention of specific companies or of certain manufacturers’ products does not imply that they are endorsed or recommended by the organizations in preference to others of a similar nature that are not mentioned. Any comments or new information on the biological or toxicological properties of or dietary exposure to the compounds evaluated in this publication should be addressed to: WHO Joint Secretary of the Joint FAO/WHO Expert Committee on Food Additives, Department of Food Safety and Zoonoses, World Health Organization, 20 Avenue Appia, 1211 Geneva 27, Switzerland.

v

Non-dioxin-like polychlorinated biphenyls First draft prepared by Susan M. Barlow,1 Antonio Agudo,2 Billy Amzal,3 Camille Béchaux,3 Genevieve Bondy,4 Peter Cressey,5 Milou M.L. Dingemans,6 Mark Feeley,4 Helen Håkansson,7 Tracy Hambridge,8 Dorothea F.K. Rawn,4 Klaus Schneider,9 Martin van den Berg6 and Yongning Wu10 Brighton, East Sussex, England, United Kingdom Catalan Institute of Oncology, L’Hospitalet de Llobregat, Spain 3 LASER Analytica, London, England, United Kingdom 4 Bureau of Chemical Safety, Food Directorate, Health Canada, Ottawa, Canada 5 Food Programme, Institute of Environmental Science and Research, Christchurch, New Zealand 6 Institute for Risk Assessment Sciences, Utrecht University, Utrecht, the Netherlands 7 Institute of Environmental Medicine, Karolinska Institutet, Stockholm, Sweden 8 Food Data Analysis Section, Food Standards Australia New Zealand, Canberra, Australia 9 Forschungs- und Beratungsinstitut Gefahrstoffe GmbH (FoBiG), Freiburg, Germany 10 China National Center for Food Safety Risk Assessment, Beijing, China 1 2

1. Explanation 1.1 Introduction 1.2 Compounds considered and nomenclature 1.3 General considerations on exposure sources and exposure measurements 2. Biological data 2.1 Biochemical aspects 2.1.1 Absorption, distribution and elimination 2.1.2 Biotransformation 2.1.3 Receptor interactions and relationship to toxicity 2.1.4 Effects on enzyme activities 2.2 Toxicological studies 2.2.1 Acute toxicity 2.2.2 Short-term studies of toxicity (a) Mice (b) Rats 2.2.3 Long-term studies of toxicity and carcinogenicity (a) Commercial mixtures of PCBs (b) Individual NDL-PCB congeners (c) Hydroxylated PCBs (d) Studies on tumour promotion 2.2.4 Genotoxicity (a) Commercial mixtures of PCBs (b) General considerations on PCB metabolites (c) Studies on individual NDL-PCBs and their metabolites 2.2.5 Reproductive and developmental toxicity

6 6 7 10 11 11 11 14 16 18 20 20 20 20 21 34 34 35 39 39 43 43 43 45 48 1

Safety evaluation of certain food additives and contaminants Eightieth JECFA

2.3

(a) Reproductive studies on commercial mixtures of PCBs 48 (b) Reproductive studies on individual NDL-PCBs and their metabolites 49 (c) Developmental toxicity studies on commercial mixtures of PCBs 50 (d) Developmental toxicity studies on individual NDL-PCBs and their metabolites 51 (e) Developmental neurotoxicity: mechanistic aspects 56 (f) Developmental neurotoxicity studies 58 2.2.6 Special studies 69 (a) Adult neurotoxicity 69 (b) Immunological effects 74 Observations in humans 79 2.3.1 Biomonitoring 79 (a) Biomarkers for NDL-PCBs 79 (b) Concentrations of NDL-PCBs in blood, plasma and serum 80 (c) Concentrations of NDL-PCBs in adipose tissue 89 (d) Concentrations of NDL-PCBs in breast milk 92 (e) Age trends for body burden of NDL-PCBs 95 (f) Time trends for body burden of NDL-PCBs 95 (g) Concentrations of NDL-PCBs in more highly exposed populations 96 (h) Concentrations of NDL-PCB metabolites in blood 99 2.3.2 General considerations on health effects 100 2.3.3 Mortality 103 2.3.4 Developmental toxicity: birth/gestational outcomes 104 (a) Studies on exposure to undefined PCB congeners 105 (b) Studies with specific results for NDL-PCBs 107 2.3.5 Neurotoxicity and behaviour 110 (a) Neurodevelopmental effects (infants and children) 110 (b) Neurophysiological and neuropsychological effects in adolescents 118 (c) Neurophysiological and neuropsychological effects in adults 119 2.3.6 Cancer 121 (a) Prospective studies: nested case–control studies of PCBs in blood or adipose tissue 122 (b) Case–control studies 128 2.3.7 Endocrine and metabolic effects 137 (a) Thyroid function and thyroid diseases 137 (b) Diabetes 141 (c) Obesity and metabolic-related conditions 147 (d) Insulin resistance and metabolic syndrome 148 2.3.8 Reproductive and sexual function 149 (a) Male reproductive function 150 (b) Female reproductive function 153 (c) Effects on reproductive and sexual function assessed in both males and females 154 2.3.9 Immune function and related outcomes 155

WHO Food Additives Series No. 71-S1, 2016 2

Supplement 1: Non-dioxin-like polychlorinated biphenyls

2.3.10 Cardiovascular outcomes (a) Blood pressure or prevalence of hypertension (b) Other cardiovascular outcomes 2.3.11 Respiratory outcomes 2.3.12 Hepatic effects 2.3.13 Musculoskeletal effects 2.3.14 Endometriosis 2.3.15 Other health effects 2.3.16 Summary of epidemiological studies 3. Analytical methods 3.1 Chemistry 3.2 Description of analytical method 3.2.1 Introduction 3.2.2 Screening tests 3.2.3 Quantitative methods 3.2.4 Quality assurance considerations 3.2.5 Reference methods 4. Sampling protocols 5. Effects of processing 6. Prevention and control 7. Levels and patterns of contamination of food commodities 8. Dietary exposure assessment 8.1 Introduction and background 8.2 Methods 8.2.1 Overview of the methods for the NDL-PCBs exposure assessment 8.2.2 How the GEMS/Food concentration data were used for the dietary exposure assessment 8.2.3 National estimates of dietary exposure using the FAO/WHO Chronic Individual Food Consumption database – summary statistics (CIFOCOss) 8.2.4 International estimates of dietary exposure using the GEMS/Food cluster diets 8.3 Estimates of dietary exposure 8.3.1 National estimates of dietary exposure (a) Austria (b) Belgium (c) China (d) Cyprus (e) Czech Republic (f) Denmark (g) Egypt (h) Finland (i) France (j) Germany (k) Greece (l) Ireland

157 158 160 160 163 163 164 165 165 169 169 171 171 172 172 178 179 180 182 185 188 200 200 200 200 202 206 207 209 209 216 216 217 218 218 219 219 219 220 221 221 222 3

Safety evaluation of certain food additives and contaminants Eightieth JECFA

(m) Italy (n) Netherlands (o) Norway (p) Republic of Korea (q) Serbia (r) Slovakia (s) Spain (t) Sweden (u) Turkey (v) United Kingdom 8.3.2 National estimates of dietary exposure for individual indicator PCBs (a) Estimated dietary exposure to the six indicator PCBs individually (b) PCB 128 8.3.3 International estimates of dietary exposure 8.3.4 Dietary exposures for infants (a) Estimated dietary exposure for breastfed infants (b) Estimated dietary exposure for fully formula-fed infants 8.3.5 Dietary exposures for specific population subgroups 8.4 Other routes of exposure 8.5 Temporal trends in dietary exposure 8.6 Limitations and uncertainties associated with the exposure estimates 9. Modelling of body burden from dietary exposure 9.1 Dynamic modelling of exposure from food 9.1.1 Model parameters (a) Absorption (b) Half-life (c) Dietary exposure 9.2 Estimates of body burden 9.3 Contribution of congeners to the total body burden 9.4 Comparison with biomonitoring data 9.5 Uncertainty around the estimated body burdens (scenario 2) 10. Dose–response analysis 11. Comments 11.1 Toxicokinetics and mode of action 11.2 Toxicological data 11.2.1 Acute toxicity and short-term studies of toxicity 11.2.2 Long-term studies of toxicity and carcinogenicity 11.2.3 Genotoxicity 11.2.4 Reproductive and developmental toxicity 11.2.5 Immunological studies 11.3 Observations in humans 11.3.1 Biomonitoring and modelling of body burden 11.3.2 Epidemiology 11.4 Analytical methods 11.5 Sampling protocols

222 223 224 224 224 225 225 225 226 226 226 226 227 229 234 234 236 236 238 239 240 243 243 244 244 244 245 245 252 252 253 254 258 258 260 260 261 262 263 263 264 264 265 267 268

WHO Food Additives Series No. 71-S1, 2016 4

Supplement 1: Non-dioxin-like polychlorinated biphenyls

11.6 11.7 11.8 11.9

Effects of processing Prevention and control Levels and patterns of food contamination Dietary exposure assessment 11.9.1 National estimates of dietary exposure 11.9.2 International estimates of dietary exposure 11.9.3 Dietary exposure of infants 11.9.4 Contribution of individual congeners to total exposures from all sources 11.10 Estimation of margins of exposure

268 268 269 270 271 273 273 273 274 277 277 279 336 347

12. Evaluation 12.1 Recommendations 13. References Appendix 1. Indicator PCB concentrations for each food category tested by country Appendix 2. Mean indicator PCB concentrations in foods reported by countries responding to JECFA call for data Appendix 3. Summary of the concentration data for food groups and major contributors to dietary exposure for the national dietary exposure estimates for the sum of the six indicator PCB congeners estimated by the Committee using the CIFOCOss consumption data and concentration data from the GEMS/Food database Appendix 4. Details of the national dietary exposure for the six indicator PCB congeners individually as estimated by the Committee using the CIFOCOss consumption data and concentration data from the GEMS/Food database Appendix 5. Dietary exposure calculations for each cluster, including consumption, concentration and dietary exposure Appendix 6. GEMS/Food cluster diets 2012 – Countries by cluster

353 372 381 402

5

Safety evaluation of certain food additives and contaminants Eightieth JECFA

1. Explanation 1.1 Introduction

WHO Food Additives Series No. 71-S1, 2016 6

Polychlorinated biphenyls (PCBs) are chemically stable aromatic chlorinated hydrocarbons. They were first produced commercially around 1930 and for the next 5 decades found a wide range of industrial applications as a result of their physicochemical properties of low electrical conductance, fire resistance, resistance to thermal breakdown and chemical inertness. Their main uses included dielectric fluids in electrical equipment such as transformers and capacitors, heat transfer agents in mechanical operations, plasticizers (e.g. in carbonless copy paper), lubricants, inks and surface coatings. The manufacture, distribution and use of PCBs have been widely discontinued or banned, but PCBs may be found in equipment still in use today. Environmental contamination by PCBs from open, partially closed or closed uses and from disposal has been widespread (UNEP, 1999). The most abundant PCBs are readily biodegradable. However, some PCBs are very persistent in the environment; hence, they are present as contaminants, especially in fatty foods, and they bioaccumulate in the adipose tissue of exposed animals and humans. As a consequence of environmental contamination by PCBs and their toxicity, many countries restricted the marketing and use of PCBs in the 1970s and 1980s. In 2001, PCBs were classified as persistent organic pollutants (POPs) under the Stockholm Convention on POPs; signatories agreed to ban all production of PCBs, to promote control and reduction of exposures and risks and to eliminate all uses by 2025 (Stockholm Convention, 2009). The Committee was requested to undertake an assessment of the nondioxin-like polychlorinated biphenyls (NDL-PCBs) by the Codex Committee on Contaminants in Foods. The Committee has not previously evaluated NDL-PCBs specifically. The Committee previously reviewed PCBs at its thirty-fifth meeting, when it concluded that it was impossible to establish a precise numerical value for a tolerable intake in humans because of limitations in the available data and the ill-defined nature of the materials that were used in feeding studies (Annex 1, reference 88). Dioxin-like PCBs (DL-PCBs), together with polychlorinated dibenzo-p-dioxins (PCDDs) and polychlorinated dibenzofurans (PCDFs), were reviewed by the Committee at its fifty-seventh meeting (Annex 1, reference 154). NDL-PCBs were comprehensively reviewed in 2005 by the European Food Safety Authority (EFSA, 2005) and by the United States Agency for Toxic Substances and Disease Registry (ATSDR, 2000, 2011), and the present Committee used these reviews as a starting point for its evaluation, taking particular account of new studies published subsequent to the reviews.

Supplement 1: Non-dioxin-like polychlorinated biphenyls

1.2 Compounds considered and nomenclature

PCBs are a class of chemicals that have a biphenyl structure of two linked benzene rings in which 1–10 chlorine atoms substitute the hydrogen atoms on the rings (Erickson, 1986) (Fig. 1). There are 209 possible congeners in total, based on the substitution positions along the phenyl rings. Fig. 1 General structure of PCBs

where x + y = 1–10

PCBs were manufactured as complex mixtures of congeners by the progressive chlorination of batches of biphenyl until a target percentage of chlorine by weight was achieved. Of the 209 congeners that are theoretically possible, only about 130 have been identified in commercial products that were marketed. Commercial PCBs were sold not as specified compositions of congeners, but on the basis of their physical properties, in particular per cent chlorination and molecular weight. The absolute congener composition of commercial PCB mixtures, as well as the content of impurities, such as PCDFs, naphthalenes and quaterphenyls, varied from batch to batch. To standardize the identification of the individual PCB congeners, a numbering system was developed by Ballschmiter & Zell (1980), following the International Union of Pure and Applied Chemistry (IUPAC) rules for characterization. A couple of octachlorinated congeners were initially misnumbered, but these were subsequently corrected (Ballschmiter, Schäfer & Buchert, 1987). In this scheme, a number, called the “BZ number”, is attributed to each individual congener. This number correlates the structural arrangement of the PCB congener and ascending order of number of chlorine substitutions within each sequential homologue, as shown in Table 1. Thus, congeners are numbered from PCB 1 to PCB 209, a useful shorthand nomenclature. However, it is important to note that it obscures the chemical identity of the congener and does not strictly follow the IUPAC rules. 7

8 Safety evaluation of certain food additives and contaminants Eightieth JECFA 3,4 12 33 35 37 56 66 70 71 77 3,5 14 34 36 39 58 68 72 73 79 80 2,3,4 21 41 55 60 82 85 87 89 105 107 128 2,3,5 23 43 57 63 83 90 92 94 109 111 130 133 2,3,6 24 45 59 64 84 91 95 96 110 113 132 135 136 2,4,5 29 48 67 74 97 99 101 102 118 120 138 146 149 153 2,4,6 30 50 69 75 98 100 103 104 119 121 140 148 150 154 155 3,4,5 38 76 78 81 122 123 124 125 126 127 157 162 164 167 168 169 2,3,4,5 61 86 106 114 129 137 141 143 156 159 170 172 174 180 182 189 194 2,3,4,6 62 88 108 115 131 139 144 145 158 161 171 175 176 183 184 191 196 197 2,3,5,6 65 93 112 117 134 147 151 152 163 165 177 178 179 187 188 193 199 201 202 2,3,4,5,6 116 142 160 166 173 181 185 186 190 192 195 198 200 203 204 205 206 207 208 209

WHO Food Additives Series No. 71-S1, 2016

Table 1 PCB congeners showing the BZ numbera and correspondence between the positions of chlorine atoms on each phenyl ring of the PCBsb

Position of chlorine atom on each ring None 2′ 3′ 4′ 2′,3′ 2′,4′ 2′,5′ 2′,6′ 3′,4′ 3′,5′ 2′,3′,4′ 2′,3′,5′ 2′,3′,6′ 2′,4′,5′ 2′,4′,6′ 3′,4′,5′ 2′,3′,4′,5′ 2′,3′,4′,6′ 2′,3′,5′,6′ 2′,3′,4′,5′,6′

2 1 4

3 2 6 11

4 3 8 13 15

2,3 5 16 20 22 40

2,4 7 17 25 28 42 47

2,5 9 18 26 31 44 49 52

2,6 10 19 27 32 46 51 53 54

The revised PCB numbering system, including the revised numbering of congeners 107–109 and 199–201. For a number of PCB congeners, the indicated (truncated) structural names do not strictly adhere to the IUPAC rules (primed and unprimed numbers are interchanged). A comprehensive survey of PCB nomenclature, including IUPAC names, is given in Mills, Thal & Barney (2007). The names of the PCB congeners can be obtained directly from the table. For example, PCB 74 is 2,4,4′,5-tetrachlorobiphenyl, PCB 99 is 2,2′,4,4′,5-pentachlorobiphenyl, PCB 138 is 2,2′,3,4,4′,5′-hexachlorobiphenyl, PCB 153 is 2,2′,4,4′,5,5′-hexachlorobiphenyl, PCB 170 is 2,2′,3,3′,4,4′,5-heptachlorobiphenyl, PCB 180 is 2,2′,3,4,4′,5,5′heptachlorobiphenyl, PCB 194 is 2,2′,3,3′,4,4′,5,5′-octachlorobiphenyl, PCB 206 is 2,2′,3,3′,4,4′,5,5′,6-nonachlorobiphenyl and PCB 209 is 2,2′,3,3′,4,4′,5,5′,6,6′-decachlorobiphenyl. The PCB congeners will be referred to only by number in the remainder of this monograph. b Dioxin-like PCBs are indicated in grey shading and bold type. Source: Adapted from IARC (2015)

a

Supplement 1: Non-dioxin-like polychlorinated biphenyls

International bodies have identified seven PCBs that can be used to characterize the presence of PCB contamination. Six of these seven are NDLPCBs (PCB 28, PCB 52, PCB 101, PCB 138, PCB 153 and PCB 180), and one is a DL-PCB (PCB 118). These seven PCBs are often called “indicator PCBs” (European Commission, 1999). In this monograph, which is concerned only with NDL-PCBs, the term “indicator PCBs” includes only the six NDL-PCBs. PCBs exhibit different toxicological effects depending on the site of chlorine substitution on the phenyl rings. Congeners with no chlorine substitution in the ortho position (PCBs 77, 81, 126 and 169) or those congeners having only ortho substitution in one position (i.e. mono-ortho-substituted PCBs 105, 114, 118, 123, 156, 157, 167 and 189) have toxicological activity similar to that of the PCDDs and PCDFs owing to their ability to adopt a similar planar structure (see Fig. 2a) and to bind strongly to the aryl hydrocarbon receptor (AhR). Hence, these 12 PCBs are referred to as DL-PCBs. The remaining 197 congeners – that is, those not conforming to the planar structure (see Fig. 2b) – are referred to as NDL-PCBs. The NDL-PCBs have different toxicological activity compared with the DL-PCBs and PCDDs/PCDFs, the end-points most sensitive to NDL-PCB exposure being toxicity to the liver and thyroid (Bjermo et al., 2013). A few of the NDL-PCBs have hybrid activity, showing both dioxin-like and non-dioxin-like activities. In the present evaluation, only those congeners with non-dioxin-like activity are considered. Fig. 2 Examples of a) a planar PCB (PCB 77) and b) a non-planar, di-ortho-substituted PCB (PCB 101) a) b)

Commercial mixtures of PCBs have been marketed in the past with various tradenames, depending on the country (e.g. Aroclors, Kanechlors, Clophens, Fenclors, Phenoclors). They comprise mainly coplanar DL-PCBs and are not further discussed, except in some toxicity studies in which their effects are compared with those of NDL-PCBs. 9

Safety evaluation of certain food additives and contaminants Eightieth JECFA

1.3 General considerations on exposure sources and exposure

measurements

The major source of exposure to PCBs for the general population is through consumption of contaminated foods in the diet, which accounts for more than 90% of total exposure. NDL-PCBs account for the majority of the total PCB contamination in food, the remainder being DL-PCBs. Dermal and inhalation routes of exposure are of minor importance, except for occupationally exposed individuals (ATSDR, 2000; EFSA, 2005). Some lower chlorinated PCB congeners are readily metabolized, but some higher chlorinated congeners are more stable and accumulate within the foodchain, particularly in foods of animal origin. Fish and fish products, including fish oils, generally contain the highest concentrations of PCBs, followed by milk, eggs and dairy products and meat and meat products. Cereals and cereal products, fruits and vegetables contain only low amounts of PCBs. Breastfeeding is a major route of exposure for infants. Ingestion of contaminated soil or dust can be a minor route of exposure for children (EFSA, 2005; Elabbas et al., 2013). The Stockholm Convention on POPs recommends measurement of the six indicator PCBs (PCB 28, PCB 52, PCB 101, PCB 138, PCB 153 and PCB 180) to characterize contamination by PCBs (UNEP, 2013). These are all NDL-PCB congeners, and they were chosen because they are found at high concentrations in the environment, in food or in human fluids/tissues.

WHO Food Additives Series No. 71-S1, 2016 10

2. Biological data 2.1 Biochemical aspects 2.1.1 Absorption, distribution and elimination

PCB congeners are lipid soluble. They are generally well absorbed from the gastrointestinal tract by passive diffusion in laboratory rodents, monkeys and humans. Absorption and excretion of PCBs from the diet are congener specific; the major determinants of the amount of a particular congener that is absorbed and excreted are the existing concentration in blood and the body burden of the congener at the time of the exposure. In rats, lower chlorinated congeners with six or fewer chlorine atoms have greater than 90% absorption, and higher chlorinated congeners have about 75% absorption (Albro & Fishbein, 1972; ATSDR, 2000; EFSA, 2005). Dietary exposures are to mixtures of PCB congeners. The profile of PCB congeners in human serum immediately following an exposure reflects that of the exposure source, but the profile begins to change within 4–24 hours as a result of selective metabolism, excretion and deposition. Thus, in most cases, the PCB profile in adults represents a steady-state body burden that does not match the profile of commercial PCB mixture formulations (ATSDR, 2000). PCBs are rapidly distributed to all body compartments and particularly to highly perfused areas, such as liver and muscle. The toxicokinetics of PCBs is similar in humans and experimental animals. However, the rates of metabolism and excretion may be slower in humans, as indicated by the longer half-lives of certain congeners in humans (Chen et al., 1982; Bühler, Schmid & Schlatter, 1988). Some PCBs have apparent half-lives in blood as short as a week or so, but the high lipid solubility of many PCBs results in much longer half-lives, with retention and accumulation in adipose tissue. Higher chlorinated PCB congeners with only isolated non-chlorinated carbons show the longest halflives, and therefore the greatest accumulation; for example, PCB 138, PCB 153 and PCB 180 have half-lives of several years in humans, as shown in Table 2. It should be noted that Table 2 shows apparent half-lives. These reflect the overall effect of intrinsic elimination, ongoing exposure and body weight changes on concentrations as a function of time (Ritter et al., 2011). As can be seen from Table 2, apparent half-lives are subject to considerable variability, much more so than intrinsic half-lives, which reflect only interindividual variability of intrinsic elimination at similar concentrations. Ritter et al. (2011) proposed intrinsic halflives at background levels for several PCBs and recommended the following values for five of the six indicator PCBs: PCB 28, 5.5 years; PCB 52, 2.6 years; 11

Safety evaluation of certain food additives and contaminants Eightieth JECFA

PCB 138, 10.8 years; PCB 153, 14.4 years; and PCB 180, 11.5 years. Ritter et al. (2011) recommended that these intrinsic half-lives be used to translate between exposure and body concentration when pharmacokinetic models are used. Table 2 Indicative apparent half-lives of the six indicator PCBs Study Yakushiji et al. (1984) Brown et al. (1989) Wolff, Fischbein & Selikoff (1992) Ritter et al. (2011) Wolff, Fischbein & Selikoff (1992) Ritter et al. (2011) Wolff, Fischbein & Selikoff (1992)a Chen et al. (1982) Chen et al. (1982)b Yakushiji et al. (1984) Brown et al. (1989) Wolff, Fischbein & Selikoff (1992) Ryan et al. (1993) Masuda (2001)c Masuda (2001)c Ritter et al. (2011) Chen et al. (1982) Chen et al. (1982)b Yakushiji et al. (1984) Brown et al. (1989) Ryan et al. (1993) Masuda (2001)c Masuda (2001)c Ritter et al. (2011) Wolff, Fischbein & Selikoff (1992) Ryan et al. (1993) Masuda (2001)c Masuda (2001)c Ritter et al. (2011) a b c

WHO Food Additives Series No. 71-S1, 2016

n 8 194 165 229 165 229 165 17 17 8 194 165 16 8 8 229 17 17 8 194 16 8 8 229 165 16 8 8 229

PCB 28 3.0 1.4 4.8 5.6 – – – – – – – – – – – – – – – – – – – – – – – – –

PCB 52 – – – – 5.5 2.6 – – – – – – – – – – – – – – – – – – – – – – –

Estimated half-life (years) PCB 101 PCB 138 – – – – – – – – – – – – 5.7 – – 32 – 20 – 16.3 – 6–7 – 16.7 – 3.4 – 4.5 – 12.8 – 8.4 – – – – – – – – – – – – – – – – – – – – – – – – – –

PCB 153 – – – – – – – – – – – – – – – – 47 26 27.5 12.4 3.8 4.2 9.1 13.8 – – – – –

PCB 180 – – – – – – – – – – – – – – – – – – – – – – – – 9.9 4.3 6.0 16.7 5.5

Co-elution with PCB 99. Recalculated by Shirai & Kissel (1996). Same patients (Yusho), but observations are from different time intervals after the exposure event. Source: Adapted from Ritter et al. (2011)

12

Supplement 1: Non-dioxin-like polychlorinated biphenyls

It can be seen from Table 2 that estimates of PCB half-lives in humans, derived from successive body burden measurements, vary widely. As discussed by Shirai & Kissel (1996), differences in physiological processes among individuals and in congener properties are to be expected, but these factors do not appear to explain all the variation. Very short half-lives (<1 year) are unlikely for congeners most frequently found in human blood, because the exposures required to sustain observed body burdens are too large. Very long half-lives (>10 years) may be artefacts of confounding by ongoing exposures, which is a common effect at low body burdens. PCB parent compounds and methyl sulfone PCB metabolites are lipophilic and are associated with lipoproteins in plasma and persistence in tissue lipids. Methyl sulfone metabolites are slightly less lipophilic than their parent compounds and are present only at low concentrations in human blood. However, for some methyl sulfone metabolites, their accumulation is cell and tissue specific (liver and lung), and some are present in adipose tissue at concentrations higher than those of their respective parent compounds. In contrast, the more polar hydroxy metabolites of PCBs are transported via blood proteins and are more readily excreted. The highest amounts of PCBs are usually found in the liver, fat, skin and breast milk (ATSDR, 2000; EFSA, 2005; Elabbas et al., 2013). Distribution of PCBs from the maternal to the fetal compartment is by passive diffusion across the placenta, and there is a correlation between maternal and cord serum concentrations (ATSDR, 2000). However, body burdens are lower in the fetus than in the mother because of the lower blood lipid and body fat content in the fetus (EFSA, 2005). In an ex vivo human placental transfer model, PCB 52 and PCB 180 were shown to transfer across the placenta within 2.5 hours, transfer of PCB 180 being more rapid (Correia Carreira et al., 2011). Hydroxy metabolites of PCBs are efficiently transferred from maternal to fetal blood via the placenta (Grimm et al., 2015). Postnatally, the amounts of PCB parent compounds and methyl sulfone PCB metabolites transferred to suckling animals are higher than the amounts transferred to the fetus, whereas transfer of the hydroxy-PCB metabolites through maternal milk is low (ATSDR, 2000; EFSA, 2005). In the breastfed human infant exposed to typical levels of NDL-PCBs in maternal milk, a rate of absorption of greater than 90% of the PCB content has been demonstrated (McLachlan, 1993; Abraham et al., 1994; Dahl et al., 1995; ATSDR, 2000). In human milk, concentrations of PCBs are highest in primiparous women and generally decline with duration of breastfeeding (ATSDR, 2000). The major routes of excretion are through the faeces for the PCB parent compounds and lipophilic methyl sulfone metabolites and through the urine and faeces for the hydroxy metabolites. For the majority of PCB excretion, biotransformation is required (ATSDR, 2000). There is significant elimination 13

Safety evaluation of certain food additives and contaminants Eightieth JECFA

of unchanged PCBs and their methyl sulfone metabolites via breast milk (EFSA, 2005). 2.1.2 Biotransformation

WHO Food Additives Series No. 71-S1, 2016 14

The metabolism of PCBs has been recently reviewed by Grimm et al. (2015). Rates of metabolism vary greatly across species. Rates of PCB metabolism also vary with the number and position of the chlorine atoms in the different congeners. In all species studied, PCB congeners with adjacent unsubstituted (vicinal) carbon atoms in the meta and para positions are more readily metabolized, whereas congeners without such adjacent unsubstituted carbon atoms are generally metabolized and cleared very slowly. PCBs with higher numbers of chlorine atoms are generally metabolized more slowly, and the PCBs that are not readily metabolized and cleared concentrate in adipose tissue. In humans, PCB 153 is often the most prevalent congener detected because of its occurrence in exposure media and its slow rate of biotransformation (Matthews & Dedrick, 1984; ATSDR, 2000). Biotransformation of PCBs involves oxidation by cytochrome P450 (CYP) enzymes. Exposure to PCBs generally induces the enzymes that metabolize them. NDL-PCBs are metabolized by CYP2B or by CYP2C and CYP3A, whereas DL-PCBs are metabolized by CYP1A (James, 2013; Quinete et al., 2014). NDL-PCBs have several routes of metabolism. They can be oxidized across the aromatic ring, the meta and para positions being the preferred sites, to one or more unstable, intermediate arene oxides. These can then spontaneously rearrange to produce a hydroxy metabolite. Arene oxides of PCBs are reactive electrophilic intermediates that may also form adducts to biomacromolecules (DNA and proteins) and to lipids. PCBs that oxidize to more stable arene oxides are subsequently reduced by epoxide hydrolase to dihydroxy metabolites, also known as dihydrodiols. Dihydrodiols can then be aromatized and form catechol metabolites, which are in equilibrium with their oxidized form, the corresponding hydroquinone and quinone. Hydroquinones and quinones are reactive intermediates with the potential for adduct formation (EFSA, 2005). Approximately 40 different hydroxy-PCBs have been identified in human blood. Hydroxy-PCB concentrations in human plasma or serum are in a range similar to those of many parent PCB congeners, except for those PCBs that are the most prevalent or persistent; for example, plasma concentrations of the most abundant hydroxy-PCB congeners reach about 30% of those determined for PCB 153, with a variation among studies of 11–82%. Thus, hydroxy-PCBs present at the highest concentrations always exceed the concentrations of a large number of individual PCB congeners (Grimm et al., 2015).

Supplement 1: Non-dioxin-like polychlorinated biphenyls

In the case of PCB congeners that do not easily form arene oxides, there is an alternative metabolic pathway of direct insertion of a hydroxyl group to form a monohydroxy metabolite, usually at an open meta position (Bandiera, 2013). Hydroxy metabolites are excreted as such or can be conjugated with glucuronide or sulfate by uridine diphosphate-glucuronosyltransferase (UGT), although there is little evidence that the higher chlorinated metabolites are conjugated (James, 2013). Of the 50 or so potential hydroxy metabolites, only five are retained in the blood and are bound to transthyretin, which normally binds thyroxine (T4). Hydroxy-PCBs that are substituted in the para position with chlorine atoms on each side of the hydroxyl group are known to be strongly retained in human blood (Letcher, Klasson-Wehler & Bergman, 2000), binding to transthyretin with an affinity that is, in general, greater than that of the natural ligand, T4 (Lans et al., 1993). The concentrations of hydroxy metabolites in blood are around 5–10 times lower than those of the most persistent PCB congeners (EFSA, 2005; Quinete et al., 2014). Another route of metabolism for PCBs with non-chlorinated meta/para positions on at least one of the phenyl rings is rapid metabolism to a methyl sulfone. This occurs in a multistep pathway involving glutathione conjugation catalysed by glutathione S-transferase (GST), degradation via the mercapturic acid pathway, and excretion into the bile and the large intestine, followed by cleavage by microbial C–S lyase. The thiols formed are methylated, reabsorbed and further oxidized on the sulfur atom to the corresponding methyl sulfone, which can then be distributed to the tissues through the blood. This enterohepatic recirculation may account for some of the long retention times of methyl sulfone metabolites (ATSDR, 2000; James, 2013). Fifty or more methyl sulfone PCB metabolites have been detected in human serum, but so far the majority of these have not been structurally identified (Grimm et al., 2015). Methyl sulfone PCB metabolites are present primarily in body lipids and accumulate with high selectivity in certain tissues, such as the liver and lung. In humans, those that have been identified are generally present only at low concentrations of 1% or less, compared with parent PCB concentrations (Grimm et al., 2015). In a recent study, the association between serum levels of 16 PCBs and genotype was analysed in 922 individuals 70 years of age from Uppsala, Sweden, focusing on CYP2B6 variation, to determine whether differences in metabolism might identify susceptible persons. The study included seven DL-PCBs and nine NDL-PCBs; the NDL-PCBs were PCB 74, PCB 99, PCB 138, PCB 153, PCB 170, PCB 180, PCB 194, PCB 206 and PCB 209. PCB concentrations were similar, or comparable, to those in other general European populations. The relationship was complex, with effects on mapping to CYP2B6 mediated predominantly through PCB 99, an NDL-PCB, and PCB 118, a DL-PCB. There were weaker associations with PCB 138 and PCB 153, but these were extinguished after adjusting for PCB 15

Safety evaluation of certain food additives and contaminants Eightieth JECFA

99 levels, suggesting that the association for these two PCBs is mediated through PCB 99 (Ng et al., 2015). Both hydroxylated and methyl sulfone PCB metabolites can have biological activity. For example, methyl sulfone metabolites can have antiestrogenic activity, bind to glucocorticoid receptors, reduce blood thyroid hormone levels and affect reproduction (Letcher et al., 2002). Thus, for those parent PCB congeners that are rapidly metabolized to persistent methyl sulfone metabolites, it is more relevant to assess the effects of those metabolites with the highest retention potential than to assess the effects of parent congeners, as the latter are present in only trace or non-detectable amounts (EFSA, 2005; James, 2013). Similarly, the most relevant hydroxy-PCBs are those that are persistent – that is, generally those with chlorine atoms on the adjacent carbons to the hydroxyl group and containing five or more chlorine atoms – which can exert toxicological effects on the thyroid (ATSDR, 2000; Quinete et al., 2014). 2.1.3 Receptor interactions and relationship to toxicity

Interactions with several nuclear receptors have been reported for PCBs, and these strongly depend on the number of chlorine atoms and their positions in the molecule. The binding of PCBs to AhR is by far the best studied, and structure– activity relationships (SARs) for the binding of PCBs are well known. PCBs lacking ortho-substituted chlorine atoms (e.g. PCBs 77, 81, 126 and 169) have the highest binding affinity for AhR and induce the typical dioxin-like activity seen at very low dose levels (Safe, 1984, 1993). With increasing chlorine substitution in the ortho position, the affinity for AhR rapidly decreases. As a result, congeners with two or more ortho chlorine atoms are considered to be NDL-PCBs (e.g. PCB 153). PCBs with one ortho chlorine atom (e.g. PCBs 105, 114, 118, 123, 156, 157, 167 and 189) do still bind to AhR and exert biological and toxicological effects that are similar to those of dioxins, but they also share biological and toxicological properties with the NDL-PCBs. These SARs between dioxin-like compounds, including some PCBs, and AhR are the basis for the WHO toxic equivalency factors (TEFs) approach that is now widely used for risk assessment (van den Berg et al., 1998). It is important to note that NDL-PCB congeners are not included in this WHO TEF concept, with the exception of some so-called mono-ortho-substituted PCBs that exhibit (moderate) AhR-mediated effects (van den Berg et al., 2006). Although there are observations of AhR-mediated responses to PCBs containing multiple ortho-substituted chlorines, there is an uncertainty regarding the possible dominating role of low-level contamination with potent dioxin-like agonists. The constitutive androstane receptor (CAR) and pregnane X receptor (PXR) are nuclear hormone receptors, and NDL-PCBs, at levels approximating

WHO Food Additives Series No. 71-S1, 2016 16

Supplement 1: Non-dioxin-like polychlorinated biphenyls

human serum levels, can directly activate both receptors, with subsequent gene transcription (Al-Salman & Plant, 2012; Gahrs et al., 2013). In particular, PCBs containing multiple ortho chlorine substitutions can have a profound agonistic effect on these nuclear receptors, which indicates that this substitution pattern plays a dominant role in binding to PXR or CAR. So far, the biological implications of chronic activation of PXR and/or CAR by NDL-PCBs remain unclear. However, it should be noted that activation of both receptors is associated with adverse health effects, such as metabolic dysfunction and changed hormone metabolism (Kretschmer & Baldwin, 2005; Al-Salman & Plant, 2012). Studies on possible interactions of PCBs with other (nuclear) receptor proteins are much more limited and usually involve a few congeners (Luthe, Jacobus & Robertson, 2008). For peroxisome proliferator–activated receptors (PPARs), it has been reported that only DL-PCBs act as antagonists (Ariyoshi et al., 1998; Robertson et al., 2007). Thus, based on these observations, it can be expected that NDL-PCBs may not interact with this receptor. In contrast, NDL-PCBs with multiple ortho chlorine atoms show a specific binding affinity to and activation of ryanodine receptors (RyRs). These receptors play a crucial role in calcium (Ca2+) signalling and neurotoxicity (Pessah et al., 2006) and are involved in numerous cellular and subcellular neuronal processes, such as exocytosis, cell death and mitochondrial function (Llansola et al., 2010; Pessah, Cherednichenko & Lein, 2010). This mechanism of action is thought to be one of the major pathways leading to the neurotoxicity of NDLPCBs. Furthermore, the observed SARs between NDL-PCBs and these RyRs may potentially provide an alternative TEF system for these compounds, with neurotoxicity as an end-point (Pessah et al., 2006). A comparable SAR was found for NDL-PCBs and a decrease in dopamine levels (Seegal, Bush & Shain, 1990; Shain, Bush & Seegal, 1991). Although the actual mechanism is still unknown, it may be related to decreased dopamine synthesis, an inhibition of tyrosine hydroxylase or L-aromatic amino acid decarboxylase, or a decreased uptake of dopamine into vesicles (Angus et al., 1997; Choksi et al., 1997; Mariussen, Morch Andersen & Fonnum, 1999). A wide range of NDL-PCBs and their most common hydroxylated and methyl sulfone metabolites have also been studied for their agonistic and antagonistic effects on the glucocorticoid receptor (GR). Although the parent PCBs can interact with GR, the inhibitory potency of the hydroxy-PCBs is much higher. These observations point towards an indirect mechanism of action via interactions of this receptor with metabolites of NDL-PCBs (Antunes-Fernandes et al., 2011). In addition, methyl sulfone PCB metabolites also act as agonists with GR in a structure-dependent way (Johansson, Nilsson & Lund, 1998; Johansson et al., 1998b). This interaction occurs at relatively low dose levels and is important, owing to the role of GR in many endocrine and physiological processes. 17

Safety evaluation of certain food additives and contaminants Eightieth JECFA

2.1.4 Effects on enzyme activities

PCBs have been identified as potent inducers of phase I and II enzymes in many vertebrate species, including humans (Safe, 1984). This enzyme induction is directly related to their binding to specific nuclear receptors (AhR, PXR, CAR). Traditionally, the differentiation between DL- and NDL-PCBs was based on the type of cytochrome P450 isoforms that are induced (Safe, 1993; Connor et al., 1995). Furthermore, metabolism of PCBs has usually been considered to be a detoxification process, because it facilitates (slow) elimination of these compounds from the body. However, during the last decades, it has become clear that common hydroxy-PCB and methyl sulfone PCB metabolites have additional mechanistic actions that can cause endocrine or toxic effects on, for example, thyroid hormone homeostasis, neuronal development and functioning, the adrenals and steroidogenesis. As mentioned above, the DL-PCBs show a high binding affinity to AhR, and this results, among other things, in the induction of CYP1A1, CYP1A2 and CYP1B1 in various tissues of the body. The lack of induction by NDL-PCBs of CYP1A1, CYP1A2 and CYP1B1 has generally been used to structurally define this category of congeners in comparison with DL-PCBs (Safe, 1984; van den Berg et al., 2006). NDL-PCBs bind to PXR and CAR, which can result in the induction of the CYP3A and CYP2B isoforms in rodents and humans (Petersen et al., 2007; Al-Salman & Plant, 2012). These enzymes also play a major role in the biotransformation and elimination of PCBs, which depend on the number and position of the chlorine atoms that are present in the molecules. To date, the involvement of CYP1A1, CYP1A2, CYP2A6, CYP2B6, CYP2C8, CYP2C9, CYP2C19 and CYP3A4 in the biotransformation of PCBs has been reported (Ariyoshi et al., 1995; McGraw & Waller, 2006; Warner, Martin & Wong, 2009; Yamazaki et al., 2011). The metabolism of NDL-PCBs in humans and rodents is generally thought to involve CYP2A, CYP2B, CYP2C and CYP3A isoforms, which play a major role in the formation of hydroxy-PCB metabolites. The relationship between adverse health effects and the induction of CYP2B and CYP3A enzymes still remains unclear, but it should be noted that these enzymes play a significant role in steroid metabolism and bioactivation of xenobiotics to genotoxic compounds. In addition, some of these hydroxy-PCB metabolites interfere significantly with thyroid hormone homeostasis at the receptor and transport protein (transthyretin) level, for which the presence of a hydroxyl group in the para position is important (Brouwer et al., 1998). In addition to the interaction of NDL-PCBs with the above cytochrome P450 isoforms that are involved with xenobiotic metabolism, several studies have reported effects on cytochrome P450 isoforms that are involved with the endogenous synthesis of (sex) hormones. These interactions with steroidogenic

WHO Food Additives Series No. 71-S1, 2016 18

Supplement 1: Non-dioxin-like polychlorinated biphenyls

cytochrome P450 isoforms in various cell types have been reported for both DLand NDL-PCBs and include parent compounds as well as hydroxy metabolites or methyl sulfone metabolites (Johansson, Nilsson & Lund, 1998; Johansson et al., 1998; Heneweer et al., 2005; Xu et al., 2006; Li, 2007; Antunes-Fernandes et al., 2011). In vitro studies have indicated effects of NDL-PCBs and their metabolites on steroidogenic pathways involved with corticosteroid and sex hormone synthesis, including upregulation of CYP11A, CYP11B1, CYP11B2, CYP17, CYP19, CYP21, 3β-hydroxysteroid dehydrogenase type 1 (3β-HSD1), 3β-HSD2 and 17β-HSD1 (Xu et al., 2006). PCBs are also known to induce various isoforms of phase II enzymes, such as UGTs, GSTs and sulfotransferases (SULTs). The induction of various isoforms of these enzymes depends on the binding and activation of the nuclear receptors mentioned previously. Although not studied in detail, it can be expected that NDL-PCBs binding to PXR and CAR may cause the induction of UGTs, GSTs and SULTs (Gardner-Stephen et al., 2004; Chai, Zeng & Xie, 2013; RungeMorris, Kocarek & Falany, 2013). In addition, DL-PCBs binding to AhR are also capable of inducing these phase II enzymes, and similar isoforms of UGT1A can be induced via both types of receptor (Zhou, Zhang & Xie, 2005). This indicates that UGT activation cannot be used as a discriminative marker between DL- and NDL-PCBs. UGT plays an important role in the metabolism and elimination of hydroxy-PCBs. Moreover, it has been found that the specific role of the UGT1A1, UGT1A6 and UGT2B1 isoforms depends on the position of the hydroxy groups in the PCB molecule and type of tissue (Tampal et al., 2002; Daidoji et al., 2005). From an endocrine point of view, it should be noted that UGT induction by these PCBs can lead to an increased elimination of thyroid hormones via the liver. This can have a distinct impact on neuroendocrine and neurobehavioural function (Vansell & Klaassen, 2002; Kato et al., 2004; Richardson et al., 2008). SULTs are another group of phase II enzymes for which an interaction with PCBs has been found. From a mechanistic point of view, there is a relationship with AhR as well as PXR and CAR. It appears that dioxin-like compounds are capable of downregulating SULT1A1 and SULT2A expression via an AhR-mediated process (Runge-Morris, Kocarek & Falany, 2013). Such an effect could also be expected for DL-PCBs, but not for the NDL-PCBs, which lack AhR agonistic properties. PXR and CAR also play a role in the expression of SULT, and therefore NDL-PCBs can likely interact with SULT, but so far experimental evidence is lacking. However, evidence for interaction with these enzymes is available for hydroxy-PCB metabolites, which can significantly inhibit the activity of several SULT isoforms. The extent of interaction depends on the position of the hydroxyl groups in the molecule; the presence of a hydroxyl group in the para or meta position is important, with ortho-hydroxy-PCBs being much 19

Safety evaluation of certain food additives and contaminants Eightieth JECFA

weaker inhibitors (van den Hurk et al., 2002; Liu et al., 2006, 2009; Ekuase et al., 2011, 2014). Limited information is available on the interaction of PCBs with GSTs. PXR and CAR are both involved in the regulation of these enzymes (Chai, Zeng & Xie, 2013). Consequently, interaction of NDL-PCBs with GSTs is a mechanistic possibility, but is not yet supported by experimental evidence. However, two studies indicate that induction of GSTs is most relevant for DL-PCBs, although their induction potency decreases with increasing ortho chlorine substitution (Aoki et al., 1992; Dragnev et al., 1995). 2.2 Toxicological studies

The focus of the toxicological information in this monograph is on data from oral toxicity studies on individual NDL-PCB congeners. Test samples of individual NDL-PCB congeners may potentially be contaminated with dioxinlike compounds, and so the purity of test samples has been described when such information was available. Studies on technical or reconstituted mixtures are not reviewed in detail. As EFSA (2005) noted in its opinion, experimental studies using technical or reconstituted mixtures or human data on exposure to the mixtures that occur in food and in the environment are not suitable for the evaluation of the effects of NDL-PCBs because, in most instances, no distinction can be made between the effects caused by NDL-PCBs and those caused by DLPCBs or PCDDs/PCDFs. It should also be noted that environmental mixtures may also contain other PCB congeners or other contaminants that are able to induce the same types of effects. 2.2.1 Acute toxicity WHO Food Additives Series No. 71-S1, 2016 20

No reports on the acute toxicity of individual NDL-PCB congeners were found. The acute toxicity of single oral doses of commercial PCB mixtures in the rat and mink is low, with median lethal dose (LD50) values ranging from 1000 to 4000 mg/kg body weight (bw) (ATSDR, 2000). 2.2.2 Short-term studies of toxicity (a) Mice (i) PCB 153

The effects of PCB 153 (purity 99.9%) on liver weight, histology and gene expression were investigated in mice and compared with those of 2,3,7,8-tetrachlorodibenzop-dioxin (TCDD), a classical AhR ligand (Kopec et al., 2010). Groups of five immature female ovariectomized mice were given a single dose of PCB 153 at 300 mg/kg bw or sesame oil vehicle (controls) by oral gavage and sacrificed 4, 12,

Supplement 1: Non-dioxin-like polychlorinated biphenyls

24, 72 or 168 hours later. Other mice were given PCB 153 at 1, 3, 10, 30, 100 or 300 mg/kg bw or sesame oil and sacrificed after 24 hours. Significant increases in relative liver weights were induced with PCB 153 at 300 mg/kg bw between 24 and 168 hours, accompanied by slight vacuolation and hepatocellular hypertrophy. Comparative analysis with TCDD suggested that the differential gene expression elicited by PCB 153 was not mediated by AhR. Protein expression of CAR/PXRregulated genes, including CYP2B10, CYP3A11, Ces2, Insig2 and Abcc3, was dose-dependently induced by PCB 153. In a follow-up study (Kopec et al., 2011), groups of five immature female ovariectomized mice were given TCDD at 30 μg/kg bw, PCB 153 (purity 99.9%) at 300 mg/kg bw, a mixture of TCDD at 30 μg/kg bw with PCB 153 at 300 mg/ kg bw (MIX) or sesame oil vehicle as single oral gavage doses and were sacrificed after 4, 12, 24, 72 or 168 hours. In the 24-hour dose–response study, animals were gavaged with TCDD (0.3, 1, 3, 6, 10, 15, 30 or 45 μg/kg bw), PCB 153 (3, 10, 30, 60, 100, 150, 300 or 450 mg/kg bw), MIX (0.3 + 3, 1 + 10, 3 + 30, 6 + 60, 10 + 100, 15 + 150, 30 + 300 or 45 μg/kg bw TCDD + 450 mg/kg bw PCB 153, respectively) or vehicle. All three treatments significantly increased relative liver weights, with MIX eliciting significantly greater increases compared with TCDD and PCB 153 alone. MIX induced hepatocellular hypertrophy, vacuolation, inflammation, hyperplasia and necrosis, a combination of TCDD and PCB 153 responses. Hepatic triglycerides were significantly increased by MIX and TCDD treatments, but not by PCB 153. Hepatic PCB 153 levels were also significantly increased by TCDD co-treatment. Microarray analysis for gene expression changes identified more than 100 unique, differentially expressed genes elicited by each of TCDD (n = 167), PCB 153 (n = 185) and MIX (n = 388). Thus, TCDD and PCB 153 cotreatment elicited specific, non-additive gene expression effects consistent with the liver changes observed. (b) Rats (i) Twenty-eight-day studies PCB 52 The short-term toxicity of PCB 52 in rats was investigated (unpublished data1 provided to WHO by study authors of the Assessing the Toxicity and Hazard of Non-dioxin-like PCBs Present in Food [ATHON] project; see also the ATHON Final Report at http://cordis.europa.eu/publication/rcn/11432_en.html and Elabbas et al., 2013). The experimental protocol followed Organisation 1

Liver and thyroid pathology data were made available for the JECFA meeting from the European Union project ATHON (Assessing the Toxicity and Hazard of Non-dioxin-like PCBs Present in Food). Results from the study are still under evaluation and are not yet published. 21

Safety evaluation of certain food additives and contaminants Eightieth JECFA

for Economic Co-operation and Development (OECD) Test Guideline 407 (Repeated Dose 28-day Oral Toxicity Study in Rodents). In order to improve the assessment of dose–response relationships at the lower end of the study dose range, the number of dose groups was increased to eight, whereas the number of rats of each sex per dose group was reduced to five. The protocol was optimized for dose–response evaluation by the benchmark dose (BMD) modelling approach and used a loading/maintenance dose protocol in order to reach steady-state conditions more rapidly. It was also enhanced to detect effects on the endocrine system. Animals were 6 weeks of age at the start of treatment. The purity of the PCB 52 used was 99.9%, and the analysed level of dioxin-like impurities, as represented by the sum of WHO toxic equivalents (TEQ), was 0.5 ng TEQWHO/g PCB 52. Groups of five male and five female rats were administered PCB 52 dissolved in corn oil or corn oil only (controls) by oral gavage at 4 mL/ kg bw. Loading doses were administered on days 0–4, and maintenance doses were administered 3 times a week over 3 weeks. The total doses of PCB 52 administered over the 28-day period were 0, 3, 10, 30, 100, 300, 1000 and 3000 mg/kg bw. Selection of the highest dose was based on a pilot study. The rats were observed for clinical signs twice daily on weekdays and once daily on weekends and were weighed every second day during the loading dose period and at least once weekly thereafter. Feed consumption and water consumption per cage were recorded once weekly. For determination of the stage of the estrous cycle, vaginal smears were collected from female rats daily starting from day 23 of the study. This was done to ensure that the females were at the diestrous stage during necropsy. A complete necropsy (macroscopic observations, tissue sampling for molecular biology, biochemistry, histopathology, analytical chemistry and organ weights) was performed on each rat. In addition, perirenal adipose tissue and liver were stored at −20 °C for determination of PCB 52 tissue concentration. Activities of UGT, pentoxyresorufin-O-deethylase (PROD) and ethoxyresorufinO-deethylase (EROD) and messenger RNA (mRNA) expressions of CYP1A1, CYP1A2, CYP1B1, CYP2B1 and CYP3A1 were measured in the liver. Hepatic retinoids, DNA damage markers and bone densitometry parameters were analysed. Observations were evaluated for exposure-related changes by analysis of variance (ANOVA). All significant exposure-related findings were further evaluated by dose–response modelling in order to establish the critical effect doses (CEDs) and the lower bounds of the confidence interval on the critical effect dose (CEDLs) at the default (5%) or end-point-specific critical effect sizes (CESs). The main effects observed were on liver and thyroid. Slightly increased relative liver weights were observed in females at the highest dose of 3000 mg/ kg bw. Blind reading of the histopathology across the full dose range showed a significant, dose-dependent increase in centrilobular hepatocellular hypertrophy

WHO Food Additives Series No. 71-S1, 2016 22

Supplement 1: Non-dioxin-like polychlorinated biphenyls

in male livers (Table 3). In females, hepatocellular hypertrophy was generally more localized in the periportal area, but without a dose-dependent distribution (Table 3). In male rats, CEDs could be calculated for the progression of the average animal from score 1 to 2 (BMD = 0.056 mg/kg bw) and from score 2 to 3 (CED = 0.658 mg/kg bw). In the thyroid, an increase in reduced follicle size was observed in both sexes (Table 4). Dose–response analysis revealed that this decrease occurred at a lower dose in male rats (CED = 6.5 mg/kg bw at CES = 5%) than in female rats (CED = 325 mg/kg bw at CES = 5%). The thyroid follicle size effect was not accompanied by follicular cell activation. The basal (control) average content of large follicles was lower in males than in females. Plasma free T4 was dose-dependently decreased in males at and above 300 mg/kg bw. Serum free triiodothyronine (T3) was not affected in either sex. There were no effects on reproductive organs, other endocrine organs or hormone levels in males or females. PCB 180 The short-term toxicity of PCB 180 was investigated in a 28-day repeated-dose toxicity study in young adult rats (Roos et al., 2011; Viluksela et al., 2014) as part of the ATHON project. The experimental protocol followed OECD Test Guideline 407 (Repeated Dose 28-day Oral Toxicity Study in Rodents), which was enhanced for detection of endocrine, neurotoxicity, retinoid, bone and DNA damage endpoints. In order to improve the assessment of dose–response relationships at the lower end of the study dose range, the number of dose groups was increased to eight, whereas the number of rats of each sex per dose group was reduced to five. The animals were 6 weeks of age at the start of treatment. The purity of PCB 180 was 98.9%, and the analysed level of dioxin-like impurities, as represented by the sum of WHO-TEQ, was 2.7 ng TEQWHO/g PCB 180. PCB 180 was dissolved in purity-controlled (0.2 pg TEQWHO/g) corn oil, corn oil also serving as control, and administered by oral gavage in a volume of 4 mL/kg bw. Groups of five male and five female rats were given total doses of PCB 180 of 0, 3, 10, 30, 100, 300, 1000 or 1700 mg/kg bw using a loading dose/maintenance dose regimen. To rapidly achieve the kinetic steady state, the total dose was divided into six daily loading doses and three weekly maintenance doses. Loading doses were administered on days 0–5 of the study, and maintenance doses were administered on days 10, 17 and 24. The rats were observed for clinical signs twice daily on weekdays and once daily on weekends, and they were weighed every second day during the loading dose period and at least once weekly thereafter. Feed consumption and water consumption per cage were recorded once weekly. For determination of the stage of the estrous cycle, vaginal smears were collected from female rats daily starting from day 23 of the study. This was done to ensure that the females 23

24 Incidence of hypertrophyc Safety evaluation of certain food additives and contaminants Eightieth JECFA

WHO Food Additives Series No. 71-S1, 2016

Table 3 Histological liver changes in rats treated with individual PCB congeners in 28-day repeated-dose toxicity studiesa

PCB congener no. 52d

180e

Males Females Centrilobular 0 3 10 30 100 300 1 000 1 700 3 000 0 3 10 30 100 300 1 000 1 700 3 000 hypertrophy mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg stagesb bw bw bw bw bw bw bw bw bw bw bw bw bw bw bw bw bw bw 0 1 1 2 4 3 – – – 1 1 2 1 2 4 5 2 1 1 1 – 1 1 1 1 – 1 2 1 1 2 – – 3 2 3 3 3 – 1 3 3 1 3 2 1 3 1 1 – – na na 3 – – – – – 1 1 4 – – – – – – – – 4 – – – – – – – – – – – – – – – – 5 – – – – – – – – – – – – – – – – 0 5 5 3 – – – – – 5 5 4 5 5 1 – – 1 – – – – – – – – – – 1 – – 3 4 1 2 – – – – – – – – – – – – – 1 1 3 na na 3 – – 2 5 1 5 1 1 – – – – – – – – 4 – – – – 4 – 4 3 – – – – – – – – 5 – – – – – – – – – – – – – – – –

na: dose not used a Doses given are total doses during study. b Hypertrophy stages 0–5 represent absent, slight, mild, moderate and severe, respectively. No other histopathology was observed. c Scores are actual counts per group, with n = 5 in all groups, except for PCB 52 females, 10 mg/kg bw, n = 4. d CED05 for mild hypertrophy in males was calculated at 0.056 mg/kg bw; no effect in females. e CED05 for mild hypertrophy was 14.8 and 205 mg/kg bw in males and females, respectively. Source: Unpublished data provided to WHO by ATHON project study authors (PCB 52); Roos et al. (2011) and Viluksela et al. (2014) (PCB 180)

Table 4 Histological thyroid changes in rats treated with individual PCB congeners in 28-day repeated-dose toxicity studiesa

PCB congener no. 52b

Incidence of thyroid change Males Females 0 3 10 30 100 300 1 000 1 700 3 000 0 3 10 30 100 300 1 000 1 700 3 000 mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg bw bw bw bw bw bw bw bw bw bw bw bw bw bw bw bw bw bw 2 5 5 4 5 5 4 na 5 2 2 2 2 2 2 3 na 4

180e

4 1 – 1 5 5 2 5 4

3 2 – 3

3 2 – 4

4 1 – 3

3 2 – 2

5 – – 5

5 – – 5

na

3 2 – na

1 4 – 2

ns

ns

ns

ns

ns

ns

ns

4

4

3

1 4 – na

Thyroid histopathology Reduced follicle sizec Epithelial hypertrophyd Stage 0 Stage 1 Stage 2 Reduced follicle sizec Epithelial hypertrophyd Stage 0 Stage 1 Stage 2 – – 5 – 1 4 – – 5 – 3 2 – 2 3 na 5 – – 2 3 – 3 2 – 2 3 – 2 3 – 1 3 1 2 3 – – 4 1

– 3 2

– 1 4

– 3 2

na

Supplement 1: Non-dioxin-like polychlorinated biphenyls

na: dose not used; ns: dose not scored a Doses given are total doses during study. b CED05 for reduced follicle size was calculated to be 6.5 and 325 mg/kg bw for males and females, respectively. Thyroid epithelial hypertrophy was not affected in either sex (only control and top dose scored in females). c Reduced follicle size is represented as the count of animals scoring below the average of control. d Thyroid epithelial hypertrophy stages 0–2 represent absent, mild and moderate, respectively. Scores are actual counts per group, with n = 5 in all groups. No other histopathology was observed. e CED05 for reduced follicle size was calculated to be 131 mg/kg bw for females; no effect in males. Thyroid epithelial hypertrophy was increased in females only, but no CED was calculated. Source: Unpublished data provided to WHO by ATHON project study authors (PCB 52); Roos et al. (2011) and Viluksela et al. (2014) (PCB 180)

25

Safety evaluation of certain food additives and contaminants Eightieth JECFA

were at the diestrous stage during necropsy. All observations were evaluated for exposure-related changes by ANOVA, and significant exposure-related findings were further evaluated by dose–response modelling in order to establish CEDs and CEDLs at the default (5%) or end-point-specific CESs. Body weight gain was reduced at 1700 mg/kg bw during the loading dose period in both sexes, but recovered thereafter. There were no observed general pathologies, with the exception of hyperplasia of the mammary glands, which was observed in six out of 10 assessed males, including controls. The most sensitive end-point was altered open-field behaviour in females. On study day 24 (test day 1), there were statistically significant, doserelated increases in the percentages of time and distance moved in the inner zone (CED = 0.35 mg/kg bw, 1.55 mg/g lipid, and CED = 0.87 mg/kg bw, 4.12 mg/g lipid, respectively, for CES = 5%). Conversely, there were decreases in the percentages of time and distance moved in the outer zone of the open field. These differences ameliorated across the 5 days of testing, as demonstrated by significant interactions between exposure and test days for both measures, indicating differences in habituation between groups. As a consequence, dose–response relationships were no longer statistically significant on day 28. The increased activity and distance moved in the inner zone of an open field suggest altered emotional responses to an unfamiliar environment and impaired behavioural inhibition. No significant dose–response relationships in open-field behaviour were found in exposed males. Absolute liver weights were dose-dependently increased at doses of 300 mg/kg bw and higher in both sexes. The increases were greater in males than in females, with CED values of 11.6 and 225 mg/kg bw (42.3 and 512 µg/g lipid, respectively) and maximum increases of 66% and 45% in males and females, respectively. The liver of exposed animals showed centrilobular hypertrophy (see Table 3). BMD analysis showed CEDLs of 9 and 138 mg/kg bw in males and females, respectively, for progression of hepatocellular hypertrophy to stage 1. Males were more sensitive to the induction of hypertrophy, in terms of both the CEDL and severity. The centrilobular hypertrophy observed in the liver was associated with the induction of hepatic cytochrome P450 enzymes, including significant increases in liver PROD activity, CYP2B1 and CYP3A1 mRNA levels, as well as CYP2B1/2 and CYP3A1 protein levels in both male and female rats, with males being more sensitive. A significant induction of hepatic EROD activity was observed in both male and female rats, with males being more sensitive. CYP1A1 mRNA and protein levels were induced at the higher dose levels. In contrast, CYP1B1 and CYP1A2 were not affected at the mRNA level, although slight inductions of protein levels were observed at the higher doses. These findings suggest that PCB 180 acts as a CAR and PXR agonist and as a weak inducer of AhR-mediated CYP1A1 expression and activity. A significant dose-related

WHO Food Additives Series No. 71-S1, 2016 26

Supplement 1: Non-dioxin-like polychlorinated biphenyls

decrease in liver retinoids was also observed after exposure to PCB 180 regarding both amount, with CEDs of 257 mg/kg bw in females and 148 mg/kg bw in males, and concentration, with CEDs of 123 mg/kg bw in females and 21.6 mg/kg bw in males. Taken together, quantification of liver histopathology, hepatic cytochrome P450 enzyme assays and hepatic retinoid levels were confirmative, as males were more sensitive than females to PCB-induced changes for all three end-points. In contrast, expressions of the tumour suppressor protein p53 and the DNA damage signalling proteins p53 Ser15, γH2AX Ser139 and pChk2 Thr68 were dosedependently increased in livers of female rats, whereas expression of pMdm2 Ser166 was not affected. These in vivo results are in line with findings in the human hepatocellular carcinoma cell line HepG2 (Al-Anati, Högberg & Stenius, 2009). Other dose-dependent changes due to PCB 180 exposure included decreased serum thyroid hormone levels with associated histopathological changes (see Table 4). The weight of the thyroid gland was dose-dependently increased in males, but decreased in females. Plasma free T4 level was dosedependently decreased in both sexes, compared with controls: in males, it was decreased at doses of 100 mg/kg bw and higher, reaching statistical significance at 300 mg/kg bw and higher; in females, it was decreased at 300 mg/kg bw, reaching statistical significance at 1700 mg/kg bw. The effect was more pronounced in males than in females, with a maximal reduction of 69% in males compared with 50% in females. Plasma free T3 concentrations were significantly decreased only in males at 1000 mg/kg bw. Histopathology revealed increased thyroid follicular cell vacuolation, suggestive of thyroid activation, and there was a dose-dependent decrease of large follicles in females, indicating depletion of follicle contents. Male thyroids had a lower proportion of large follicles at the control and low doses compared with females, comparable with the level in females exposed to high doses. The area of large follicles was estimated as a percentage of the total area of the section, and the data for the females were used for dose–response and BMD analyses. These analyses revealed a CED of 131 mg/kg bw, with a corresponding CEDL of 81 mg/kg bw. The follicular epithelial cells showed hypertrophy, which increased in a dose-dependent way in females. Males had a higher basal score for hypertrophy, and no significant increase was observed with treatment. In the adrenals, cells in the zona fasciculata showed dose-dependent hypertrophy. Females were more sensitive than males, showing progression of hypertrophy to further stages, and they also had a lower CED (2.0 mg/kg bw) than did males (594 mg/kg bw). There was also hypertrophy and vacuolation in cells of the zona reticularis, with a significant dose–response relationship in females. The inner zones of the cortex occasionally also showed hyperaemia, with a significant dose–response relationship in females, but not in males. 27

Safety evaluation of certain food additives and contaminants Eightieth JECFA

Cauda epididymal sperm counts were analysed in the controls and the highest-dose group, but no differences were observed. Serum testosterone levels in males and estradiol and progesterone levels in females were not affected by the treatment. However, serum follicle stimulating hormone (FSH) and luteinizing hormone (LH) levels showed significant decreasing trends in males. LH levels were not affected in females. (ii) Ninety-day studies

PCB 28 The short-term toxicity of PCB 28 was investigated in a 90-day dietary exposure study in rats (Chu et al., 1996a). The purity of PCB 28 was greater than 99%, and contamination with PCDDs and PCDFs was found to be less than 0.1 mg/ kg. Groups of 10 male and 10 female weanling rats were administered PCB 28 in the diet at 0, 0.05, 0.50, 5.0 or 50.0 mg/kg feed for 13 weeks. The corresponding calculated exposures to PCB 28 were 0, 2.8, 36, 359 and 3783 µg/kg bw per day for male rats and 0, 2.9, 37, 365 and 3956 µg/kg bw per day for female rats. Corn oil was used to dissolve the test substance prior to mixing with the diet, and control groups received the diet containing an equivalent amount of corn oil (4% weight per weight [w/w]) only. Feed consumption and body weights were determined weekly. Observations for clinical signs of toxicity were made daily. Brain, liver, heart, lung, spleen, thymus and kidney weights were recorded at termination of the experiment. Haematology, clinical chemistry and full histopathology were performed. Liver aminopyrine N-demethylase (APDM), UGT and EROD activities and uroporphyrin levels were determined. Biogenic amines (dopamine, norepinephrine, serotonin, 3,4-dihydroxyphenylacetic acid, 3-methoxy-4hydroxyphenylacetic acid, 5-hydroxyindoleacetic acid) and total protein were analysed in several sections of the right hemisphere of the brain. Vitamin A content of liver, lung and kidneys and ascorbic acid in urine samples were analysed. PCB 28 residues were analysed in several tissues. Growth rate and feed consumption were not affected by treatment, and no clinical signs of toxicity were observed. Liver and thyroid gland showed treatment-related histopathological changes in both male and female rats. Liver pathology was mild to moderate in nature, and treatment-related changes generally occurred at 5.0 mg/kg feed and above (Table 5). Several thyroid gland changes were present at 0.05 mg/kg feed; however, the authors judged that only the thyroid gland changes at 5.0 mg/kg feed and above were biologically significant (see Table 6). Morphological changes were also seen in the kidney and thymus of treated groups; these changes were mild even at the highest dose. No treatment-related histopathological changes were found in other tissues, and no

WHO Food Additives Series No. 71-S1, 2016 28

Table 5 Histological changes in liver of rats treated with individual PCB congeners in 90-day repeated-dose toxicity studies Incidence of histological changes in livera

PCB congener no. 28 5.0 mg/kg bw Control 3/10 (0.4) 4/10 (0.4) 7/10 (0.6) 3/10 (0.2) 3/9 (0.2) 8/10 (1.0) 6/9 (0.6) 6/10 (0.6) 5/9 (1.3) 2/10 (0.5) 5/9 (0.8) 8/10 (0.9) 8/10 (1.5) 5/10 (1.1) 4/10 (0.3) 8/10 (1.6) 3/10 (0.4) 6/10 (0.8) 9/10 (0.6) 10/10 (2.4) 9/9 (2.3) 8/9 (1.0) 6/9 (1.3) 5/9 (0.7) 50.0 mg/kg bw 0.05 mg/kg bw 5.0 mg/kg bw 5/10 (1.1) 7/10 (1.1) 3/10 (0.3) 10/10 (2.4)

Control

0.05 mg/kg feed

Males 0.50 mg/kg bw

Females 0.50 mg/kg bw

50.0 mg/kg bw 8/10 (1.1) 2/10 (0.4) 6/10 (0.4) 6/10 (0.7)

4/9 (0.6)

8/10 (1.4)

0/9

1/10 (0.3)

2/9 (0.4)

3/10 (0.3)

5/9 (0.7)

10/10 (2.0)

128 0/10 7/10 (0.45) 9/10 (1.7) 9/10 (2.0) 7/10 (2.1) 1/10 (0.05) 8/10 (0.4) 7/10 (0.9) 0/10 0/10 5/10 (0.7) 3/10 (0.2) 8/10 (0.6) 4/10 (0.2)

0/10

1/10 (0.10)

10/10 (1.3) 3/10 (0.35)

8/10 (1.2) 3/10 (0.4)

7/10 (0.8) 7/10 (0.7)

2/10 (0.25)

3/10 (0.15)

10/10 (0.9)

10/10 (0.9)

7/10 (0.35)

10/10 (0.6)

10/10 (1.2)

8/10 (1.8)

153 7/10 (0.6) 10/10 (1.9) 10/10 (3.0) 1/10 (0.3) 9/10 (1.3) 5/10 (1.3)

3/10 (0.4)

9/10 (0.4)

6/10 (0.6) 4/10 (0.2)

3/10 (0.3) 7/10 (0.5)

4/10 (0.5) 10/10 (1.5)

6/10 (0.8) 10/10 (2.5)

10/10 (0.9) 3/10 (0.9)

Liver histopathology Liver cytoplasm Increased portal density Periportal vacuolation Midzonal vacuolation Increased perivenous homogeneity Liver cytoplasm Increased portal density Midzonal vacuolation Increased perivenous homogeneity Liver cytoplasm Midzonal vacuolation Increased perivenous homogeneity

6/10 (0.5)

10/10 (0.8)

Supplement 1: Non-dioxin-like polychlorinated biphenyls

a

Data denote (number of animals showing changes/number of animals examined). Figures in parentheses denote the severity of histology gradings, which were as follows: 1 = minimal; 2 = mild; 3 = moderate; 4 = severe. The overall scores are obtained by dividing the sum of total scores by the number of tissues examined. Source: Chu et al. (1996a) (PCB 28); Lecavalier et al. (1997) (PCB 128); Chu et al. (1996b) (PCB 153)

29

30 Incidence of histological change in thyroida Males 0.50 mg/kg feed 9/10 (1.7) 9/10 (1.0) 5.0 mg/kg feed 9/10 (1.7) 8/10 (0.7) Control 4/10 (0.6) 2/10 (0.2) 4/10 (0.4) 5/10 (0.2) 5/10 (0.3) 0/10 3/10 (0.45) 9/10 (0.8) 9/10 (0.7) 8/10 (0.7) 4/10 (0.1) 6/10 (0.68) 10/10 (0.6) 6/10 (0.6) 8/10 (0.9) 4/10 (0.1) 6/10 (0.75) 9/9 (1.2) 9/9 (1.2) 9/9 (1.3) 2/9 (0.1) 8/10 (1.20) 10/10 (1.5) 10/10 (1.1) 9/10 (1.7) 4/10 (0.2) 7/10 (1.50) 10/10 (2.5) 10/10 (1.5) 10/10 (2.0) 3/10 (0.5) 9/10 (1.58) 50.0 mg/kg feed 8/10 (1.8) 7/10 (0.7) 0.05 mg/kg feed 8/10 (1.6) 9/10 (0.8) 5.0 mg/kg feed 8/9 (1.6) 5/9 (0.4) 9/10 (1.4) 10/10 (1.4) 9/10 (1.0) 2/10 (0.2) 8/10 (1.25) Females 0.50 mg/kg feed 8/10 (1.3) 7/10 (0.7) 50.0 mg/kg feed 7/10 (1.6) 9/10 (1.7) 9/10 (1.4) 10/10 (1.3) 10/10 (1.6) 3/10 (0.3) 8/10 (1.33) Safety evaluation of certain food additives and contaminants Eightieth JECFA 10/10 (1.8) 3/10 (0.10) 4/10 (0.40) 8/10 (1.3) 5/10 (0.7) 2/10 (0.3) 10/10 (1.6) 3/10 (0.2) 9/10 (1.4) 10/10 (1.6) 10/10 (1.3) 10/10 (2.0) 10/10 (2.0) 10/10 (1.8) 1/10 (0.2) 8/10 (0.8) 2/10 (0.3) 10/10 (2.3) 5/10 (0.63) 9/10 (0.93) 9/10 (1.4) 5/10 (0.8) 10/10 (2.4) 7/10 (0.68) 10/10 (1.8) 10/10 (2.5) 5/10 (0.8) 5/10 (0.48) 0/10 0/10 2/10 (0.3) 2/10 (0.1) 7/10 (0.43) 0/10 6/10 (0.20) 4/10 (0.8) 5/10 (0.7) 5/10 (0.5) 10/10 (1.1) 7/10 (0.9) 9/10 (0.6) 0/10 8/10 (0.78) 5/10 (1.0) 9/10 (1.3) 8/10 (1.2) 10/10 (1.2) 10/10 (1.4) 10/10 (1.5) 1/10 (0.03) 10/10 (1.58) 5/10 (0.9) 10/10 (0.7) 10/10 (1.4) 10/10 (1.7) 8/10 (1.2) 8/10 (1.7) 2/10 (0.30) 9/10 (1.75) 10/10 (1.9) 10/10 (1.6) 10/10 (2.3) 10/10 (1.4) 10/10 (2.5)

WHO Food Additives Series No. 71-S1, 2016

Table 6 Histological changes in thyroid of rats treated with individual PCB congeners in 90-day repeated-dose toxicity studies

PCB congener no. 28b

Control 3/9 (0.6) 6/9 (0.9)

0.05 mg/kg feed 7/10 (1.2) 8/10 (0.7)

128c

4/9 (0.6) 8/9 (1.2) 0/9 0/9 3/10 (0.65)

5/10 (0.8) 10/10 (1.0) 2/10 (0.2) 2/10 (0.1) 9/10 (1.50)

153d

Thyroid histopathology Reduced follicle size Follicle collapse/angularity Thyroid epithelium Increased height Cytoplasmic vacuolation Nuclear vesiculation Colloid density Reduced follicle size Thyroid epithelium Increased height Papillary proliferation Nuclear vesiculation Reduced follicle size Follicle collapse Thyroid epithelium Increased height Cytoplasmic vacuolation Nuclear vesiculation

10/10 (1.2) 0/10 1/10 (0.18) 1/10 (0.08) 1/10 (0.03)

9/10 (1.3) 1/10 (0.13) 0/10 3/10 (0.35) 1/10 (0.05)

0/10 8/10 (0.8) 0/10

2/10 (0.1) 10/10 (0.6) 3/10 (0.1)

a

Data denote (number of animals showing changes/number of animals examined). Figures in parentheses denote the severity of histology gradings, which were as follows: 1 = minimal; 2 = mild; 3 = moderate; 4 = severe. The overall scores are obtained by dividing the sum of total scores by the number of tissues examined. Comments and conclusions by the authors of the study on PCB 28 (Chu et al., 1996a): Low magnification: treatment-related follicle size and colloid density and collapse of follicles. At cytological level: increased epithelial height, cytoplasmic vacuolation and nuclear vesiculation. 0.05 mg/kg feed: moderate and irregular reduction in follicle size + vacuolation + mildly reduced colloid density. 0.50 mg/kg feed: no comment. 5.0 and 50.0 mg/kg feed: biologically significant changes in both sexes.

b

Supplement 1: Non-dioxin-like polychlorinated biphenyls

50.0 mg/kg feed: marked reduction in follicle size and colloid density. Conclusions: Dose/treatment-related effect. Severity at 0.05 and 0.50 mg/kg feed was considered minimal. A broader spectrum of histological changes was observed only at 5.0 mg/kg feed and above. NOAEL = 0.50 mg/kg feed (equal to 36 µg/kg bw per day). c Comments and conclusions by the authors of the study on PCB 128 (Lecavalier et al., 1997): Treatment-related histological changes consisted of reduced follicle size with papillary proliferation and nuclear vesiculation of the epithelium. Incidence and severity were dose dependent and were rated minimal in the low-dose groups. Changes became progressively more severe as the dose increased. Conclusions: More severe changes occurred at 5.0 mg/kg feed and above. NOAEL = 0.50 mg/kg feed (equal to 42 µg/kg bw per day). d Comments and conclusions by the authors of the study on PCB 153 (Chu et al., 1996b): Minimal dose-dependent reduction in follicle size in the low-dose group (0.05 mg/kg feed). Changes of moderate degree in the high-dose group. Dose-dependent reduction in follicle size and increased epithelial height and nuclear vesiculation. Epithelial cells changed from low cuboidal to columnar shape, with females being more sensitive and effects occurring from 0.50 mg/kg feed. Cytoplasmic vacuolation occurred also at high rate in control animals; thus, considered not to be treatment related. Conclusions: NOAEL = 0.50 mg/kg feed (equal to 34 µg/kg bw per day). Source: Chu et al. (1996a) (PCB 28); Lecavalier et al. (1997) (PCB 128); Chu et al. (1996b) (PCB 153)

haematological changes were observed. An increase in urinary ascorbic acid and minimal induction of hepatic EROD activity were observed in both sexes of the 50.0 mg/kg feed group. A significant decrease in dopamine concentrations in the substantia nigra region of the brain was observed in female rats at 0.50 mg/kg feed and above. Vitamin A content in liver, lung and kidneys was not affected by PCB 28 treatment. PCB 28 residues were found in fat, liver, kidney, brain and spleen and accumulated in a dose-dependent fashion, with the PCB concentrations in fat being at least 20- to 40-fold higher than those in other tissues. Based on these data, the authors of the study concluded that the no-observed-adverse-effect level (NOAEL) for PCB 28 was 0.50 mg/kg feed (equal to 36 µg/kg bw per day). PCB 128 The short-term toxicity of PCB 128 was investigated in a 90-day dietary exposure study in rats (Lecavalier et al., 1997). The purity of PCB 128 was greater than 99%, and no detectable levels of PCDDs or PCDFs were found (limit of detection [LOD] of 1 mg/kg). Groups of 10 male and 10 female weanling rats were administered PCB 128 in the diet at 0, 0.05, 0.50, 5.0 or 50.0 mg/kg feed for a period of 13 weeks. The corresponding calculated exposures to PCB 128 were 0, 4.2, 42, 425 and 4210 µg/kg bw per day for male rats and 0, 4.5, 45, 441 and 4397 µg/kg bw per day for female rats. Corn oil was used to dissolve the test substance prior to mixing with the diet, and control groups received the diet containing an equivalent amount of corn oil (4% w/w) only. Feed consumption and body weights were determined weekly. Observations for clinical signs of toxicity were made daily. Brain, liver, heart, lung, spleen, thymus and kidney weights were recorded at termination of the experiment. Haematology, clinical chemistry and full histopathology were performed. Liver aniline hydroxylase (AH), APDM, PROD and EROD activities and total protein and uroporphyrin levels were determined. Selected biogenic amines and metabolites (i.e. dopamine, norepinephrine, serotonin, 3,4-dihydroxyphenylacetic acid, 3-methoxy-4-hydroxyphenylacetic acid, 31

Safety evaluation of certain food additives and contaminants Eightieth JECFA

5-hydroxyindoleacetic acid) and total protein were analysed in several sections of the right hemisphere of the brain. Vitamin A content of liver, lung and kidneys and ascorbic acid in urine samples were analysed. PCB 128 residues were analysed in several tissues. Growth rate and feed consumption were not affected by treatment, and no clinical signs of toxicity were observed. The highest-dose group of female rats had an increased liver to body weight ratio. Liver and thyroid gland showed treatment-related histopathological changes in both male and female rats, with female rats seemingly more sensitive (see Tables 5 and 6). Several liver and thyroid changes were present at 0.05 mg/kg feed, although they were mostly of minimal severity; the authors of the study judged that more severe changes occurred at 5.0 mg/kg feed and above. An increase in urinary ascorbic acid was observed in the 50.0 mg/kg feed group of both sexes. Hepatic EROD activity was induced in both male and female rats starting from doses of 0.5 mg/kg feed in female rats, whereas APDM activity was increased at only the highest dose level in both male and female rats. PROD activity data were not presented. Decreased dopamine concentrations were found in the frontal cortex and hippocampus of female rats. Although a dose-related trend in dopamine levels in the hippocampus was observed, there was no dose-related trend in the levels in the frontal cortex, as even the lowest dose, 0.05 mg/kg feed, resulted in the same level of dopamine reduction. Hepatic vitamin A content was decreased at the highest dose level in female rats only. PCB 128 residues were found in fat, liver, kidney, brain, spleen and serum and accumulated in a dose-dependent fashion, with the PCB concentrations in fat being markedly higher than those in other tissues. PCB 128 concentrations in fat, kidneys and brain were higher in female rats than in male rats by a factor of about 2. Based on these data, the authors of the study concluded that the NOAEL for PCB 128 was 0.50 mg/kg feed (equal to 42 µg/kg bw per day). WHO Food Additives Series No. 71-S1, 2016 32

PCB 153 The short-term toxicity of PCB 153 was investigated in a 90-day dietary exposure study in rats (Chu et al., 1996b). The purity of PCB 153 was greater than 99% (authors wrote in error <99%) and contained less than 1% (authors wrote in error >1%) of PCDDs/PCDFs (LOD 1 mg/kg). Groups of 10 male and l0 female weanling rats were administered PCB 153 in the diet at 0, 0.05, 0.50, 5.0 or 50.0 mg/kg feed for a period of 13 weeks. The corresponding calculated exposures to PCB 153 were 0, 3.6, 34, 346 and 3534 µg/kg bw per day for male rats and 0, 4.2, 42, 428 and 4125 µg/kg bw per day for female rats. Corn oil was used to dissolve the test substance prior to mixing with the diet, and control groups received the diet containing an equivalent amount of corn oil (4% w/w) only.

Supplement 1: Non-dioxin-like polychlorinated biphenyls

Feed consumption and body weights were determined weekly. Observations for clinical signs of toxicity were made daily. Brain, liver, heart, lung, spleen, thymus and kidney weights were recorded at termination of the experiment. Haematology, clinical chemistry and full histopathology were performed. Liver AH, APDM and EROD activities and total protein and uroporphyrin levels were determined. Selected biogenic amines and metabolites (i.e. dopamine, norepinephrine, serotonin, 3,4-dihydroxyphenylacetic acid, 3-methoxy-4-hydroxyphenylacetic acid, 5-hydroxyindoleacetic acid) and total protein were analysed in several sections of the right hemisphere of the brain. Vitamin A content of liver, lung and kidneys and ascorbic acid in urine samples were analysed. PCB 153 residues were analysed in several tissues. Growth rate and feed consumption were not affected by treatment. Clinical signs of toxicity were not observed. Enlarged, fatty liver was observed in treated male rats at all dose levels and in females only at the highest dose level. Liver and thyroid gland showed treatment-related histopathological changes in male and female rats. Several liver and thyroid changes were present at 0.05 mg/ kg feed, although they were mostly of minimal severity; the authors judged that more severe changes occurred at 5.0 mg/kg feed and above (Table 5). Details of histopathological changes in the thyroid are shown in Table 6. An increase in urinary ascorbic acid was observed in both sexes of the 50.0 mg/kg feed group, and increased liver uroporphyrin levels were observed in females only at the highest dose level. Hepatic AH, APDM and EROD activities were induced in both sexes starting from doses of 0.05 and 0.50 mg/kg feed in female and male rats, respectively. Changes in brain biogenic amines and intermediate products were observed mainly in female rats. Observations included decreased dopamine and 5-hydroxytryptamine concentrations in the frontal cortex region at 5.0 and 50.0 mg/kg feed and decreased dihydroxyphenylacetic acid in the caudate nucleus region, which was significantly reduced at 0.50 mg/kg feed. Thus, female rats appeared to be more sensitive than males to the neurotoxic effects of PCB 153. Treatment-related reductions in hepatic and pulmonary vitamin A were seen in the highest-dose group of both sexes. PCB 153 residues were found in fat, liver, kidney, brain, spleen and serum and accumulated in a dose-dependent fashion, with the PCB concentrations in fat being markedly higher than those in other tissues. PCB 153 concentrations in all tissues except for liver were higher in females than in males. Based on these data, the authors of the study concluded that the NOAEL for PCB 153 was 0.50 mg/kg feed (equal to 34 µg/kg bw per day). (iii) PCB metabolites

Reduced thyroid hormone levels were found in serum of rats treated with four consecutive daily doses of each of the 3-methylsulfonyl metabolites of 33

Safety evaluation of certain food additives and contaminants Eightieth JECFA

the congeners PCB 132, PCB 141 and PCB 149 and with the 4-methylsulfonyl metabolite of PCB 149. All four metabolites (20 µmol/kg bw, intraperitoneal injection, once per day for 4 days) reduced the serum concentration of total T4 by 22–44% on days 2, 3, 4 and 7 after the last dose. Concentrations of total T3 were reduced by 37% on day 7 after treatment with the 4-methylsulfonyl metabolite of PCB 149. A 30% increase in thyroid weight was seen after treatment with the 3-methylsulfonyl metabolite of PCB 141. The data suggest that these 3- and 4-methylsulfonyl metabolites act on the thyroid, but probably through different mechanisms (Kato et al., 1998). A similar study conducted with the 3-methylsulfonyl metabolites of PCB 49, PCB 70, PCB 87 and PCB 101 and the 4-methylsulfonyl metabolite of PCB 101 found that all five methylsulfonyl metabolites of PCBs influence thyroid hormone metabolism (Kato et al., 1999). A further study by this group demonstrated that the 3-methylsulfonyl metabolites of PCB 49, PCB 70, PCB 89, PCB 101, PCB 132, PCB 141 and PCB 149 and the 4-methylsulfonyl metabolite of PCB 101 induced hepatic UGT in male rats. The increase in hepatic glucuronidation of T4 after the administration of the eight test compounds was the probable cause of the reduced serum concentration of T4 (Kato et al., 2000). 2.2.3 Long-term studies of toxicity and carcinogenicity

Studies on the carcinogenicity of commercial mixtures of PCBs, individual PCB congeners, combinations of specified congeners and PCB metabolites have been described in detail by IARC (2015). (a) Commercial mixtures of PCBs

WHO Food Additives Series No. 71-S1, 2016 34

Early studies on commercial PCB mixtures containing both DL- and NDL-PCBs administered orally in the diet were summarized in Ahlborg, Hanberg & Kenne (1992). Various Aroclor, Kanechlor and Clophen mixtures have been tested in mice and in several strains of rats at doses of 5 mg/kg bw per day or more. They induced hepatocellular carcinomas and/or hepatic foci and neoplastic nodules in both sexes of both species. Several of these early bioassays for cancer were re-evaluated by a panel of pathologists, applying the then-accepted classification for liver pathology (IEHR, 1991). The outcome of the evaluation was that only Aroclor 1260 and Clophen A60 were considered to give a clear carcinogenic response. However, in many of the other studies, there were high incidences of hepatocellular proliferative lesions in both male and female rats that were related to the administration of the various mixtures. PCB mixtures are also effective tumour promoters in mouse and rat liver when given in conjunction with known carcinogens, such as 2-acetylaminofluorene, benzene hexachloride, 3′-methyl-

Supplement 1: Non-dioxin-like polychlorinated biphenyls

4-dimethylaminoazobenzene, N-methyl-N′-nitrosoguanidine or nitrosamines (Ahlborg, Hanberg & Kenne, 1992; ATSDR, 2000; IARC, 2015). Later studies on commercial mixtures have been summarized in ATSDR (2000) and IARC (2015). They confirmed that several commercial mixtures increased the incidences of hepatocellular adenomas and carcinomas, hepatocholangiomas and hepatocholangiocarcinomas, and thyroid follicular cell adenomas and carcinomas in rats. Increases in hepatocellular carcinomas were also confirmed in three mouse studies (liver was the only tissue examined and reported on). In its opinion, EFSA (2005) commented on whether the liver carcinogenicity of technical PCB mixtures is due to the dioxin-like compounds present in these mixtures. For example, in a comparative chronic carcinogenicity study in rats administered four different Aroclor mixtures (1016, 1242, 1254 and 1260), the total TEQ doses associated with dioxin-like constituents within the technical mixtures, but not the doses of total PCBs, were mainly, if not exclusively, responsible for the development of liver neoplasms (Mayes et al., 1998). EFSA (2005) also did a quantitative comparison of the data from Mayes et al. (1998) with those from a chronic carcinogenicity study with TCDD in female rats (Kociba et al., 1978), which showed that for induction of hepatic neoplasms in female rats, the dose–response curves for the total TEQs in the various technical PCB mixtures were similar to that for TCDD. EFSA (2005) commented that these findings suggest that in rats, NDL-PCBs, administered together with DL-PCBs in technical mixtures, play a minor role – if any – as carcinogens. (b) Individual NDL-PCB congeners

Few studies on the chronic toxicity and carcinogenicity of individual NDL-PCB congeners have been published to date. (i) PCB 153

The United States National Toxicology Program (NTP) carried out a carcinogenesis bioassay on PCB 153, which is the PCB that is present at the highest concentrations in human samples on a molar basis (NTP, 2006a). Groups of 80–82 female rats were given PCB 153 in corn oil:acetone (99:1) by oral gavage for 14, 31 or 53 weeks or 2 years. Analytical checks of the test material for purity showed an overall purity of greater than 99%, no contamination with DL-PCBs and only minor contamination with PCB 101 (0.21%) and PCB 180 (0.002%). PCB 153 dose levels of 0 (vehicle only), 10, 100, 300, 1000 and 3000 µg/kg bw per day were administered on 5 days/week for up to 105 weeks. A stop-exposure group of 50 female rats was administered 3000 µg/kg bw per day for 30 weeks and then vehicle only for the remainder of the study. In addition to the usual 35

Safety evaluation of certain food additives and contaminants Eightieth JECFA

histopathological observations, PCB concentrations in fat, liver, lung and blood were measured, and specific end-points that might be affected by PCBs, such as cytochrome P450 levels in liver and lung, thyroid hormone levels and hepatic cell proliferation, were investigated. Detectable levels of PCB 153 in fat were found at 14, 31 and 53 weeks and at the end of the 2-year study in the vehicle controls. Measurable concentrations of PCB 153 were also found in the lungs of vehicle control rats at 31 and 53 weeks and at 2 years. No measurable concentrations of PCB 153 were found in the liver of vehicle controls at any time point. The finding of PCB 153 in fat and lung of the vehicle control group is likely attributable to the presence of low levels of PCB 153 in the laboratory chow of rats (Feeley & Jordan, 1998; Jordan & Feeley, 1999). NTP (2006a) estimated that the background intake of PCB 153 from the chow was 1–3 orders of magnitude less than the lowest dose of PCB 153 administered. In the groups dosed with PCB 153, concentrations of PCB 153 in fat and liver increased with increasing dose and exposure duration. In fat, these doses resulted in a linear increase in concentrations; at the end of the study, concentrations were approximately 440, 20 000, 160 000, 520 000, 1 600 000 and 4 300 000 ng/g lipid for the 0, 10, 100, 300, 1000 and 3000 µg/kg bw per day dose groups, respectively. Concentrations of PCB 153 in lung and blood increased with increasing dose at each time point, and concentrations in blood increased with duration of exposure. In liver, lung and blood of rats in the 3000 µg/kg bw per day stop-exposure group, PCB 153 concentrations were slightly above or below those found in the 1000 µg/ kg bw per day group. There were no dose-related effects on survival. Mean body weights were unaffected by treatment except in the 3000 µg/kg bw per day core study rats, which had lower mean body weight than vehicle controls after week 69 of the study. Absolute and/or relative liver weights in the rats given 1000 or 3000 µg/kg bw per day were significantly higher than those of vehicle controls at 14 and 31 weeks. At week 53, absolute and relative liver weights were significantly higher in rats given 100 µg/kg bw per day or more compared with vehicle controls. Absolute kidney weights of all groups exposed to PCB 153 and the relative kidney weight of 3000 µg/kg bw per day rats were significantly increased at week 53. There were no effects on thyroid weight. There were significant increases in the incidences of hepatocyte hypertrophy in the rats given 1000 or 3000 µg/kg bw per day at 14 weeks and in all groups administered 300 µg/kg bw per day or more at 31 and 53 weeks. At 2 years, the incidence of hepatocyte hypertrophy was significantly increased in all dosed groups. There were significant increases in diffuse fatty change in the liver in the groups given 300 µg/kg bw per day or more and in bile duct hyperplasia in those given 300 or 3000 µg/kg bw per day (core and stop-exposure groups). Oval cell hyperplasia and pigmentation of the liver were significantly increased in

WHO Food Additives Series No. 71-S1, 2016 36

Supplement 1: Non-dioxin-like polychlorinated biphenyls

the 3000 µg/kg bw per day core study group. At 2 years, two cholangiomas in the 1000 µg/kg bw per day group and two cholangiomas in the 3000 µg/kg bw per day stop-exposure group were seen. A single hepatocellular adenoma was seen in the 3000 µg/kg bw per day core study group. Hepatic cell proliferation, as measured by hepatocellular labelling index, was not significantly different between the vehicle control and dosed groups at any of the interim evaluations. At 2 years, there were significant increases in chronic active inflammation in the ovary and oviduct in the 1000 and 3000 µg/kg bw per day core study groups. Suppurative inflammation of the uterus in the 1000 µg/kg bw per day group and chronic active inflammation in the 3000 µg/kg bw per day core study group were significantly increased compared with vehicle controls. Sporadic incidences of minimal to mild follicular cell hypertrophy of the thyroid gland were seen at 53 weeks in all groups, except at 10 µg/kg bw per day. At 2 years, the incidences of minimal to mild follicular cell hypertrophy were significantly increased in the 300 µg/kg bw per day group and in the 3000 µg/kg bw per day (core and stop-exposure) groups. In the thyroid hormone assessments, serum total T4, free T4 and total T3 concentrations were significantly lower in the 3000 µg/kg bw per day group than in vehicle controls at the 14-week interim evaluation. At the 31-week interim evaluation, no significant differences were observed in serum total T4, free T4, total T3 or thyroid stimulating hormone (TSH) concentrations. At the 53-week interim evaluation, serum total T4 and free T4 concentrations in the 3000 µg/kg bw per day group were significantly lower than in vehicle controls. At PCB 153 dose levels of 100 μg/kg bw per day or more, large, significant and dose-related increases in hepatic PROD activities were found, with activities increased by 136-, 140- and 40-fold, compared with vehicle controls, at 14, 31 and 53 weeks, respectively. Increased PROD activity is characteristic of NDL-PCBs that induce the CYP2B subfamily of cytochrome P450 enzymes. However, there was also up to a 2-fold increase in EROD and acetanilide-4-hydroxylase (AHH) activities in the groups of PCB 153–dosed animals compared with vehicle controls at 14 and 31 weeks. Increases in EROD and AHH activities are characteristic of induction of the CYP1A subfamily of cytochrome P450 enzymes by compounds with dioxin-like activity that bind to AhR. However, this should be compared with the much higher increases of 50- to 100-fold in EROD activity and of 5-fold in AHH activity that were induced in a parallel NTP bioassay on PCB 126, a DL-PCB (NTP, 2006c), indicating that if there was some contamination with DLPCBs, it was very low. The overall NOAEL for effects on liver and thyroid for PCB 153 appears to be 10 μg/kg bw per day, equivalent to approximately 7 μg/kg bw per day if adjusted for 5 days/week dosing. 37

Safety evaluation of certain food additives and contaminants Eightieth JECFA

NTP (2006a) concluded that under the conditions of this 2-year gavage study, there was equivocal evidence of carcinogenic activity of PCB 153 in female rats, based on the occurrences of cholangioma of the liver. Although the numbers of cholangiomas were small, the occurrence of bile duct hyperplasia and oval cell hyperplasia could have contributed to cholangioma formation, and so the tumours may have been treatment related (NTP, 2006a). (ii) PCB 153 plus PCB 126

In a parallel carcinogenesis bioassay in female rats, NTP (2006b) evaluated the effects of combined treatment with PCB 153, an NDL-PCB, and PCB 126, a DL-PCB. This mixture study was conducted because previous studies had demonstrated interactions between PCB 153 and dioxin-like compounds on pharmacokinetics and biological effects. Groups of 80–81 female rats were given mixtures of PCB 126 and PCB 153 in corn oil:acetone (99:1) by oral gavage for 14, 31 or 53 weeks or 2 years. Both PCBs had an overall purity of greater than 99%. The mixture was given either as a constant ratio of PCB 126 (in ng/kg bw per day) to PCB 153 (in µg/ kg bw per day) of 10/10, 100/100, 300/300 or 1000/1000 or varying ratios of PCB 126 at 300 ng/kg bw per day to PCB 153 at 100, 300 or 3000 µg/kg bw per day. Dosing was on 5 days/week for up to 105 weeks. The main non-neoplastic finding was a significant increase in the incidence and severity of hepatotoxicity at 14, 31 and 53 weeks and 2 years. There were also numerous increases in the incidences of non-neoplastic lesions, notably in the lung, pancreas, adrenal cortex, thyroid gland, thymus, kidney, nose and forestomach, at 14, 31 and 53 weeks and/or 2 years. Neoplastic findings were increased incidences of hepatocholangiocarcinoma, hepatocholangioma and hepatocellular neoplasms (predominantly adenomas), squamous neoplasms of the lung (predominantly cystic keratinizing epithelioma) and gingival squamous cell carcinoma of the oral mucosa. Increased incidences of pancreatic acinar neoplasms were also considered to be treatment related. Increased incidences of uterine squamous cell carcinoma were possibly treatment related. NTP (2006b) concluded that under the conditions of this 2-year gavage study, there was clear evidence of carcinogenic activity of a constant ratio binary mixture of PCB 126 and PCB 153 in female rats. NTP (2006b) did not draw any direct comparisons with the NTP study on PCB 126 alone (NTP, 2006c) or discuss in general terms the impact of PCB 153 on the overall tumour response. ATSDR (2000) commented that some of the effects observed in other studies using an initiation–promotion model in which PCB 153 was administered as a promoting agent are consistent with the results of this NTP (2006b) study,

WHO Food Additives Series No. 71-S1, 2016 38

Supplement 1: Non-dioxin-like polychlorinated biphenyls

noting that in initiation–promotion studies, PCB 153 induces hepatic EROD and PROD activities and hepatocyte hypertrophy, but does not affect hepatocyte proliferation. It has been suggested that the development of thyroid follicular cell tumours in PCB-treated rats is attributable to a non-genotoxic mechanism whereby decreasing thyroid hormone levels after PCB treatment result in increased TSH levels, but other mechanisms may also be involved (EFSA, 2005; Knerr & Schrenk, 2006). This mode of action is considered a risk factor for the development of thyroid cancer in rodents, but not in humans (Capen, 1997; EFSA, 2005). (c) Hydroxylated PCBs

Two hydroxylated PCBs, 2′,4′,6′-trichloro-4-biphenylol (hydroxy-PCB 30) and 2′,3′,4′,5′-tetrachloro-4-biphenylol (hydroxy-PCB 61), which have relatively high estrogenic activity in vitro and in vivo, have been tested in the neonatal mouse model to examine the relationship between the estrogenicity and carcinogenicity of hydroxylated PCB congeners. The model used was the BALB/cCrgl female mouse, which has a low incidence of mammary tumours and well-documented neonatal responses to 17β-estradiol. Groups of 24–43 mice were given subcutaneous injections of 17β-estradiol and/or one or both of the hydroxylated PCBs every 24 hours for 5 days, beginning within 16 hours of birth, in the following doses and combinations: 5 µg 17β-estradiol alone; 2.5 µg 17β-estradiol plus 100 μg hydroxyPCB 30; 20 μg hydroxy-PCB 30; 200 μg hydroxy-PCB 30; 40 μg hydroxy-PCB 61; 400 μg hydroxy-PCB 61; 10 μg hydroxy-PCB 30 plus 10 μg hydroxy-PCB 61; or 100 μg hydroxy-PCB 30 plus 100 μg hydroxy-PCB 61. Negative controls were given the sesame oil vehicle. The mice were followed to 20 months of age. Mice treated with hydroxy-PCB 30 (200 µg/day) or 17β-estradiol (5 µg/day) showed similar increased incidences of cervicovaginal tract carcinomas (43% and 47%, respectively). Mice treated with hydroxy-PCBs as a mixture showed a change in the type of cervicovaginal tract tumour, shifting from predominantly squamous cell carcinoma to adenosquamous cell carcinoma. The authors of the study concluded that the individual hydroxylated PCBs tested were estrogenic and tumorigenic in mice exposed during development of the reproductive tract and that the results suggested that mixtures may act differently from individual compounds (Martinez, Stephens & Jones, 2005). (d) Studies on tumour promotion

The majority of rodent carcinogenicity studies on PCBs have used commercial PCB mixtures, and these have been shown to cause hepatocellular carcinomas in rats and mice and thyroid follicular cell adenomas in rats (ATSDR, 2000). 39

Safety evaluation of certain food additives and contaminants Eightieth JECFA

These studies are indicative that PCB mixtures can be complete carcinogens, where some congeners are acting as initiators and others as promoters (Ruiz et al., 2008). PCBs are considered to show tumour promotion activity in mice and co-carcinogenic effects in rats. Tumour promotion experiments have shown that after initiation with a genotoxic carcinogen, technical PCB mixtures and individual DL- and NDL-PCBs can act as liver tumour promoters in rodents. Some individual NDL-PCB congeners have been tested in tumour promotion studies after initiation with diethylnitrosamine (DEN). PCB 52 was reported to have (weak) promotor activity in rat liver in four studies, and PCB 153 was reported as having promotor activity in rat liver in two studies (Ahlborg, Hanberg & Kenne, 1992). In a medium-term tumour promotion assay in female rats, in which DEN was given as an initiator followed by partial hepatectomy, PCB 126 (a DL-PCB) alone, PCB 153 alone or a combination of the two PCB congeners was given by oral gavage 3 times per week for 8 weeks. When PCB 153 alone was given at doses of 10–10 000 µg/kg bw per day, it caused small, but statistically significant, increases in GST-positive liver cell foci area at 5000 and 10 000 µg/kg bw per day, doses that also caused significant increases in liver weight. PCB 126 alone caused a greater response than PCB 153 at the highest dose at which it was tested, 10 µg/kg bw per day (Dean et al., 2002). The promoting activity of PCB 153 in male mice initiated with DEN and the question of whether the deletion of the nuclear factor NF-κB p50 subunit influences liver carcinogenesis have been investigated. Four groups of 14–17 wild-type and transgenic mice were injected intraperitoneally with DEN at 9 weeks of age. After a 2-week recovery period, both wild-type and NF-κB p50−/− mice were injected intraperitoneally with PCB 153 at a dose of 0 (corn oil) or 300 µmol/kg bw every 14 days for a total of 20 injections. Mice were then maintained for an additional 15 weeks before being killed. The incidence of hepatocellular tumours, mainly classified as carcinomas, was higher in wild-type mice treated with PCB 153 than in wild-type mice receiving corn oil only. The deletion of p50 decreased the incidence of hepatocellular tumours in mice treated with PCB 153 or corn oil only (Glauert et al., 2008). Groups of 4–18 male mice were treated with an initiating intraperitoneal dose of DEN and then given a low or a high oral gavage dose of PCB 126, a DLPCB, or PCB 153, an NDL-PCB, or a low-dose mixture of the two PCBs between 3 and 24 weeks after initiation (Rignall et al., 2013). The low dose was adjusted to induce approximately 150-fold increases in cytochrome P450 (CYP1A1 for PCB 126 and CYP2B10 for PCB 153); the high dose was twice the low dose. To keep liver PCB levels constant, mice were given initial loading doses (low-dose groups: PCB 126 at 62 µg/kg bw, PCB 153 at 67.5 mg/kg bw; high-dose groups: PCB 126 at 124 µg/kg bw, PCB 153 at 135 mg/kg bw) followed by weekly maintenance doses (low-dose groups: PCB 126 at 9.5 µg/kg bw, PCB 153 at 10.5 mg/kg bw;

WHO Food Additives Series No. 71-S1, 2016 40

Supplement 1: Non-dioxin-like polychlorinated biphenyls

high-dose groups: PCB 126 at 19 µg/kg bw, PCB 153 at 21 mg/kg bw), calculated on the basis of the half-lives of the two PCBs. When given individually, the PCBs each produced dose-dependent increases in mRNA, protein and activity for the respective cytochrome P450 enzymes that they induce. Combined treatment caused more than additive effects on CYP1A1 mRNA expression, protein level and EROD activity. Changes in the levels of several proteins were detected by proteome analysis in livers of PCB-treated mice. The individual PCBs caused no significant increase in the number of glucose-6-phosphatase-deficient neoplastic lesions in the liver, whereas a moderate significant effect occurred in the combination group. These results suggested weak, but significant, responseadditive effects of the two PCBs when given in combination and that cytochrome P450 biomarkers tend to overestimate the carcinogenic response produced by the PCBs in mouse liver. A variety of mechanisms have been suggested for the tumour promotion activity of PCBs in the liver (reviewed by ATSDR, 2000; Knerr & Schrenk, 2006; Elabbas et al., 2013; IARC, 2015). One proposed mechanism is induction of oxidative stress, and there is evidence that some PCB metabolites are inducers of oxidative stress. Suppression of apoptosis in preneoplastic cells has also been proposed as a mechanism for the tumour promotion activity of PCBs in the liver. For example, in vitro studies have shown that six out of 20 NDL-PCB congeners tested lowered basal hepatic levels of the tumour suppressor protein p53 and attenuated the p53 response after treatment with inducers (Al-Anati, Högberg & Stenius, 2009), although one congener, PCB 180, has been shown to increase expression of p53 (Viluksela et al., 2014). Another proposed mechanism is inhibition of gap junctional intercellular communication (GJIC), which has been shown for NDL-PCBs in vivo and in vitro in rodent and human cells (IARC, 2015). DL-PCBs alter p53 signalling, but they have no acute effect on GJIC (Machala et al., 2003; Elabbas et al., 2013). The monochlorinated to hexachlorinated NDL-PCBs, on the other hand, are acute inhibitors of GJIC in vitro, with trichlorobiphenyls to hexachlorobiphenyls that have chlorine substitutions at the ortho position (e.g. PCB 136, PCB 153) being particularly potent inhibitors; heptachlorobiphenyls and octachlorobiphenyls have minimal or no GJIC inhibitory activity (Machala et al., 2003, 2004). The assay for GJIC inhibition showed good predictability for tumour promotion of ortho-substituted PCBs. Recently, inhibition of GJIC has been confirmed using single doses of ultrapure NDL-PCB congeners; among the six indicator PCBs, the inhibitory activity of PCB 52 was moderate, whereas that of the other five indicator PCBs was weak (Hamers et al., 2011). IARC (2015) noted that different cell- and connexin-specific mechanisms of action probably account for the inhibitory effects of PCBs on GJIC. PCB 153 decreased the number of gap junction plaques 41

Safety evaluation of certain food additives and contaminants Eightieth JECFA

and decreased levels of connexin 43 (constitutive protein of gap junctions) in liver epithelial cells; it enhanced proteasomal and lysosomal degradation of connexin 43 and inhibited trafficking of connexin 43 to the plasma membrane (Šimečková et al., 2009a). Additional non-genomic effects of NDL-PCBs on membrane-associated proteins, which are closely related to tumour promotion and progression, have been described by IARC (2015). They include the following observations. Initiation of male mice with DEN followed by exposure to PCB 153 appears to strongly select for Catnb-mutated, glutamine synthetase–positive tumours of the liver (Strathmann et al., 2006). In a rat liver progenitor WB-F344 cell line, PCB 153 was found to decrease levels of several proteins at adherens junctions involved in cell–cell communication and intracellular signalling, including E-cadherin, β-catenin and plakoglobin (Šimečková et al., 2009b). Oral administration of PCB 126 (a DL-PCB), mono-ortho-substituted PCB 118 (a DL-PCB) or PCB 153 (an NDL-PCB) differentially altered expression of the tight junction proteins claudin 5, occludin and ZO-1 in brain capillaries in C57/B16 mice and increased the permeability of the blood–brain barrier. In addition, the rates of formation and progression of brain metastases by luciferase-tagged melanoma cells were enhanced by a single oral gavage dose of PCB 118 or PCB 126 at 150 µmol/kg bw, with a lesser but still significant enhancement by PCB 153 (Seelbach et al., 2010). PCB 104 (an NDL-PCB) induced endothelial hyperpermeability of human microvascular endothelial cells, HMEC-1, and transendothelial migration of human breast cancer cells, MDA-MB-231. These effects were associated with overexpression of vascular endothelial growth factor (Eum et al., 2004). PCB 104 and PCB 153 induced expression of intercellular adhesion molecule-1 (ICAM1), vascular cell adhesion molecule-1 (VCAM-1) and monocyte chemoattractant protein-1 (MCP-1) in the liver, lung and brain of male mice, and PCB 104 also increased levels of matrix metalloproteinase-7 (MMP-7) mRNA in the liver and brain; these are proinflammatory mediators that can contribute to metastasis (Sipka et al., 2008). A mixture of seven NDL-PCBs (PCBs 28, 52, 101, 138, 153, 180 and 209) increased cell motility of human non-metastatic MCF-7 cells and human metastatic breast cancer MDA-MB-231 cells in vitro via production of reactive oxygen species and activation of the Rho-associated kinase (ROCK); in an in vivo study in mice, PCBs significantly promoted disease progression, leading to enhanced capability of metastatic breast cancer cells to metastasize to bone, lung and liver (Liu, Li & Du, 2010). Human skin keratinocytes were exposed to a synthetic mixture of volatile PCBs or to PCB 28 or PCB 52, both prominent airborne NDL-PCB congeners, for up to 48 days. Compared with untreated control cells, the PCB mixture and the two congeners significantly inhibited telomerase activity from day 18, whereas telomere length was reduced by PCB 52 from day 18 and by PCB 28 and by the mixture from day 30 onwards

WHO Food Additives Series No. 71-S1, 2016 42

Supplement 1: Non-dioxin-like polychlorinated biphenyls

(Senthilkumar et al., 2011). In a similar study on PCB 153, telomerase activity, telomere length and cell growth were significantly reduced, whereas intracellular superoxide levels were increased, compared with controls, from day 6 to day 48, suggesting that superoxide may be one of the factors regulating telomerase activity, telomere length and cell growth (Senthilkumar, Robertson & Ludewig, 2012). 2.2.4 Genotoxicity

Genotoxicity studies on commercial PCB mixtures and on individual PCB congeners have been summarized by ATSDR (2000) and IARC (2015). (a) Commercial mixtures of PCBs

For commercial mixtures, the experimental in vitro test systems used have included gene mutation in bacterial and yeast cells, gene mutation, chromosomal aberration and micronucleus formation in animal and human cells, DNA strand breaks in animal and human cells, unscheduled DNA synthesis (UDS) in mammalian cells and DNA adduct formation in mammalian and human cells. The majority of such studies on commercial mixtures showed no genotoxicity, with only a few giving positive results. In vivo studies on commercial mixtures have included investigation of UDS and DNA adducts in rat, mouse and monkey hepatocytes, chromosomal aberrations in rat and mouse bone marrow and spermatogonia, micronucleus formation in mouse bone marrow, dominant lethal mutations in rats, gene mutations in transgenic mice and DNA strand breaks (comet assay) in various mouse tissues. As with the in vitro studies, the majority of in vivo studies on commercial mixtures showed no genotoxicity, with only a few tests giving positive or weakly positive results (IARC, 2015). Other reviews have concluded that as commercial PCB mixtures exhibit no or minimal mutagenic activity in most assay systems, this suggests that PCBs are not potent genotoxicants (ATSDR, 2000) and that those exhibiting carcinogenicity are probably acting as indirect, non-genotoxic carcinogens (Safe, 1989; Knerr & Schrenk, 2006). (b) General considerations on PCB metabolites

Some PCB metabolites, particularly those from lower chlorinated congeners, may be more reactive towards DNA than their parent PCB congeners. Robertson & Gupta (2000), for example, showed that metabolism of PCBs can generate electrophilic metabolites and reactive oxygen species that could damage DNA. Srinivasan et al. (2001) showed that dihydroxylated PCBs and PCB quinones, after reaction with glutathione, produced reactive oxygen species in a human cell line and oxidative DNA damage in the form of DNA strand breaks in bacteria 43

Safety evaluation of certain food additives and contaminants Eightieth JECFA

WHO Food Additives Series No. 71-S1, 2016 44

(Escherichia coli). Pereg et al. (2001) pointed out that the issue of covalent binding of PCB metabolites to macromolecules was unclear; in vitro studies showed macromolecular binding, including binding to DNA, but conflicting results were obtained in vivo from studies that used different animal models and techniques of differing sensitivity for detection of binding and adduct formation. Formation of a guanine adduct has been reported after incubation of calf thymus DNA with quinones of lower chlorinated PCBs (Zhao et al., 2004). Wangpradit et al. (2009) showed that an enzyme produced in human ovary, breast and prostate tissue, prostaglandin H synthase, which has both cyclooxygenase and peroxidase activities, oxidizes three dihydroxy metabolites of PCB 3 to the corresponding quinones, which are reactive electrophilic species with a potential for protein and DNA damage. Cu2+-mediated activation of PCB catechol and hydroquinone metabolites induces oxidative damage and polar DNA adducts (Spencer et al., 2009). Overall, these studies show that liver enzymes metabolize lower chlorinated PCB congeners to reactive intermediates, such as epoxides, quinones and reactive oxygen species, that have the potential to modify nucleotides and DNA, hence causing mutations, and that oxidative damage may be involved in the production of liver tumours after exposure to PCBs (Ludewig & Robertson, 2013). Recently, lower chlorinated PCB 3 metabolites have been shown to induce gene mutations, chromosome breaks, chromosome loss and polyploidy in cells in culture, providing the first evidence of their potential for mutagenicity (Robertson & Ludewig, 2011). The mutagenicity in vitro of PCB 3 metabolites indicates the need for further studies to assess the risks associated with human exposure to lower chlorinated hydroxy-PCBs. Ruiz et al. (2008) used a quantitative structure–activity relationship (QSAR) approach to assess the potential for mutagenesis and carcinogenesis of all 209 individual PCB congeners (both DL- and NDL-PCBs) and their possible hydroquinone and benzoquinone metabolites. Their analysis concluded that monochlorinated and dichlorinated PCBs and their metabolites are predicted to have a high probability of being mutagenic. The probability of mutagenicity decreased with increasing numbers of chlorine atoms, and higher chlorinated PCBs were predicted to be non-mutagenic. The predictions were in agreement with experimental data on DNA adduct formation. The higher chlorinated PCBs were, however, predicted to be carcinogenic, particularly in the female mouse model. The benzoquinone metabolites of PCBs could be carcinogenic (but the weight of evidence is poor), and hydroquinone metabolites were less likely than benzoquinone metabolites to be carcinogenic.

Supplement 1: Non-dioxin-like polychlorinated biphenyls

(c) Studies on individual NDL-PCBs and their metabolites

The results of in vitro and in vivo studies in which individual NDL-PCBs and/or their metabolites from the indicator group have been tested are shown in Table 7. Table 7 Genotoxicity studies on individual NDL-PCB congeners and their metabolites Results Without With metabolic metabolic activation activation − − − − − + − + − + + + + + + + − + − + − NT + − + + − NT NT NT NT NT NT NT NT NT NT NT NT NT NT NT NT NT NT NT NT NT NT NT NT NT NT −

PCB congener no. / metabolite In vitro studies PCB 3 2′-OH-PCB 3 3′-OH-PCB 3 4′-OH-PCB 3 PCB 3-3′,4′ hydroquinone PCB 3-3′,4′ quinone PCB 3-2′,5′ hydroquinone PCB 3-2′,5′ quinone PCB 3 2′-OH-PCB 3 3′-OH-PCB 3 4′-OH-PCB 3 PCB 3-3′,4′ hydroquinone PCB 3-3′,4′ quinone PCB 3-2′,5′ hydroquinone PCB 3-2′,5′ quinone PCB 3 PCB 3-3′,4′ hydroquinone PCB 3-3′,4′ quinone PCB 3-2′,5′ hydroquinone PCB 3-2′,5′ quinone PCB 3 PCB 3-2,5 hydroquinone PCB 3-2,5 quinone PCB 52

Test system Chinese hamster V79 cells – gene mutation

Effective dose Reference Zettner et al. (2007)

0.6 µmol/L 0.5 µmol/L 50 µmol/L 100 µmol/L 75 µmol/L 15 µmol/L 5 µmol/L 2.5 µmol/L 1 µmol/L 2.5 µmol/L 5 µmol/L Zettner et al. (2007)

Chinese hamster V79 cells – micronucleus formation

Chinese hamster V79 cells – sister chromatid exchange or polyploidy Comet assay – HL-60 cells – Jurkat cells – HL-60 cells – Jurkat cells Salmonella typhimurium TA1538 – reverse mutation S. typhimurium TA98, TA100, TA1535, TA1537 – reverse mutation

Flor & Ludewig (2010)

10 µmol/L 5 µmol/L 5 µmol/L Wyndham, Devenish & Safe (1976) Hsia, Lin & Allen (1978)

PCB 52 4-OH-PCB 52 3,4-epoxy-PCB 52

NT NT NT

− − −

45

Safety evaluation of certain food additives and contaminants Eightieth JECFA

Table 7 (continued) Results Without With metabolic metabolic activation activation + + + (+) NT NT NT NT

PCB congener no. / metabolite PCB 52 4-OH-/3-OH-2,2′,5,5′-tetrachlorobiphenyl (4:1) 3,4-epoxy-PCB 52 PCB 52

Test system Mouse fibroblast L-929 cells – DNA strand breaks (alkaline sedimentation) Human lymphocytes (6 donors) – DNA strand breaks (comet assay) Human lymphocytes (4 donors) – sister chromatid exchange Human lymphocytes (5–9 donors) – chromosomal aberrations Fish fibroblast RTG-2 cells – DNA strand breaks (comet assay) Fish fibroblast RTG-2 cells – micronucleus formation Human lymphocytes – sister chromatid exchange Human lymphocytes – micronucleus formation Fish fibroblast RTG-2 cells – DNA strand breaks (comet assay) Fish fibroblast RTG-2 cells – micronucleus formation Human lymphocytes (5–9 donors) – chromosomal aberrations (structural)

Effective dose Reference 20 µg/mL 20 µg/mL 10 µg/mL 0.3 µg/mL Stadnicki & Allen (1979)

Sandal, Yilmaz & Carpenter (2008)

PCB 52

−

NT

Sargent, Rollof & Meisner (1989)

−

NT

PCB 101

+

NT

16 µg/mL

Marabini, Calò & Fucile (2011)

PCB 101

+

NT

16 µg/mL

3′-MeSO2-2,2′,4′,5,5′-pentachlorobiphenyl (metabolite of PCB 101)

−

NT

−

NT

WHO Food Additives Series No. 71-S1, 2016

PCB 138

+

NT

25 µg/mL

Marabini, Calò & Fucile (2011)

PCB 138

−

NT

PCB 153

+

NT

1 µg/mL

Sargent, Rollof & Meisner (1989)

46

Supplement 1: Non-dioxin-like polychlorinated biphenyls

PCB congener no. / metabolite PCB 153

PCB 153

PCB 153

PCB 153

PCB congener no. In vivo studies PCB 52

Test system Human breast epithelial MCF-10A cells – micronucleus formation Human hepatoma HepG2 cells – micronucleus formation Fish fibroblast RTG-2 cells – DNA strand breaks (comet assay) Fish fibroblast RTG-2 cells – micronucleus formation Test system Female SpragueDawley rat, 70% hepatectomy, bone marrow cells – chromosomal aberrations (numerical and structural) Female SpragueDawley rat, liver cells after 70% hepatectomy – chromosomal aberrations (numerical) Female SpragueDawley rat, liver or brain – DNA adducts, M1dG secondary oxidative DNA lesion

Results Without With metabolic metabolic activation activation + NT

Effective dose Reference 0.4 µg/mL Venkatesha et al. (2008)

+

NT

36 µg/mL

Wei et al. (2009) Marabini, Calò & Fucile (2011)

+

NT

11 µg/mL

+

NT

11 µg/mL

Results −

Dose 10 mg/kg in feed, 1 year

Reference Meisner et al. (1992)

PCB 52

−

10 mg/kg in Sargent et al. feed for 7 or 10 (1992) months

PCB 153

−

1 mg/kg bw orally × 5 per week for 53 weeks

Jeong et al. (2008)

diOH: dihydroxy; MeSO2: methyl sulfone; NT: not tested; OH: hydroxy; +: considered to be positive; (+): considered to be weakly positive in an inadequate study; –: considered to be negative Source: Adapted from IARC (2015)

47

Safety evaluation of certain food additives and contaminants Eightieth JECFA

No studies on PCB 28 or PCB 180 have been found. PCB 52 and its two major metabolites showed no activity in tests for gene mutation in five strains of Salmonella typhimurium. PCB 52 was negative in a test for chromosomal aberrations in human lymphocytes. PCB 52 and its major metabolites were positive or weakly positive in tests for DNA strand breaks in mouse fibroblasts and human lymphocytes, but negative for sister chromatid exchange in human lymphocytes. PCB 52 was negative in two in vivo studies on chromosomal aberrations in bone marrow and liver. PCB 101 has not been tested in any definitive genotoxicity studies in vitro or in vivo; in other tests, PCB 101 was positive for DNA strand breaks and micronucleus formation in fish cells. Its major methyl sulfone metabolite was negative in a test for micronucleus formation in human lymphocytes and negative in an indicator test for sister chromatid exchange. There are no in vivo tests on the parent congener or its metabolites. PCB 138 has been tested in fish cells only. It was positive for strand breaks but negative for micronucleus formation. PCB 153 was positive in three separate tests for chromosomal aberrations and for micronucleus formation in human cells. It was also positive in tests for DNA strand breaks and micronucleus formation in fish cells. In an in vivo study, PCB 153 did not form DNA adducts in rat liver or brain. 2.2.5 Reproductive and developmental toxicity (a) Reproductive studies on commercial mixtures of PCBs

Early reproductive studies were conducted mainly on commercial mixtures. In male rodents, oral exposure to PCB mixtures caused reductions in spermatozoa, male fertility, and ventral prostate and seminal vesicle weights. In female rodents, adverse effects on the estrous cycle and ovulation and reduction in weights of female reproductive organs were reported (Ahlborg, Hanberg & Kenne, 1992). The ATSDR (2000) review concluded that there was evidence for adverse effects on weights of male reproductive organs and spermatozoa in rodents and monkeys, but the evidence was limited. For female animals, the review concluded that reproductive toxicity had been established in a number of oral studies with commercial PCB mixtures. The effects observed included prolonged estrus, decreased sexual receptivity and reduced implantation rate in adult rats and/or their offspring exposed during gestation and lactation, decreased conception rate in mice, partial or total reproductive inhibition in mink, and prolonged menstruation and decreased fertility in monkeys. Mink and monkeys were considered to be particularly sensitive, with effects occurring at doses in the range of 0.1–1 mg/kg bw per day in intermediate-duration studies and as low as 0.02 mg/kg bw per day in monkeys following chronic exposure (ATSDR, 2000; EFSA, 2005).

WHO Food Additives Series No. 71-S1, 2016 48

Supplement 1: Non-dioxin-like polychlorinated biphenyls

(b) Reproductive studies on individual NDL-PCBs and their metabolites (i) PCB 28

The direct effects of two NDL-PCBs, PCB 28 and PCB 30, on fresh spermatozoa of bulls aged 2–4 years were evaluated in vitro. Median inhibitory concentrations (IC50 values) for cytotoxicity of 8.45 and 5.45 ng/mL were measured for PCB 28 and PCB 30, respectively, and divisions or multiples of the IC50 were used as the doses to assess effects on sperm motility and viability. Total motility, progressive motility and viability decreased in a dose- and duration-dependent manner. Total motility, at the IC50 dose following 2 hours of exposure, decreased by 72% for PCB 28 and by 61% for PCB 30. Motility results were in accordance with the viability and morphology data showing that total abnormalities (especially acrosome reaction rate) were increased (Yurdakok et al., 2015). (ii) PCB 132

Groups of 12 pregnant rats were given PCB 132 at a dose of 0 (corn oil), 1 or 10 mg/ kg bw as a single intraperitoneal injection on gestation day (GD) 15 to investigate effects on reproductive parameters of male offspring. The male offspring were killed, and the epididymal sperm counts, motility, velocity, reactive oxygen species generation, sperm–oocyte penetration rate, testicular histopathology, apoptosisrelated gene expression and caspase activation were assessed on postnatal day (PND) 84. Cauda epididymal weight was reduced at both doses, but the reduction was not dose related and was significant only at 1 mg/kg bw. There was no effect on testis histopathology. Epididymal sperm count and motile epididymal sperm count were significantly reduced at both doses in a dose-related manner. Doserelated increases in reactive oxygen species and reductions in sperm–oocyte penetration rate were observed in spermatozoa, reaching statistical significance at the higher dose. In the low-dose group, p53 was significantly induced and caspase-3 was inhibited. In the high-dose group, activation of caspase-3 and caspase-9 was significantly increased, whereas the expressions of the apoptosisrelated genes Fas, Bax, bcl-2 and p53 were significantly decreased. The authors commented that these intraperitoneal doses would give rise to levels of PCB 132 in tissues that are several orders of magnitude higher than any present in human tissues (Hsu et al., 2007). (iii) PCB 153

PCB 153 has been shown to affect bovine oocytes cultured in vitro. Concentrations of 0.84, 8.4 and 84 ng/mL had no effect on oocyte maturation, but resulted in a reduced proportion of oocytes undergoing cleavage at the highest concentration. There were no differences in blastocyst development (Krogenaes et al., 1998). 49

Safety evaluation of certain food additives and contaminants Eightieth JECFA

The effect of three different mixtures of PCB 153 and PCB 118, a DLPCB, on fetal testis development in sheep has been investigated. Pregnant ewes 2–7 years of age were treated by oral gavage 3 times a week, from mating until euthanasia (day 134). Originally, two exposure groups were planned to receive either 98 μg/kg bw per day of PCB 153 or 49 μg/kg bw per day of PCB 118, with a third group receiving corn oil only. However, there were two episodes of crosscontamination, resulting in a mixed exposure to PCBs in all three groups. This produced three groups of fetuses with distinct PCB levels in adipose tissue: high PCB 153/low PCB 118 (n = 13), low PCB 153/high PCB 118 (n = 14) and low PCB 153/low PCB 118 (n = 14). The corresponding levels of PCB 153 and PCB 118 in adipose tissue (in ng/g lipid) were 47 607/460 for the high PCB 153/low PCB 118 group, 2545/6118 for the low PCB 153/high PCB 118 group, and 1565/1969 for the low PCB 153/low PCB 118 group. Fetal body weight was significantly lower in the high/low and low/high PCB groups compared with the low/low PCB group, but there were no significant differences in testis weight corrected for body weight. Circulating testosterone level and testis morphology showed no differences between the three groups. Proteomic investigations showed 26 statistically significant spot alterations in proteins involved in 15 functional pathways and 10 structural pathways in the testes of fetuses exposed to high PCB 153/low PCB 118 or low PCB 153/high PCB 118, relative to low PCB 153/low PCB 118, the pathways involved indicating effects on stress response, protein synthesis or cytoskeleton regulation (Krogenaes et al., 2014). (c) Developmental toxicity studies on commercial mixtures of PCBs The majority of developmental toxicity studies on NDL-PCBs have focused on outcomes related to developmental neurotoxicity, and these are summarized below in section 2.2.5(f). Below, only toxicity studies investigating multiple developmental outcomes other than neurotoxicity are summarized. Early developmental toxicity studies were conducted mainly on commercial mixtures and have been summarized elsewhere (Ahlborg, Hanberg & Kenne, 1992; ATSDR, 2000; Ulbrich & Stahlmann, 2004). Studies on mice, rats, rabbits, guinea-pigs, mink and monkeys showed one or more of reduced implantations, increased resorptions, increased abortions, reduced fetal growth and delayed development, increased length of gestation, reduced litter size, reduced birth weight, reduced perinatal survival, reduced growth of offspring and delayed developmental landmarks. Postnatal effects on learning and behaviour were seen in rat, mouse and monkey. In the mouse, commercial mixtures caused teratogenicity (hydronephrosis and cleft palate), which was considered to be attributable to the DL-PCB congeners (interaction with AhR). Doses causing developmental toxicity were as low as 0.1 mg/kg bw per day, with

WHO Food Additives Series No. 71-S1, 2016 50

Supplement 1: Non-dioxin-like polychlorinated biphenyls

some behavioural effects in monkeys reported at lower doses, down to 0.006 mg/ kg bw per day. Developmental toxicity in rodents can occur in the absence of maternal toxicity, but maternal toxicity was also evident in monkeys, even at low doses. (d) Developmental toxicity studies on individual NDL-PCBs and their metabolites (i) PCB 28

PCB 28 was given orally by gavage to groups of 6–9 pregnant rats on GDs 10–16 at a dose of 0, 8 or 32 mg/kg bw per day. In the offspring from the high-dose group, there was reduced birth weight and increased liver weight at weaning. High-dose female offspring also showed slower learning in a T-maze (see section 2.2.5(f)). There was no effect on plasma T4 concentrations (Ness et al., 1993; Schantz, Moshtaghian & Ness, 1995). (ii) PCB 52

In a study conducted under the ATHON project (Elabbas et al., 2013), PCB 52 was given orally by gavage to groups of 5–6 pregnant rats between GD 7 and PND 10. PCB 52 was dissolved in corn oil, and the total dose levels were 0, 30, 100, 300, 1000 and 3000 mg/kg bw. Controls received only the vehicle. The total dose was divided into 10 equal subdoses given on GDs 7, 9, 11, 14 and 16 and PNDs 1 and 10. Litter size was adjusted to five males and five females on PND 1, pups were weaned on PND 28, and samples were taken from one male and one female offspring on PNDs 7, 35 and 84. Parameters monitored were “development of the reproductive system, bone and teeth, sex steroidogenesis, limited histopathology, biochemistry and molecular biology”. Body weight development, mortality and developmental milestones (eye opening, tooth eruption) of the offspring after perinatal exposure were not affected by treatment. An increase in liver weight in offspring was observed (no further details were provided). The developmental neurotoxicity findings from this study are described in section 2.2.5(f). Other findings from this study have not yet been published. (iii) PCB 153

PCB 153 was given orally by gavage to groups of 6–9 pregnant rats daily from GD 10 to GD 16 at a dose of 0, 16 or 64 mg/kg bw per day. Postnatally, offspring showed reduced body weight during the preweaning period and increased liver weight at weaning in the high-dose group. High-dose males showed decreased response latency in a radial arm maze, and high-dose female offspring showed slower learning in a T-maze (see section 2.2.5(f). Plasma T4 but not T3 concentrations were reduced at both doses (Ness et al., 1993; Schantz, Moshtaghian & Ness, 1995). 51

Safety evaluation of certain food additives and contaminants Eightieth JECFA

Groups of 10 pregnant rats were exposed to PCB 153 daily at a dose of 0, 16 or 64 mg/kg bw per day orally by gavage from GD 10 to GD 16. The dams were allowed to litter out and raise their offspring. At 1 or 3 weeks of age, the offspring were examined for changes in developmental parameters. There were no treatment-related effects on body weight, body length, tail length, anogenital distance or weights of kidneys, testes, ovaries or uterus. Liver weight was significantly increased in the high-dose male offspring at 1 and 3 weeks of age, but not in female offspring. There was a dose-dependent decrease in concentrations of T4 in plasma in males and females combined at 1 and 3 weeks of age, which was statistically significant in the high-dose group. Concentrations of T3 in plasma were reduced in females in a dose-related manner at 3 weeks of age, and the reduction was statistically significant in the high-dose group. TSH level was not affected. There were no changes in concentrations of growth hormone or insulinlike growth factor-1 (IGF-1) in plasma in any dose group (Kobayashi et al., 2008). In a subsequent study using 10 pregnant rats per dose group but lower doses of 0, 1 and 4 mg/kg bw per day, the same parameters were assessed in the offspring at 1, 3 or 9 weeks of age. There were no effects on any parameter, apart from a doserelated increase in T3 concentration in plasma in males at 1 week of age; at 3 and 9 weeks of age, there were no effects of treatment on T3 concentration in males. There were no effects on T3 concentration in female offspring at any age, nor were T4 or TSH levels affected in any group (Kobayashi et al., 2009). Lyche et al. (2004a) investigated the possible effects of gestational exposure to environmental levels of PCB 153, an NDL-PCB, or PCB 126, a DLPCB, on the hypothalamic–pituitary–gonadal axis in goat kids. Groups of 10 pregnant does were given corn oil vehicle only, PCB 153 (estimated dose 98 µg/kg bw per day) or PCB 126 (estimated dose 49 ng/kg bw per day) orally, 3 times per week, from GD 60 until delivery. Pre-pubertal and post-pubertal concentrations of LH, FSH, prolactin and progesterone in plasma were analysed. LH, FSH, prolactin and progesterone concentrations were also measured during an induced estrous cycle. There were no effects on body weight of the kids. In female offspring exposed to PCB 153, pre-pubertal LH concentration was significantly lower than in controls, there was a significant delay in onset of puberty (by an average of about 8 days) and there was a significant increase in progesterone level during the luteal phase of an induced estrous cycle at 9 months of age. There was no effect on prolactin or FSH level. PCB 126 did not produce any effects on the hypothalamic–pituitary–gonadal axis. The mean concentrations of PCBs in adipose tissue in the goat offspring at 9 months of age were 5.8 μg/g (lipid weight) and 0.49 ng/g (lipid weight) for PCB 153 and PCB 126, respectively. In a further study by the same research group (Zimmer et al., 2009), groups of 10 pregnant goats were treated with corn oil only, PCB 153 or PCB 126, as described above. In male offspring exposed to PCB 153, mean basal cortisol

WHO Food Additives Series No. 71-S1, 2016 52

Supplement 1: Non-dioxin-like polychlorinated biphenyls

concentrations were significantly lower around the onset of puberty and during their first breeding season, compared with controls. Male goat kids exposed to either PCB congener showed a greater and more prolonged rise in plasma cortisol levels compared with controls when animals were subjected to mild stress at 9 months of age. Neither the basal maternal cortisol level in plasma nor adrenal masses in goat kids were affected by PCB exposure. Lundberg et al. (2006) studied the effects of perinatal exposure to PCB 153 and PCB 126 in female goat offspring. The pregnant goats were exposed to 98 μg/kg bw per day of PCB 153 or 49 ng/kg bw per day of PCB 126 in corn oil from GD 60 until delivery. The offspring were also exposed to the PCBs during the lactation period of 6 weeks. The diaphyseal bone was analysed at a distance of 18% and 50% of the total bone length, and the metaphyseal bone at a distance of 9%. Also, biomechanical three-point bending of the bones was conducted, with the load being applied to the mid-diaphyseal peripheral quantitative computed tomography measure point (50%). Compared with non-exposed goats, PCB 153 exposure significantly decreased the total cross-sectional area, decreased the cross-sectional area of the marrow cavity, decreased the moment of resistance at the diaphyseal 18% measure point and increased the trabecular bone mineral density at the metaphyseal measure point. PCB 126 exposure did not produce any observable changes in bone tissue. The biomechanical testing showed no significant differences between the exposed and control groups for either congener. (iv) PCB 180

In a study conducted under the ATHON project (Elabbas et al., 2013), PCB 180 was given orally by gavage to groups of 5–6 pregnant rats between GD 7 and GD 10. PCB 180 was dissolved in corn oil, and the total dose levels were 0, 10, 30, 100, 300 and 1000 mg/kg bw. Controls received only the vehicle. The total dose was divided into four equal subdoses given on GDs 7–10. Litter size was adjusted to four males and four females on PND 1, pups were weaned on PND 28, and samples were taken from one male and one female offspring on PNDs 7, 35 and 84. Parameters monitored were “development of the reproductive system, bone and teeth, sex steroidogenesis, limited histopathology, biochemistry and molecular biology”. Maternal body weight development was slightly decreased at the highest dose level. Offspring body weight was decreased during the first weeks of life after high-dose exposure, but recovered thereafter. Neonatal mortality was slightly and dose-dependently increased at the two highest dose levels. Analysis of developmental milestones revealed slight delays in balano-preputial separation and vaginal opening. Tooth eruption, eye opening and anogenital distance were not affected. Offspring liver weights were dose-dependently increased on PNDs 53

Safety evaluation of certain food additives and contaminants Eightieth JECFA

7, 35 and 84, but not on PND 1. The developmental neurotoxicity findings from this study are described in section 2.2.5(f). Other findings from this study have not yet been published. (v) Mixture of PCBs 138, 153, 180 and 126

The effects of a single dose of a reconstituted mixture of NDL- and DL-PCBs were studied in rat dams and offspring after maternal exposure (Cocchi et al., 2009; Colciago et al., 2009). The mixture contained equal concentrations by weight of three NDL-PCBs (PCB 138, PCB 153 and PCB 180), together with a DL-PCB (PCB 126), which was added at a ratio of 1:10 000 of the total mixture. Groups of animals were given 10 mg/kg bw per day of the mixture by subcutaneous injection, daily on GDs 15–19 and twice a week during lactation until PND 21. The dose averaged over the treatment period was estimated to be 3.7 mg/kg bw per day. Offspring were sacrificed on PND 21 and PND 60. Maternal body weight, litter size, sex ratio of the offspring and postnatal mortality were not affected by the treatment. Offspring body weight after weaning was significantly decreased in both males and females. The relative (to body weight) weights of ovary, testis and prostate were not affected. Plasma testosterone concentrations in females in diestrus or in males were not affected on PND 60. No treatment-related effects were observed in free T3 or T4 concentrations in plasma on PND 21 or PND 60. Concentrations of IGF-1 in plasma were significantly higher in mixture-treated males and females than in controls on PND 21, but the difference diminished by PND 60. Alterations observed in the hypothalamic–pituitary growth hormone axis included increased somatostatin expression in the hypothalamic periventricular nucleus in both sexes and in the lateral arcuate nucleus in males only. Expression of growth hormone in the anterior pituitary was decreased in males only. These changes were observed on PND 60. Treatment-related changes were observed in structure, geometry and mineral content of long bones in the offspring examined on PND 60. The width of uncalcified epiphyseal cartilage in the tibia was significantly increased in both sexes. There was decreased whole bone planar area in both sexes and decreased bone mineral content in males. Using peripheral quantitative computed tomography, changes were observed only in the mid-diaphysis of tibiae of male offspring. These included decreased total bone and trabecular bone area, decreased bone mineral content and decreased periosteal and endosteal diameter and bone thickness. The bone strength index was also decreased in males (Cocchi et al., 2009). Treatment with the mixture resulted in several alterations in the normal dimorphic expression pattern of the androgen activating enzymes, aromatase

WHO Food Additives Series No. 71-S1, 2016 54

Supplement 1: Non-dioxin-like polychlorinated biphenyls

and 5α-reductases 1 and 2, in the hypothalamus. Expression of aromatase was significantly increased in males and slightly decreased in females on PND 21. Hypothalamic expression of 5α-reductase 1 was decreased in females on PND 21, and the decrease persisted until PND 60. In males, expression of 5α-reductase 1 was decreased only on PND 60. In contrast, expression of 5α-reductase 2 was significantly increased in females, but slightly decreased in males, on PND 60. In spite of reduced body weight, the onset of puberty was shown to occur earlier in mixture-treated females than in controls, as indicated by the age and body weight at vaginal opening. In contrast, testicular descent in males was delayed. Studies on copulatory behaviour of adult male rats revealed slight but significant delays in normal male sexual behaviour; female sexual behaviour was not affected. Learning and memory tests showed no changes in the spatial memory test in either sex, but there was a significantly prolonged latency in the passive avoidance test in males. No changes were observed in spontaneous locomotor activity or in depression and anxiety behaviour (Colciago et al., 2009). (vi) 4-Hydroxy-PCB 107

Exposure of pregnant rats to 14C-labelled 4-hydroxy-PCB 107, given orally by gavage from GD 10 to GD 16, resulted in accumulation of 4-hydroxy-PCB 107 in fetal livers, brain and plasma measured at GD 17 and GD 20 (Meerts et al., 2002). Thyroid hormone status and metabolism were studied in groups of pregnant rats given 4-hydroxy-PCB 107 (14C-labelled or unlabelled) orally by gavage at 5 mg/kg bw on GDs 10–16. Fetuses were studied at GD 17 and GD 20. The test compound accumulated in the fetal compartment, with fetal/maternal ratios of 11.0, 2.6 and 1.2 in liver, cerebellum and plasma, respectively, at GD 20. Radiolabel was bound to plasma transthyretin in dams and fetuses. Concentrations of total T4 and free T4 in fetal plasma were significantly decreased at GD 17 and GD 20 (89% and 41%, respectively, at GD 20), whereas concentrations of TSH in fetal plasma were increased more than 2-fold at GD 20. No effects were seen on T3 concentrations in fetal brain (Meerts et al., 2002). The same group investigated the effects of exposure of rats to 4-hydroxyPCB 107 given orally by gavage on GDs 10–16 at a dose of 0 (corn oil), 0.5 or 5 mg/kg bw per day on postnatal development, sex steroid hormone levels in offspring and reproduction of the F1 females. F0 group sizes were 8–11 dams. There were no effects on F0 maternal toxicity or reproductive parameters. After birth, F1 litters were culled to four of each sex per litter, weaned at PND 21 and allowed to mature until 11 months of age. Body weights, organ weights and developmental landmarks, including sexual maturation (vaginal opening and preputial separation), were unaffected by treatment. Half the cohort of offspring was mated at approximately 260 days of age, and the pregnant females were 55

Safety evaluation of certain food additives and contaminants Eightieth JECFA

killed on GD 20; there was no effect on reproduction in this F1 generation. In the unmated cohort, there was a statistically significant and dose-related prolongation of the estrous cycle in female offspring from both treated groups when monitored between PND 210 and PND 231. The mean plasma estradiol concentration in 11-month-old female offspring was significantly increased by 230% in the group given 4-hydroxy-PCB 107 at 5 mg/kg bw per day, and the estradiol:progesterone ratio was also increased in this group (both only at the pro-estrous stage). No effects on male or female reproductive organ weights or on testosterone levels at PNDs 310–325 could be detected (Meerts et al., 2004). (e) Developmental neurotoxicity: mechanistic aspects

Multiple studies with NDL-PCBs and rodents have shown that these compounds cause neurobehavioural effects, especially after prenatal and/or postnatal exposure. Common end-points that are affected by NDL-PCBs include locomotor activity, spontaneous behaviour, habituation capability, spatial learning and anxiety. Studies with mice and rats maternally exposed to NDL-PCBs before or after birth showed an increase, as well as a decrease, in spontaneous activity (Holene et al., 1998; Gralewicz et al., 2009; Boix, Cauli & Felipo, 2010; Boix et al., 2011). Additionally, exposure to NDL-PCBs can cause impairment of spatial learning and memory abilities in rodents (Piedrafita et al., 2008a,b; Boix, Cauli & Felipo, 2010). It should be noted that these neurodevelopmental studies usually involved a limited number of dose levels. This makes it impossible to determine relative effect potencies and QSARs for the NDL-PCBs. In vivo studies on neurobehavioural or developmental effects using hydroxylated metabolites of NDL-PCBs are lacking. This is despite the fact that in vitro studies point towards a potentially important role for hydroxylated metabolites in neurotoxic mechanisms. Only one perinatal study used a hydroxylated PCB (4-hydroxy-PCB 106) in rats, which indicated hyperactivity in the male offspring (Lesmana et al., 2014). Although the number of studies is limited, the results of neurobehavioural studies with NDL-PCBs in rodents indicate comparable patterns of effects, albeit with congener-specific differences. Based on the limited number of in vivo studies, a distinct mechanism of action for the neurodevelopmental effects of NDL-PCBs cannot be established. However, results from in vitro or ex vivo studies point towards some possible mechanistic pathways that can involve disruption of, for example, (developing) neuronal cells by effects on either calcium or thyroid hormone homeostasis. The in vitro studies provide useful information with respect to the potential SARs for the in vivo situation. A SAR study with trichlorinated, tetrachlorinated and some pentachlorinated NDL-PCB congeners identified ortho-substituted congeners as the most potent for disturbing calcium homeostasis, whereas para substitution

WHO Food Additives Series No. 71-S1, 2016 56

Supplement 1: Non-dioxin-like polychlorinated biphenyls

was associated with lower activity. However, the more highly chlorinated NDLPCBs showed no or only subtle effects on calcium homeostasis (Langeveld, Meijer & Westerink, 2012). Another SAR study on the effects of NDL-PCBs on calcium sequestration within brain microsomes and mitochondria identified similar relationships, with congeners having chlorine substitutions at the ortho or ortho-lateral (meta, para) positions again being the most potent (Kodavanti et al., 1996). Both studies are also in line with earlier SAR results obtained for the effects of NDL-PCBs on dopamine content in PC12 cells or activation of RyR1 (Shain, Bush & Seegal, 1991; Pessah et al., 2006). Moreover, metabolic hydroxylation of PCBs appears to yield structures with higher activity towards this RyR (Pessah et al., 2006). With respect to risk assessment of NDL-PCBs, it is interesting to note that a preliminary “neurotoxic equivalency scheme” for PCBs was developed based on in vitro neurotoxicity SAR data, including phorbol ester binding, dopamine release, inhibition of calcium uptake and RyR binding (Simon, Britt & James, 2007). It would be very useful to include in this concept newly available (structure–activity) data on neurotoxicity end-points. For example, a recent SAR study on dopamine transporter interactions of NDL-PCBs showed that most NDL-PCBs (including all NDL-PCBs highlighted in this review) interact with the dopamine transporter with different potencies (Wigestrand et al., 2013). Thyroid hormones, including T3 and T4, are also essential for the development of the prenatal and postnatal nervous system (Murk et al., 2013), and disturbance of thyroid hormone homeostasis in early life stages may affect the central nervous system and express itself, for example, as lower cognitive function (Schroeder & Privalsky, 2014). Thyroid hormones bind to thyroid hormone receptors and regulate, among other things, neurogenesis (Preau et al., 2015) and neuronal differentiation (Ibhazehiebo et al., 2011). Thyroid hormones with ortho iodine atoms clearly bear a structural resemblance to NDL-PCBs, whereas the hydroxy group at the para position of thyroid hormones resembles some of the more common 4-hydroxylated metabolites of PCBs and PBDEs. This structural resemblance is a mechanistic reason for interference of NDL-PCBs or their hydroxylated metabolites with the functioning of thyroid hormones. A wide range of in vivo studies has shown effects of NDL-PCBs on thyroid hormone homeostasis. Although results are not unequivocal, most studies showed a decrease in total and free T4 concentrations following exposure to NDL-PCBs (Craft, DeVito & Crofton, 2002; Khan et al., 2002; Kato et al., 2004, 2007, 2010, 2012). In addition, hydroxylated metabolites of NDL-PCBs can also competitively inhibit the binding of T4 to the transport proteins transthyretin and thyroxinebinding globulin, which offers another possibility for interference with thyroid hormone homeostasis and subsequent effects on the neuronal system (Meerts et al., 2002, 2004). Again, the position of the hydroxy group plays an important role 57

Safety evaluation of certain food additives and contaminants Eightieth JECFA

here, with hydroxy groups present at the para and ortho positions being more active than those in the meta position (Chauhan, Kodavanti & McKinney, 2000; Meerts et al., 2002). (f) Developmental neurotoxicity studies

Multiple rodent studies have shown that both prenatal and postnatal exposures to NDL-PCBs cause effects on neurobehaviour. Behavioural end-points that were affected by NDL-PCBs include spontaneous (locomotor) activity, habituation capability, spatial learning and anxiety-like behaviour in rodents. Both increases and decreases in spontaneous activity have been observed in studies with developmentally exposed mice and rats. Despite the fact that in vitro studies indicate a role for NDL-PCBs in neurodevelopmental toxicity mechanisms, in vivo studies on the effects of exposure to hydroxylated metabolites of NDLPCBs on neurobehavioural or developmental end-points are very scarce. Several studies consist of, or include, specific investigations on the effects of NDL-PCBs on one or several neurotransmitter systems in vivo, which are discussed below. (i) Behavioural studies

The effects of single doses of each of PCB 28 and PCB 52 (NDL-PCBs) and PCB 118 and PCB 156 (DL-PCBs) given orally on PND 10 were studied in male mice. Mice were given the following doses: PCB 28 – 0.18, 0.36 or 3.6 mg/kg bw; PCB 52 – 0.20, 0.41 or 4.1 mg/kg bw; PCB 118 – 0.23, 0.46 or 4.6 mg/kg bw; and PCB 156 – 0.25, 0.51 or 5.1 mg/kg bw. Spontaneous motor activity (locomotion, rearing and total activity) was measured at the age of 4 months over three 20-minute periods in an automated system resembling home cages. Learning and memory functions at the age of 5 months were assessed by performance in the Morris swim maze and radial arm maze. Acute toxicity and effects on body weight were not observed. A significant dose-related change in motor activity was observed in mice exposed to PCB 28 and PCB 52, showing hypoactivity at the lowest dose levels. At the highest dose levels, the pattern of activity was reversed (activity decreased with each 20-minute recording in control animals, whereas the activity was low in exposed animals compared with controls in the first recording and high in the third recording). No effects were observed on these variables in mice exposed to the DL-PCBs, PCB 118 or PCB 156. Effects on water maze performance were observed in mice exposed to PCB 52 at 4.1 mg/kg bw. No effects were observed with the other NDL-PCBs or at lower dose levels of PCB 52. Similarly, impaired radial arm maze performance was observed in mice exposed to PCB 52 at 4.1 mg/kg bw (no effects of PCB 28 or at lower doses of PCB 52). An increase in binding to nicotinic binding sites was observed, which indicates an increased expression of cholinergic nicotinic receptors in mice exposed to the highest dose

WHO Food Additives Series No. 71-S1, 2016 58

Supplement 1: Non-dioxin-like polychlorinated biphenyls

of PCB 28 (no effects of PCB 52). No effects on muscarinic acetylcholine receptor expression or neurotransmitter levels were observed after exposure to PCB 28 or PCB 52. The authors concluded that postnatal (neonatal) oral exposure to lower chlorinated NDL-PCBs results in persistent neurotoxic effects in the adult animal (Eriksson & Fredriksson, 1996). Effects of lactational exposure to a mixture of the six indicator PCBs (PCBs 28, 52, 101, 138, 153 and 180) on neurobehavioural end-points were assessed in mice. Lactating dams received oral gavage doses of 1, 10 or 100 ng/kg bw per day of the mixture on PNDs 0–21. The six PCBs in each mixture were in the proportion 37%, 32%, 11%, 12%, 6% and 2% for PCBs 153, 138, 180, 101, 52 and 28, respectively, reflecting their relative proportions in marine contaminated fish. Offspring (10 of each sex per group) were assessed between PND 0 and PND 275. No effects on body weight of dams or pups or on maternal behaviour were found. Neonatal female offspring exposed to the PCB mixture at 100 ng/ kg bw per day exhibited significantly longer turning reflexes on PNDs 7 and 9, whereas a reduction in general locomotor activity was seen at 1 and 10 ng/kg bw per day in male offspring only. These effects disappeared with age. Changes in visuomotor integration (water escape pole climbing test) were observed on PND 32 in the males at 1 and 100 ng/kg bw per day. Anxious behaviour was detected at PND 40 (elevated plus maze) and PND 160 (light/dark choice test) in both sexes of offspring. The authors proposed that the effects may be related to the observed overexpression of RyR3 in the cerebellum. No other effects were detected in an applied battery of developmental, behavioural and cognitive tests or gene expression of neurotransmitter receptors. The authors concluded that exposure to this mixture of NDL-PCBs results in sex-specific neurodevelopmental effects (Elnar et al., 2012). Male rat offspring were lactationally exposed to PCB 153. Dams were orally dosed with PCB 153 at 5 mg/kg bw every second day from day 3 to day 13 after delivery. No effects were observed on body weight of the dams or physical development of the pups. Effects of exposure to PCB 153 on operant conditioning were studied using a two-component schedule of reinforcement. Exposed offspring demonstrated increased activity (“burst” of lever presses) and attention deficits. In addition to the behavioural testing, xenobiotic metabolizing enzymes (EROD, aldrin epoxidase, GST) and residues of PCB 153 were also determined in the offspring at PNDs 14, 28 and 112 in brain, stomach and liver. The authors concluded that lactational exposure to PCB 153 results in neurotoxic effects (Holene et al., 1998). The effects of perinatal exposure to PCB 153 on synaptic plasticity in rat offspring from dams given PCB 153 orally from GD 3 to PND 21 at a dose of 1.25, 5 or 20 mg/kg bw per day were studied using long-term potentiation in hippocampal slices taken from the offspring at PND 30. There were no effects 59

Safety evaluation of certain food additives and contaminants Eightieth JECFA

of PCB 153 exposure on body weight of the dams, fertility, birth number or postnatal growth. Reduced long-term potentiation was observed in slices from PCB-exposed rats at all dose levels. The authors concluded that developmental exposure to PCB 153 impairs synaptic plasticity in the hippocampus, which is correlated with learning ability (Hussain et al., 2000). However, it should be noted that chemical analysis revealed contamination of this PCB with very low levels of PCDFs (possibly affecting the outcome of this study). Female rats were exposed during pregnancy and lactation to PCB 153, given orally at 1 mg/kg bw per day from GD 7 to PND 21. Effects of performance in a Y-maze conditional discrimination task were studied at 3 and 7 months of age. Impaired learning ability was observed in 3-month-old rats (both males and females) exposed to PCB 153, but no effect was observed in 7-month-old rats. The function of the glutamate–nitric oxide–cyclic guanosine monophosphate (cGMP) pathway in the brain (assessed in the same rats by brain microdialysis) was impaired in 3-month-old males and females after exposure to PCB 153. No effects on this pathway were observed after 8 months. The authors concluded that perinatal exposure to PCB 153 affects learning ability in young rats, at least through impairment of the glutamate–nitric oxide–cGMP pathway function (Piedrafita et al., 2008a,b). Other mechanisms may also play a role. Effects of prenatal and postnatal exposure to PCB 153 were studied in rat offspring from dams given PCB 153 at 1 or 5 mg/kg bw per day by oral gavage from GD 7 to PND 21. No effects were observed on general appearance, home cage behaviour or body weight of the rats. A wide range of neurobehavioural endpoints was studied in the offspring as adults, including spontaneous locomotor activity (open-field test), spatial short-term memory (radial maze), long-term memory (passive avoidance), sensitivity to pain and vulnerability to stress (hot plate test), efficiency of the sensorimotor response (startle response test) and motor coordination (rotarod). Increased locomotor activity was observed in female offspring at both dose levels of PCB 153. Furthermore, increased habituation of the startle reflex was observed in males and females, but only in the low-dose group. In addition, impairment of motor coordination was observed in males exposed to the high dose of PCB 153 (Gralewicz et al., 2009). Effects of prenatal and postnatal exposure to highly purified PCB 52, PCB 138 or PCB 180 on cognitive function or motor coordination were investigated in rat offspring (3–4 months of age) from dams that were exposed to these PCBs at 1 mg/kg bw per day via feed from GD 7 to PND 21. Cognitive function was assessed as the ability to learn a Y-maze conditional discrimination task. Impaired performance was observed in both male and female rats exposed to PCB 138 and PCB 180. No effects were observed in rats exposed to PCB 52. The authors proposed that this effect on cognition is associated with reduced function of the glutamate–nitric oxide–cGMP pathway and reduced expression

WHO Food Additives Series No. 71-S1, 2016 60

Supplement 1: Non-dioxin-like polychlorinated biphenyls

of the NR1 subunit of N-methyl-D-aspartate (NMDA) receptors. In addition, motor coordination was assessed using the rotarod, which demonstrated that only PCB 52 impaired motor coordination. The authors concluded that this effect is associated with an extracellular increase of the neurotransmitter gamma-aminobutyric acid in the cerebellum. Overall, the authors concluded that exposure to NDL-PCBs during pregnancy and lactation induces long-lasting effects on cognitive function or motor coordination, with different effects for different NDL-PCBs (Boix, Cauli & Felipo, 2010). In a second study with these PCBs (PCB 52, PCB 138 and PCB 180), the effects of prenatal and postnatal exposure on motor activity were studied in adult rat offspring (4 months of age) from dams exposed to each PCB at 1 mg/kg bw per day via feed from GD 7 to PND 21. Motor activity was studied in an open-field activity chamber. No effects were observed after exposure to PCB 52, but exposure to PCB 180 resulted in reduced motor activity in males. PCB 138 reduced motor activity in both males and females. Effects on the extracellular neurotransmitters dopamine (by PCB 180 in males and females) and glutamate (by all three PCBs in males and females) and modulation of these through metabotropic glutamate receptors (by PCB 138 or PCB 180 in males and females) were observed in the nucleus accumbens of the exposed rats. Brain (striatum) and perirenal PCB levels were determined at 4 months of age. The authors concluded that different NDLPCBs have different effects on motor activity and that these effects are also sex dependent (Boix et al., 2011). Effects of prenatal exposure to PCBs on spatial learning and memory were investigated in rats. Dams were exposed via oral gavage to PCB 28, an NDL-PCB, at 8 or 32 mg/kg bw per day, PCB 118, a DL-PCB, at 4 or 16 mg/kg bw per day or PCB 153, an NDL-PCB, at 16 or 64 mg/kg bw per day, from GD 10 to GD 16. Overt toxicity was not observed in the dams. Birth weight was reduced in offspring exposed to PCB 28 (females) and PCB 118 (males and females), but no effects were seen in other dose or PCB groups. Mildly decreased weight gain during nursing was observed in offspring exposed to PCB 28 (high dose only) or PCB 118 (both dose levels). No effects were observed for PCB 153. Working memory and reference memory in 3-month-old rats were investigated using an eight-arm radial maze. In general, no effects of these dose levels on the number of errors were observed, although a smaller latency (to enter a radial arm) was observed only in males at the high dose of PCB 153. Spatial learning was investigated in a T-maze, delayed spatial alternation task. Impaired performance in this learning task was observed in female rats exposed to high doses of PCB 28, PCB 118 and PCB 153, but not in males. The effect pattern in the learning tasks suggests the occurrence of learning or attentional deficits, rather than a memory deficit. The authors concluded that perinatal exposure to NDL-PCBs results in persistent effects on learning and that these effects may be sex specific (Schantz, Moshtaghian & Ness, 1995). 61

Safety evaluation of certain food additives and contaminants Eightieth JECFA

In a study on exposure to individual PCBs, male rat offspring were directly exposed on three occasions at “around” PND 8, PND 14 and PND 18 (exact time of birth was not observed) to PCB 52, PCB 153 or PCB 180 given by oral gavage at 0 or 10 mg/kg bw. Activity level and stimulus control were measured using an operant visual discrimination task (time of testing not stated). PCB exposure did not produce behavioural changes during training when responding was frequently reinforced using a variable-interval 3-second schedule. When correct responses were reinforced on a variable-interval 180-second schedule, animals exposed to PCB 153 or PCB 180 were less active than controls or those exposed to PCB 52. Stimulus control was better in animals exposed to PCB 180 than in controls and in the PCB 52 group. The PCB 153 and PCB 180 groups also had fewer responses with short inter-response times compared with the PCB 52 group. No effects of exposure to PCB 52 were found when compared with controls (Johansen et al., 2011). In a further study by the same group on PCB 153, male and female offspring from two different strains of rat – spontaneously hypertensive rats (SHR/ NCrl), an animal model of attention deficit hyperactivity disorder (ADHD), and Wistar Kyoto (WKY/NHsd) controls – were directly exposed on three occasions at “around” PND 8, PND 14 and PND 18 to PCB 153 given by oral gavage at 0, 1, 3 or 6 mg/kg bw. The rats were tested between PND 37 and PND 64 in the same operant procedure as in the study described above. Exposure to PCB 153 was associated with pronounced and long-lasting behavioural changes in SHR/ NCrl rats; 1 mg/kg bw tended to reduce ADHD-like behaviours and produced opposite behavioural effects compared with 3 and 6 mg/kg bw, especially in the females. In the WKY/NHsd controls and for the three doses tested, PCB 153 exposure produced a few specific behavioural changes only in males. The authors concluded that the data suggest that PCB 153 exposure interacts with strain and sex and also indicate a non-linear dose–response relationship for the behaviours observed (Johansen et al., 2014). The effects of prenatal exposure to PCB 95 on neurobehavioural functions and binding to RyR in different brain regions were studied in rats. Dams were given PCB 95 (8 or 32 mg/kg bw per day) on GDs 10–16 via oral gavage. At these dose levels, reproductive and developmental parameters were not affected. Locomotor activity was evaluated in the open field at 35 and 100 days of age, and hypoactivity was observed, but only at 100 days of age. Spatial learning and memory were assessed using a working memory task in an eight-arm radial maze at 60 days of age and a T-maze task at 140 days of age. Exposure to PCB 95 was associated with improved performance in the radial maze task for both sexes, but no effects were observed in the T-maze task. After the neurobehavioural testing, [3H]ryanodine binding was assayed in homogenates of the cerebral cortex, hippocampus and cerebellum (PND 181). Region-specific changes were

WHO Food Additives Series No. 71-S1, 2016 62

Supplement 1: Non-dioxin-like polychlorinated biphenyls

observed in ryanodine binding, with a decrease in hippocampus, an increase in cerebral cortex and an increase in cerebellum at the lowest dose only. The authors concluded that these changes may be related to the observed neurobehavioural effects of PCB 95 (Schantz et al., 1997). Other mechanisms may also play a role. The effects of prenatal and postnatal exposure to PCB 52 or PCB 180 on the dopamine neurotransmitter system were indirectly studied in rats. The purity of the PCBs was determined to be less than 0.5 ng TEQ/g for PCB 52 and 2.7 ng TEQ/g for PCB 180. The study utilized catalepsy induced by the dopamine receptor blocker, haloperidol. Dams were given PCB 52 or PCB 180 by oral gavage. PCB 52 was given to groups of seven rats per dose at total dose levels between 30 and 3000 mg/kg bw, divided over 10 different administrations between GD 7 and PND 10 (i.e. individual doses of 0, 3, 10, 30, 100 and 300 mg/kg bw). PCB 180 was given at total dose levels between 10 and 3000 mg/kg bw, given as four daily administrations between GD 7 and GD 10 (i.e. individual doses of 0, 2.5, 7.5, 25, 75 and 250 mg/kg bw). These PCBs were extensively charcoal cleaned to remove dioxin-like impurities. Maternal body weight and developmental milestones in the offspring were not affected by PCB 52. Mild reduction in maternal body weight and delayed sexual development in the offspring were observed in the PCB 180–exposed rats at the highest dose. At the age of 80 days, the offspring were transported to the neurobehavioural testing facility and allowed a 4-week adaptation period. Injecting the adult offspring at 180 days of age with haloperidol resulted in mildly increased latencies to movement onset in females after exposure to PCB 52, but no dose-dependent effects were observed in males. In contrast, exposure to PCB 180 resulted in effects in both sexes, showing in particular a reduced latency to movement that was most pronounced in the male offspring. The authors concluded that the observed changes can be related to PCB-induced changes in the dopaminergic system, showing relatively weak effects of PCB 52 in the females, whereas effects of PCB 180 were more dominant in males. Regarding the cause of these sex-dependent differences, the hypothesis was posed that this may be related to (lack of) interaction with estrogenic processes (Lilienthal et al., 2014). In an earlier report on the same study, the effects of prenatal and postnatal exposure to PCB 52 or PCB 180 on sexually dimorphic sweet preference were studied in Sprague-Dawley rats. Sweet preference can be studied by measuring saccharin consumption, and female rats typically consume more saccharin solution than males. In parallel, Long-Evans rats received a daily oral dose of PCB 74 (total 229 mg/kg bw) or PCB 95 (total 248 mg/kg bw) from GD 10 to PND 7 (no dosing at PND 0). No effects on sweet preference were observed for PCB 52 in 100- to 120-day-old male or female offspring. Increased saccharin consumption (interpreted by the authors to be supernormal behaviour) was observed in 100- to 120-day-old female offspring exposed to PCB 180 (no effects in males). 63

Safety evaluation of certain food additives and contaminants Eightieth JECFA

WHO Food Additives Series No. 71-S1, 2016 64

Decreased sweet preference was observed in 80-day-old female offspring exposed to PCB 74 (no effects in males). Increased saccharin consumption was observed in 80-day-old males exposed to PCB 95 (no effects in females); this is the only clear indication of reduction in sexually dimorphic behaviour (feminization of males). The authors concluded that different NDL-PCBs exhibit different effects on sexually dimorphic behaviour (Lilienthal et al., 2013). The effects of prenatal and postnatal exposure to PCB 74 or PCB 95 on the dopamine neurotransmitter system were studied in rats using catalepsy induced by the dopamine receptor blocker, haloperidol. Brainstem auditory evoked potentials (BAEPs) were also studied (see section 2.2.5(f)(iii)). Dams were given the PCBs by oral gavage at dose levels of 12 mg/kg bw per day (PCB 74) or 13 mg/ kg bw per day (PCB 95), from GD 10 to PND 7. In both PCB-treated groups, free T4 concentrations in serum were significantly reduced in male offspring. There was a slight, but statistically significant, reduction in latency to movement onset in female offspring exposed to PCB 74; male offspring exposed to PCB 74 and offspring exposed to PCB 95 were not affected in this test (Lilienthal et al., 2015). The effects of lactational exposure to a hydroxylated metabolite of PCB 106 were studied in rats. Lactating dams were orally exposed to 4-hydroxyPCB 106 at 0.5, 5 or 50 mg/kg bw every second day from PND 3 to PND 13. No effects were observed on body weight of the dams, lactation or physical development of the offspring. Motor activity was assessed at PND 28 in an openfield activity chamber in 5-minute sessions. Circadian locomotor activity was recorded at PNDs 28–31 in standard cages for 72 hours. Brain (striatum) samples were collected at PND 31 to determine dopamine levels and dopamine receptor expression. Exposure to this hydroxy-PCB metabolite resulted in hyperactivity (increased locomotor activity) in males only at 0.5 and 5 mg/kg bw, with no effects seen at 50 mg/kg bw. Again in males only, a spontaneous hyperlocomotion was observed during circadian locomotor activity recordings. Furthermore, effects on dopamine levels and/or expression of dopamine receptors in the brain (striatum) were observed at all dose levels. The authors concluded that postnatal exposure to 4-hydroxy-PCB 106 results in neurobehavioural effects that relate to changes in dopamine levels and receptor expression (Lesmana et al., 2014). Other mechanisms may also play a role. (ii) Neurochemical studies

The effects of prenatal exposure to PCB 47 on dopamine function were investigated in rats. Dams were exposed to 1, 10 or 20 mg/kg bw per day from GD 6 through weaning by incorporation of the PCB congener into cookies. No effects were detected on reproduction or on body weight of the offspring. The concentrations of dopamine and its metabolites were measured in different brain regions (frontal

Supplement 1: Non-dioxin-like polychlorinated biphenyls

cortex, caudate nucleus and substantia nigra) on PND 35, PND 60 and PND 90. Decreased concentrations of dopamine were observed in the frontal cortex and caudate nucleus. Based on these results, the authors posed the hypothesis that these reductions are a consequence of inhibition of dopamine synthesis by PCB 47 during brain development (Seegal, Brosch & Okoniewski, 1997). The effects of prenatal and postnatal exposure to PCB 153 on the expression and affinity of dopamine receptors were investigated in rats. Dams were orally exposed to PCB 153 (at 5 mg/kg bw) every other day from GD 7 to PND 21. No adverse effects were detected in the dams, and this exposure had no influence on reproduction or on development of the offspring. In addition, D1-like and D2-like dopamine receptor densities and affinities in offspring were measured on PND 21 and PND 36 using saturation binding studies. In male offspring, the density of D1 receptors was decreased in the cortex and striatum on PND 21, but this was not detected in females or on PND 36 in both sexes. Density of D2 receptors was increased with reduced affinity in the cortex of male offspring at PND 21 and PND 36. In female offspring, the D2 receptor density was increased at PND 36, whereas D2 receptor affinity was reduced on PND 21 and PND 36. No effects were observed in the striatum in either sex. Chemical analysis revealed measurable and similar levels of PCB 153 in the cortex and striatum, indicating that it is transferred across the blood–brain barrier. The authors concluded that perinatal exposure to PCB 153 affects both D1 and D2 receptor expression and affinity and that some of these effects are specific for sex, age and brain areas (Coccini et al., 2011). In a similar study from the same group, the effects of prenatal exposure to PCB 153 at 20 mg/kg bw per day from GD 10 to GD 16 were investigated in rats. Dams were dosed by oral gavage. Effects on cholinergic muscarinic receptor (MR) density in the cerebral cortex, cerebellum, hippocampus and striatum were investigated at PND 21. Overt toxicity was not observed in dams or offspring, but PCB 153 decreased MR density in the cerebellum, while increasing MR density in the cortex. No effects on MR density were detected in the striatum or hippocampus. These results show that PCB 153 can affect MR expression in different brain regions (Coccini et al., 2006). Coccini et al. (2007) also showed that perinatal exposure to PCB 153 given at 5 mg/kg bw per day, via incorporation in sweet jelly placed underneath the normal chow, from GD 7 to PND 21 decreased MR density in the cerebellum of males and in the cerebral cortex in both sexes at PND 36, but MR affinity was not affected. In the cerebral cortex, a decrease in the MR subtypes ACh M1 and ACh M3 immunopositive neurons was also observed. Again, no overt toxicity was observed in the dams or offspring. It should be noted that some neurochemical effects persisted from PND 21 to PND 36, whereas others were not observed before PND 36, although at this time point the brain levels of PCB 65

Safety evaluation of certain food additives and contaminants Eightieth JECFA

153 had already declined significantly. Based on these observations, the authors concluded that PCB 153 can induce delayed neurotoxicity after prenatal and postnatal exposure (Coccini et al., 2007). The effects of prenatal exposure to PCB 153 on monoamine oxidase B activity and content of dopamine, serotonin, 5-hydroxyindoleacetic acid and homovanillic acid in different brain regions (striatum, hippocampus, cerebellum and cerebral cortex) were investigated in PND 21 offspring. PCB 153 was given at a dose of 20 mg/kg bw to rats via daily oral gavage of dams from GD 10 to GD 16. No effects on monoamine oxidase B activity were detected in females, but reduced monoamine oxidase B activity was detected in the cerebellum of the males. PCB 153 also decreased serotonin levels in the cerebral cortex in both sexes. Additionally, dopamine, 5-hydroxyindoleacetic acid and homovanillic acid contents were reduced in the striatum of exposed males and females. The authors concluded that prenatal exposure to PCB 153 results in sex-specific changes in dopaminergic and serotonergic systems (Castoldi et al., 2006). The effects of prenatal exposure to PCB 153 on brain neurotransmitter levels were investigated in female rats. Groups of 10 dams were given PCB 153 by oral gavage from GD 10 to GD 16 at 0, 16 or 64 mg/kg bw per day. Brain neurotransmitters and metabolites were measured in different brain regions or whole brain in groups of 5–9 female offspring at 1, 3, 6 and 9 weeks of age and after 1 year. At 1–3 weeks of age, brain levels of dopamine, 3,4-dihydroxyphenylacetic acid, homovanillic acid, serotonin and 5-hydroxyindoleacetic acid were increased in the offspring. At 9 weeks of age, dopamine turnover was reduced in forebrain and hindbrain, whereas 5-hydroxyindoleacetic acid levels were increased in all brain areas. At 1 year of age, reductions in the levels of dopamine, 3,4-dihydroxyphenylacetic acid and homovanillic acid could still be observed in the hippocampus, hypothalamus and medulla oblongata. The turnover of serotonin was increased at 1–9 weeks of age, and the turnover of dopaminergic neurons was reduced at 9 weeks and 1 year of age. Decreases in dopamine, 3,4-dihydroxyphenylacetic acid and homovanillic acid levels were also observed in PCB-exposed dams at 15 weeks of age (3 weeks after parturition). In addition, reduced levels of dopamine and related metabolites were observed in the brain of the dams. The authors concluded that prenatal exposure to PCB 153 results in long-term changes in neurotransmitters and metabolites in the brain, which are dependent on developmental stage (Honma et al., 2009). In summary, the results of neurodevelopmental studies with NDL-PCBs in rodents indicate comparable patterns of effects, albeit with congener-specific differences. It should also be noted that such studies have usually involved a limited number of dose levels. This aspect hampers the determination of relative effect potencies and QSARs for the NDL-PCBs. Based on in vivo studies, a distinct mechanism of action for the neurodevelopmental effects of NDL-PCBs cannot be

WHO Food Additives Series No. 71-S1, 2016 66

Supplement 1: Non-dioxin-like polychlorinated biphenyls

established. However, results from in vitro or ex vivo studies using neuronal cells indicate a number of mechanistic pathways that involve disruption of intracellular calcium or thyroid hormone homeostasis; these have been discussed previously in section 2.2.5(e). In the majority of neurodevelopmental studies, total dose levels above 1 mg/kg bw or even much higher were tested. In these studies, the maternal dose levels of NDL-PCBs were at least 1–2 orders of magnitude higher than estimated adult human daily exposure levels (Duarte-Davidson & Jones, 1994). There is one study in which effects were detected at doses relevant to humans (Elnar et al., 2012). Although PCB levels in human milk are known, it is complicated to derive margins of exposure (MOEs) in comparison with the results of available neurodevelopmental toxicological studies, as the exposure of the offspring is commonly not measured. However, the lowest effect concentrations in studies in which mouse pups were individually exposed postnatally to a single oral dose (Eriksson & Fredriksson, 1996) are very comparable to, or even up to 1 order of magnitude lower than, those present in human milk (Fürst, 2006). (iii) Effects on the auditory system

Exposure to Aroclors during the prenatal and preweaning developmental period in rats has been shown to cause delayed development and reduced amplitude of the auditory startle reflex, increased susceptibility to audiogenic seizures as adults and permanent auditory deficits in the low-frequency range attributable to loss of outer hair cells in the cochlear organ of Corti (Overmann et al., 1987; Goldey et al., 1995; Crofton et al., 2000; Powers et al., 2006; Poon et al., 2015). These effects can be attenuated by postnatal administration of T4 (Goldey & Crofton, 1998). These effects may be partially attributable to the DL-PCBs, such as PCB 126, present in such mixtures (Crofton & Rice, 1999). The study of Meerts et al. (2004) showed that whereas developmental exposure to Aroclor 1254 raised auditory thresholds, exposure to 4-hydroxy-PCB 107 did not. However, the studies described below show that individual NDL-PCB congeners can have effects on the development of the auditory system. Female rats were exposed to PCB 95 at 0 or 6 mg/kg bw per day from GD 5 to PND 21 by applying the PCB in corn oil to a cornflake, which was rapidly consumed. Control pups were from a minimum of two separate litters, and PCBexposed pups were sampled from three litters. Postnatally, offspring were divided into three groups; one group was raised in a normal auditory environment, a second group was exposed continuously from PND 9 to PND 35 or from PND 9 to PND 40 to a tone (25-millisecond tone, 5-millisecond ramps) and the third group was exposed continuously from PND 9 to PND 35 or from PND 9 to PND 40 to noise (50-millisecond noise pulses, 5-millisecond ramps) at a sound pressure 67

Safety evaluation of certain food additives and contaminants Eightieth JECFA

of 65–70 dB. Six control or PCB-exposed rats in the normal auditory exposure group, five from each of the tone-reared groups and five from each of the noisereared groups were mapped. Developmental exposure to PCB 95 did not produce any effect on litter size, sex ratio or weight gain compared with the control group. Hearing sensitivity and brainstem auditory responses of the pups were normal. However, there was abnormal development of the primary auditory cortex in PCB-exposed pups, which was irregularly shaped and marked by internal nonresponsive zones. Its topographic organization was grossly abnormal or reversed in about half of the exposed pups, the balance of neuronal inhibition to excitation for A1 neurons was disturbed, and development was significantly altered in pups also exposed to tonal stimuli or noise stimuli from PND 9 to PNDs 35–40, the critical period of plasticity that underlies the normal postnatal auditory system (Zhang, Bao & Merzenich, 2002; Kenet et al., 2007). Auditory function in Sprague-Dawley rats was assessed following developmental exposure to PCB 52 or PCB 180 (Lilienthal et al., 2011). Pregnant rats received repeated oral doses of PCB 52 (total doses of 0, 30, 100, 300, 1000 or 3000 mg/kg bw) or PCB 180 (total doses of 0, 10, 30, 100, 300 or 1000 mg/kg bw). The purity of the PCBs was determined to be less than 0.5 ng TEQ/g for PCB 52 and 2.7 ng TEQ/g for PCB 180. BAEPs were recorded in adult male and female offspring after stimulation with clicks or pure tones in the frequency range from 0.5 to 16 kHz. Significant elevation of BAEP thresholds was detected in the lowfrequency range after developmental exposure to PCB 52. Calculation of BMDs revealed lowest values in the frequency range of 0.5–2 kHz. Effects were more pronounced in male offspring than in female offspring. Latencies of waves II and IV over a range of frequencies were prolonged in exposed males, whereas only wave IV was affected in females. PCB 180 increased BAEP thresholds only at 0.5 and 4 kHz in female offspring, and wave IV latency was prolonged only at 0.5 kHz in female offspring. In a follow-up study from the same laboratory, further experiments were performed with PCB 74 and PCB 95 using Long-Evans rats (Lilienthal et al., 2015). Rat dams were given equimolar doses of either congener (40 μmol/kg bw, i.e. 11.68 mg/kg bw of PCB 74 or 13.06 mg/kg bw of PCB 95) in corn oil by oral gavage from GD 10 to PND 7. Control dams were given vehicle only. Adult offspring were tested for cataleptic behaviour after induction with haloperidol (see section 2.2.5(f)(i)) and BAEPs. Pronounced changes were observed in BAEPs at low frequencies in PCB 74–exposed offspring, with elevated thresholds in both sexes. PCB 95 increased thresholds in males, but not females. Small effects on latency of the late wave IV were detected in both sexes after developmental exposure to PCB 74 or PCB 95. In summary, the results confirm that developmental exposure to individual NDL-PCB congeners can affect auditory function and that different

WHO Food Additives Series No. 71-S1, 2016 68

Supplement 1: Non-dioxin-like polychlorinated biphenyls

congeners exhibit different potencies. PCB 74 was the most potent congener of the NDL-PCBs tested in terms of inducing threshold increases. The effects of PCB 95 and PCB 52 were similar but were less than those of PCB 74. In contrast, increases by PCB 180 were smaller. Effects of PCB 74 and PCB 52 were more expressed in male offspring than in female offspring, and PCB 95 elevated thresholds only in males. The modest effects of PCB 180 on BAEP thresholds were found only in females. As all congeners resulted in similar reductions in circulating thyroid hormone levels, other factors involved in the development of cochlear and neural structures of the auditory system are likely to contribute to the observed effects; the authors of the studies discussed the potential role of retinoids or the RyR (Lilienthal et al., 2011, 2015). 2.2.6 Special studies (a) Adult neurotoxicity

This section covers only those studies in which PCB treatments have been given to adult animals; studies in which animals have been exposed during gestation, the perinatal period or postnatally as juveniles are discussed in section 2.2.5(f) above. Early studies on neurobehavioural effects of PCBs in adult animals exposed for various durations to commercial mixtures, defined experimental mixtures or single congeners have been reviewed by ATSDR (2000). Neurochemical effects of PCBs have also been investigated in rats, mice and monkeys exposed to commercial PCB mixtures or to individual PCB congeners. Some studies have assessed both neurochemical and neurobehavioural effects of PCBs in an attempt to link a biochemical alteration to a particular neurobehavioural deficit. The ATSDR (2000) review considers the studies in terms of effects on motor activity and effects on higher functions, such as learning and memory. In these early studies, effects on higher functions were reported only for exposures occurring during the prenatal, perinatal or postnatal juvenile period. Later in vitro studies illustrate the potential for effects of NDL-PCBs on functions such as learning and memory in adult animals. Single or repeated administration of relatively high doses of Aroclor 1254 to adult mice or rats generally decreased spontaneous motor activity (ATDSR, 2000). This may be attributable to reductions in brain dopamine levels in adult animals (Seegal, Bush & Brosch, 1991a). The uptake and release of dopamine or other neurotransmitters are dependent on, among other things, the maintenance of intracellular calcium homeostasis, and this has been investigated in a number of SAR studies by the research group of P.R. Kodavanti and H.A. Tilson. They used rat cerebellar granule cells cultured in vitro or microsomal or mitochondrial organelles isolated from brain tissue and treated with various individual PCB 69

Safety evaluation of certain food additives and contaminants Eightieth JECFA

congeners or Aroclor mixtures. Perturbed signal transduction mechanisms involving alterations in several aspects of cellular calcium homeostasis were observed. This has been shown to have consequences for inositol phosphate signalling by inhibiting agonist-stimulated inositol phosphate accumulation. The research group also observed perturbations in protein kinase C (PKC) translocation, which has been confirmed in vivo in rats treated with an Aroclor 1254 mixture of PCBs. PKC signalling plays a significant role in motor behaviour, learning and memory. The effects in vitro on calcium homeostasis and PKC translocation were seen at relatively low concentrations (5–50 µmol/L), whereas higher concentrations (>200 µmol/L) were required to produce cytotoxicity. The SARs for these perturbations of signal transduction and second messenger systems for the 24 PCB congeners tested were consistent with a chlorination pattern that favoured non-coplanarity (NDL-PCBs), whereas the PCB congeners with a chlorination pattern that favoured coplanarity (DL-PCBs) were less active. The studies indicated that the effects of most PCB congeners in vitro may be related to an interaction at specific sites having preference for low lateral substitution or lateral content (meta or para) in the presence of ortho substitution (Kodavanti et al., 1993, 1995, 1996, 1998a; Shafer et al., 1996; Kodavanti & Tilson, 1997; Tilson & Kodavanti, 1998; Tilson et al., 1998). Other mechanisms that may be responsible for reductions in brain dopamine levels have been investigated. In vitro studies have shown that reductions in brain dopamine levels may be related to inhibition of tyrosine hydroxylase activity, the rate-limiting enzyme for catecholamine synthesis in the brain, although this was not seen in the rat in vivo (Choksi et al., 1997). A later study investigated whether dopamine reductions may involve inhibition of the dopamine transporter (DAT) and/or the vesicular monoamine transporter (VMAT), which are responsible for the uptake of extracellular dopamine and the packaging of nerve terminal cytosolic dopamine into synaptic vesicles, respectively. The results suggested that elevations in 3,4-dihydroxyphenylacetic acid, reflective of increases in nerve terminal cytosolic dopamine due to VMAT inhibition, rather than elevations in media dopamine due to DAT inhibition, were largely responsible for the observed decreases in tissue dopamine content (Bemis & Seegal, 2004). To investigate whether the chirality of PCB congeners implicated in neurotoxic effects is important, the effects of racemic PCB 84 and two of its enantiomers on PKC translocation in cerebellar granule cells and calcium sequestration in microsomes isolated from adult rat cerebellum were studied. Both (+)- and (−)-PCB 84 enantiomers affected PKC translocation in a concentration-dependent manner, with (−)-PCB 84 being slightly more potent; racemic PCB 84 was significantly more potent and efficacious than either of the pure enantiomers alone. Microsomal calcium uptake was inhibited by both (−)-

WHO Food Additives Series No. 71-S1, 2016 70

Supplement 1: Non-dioxin-like polychlorinated biphenyls

and (+)-PCB 84 enantiomers to a similar extent, whereas racemic PCB 84 was more potent (Lehmler et al., 2005). The same research group has investigated nitric oxide synthases (NOS), which play a key role in motor activity in the cerebellum, hormonal regulation in the hypothalamus and long-term potentiation, learning and memory processes in the hippocampus. In in vitro studies on tissue taken from the cerebellum, hippocampus and hypothalamus of rats aged 90–120 days, two specific dichloroPCB congeners, some pentachloro- and hexachloro-PCB congeners, and some hydroxy metabolites of tetrachloro-, pentachloro- and hexachloro-PCBs were tested. Only dichloro-ortho-PCB (PCB 4) inhibited both neuronal and membrane NOS, whereas the non-ortho, para-substituted PCB 15 and pentachloro- or hexachloro-PCB congeners did not. Hydroxy substitution of one or more chlorine molecules significantly increased the potency of both ortho- and non-orthohexachlorobiphenyls. The authors concluded that selective sensitivity of NOS to dichloro-ortho-PCB and hydroxy metabolites suggests that the inhibition of NOS could play a role in the neuroendocrine effects as well as learning and memory deficits caused by exposure to PCBs (Sharma & Kodavanti, 2002). Investigation of the distribution of individual PCB congeners following once daily gavage treatment of adult rats with Aroclor 1254 at 0 or 30 mg/kg bw per day, 5 days/week, for 4 weeks showed that in all the tissues, the lower chlorinated (tetra- and penta-) congeners accumulated less than their respective proportions in the parent Aroclor 1254 mixture. Higher chlorinated (hexa- to nona-) congeners accumulated more than the proportion of these congeners found in the Aroclor 1254 mixture. This shift towards accumulation of higher chlorinated congeners was more pronounced in the brain than in liver and fat. Predominant congeners (5–32% of total PCBs) detected in different brain regions, blood, liver and fat were as follows: PCB 163 + PCB 138 (coeluted), PCB 153 + PCB 132 (coeluted), PCB 156 + PCB 171 (coeluted), PCB 118, PCB 99 and PCB 105. These congeners together accounted for about two thirds of the total PCB load in the brain. Of these, all but PCBs 156, 118 and 105 are NDL-PCBs. The total PCB concentrations accumulated in the brain were as high as 50 µmol/L (based on average relative molecular weight of 326.4 for Aroclor 1254), and it is at these concentrations that intracellular second messengers were significantly affected in neuronal cultures and brain homogenate preparations in vitro. These results indicated that concentrations that altered calcium disposition and second messenger systems in vitro are achievable in brain in vivo following repeated exposure (Kodavanti et al., 1998b). The effects of a commercial mixture (Aroclor 1254), a DL-PCB (PCB 126) and an NDL-PCB (PCB 99) on the expression of NMDA receptors and the subsequent toxic effects have been studied in vitro using a human SHS5-SY neuroblastoma cell line. NMDA receptors are ionotropic receptors gated by the 71

Safety evaluation of certain food additives and contaminants Eightieth JECFA

neurotransmitter glutamate, which allow calcium flux into the cell, and they play an important role in the physiology and pathophysiology of the central nervous system (Waxman & Lynch, 2005). All three PCB treatments increased caspase-3, which plays a central role in cell apoptosis, and induced apoptosis and cell death in a dose-related manner at concentrations of 10–50 µmol/L. The mechanisms involved in cell death were mainly mediated through the NMDA receptors. The authors speculated that this may be induced by a rapid increase in intracellular calcium concentrations, followed by a series of events eventually leading to apoptosis and necrosis. NMDA receptor antagonists and an intracellular calcium chelator gave partial protection to cells against the effects of the PCBs, indicating that there are likely to be other parallel mechanisms leading to cell death. In this study, the NDL-PCB, PCB 99, was found to be more neurotoxic than the DL-PCB or the PCB mixture (Ndountse & Chan, 2009). Early neurochemical studies showed the potential for commercial PCB mixtures to selectively alter dopamine, noradrenaline and serotonin concentrations in some regions of the adult rat brain following a single high dose given by gavage and in primate brain following 20 weeks of exposure. The concentrations of specific congeners in the affected brain regions indicated that it was non-coplanar NDL-PCBs that may be responsible (Seegal, Bush & Brosch, 1985, 1991a,b; Seegal, Brosch & Bush, 1986). The same group investigated SARs for the effects of individual PCB congeners on dopamine in vitro in PC12 cells, a clonal cell line derived from a phaeochromocytoma of the rat adrenal medulla, which, when cultured in the presence of nerve growth factor or other compounds, differentiate to resemble sympathetic neurons morphologically and functionally. The study showed that (1) congeners with ortho- or ortho,para-chlorine substitutions were most potent; (2) chlorination in a meta position decreased cell dopamine content in ortho-substituted congeners, but had little effect in ortho,para-substituted congeners; and (3) increasing congener chlorination did not correlate with a decrease in potency, although total chlorination of a ring appeared to reduce potency. An experiment with PCB 4 indicated that it was the congener and not its metabolites that was the toxicant. Thus, PCB congeners decrease cell dopamine content by interaction at specific sites that have preference for ortho- or ortho,para-substituted congeners (Shain, Bush & Seegal, 1991). In in vivo studies with individual NDL-PCBs, a reduction in dopamine concentration in the substantia nigra region was observed in female but not male rats after 13 weeks of treatment with PCB 28 at dietary exposure levels of 0.5, 5 and 50 mg/kg feed, giving a lowest-observed-adverse-effect level (LOAEL) of 36 µg/kg bw per day (Chu et al., 1996a). Similarly, reductions in dopamine and serotonin concentrations in the frontal cortex of the rat brain, mainly in females, were observed after treatment for 13 weeks with PCB 153 in the diet at 5 and 50

WHO Food Additives Series No. 71-S1, 2016 72

Supplement 1: Non-dioxin-like polychlorinated biphenyls

mg/kg feed, but not at 0.5 mg/kg feed, giving a NOAEL of 34 µg/kg bw per day (Chu et al., 1996b). The effect of oral exposure to a commercial mixture, Aroclor 1254, on central and systemic vasopressin release following the stimulus of acute dehydration in the rat has been investigated by Coburn and co-workers (Coburn, Gillard & Currás-Collazo, 2005; Coburn, Currás-Collazo & Kodavanti, 2007). Vasopressin has multiple functions, including maintenance of body fluid homeostasis, cardiovascular control, learning and memory, and nervous system development. Central vasopressin release from magnocellular neuroendocrine cells (MNCs) in the supraoptic nucleus (SON) of the hypothalamus occurs within several hours after acute dehydration and is an important autoregulatory mechanism. Male rats were fed daily for 15 days with a cheese puff injected with corn oil vehicle or Aroclor 1254 to give an exposure of 0 or 30 mg/kg bw per day. On the 15th day, acute dehydration was produced by intraperitoneal injection of sodium chloride in half the animals, whereas the other half received physiological saline as normosmotic controls. Water was withheld until sacrifice 4.5–6 hours later. Intranuclear vasopressin release from SON tissue in vitro and systemic vasopressin release were measured. The SON from dehydrated rats not receiving PCBs released significantly more vasopressin than did the SON from normosmotic control rats. In contrast, whereas PCB exposure had no effect on baseline water intake, weight gain or plasma osmolality responses to dehydration, the SON did not respond with increased vasopressin release during dehydration. Dehydrated PCB-fed rats showed a significantly higher increase in plasma vasopressin. The study indicated subtle disruption of the MNC system (Coburn, Gillard & CurrásCollazo, 2005). In subsequent work on the release of vasopressin from SON tissue in vitro in response to specific PCB congeners, it was shown that PCB 47 (an NDL-PCB) but not PCB 77 (a DL-PCB) reduced vasopressin release (Coburn, Currás-Collazo & Kodavanti, 2007). More recently, Coburn et al. (2015) studied NOS activity in the SON of hyperosmotic rats as a potential target of PCBinduced disruption of neuroendocrine processes necessary for osmoregulation. Vasopressin responses to hyperosmotic stimulation are regulated by nitric oxide signalling. Male rats were exposed to Aroclor 1254 (30 mg/kg bw per day) in utero, and NADPH-diaphorase (also known as NOS) activity was assessed, under normosmotic and hyperosmotic conditions, in SON sections at three ages: PND 10, early adult (3–5 months) and late adult (14–16 months). The study showed that developmental but not adult exposure to PCBs significantly reduced NOS responses to hyperosmolality in neuroendocrine cells, compared with controls. The reduced NOS activity produced by in utero exposure persisted in stimulated late adult rats concomitant with reduced osmoregulatory capacity. Rats receiving PCB exposure as early adults orally for 14 days displayed normal responses. These findings suggested that developmental exposure to PCBs permanently 73

Safety evaluation of certain food additives and contaminants Eightieth JECFA

compromises NOS signalling in the activated neuroendocrine hypothalamus, with potential osmoregulatory consequences (Coburn et al., 2015). A 28-day study in which rats, aged 6 weeks at the start of treatment, were given loading and maintenance doses of PCB 180 (total doses of 0–1700 mg/ kg bw) by oral gavage is described in section 2.2.2(b). It showed that the most sensitive end-point was altered open-field behaviour in females (Viluksela et al., 2014). (b) Immunological effects (i) Mice

C57BL/6J mice (3–4 weeks old, minimum four animals per dose group) were treated with a single intraperitoneal injection of corn oil alone, TCDD alone (0.0037 µmol/kg bw), PCB 153 alone (100, 400 or 1000 µmol/kg bw) or TCDD plus PCB 153 (0.0037 µmol/kg bw of TCDD plus 100, 400 or 1000 µmol/kg bw of PCB 153) (Biegel et al., 1989). TCDD and PCB 153 were synthesized to greater than 99% purity, as determined by gas chromatography (GC). Five days after treatment, each mouse was injected with sheep red blood cells (SRBCs; 4 × 108 cells), and 5 days later, splenic plaque-forming cell (PFC) responses to SRBCs were assessed. TCDD alone significantly inhibited PFC responses by 75% relative to controls, whereas PCB 153 alone had no significant effect at any dose. Co-administration of TCDD and PCB 153 (100 µmol/kg bw) had the same effect on PFC responses as for TCDD alone. PCB 153 (400 or 1000 µmol/kg bw) in combination with TCDD antagonized the inhibitory effects of TCDD on PFCs. Hepatic EROD was induced by TCDD alone, but not PCB 153. In combination with TCDD, PCB 153 (400 or 1000 µmol/kg bw) partially antagonized EROD induction relative to TCDD alone. Radiolabelled PCB 153 partially displaced TCDD from AhR. The results indicate that PCB 153 antagonizes AhR-mediated TCDD-induced inhibition of T cell–dependent PFC responses without substantively interacting with AhR. Female B6C3F1 mice (8 weeks old, eight mice per group) were administered a single oral gavage dose of corn oil, PCB 153 (3.58, 35.8 or 358 mg/kg bw), TCDD (0.1, 1.0 or 10 µg/kg bw) or all dose combinations of PCB 153 and TCDD (Smialowicz et al., 1997). TCDD and PCB 153 were more than 98% pure. TCDD alone (1.0 or 10 µg/kg bw) significantly reduced spleen and thymus weights and significantly increased liver weight; PCB 153 alone had no comparable effects at any dose, with the exception of the high-dose group, in which an increase in liver weight was observed. Neither TCDD nor PCB 153 had a significant effect on body weight. When co-administered, PCB 153 and TCDD (10 µg/kg bw) significantly reduced spleen and thymus weights, irrespective of PCB 153 dose. Liver weights were elevated by TCDD (10 µg/kg bw), irrespective

WHO Food Additives Series No. 71-S1, 2016 74

Supplement 1: Non-dioxin-like polychlorinated biphenyls

of PCB 153 dose, and by the high dose of PCB 153, irrespective of TCDD dose. To assess PFC responses, mice were immunized 7 days after TCDD and/or PCB 153 exposure with a single intravenous injection of SRBCs (0.2 mL, 2 × 108 cells). Splenocyte primary PFC responses were assessed 4 days after immunization. TCDD (1.0 or 10 µg/kg bw) significantly suppressed PFC responses, whereas PCB 153 (358 mg/kg bw) significantly enhanced PFC responses. When coadministered with TCDD, PCB 153 (358 mg/kg bw) antagonized suppression of PFC responses by TCDD (0.1 or 1.0 µg/kg bw), whereas PCB 153 (358 mg/kg bw) and TCDD (10 µg/kg bw) suppressed splenocyte PFC responses. Taken together, the results indicate that PCB 153 acted as a functional antagonist by inducing a counterbalancing effect on immune responses in the opposite direction to that of TCDD and not via competition for AhR. Male C57BL/6 and DBA/2 mice (6–8 weeks old, five mice per group) were treated with a single intraperitoneal dose of corn oil or one of the following ortho-substituted NDL-PCBs: PCB 206, PCB 207, PCB 208 or PCB 209 (10, 20 or 100 µmol/kg bw in C57BL/6 mice; 25, 100 or 400 µmol/kg bw in DBA/2 mice) (Harper et al., 1993). PCB purity was greater than 99%, as determined by gas– liquid chromatography. Four days after treatment, mice were immunized with SRBCs (4 × 108 cells in 200 µL) or trinitrophenyl-lipopolysaccharide (50 µg in 200 µL). Splenocyte PFC responses were assessed 4 days after immunization. Inhibition of SRBC PFC responses was significant in mice exposed to PCB 207 and PCB 208 (all doses; both strains), PCB 206 (20 and 100 µmol/kg bw, C57BL/6 mice; 100 and 400 µmol/kg bw, DBA/2 mice) and PCB 209 (100 µmol/kg bw, C57BL/6 mice; 400 µmol/kg bw, DBA/2 mice). In C57BL/6 mice exposed to PCBs at 100 µmol/kg bw, PFC responses were inhibited by an average of 72– 81% relative to controls. Inhibition of SRBC PFC responses was less pronounced in DBA/2 mice. Minimal to no PFC response inhibition was observed in mice exposed to the T cell–independent antigen trinitrophenyl-lipopolysaccharide. Significant hepatic EROD induction was observed in C57BL/6 mice exposed to PCB 206 (25 and 100 µmol/kg bw). No significant induction of hepatic EROD was observed in C57BL/6 mice for any other PCB or in DBA/2 mice, suggesting an AhR-independent mechanism of action for inhibition of PFC response for higher chlorinated NDL-PCB congeners. A single gavage dose of PCB 104 or PCB 153 (150 µmol/kg bw) in male C57BL/6 mice increased mRNA expression for the proinflammatory mediators ICAM-1, VCAM-1 and MCP-1 in liver, lungs and/or brain (Sipka et al., 2008). A 28-day repeated-dose toxicity study compared the effects of polyhalogenated seafood contaminants on female BALB/c mice (Maranghi et al., 2013). Groups of 10 mice were fed diets containing one of the following: PCB 153, 1.3 µg/g; TCDD, 0.6 ng/g; hexabromocyclododecane [HBCD], 1.3 mg/g; or 2,2′,4,4′-tetrabromodiphenyl ether (BDE-47), 3.0 µg/g. Mice were 22 days old at 75

Safety evaluation of certain food additives and contaminants Eightieth JECFA

the beginning of the study. Test substances (purity not indicated) were dissolved in 100% dimethylsulfoxide (DMSO; final concentration 0.4 mL/kg feed) and added to AIN-93G rodent diet that also contained freeze-dried Atlantic salmon. Control mice (n = 15) received diet also containing freeze-dried salmon; the presence of DMSO was not specified. Feed intake was restricted during the study, beginning with 2.25 g feed per day and increasing biweekly with body weight. This allowed for reasonably accurate achievement of the following exposure levels based on feed intake: PCB 153, 195 µg/kg bw per day; TCDD, 90 ng/kg bw per day; HBCD, 199 mg/kg bw per day; and BDE-47, 450 µg/kg bw per day. No significant changes in spleen, thymus or liver weight were observed in mice exposed to PCB 153. Histopathological evidence of immune effects in mice due to PCB 153 exposure included inflammatory infiltration in liver and spleen and thymic changes suggestive of accelerated involution. Increased hepatocyte vacuolation, pyknotic nuclei and periportal lymphocytic infiltration were observed in livers from mice exposed to PCB 153 compared with controls. (ii) Rats

Female SD rats (starting body weight 150 g; nine rats per group) were fed control diet or diets containing PCB 153 (10, 30 or 100 mg/kg diet), TCDD (0.5 or 5 µg/ kg diet) or combinations of PCB 153 and TCDD for 13 weeks (van der Kolk et al., 1992). No PCDDs or PCDFs were detected in PCB 153 using high-resolution gas chromatography with mass spectrometry (HRGC-MS; LOD 0.5 ng TEQ/g). There was no significant effect of PCB 153 alone on body weight at any dose; TCDD (5 µg/kg diet) significantly reduced body weight gain. PCB 153 alone (100 mg/kg diet) and TCDD alone (5 µg/kg diet) significantly increased relative liver weight. Whereas TCDD (5 µg/kg diet) significantly reduced thymus weight, no significant changes in thymus weight due to PCB 153 were observed at any dose. When PCB 153 and TCDD were co-administered, no interactive effects on thymus were evident, although changes in liver weight were additive. PCB 153 alone induced PROD but not EROD activity, whereas TCDD alone induced both PROD and EROD activities. Together, TCDD antagonized PCB 153–induced PROD activity. No interactive effects of PCB 153 on TCDD-induced EROD activity were observed. Thus, exposure to TCDD but not PCB 153 led to thymic atrophy associated with AhR induction. Induction of PROD activity was not associated with thymic atrophy in rats exposed to PCB 153. Exposure to PCB 153 in the 2-year NTP chronic toxicity study caused inflammation in female rats (NTP, 2006a). Inflammatory lesions in female reproductive organs were observed in Harlan Sprague-Dawley rats exposed to PCB 153 by gavage at 10, 100, 300, 1000 or 3000 µg/kg bw per day, 5 days/week, for 14, 31 or 53 weeks or 2 years (NTP, 2006a; Strauss & Heiger-Bernays, 2012).

WHO Food Additives Series No. 71-S1, 2016 76

Supplement 1: Non-dioxin-like polychlorinated biphenyls

Dose-dependent increases in the incidence of chronic active inflammation were observed in the ovary and oviduct of female rats in the 1000 and 3000 μg/kg bw per day dose groups. The incidence of suppurative uterine inflammation and chronic active uterine inflammation increased relative to controls in the 1000 and 3000 μg/kg bw per day dose groups, respectively (NTP, 2006a). Bone marrow hyperplasia, also indicative of chronic inflammation, was seen in female rats in the 3000 µg/kg bw per day dose group (NTP, 2006a; Strauss & Heiger-Bernays, 2012). (iii) Non-rodent species

The effects of perinatal exposure to PCB 153, an NDL-PCB, or PCB 126, a DL-PCB, on females and their offspring were examined in goats (Lyche et al., 2004b, 2006). Adult female goats (Norwegian breed) were exposed to PCB 153 or PCB 126 at an estimated dose of 98 µg/kg bw per day or 49 ng/kg bw per day, respectively, for 3 days/week from GD 60 to parturition (approximately 90 days, assuming a 150day gestation period). Their offspring were exposed indirectly during gestation and lactation. Significant changes in immune parameters were observed in does and kids exposed to PCB 153. Two weeks after parturition, increased white blood cell, neutrophil and lymphocyte numbers were detected in blood from kids exposed perinatally to PCB 153. Decreased blood lymphocyte proliferation stimulated by the mitogens concanavalin A and phytohaemagglutinin were also observed. There were no effects on these parameters in PCB-exposed kids at 4 or 8 weeks after parturition (Lyche et al., 2004b). PCB 153 significantly affected maternal humoral responses to vaccination, resulting in reduced transfer of specific antibodies to kids. Perinatal PCB 153 exposure also disrupted antibody responses to vaccination in kids (Lyche et al., 2006). In kids exposed perinatally to PCB 126, blood monocyte numbers were significantly lower at 2, 4 and 8 weeks after parturition, but no further effects were observed (Lyche et al., 2004b). Maternal and juvenile responses to vaccination were also disrupted by PCB 126 (Lyche et al., 2006), but the pattern of disruption differed from that of PCB 153, suggesting that the immunomodulatory effects of PCBs are due to AhR- and non-AhR-mediated events. (v) In vitro studies

Non-coplanar, ortho-substituted PCBs (NDL-PCBs) inhibit mitogen-stimulated murine splenocyte (mixed spleen cell) proliferation more effectively than coplanar PCBs (DL-PCBs) in vitro (Stack et al., 1999; Smithwick et al., 2003; Mori et al., 2006, 2008). Proliferation stimulated by lipopolysaccharides and concanavalin A was inhibited in mouse splenocytes by NDL-PCBs (Smithwick et al., 2003; Mori et al., 2006, 2008), suggesting that pathways common to multiple 77

Safety evaluation of certain food additives and contaminants Eightieth JECFA

immune cell types were affected. Anti-proliferative effects of NDL-PCBs were AhR independent (Stack et al., 1999; Smithwick et al., 2003). Thymocyte viability was reduced by ex vivo exposure to the orthosubstituted, non-coplanar NDL-PCB, PCB 52, at micromole per litre concentrations, but not by the coplanar DL-PCB, PCB 77 (Yilmaz et al., 2006). Thymocyte death was associated with disrupted calcium homeostasis and increased membrane fluidity, indicating that ortho-substituted PCBs disrupt cellular membranes. PCB 52 and PCB 77, at micromole per litre concentrations, significantly stimulated interferon gamma and inhibited interleukin-10 (IL10) production in concanavalin A–stimulated murine thymocytes; PCB 52 was approximately 10-fold more potent (Sandal et al., 2005). However, exposure to PCB 153, an NDL-PCB, did not alter the percentages of human lymphocytes producing interferon gamma or IL-4 in culture (Gaspar-Ramírez et al., 2012). Lymphocyte proliferation was significantly modulated in vitro in marine mammal lymphocyte cultures (Mori et al., 2006, 2008). Concanavalin A– and lipopolysaccharide-stimulated mouse lymphocytes were consistently inhibited by TCDD, PCB 169 (a DL-PCB) and PCB 138, PCB 153 and PCB 180 (NDLPCBs); however, B-cell proliferation and T-cell proliferation were unchanged or stimulated in lymphocyte cultures from most of the marine mammals tested. Non-coplanar NDL-PCB congeners with low affinity for AhR have more pronounced effects in vitro on human and rodent granulocytes than do coplanar DL-PCBs, including activation of quiescent neutrophils and enhancement or inhibition of activated neutrophils (Ganey et al., 1993; Brown & Ganey, 1995; Olivero-Verbel & Ganey, 1998; Voie, Wiik & Fonnum, 1998; Bezdecny, Roth & Ganey, 2005). However, non-coplanarity alone does not fully account for the effects of PCBs on granulocyte functions. For example, chlorine substitutions in the ortho or meta position are required to stimulate superoxide production in neutrophils in vitro (Brown et al., 1998). Numbers of substitutions at the ortho position, congener size and absolute hardness have also been correlated with changes in respiratory burst activity in human granulocytes in vitro (Voie et al., 2000). In marine mammals, the effects of TCDD, PCB 169 (a DL-PCB) and PCB 138, PCB 153 and PCB 180 (NDL-PCBs) on leukocyte function were congener and species specific (Levin et al., 2004, 2005; Levin, Morsey & DeGuise, 2006). Bottlenose dolphin (Tursiops truncatus) monocytes and neutrophils were more sensitive to PCBs than monocytes and neutrophils from beluga whales (Delphinapterus leucas). In dolphins and belugas, PCB 169 (a DL-PCB) and TCDD alone were not inhibitory, whereas the individual NDL-PCBs, PCB 138, PCB 153 and PCB 180, and mixtures containing these congeners were inhibitory (Levin et al., 2004). In a larger cross-section of marine mammals, modulation of leukocyte phagocytosis by the same group of test substances was dependent on the presence of at least one NDL-PCB in all species but one (harbour seal, Phoca

WHO Food Additives Series No. 71-S1, 2016 78

Supplement 1: Non-dioxin-like polychlorinated biphenyls

vitulina); mouse leukocytes were unaffected (Levin et al., 2005). Both NDL- and DL-PCBs modulated respiratory burst in leukocytes from marine mammals, mice and humans (Levin, Morsey & DeGuise, 2006). The results highlight speciesspecific differences in granulocyte responses to PCBs in vitro. 2.3 Observations in humans 2.3.1 Biomonitoring (a) Biomarkers for NDL-PCBs

The most commonly used biomarkers of PCB exposure in humans are PCB concentrations in adipose tissue, serum, plasma and milk. The presence of PCBs in human tissues and fluids may reflect exposure from one or more sources, including air, water, food, soil and dust, with food being the major contributor. Exposures from all sources are nearly always to mixtures of PCBs rather than to individual congeners. There is a strong correlation between concentrations of PCBs in serum and adipose tissue, when expressed on a lipid basis. For example, in non-occupationally exposed subjects, a strong correlation was found between concentrations of nine NDL-PCBs (PCBs 74, 99, 138, 146, 153, 170, 180, 183 and 187) in serum and adipose tissue (Stellman et al., 1998). Similarly, a strong correlation was also found between concentrations of PCB 153 and PCB 180 in serum lipid and breast or gluteal adipose tissue in another study in women undergoing surgery for breast cancer (Rusiecki et al., 2005). Concentrations of PCBs in both serum and adipose tissue are widely regarded as useful biomarkers of PCB body burden. The concentrations of PCBs in serum or plasma can be significantly influenced by serum lipid content, owing to partitioning of PCBs between adipose tissue and serum lipids. Thus, PCB concentrations in serum lipid or plasma lipid are regarded as better indicators of body burden than PCB concentrations that have not been corrected for lipid content (Brown & Lawton, 1984). In general, concentrations in blood lipids reflect more recent exposures, as well as the full spectrum of PCB congeners to which a person is exposed, whereas the pattern of PCB congeners in adipose tissue reflects long-term exposures. PCB concentrations in human milk largely reflect the pattern and amounts of the congeners present in maternal adipose tissue (ATSDR, 2000; EFSA, 2005). PCB residue data in humans suggest that assessment of tissue or body burdens of PCBs should be based on individual congeners or groups of congeners, rather than on profiles of commercial PCB formulations. Numerous publications have reported that PCB 138, PCB 153 and PCB 180, all NDL-PCBs, are the most consistently detected and quantitatively dominant congeners found in human tissues. If only one congener is to be used as a marker of total PCB exposure, 79

Safety evaluation of certain food additives and contaminants Eightieth JECFA

then PCB 153 is a good candidate, because it is very stable and often it is the most abundant congener. PCB 153 has been shown to have a high correlation with the total amount of PCBs in human breast milk, human plasma and human serum. However, it has also been noted that the correlations are lower if a more complete profile of congeners is considered and that either total PCBs or PCB 153 as a marker of the total could be misleading indicators of the differential exposure to other individual or groups of congeners of toxicological significance (ATSDR, 2000; Glynn et al., 2000). In the summary of biomonitoring data that follows, the focus is on the most abundant NDL-PCB congeners, PCB 138, PCB 153 and PCB 180. These three PCBs are the most frequently detected PCBs in population biomonitoring studies (Glynn et al., 2000) and generally account for 65–80% of the measured total sum of PCBs (Needham et al., 2005). (b) Concentrations of NDL-PCBs in blood, plasma and serum

The most comprehensive, ongoing survey of NDL-PCB levels in human serum is the United States National Health and Nutrition Examination Survey (NHANES). From 1999, the NHANES became a continuous, rolling survey, and blood samples were taken randomly from the population in the USA for measurement of a large number of environmental chemicals, including PCBs. Data on NDL-PCBs are currently available for survey cycles carried out in 1999–2000, 2001–2002, 2003– 2004, 2005–2006 and 2007–2008 (CDC, 2015). Over time, the concentrations of 31 individual NDL-PCB congeners have been measured (Table 8), although not every congener has been measured in every survey cycle or in every population subgroup within each survey cycle. In each survey cycle, blood samples for measurement of environmental chemicals were taken from approximately one third of the participants – that is, from over 1800 children and adults, ranging from children aged 12 years or older up to adults aged 74 years. In the survey cycles between 1999 and 2004, PCBs were measured in individual samples, and the geometric means and selected percentiles were estimated. Such measurements in individuals tend to have a log-normal distribution, with central tendency best estimated using a geometric mean. From 2005 onwards, a weighted pooled sample design was used because of the need to increase sensitivity by using larger volumes and to reduce costs. The measured value for a pooled sample is comparable to an arithmetic average of measurements in individuals. Consequently, a pooled sample result using an arithmetic mean is expected to be higher than the geometric mean of multiple individual results. Within each survey cycle, data are stratified according to sex, race/ethnicity (non-Hispanic, Hispanic, Mexican American) and age group (12– 19, 20–39, 40–59, 60+ years).

WHO Food Additives Series No. 71-S1, 2016 80

Supplement 1: Non-dioxin-like polychlorinated biphenyls

Table 8 NDL-PCB congeners measured in pooled serum samples in the United States NHANES surveys in 1999–2008 PCB chemical name 2,4,4′-Trichlorobiphenyl 2,2′,3,5′-Tetrachlorobiphenyl 2,2′,4,5′-Tetrachlorobiphenyl 2,2′,5,5′-Tetrachlorobiphenyl 2,3′,4,4′-Tetrachlorobiphenyl 2,4,4′,5-Tetrachlorobiphenyl 2,2′,3,4,5′-Pentachlorobiphenyl 2,2′,4,4′,5-Pentachlorobiphenyl 2,2′,4,5,5′-Pentachlorobiphenyl 2,3,3′,4′,6-Pentachlorobiphenyl 2,2′,3,3′,4,4′-Hexachlorobiphenyl 2,2′,3,4,4′,5′-Hexachlorobiphenyl and 2,3,3′,4,4′,6-hexachlorobiphenyl 2,2′,3,4′,5,5′-Hexachlorobiphenyl 2,2′,3,4′,5′,6-Hexachlorobiphenyl 2,2′,3,5,5′,6-Hexachlorobiphenyl 2,2′,4,4′,5,5′-Hexachlorobiphenyl 2,2′,3,3′,4,4′,5-Heptachlorobiphenyl 2,2′,3,3′,4,5,5′-Heptachlorobiphenyl 2,2′,3,3′,4,5′,6′-Heptachlorobiphenyl 2,2′,3,3′,5,5′,6-Heptachlorobiphenyl 2,2′,3,4,4′,5,5′-Heptachlorobiphenyl 2,2′,3,4,4′,5′,6-Heptachlorobiphenyl 2,2′,3,4′,5,5′,6-Heptachlorobiphenyl 2,2′,3,3′,4,4′,5,5′-Octachlorobiphenyl 2,2′,3,3′,4,4′,5,6-Octachlorobiphenyl 2,2′,3,3′,4,4′,5,6′-Octachlorobiphenyl and 2,2′,3,4,4′,5,5′,6-octachlorobiphenyl 2,2′,3,3′,4,5,5′,6-Octachlorobiphenyl 2,2′,3,3′,4,4′,5,5′,6-Nonachlorobiphenyl 2,2′,3,3′,4,4′,5,5′,6,6′-Decachlorobiphenyl PCB congener no. 28 44 49 52 66 74 87 99 101 110 128 138 & 158 146 149 151 153 170 172 177 178 180 183 187 194 195 196 & 203 199 206 209

The results have been summarized by the United States Centers for Disease Control and Prevention (CDC, 2009, 2013, 2015). Comparison of results from the earlier years of the survey up to 2004 with those from later years is not possible because of the switch from geometric means of individual results to arithmetic means from pooled samples. In the years up to and including 2004, serum levels of NDL-PCBs were roughly similar over the three surveys (CDC, 2009). The results for the six indicator PCBs (PCBs 28, 52, 101, 138 (+158), 153, 180) are shown in Table 9. 81

Safety evaluation of certain food additives and contaminants Eightieth JECFA

Table 9 Mean concentrations of the six indicator PCBs in serum in the NHANES survey NHANES survey year 1999–2000 2003–2004b 2005–2006c 2007–2008c 52 1999–2000 2001–2002 2003–2004 2005–2006c 2007–2008c 101 1999–2000 2001–2002 2003–2004 2005–2006c 2007–2008c 138 + 158 1999–2000 2001–2002 2003–2004 2005–2006c 2007–2008c 153 1999–2000 2001–2002 2003–2004 2005–2006c 2007–2008c 180 1999–2000 2001–2002 2003–2004 Concentration (ng/g of lipid, lipid adjusted) Mean 95th percentile WGM <LOD –a 4.90 11.3 WAM 2.74–3.81 – 0.715–5.42 – WGM <LOD –a 16.5 –a 2.66 7.60 WAM 0.348–0.807 – –a – WGM <LOD –a <LOD –a 1.65 5.83 WAM 0.632d – 0.608–0.954 – WGM 71.2 –a 19.9 94.6 15.1 75.3 WAM 1.92–81.1 – 1.84–79.9 – WGM 114 –a 27.2 126 19.8 97.1 WAM 2.83–130 – 2.88–102 – WGM 79.3 –a 19.2 87.3 15.1 81.5 Concentration (ng/g of serum, whole weight) Mean 95th percentile <LOD 0.030 0.016–0.025 0.004–0.018 –a –a 0.016 0.003–0.005 –a –a –a 0.010 0.005 –a –a 0.122 0.092 0.014–0.541 0.014–0.477 –a 0.176 0.121 0.015–0.865 0.018–0.614 –a 0.118 0.092 <LOD 0.067 – – <LOD 0.090 0.043 – – <LOD <LOD 0.033 – – 0.460 0.650 0.477 – – 0.750 0.860 0.624 – – 0.540 0.610 0.534

PCB congener no. 28

WHO Food Additives Series No. 71-S1, 2016 82

Supplement 1: Non-dioxin-like polychlorinated biphenyls

PCB congener no.

NHANES survey year 2005–2006c 2007–2008c

Concentration (ng/g of lipid, lipid adjusted) Mean 95th percentile WAM 1.78–88.6 – 2.02–82.5 –

Concentration (ng/g of serum, whole weight) Mean 95th percentile 0.010–0.589 0.010–0.501 – –

LOD: limit of detection; WAM: weighted arithmetic mean; WGM: weighted geometric mean a Mean not calculated, as proportion of results below LOD was too high. b Values are the WGM for the whole sample. c The range gives the lowest and highest values from the various sex/race/ethnicity/age subgroups of the population in which the PCB congener was measured. Values tended to increase with age. d Only one subgroup included in the survey. Source: CDC (2015)

CDC (2009, 2013) compared the NHANES results with those of earlier surveys in other countries. The NHANES results confirm earlier reports that PCB 138, PCB 153 and PCB 180 are the most frequently detected PCBs, accounting for 65–80% of total PCBs in human serum, and that concentrations of the di-orthosubstituted PCBs are usually higher than those of the mono-ortho-substituted PCBs, which in turn are higher than those of the coplanar PCBs (Patterson et al., 1994, 2009; Glynn et al., 2000; Longnecker et al., 2000; Heudorf, Angerer & Drexler, 2002; Bates et al., 2004; Apostoli et al., 2005; Needham et al., 2005; Turyk et al., 2006; CDC, 2013). The NHANES 2001–2002 results also illustrate declines over time, being generally lower than lipid-adjusted serum concentrations measured in selected populations during the 1980s–1990s (Patterson et al., 1994; Glynn et al., 2000; Longnecker et al., 2000; Link et al., 2005; Hagmar et al., 2006; CDC, 2013). CDC (2013) noted that surveys in other countries have reported higher or lower concentrations of PCBs in serum than those found in the NHANES surveys. For example, in a 1998 study of 624 urban Germans aged 0–65 years (Heudorf, Angerer & Drexler, 2002), 95th percentile levels for PCB 138, PCB 153 and PCB 180 were similar to or up to 2 times higher than 95th percentile levels in the NHANES 1999–2000 subsample. In two separate Italian studies of a regional reference population and a convenience sample in 2001–2003, median levels of PCB 138, PCB 153 and PCB 180 in serum were about 5 times higher than those from NHANES 1999–2000 (Apostoli et al., 2005; Needham et al., 2005; Turci et al., 2006; CDC, 2013). In some other countries, PCB concentrations in serum from comparable populations were 10 or more times higher than in NHANES subsamples from 1999–2000 and 2001–2002 (Jursa et al., 2006; Petrik et al., 2006). In contrast, PCB concentrations in serum from a representative population of New Zealand residents in 1996–1997 (Bates et al., 2004) were slightly lower than those from NHANES 1999–2000. 83

Safety evaluation of certain food additives and contaminants Eightieth JECFA

Biomonitoring results for the six indicator PCBs in human blood, serum and plasma from other studies around the world published since 2004 are summarized in Table 10. Table 10 Summary of studies on mean blood, serum and plasma concentrations of the six indicator PCBsa No. of Country subjects Italy 162 M + F Type of blood Age range sample (years) analysed <30–>50 43 (mean) Individual serum Mean concentration (variation or range shown in italics) PCB 28 2.22 PCB 52 0.713 PCB 101 0.622 PCB 138 71.4 PCB 153 105 PCB 180 80.0 in ng/g lipid

Study Turci et al. (2004) Variation not stated Rusiecki et al. (2005) Variation as SD Jaraczewska et al. (2006) Variation as SD Jaraczewska et al. (2006) Variation as SD Petrik et al. (2006) Variation as total range Petrik et al. (2006) Variation as total range Porpora et al. (2006) Variation as SD Thomas et al. (2006) Variation as total range Turci et al. (2006) Variation as 95th percentile Weiss et al. (2006) Variation as total range Park et al. (2007) Variation as SD

India

29 F

24–65

Individual serum Individual maternal serum

–

–

–

–

16.08 11.90 20.7 10.6 11.3 6.8 267

5.81 3.99 17.2 9.0 13.6 10.4 246

Poland

22 F

22–38

–

–

–

13.7 6.2 3.6 2.8 165 8.5–3 500

Poland

22 F

0

Individual – cord blood serum Individual serum –

–

–

Slovakia

402 M 636 F

>18

–

–

Slovakia

112 M 106 F

8–10

Individual serum

–

–

–

65.6 <8.9–343

38.8–5 193 42.4– 4 809 109 90.0 <13.7– 627 95 55 41d <9.3–200 170 387 260 70–620 <8.6–584 45 28 33d <4.7–200 121 298 –

Italy

40 F

20–40

WHO Food Additives Series No. 71-S1, 2016

Individual whole blood Individual serum

5.6 8.3 2.1d

3.0 5.5 <0.25d

3.1 3.2 <1.3d <1.3–9 –

53 26 27d <7.1–110 128 277

United 41 M + F Kingdom

22–80

Italy

175 M 151 F

37–43 (means)

Individual serum

<1.0–14 <0.25– 4.7 11.3 26.1 48.5 85.1 –

Sweden

53 F

52–81

Individual serum

–

–

–

Republic of Korea

47 M 40 F

21–>50

Individual serum

2.46 2.40

1.06 1.94

1.45 2.32

34.00 27.09

54.90 54.06

28.42 27.83

84

Supplement 1: Non-dioxin-like polychlorinated biphenyls

Study Sandanger et al. (2007) Variation as total range Zhao et al. (2007) Variation as total range Cerná et al. (2008) Variation as total range Kang et al. (2008) Variation as total range Todaka et al. (2008) Variation as SD

No. of Country subjects Canada 109 F

Type of blood Age range sample (years) analysed 48–76 Individual plasma

Mean concentration (variation or range shown in italics) PCB 28 – PCB 52 – PCB 101 – PCB 138 25.9 5.45–101 PCB 153 47.1 14.4–177 12.22 0.10– 67.11 423b 64–2 280 26.6 PCB 180 40.4 1.39–206 22.66 0.99– 102.59 374b 43–2 210 18.1 4.2–82.0 68 34 53 26 303

China, Pingqiao District Czech Republic

60 F

0

Individual – whole cord blood, lipid extraction Individual serum 14 3–126

1.56 0.05– 5.27 5 1–115 6.6 <1.0– 25.6 1.3 0.8 1.3 0.9 <0.4

3.13 0.05– 38.48 –

34.42 5.65– 231.21 186b 13–965

124 M 78 F

33 (mean)

Republic of Korea

20 M 20 F

27–58

Individual serum

6.0 <0.6– 24.6 2.6 1.5 2.6 1.8 <0.3

5.2 0.9–19.2 2.0 1.2 1.8 1.3 <0.4

16.4

<7.3–67.8 6.6–106 43 19 39 20 108 97 40 85 41 242

Japan

51 M

60–79

Todaka et al. Japan (2008) Variation as SD Italy, TurrioBaldassarri et general al. (2008) population of Variation not Brescia stated Axelrad, USA Goodman & Woodruff (2009) Variation as 95th percentile Koppen et al. Belgium (2009) Variation as total range Porpora et al. Italy (2009) Variation as 95% CI Röllin et al. South (2009) Africa Todaka et al. Japan (2010) Variation as SD Windham et al. USA (2010) Variation as SD

76 F

60–66

52 M 42 F

51 (mean)

Individual whole blood, lipid extraction Individual whole blood, lipid extraction Pooled serum

496 F

16–39

Individual serum

–

–

–

–

14d

–

41 1 054 – 1 071 F 0 Individual – cord blood serum Individual serum 3.4 2.5–4.5 96 F 119 F Primiparous 611 F Not provided 21–40 Individual plasma Individual whole blood, lipid extraction Individual serum – 1.3 0.7 – – – 21.3 2.3–156.7 1.6 1.3–1.9 – 0.8 0.6 – 1.6 1.3–1.9 – 0.8 0.6 – 33.8 29.0–39.3 3.6 14.8 8.0 10.0 c

37.7 2.3–230.2 61.8 51.8–73.8 3.2 26.8 15.4 14.3 18.4

26.0 2.1–153.1 34.4 29.0–40.8 – 17.3 11.6 9.3 13.4

78 F

18–45

6–10

13.1

85

Safety evaluation of certain food additives and contaminants Eightieth JECFA

Table 10 (continued) No. of Country subjects France 1 055 F Type of blood Age range sample (years) analysed <30–49 Individual serum Mean concentration (variation or range shown in italics) PCB 28 – PCB 52 – PCB 101 – PCB 138 – PCB 153 88.4d 217.6 Spain 284 M 323 F 6–75 Individual serum 0.0d 0.0–1.1 Spain 1 259 F Pregnant 31 (mean) Individual serum – 0.0d 0.0–0.0 – 0.0d 0.0–7.1 – 0.0d 0.0–132.6 21.83 20.99– 22.70 24.9 15.1 12.8 6.4 22.04d 14.13 33.7 11.3 201 F 58–78 Individual serum – – – 62 26–111 (median value 56) 20 (sex information only available in supplementary table) 386 Individual (age serum information only available in supplementary table) 18–74 Individual serum 4.8 6.6 1.2 2.5 1.6 2.6 28 20 21.8d 0.0–111.1 38.92 37.48– 40.42 43.2 27.1 24.0 12.5 33.52d 22.22 59.0 36.5 124 63–195 (median value 114) 26 22 6.7d 0.0–36.3 26.99 25.92– 8.12 31.4 22.1 15.3 9.7 23.97d 22.97 26.7 10.2 83 46–130 (median value 79) 14 9.2 PCB 180 –

Study Bachelet et al. (2011) Variation as 90th percentile HenríquezHernández et al. (2011) Variation as 5th–95th percentile Ibarluzea et al. (2011) Variation as 95th percentile Kalantzi et al. (2011) Variation as SD Todaka et al. (2011) Variation as SD Amodio et al. (2012) Variation as interquartile range Arrebola et al. (2012) Variation as SD

Greece

34 M 27 F 97 F Secundiparous 50 M 51 F

20–65

Individual serum

–

–

0.13 0.38 0.7 0.4 4.15d 1.51 –

Japan

22–41

Italy (Sicily)

<30–>69

Individual 1.1 whole blood, lipid 0.6 extraction Individual 4.41d serum 1.76 Individual serum –

0.7 0.5 4.23d 1.59 –

WHO Food Additives Series No. 71-S1, 2016

Plurinational State of Bolivia Bergkvist et al. Sweden (2012) Variation as 5th–95th percentile of median Eguchi et al. India (2012) Variation as SD

22 M 90 F

18–70

Fréry et al. (2012) Variation as 95% CI Porta et al. (2012) Variation as 95% CI

France

2.2 1.9–2.5

1 0.2–3.1 –

1.1 0.9–1.3 –

70 60–80 44.0 40.0–48.4

110 100–130 63.4 57.5–70.0

90 80–110 54.4 82.4– 103.5

Spain

94 M 137 F

18–>65

Individual serum

–

86

Supplement 1: Non-dioxin-like polychlorinated biphenyls

Study Rylander et al. (2012) Variation as total range Nøst et al. (2013) Variation not stated Aylward et al. (2014) Variation as SD Ben Hassine et al. (2014) Variation as SD Chovancová et al. (2014) Variation as total range Esposito et al. (2014) Variation as total range Huetos et al. (2014) Variation as 95% CI Huetos et al. (2014) Variation as 95% CI ArtachoCordón et al. (2015) Variation as SD Ulutaş et al. (2015) Variation as SD Whitehead et al. (2015) Variation as 25th–90th percentile Zubero et al. (2015) Variation as 95% CI Mrema et al. (2014) Variation as total range

No. of Country subjects Norway 273 F

Type of blood Age range sample (years) analysed 48–62 Individual plasma

Mean concentration (variation or range shown in italics) PCB 28 – PCB 52 – PCB 101 – PCB 138 66 PCB 153 87 PCB 180 70

Norway

53 M

29–82

Individual serum

–

–

–

<LOD–164 <LOD–211 <LOD– 182 – 120–240 –

USA

43 M

55–80

Individual serum Individual serum Pooled serum

–

–

–

54.0 33.2 26.1 10.7 –

69.7 39.9 51.9 19.7 236 16.2–628

63.1 38.2 34.6 11.3 –

Tunisia

32 M 81 F 20 M 20 F

20–81

–

–

0.44 3.3 –

Slovakia

24–62

–

–

Italy

32 M 26 F

18–64 24–64

Individual serum

4.4

1.0

3.2 0.6–4.9 –

37 6.7–117.7 45.77 41.07– 50.47 47.14 37.69– 56.59 28.59 16.64 0.59 0.78 6.8 g

56 10.1– 195.3 62.49 54.88– 70.10 60.33 50.09– 70.56 119.07 35.96 0.64 0.75 11 g

73 5.6–1 138 83.47 72.23– 94.72 74.63 61.07– 88.19 31.74 11.00 0.39 0.47 10g 6.1–25 92.4 84.0– 101.7 189–403d 10–1 616d

0.8–14.9 0.2–7.3 Spain 963 M 18–>50 Individual serum – –

Spain

917 F

18–>50

Individual serum

–

–

–

Tunisia

54 F

38–50

Individual serum

–

–

–

Turkey

57 F

20–41

USA

48 F

<34–>41

Individual 0.17 whole blood, lipid 0.34 extraction Individual 4.1g serum 3.1–6.8 Individual serum –

0.06 0.28 –

0.10 0.34 1.7 g

<1.6–2.9 – –

4.8–14 64.8 8.7–71.6

6.5–28 92.4 83.2– 102.7 265–668d 13–2 205d

Spain

60 M 102 F

20–69

(in pmol/g serum lipid)

Italy

182 M 185 F

19–70

Individual serum

–

–

–

174–532d 13–1 612d

87

Safety evaluation of certain food additives and contaminants Eightieth JECFA

Table 10 (continued) No. of Country subjects Type of blood Age range sample (years) analysed 9–11 Individual whole blood Mean concentration (variation or range shown in italics) PCB 28 – PCB 52 – PCB 101 – PCB 138 0.05 0.04 0.150 0.552 PCB 153 0.10 0.09 0.277 0.995 PCB 180 0.06 0.06 0.260 1.022 (in µg/L of whole blood)

Study

Link et al. Germany 803 M + F (2012) Variation as SD Fromme et al. Germany 33 M (2015) 37 F Variation as 95th percentile Hirai et al. Japan (2005) Variation as SD Apostoli et al. (2005) Variation as SD Redding et al. (2008) Variation not stated Italyf 12 M 12 F

4–76

Individual 0.003 whole blood, lipid 0.008 extraction Individual 0.006 3 whole blood, lipid 0.003 4 extraction Individual serum Plasma, serum & whole blood –

0.001 0.002

0.001 0.003

(in ng/g whole blood)

25–46

0.002 8 0.003 2 –

0.005 7 0.004 9 –

0.064 7 0.032 4 (in µg/L)

0.171 0 0.088 5 1.43 1.25 0.1–1.7

0.089 7 0.049 1 1.64 1.69 –

165 M 146 F

20–79

0.83 0.70 –

Dirtu et al. (2010) Variation as total range Dirtu et al. (2010) Variation as total range Grimalt et al. (2010) Variation as SD Grimalt et al. (2010) Variation as SD Turci et al. (2010) Variation as total range Turci et al. (2010) Variation as total range Cao et al. (2011) Variation as maximum

Summary of studies from various countries, 2002– 2008 Belgium 16 M 4F

–

–

–

Not stated

Individual serum

–

–

–

0.73d

1.12d

0.71d 0.15– 1.71d 0.88d

Romania

31 M 22 F

Not stated

Individual serum

–

–

–

0.19–1.59d 0.240– 2.41d 0.57d 0.68d

Spain

410 F

0

Spain

285 M + F

4

19 M Italy (Novafel- 17 F tria) Italy (Pavia) 36 M 23 F

42 (mean)

42 (mean)

China

1210 F

28 (mean)

Individual 0.014 cord blood serum 0.070 Individual 0.024 serum 0.33 Individual – serum <LOD– 0.550d Individual – serum <LOD– 0.220 Individual 0.25 cord blood serum 6.68

0.021 0.076 0.036 0.30 0.038d <LOD– 0.500d 0.105 <LOD– 1.10 0.39 17.67

0.032 0.097 0.088 0.14 0.040d <LOD– 0.96d – <LOD– 0.11 0.20 6.40

0.09–4.71d 0.09–3.83d 0.07– 7.72d 0.17 0.21 0.20 0.13 0.24 0.53 0.220d 0.24 0.35 0.67 0.560d 0.36 0.20 0.47 0.305d

WHO Food Additives Series No. 71-S1, 2016

0.11–1.64d 0.50–3.42d 0.05– 0.95d 0.400 0.560 0.470 <LOD– 1.10 0.09 2.15 <LOD– 2.10 0.20 3.75 <LOD– 1.55 –

88

Supplement 1: Non-dioxin-like polychlorinated biphenyls

Study Schettgen et al. (2011) Variation as 95th percentile Lind et al. (2012) Variation as 25th–75th percentile Pandelova & Schramm (2012) Variation not stated Lopes et al. (2014) Variation as SD Lopes et al. (2014) Variation as SD

No. of Country subjects Germany 105 M + F

Type of blood Age range sample (years) analysed 5–84 Individual plasma, 7 age groups 70 Individual serum

Mean concentration (variation or range shown in italics) PCB 28 0.011– 0.016d 0.025– 0.037d – PCB 52 <0.01d <0.01– 0.017d – PCB 101 <0.01– 0.011d <0.01– 0.021d – PCB 138 0.117– 0.874d 0.379– 1.655d 0.819d 0.619– 1.116d PCB 153 0.149– 1.338d 0.536– 2.223d 1.428d 1.114– 1.848d 0.22 PCB 180 0.098– 1.292d 0.350– 1.876d 1.165d 0.918– 1.488d 0.12

Sweden

506 M 520 F

Germany 53 M 53 F

20–29

Individual plasma

–

–

–

0.34

Portugal

68 F Maternal serum 68 F

17–41

Individual maternal serum

–

–

–

0.16 0.08 0.11 0.05

1.01 0.24 0.97 0.18

0.46 0.24 0.36 0.13

Portugal

0

Individual – cord blood serum

–

–

CI: confidence interval; F: females; LOD: limit of detection; M: males; SD: standard deviation a In time-series studies, the values given are for the most recent year. b From a polluted area formerly manufacturing PCBs. c PCB 138/158, when co-eluted. d Median values.

(c) Concentrations of NDL-PCBs in adipose tissue

Adipose tissue concentrations have been measured in the general population and in case–control studies on particular diseases. For details on the latter, see sections 2.3.4–2.3.15. The results of studies on NDL-PCB concentrations in adipose tissue are summarized in Tables 11 and 12. Table 11 gives the concentrations of the six indicator PCBs in adipose tissue. Table 12 summarizes some studies in which both DL- and NDL-PCBs have been measured, giving overall values for the sum of all PCBs measured together with the sum of the three indicator PCB congeners that make the highest contribution to total PCB levels in food and in adipose tissue – that is, PCBs 138, 153 and 180. The majority of studies found a positive correlation of adipose tissue levels with age, but no significant difference between males and females.

89

Safety evaluation of certain food additives and contaminants Eightieth JECFA

Table 11 Summary of studies on concentrations of the six indicator PCBs in adipose tissue Study Covaci et al. (2002) Variation as total range Costabeber Spain & Emanuelli (2003) Variation as SD De Saeger et Belgium al. (2005) Variation as total range Kiviranta et al. (2005) 123 F 15–87 51 (mean) Country Belgium No. of subjects 11 M 9F Age range (years) 19–77 30–65 Mean concentrationa (ng/g lipid) (variation or range shown in italics) PCB 28 PCB 52 PCB 101 PCB 138 PCB 153 PCB 180 0.3b 0.8b 105.1b 211.1b 148.3b 1.3b nd–11.8b (LOD 0.2– 0.5) 39 50 <LOQ (LOQ 10) <LOQ–16 4.61 6.92 – nd–2.2b (LOD 0.2– 0.5) 11 14 <LOQ (LOQ 10) <LOQ–17 0.784 0.779 – nd–8.8b (LOD 0.2– 0.5) 2 4 <LOQ (LOQ 10) <LOQ 1.40 2.02 – 29.0–227.5b 67.8–399.2b 50.1–342.6b 102 57 140 <LOQ–348 74.7 52.9 – 121 66 271 10–665 135 94.9 11.54 19.14 13.09 8.13 310 55–848 278 18–1 294 131 92 178 134 83 213 <LOQ–570 106 74.8 4.58 4.87 6.12 6.01 232 11–931 201 13–1 084 98.1 74.4 185

57 M 47 F

2–90 17–91

Finland

214 M 206 F 34 F

13–81

Variation as SD Rusiecki et al. India (2005) Variation as SD Li et al. (2006) Singapore Variation as SD Naert et al. Belgium (2006) Variation as total range Vaclavik et al. (2006) Variation as total range Covaci et al. (2008)

24–65

36 F 31 M 22 F

22–40 19–83 22–84

– –

– –

– –

9.36 5.49 181 19–543

Denmark

402 F

50–65

–

–

–

140 7–629

Belgium

18 M 7F 20 F

9–70

nd (LOD 1–4) 5.71

nd (LOD 1–4)

–

83.0 58.5 88.3

WHO Food Additives Series No. 71-S1, 2016

Variation as SD Fernandez et Spain al. (2008) Variation as SD Shen et al. (2008) Variation as total range Tan et al. (2008) Variation as SD Kalantzi et al. (2009) Variation as SD Pulkrabová et al. (2009)

24–81

China

20 M 4F

26–73 33–62

12.9 1.0–189

nd 0.242 (LOQ 0.002– 0.2) 0.239 0.6 1.34 0.15–3.46 – 0.16–4.48 –

46.8 50.7 3.14–251 10.3c 12.4c 29.3 26.5 121.6

89.9 52.5 3.13–247 15.4c 20.2c 20.8 20.1 233.6

89.1 36.3 1.17–230 9.87 14.3 24.2 23.3 245.0

Singapore

88 F

18–40

0.66c 1.13c –

Brazil

25 F

40–71

–

–

Czech Republic

5M 93 F

17–60

2.0

1.6

4.2

90

Supplement 1: Non-dioxin-like polychlorinated biphenyls

Study Variation as 5th–95th percentile Arrebola et al. (2010) Variation as SD Arrebola et al. (2010) Variation as SD Arrebola et al. (2010) Variation as SD Arrebola et al. (2010) Variation as SD Çok et al. (2010) Variation as SD Çok et al. (2010) Variation as SD Bräuner et al. (2011) Variation as 5th–95th percentile Arrebola et al. (2012) Variation as SD

Country

No. of subjects

Age range (years)

Mean concentrationa (ng/g lipid) (variation or range shown in italics) PCB 28 PCB 52 PCB 101 PCB 138 PCB 153 PCB 180 0.9–4.0 0.6–3.6 0.8–14.0 46.7–240.8 82.8–492.4 61.6–483.8 – – – 42.24d 8.13 33.81d 9.08 23.41d 12.51 46.26d 7.55 64.6 78.8 68.3 81.2 140b 77–280 206.2d 3.26 174.20d 2.97 97.89d 6.54 166.80d 4.86 55.2 33.9 58.2 66.2 310b 200–540 0.9d 7.4 156.28d 4.85 106.22d 5.66 67.00d 9.06 112.89d 6.24 94.1 131.6 48.2 23.5 220b 150–340 1.7d 19.5

Spain (urban) Spain (urban)

105 M

16–81

81 F

16–85

–

–

–

Spain 92 M (semirural) Spain 109 F (semirural) Turkey 25 M (infertile) 21 M (fertile) 126 M 119 F

16–81

–

–

–

16–85

–

–

–

21–46

16.5 47.9 25.1 67.9 –

72.1 93.4 24.2 36.2 –

57.9 59.4 49.1 55.3 –

Turkey

21–46

Denmark

50–64 50–64

22 M Plurinational State 90 F of Bolivia

18–70

–

–

–

3.6d 15.7

F: females; M: males; LOD: limit of detection; LOQ: limit of quantification; nd: not detected; SD: standard deviation; –: not measured a Except where otherwise indicated. b Median value. c PCBs 28 + 31; PCBs 138 + 158; PCBs 153 + 132. d Geometric mean.

Table 12 Summary of studies on DL- and NDL-PCB concentrations in adipose tissue Study Smeds & Saukko (2001) Covaci et al. (2002) Country Finland Belgium No. of subjects 17 M 10 F 11 M 9F 123 F 57 M 47 F Age range (years) 19–88 19–95 19–77 30–65 15–87 2–90 17–91 No. of PCBs analysed 7 groups 35 11 7 ∑PCBs (ng/g) 504a 334b 560a 658a Range PCBs 138, 153 + 180 (ng/g lipid) – 105–211b 102–134a 140–271a

Costabeber & Emanuelli Spain (2003) De Saeger et al. (2005) Belgium

91

Safety evaluation of certain food additives and contaminants Eightieth JECFA

Table 12 (continued) Study Johnson-Restrepo et al. (2005) Kiviranta et al. (2005) Rusiecki et al. (2005) Li et al. (2006) Naert et al. (2006) Vaclavik et al. (2006) Kunisue et al. (2007) Covaci et al. (2008) Fernandez et al. (2008) Shen et al. (2008) Country USA Finland India Singapore Belgium Denmark Japan Belgium Spain China No. of subjects 40 M 12 F 214 M 206 F 34 F 36 F 31 M 22 F 402 F 18 M 10 F 18 M 7F 20 F 20 M 4F 83 F 25 F 5M 93 F 179 M 178 F 46 M Age range (years) 18–51 13–81 24–65 22–40 19–83 22–84 50–65 25–81 53–109 9–70 24–81 26–73 33–62 18–40 40–71 17–60 16–81 16–85 21–46 50–64 50–64 18–70 No. of PCBs analysed 37 37 2c 7 7 10 62 7 37 10 41 4 7 3 7 10 3 ∑PCBs (ng/g) 144 437b – 57a 605b 872b 1 300 850 334a 687b 154a 45b 79.6a 595b 323a 370a 382a,d 351a,e – Range PCBs 138, 153 + 180 (ng/g lipid) – 63–116b 5–12a 6–13a 181–310 130–266b – 83–131a 81–168b 36–52a 6–11b 30–339a 110–230b 34–150 43–173 55–94a 140–310b 130–280b 41–105

Tan et al. (2008) Singapore Kalantzi et al. (2009) Brazil Pulkrabová et al. (2009) Czech Republic Arrebola et al. (2010) Spain Çok et al. (2010) Bräuner et al. (2011) Arrebola et al. (2012) Turkey Denmark

126 M 119 F Plurinational 22 M State of 90 F Bolivia

WHO Food Additives Series No. 71-S1, 2016 92

F: females; M: males a Mean value or range of means. b Median value. c PCB 138 not measured. d Mean for infertile men. e Mean for fertile men. Source: Expanded and adapted from Arrebola et al. (2010)

(d) Concentrations of NDL-PCBs in breast milk

Human milk is an important source of exposure to PCBs for breastfed infants. Owing to its high fat content, human milk can accumulate large amounts of PCBs, which is easier for detection, making it an ideal matrix for the determination of PCBs. Unlike human blood, human milk can be sampled using non-invasive techniques. It can also be used as a good indicator of the body burden of lipophilic non-metabolized PCBs, as fat is mobilized for the production of milk during lactation. Animal studies have revealed that large amounts of PCBs can

Supplement 1: Non-dioxin-like polychlorinated biphenyls

be eliminated through lactation (Lindell, 2012). However, it is less clear that this is the case in humans (see below). WHO introduced worldwide measurement campaigns to determine the exposure of infants to the 12 DL-PCBs and the six indicator PCBs over time. To date, there have been five rounds of the UNEP/WHO (2013) survey from 2000 up to 2012. For the sum of the six indicator PCBs, the median concentrations over the 12 years of the survey are between 10.8 and 30.7 ng/g lipid, and maxima are between 37.1 and 65.8 ng/g lipid (UNEP/WHO, 2013). The results of studies on PCB concentrations in human milk are summarized in Table 13. Investigations have included the measurement of both DL- and NDL-PCB congeners. In Table 13, only results relating to the six indicator PCBs are presented, including the congeners that make the highest contribution to total PCB levels in food and in human milk – that is, PCBs 138, 153 and 180. The study of Link et al. (2012) showed that higher amounts of NDLPCBs were found in whole blood of children aged 9–11 years if they had been breastfed, compared with children who had not been breastfed. The differences between the two groups were 4-fold, 2-fold and 2-fold for PCBs 138, 153 and 180, respectively. PCB 128, although not one of the routinely reported NDL-PCB congeners, has been measured in human milk at mean concentrations ranging from 0.2 to 4.0 ng/g lipid (Koopman-Esseboom et al., 1994; Todaka et al., 2010; Ryan & Rawn, 2014).

Table 13 Summary of studies on concentrations of the six indicator PCBs in human milk Study EFSA (2005) Country Summary of studies from 18 European countries Summary of studies from various countries, 1998–2007b Sweden No. of subjects 58 pooled samples – PCB 28 4.6 PCB 52 0.51 Mean concentration (ng/g lipid)a PCB 101 PCB 138 PCB 153 0.86 64.0 81.7 PCB 180 58.5

Redding et al. (2008)

–

–

–

–

100–10 100 – (ng/mL whole milk) 74 40 77–130 3.4–11.4 35 30 48–96 1.6–6.7

Glynn et al. (2001) Jaraczewska et Poland al. (2006) Ingelido et al. Italy (2007) Çok et al. Turkey (2009, 2011, 2012)

27 22 39 198

3.5 – 2.1–6.2 1.2–3.0

0.45 – 0.20–0.33 0.2–0.7

0.58 0.8 0.54–1.1 0.3–0.8

37 26 58–98 1.6–5.8

93

Safety evaluation of certain food additives and contaminants Eightieth JECFA

Table 13 (continued) Study Lignell et al. (2009) Cerná et al. (2010) Glynn et al. (2011) Croes et al. (2012) Shen et al. (2012) Focant et al. (2013) Malarvannan et al. (2013) Vigh et al. (2013) a

Country Sweden Czech Republic Sweden Belgium China France Philippines Hungary

No. of subjects 325 90 203 84 23 (urban) 51 (rural) 44 30 22 (days 5, 12 and 84 postpartum)

PCB 28 2.8 2.1–19.2 <0.6–1.8d 0.88 2.68 2.69 7.8 1.4 1.0–1.4

PCB 52 – 0.32–2.87 – 0.23 0.34 0.41 15.8 0.74 0.8–1.4

Mean concentration (ng/g lipid)a PCB 101 PCB 138 PCB 153 – 29 58 0.78–2.80 – 0.39 0.23 0.17 8.3 2.0 2.2–2.8 125–597c 18–24d 18.7 14.21 7.73 40.0 15 7.2–8.8 93–903c 31–48d 32.8 11.2 6.59 83.0 22 10.5–12.8

PCB 180 28 80–539c 15–23d 17.2 2.40 1.48 48.3 4.9 3.5–4.3

Except where otherwise indicated. Excludes results from Inuit populations. The highest levels for PCBs 138, 153 and 180 were found in residents in an industrial area in the vicinity of a former plant that produced PCB-based paints in the 1970s and 1980s. d Median values. b c

Early reviews suggested that infant exposure to PCBs may be reduced with duration of lactation and with successive pregnancies. However, whereas time-series data indicate that body burdens of PCBs are lowered by lactation, a study has shown that this occurs slowly, averaging only 1% per month over longer lactation periods (Hooper et al., 2007). The authors concluded that 6 months of breastfeeding would decrease PCB 153 levels in mothers by approximately 4% and that decreases are not higher earlier in lactation, although they may be lower later. Thus, 6–12 months of breastfeeding would not greatly reduce a mother’s body burden of these chemicals, and prenatal and lactational exposures for a second child would not be markedly lower than those for the first child. Redding et al. (2008) developed a model for predicting the concentrations of the most prevalent congener, PCB 153, in human milk. They first transformed an earlier physiologically based pharmacokinetic (PBPK) model for lactational transfer of PCB 153 in mice into a PBPK model for a non-pregnant human female at an average childbearing age of 25 years. They then predicted the body burden buildup of PCB 153 from birth over a 25-year period based on realistic exposure levels found in foods, as reported for the Japanese population, incorporating agerelated physiological changes during the first 25 years of life. The PBPK model was then transformed to a lactating 25-year-old woman by incorporating the

WHO Food Additives Series No. 71-S1, 2016 94

Supplement 1: Non-dioxin-like polychlorinated biphenyls

physiological changes related to pregnancy and childbirth. Using this model, the authors predicted levels of PCB 153 in milk using three sets of values (minimal, median, maximal) for the most sensitive parameters based on actual data reported in the literature. They found that the range and spread of the model predictions of PCB 153 concentrations in breast milk agreed well with the human biomonitoring data found worldwide. The mean predicted value was 3 µg/L whole milk. (e) Age trends for body burden of NDL-PCBs

Owing to the accumulation of poorly metabolized PCBs, the total body burden of PCBs, including NDL-PCBs, and their metabolites generally increases with age, and this is reflected in blood and adipose tissue biomarker concentrations (Glynn et al., 2003, 2007; EFSA, 2005; Turci et al., 2006; Park et al., 2007; Sandanger et al., 2007; Arrebola et al., 2010; Ibarluzea et al., 2011; Rylander et al., 2011; Quinn & Wania, 2012; Nøst et al., 2013; Mrema et al., 2014; Quinete et al., 2014; Fromme et al., 2015; Zubero et al., 2015). Apostoli et al. (2005), for example, found an increase of 1.7 ng/mL in concentrations of PCBs (total 24 congeners) in whole blood for every 10-year increase in age. Increases in cross-sectional body burden trends with age have been variously attributed to bioaccumulation with longer exposure (Naert et al., 2006; Fernandez et al., 2008), age-dependent changes in metabolism (Ahlborg et al., 1995; Fangstrom et al., 2005) or older individuals having lived during periods of greater environmental exposure (Quinn & Wania, 2012). (f) Time trends for body burden of NDL-PCBs

Since the widespread bans on PCB manufacture and use in the 1970s and 1980s, levels in environmental media have been declining. In smaller studies of various populations in the USA that were not occupationally exposed to PCBs or consuming fish from PCB-contaminated waters, the concentrations of PCBs in adipose tissue and in human milk mostly decreased over time between 1973 and 1996 (ATSDR, 2000). In the last two decades, a clear decrease in concentrations of PCBs in blood has been observed in Europe. Overall, mean concentrations of PCB 138, PCB 153 and PCB 180 in whole blood appear to have decreased by approximately 80% in 20 years (Link et al., 2005; Hagmar et al., 2006; Agudo et al., 2009; AMAP, 2009). However, compared with North America (CDC, 2005, 2009), concentrations of PCB 138, PCB 153 and PCB 180 in serum were higher by 2- to 5-fold in Germany in 1998 (Heudorf, Angerer & Drexler, 2002) or Italy in 2001–2003 (Turci et al., 2004; Apostoli et al., 2005; Needham et al., 2005). Some examples of the decline in PCB concentrations over time are given below. A marked decline in PCB levels in blood was shown in the German Environmental Surveys carried out in 1998 (Becker et al., 2002) and 2003–2006 95

Safety evaluation of certain food additives and contaminants Eightieth JECFA

WHO Food Additives Series No. 71-S1, 2016 96

(Becker et al., 2008). Mean concentrations for the sum of PCBs in blood were 1.3–1.7 µg/L in 1998 and 0.3 µg/L in the later survey, with a large difference (a factor of 5.6) between age groups 18–25 and 66–69 years. A study of 803 German children aged 9–11 years in Baden-Württemberg, in which PCBs 138, 153 and 180 were measured in whole blood in the years 1996– 1997, 1998–1999, 2000–2001, 2002–2003, 2004–2005 and 2008–2009, showed that the median concentration of the sum of the three congeners decreased over this time period from 470 to 180 ng/L. For the individual PCBs, median concentrations decreased from 150 to 40 ng/L for PCB 138, from 180 to 70 ng/L for PCB 153 and from 150 to 30 ng/L for PCB 180 (Link et al., 2012). In a series of 14 annual surveys conducted between 1995 and 2008 in Germany, declines in mean concentrations of three indicator PCBs in plasma (uncorrected for lipid content) were observed in male and female students; the concentrations declined from 0.62 to 0.34 µg/L for PCB 138, from 0.53 to 0.22 µg/L for PCB 153 and from 0.32 to 0.12 µg/L for PCB 180 (Pandelova & Schramm, 2012). In a study in 53 Norwegian men who were repeatedly sampled over time, mean PCB 153 concentrations in serum steadily declined over the five survey periods, from a range of 280–440 ng/g lipid (mean depended on the age cohort) in 1979 to 120–240 ng/g lipid in 2007 (Nøst et al., 2013). An Australian study also reported more recent reductions in NDL-PCB exposures over time using pooled serum biomonitoring data from 2003 and 2009, PBPK modelling and reconstruction of historical intake and elimination rates for eight NDL-PCB congeners (PCBs 74, 99, 138, 146, 153, 170, 180 and 197), although PCB 74 and PCB 99 exposures declined more slowly than the others (Bu et al., 2015). A study from the Czech Republic also confirmed a downward trend in NDL-PCB concentrations (PCBs 28, 52, 101, 138, 153 and 180) measured in human milk in biomonitoring conducted between 1994 and 2009 (Cerná et al., 2012). In the WHO surveys on human milk (UNEP/WHO, 2013), concentrations of the six indicator PCBs were found to be steadily decreasing, with levels mostly in the range of 100–500 ng/g lipid in 2000–2003 and decreasing to around 5–55 ng/g lipid in 2008–2012. (g) Concentrations of NDL-PCBs in more highly exposed populations

High exposure to PCBs can be due to consumption of foods containing high levels of PCB contamination or living in an area polluted from earlier PCB manufacturing and disposal. In populations that are exposed to high levels of PCBs through the diet (e.g. from high consumption of fish from PCB-contaminated waters, recreational/ sport fish or whale blubber), the concentrations of PCBs in blood, tissue and milk

Supplement 1: Non-dioxin-like polychlorinated biphenyls

will be higher than for populations not so exposed (ATSDR, 2000; EFSA, 2005). In a Lake Ontario (USA) population, median PCB concentrations in serum in non-consumers and consumers of sport fish, respectively, were as follows: 42 and 47 ng/g lipid for PCBs 28 + 31; 15 and 17 ng/g lipid for PCB 52; 4.0 and 4.5 ng/g lipid for PCB 101; 52 and 69 ng/g lipid for PCBs 138 + 163; and 51 and 65 ng/g lipid for PCB 180 (Bloom et al., 2005). In the Faroe Islands population, which includes pilot whale in their diet, arithmetic mean concentrations of PCBs 28, 138 + 158, 153 and 180 in serum were, respectively, 8, 245, 378 and 241 ng/g lipid in 7-year-olds and 8, 81, 250 and 147 ng/kg lipid in 14-year-olds (Barr et al., 2006). In indigenous populations from three areas of the Russian Arctic, mean serum concentrations of the indicator PCBs 28 (+ 31), 52, 101, 138, 153 and 180 in 72 males and 137 females aged 6–77 years were as follows: 9.7–50 ng/g lipid for PCB 28 (+ PCB 31); 13–56 ng/g lipid for PCB 52; 22–73 ng/g lipid for PCB 101; 32–58 ng/g lipid for PCB 138; 65–123 ng/g lipid for PCB 153; and 23–79 ng/g lipid for PCB 180. Levels were generally higher in older compared with younger people and in men compared with women (Rylander et al., 2011). In the Inuit population of Greenland, the concentrations of the indicator PCBs 138, 153 and 180 in autopsy samples of adipose tissue from 22 women with mean age 61 years and 19 men with mean age 59 years, collected between 1992 and 1994, were 19-, 21- and 16-fold higher than the concentrations of the respective congeners measured by the same analytical method in Canadians from Quebec City. In the Greenland subjects, the sum of the three most abundant PCB congeners (PCBs 138, 153 and 180) represented 63–68% of the total PCBs in the tissue samples (Dewailly et al., 1999). Similarly, in a review of 30 studies from 15 different countries that measured PCB 153 levels in whole human milk between 1998 and 2007, the mean concentration in milk from Inuit Canadians was 17 µg/L, compared with a range of means from 0.1 to 10.1 µg/L in the other studies. Mean concentrations in milk for non-Inuit Canadians in two of the studies were 3–4 µg/L (Redding et al., 2008). PCB concentrations have been measured in residents of industrial areas polluted with PCBs due to local manufacturing. In two areas of Slovakia in which commercial PCBs were manufactured between 1959 and 1984, 196 men and 119 women between 20 and 75 years of age, living in the areas in 2001, had mean concentrations in serum of 313, 604 and 630 ng/g lipid for PCBs 138, 153 and 180, respectively (Jursa et al., 2006). In a larger study of one of the two contaminated areas in Slovakia, 434 men and 575 women over 18 years of age, living in the area in 2001, had mean concentrations in serum of 572, 912 and 913 ng/g lipid for PCBs 138, 153 and 180, respectively (Petrik et al., 2006). In children 8–10 years of age from the same contaminated area, mean concentrations of PCBs in serum were lower than those of adults, at 149, 236 and 212 ng/g lipid for PCBs 138, 153 97

Safety evaluation of certain food additives and contaminants Eightieth JECFA

and 180, respectively (Petrik et al., 2006). All levels in adult and child residents of contaminated areas were higher than those in adults and children from noncontaminated areas of Slovakia (see Table 10). In Italy, only one manufacturing plant in the industrial town of Brescia produced PCBs between 1958 and 1983. The town has about 200 000 residents. Widespread high levels of PCBs in soil, above the legal limit, were found in the area around the plant, with approximately 11 000 residents living in this contaminated area. It included an agricultural area to the south of the plant that contained several small farms, and PCBs from farm produce were known to have entered the local food-chain. In one study, blood samples were collected from 527 male and female Brescia residents in 2003 or 2004. Mean concentrations in serum were 188 ng/g lipid for PCB 138, 332 ng/g lipid for PCB 153 and 424 ng/g lipid for PCB 180 (Zani et al., 2013). In another study, blood samples were collected in 2004 or 2005 from residents in six different pooled sera groups from known contaminated areas and one pooled sera group from the general population of Brescia. The ranges of mean concentrations for the seven areas were as follows: <0.3–5.7 ng/g lipid for PCB 28; 0.2–<7.2 ng/g lipid for PCB 52; 0.2–<9.2 ng/g lipid for PCB 101; 66–925 ng/g lipid for PCBs 138 + 163; 141–2622 ng/g lipid for PCB 153; and 190–4221 ng/g lipid for PCB 180. The highest concentrations were found in former workers at the PCB plant and in consumers of contaminated food from the local farms (TurrioBaldassarri et al., 2008). Further examples of PCB concentrations in serum in groups exposed to high levels in the diet are shown in Table 14.

WHO Food Additives Series No. 71-S1, 2016

Table 14 PCB concentrations in plasma/serum after consumption of PCB-contaminated fish or marine mammals Region Nunavik, northern Canada Geometric mean PCB congener no. concentration (ng/g lipid)a 138 58 Range: 10–387 153 105 Range: 19–709 180 44 Range: 8–384 489 male, 264 female Native 28 0.04 American adults 52 0.04 – 101 138 + 163 + 164 0.52 153 0.63 180 0.47 Sample 159 Inuit females Reference Muckle et al. (2001)

St Lawrence River, USA

DeCaprio et al. (2005)

98

Supplement 1: Non-dioxin-like polychlorinated biphenyls

Region St Lawrence River, USA

Sample Native American adolescents

Great Lakes, USA

Fishing ship male captains 293 fish consumers Referents consuming little or no sport fish

Geometric mean PCB congener no. concentration (ng/g lipid)a 52 0.03 (NBF) 0.03 (BF) 101 [+90] 0.05 (NBF) 0.05 (BF) 138 [+ 163 + 164] 0.06 (NBF) 0.08 (BF) 153 0.07 (NBF) 0.11 (BF) 180 0.03 (NBF) 0.05 (BF) 96 µg/L 138 + 163 153 + 132 + 105 97 µg/L 95 µg/L 180 138 + 163 95 µg/L 153 + 132 + 105 97 µg/L 92 µg/L 180 77 µg/L 138 + 163 153 + 132 + 105 84 µg/L 180 84 µg/L

Reference Schell et al. (2008)

Knobeloch et al. (2009)

BF: breastfed as infants; NBF: not breastfed as infants a Unless otherwise indicated. Source: Adapted from IARC (2015)

(h) Concentrations of NDL-PCB metabolites in blood

Quinete et al. (2014) reviewed the biomonitoring literature from 1950 to 2013 reporting on concentrations of hydroxylated and methyl sulfone PCB metabolites, which are the major metabolites of PCBs, in human blood. The review describes the predominant hydroxy-PCB congeners in human blood as 4-hydroxy-PCB 187 and 4-hydroxy-PCB 107, followed by 4-hydroxy-PCB 146, 3-hydroxy-PCB 153 and 3-hydroxy-PCB 138; 4-hydroxy-PCB 172 and 3-hydroxy-PCB 180 are less frequently detected. Although these congeners have dominated in a number of studies, the one that occurs most commonly varies; this suggests that different PCB sources and metabolism capability may be responsible for the differences in metabolite fingerprints. Median concentrations of total hydroxy-PCBs varied between 0.0002 and 1.6 ng/g wet weight (ww) in serum or plasma. In comparison, higher concentrations of hydroxy-PCBs have been reported in blood from Canadian Inuit (up to 12 ng/g ww), in pregnant women from the Faroe Islands (median 5 ng/g ww), in women from Slovakia (median 2 ng/g ww) and in men from Latvia (median 5 ng/g ww). These levels are consistent with populations exposed to high levels of PCBs through their diet, which includes high consumption of fish and sea mammals (Canada and Faroe Islands), or living in PCB-polluted areas (Slovakia and Latvia). Overall, the median Σhydroxy99

Safety evaluation of certain food additives and contaminants Eightieth JECFA

PCBs/ΣPCBs ratio in human blood from different locations worldwide ranged from 0.06 to 0.51 (Quinete et al., 2014). The occurrence and levels of lower chlorinated hydroxy-PCBs (1–3 chlorine atoms) in human blood have been much less studied than those of the parent congeners. Quinete et al. (2014) mentioned only two reports on concentrations of lower chlorinated hydroxy-PCBs (trichlorinated and tetrachlorinated hydroxy-PCBs only) in human blood, the first on one male in Japan (Kunisue & Tanabe, 2009) and the other from a study in India (Eguchi et al., 2012). An earlier report from Weiss et al. (2006) studied 3-hydroxy-PCB concentrations in serum of middle-aged and elderly Swedish women who were expected to be relatively highly exposed; the summed median concentration was 1.7 ng/mL. 2.3.2 General considerations on health effects

Based on the available experimental information reviewed above, it seems likely that PCB congeners can cause serious toxicological outcomes, such as carcinogenicity, immunotoxicity, developmental toxicity and neurodevelopmental disorders. In contrast to DL-PCB congeners, for which there are sufficient data to perform a health risk assessment, such information is lacking for NDL-PCB congeners. Humans are always exposed to complex mixtures of individual DLPCB and NDL-PCB congeners, whose relative contribution to toxicity is unclear, and it has not yet been possible to ascribe the observed effects in epidemiological studies to any individual PCB or subgroup of NDL-PCB congeners. In 2005, EFSA stated that the available NDL-PCB data were inadequate and that more toxicological data were needed to identify any NDL-PCB hazard and to assess any risk involved with the consumption of NDL-PCBs (EFSA, 2005). The available experimental animal data on carcinogenicity and immunotoxicity were considered unsuitable for risk assessment, as few congeners were tested, the test compounds were not always examined for chemical purity and the mixtures used were not relevant for human exposure situations. The neurobehavioural data, although considered insufficient, were regarded as relevant, as behavioural effects observed in laboratory animals following perinatal PCB exposure had also been observed in humans linked to PCB exposure in epidemiological studies. EFSA (2005) cited BMD calculations made by others based on human studies on developmental neurotoxicity and immunotoxicity after perinatal exposure to total DL- and NDL-PCBs. EFSA (2005) noted that the lower 95% confidence limit (BMDL) of approximately 1 μg PCB per gram lipid is only about 4 times higher than the current median concentration in human milk, but that the existing epidemiological studies do not allow an estimation of the toxicity that may specifically be attributed to the NDL-PCBs. Moreover, EFSA (2005) identified

100

WHO Food Additives Series No. 71-S1, 2016

Supplement 1: Non-dioxin-like polychlorinated biphenyls

the developing fetus and the neonate as potential “at risk” populations because of increased susceptibility. Lack of congener-specific exposure and toxicity data has limited the ability to conclude which congeners are responsible for the observed effects. An overall perspective of the toxicology and epidemiology of PCBs was evaluated in the United States ATSDR toxicological profile on PCBs (ATSDR, 2000). Information on the health effects of PCBs is available from studies of people exposed in the workplace, by consumption of contaminated rice oil in Japan (the Yusho incident) and China (the Yucheng incident in the Province of Taiwan), by consumption of contaminated fish and via general environmental exposures, as well as by consumption of food products of animal origin. Health effects that have been associated with exposure to PCBs in humans include liver, thyroid, dermal and ocular effects, immunological alterations, neurodevelopmental changes, reduced birth weight, reproductive toxicity and cancer. Although PCBs may have contributed to adverse health effects in human populations, it cannot be determined with certainty which congeners may have caused the effects. The following health effects in humans were summarized by the original ATSDR report on PCBs (ATSDR, 2000). Concerning hepatic effects, the findings of human studies are not consistent, but there seems to be an increase in levels of some liver enzymes in the serum of people exposed to PCBs. As for endocrine effects, direct evidence linking PCB exposure to thyroid morbidity in humans is limited, although there is evidence that PCBs can produce both agonistic and antagonistic estrogenic responses. Skin irritation and ocular effects including hypersecretion and abnormal pigmentation have been observed following occupational exposure and accidental ingestion of rice oil contaminated with PCBs (and other halogenated chemicals, such as PCDFs). There seems to be an overall consistency in the data supporting sensitivity of the immune system to PCBs, particularly in infants exposed in utero or via breastfeeding, the most common effect reported being an increased susceptibility to respiratory tract infection. Neurological effects have been studied mainly in newborns and young children; abnormal reflexes, deficits in memory and poorer performance in tests of mental development index have been reported to be associated with PCB exposure. Regarding reproductive effects, miscarriages and shorter length of the menstrual cycle have been reported in occupationally exposed women. The ability of PCBs to cause reproductive effect in males is less clear. Studies on children of environmentally exposed women have produced mixed results for developmental effects such as anthropometric measures at birth and growth during infancy. Finally, carcinogenicity of PCBs in humans has been investigated in retrospective occupational studies in workers exposed during capacitor manufacture and repair and in case–control studies of the general population assessing PCB levels in serum or adipose tissue. Based on indications of PCBrelated cancer at several sites, particularly the liver, biliary tract, intestines and 101

Safety evaluation of certain food additives and contaminants Eightieth JECFA

skin (melanoma), the human studies provide suggestive evidence that PCBs are carcinogenic. An addendum to the ATSDR (2002) report provided further evidence of potential effects of PCBs on human health (ATSDR, 2011). An association has been reported between increased PCB 153 exposure and osteoporotic fractures, but there is no clear association with bone mineral density. An association of PCB exposure and the risk of type 2 diabetes has also been reported. A study of an occupational cohort involving exposed workers indicated that exposure to PCBs likely has an effect on neurodegenerative diseases. Prospective or case–control studies on children indicate that long-term exposure to PCBs can impair dental health and hearing. Increased exposure to PCBs during prenatal and postnatal life could also be associated with risk of cryptorchidism and low testosterone and estradiol levels in male children. Finally, the International Agency for Research on Cancer (IARC) recently reassessed the carcinogenicity of PCBs (Lauby-Secretan et al., 2013; IARC, 2015); human data included more than 70 independent epidemiological studies with informative data on the carcinogenicity of PCBs. Excess risk for melanoma was reported in several studies, mainly cohort studies of workers in the manufacture of capacitors and transformers and in electric power and equipment maintenance, with a significant linear exposure–response trend. An association between melanoma incidence and PCB exposure was noted consistently in occupational studies in different industries in North America and Europe, in studies of the general population and with cohort and case–control designs. Increased risks for non-Hodgkin lymphoma and breast cancer were also reported, both of which are biologically plausible. However, the associations were not consistent and were considered as providing limited evidence. Data for cancers at other sites were too sparse to allow any conclusions to be drawn. On the basis of sufficient evidence of carcinogenicity in humans and experimental animals, PCBs were classified as carcinogenic to humans (group 1). Additionally, DL-PCBs were also classified in group 1 on the basis of extensive evidence of an AhR-mediated mechanism of carcinogenesis that is identical to that of TCDD and sufficient evidence of carcinogenicity in experimental animals. However, according to the IARC evaluation, the carcinogenicity of PCBs cannot be solely attributed to the carcinogenicity of the DL-PCBs (Lauby-Secretan et al., 2013; IARC, 2015). As the most recent assessment of the overall health effects of NDL-PCBs was published by EFSA in 2005, we have considered in this review the studies published from 2005 onwards. Given the huge number of studies available and keeping in mind that our primary aim is to assess the health effects of NDL-PCBS, we will only briefly summarize the studies that did not provide specific results for NDL-PCBs, for either individual congeners or groups of NDL-PCBs. Moreover, for studies reporting results for NDL-PCBs, we will provide little detail of cross-

102

WHO Food Additives Series No. 71-S1, 2016

Supplement 1: Non-dioxin-like polychlorinated biphenyls

sectional studies, as they do not provide strong evidence of a causal relationship. Within the relevant studies, more attention will be paid to prospective cohorts than to case–control studies. Only studies with specific results for individual NDL-PCB congeners or groups of NDL-PCBs reporting a measurement of the level of exposure will be described in detail. 2.3.3 Mortality

Cooking oil contaminated by Kanechlor has been the source of two accidental mass poisonings, one in western Japan and the other in Taiwan, China. Commercial PCB mixtures were used as heat transfer media in oil tanks; leakage of the pipes caused exposure to the PCB mixture and PCB pyrolytic products, mainly PCDFs and polychlorinated quaterphenyls. In 1968, approximately 1800 people who ingested rice oil contaminated by Kanechlor 400 and its pyrolytic products, mainly in Fukuoka and Nagasaki prefectures in Japan, developed a strange skin disease, including acneform eruption, follicular accentuation and pigmentation, as well as eye discharge and swelling of eyelids. In 1979, about 10 years after the incident in western Japan, a similar food poisoning incident occurred in three counties (Taichung, Changhua and Miaoli) of central Taiwan, China. About 2000 residents from these counties had ingested rice oil contaminated with Kanechlor 500 and its pyrolytic products and showed clinical manifestations similar to those described for Japanese patients. These two accidental mass poisoning diseases were later called “Yusho”, oil disease in Japanese, and “Yucheng”, oil disease in Chinese (ATSDR, 2000; IARC, 2015). Mortality data among those registered from both groups of patients were identified, and analyses of these cohorts have been published with different follow-up periods. Many of these studies reported data according to different causes of death, mainly cancer, and some of them are referred to in the corresponding sections of this report. Recently updated analyses on overall mortality in the two cohorts have been published. After a 40-year follow-up, Yusho patients showed no significant difference in relative survival compared with the general Japanese population; overall 15-year relative survival was 0.99 (95% confidence interval [CI] 0.98–1.01) (Onozuka, Hirata & Furue, 2011). With the same data, net survival at 15 years was 97.4% (95% CI 94.2–99.6) in males and 100% (95% CI 100.0–100.0) in females (Onozuka, Hirata & Furue, 2014). As for Yucheng patients, after a 38-year follow-up, they had a standardized mortality rate (SMR) of 1.2 (95% CI 1.1–1.3) compared with their neighbourhood referents (Li et al., 2013). Apart from incidental contamination, chronic exposure to PCBs may occur through a diet rich in foods with a high content of PCBs; such exposure has been observed in northern Europe in populations with a high consumption of fish. No association with overall mortality was reported in studies in Swedish 103

Safety evaluation of certain food additives and contaminants Eightieth JECFA

WHO Food Additives Series No. 71-S1, 2016 104

populations from the east (Baltic) or west coast (Mikoczy & Rylander, 2009), the Baltic coast in Finland (Turunen et al., 2008) or the Great Lakes area of the USA (Wisconsin) (Tomasallo et al., 2010). Studies have been published on mortality among cohorts of workers exposed to PCBs in several industrial activities and occupations, mostly in North America and Europe, including capacitor manufacture, transformer manufacture and repair, and electric power–related occupations. Most of these studies based the exposure assessment on historical reconstructions of jobs and activities and were mainly designed to assess cancer mortality. Recently, an update on long-term employees of three electrical capacitor manufacturing plants in the USA showed that all-cause mortality was not elevated compared with national mortality rates, with an SMR of 0.99 (95% CI 0.97–1.01) (Ruder et al., 2014). In two Italian capacitor manufacturing plants, the SMR for all-cause mortality was 0.98 (95% CI 0.89–1.08) compared with mortality rates of the Lombardy region (Pesatori et al., 2013). Overall, there seems to be a small excess mortality of Yusho and Yucheng patients. However, it must be considered that patients from both countries had been exposed to PCDFs in addition to PCBs and that the mean total PCB and PCDF concentrations in the Yusho and Yucheng victims remained several times higher compared with the general population for many years after exposure (IARC, 2015). Compared with cohorts of Yusho or Yucheng patients, who consumed food contaminated with a high level of PCBs for a short period, potential exposure of the general population to PCBs through diet is a long-term, low-level exposure. It should also be borne in mind that fish or local vegetables contaminated by PCBs are often also contaminated by other compounds, such as 2,2-bis(p-chlorophenyl)-1,1,1-trichloroethane (DDT), PCDFs, PCDDs or heavy metals. In a similar way, workers from cohorts occupationally exposed to PCBs could have been often exposed to other chemicals, and therefore it is difficult to draw definitive conclusions on the causation of possible observed mortality excesses. Finally, in all these studies, PCB exposure involved exposure to complex mixtures of PCB congeners, often unknown, but most likely including both DLand NDL-PCB congeners. 2.3.4 Developmental toxicity: birth/gestational outcomes

This section covers birth outcomes related to fetal development, mostly determined by means of anthropometric measurements in newborns and infants, including childhood growth in some cases. Effects concerning specific functions, systems or organs are included in other sections. The biggest group of specific developmental effects concerns the nervous system (see section 2.3.5 below), but there are several studies reporting potential effects of PCBs on neonates

Supplement 1: Non-dioxin-like polychlorinated biphenyls

and infants regarding thyroid function and thyroid hormones, reproductive effects (mostly in males), including malformations and sex hormone levels, and immunotoxic effects, including response to vaccines, immunoglobulin levels and infections. (a) Studies on exposure to undefined PCB congeners

PCB concentrations in maternal serum in the Great Lakes sport-caught fish cohort (USA) were not associated with birth weight of newborns; no individual congeners measured were specified (Weisskopf et al., 2005). Serum specimens from pregnant women participating in the Child Health and Development Study in the San Francisco Bay area, California, USA, were collected in the second or third trimesters, and 11 PCB congeners were measured. In male infants, higher PCB exposure (sum of all congeners) was associated with reduced birth weight and head circumference, whereas in female infants, it was associated with shorter gestation (Hertz-Picciotto et al., 2005). In pregnant women enrolled in the United States Collaborative Perinatal Project in 12 centres, 11 PCB congeners were measured in blood collected during pregnancy (third trimester), at delivery and 5 weeks postpartum; higher levels of PCBs (sum of all congeners) were associated with higher (but non-significant) risk of preterm birth, whereas birth weight and length of gestation were unrelated to PCB levels in blood (Longnecker et al., 2005). In a cohort of infants born to mothers residing near a PCB-contaminated harbour in New Bedford, Massachusetts, USA, small negative (non-significant) associations were observed for PCB concentrations in cord blood (sum of PCBs 118, 138, 153 and 180) and birth weight, birth length and head circumference (Sagiv et al., 2007). In breast milk samples from mothers from Melbourne, Australia, no significant association was found between PCB concentration (no individual congeners specified) in breast milk and low birth weight, small for gestational age, and previous miscarriage or stillbirth; there was an elevated but non-significant association with prematurity (Khanjani & Sim, 2007). Among mothers from a multiethnic cohort in New York City, New York, USA, concentrations of PCBs (sum of PCBs 118, 128, 153 and 180) in maternal plasma from blood collected during pregnancy were not associated with lower birth weight, head circumference, birth length or gestational age (Wolff et al., 2007). Within the PCBRISK project in two regions from Slovakia, concentrations of PCBs (17 congeners) were measured in serum from blood collected from mothers at delivery and from children at 8–9 years of age. There was no significant association between PCB levels (sum of six congeners, including two DL-PCBs) and birth weight (Sonneborn et al., 2008); at 8–9 years of age, PCB exposure (no individual congeners specified) was significantly related to developmental enamel defects of permanent teeth, but no association with caries susceptibility, gingival 105

Safety evaluation of certain food additives and contaminants Eightieth JECFA

health or number of teeth was observed (Jan et al., 2007). In a prospective cohort, the Michigan Department of Community Health (USA), PCBs were measured in serum from maternal blood collected at enrolment (before delivery); no significant association was found between estimated PCB levels in maternal serum and gestational age or infant birth weight in adjusted models (Givens et al., 2007). In a selected group of women from the Danish National Birth Cohort, maternal PCB concentrations (seven congeners, including three DL-PCBs) were measured in blood collected during pregnancy; PCB concentrations (sum of seven congeners) in maternal plasma were inversely associated with birth weight and placental weight (Halldorsson et al., 2008). In a prospective birth cohort in a random sample of mother–infant pairs from Flanders, Belgium, PCB levels (PCBs 118, 138, 153, 170 and 180) were measured in cord blood, and children were followed up to 3 years of age; increasing PCB concentrations (sum of all congeners) were associated with lower birth weight, but higher body mass index (BMI) (measured as z-score) (Verhulst et al., 2009). In women aged 26–42 years who were resident in the province of Brescia, Italy, and undergoing a planned caesarean section at the University Hospital, concentrations of 30 PCBs were analysed in cord and maternal serum, placenta and maternal subcutaneous adipose tissue; no significant associations were reported between concentrations of PCBs (sum of all congeners) from any origin and preterm birth, small weight for gestational age or small length for gestational age (Bergonzi et al., 2011). Within the Child and Environment (INfancia y Medio Ambiente or INMA) birth cohort in Spain, levels of PCBs (PCBs 118, 138, 153 and 180) in serum were measured in women during pregnancy; rapid growth during the first 6 months was defined as a change in weight-for-age z-scores above 0.67, and elevated BMI at 14 months as a z-score at or above the 85th percentile. Neither the sum of the PCB congener concentrations nor individual PCB concentrations (data not shown) were associated with rapid growth (Mendez et al., 2011). Finally, the concentrations of all 209 PCB congeners were measured in the maternal and fetal blood of a small cohort of pregnant women in Fukuoka, Japan, as well as several indicators of placental function. Significant associations between PCB exposure (overall) and both placental growth factor and syncytiotrophoblast volume were identified (Tsuji et al., 2013). In all these studies, exposure was to complex mixtures of PCB congeners, often unknown, but in most instances including both DL- and NDL-PCB congeners. In some studies, it is likely that the highest contribution to the sum of PCB concentrations was from NDL-PCB congeners, but the mixtures also included PCB 118 or another DL-PCB congener. Therefore, they do not provide any relevant information regarding the potential effects of NDL-PCBs on developmental health. It should also be noted that there is remarkable inconsistency across many of these studies.

106

WHO Food Additives Series No. 71-S1, 2016

Supplement 1: Non-dioxin-like polychlorinated biphenyls

(b) Studies with specific results for NDL-PCBs

Within the Columbia-Presbyterian cohort of the National Collaborative Perinatal Project, in a subset of mother–child pairs (born 1959–1962) of African American mothers, concentrations of 24 PCB congeners were measured in maternal blood collected during the third trimester. Regression models for repeated measures (at ages 4, 7 and 17 years) were used to investigate associations between PCB concentrations and height and weight through 17 years. After adjusting for maternal pre-pregnancy weight, preterm status, and triglyceride and cholesterol levels in serum, there was a significant negative association between the sum of seven di-ortho-substituted congeners (PCBs 99, 101, 138, 146, 153, 170 and 180) and five tri-ortho-substituted congeners (PCBs 174, 183, 187, 199 and 203) and weight (not height) through to 17 years of age in girls; no significant associations were found for boys. For girls, for di-ortho-substituted PCBs, the β was −10.6 (95% CI −18.2 to −3.0); as PCB concentrations were expressed in natural log scale, this means that for a doubling of maternal di-ortho-substituted PCB concentrations, there was an average 7.35% decrease in weight (Lamb et al., 2006). Umbilical cord blood samples were collected from babies of mothers admitted to the National Hospital of Singapore for caesarean section during 2006; concentrations of 41 PCB congeners were measured and compared with several neonatal variables. Using partial least-squares regression, there was a significant inverse coefficient plot for the relationship between the concentration of PCB 153 in serum and Apgar score at 1 minute; there were also inverse (but nonsignificant) associations between the concentration of PCB 153 and birth weight, length at birth and head circumference. In this analysis, all the quantitative neonatal variables had significant inverse associations with the DL-PCB, PCB 118 (Tan et al., 2009). A study involving pregnant women (mean age 26 years) from Greenland (Inuit), Warsaw, Poland, and Kharkiv, Ukraine, was carried out in 2002–2004. The concentration of PCB 153 was analysed in maternal blood collected during pregnancy (at 24 or 33 weeks). There was no significant association between the concentration of PCB 153 and gestational age or risk of preterm birth. An increase in the PCB 153 concentration in the serum of Inuit mothers by one unit on the log scale was associated with a significant decrease in infant birth weight of 72.7 g (95% CI 27–118) and gestational age by 0.2 week (95% CI 0–0.4); decreases observed in the cohorts in Kharkiv and Warsaw were not statistically significant. These associations were adjusted for smoking status (Greenland), age of mother (Kharkiv and Warsaw) and mother’s BMI (Warsaw) (Wojtyniak et al., 2010). In the Child Development Study of the New York State Angler Cohort Study (New York, USA), blood samples were collected from women as they began trying 107

Safety evaluation of certain food additives and contaminants Eightieth JECFA

to become pregnant (preconception), after a positive pregnancy test (prenatal) and at about 6 weeks post-delivery (postnatal) in 1996–1999. Concentrations of 76 PCB congeners were measured in maternal serum. No association was found between birth weight and preconception or prenatal concentrations of a group of coeluting and single estrogenic NDL-PCBs (PCBs 4–10, 5–8, 15–17, 18, 31, 44, 47, 48, 52, 70, 99, 101, 136 and 153). However, a significant reduction in birth weight was observed to be associated with preconception concentrations of anti-estrogenic PCBs, which include the coeluting and single DL-PCBs (PCBs 77–110, 105, 114, 126, 156–171 and 169) (Murphy et al., 2010). Within the same study, concentrations of PCBs were also measured in breast milk and in child serum at age 2 years. A non-significant decrease in weight, length and head circumference at age 2 years was associated with PCB 153 (for both prenatal and breast milk concentrations), whereas increasing concentrations of PCB 77 (a DLPCB congener) were associated with a significant decrease in length z-score at 2 years (Jackson et al., 2010). Umbilical cord blood samples were collected from Guiyu and the control area of Chaonan, China, and concentrations of 28 PCB congeners were measured. Increased concentrations of PCB 153 were significantly associated with decreases in birth weight, Apgar score and gestational age, with correlation coefficients (P-value) of −0.30 (0.02), −0.20 (0.03) and −0.32 (0.01), respectively. There were also significant inverse correlations with concentrations of PCB 138 and PCB 180 (Wu et al., 2011). Within the Environmental Health Risks in European Birth Cohorts (ENRIECO) project, concentrations of PCB 153 were measured in blood collected during pregnancy, cord serum or plasma, or breast milk from about 7990 women enrolled in 15 study populations from 12 European birth cohorts from 1990 to 2008. To facilitate comparisons, the exposure was expressed as wet weight cord serum levels, which directly reflect fetal exposure at the time of delivery. Linear regression of birth weight on estimates of concentration of PCB 153 in cord serum was performed, adjusted for child’s gestational age and sex, mother’s region, maternal BMI, height, smoking status during pregnancy, socioeconomic status, mother’s age, parity and ethnicity, and time of sampling. Summary estimates were obtained by meta-analysis. Birth weight decreased with increasing concentration of PCB 153 in cord serum from 12 of 15 study populations. The meta-analysis including all cohorts indicated a birth weight decline of 150 g (95% CI 50–250) per 1 μg/L increase in PCB 153 concentration (Govarts et al., 2012). Gestational weight gain, which is associated negatively with PCB levels in maternal and cord blood and positively with birth weight, could substantially confound this association. In a reanalysis of previous data, a modified PCB 153 concentration was obtained by means of a pharmacokinetic model that simulates lifetime exposure and bioaccumulation in the mother and

108

WHO Food Additives Series No. 71-S1, 2016

Supplement 1: Non-dioxin-like polychlorinated biphenyls

the fetus. A 118 g decrease in birth weight (95% CI 106–129) for each 1 μg/L increase in simulated PCB 153 level in cord plasma was obtained; after further adjustment for simulated gestational weight gain, the estimated decrease in birth weight was reduced to 6 g (not significant) (Verner et al., 2013). Considering these results, authors of the meta-analysis used data from seven of the 12 cohorts that had actual (not simulated) data on gestational weight gain; moreover, using these data, they estimated change in fat mass. In the reanalysis of these seven cohorts, it was observed that the magnitude of decreased birth weight (per 1 μg/L increase in PCB 153 concentration) was reduced by 48% after including absolute gestational weight gain and by 31% after including change in fat mass, with a marginally significant P-value of 0.05 (Govarts et al., 2014). In a sample of 600 infants born between 1960 and 1963 and randomly selected from the Child Health and Development Study in the San Francisco Bay area, California, USA, concentrations of 11 PCB congeners were measured in serum collected within 3 days of delivery. No significant associations were observed between birth weight or length of gestation and concentrations of PCB 138, 153, 170 or 180 in serum, assessed by means of linear regression adjusted for maternal race, age, employment status, infant sex, laboratory and natural log-transformed triglyceride and cholesterol levels in serum. However, for the sum of the concentrations of di-ortho-substituted PCBs (PCBs 99, 138, 153, 170 and 180) in serum, the adjusted means of gestational age for the third and first tertiles were, respectively, 39.8 and 40.2 weeks (P-value < 0.05 for the difference of means). For the group of CYP1A and CYP2B enzyme inhibitors (PCBs 99, 153, 180 and 203), the corresponding adjusted means were 39.8 and 40.3 weeks (P-value < 0.05 for the difference of means) (Kezios et al., 2012). In the Persistent Organic Pollutants in Uppsala Primiparas (POPUP) study, first-time mothers living in Uppsala County, Sweden, in 1996–2010 agreed to donate breast milk sampled during the third week after delivery; the samples were analysed for PCB 138, PCB 153 and PCB 180 (Lignell et al., 2013). In the multivariate model (adjusted for sum of the concentrations of the polybrominated diphenyl ethers [PBDEs] BDE-47, BDE-99, BDE-100 and BDE-153, age of the mother, pre-pregnancy BMI, weight gain during pregnancy, maternal education, smoking and infant sex), birth weight was significantly increased (P-value = 0.02) by 137 g for each unit increase in the sum of the natural log concentrations of PCBs 138, 153 and 180; when the analysis was stratified by sex, the significant association was restricted to male infants. The association between dietary exposure to dioxins and PCBs during pregnancy and birth size was studied within the Norwegian Mother and Child Cohort Study, including more than 50 000 women recruited all over Norway from 1999 to 2008. The maternal dietary exposure was estimated using a validated food frequency questionnaire and available data on concentrations of 109

Safety evaluation of certain food additives and contaminants Eightieth JECFA

PCBs in Norwegian foods (Papadopoulou et al., 2013). An inverse dose–response association between dietary exposure to NDL-PCBs and fetal growth was obtained after adjustment for maternal age, energy intake, maternal education, pre-pregnancy BMI, parity, weight gain and smoking during pregnancy, gestational age and child sex. Comparing the highest with the lowest quartile of estimated NDL-PCB dietary exposure during pregnancy, there were significant reductions of 40.9 g (95% CI 30–52) in birth weight, 0.21 cm (95% CI 0.15–0.26) in birth length and 0.06 cm (95% CI 0.02–0.09) in head circumference. In this study, there were similar associations for estimated dietary exposure to dioxins and DL-PCBs. The parameter estimates cannot be directly related to levels of NDL-PCBs in serum (or other body tissues or fluids). In a cohort of newborns from 14 Inuit communities of Nunavik, Canada, PCB 153 concentration was measured in umbilical cord plasma and child blood. Mothers (1996–2000) were recruited at their first prenatal visit, and children were evaluated at birth and at age 8–14 years; weight, height and head circumference were measured at birth and during childhood. In a cross-sectional analysis, PCB 153 concentrations in cord blood were not associated with anthropometric measurements at birth or school age, but PCB 153 concentrations in child blood were associated with reduced weight, height and head circumference during childhood (mean age of 11 years) (Dallaire et al., 2014). 2.3.5 Neurotoxicity and behaviour (a) Neurodevelopmental effects (infants and children)

In pregnant women enrolled in the United States Collaborative Perinatal Project in 12 centres, concentrations of 11 PCB congeners (PCBs 28, 52, 74, 105, 118, 138, 153, 170, 180, 194 and 203) were measured in blood collected during pregnancy (third trimester, eighth week), and cognitive test (intelligence quotient [IQ]) scores on the Wechsler Intelligence Scale for Children were assessed at 7 years of age. Overall, there was a non-significant increase in IQ associated with an increase in PCB level (sum of concentrations of all 11 congeners) (Gray et al., 2005). Maternal blood was collected at week 31 of gestation in 1998–2002 from women from the Mount Sinai Children’s Environmental Health Center (New York City, New York, USA), and the Brazelton Neonatal Behavioral Assessment Scale was administered before hospital discharge; no adverse associations were found between PCB levels (sum of the concentrations of PCBs 118, 138, 153 and 180) and any behaviour (S.M. Engel et al., 2007). In a cohort of infants born to mothers residing near a PCB-contaminated harbour in New Bedford, Massachusetts, USA, concentrations of 51 PCB congeners were measured in cord serum samples. The sum of the concentrations of PCB congeners (PCBs 118, 138, 153 and 180) showed a significant inverse association with some

110

WHO Food Additives Series No. 71-S1, 2016

Supplement 1: Non-dioxin-like polychlorinated biphenyls

Brazelton Neonatal Behavioral Assessment Scale measures, such as the quality of alert responsiveness score (Sagiv et al., 2008). In the same study, the sum of the concentrations of the four PCB congeners (PCBs 118, 138, 153 and 180) in cord serum was associated with a higher risk of atypical behaviour corresponding to ADHD, measured with the Conners’ Rating Scale for Teachers (Sagiv et al., 2010). In these studies, PCB exposure involved complex mixtures of PCB congeners. In some studies, it is likely that the highest contribution to the sum of PCBs was from NDL-PCB congeners, but the mixtures included PCB 118 or another DLPCB congener. Therefore, these studies do not provide any relevant information regarding the potential neurodevelopmental effects of NDL-PCBs. (i) Case–control studies

The population-based case–control study of the Texas Neural Tube Defect (NTD) Project was carried out among pregnant Mexican American women who resided in the 14 Texas–Mexico border counties during 1995–2000. Cases (infants or fetuses) had diagnoses of anencephalus (ICD-9-CM code 740), spina bifida (ICD-9-CM code 741) or encephalocele (ICD-9-CM code 742.0); controls were randomly selected from study area women delivering normal live births during the same period. Concentrations of 25 PCB congeners were measured in serum samples collected on the first anniversary date of conception (plus or minus 1 month). No significant association was reported for the most common PCB congeners and the risk of NTD; the odds ratio (OR) for PCB 153 was 0.7 (95% CI 0.4–1.2) (Suarez et al., 2005). Several issues must be considered when interpreting the results of this study. First, the concentration of PCB 153 (18 ng/g lipid) was low compared with the range of 30–45 ng/g reported for 10 other study populations. Second, specimens were collected 1 year post-conception, which may or may not mirror the PCB burden during the relevant window of neural tube formation; postpartum events (e.g. breastfeeding) could potentially alter the PCB levels in serum. Of greater concern is the reduced participation in the blood collection, with a slight tendency for women of higher socioeconomic status to participate in the blood collection for PCB analysis compared with those enrolled in the original case–control study. (ii) Cohort studies

The Oswego, New York, USA, study tracks a cohort of children who were born between 1991 and 1994. Immediately after birth, a sample of umbilical cord blood and placental tissue was obtained. Concentrations of 68 PCB congeners were measured in cord blood, although a sum of the concentrations of the most persistent and higher chlorinated congeners (heptachlorinated, octachlorinated and nonachlorinated homologues: PCBs 170 + 190, 172, 174, 177, 179, 180, 183, 111

Safety evaluation of certain food additives and contaminants Eightieth JECFA

185, 187 + 181, 194, 195, 199, 203 + 196, 206) was used as a measure of prenatal PCB exposure in cord blood; all of them are NDL-PCBs. Moreover, concentrations of PCB 153, PCB 138 and PCB 180 were measured in placental tissue. In this cohort, several tests to assess behavioural processes and cognitive development were performed in children; the Neurobehavioural Evaluation System 2 (NES2) Continuous Performance Test (CPT) was applied at 8 years, and the Extended Continuous Performance Test (E-CPT) at 9.5 years. After taking into account many measured covariables, including maternal IQ, maternal sustained attention and maternal response inhibition, results revealed associations between PCB level in cord blood and impulsive responding at both testing ages. At 8 years of age, there was a dose–response relationship between total commission errors on the NES2 CPT and PCB levels (P-value = 0.026). E-CPT testing at 9.5 years clearly indicated that the PCB-related impulsive responding was due to impaired response inhibition (the association with commission errors was significant only at higher levels [70–90%] of probability of targets appearing) (Stewart et al., 2005). To assess inappropriate responding on intermittent reinforcement schedules, the Differential Reinforcement of Low Rates task protocol was used to evaluate the inter-response times. The sum of the concentrations of higher chlorinated PCBs was associated with shorter inter-response times; the covariate-adjusted standardized β was −0.170 (P-value = 0.03) (Stewart et al., 2006). These results were significant taking into account multiple potential confounders, including several non-PCB contaminants: prenatal methylmercury, 1,1-dichloro-2,2-bis(pchlorophenyl) ethylene (DDE), hexachlorobenzene (HCB), and prenatal and postnatal lead. In the same study, the IQ score was measured at 9 years by means of the Wechsler Intelligence Scale for Children and was compared with PCB levels in placental tissue. In the analysis using a maximal covariate set, there were inverse (but non-significant) associations between the full-scale IQ and placental concentrations of PCBs 138, 153 and 180. However, a 1 ng/g increase in PCB 153 concentration was associated with a decrease of 0.234 standard units of the verbal IQ (P-value = 0.006) and with 0.253 standard units of the freedom from distractibility score (P-value = 0.008) (Stewart et al., 2008). In the Cord Blood Monitoring Program, which took place in Inuit communities of Nunavik in northern Quebec, Canada, concentrations of 14 PCB congeners were measured in umbilical cord plasma of newborns between 1993 and 1996, as well as at the time of the neurological examination at age 4–6 years (mean 5.4). Gross motor functions as well as fine neuromotor performance were assessed using quantitative measures of postural hand tremor, reaction time, sway oscillations, and alternating and pointing movements. Negative effects of PCB concentrations in cord blood on neuromotor development were not clearly observed. However, PCB 153 concentration in blood at testing time (crosssectional relationship) was associated with transversal sway in balance conditions;

112

WHO Food Additives Series No. 71-S1, 2016

Supplement 1: Non-dioxin-like polychlorinated biphenyls

adjusted for several maternal and child covariates, the Pearson correlation coefficient was 0.15, and the standardized β was 0.22 (both with P-value < 0.05) for the relationship between the log-transformed concentration of PCB 153 and 1 standard deviation of the transversal sway (in millimetres) (Després et al., 2005). In the same children, pattern reversal visual evoked potentials (VEPs) were also assessed; latency and amplitude of VEPs are useful to detect subclinical effects on visual processing. The following standard VEP components were examined: N75 (negative deflection at 75 milliseconds), P100 (positive deflection at 100 milliseconds) and N150 (negative deflection at 150 milliseconds); for each component, the latency was determined from the stimulus onset to the maximal waveform peak, whereas the amplitude was calculated from peak-to-peak procedure (N75-to-P100 and P100-to-N150). When PCB 153 concentrations in cord blood and at the testing time were included in a multivariate model, there was no association between VEPs and PCB 153 levels in cord blood. After adjustment for maternal alcohol and marijuana consumption, maternal nonverbal reasoning abilities, haemoglobin concentrations at testing time, highest grade completed by the primary caregiver, number of children and adults at home, sex, parity and height at birth, PCB 153 concentration in plasma at testing time was significantly related to longer P100 latency at 95% and 12% contrast and N150 latency at 12% contrast, with βs of 2.5 (P-value < 0.01), 3.22 (P-value < 0.05) and 5.58 (P-value < 0.05), respectively. The amplitudes of the N75-to-P100 and the P100-to-N150 at 95% contrast were significantly reduced as a function of increased PCB 153 concentration in the plasma of the children, with respective βs of −3.74 and −3.17 (both with P-value < 0.05) (Saint-Amour et al., 2006). In a cohort of healthy mother–infant pairs recruited between 1993 and 1995 from the obstetrical wards of three hospitals in Düsseldorf, Germany, concentrations of PCBs 138, 153 and 180 were measured in cord blood, early breast milk (around 2 weeks) and 42-month serum. The Kaufman Assessment Battery for Children (K-ABC) was administered to children at 42 and 72 months (6 years) of age. This test yields a Mental Processing Composite, which is comparable to a conventional IQ (mean 100, standard deviation 15). The effect size on the test was assessed by comparing the scores corresponding to the sum of the concentrations of PCBs 138, 153 and 180 of the 95th and the 5th percentiles. For PCBs in milk, the effect size was −8.5 (P-value = 0.028) for test score administered at 42 months; there was a similar effect size (−10.3, P-value = 0.025) for the PCBs in serum collected at 42 months. For PCBs in both milk and serum, the effect sizes on the test at 72 months were also negative, but no longer significant. These effects were adjusted for sex, parental education, maternal IQ, birth weight, breastfeeding (duration) and HOME score (an indicator of quality of the home environment) (Winneke et al., 2005). Mental development and motor development were assessed by means of the Bayley Scales of Infant 113

Safety evaluation of certain food additives and contaminants Eightieth JECFA

Development (BSID) at the ages of 12 and 24 months in the same study, as well as among a subset of the Duisburg, Germany, cohort, including healthy mother– infant pairs recruited in Duisburg in which concentrations of 18 PCB congeners were measured in maternal blood during pregnancy and in maternal milk. The test was applied at ages 10 and 30 months (Düsseldorf) or at ages 12 and 24 months (Duisburg). The PCB concentration (sum of PCBs 138, 153 and 180) showed a negative relationship with both mental and motor development; the associations were non-significant. The more significant result was for mental development at 30 months in the Düsseldorf cohort: the per cent change in the BSID score (mental development) was −3.3 (95% CI −9.3 to 0.3) for doubling the PCB concentration, adjusted for parental education and occupation, alcohol and smoking during pregnancy, duration of pregnancy, mother’s age at delivery, thyroid disease during pregnancy, medication during pregnancy, Apgar score at 5 minutes, neonatal jaundice, duration of breastfeeding, child’s home environment (HOME score) and lead concentrations in mother’s blood (Wilhelm et al., 2008). Mother–infant pairs for a birth cohort were recruited in 2002–2004 from two districts in eastern Slovakia (Michalovce, with high PCB contamination from a chemical manufacturing plant, and Svidnik, with lower levels of PCBs). The concentrations of 15 PCB congeners were determined in the maternal and cord serum samples collected at delivery. The BSID-II was administered to children at 16 months of age. Two scales of the BSID, the mental development index (MDI) and the psychomotor development index (PDI), were used for the constructs of cognitive and motor development, respectively. Higher concentrations of NDL-PCBs (PCBs 138, 153, 170 and 180) in either maternal or cord serum were associated with lower levels of both MDI and PDI scores, and this association was significant for PCB concentrations in cord serum in relation to PDI. In multiple linear regression models adjusting for district, HOME score, sex and Raven score (a non-verbal intelligence test applied to the mothers after delivery), the β (score change in PDI) was −2.14 (P-value < 0.05) for a one-unit change in natural log concentration of PCB 153 and −1.95 (P-value < 0.05) for a one-unit change in natural log sum of the concentrations of PCB congeners 138, 153, 170 and 180. In this study, a strong inverse relationship was found between concentrations of DL-PCBs (in maternal and cord serum) and both PDI and MDI scores (Park et al., 2010). In the same subjects, hydroxylated PCBs were measured; some studies have shown that concentrations in cord plasma are approximately 50% of maternal levels, whereas they are only 30% of the parent PCB congeners, suggesting an active transport of hydroxy-PCBs across the placenta. In the same analysis conducted previously, no significant association was found for 3-hydroxy-PCB 153 in maternal or cord serum and either MDI or PDI. However, the 4-hydroxy-PCB 107 concentration in cord serum was significantly associated with lower MDI (β −2.27, P-value = 0.01) and PDI (β −4.50, P-value = 0.004);

114

WHO Food Additives Series No. 71-S1, 2016

Supplement 1: Non-dioxin-like polychlorinated biphenyls

also, the 4-hydroxy-PCB 107 concentration in maternal serum was significantly associated with lower MDI (β −1.76, P-value = 0.03) but not PDI (Park et al., 2009). The Tohoku study of child development included mother–neonate pairs from an urban coastal area of the Tohoku district in Japan that were recruited during 2001–2003. Concentrations of all 209 PCB congeners were measured in cord blood. The Neonatal Behavioral Assessment Scale was administered 3 days after birth, and the K-ABC, composed of four scales used to assess their intelligence and achievement, was applied at the age of 42 months. The motor cluster of the Neonatal Behavioral Assessment Scale was negatively associated with PCB concentration (sum of all congeners) in cord serum; therefore, this result does not provide any relevant information regarding the potential effects of NDL-PCBs (Suzuki et al., 2010). In the same subjects, the K-ABC results were compared with concentrations of PCB congeners in cord blood grouped according to their degree of chlorination. The multiple regression analysis showed that the sum of the log concentrations of nine higher chlorinated PCBs was associated with a decrease of 8.172 (95% CI 1.108–15.24) in the sequential processing score and 6.793 (95% CI 0.771–12.81) in the mental processing score, after adjustment for sex, concentrations of mercury and lead in cord serum, birth order, child age at examination, tester of the K-ABC, analytical institute, drinking and smoking habits during pregnancy, the Raven scores, annual family income and duration of breastfeeding (Tatsuta et al., 2014). In a population-based birth cohort established in three Spanish regions (INMA study), maternal blood was obtained between the seventh and 26th weeks of pregnancy, and concentrations of PCBs 28, 52, 101, 118, 138, 153 and 180 were measured in serum. Neuropsychological assessment by means of the BSID was carried out at age 14 months (range 11–21 months) (n = 1391). Prenatal exposure to PCBs, particularly to PCBs 138 and 153, tended to show an impairment of psychomotor development, although the results were not significant. After adjustment for region of study, gestational age and paternal social class, decreases in psychomotor test scores of 1 point (P-value = 0.059) and 0.99 point (P-value = 0.072) were associated with 10-fold increases in concentrations of PCB 138 and PCB 153 in serum, respectively (Forns et al., 2012a). In a reanalysis of these subjects restricted to 1175 children, increasing prenatal PCB 153 concentrations were associated with lower mental and psychomotor scores, although significance was reached only for psychomotor development (β −1.36, 95% CI −2.61 to −0.11), adjusted for region of study, gestational age and paternal social class (Gascon et al., 2013). The exclusions in the second analysis were justified because the authors aimed to estimate postnatal exposure via breast milk by means of a physiologically based kinetic (PBK) model using data on breastfeeding as well as height and weight profiles through the first year of life; children for which this information 115

Safety evaluation of certain food additives and contaminants Eightieth JECFA

was not available (n = 216) were not included. In contrast, the McCarthy Scales of Children’s Abilities (MSCA) were applied at age 4 years to the subset of children from the study residents in Menorca with PCB concentrations measured in cord serum and in blood collected at 4 years. PCB 153 concentrations in cord serum were negatively associated with several MSCA scores: the β (change in the score for 1 unit in the log concentration [ng/mL] of PCB 153) for the general cognitive score was −3.23 (P-value = 0.01); and the βs for specific MSCA scores were −2.71 for verbal, −2.75 for quantitative, −3.02 for perceptual, −2.82 for memory and −2.26 for motor – all of them except the last one were significant at a P-value of <0.05. These associations were adjusted for psychologist, child’s age, maternal social class, folic acid supplementation during pregnancy, maternal cigarettes during pregnancy, paternal education, child sex and duration of breastfeeding. The association between PCB 153 concentration and general cognitive score remained significant (β −3.45, P-value = 0.017) after further adjustment for concentrations of DDE and HCB. The associations for PCBs 138 and 180 were less precise, whereas no significant associations were reported for the DL-PCB congener, PCB 118 (Forns et al., 2012b). Pregnant women from the Hokkaido Study on Environment and Children’s Health were recruited in 2002–2004 from the Sapporo Toho Hospital in Hokkaido, Japan. Concentrations of a total of 68 PCB congeners were measured in blood collected during pregnancy (third trimester). Mental development and motor development were assessed by means of the BSID-II administered to children at 6 months of age. After adjustment for gestational age (days), smoking during pregnancy, caffeine intake during pregnancy (mg/day) and blood sampling time, there were no significant associations between the BSID-II scores (MDI, PDI) and the two NDL-PCBs (PCBs 170 and 180) or the DL-PCBs, although significant negative associations were reported for some PCDD or PCDF isomers (Nakajima et al., 2006). In a Faroe Islands birth cohort established in 1994–1995, prenatal exposure was determined from maternal concentrations of PCBs in pregnancy serum (blood obtained at gestational week 34) and milk. At 7 years of age, a new blood sample was taken and used for a new assessment of PCB concentrations as well as to analyse MRs in lymphocytes and the enzyme activity of monoamine oxidase-B in platelets as potential markers of neural cell function. No correlation was found between the two biomarkers and concentrations of PCBs 28, 77, 105, 115 and 153 and 4-hydroxy-PCB 107 or the sum of the concentrations of PCBs 138, 153 and 180 (results not shown) (Coccini et al., 2009). Within the New York State Angler Cohort Study, a population-based cohort from 16 counties along Lakes Erie and Ontario in New York State, USA, established in 1991, concentrations of 62 single eluting and 12 co-eluting PCB congeners in maternal serum collected during pregnancy (around 8 weeks of

116

WHO Food Additives Series No. 71-S1, 2016

Supplement 1: Non-dioxin-like polychlorinated biphenyls

gestation) were measured, and postnatal exposure was assessed by quantifying all 209 PCBs in breast milk. At a 24-month home visit, the child psychologist administered the BSID-II to assess cognitive and psychomotor development. When examining the combined effect of prenatal and postnatal exposure to PCB 153 on BSID-II scores, there was the suggestion of an increase in MDI associated with higher levels of prenatal exposure, although the effect was not statistically significant after adjustment for maternal IQ and maternal prenatal lead level. Regarding PDI, higher levels of postnatal (breast milk) exposure were associated with a decrease in PDI; children in the third tertile of PCB 153 exposure had a decrease of 24.9 points (95% CI −44.3 to −5.5) compared with the PDI score of children in the lowest tertile (Lynch et al., 2012). Within the first Flemish Environment and Health Study, mothers and their newborns were recruited through 25 maternity hospitals from Flanders, Belgium, in 2002–2003. Concentrations of PCBs 118, 138, 153, 170 and 180 were measured in cord blood. Children were recontacted when they were 7–8 years old, when behavioural problems were assessed using the standardized Strengths and Difficulties Questionnaire, a validated screening questionnaire, which comprises five domains: emotional, conduct, hyperactivity, peer and social problems. No significant associations were found between exposure to NDL-PCBs (sum of concentrations of PCBs 138, 153 and 180 in log scale) and emotional symptoms (score ≥5), conduct problems (score ≥5), hyperactivity (score ≥7) and total difficulties (score ≥16). In the same analysis, no associations were found for dioxin-like compounds (total of PCDDs/PCDFs and DL-PCBs) (Sioen et al., 2013). Mother–infant pairs from the northern part of the Netherlands were included in an observational study between 1998 and 2000; only women whose pregnancy or delivery had not involved serious illness or complications and only infants born at term who had no congenital anomalies or diseases were included. Umbilical cord blood samples were taken immediately after delivery; concentrations of 10 PCBs and six hydroxy-PCBs were measured. When the infants were 3 months old, their motor development was evaluated by assessing the presence and performance of spontaneous movement patterns from video recordings. The Motor Optimality Score was calculated by means of a Motor Optimality List consisting of five categories; the score could range from low (5) to high (28) optimality. Although the levels of several PCB congeners were positively associated with the Motor Optimality Score or specific motor optimality categories, after adjusting for sex, gestational age and maternal smoking during pregnancy, the only significant associations were for PCB 118 (a DL-PCB congener), with an OR of 1.84 (95% CI 1.02–3.32) for the absence of antigravity movements for a 10 ng/g lipid increase, and for 4′-hydroxy-PCB 172, with an OR of 0.19 (95% CI 0.05–0.79) for more manipulation (Berghuis et al., 2013). 117

Safety evaluation of certain food additives and contaminants Eightieth JECFA

Within the birth cohort Health Outcomes and Measures of the Environment Study, pregnant women from seven prenatal clinics in the Cincinnati area, Ohio, USA, were recruited in 2003–2006. Women provided a serum sample around 16 and 26 weeks of pregnancy and within 24 hours of delivery, in which concentrations of 27 PCB congeners were measured. When children were 4 and 5 years old, mothers completed the Social Responsiveness Scale (SRS), a measure of autistic behaviours. The SRS yields a T-score (mean 50, standard deviation 10); higher scores indicate more autistic behaviours, with T-scores of 60 and higher considered to be indicative of clinically significant deficiencies in reciprocal social behaviour and T-scores of 75 and higher being consistent with a clinical diagnosis of autism spectrum disorders (Braun et al., 2014). After adjustment for demographic factors, as well as depressive symptoms during pregnancy, HOME score, gestational serum cotinine concentration and concentrations of other endocrine-disrupting chemicals in a semi-Bayesian hierarchical regression model, most PCB concentrations were associated with negligible changes in SRS scores, with the absolute change being 1.5 points or less. The largest effects were for PCBs 153, 194 and 178, although only the effect for the last congener was statistically significant. The T-score decreased by 1.5 points (β −1.5, 95% CI −4.1 to 1.2) for an increment of 2 standard deviations in the log concentration of PCB 153, whereas the change in SRS scores among children born to women with detectable versus non-detectable concentrations of PCB 178 was −2.6 (95% CI −4.4 to −0.2). (b) Neurophysiological and neuropsychological effects in adolescents

In Akwesasne Mohawk adolescents (age 10–16 years, average 13 years) resident along the St Lawrence River near the junction of New York State, USA, and Ontario and Quebec, Canada, overall exposure to PCBs (a sum of the concentrations of the 16 PCB congeners detected as individuals, pairs or triplets in 50% or more of the serum samples) showed a significant negative relationship with two separate measures of long-term memory, as well as with a measure of comprehension and knowledge (Newman et al., 2006). In the same population, no evidence of negative effects of PCB levels in blood of adolescents (a summary measure of 10 PCBs, including PCB 118, a DL-PCB congener) on ADHD-like behaviour was found (Newman et al., 2006). In these studies, PCB exposure involves exposure to mixtures of PCB congeners; even though in the last study it is likely that the highest contribution to the sum of PCBs was from NDL-PCB congeners, the study included one DL-PCB congener, and therefore these studies do not provide relevant information regarding the potential neurotoxicity of NDL-PCBs in adolescents. However, in the same adolescents, cognitive function was found to be associated with a group of NDL-PCBs (PCBs 52, 70, 84, 74, 87,

118

WHO Food Additives Series No. 71-S1, 2016

Supplement 1: Non-dioxin-like polychlorinated biphenyls

95, 99, 101 [+ 90], 110, 138 [+ 163 + 164], 153, 180 and 187) in a cross-sectional analysis (Newman et al., 2009). Two measures of long-term memory, Delayed Recall Index and Long Term Retrieval, were inversely associated with NDL-PCB concentrations in multivariate linear regression analysis. The Delayed Recall Index is measured by the Test of Memory and Learning, a measure of general and specific memory functioning, whereas the Long Term Retrieval is one of the domains of cognition measured by the Woodcock-Johnson Revised Tests. The standardized β for the Delayed Recall Index for 1 standard deviation of the sum of serum concentrations of NDL-PCBs was −0.21 (P-value < 0.05); for the Long Term Retrieval, it was −0.25 (P-value < 0.05). Similar effects were also reported for the group of DL-PCBs. In 12-year-old children from three districts in eastern Slovakia (Michalovce, Svidnik and Bratislava), concentrations of 15 PCBs (PCBs 28, 52, 101, 105, 114, 118, 123, 138, 153, 156, 157, 167, 170, 180 and 189) were measured in serum. As a marker of cochlear status, transient evoked and distortion product otoacoustic emissions were measured, and their cross-sectional association was analysed in relation to PCB concentrations (Trnovec et al., 2010). In multivariate regression models in which otoacoustic emission measures were modelled as a function of log PCB concentrations, with adjustment for sex, age and site of examination, NDL-PCB concentrations (sum of PCB congeners 138, 153, 170 and 180) were significantly associated with lower transient evoked otoacoustic emissions at 1000 Hz (β −3.47, P-value = 0.007) and 1500 Hz (β −2.01, P-value = 0.051) and with lower distortion product otoacoustic emissions at 1000 Hz (β −3.31, P-value = 0.036) in the left ear only. In this study, similar effects were also reported for the group of DL-PCBs. (c) Neurophysiological and neuropsychological effects in adults

No overall excess of Parkinson disease, amyotrophic lateral sclerosis or dementia was found among workers exposed to PCBs in three electrical capacitor manufacturing plants in the USA; however, sex-specific analyses revealed that women had an excess of amyotrophic lateral sclerosis and that highly exposed women had an excess of Parkinson disease and dementia (Steenland et al., 2006). In elderly (≥60 years) Yucheng patients (Taiwan, China), exposed women had reduced functioning in attention and digit span, visual memory span and verbal memory recalls, whereas all test results were similar to those of the reference group in exposed men (Lin et al., 2008); furthermore, there was no statistically significant difference in the Mini-Mental State Examination, but attention and digit span forward and total scores showed a significant decline in the exposed subjects (Lin et al., 2010). A neuropsychological evaluation assessed by a battery of 18 tests was carried out in Akwesasne Mohawk adults aged 18–79 years who 119

Safety evaluation of certain food additives and contaminants Eightieth JECFA

were residents along the St Lawrence River near the junction of New York State, USA, and Ontario and Quebec, Canada; overall exposure to PCBs (sum of the concentrations of 101 congeners measured in serum) was significantly related to outcome variables in the domains of executive functioning, motor functioning and memory (Haase et al., 2009). In men and women aged 50–89 years who were former capacitor workers from the General Electric Corporation at either of its two factories in New York, USA, the concentration of PCBs (as a sum of 30 congeners) in serum was not significantly correlated with deficits in neurocognitive function (Seegal et al., 2013). In all these studies, PCB exposure involved complex mixtures, and no results were provided for individual congeners; therefore, these studies do not provide any relevant information regarding the potential neurotoxicity of NDL-PCBs in adults. Potential neurotoxic effects of NDL-PCBs have been assessed in several cross-sectional studies. The neuropsychological status of older adults (aged 55– 74 years) living along contaminated portions of the upper Hudson River (New York, USA) was evaluated by means of a battery of 34 tests capable of detecting subtle deficits in cognition, motor function, affective state and olfactory function. Concentrations of 30 PCB congeners were measured in serum samples. After adjustment for potential confounders, the serum concentrations of PCBs 138, 170 and 180 (but not PCB 153) were significantly associated with a decrease in verbal learning, as measured by a California Verbal Learning Test trial, and the four congeners (i.e. including PCB 153) were associated with an increase in depressive symptoms as measured by the Beck Depression Inventory. The β (change in 1 score of the test for 1 unit log concentration of PCB) for PCB 153 was −0.338 (P-value = 0.14) for the California Verbal Learning Test and 0.975 (P-value = 0.01) for the Beck Depression Inventory (Fitzgerald et al., 2008). Cross-sectional analyses have been carried out in older adults (≥60 years) from NHANES 1999–2004 in relation to 11 PCB congeners. Using audiometry measures, hearing impairment was defined as a pure-tone average of the thresholds at 0.5, 1, 2 and 4 kHz of greater than 25 dB hearing level in the better ear, consistent with the WHO definition. There was an increased risk of hearing impairment associated with higher levels of PCB congeners 138, 153, 170 and 180 in serum, but the association was significant only for PCBs 170 and 180. The ORs for the subjects with the highest quintile of exposure to PCBs compared with subjects with PCB values below the detection level were 5.8 (95% CI 1.2–28.2) and 7.8 (95% CI 1.8–33.4) for PCB 170 and PCB 180, respectively, adjusted for age, sex, ethnicity, education, cigarette smoking, occupational/loud/ firearm noise exposure, ototoxic medication, history of diabetes/hypertension/ thyroid disease, BMI, triglyceride level and total cholesterol level. In the same study, no association was observed for the only DL-PCB congener analysed, PCB 118 (Min, Kim & Min, 2014). In an analysis restricted to older adults (aged 60–

120

WHO Food Additives Series No. 71-S1, 2016

Supplement 1: Non-dioxin-like polychlorinated biphenyls

84 years) from the NHANES 1999–2002, the Wechsler Adult Intelligence Scale Digit-Symbol Coding Test, used to detect brain impairment in a clinical setting, was not associated with the concentrations of NDL-PCBs (sum of PCBs 99, 138, 146, 153, 170, 180 and 187) in serum; in the same study, restricted to subjects aged 70–84 years, a significant inverse association was reported for the sum of the concentrations of five DL-PCB congeners in serum (Bouchard et al., 2014). 2.3.6 Cancer

IARC recently assessed the carcinogenicity of PCBs (Lauby-Secretan et al., 2013; IARC, 2015). Although the epidemiological studies included in the review date back more than 20 years, in order to be consistent with other sections of this monograph, we will review here the studies published from 2005 onwards. Only those reporting specific results for NDL-PCBs will be reviewed in detail. Studies of cancer mortality and incidence have been conducted among workers exposed to PCBs used mainly as dielectric fluids in the manufacture of capacitors (Prince et al., 2006a,b; Ruder et al., 2006; Silver et al., 2009; Hopf et al., 2010; Pesatori et al., 2013) or in transformer manufacture and repair (Caironi et al., 2005). In these studies, significant excess risk of cancer death (SMR) or standardized incidence rate (SIR) has been reported for lymphomas and cancers of the digestive tract (Pesatori et al., 2013), melanoma (Ruder et al., 2006), all cancers and stomach cancer (Prince et al., 2006b; Hopf et al., 2010) and all cancers (Silver et al., 2009). Mortality data among those exposed to two accidental mass poisonings through ingestion of rice oil contaminated with PCBs and PCDFs in Japan and Taiwan, China (Yusho and Yucheng patients, respectively), have been published, with different follow-up periods. Regarding the Yusho patients, after a 40-year follow-up, there was a significant excess mortality from all cancers and from cancers of the lung and liver (Onozuka et al., 2009; Yoshimura, 2012). In an analysis restricted to the most severely affected area, where SMRs were estimated using regional instead of national death rates, the excess risks for cancer (all, lung, liver) were not significant (Kashima, Yorifuji & Tsuda, 2011). As for Yucheng patients, the analysis follow-up to 2003 revealed that mortality from cancers overall and at several sites was similar to that of the national population (Tsai et al., 2007). However, in a second analysis with extended follow-up (1980–2008), a significantly increased mortality was reported for cancer of the stomach and neoplasms of lymphatic and haematopoietic tissue among men (Li et al., 2013). Apart from incidental food contamination, chronic exposure to PCBs may occur through a diet rich in foods with a high content of PCBs, mostly in populations with a high consumption of fish. In Swedish populations from the east (Baltic) or west coast, no association with overall mortality was found, but 121

Safety evaluation of certain food additives and contaminants Eightieth JECFA

WHO Food Additives Series No. 71-S1, 2016 122

there was a significant increase in the incidence of skin cancer on the west coast (Mikoczy & Rylander, 2009). No significant associations with cancer mortality were reported in residents on the Baltic coast in Finland (Turunen et al., 2008). Cancer mortality rates did not differ from those of the general population for those consuming fish from the Great Lakes area in the USA (Wisconsin) (Tomasallo et al., 2010). Another source of exposure to PCBs is environmental exposure. In residents in an industrialized area from Sweden, there was a significant increase in excess mortality from lymphoma following an accidental spill of oil contaminated with PCBs (Helmfrid et al., 2012). Ten years after the release of PCB contamination in the favoured fishing grounds of a village from Guam, an increase in the proportional cancer mortality rate was observed among residents in the contaminated area compared with other villages (Haddock, Badowski & Bordallo, 2011). The mortality in an area near Rome, Italy, heavily polluted over the years by industrial wastes, including high levels of PCBs, was compared with that of the general population in the Lazio region. There was a significant increase in the risk of mortality by cancer (all neoplasm), as well as for cancers of the larynx, lung and pleura, among men, but no significant associations were reported among women (Fantini et al., 2012). Occupational studies mainly assessed exposure to PCB mixtures through job–exposure matrices and historical measurements, but most did not have data on non-occupational risk factors, which are important for many cancer sites. Moreover, workers from cohorts occupationally exposed to PCBs could have been often exposed to other chemicals. Regarding Yusho and Yucheng patients, it must be taken into account that they were exposed to PCDFs in addition to PCBs; the mean total PCB and PCDF concentrations in the Yusho and Yucheng victims remained several times higher than in the general population many years after exposure. Compared with occupational or accidental exposure, potential exposure to PCBs through diet is a long-term, low-level exposure; however, fish or local vegetables contaminated by PCBs are often also contaminated by other compounds, such as pesticides, furans, dioxins and heavy metals. Overall, the studies reported above may have some value for the assessment of the potential effects of PCBs, but they do not provide any relevant information regarding the carcinogenic effects of NDL-PCBs. (a) Prospective studies: nested case–control studies of PCBs in blood or adipose tissue

Several cohort studies have analysed the potential relationship between risk of cancer and internal measurements of exposure to PCBs. The most commonly used marker of past PCB exposure is the serum or plasma concentration of a set of PCB congeners, although a few studies have measured PCB concentrations

Supplement 1: Non-dioxin-like polychlorinated biphenyls

in adipose tissue. Most studies have used a case–control design nested within a cohort, whereas a few studies have used a case–cohort approach, in which the referent group is formed by a random sample of the whole cohort selected at baseline. PCB 118, PCB 138, PCB 153 and PCB 180 are the congeners most often reported, because they are frequently analysed and prevalent in human biological samples. Usually a summary estimate of the sum of all measured PCB concentrations is used as the main indicator of exposure, but results for individual congeners or groups of congeners are provided in some instances. (i) Breast cancer

Between 1993 and 1997, 29 875 Danish women aged 50–64 years were enrolled in a prospective study of diet and cancer and followed until 2000 through linkage with the Danish Cancer Registry. During this period, 409 women were diagnosed with postmenopausal breast cancer and matched to one control by age, postmenopausal status and use of hormone replacement therapy. Concentrations of 18 PCB congeners were measured in adipose tissue biopsies. After adjustment for age, use of hormone replacement therapy, benign breast tumour, BMI, alcohol, parity, age at delivery, years of hormone replacement therapy and lactation, no significant association was found with concentrations of PCBs in the whole data set, but an inverse association was observed when the analysis was restricted to the 75 estrogen receptor–negative cases. There was a significant trend of decreased risk across quartiles of serum concentration of PCBs 138, 153, 170, 180, 183 and 187; however, the only significant OR for specific quartiles was for PCB 153 (OR 0.3, 95% CI 0.1–0.9) when comparing the fourth with the first quartile (Raaschou-Nielsen et al., 2005). Within the same study, further analyses were carried out showing no significant associations between breast cancer risk and concentrations of Wolff ’s groups (Bräuner et al., 2014). Wolff ’s groups (Wolff et al., 1997) are as follows: Group 1 (estrogenic, including PCBs 187 and 201), Group 2A (DL-PCBs 118 and 156), Group 2B (anti-estrogenic, PCBs 138 and 170) and Group 3 (cytochrome P450 inducers, PCBs 99, 153, 180 and 183). This is the largest nested case–control study on breast cancer and PCBs and the only one with PCB concentrations measured in adipose tissue. The inverse association of concentrations of some PCB congeners among women with estrogen receptor– negative tumours does not have a clear interpretation. Within the residents of Oakland, California, USA, participating in the Child Health and Development Study, where cases were identified by linkage to the California Cancer Registry and the California Vital Status Records, a nested case–control study compared serum concentrations of 16 PCBs in archived early-postpartum serum samples from 112 cases and age-matched controls (median time from blood draw to diagnosis was 17 years). No associations 123

Safety evaluation of certain food additives and contaminants Eightieth JECFA

were reported between risk of cancer of the breast and sum of total PCBs or Wolff ’s groups (results not shown). PCB 203 was associated with an increased risk when comparing highest versus lowest quartile (OR 6.3, 95% CI 1.9–21.7). No significant associations were found for common congeners, such as PCB 130, 153, 170 and 180. The OR for a 1 mmol/L increase in the concentration of PCB 153 was 1.11 (P-value = 0.91). In this study, there was also a significant decreased risk associated with PCB 167, a DL-PCB congener. These associations were adjusted for total cholesterol, total triglycerides, parity, year of blood draw, BMI and breastfeeding after current pregnancy, in addition to matching by age (Cohn et al., 2012). This is the only nested case–control study to include mostly premenopausal women, but it has limited power. Besides these studies, the IARC (2015) review included another 12 articles published from 1994 to 2002 with results from seven cohorts. In three studies, no specific results on NDL-PCBs were provided; in three studies, no statistical associations were reported for PCBs 138, 153 and 180 or Wolff ’s groups including NDL-PCB congeners. One study reported a significant increase for higher concentrations (fourth quartile) of PCB 138, but no significant associations for PCB 138, 180 or 118. (ii) Non-Hodgkin lymphoma

Among residents of Washington County, Maryland, USA, who had participated in one of two studies conducted in 1974–1989 to obtain blood samples for a serum bank (the CLUE I study), participants were followed up until 1994 by linkage with the Washington County Cancer Registry. Seventy-four cases of nonHodgkin lymphoma (NHL) (ICD-8 200 or 202) were identified during follow-up, and 147 controls matched by race, sex and age were included in a case–control study. PCB concentrations were measured in serum collected before diagnosis and corrected for lipids. In addition to matching variables, the estimates were adjusted for education and cigarette smoking. There was a significant dose– response relationship between risk of NHL and quartiles of lipid-corrected serum concentrations for the sum of 28 measured congeners. In the analysis focusing on the effect of specific congeners, there was a significant exposure–response relationship between risk of NHL and increasing concentrations of PCB 138 and PCB 153 (P-value for trend < 0.05). The ORs for the fourth versus the first quartile were 4.4 (95% CI 1.5–12.6) and 2.2 (95% CI 0.9–5.2) for PCB 138 and PCB 153, respectively. It must be noted that there was also a significant association between risk of NHL and increasing levels of the DL-PCB congener, PCB 118, in serum (L.S. Engel et al., 2007). The Nurses’ Health Study was established in 1976 and included more than 120 000 registered nurses in the USA, who were subsequently followed by

124

WHO Food Additives Series No. 71-S1, 2016

Supplement 1: Non-dioxin-like polychlorinated biphenyls

questionnaire every 2 years; 32 826 women from the cohort provided a blood sample between 1989 and 1990. Thirty women with incident NHL diagnosed between the date of blood collection and 1994 (median follow-up, 1 year) were included as cases, and 78 cohort members selected previously as controls for a study of cancer of the breast served as controls. A statistically significant exposure– response relationship was observed between the risk of NHL and increasing concentrations of lipid-corrected PCBs (sum of 21 congeners) adjusted for age, BMI and smoking status. A significant exposure–response relationship was also observed for PCB 138 (P-value for trend < 0.05), but not for PCB 153. The OR for the third versus the first tertile of PCB 138 was 3.8 (95% CI 1.1–13.6), whereas the corresponding value for PCB 153 was 3.2 (95% CI 0.9–11.8). In this study, there was also a significant exposure–response relationship (P-value for trend < 0.05) between NHL risk and levels of the DL-PCB congener, PCB 118, in serum (L.S. Engel et al., 2007). An extended follow-up of the Nurses’ Health Study cohort (median time to diagnosis, 5.8 years) included 145 cases of NHL identified by annual followup questionnaires and confirmed by review of medical records and pathology reports; two controls were selected for each case (n = 290), matched by age, race, month of blood draw and fasting status. No association was observed between risk of NHL and total concentrations of PCBs (sum of 51 congeners measured as lipid-corrected concentrations) or concentrations of specific congeners (PCB 118, PCB 138, PCB 153, PCB 180) in serum after adjustment for race, age, date of blood draw, fasting status at blood draw (matching), region, BMI, smoking, height, parity and breastfeeding. The OR for the third versus the first quartile for PCB 153 was 0.82 (95% CI 0.43–1.56). The same pattern of no association was observed in the subgroup analysis by the main subtypes of NHL: diffuse large B-cell lymphoma, follicular lymphoma and chronic lymphocytic leukaemia/ small lymphocytic lymphoma (Laden et al., 2010). It must be noted that this is a well-designed study and that the positive association observed in the initial study was not confirmed in the second, larger study, after adjustment for additional relevant confounders. Moreover, in the second study, the time since blood draw to diagnosis of cases was prolonged (5.8 years compared with 1 year), and PCB measurements were carried out at different laboratories. Within the Physicians’ Health Study, started in 1982 in the USA as a randomized trial for the primary prevention of cardiovascular disease and cancer, 14 916 participants (male physicians aged 40–84 years at enrolment) provided a blood sample in 1982–1984 (before randomization). Newly diagnosed cases of NHL until 2003, identified by means of annual questionnaires, were confirmed by review of medical records. After excluding those diagnosed within 6 months after blood collection, with a prior diagnosis of cancer, with NHL of uncommon subtypes (e.g. mantle cell lymphoma) or lacking sufficient information for subtype 125

Safety evaluation of certain food additives and contaminants Eightieth JECFA

classification, 205 cases with available blood samples were included. For each case, two controls per case were randomly selected among subjects at risk of NHL when the case occurred, matched by race, age and date of blood collection. Lipidcorrected concentrations of 51 PCB congeners in serum were determined for cases and controls. A significant association was observed between NHL risk and log of lipid-corrected concentrations of PCBs (sum of all congeners), adjusted (in addition to matching variables) for region, BMI, smoking status, alcohol intake and height. The association was also significant for the log concentrations of PCB 138, PCB 153 and PCB 180, but not for PCB 118, a DL-PCB congener. The ORs for the fifth versus the first quintile were 2.1 (95% CI 1.1–3.8) for PCB 153 and 2.4 (95% CI 1.3–45) for PCB 180 (Bertrand et al., 2010). This is a well-designed study with reasonable sample size, although it was restricted to men. The JANUS Serum Bank contains serum samples collected between 1973 and 1991 from almost 300 000 individuals undergoing routine health examinations in Norway. Through linkage with the Norwegian Cancer Registry up to 1999 (median time to diagnosis, 16.6 years), 194 histologically confirmed cases of NHL were ascertained. Information, including lipid-corrected concentrations of 36 PCB congeners, was available for 190 case–control pairs matched by age, sex, county and date of examination. In the analysis, further adjustments were made for BMI and smoking status. No significant association was observed between NHL risk and the concentration of PCBs (sum of all congeners) when comparing higher quartiles with the first quartile, although there was a significant upward dose–response trend (P-value for trend < 0.05). A statistically significant increase in risk was reported for the highest to the lowest quartile of PCB 153 concentrations (OR 2.0, 95% CI 1.0–3.9), with a significant dose–response trend (P-value for trend < 0.05). Significant trends were also reported for PCB 118 and PCB 138 (L.S. Engel et al., 2007). It is not clear why significant associations were found for three common congeners (PCBs 118, 138 and 153), but not for all PCBs combined. Among participants in the Danish diet and cancer study described above (Raaschou-Nielsen et al., 2005), 278 initially cancer-free cohort members were diagnosed with NHL up to July 2008 (mean follow-up, 9.6 years); a subcohort of 256 participants was randomly selected for analysis using a case–cohort approach. Valid measurements of concentrations of 10 PCB congeners in adipose tissue were available for 239 cases and 245 subcohort members. No association was observed between lipid-corrected concentrations of total PCBs in adipose tissue and risk of NHL, adjusted for age, sex and BMI. There was also no consistent association and no significant trend with several congeners analysed individually, including PCB 153, PCB 138 and PCB 180, as well as some DL-PCB congeners, such as PCBs 118 and 156 (Bräuner et al., 2012). It must be noted that this is the

126

WHO Food Additives Series No. 71-S1, 2016

Supplement 1: Non-dioxin-like polychlorinated biphenyls

largest nested case–control study on NHL and PCB concentrations measured in adipose tissue, but estimates were adjusted only for age, sex and BMI. (iii) Cancers of the male genital tract

A nested case–control study on the risk of testicular germ cell tumours was carried out within the Norwegian JANUS cohort described above (L.S. Engel et al., 2007). Male cases and controls were matched by region, age group (2 years) and year of blood draw. Lipid-corrected measurements of the concentrations of 34 PCBs were available for 49 cases (34 seminomas, eight non-seminomas, five of mixed histology and two of unknown histology) and 51 controls. There was no statistically significant association (adjusted for matching variables) between risk of testicular germ cell tumours and total PCB concentration or concentrations of common NDL-PCB congeners, such as PCB 153 or PCB 138. When the analysis was restricted to seminomas, the OR for the highest versus the lowest tertile of PCB 99 concentration was 4.4 (95% CI 1.0–19.8). In this study, the strongest (and most significant) association was the increased risk observed for the DL-PCB congener PCB 167, for both all testicular germ cell tumours and seminomas (Purdue et al., 2009). It must be noted that this is a well-designed study, but with a small sample size and very limited power. In a cohort of men in the United States military, concentrations of 15 PCBs were analysed in pre-diagnostic serum samples of 736 incident cases of testicular germ cell tumours and 913 controls matched to the cases on age, race and serum draw date. Of the 15 PCB congeners analysed, four (PCB 28, PCB 52, PCB 105 and PCB 128) were excluded from data analysis, as fewer than 35% of the study samples had values above the LOD. The sum of the concentrations of the 11 remaining PCBs was significantly associated with decreased risk of all testicular germ cell tumours, as well as with seminoma and non-seminoma, adjusted for age, race/ethnicity, date of serum sample collection (matching), serum DDE level, age at serum draw, BMI and height. Results were also analysed in terms of individual congeners and by Wolff ’s groups. Statistically significantly decreased risks of all testicular germ cell tumours were associated with increasing concentrations (P-value for trend < 0.05) of eight specific PCBs (PCB 118, PCB 138, PCB 153, PCB 156, PCB 163, PCB 170, PCB 180 and PCB 187), but not with three other specific PCBs (PCB 99, PCB 101 and PCB 183). Statistically significantly decreased risks of all testicular germ cell tumours were also associated with all Wolff ’s groups. Similar decreases in risk were observed for both seminomas and non-seminoma tumours. However, no significant ORs were observed when comparing tertiles of PCB concentrations (McGlynn et al., 2009a). In an analysis restricted to 568 cases and 698 controls that examined associations between testicular germ cell tumours and concentrations of 11 PCBs in relation to polymorphisms in genes encoding 127

Safety evaluation of certain food additives and contaminants Eightieth JECFA

hormone metabolizing enzymes, a statistically significantly reduced risk (P-value for trend < 0.01) for PCB 138 (and PCB 118) was found only among subjects with the major homozygous allele for HSD17B4 (Chia et al., 2010). It must be noted that this appears to be a large, well-designed and well-implemented study, but the consistent inverse associations of cancer risk with exposure to PCBs is difficult to explain biologically. The cohort of the Japan Public Health Center–based Prospective Study consisted of 65 657 men, 14 203 of whom donated blood between 1990 and 1995. In total, 201 newly diagnosed cases of cancer of the prostate (97% pathologically confirmed) were identified up to 2005. For each case, two controls were selected, matched by age (within 3 years), public health centre area, residence, date and time of day of blood collection, and duration of fasting. Lipid-corrected concentrations of 41 PCB congeners were measured in plasma. Apart from matching variables, comparisons were further adjusted for BMI, smoking, alcohol, marital status, and intakes of green tea and miso soup. No statistically significant association with all cancers of the prostate was seen for total PCBs, individual PCBs or Wolff ’s groups. For individual congeners, ORs were not presented, but no significant differences were found between cases and controls in median concentrations of PCB 153 (P-value = 0.33), PCB 138 (P-value = 0.20), PCB 180 (P-value = 0.62) or PCB 170 (P-value = 0.48) in plasma. No statistically significant differences were found for concentration of total PCBs according to stage (localized or advanced) at diagnosis of cancer of the prostate (Sawada et al., 2010). This was a welldesigned and well-conducted study showing null results; although the sample size is limited, power is reasonable for the main analysis, although limited for subgroup analyses. (b) Case–control studies (i) Non-Hodgkin lymphoma WHO Food Additives Series No. 71-S1, 2016 128

In a case–control study in Australia including 694 histologically confirmed cases of NHL and 694 controls, exposure to PCBs was coded by an expert industrial hygienist based on questionnaire information. After adjusting for age, sex, residence and ethnicity, ever exposure to PCBs was not related to increased risk of NHL (Fritschi et al., 2005). A case–control study including 495 NHL cases and 495 controls was conducted in an area of northern Italy where environmental exposure had resulted from soil contamination; PCB concentration in the soil was used to define four areas with increasing levels of exposure. Risk of NHL was elevated for subjects having resided 10 or more years in any of the three most polluted areas, and particularly in the most polluted area (Maifredi et al., 2011). These studies do not provide any relevant information regarding the carcinogenic effects of NDL-PCBs.

Supplement 1: Non-dioxin-like polychlorinated biphenyls

A population-based case–control study in Sweden included 99 cases of NHL and 99 controls matched by age, sex and health service region (Hardell et al., 2009). After adjusting for age, sex and BMI, risk of NHL was elevated for subjects with PCB concentrations above the median concentration of the sum of 35 PCBs in plasma of controls, although this difference was statistically significant only for follicular lymphoma. PCBs were grouped according to the degree of chlorination, as lower chlorinated (PCBs 52, 66 and 74), moderately chlorinated (PCBs 99 + 113, 101, 105, 110, 118, 138, 153, 156, 170 + 190, 172 + 192, 178, 180 + 193, 182 + 187 and 189) and higher chlorinated PCBs (PCBs 194 and 206); whereas lower and higher chlorinated PCBs included NDL-PCBs only, the broader group of moderately chlorinated PCBs included a mixture of DL- and NDL-PCB congeners. No significant associations were reported by subgroups in the whole data set, but when the analysis was restricted to follicular lymphoma, the OR for the above versus below median concentration of higher chlorinated PCBs in serum was 9.6 (95% CI 1.9–49). The risks of NHL associated with both lower and higher chlorinated PCBs were higher among Epstein Barr virus early antigen IgG–positive subjects, especially among cases with diffuse large B-cell lymphoma. It must be noted that in this study, no adjustment for multiple comparisons was considered, in spite of the huge number of estimates and tests carried out. A large population case–control study was conducted in Canada. Lipidadjusted concentrations of 14 PCB congeners were measured in pretreatment samples of plasma from 422 cases of NHL and 460 controls frequency-matched to cases by 5-year age groups, sex and residence. ORs were adjusted for age, sex, education, BMI, ethnicity, farming and family history of NHL. There was a significantly increased risk of NHL for the sum of the concentrations of all PCBs. Risk of NHL was found to be highest in the highest quartile of the sum of the concentrations of seven NDL-PCBs (OR 2.18, 95% CI 1.41–3.38), with a significant trend. There were also significant associations for the concentrations of PCBs 138 and 153; for the latter, the OR for the highest versus the lowest quartile was 1.79 (95% CI 1.17–2.72). A similar pattern was observed for follicular lymphoma, with significant increased risk associated with the sum of the concentrations of NDL-PCB congeners and with the concentration of PCB 153, whereas no significant associations were reported for the diffuse large B-cell lymphoma. It must be noted that there was also an increased risk associated with the sum of the concentrations of the three DL-PCB congeners analysed, for both all lymphoma and the follicular subtype (Spinelli et al., 2007). This is a highquality study and was one of the largest studies of NHL and PCBs. It accounts for relevant confounders and provides results for individual congeners and lymphoma subtypes. The participation rate for controls was less than 50%, but this is typical of most of the available population-based case–control studies, and 129

Safety evaluation of certain food additives and contaminants Eightieth JECFA

potential confounding factors (e.g. education) were comparable between cases and controls, despite differences in participation. A multicentre study included 174 NHL cases and 203 controls from France, Germany and Spain. Patients admitted to the same hospital as the cases for non-cancer diseases were selected as controls in France and Spain, whereas German controls were selected from the general population. Concentrations of nine PCB congeners were measured in plasma, and individual results were provided for PCBs 28, 118, 138, 153, 170 and 180. Moreover, PCBs were grouped in functional groups as follows: pseudo-estrogenic (PCBs 28, 52 and 153), higher chlorinated anti-estrogenic (PCBs 170, 180 and 194), phenobarbital inducers (PCBs 101, 153, 180 and 194), 3-methyl cholanthrene inducers (PCBs 118, 138 and 170), immunotoxic (PCBs 138, 153 and 180) and BRCA1 inhibitors (PCBs 101 and 138). Risk estimates were adjusted for age, sex, education and centre. Risk of NHL did not increase by quartile of PCB concentration in plasma overall, for specific congeners or for the functional PCB congener groups. When exploring risk by lymphoma subtype, no significant associations were found for diffuse large B-cell lymphoma or chronic lymphocytic leukaemia/small lymphocytic lymphoma in the whole data set. However, in the German and French groups, but not in the Spanish group, combined risk of diffuse large B-cell lymphoma associated with concentrations of immunotoxic PCBs in plasma above the median concentration showed a 3-fold increase (OR 3.2, 95% CI 0.9–11.5), increasing to 6-fold (OR 6.1, 95% CI 1.0–37.8) in the upper quartile (P-value for trend = 0.04) (Cocco et al., 2008). This study is of good quality, with a meticulous classification of lymphoma. The association of the concentration of immunotoxic PCBs with diffuse large B-cell lymphoma in two centres is noteworthy. The heterogeneity between countries may have been a result of differences in PCB exposure or distribution of confounding factors. No adjustment for multiple comparisons was considered, in spite of the huge number of estimates and tests carried out. A case–control study conducted in 1998–2000 by the United States National Cancer Institute in four areas with population-based cancer registries included 1321 NHL cases and 1057 general population controls. Pretreatment plasma samples were available in a subset of 100 cases with a histologically confirmed diagnosis and 100 controls, for which concentrations of 40 PCB congeners in plasma were measured. In this analysis, only 28 congeners detected in at least 30% of samples were included. ORs were adjusted for the matching factors, age, sex, study site and date of blood draw; other potential confounders were tested (education, race, BMI and family history of NHL), but no confounding was observed. The results showed significant upward trends in risk of NHL with increasing quartiles of concentration of the subgroup of higher chlorinated PCB congeners (P-value for trend = 0.04) (this group includes PCBs 156, 180 and 194) in plasma. Regarding individual congeners, significant increases in

130

WHO Food Additives Series No. 71-S1, 2016

Supplement 1: Non-dioxin-like polychlorinated biphenyls

risk were reported only for PCB 180 (P-value for trend = 0.01) and PCB 194 (P-value for trend = 0.04). For PCB 180, there was also a significantly increased risk for the highest versus the lowest tertile (OR 3.5, 95% CI 1.53–9.15). Some associations were stronger among the cases of diffuse large B-cell lymphoma than among the cases of follicular lymphoma (De Roos et al., 2005). Further analysis of these data explored the interaction between PCB 180 and common variants (61 single nucleotide polymorphisms) in 36 genes implicated in the immune and inflammatory response; significant increases in risk of NHL were observed for PCB 180 in plasma for IFNG (C-1615T) TT, IL16 (3′-UTR, Ex22871A > G) AA, IL8 (T-251A) TT and IL10 (A-1082G) AG/GG genotypes (Colt et al., 2009). Further analyses explored the interaction with status of HLA-DRB1*01:01 class II leukocyte surface antigen and of the extended ancestral haplotype (AH) 8.1 (HLA-A*01-B*08-DR*03-TNF-308A); risk of NHL was elevated among study subjects with concentrations of PCB 180 in blood above the median concentration and lacking the HLA-DRB1*01:01 allele or the AH 8.1 allele (Wang et al., 2011). These related studies were well conducted, and an extensive and detailed analysis was carried out; however, this is a relatively small study, particularly for subgroup analyses, with wide confidence intervals for the estimates. In a study among residents in an area of Besançon, France, close to a municipal solid waste incinerator, cases were subjects with newly diagnosed NHL during 2003–2005 at the Department of Hematology of the University Hospital (the only tertiary referral hospital in the region). Thirty-four cases (out of 53 eligible) were included; 34 controls were randomly selected from the donor registry of the regional blood bank living in the study area, matched by sex, age and date (year) of blood draw. Concentrations of a total of 18 PCB congeners were measured in serum. The OR associated with serum concentrations of NDL-PCBs (sum of PCBs 28, 52, 101, 138, 153 and 209) was 1.02 (95% CI 1.01–1.05) per 10 ng/g lipid. The OR for NDL-PCBs did not substantially change when adjusting (in addition to matching variables) for BMI or pesticides (β-hexachlorocyclohexane, p,p′DDT). Significant associations were also observed for the individual congeners PCB 130, PCB 180 and PCB 153; for PCB 153, the OR (per 10 ng/g lipid) was 1.04 (95% CI 1.00–1.09). It must be noted that in this study, the strongest (and most significant) positive associations were observed for DL-PCBs, both when analysed as the sum of 12 congeners and in the analysis for individual congeners (Viel et al., 2011). Besides the studies discussed above, IARC (2015) reviewed another three case–control studies, but in all of them the results referred to a mixture of PCBs, including both NDL- and DL-PCB congeners, and therefore they do not provide any relevant information regarding the potential effects of NDL-PCBs on the risk of NHL. 131

Safety evaluation of certain food additives and contaminants Eightieth JECFA

(ii) Breast cancer

PCB exposure (28 congeners) was measured in banked serum collected in 1981–1987 from 63 Alaskan native women who subsequently developed breast cancer and 63 age-matched controls; no association with PCB exposure (sum of all congeners) was found after adjusting for ethnicity, family history of cancer of the breast and parity (Rubin et al., 2006). In a study in Greenland among subjects of Inuit descent, higher concentrations of PCBs in serum were found for 31 cases of breast cancer compared with 115 controls; however, the OR for total PCBs (sum of 12 congeners) did not demonstrate any significant association (Bonefeld-Jorgensen et al., 2011). In a population-based study in North Carolina, USA, including 612 breast cancer cases and 599 controls, in which a significantly increased risk of breast cancer had previously been found for higher concentrations of PCBs (sum of 35 congeners) among African women after adjusting for age, menopausal status, BMI, parity/lactation, hormone replacement therapy and income, statistically significant multiplicative interactions were observed between CYP1A1 M2-containing genotypes and total PCBs among white women and between CYP1A1 M3-containing genotypes and total PCBs among African American women (Li et al., 2005). In a population-based case–control study of African American women including 355 cases and 327 controls, risk of breast cancer was not associated with PCB concentrations (measured using Aroclors 1242 and 1260 and reported as total PCBs) in serum adjusted by age, BMI and breastfeeding; moreover, PCB concentrations in serum were not associated with an increase in the risk of any subtype of cancer defined by tumour receptor status (progesterone receptor, estrogen receptor, p53 or HER-2/neu) (Gatto et al., 2007). In these studies, the results refer to a mixture of PCBs (in some instances of unknown composition), and therefore they do not provide relevant information regarding the potential effects of NDL-PCBs on breast cancer risk. In a hospital-based case–control study on breast cancer in Nagano, Japan, PCB concentrations (sum of 41 congeners) were measured in serum samples from 403 matched pairs collected from 2001 to 2005. After adjustment for total lipid concentration in serum, BMI, reproductive risk factors, smoking, diet and medical history, serum concentrations of total PCBs were associated with significantly decreased risk of cancer of the breast. For the specific congeners PCB 153 and PCB 180, the ORs comparing the highest to the lowest concentration quartiles were 0.40 (95% CI 0.18–0.91) and 0.29 (95% CI 0.13–0.66), respectively. The trend in the inverse relationship persisted when results were stratified by hormone receptor and menopausal status (Itoh et al., 2009). It is not clear how the inverse associations reported in this study may be explained. In a hospital-based case–control study in Mexico, blood samples were used to measure serum concentrations of 20 PCBs for 70 breast cancer cases

132

WHO Food Additives Series No. 71-S1, 2016

Supplement 1: Non-dioxin-like polychlorinated biphenyls

and 70 controls. Age, age at menarche, lactation, menopausal status, BMI and family history of breast cancer were adjusted for in the analysis. A significantly increased risk was apparent for the sum of the concentrations of all PCBs. When PCBs were grouped according to the Wolff ’s groups, significantly increased risks were reported for group 2B (di‐ortho-substituted, PCBs 128, 138 and 170), with OR 1.90 (95% CI 1.25–2.88); group 3 (cytochrome P450 inducers, PCBs 153 and 180), with OR 1.81 (95% CI 1.08–3.04); and group 4 (PCBs 8, 195, 206 and 209), with OR 1.57 (95% CI 1.20–2.07). Elevated ORs were reported for several PCB congeners (PCBs 128, 138, 170, 180, 187, 195, 206 and 209), but not for PCB 153 (OR 1.36, 95% CI 0.67–2.73). The ORs were generally higher in postmenopausal women. No significant association was reported for group 2A, which includes several DL-PCBs, but a significantly increased risk was observed for PCB 118, the most common DL-PCB congener (Recio-Vega et al., 2011). Although this was a small study, several increased risks were reported; however, no attempt was made to correct for multiple comparisons. It must be noted that the age distribution was different in cases and controls, suggesting potential for residual confounding by age. Furthermore, the analytical approach is unclear; actually, it is impossible to clarify, from the information provided in the paper, to what level of PCB exposure the ORs reported correspond. Besides these studies, IARC (2005) reviewed 12 more case–control studies (articles published between 1998 and 2004) on the risk of breast cancer and PCBs. No specific results on NDL-PCBs were reported in half of these studies. In four studies, statistical associations were observed for either individual congeners or groups of NDL-PCBs, and no significant association of NDL-PCBs with breast cancer risk was observed in the remaining two studies. Occupational risk factors for cancer of the male breast were investigated in a multicentre study of 104 cases and 1901 controls in eight European countries. Lifetime work history was obtained by in-person interviews, and potential occupational exposures, including to PCBs, were assessed using expert judgement. No significant association was observed for PCBs and dioxins combined (Villeneuve et al., 2010). Results from this study do not provide any relevant information regarding the effects of NDL-PCBs on cancer of the male breast. (iii) Cancer of the prostate

A hospital-based case–control study was conducted in Iowa, USA, in which 30 PCBs were measured in serum samples from 58 patients with cancer of the prostate and 99 age-matched controls. Different approaches for the grouping of PCBs were used, but specific results were shown for three groups: the moderately chlorinated PCBs included 18 congeners, among which were some DL-PCBs; 133

Safety evaluation of certain food additives and contaminants Eightieth JECFA

WHO Food Additives Series No. 71-S1, 2016 134

the phenobarbital inducers included PCBs 99, 101, 153, 180, 183 and 194 (all of them NDL-PCBs); and the phenobarbital inducers and persistent PCBs included PCBs 99, 153, 180, 183 and 203. It should be noted that there is an important degree of overlapping between the two phenobarbital inducer groups. The estimates were adjusted for age, BMI and history of prostatitis; in a second model, further adjustment was carried out for total lipids. For the group phenobarbital inducers, there was a marginally significant, increasing risk of prostate cancer across tertiles of PCB concentrations (ng/mL) (P-value for trend = 0.048), which was no longer significant (P-value for trend = 0.10) after further adjustment for total lipids. The same pattern was observed for the phenobarbital inducers and persistent PCBs group, with P-values for trend of 0.043 (not lipid adjusted) and 0.09 (lipid adjusted). In this group, the OR for the highest versus the lowest tertile (not lipid adjusted) was 2.44 (95% CI 1.01–5.90) (Ritchie et al., 2005). It should be noted that this study was small, with multiple comparisons. In a population-based case–control study in Sweden, concentrations of 37 PCBs in samples of fat tissue from 58 cases of cancer of the prostate were compared with 20 controls with benign prostate hyperplasia. The association with the sum of PCBs was not significant. The concentration of PCB 153 in fat tissue (above versus below median concentration) was associated with a significantly increased risk of cancer of the prostate (OR 3.15, 95% CI 1.04–9.54), adjusted for age and BMI. In the stratified analysis by prostate-specific antigen levels, the increased risk was significant only for men with prostate-specific antigen concentration above 10 ng/mL (OR 7.91, 95% CI 2.0–31.2) (Hardell et al., 2006a). It should be noted that this study was small, with multiple comparisons. In a case–control study among urology patients in Ontario, Canada, concentrations of 14 PCB congeners were measured in serum of 79 men with incident cancer of the prostate and 329 age-matched controls. No association was observed between the sum of the concentrations of all PCBs or concentrations of individual PCBs, including common NDL-PCB congeners (PCBs 138, 153, 170 and 180) as well as DL-PCB congeners (PCBs 118 and 156), adjusted for age, physical activity, alcohol consumption and tobacco smoking (Aronson et al., 2010). As both cases and controls underwent the same diagnostic procedures and were screened by prostate-specific antigen level and digital rectal examination, selection bias was unlikely in this study. (iv) Malignant melanoma

In a multicentre case–control study in nine European countries, 293 men and women with uveal melanoma were compared with 3198 population and hospital controls frequency-matched by country, age and sex. Exposure to transformer oils was assessed by questionnaire, based upon the subject’s reported exposure to

Supplement 1: Non-dioxin-like polychlorinated biphenyls

a named brand of oil with known PCB content. There was a strong, significantly increased risk of uveal melanoma for confirmed exposure, especially for exposure to Pyralène, among men (Behrens et al., 2010). However, the exposure to PCBs refers to a mixture of (unknown) congeners, and therefore this study does not provide any relevant information regarding the effects of NDL-PCBs on uveal melanoma. Within a large study originally set up to evaluate the effect of exposure to ultraviolet light and gene variants on risk of melanoma, with controls recruited using population-based registries, a case–control study was conducted including 80 patients with malignant melanoma of the skin and 310 controls. Lipid-adjusted concentrations of 14 PCB congeners in plasma were determined. Results were adjusted for age, sex, education, skin reaction to repeated sun exposure and total recreational sun exposure. Statistically significant associations with malignant melanoma were observed for the sum of the concentrations of all 14 PCB congeners, as well as for the 11 NDL-PCB congeners; for the latter, the OR for the highest compared with the lowest quartile was 7.02 (95% CI 2.30–21.4), with a significant trend across quartiles (P-value for trend < 0.001). Concerning specific congeners, significantly increased risks were also found for PCBs 138, 153, 170 and 187. The OR for the highest compared with the lowest quartile of PCB 153 concentration was 4.86 (95% CI 1.68–14.1). In this study, significantly increased risks were also observed for the sum of the concentrations of DL-PCBs, as well as for some specific DL-PCB congeners, such as PCBs 118 and 156 (Gallagher et al., 2011). It should be noted that in light of its appropriate design and control of relevant potential confounders, this is a high-quality study, despite the relatively small sample size. It cannot be discounted that the positive associations for all the individual PCB congeners may have been a result of correlations among congeners. In contrast, multiple comparisons were not formally addressed, but it is likely that this would not change the interpretation of the results. (v) Other cancers

In a European multicentre case–control study of 183 men with histologically confirmed carcinoma of the extrahepatic biliary tract and 1938 matched controls, self-reported job descriptions were used as indicators of occupational exposure to suspected endocrine-disrupting compounds, including PCBs. After adjustment for age, country and history of gallstones, a significantly increased risk of extrahepatic biliary tract cancer was reported for ever-exposure to PCBs. These results were based on a small number of exposed cases, and trends were inconsistent (Ahrens et al., 2007). Results from this study do not provide any relevant information regarding the effects of NDL-PCBs on cancer of the extrahepatic biliary tract. 135

Safety evaluation of certain food additives and contaminants Eightieth JECFA

In a case–control study in Sweden, concentrations of 37 PCBs were measured in serum from blood collected from 61 cases of cancer of the testis and 58 age-matched controls from the Swedish population registry, as well as from 44 mothers of cases and 45 mothers of controls. No association between cancer of the testis and the sum of PCB concentrations in serum in cases and controls was observed. However, there was a significant association between the concentration of PCBs analysed in serum from mothers of cases and increased risk of testicular cancer in the sons, both for the sum of all congeners and for the group of enzyme inducers (sum of PCBs 99, 153, 180, 183, 199 and 201, all of them NDL-PCBs); for the latter, the OR for above versus below the median concentration was 2.6 (95% CI 1.1–6.5). The increased risk was also significant for the non-seminoma cases, but not for the seminomas (Hardell et al., 2006b). The timing of blood collection of the mother, which was decades after the births of the cases, makes interpretation of these results difficult. In a case–control study in northern Mexico including 43 cases of lung cancer and 86 controls recruited from two hospitals, information on history of exposure to PCBs was collected through in-person interview, and concentrations of 20 PCB congeners were measured in serum. After adjustment for age, agricultural occupation and tobacco smoking, the association between PCB 18 concentration and cancer of the lung (OR 1.13, 95% CI 1.04–1.38) was the only significant result out of the 20 individual congeners analysed. No significant associations were found for concentrations of the sum of all PCBs or of any of the seven groups of PCBs tested (Recio-Vega et al., 2013). It should be noted that this study provides information about lower chlorinated PCBs, which are rarely measured; however, the etiological relevance of such measurements of PCBs of short half-life is questionable. In addition, methods used for subject recruitment and statistical analysis were not clearly described, and the possibility of residual confounding by age or other factors cannot be ruled out. For instance, cases were older than controls (mean age 66 versus 55 years), with a higher proportion of males (77% versus 66%); moreover, the increase in risk for history of smoking was relatively low (OR 2.72). Regarding the analysis, it is impossible to clarify, from the information provided in the paper, to what levels of PCB exposure the ORs reported correspond. A population-based case–control study in California, USA, included 184 children aged 0–7 years with acute lymphocytic leukaemia and 212 controls matched by birth date, sex, race and ethnicity. Concentrations of six PCB congeners (PCBs 105, 118, 138, 153, 170 and 180) in residential carpet dust were used as an exposure indicator. In addition to matching variables, the estimates were adjusted for age of home and breastfeeding duration. There was a significant association between the risk of acute lymphocytic leukaemia in children and PCB concentration in dust for PCB 138 and PCB 153. The OR

136

WHO Food Additives Series No. 71-S1, 2016

Supplement 1: Non-dioxin-like polychlorinated biphenyls

comparing concentrations of PCB 153 above versus below the LOD of 1 ng/g was 1.67 (95% CI 1.06–2.63); moreover, there was a significantly increased risk of acute lymphocytic leukaemia across tertiles of PCB 153 concentration (P-value for trend = 0.018). Risk estimates for the PCB congeners were not changed by adjustment for concentrations of the organochlorine pesticides DDT, DDE, α-chlordane, methoxychlor and pentachlorophenol. In this study, a significantly increased risk was also reported for the common DL-PCB congener, PCB 118 (Ward et al., 2009). This is a well-designed study, using a good method of exposure assessment; moreover, potential confounding was ruled out. 2.3.7 Endocrine and metabolic effects (a) Thyroid function and thyroid diseases

Associations of total T4 and TSH levels with PCB concentrations (22 congeners) were assessed in adults without thyroid disease who participated in the 1999– 2002 NHANES, a sample representative of the United States population. There were significant inverse associations of total PCB concentrations with T4 and TSH levels in women over 60 years of age (Turyk, Anderson & Persky, 2007). The concentrations of 15 PCBs in serum and indicators of thyroid structure and function were analysed in adults from two areas from eastern Slovakia. The sum of the concentrations of all PCBs was significantly associated with higher thyroid volume and higher concentration of thyroperoxidase antibodies (Langer et al., 2007a), higher levels of T4 and lower levels of T3 (Langer et al., 2007b), and a higher prevalence of antinuclear antibodies (Cebecauer et al., 2009), indicators of autoimmune or possible immunotoxic effects. In a cohort of women from Michigan, USA, no association was found between the cumulative incidence of thyroid disease during the 25-year follow-up and PCB levels in serum collected at baseline, measured by means of the PCB–Aroclor 1254 detection method (Yard et al., 2011). In a case–control study in the area of Fukuoka, Japan, a strong association was reported between congenital hypothyroidism (cretinism) in neonates and total PCB concentrations (congeners not specified) in breast milk of mothers (Nagayama et al., 2007). In all these studies, the PCB exposure refers to a complex mixture of PCBs, and results for individual congeners are not reported. Therefore, thse studies do not provide any relevant results regarding the potential effects of NDL-PCBs on the thyroid. (i) Cross-sectional analyses

In a cohort of 118 pregnant women 25–34 years of age and 118 newborns from the general population in central Taiwan, China, concentrations of 27 PCBs (including the six indicator congeners) were measured in placental blood, and thyroid hormone levels were measured in cord blood. A significant positive 137

Safety evaluation of certain food additives and contaminants Eightieth JECFA

correlation was observed between the sum of the concentrations of PCB 138, PCB 153 and PCB 180 and levels of total and free T4 only in female neonates. No association was found for T3, TSH or thyroxine-binding globulin (Wang et al., 2005). In a study on adults resident in three areas in eastern Slovakia, concentrations of nine PCBs (including the six indicator congeners) in serum and thyroid volume, TSH level and thyroid peroxidase antibodies were measured. A significantly increased thyroid volume was observed for subjects with the highest concentration of PCB 101 (P-value < 0.02); a non-significantly increased thyroid volume (P-value > 0.06) was also observed for higher concentrations of PCBs 153 and 180. No association was reported with TSH level or thyroid peroxidase antibodies (Langer et al., 2005). In south-western Quebec, Canada, maternal blood was collected during the first or second trimester from pregnant women, and cord blood was also collected at delivery. Concentrations of 14 PCB congeners and TSH, total T3 and free T4 were measured in serum from both blood samples. In maternal blood during pregnancy, there was a significant negative correlation between total T3 level and concentrations of PCB 138, PCB 153 and PCB 180 in serum. No associations were observed for TSH or free T4 level. No associations were observed between PCB concentrations and levels of thyroid hormones measured in cord blood (Takser et al., 2005). It should be noted that although prenatal and postnatal measurements were carried out in this study, only cross-sectional analyses were reported. Lipid-adjusted concentrations of PCB 153 and TSH levels in serum were measured in middle-aged and elderly men from the east (Baltic) coast of Sweden. No association was observed between PCB concentrations and the concentration of TSH (Rylander et al., 2006). Concentrations of PCBs 118, 138, 153, 170 and 180 and thyroid hormones were measured in the cord blood of neonates from Antwerp, Belgium. Significant inverse correlations were observed between PCBs 138, 170 and 180 and both free T3 and free T4 levels and between PCB 153 and the level of free T4. No association was observed between any of the PCB congeners measured and TSH level (Maervoet et al., 2007). Concentrations of 57 PCB congeners in serum, as well as free T4, total T3 and TSH, were measured in adult men recruited from an infertility clinic from 2000 to 2003 in Boston, Massachusetts, USA. In multivariate linear regression, significant associations were found between increasing concentrations of PCB 138 and PCB 153 and lower levels of total T3 and higher levels of TSH; no significant association was reported between PCB concentrations and the level of free T4 (Meeker, Altshul & Hauser, 2007).

138

WHO Food Additives Series No. 71-S1, 2016

Supplement 1: Non-dioxin-like polychlorinated biphenyls

Maternal concentrations of thyroid hormones and PCBs (34 congeners) were measured in blood collected at around 26 weeks of gestation (n = 320) or before delivery (n = 14) from a birth cohort study (CHAMACOS) in pregnant women and children in the Salinas Valley, California, USA. Multiple linear regression analysis showed a significant inverse association between free T4 level and concentrations of PCB 28, PCB 52 and PCB 105; there was an inverse association as well with concentrations of PCBs 138, 183 and 180, but it was not statistically significant. No significant associations were reported between concentrations of NDL-PCBs and total T4 or TSH level (Chevrier et al., 2008). The cross-sectional relationship between concentrations of thyroid hormones and seven PCB congeners was analysed in children from a general population birth cohort in Menorca, Spain, with blood samples collected at age 4 years. Concentrations of PCB 138 and PCB 153 in blood were significantly related to lower total T3 levels, whereas no relationship was found between any of the measured NDL-PCBs and free T4 or TSH level (Álvarez-Pedrerol et al., 2008). Concentrations of thyroid hormones and 18 PCBs were measured in serum from youth members, aged 13 years, of the Akwesasne Mohawk Nation who were resident in the proximity of the St Lawrence River in New York State, USA, and in Ontario and Quebec, Canada. In multiple regression analysis, a significant association was observed between increased levels of TSH and higher concentrations of PCB 138 (measured as the sum of PCB congeners 138 + 163 + 164) and PCB 153, whereas free T4 level was significantly associated with lower concentrations of PCB 52, PCB 138 (+ 163 + 164), PCB 101 (+ 90) and PCB 153. No significant associations were reported between any of the measured NDLPCB congeners and total T3 or total T4 level (Schell et al., 2008). Concentrations of 16 PCB congeners and thyroid hormones were measured in serum from adult freshwater fish consumers, aged 18–74 years, from two lakeside regions in Quebec, Canada. The following significant associations were found: increased concentrations of PCB 138 were related to lower levels of T3 in women and T4 in men; increased concentrations of PCB 153 were related to lower levels of T4 in women; and increased concentrations of PCB 138 and PCB 180 were related to higher levels of TSH (Abdelouahab et al., 2008). Between 2004 and 2005, umbilical cord blood from women delivering at the Johns Hopkins Hospital, Baltimore, Maryland, USA, was collected for the measurement of the concentrations of six PCBs (PCBs 74, 99, 118, 138 [coeluted with 158], 153 and 180) and thyroid hormones in serum. In the full study population, none of the PCBs was associated with a change in average levels of thyroid hormones. For infants born by spontaneous, vaginal, unassisted deliveries, concentrations of PCBs 138 (+ 158), 153 and 180 in serum were associated with lower levels of total T4 and free T4, but no association was observed with TSH level (Herbstman et al., 2008). 139

Safety evaluation of certain food additives and contaminants Eightieth JECFA

A cross-sectional study was conducted in 2004 among the adult (≥18 years of age) Inuit population of Quebec, Canada, in which concentrations of thyroid hormones, thyroxine-binding globulin and 14 PCBs were measured in serum. In multiple linear regression models, increased concentrations of PCBs 138, 153 and 180 were significantly associated with lower levels of free T4 and total T3; and increased concentrations of PCBs 138 and 180 were significantly associated with higher levels of TSH. No significant associations were found for any of these three congeners and thyroxine-binding globulin (Dallaire et al., 2009). (ii) Prospective studies

In the CHAMACOS cohort described above, concentrations of 34 PCB congeners measured in blood collected at around 26 weeks of gestation in Mexican American women from the Salinas Valley, California, USA, were compared with TSH levels in their children’s blood collected shortly after birth. Linear regression models, adjusted for neonatal age at time of heel stick for TSH measurement, gestational age at birth, infant birth weight, sex and mother’s pre-pregnancy BMI, were used to assess the relationship between the concentrations of TSH and the 19 PCB congeners with a detection frequency of 75% and above, both expressed on the log scale. A significant positive association with TSH level was found with PCBs 101, 138, 153 and 180. Each 10-fold increase in the concentration of PCB 101 was associated with an increase of 23% in TSH concentration (95% CI 7–45%); the corresponding figures for other indicator congeners were 23% (2–51%) for PCB 138, 20% (0–48%) for PCB 153 and 23% (2–48%) for PCB 180. There were positive associations with PCBs 28 and 52, but they were not statistically significant. In this analysis, PCB congeners were grouped according to previously proposed structure-based and mechanism-based groupings; in addition, another group was based on the congeners’ ability to induce a group of enzymes (UGT, CYP1A/EROD, CYP2B/PROD). This group showed a positive association with TSH level, but it includes two DL-PCB congeners. Among the mechanismbased groups, only group 3 of the Wolff classification (phenobarbital, CYP1A and CYP2B inducers), including PCBs 99, 153, 180 and 183, was significantly associated with TSH level. Among the 19 congeners analysed individually, two DL-PCB congeners were included (PCB 118 and PCB 166); none of them showed a significant association with TSH level (Chevrier et al., 2007). In the cohort from the Johns Hopkins Hospital 2004–2005 study described above, in addition to umbilical cord blood, blood for a second T4 test was collected during a routine paediatric visit, on average at 18 days of age (range 5–117 days), for a subset of 139 infants. In the analysis, restricted to 92 infants born by spontaneous, vaginal, unassisted deliveries, concentrations of

140

WHO Food Additives Series No. 71-S1, 2016

Supplement 1: Non-dioxin-like polychlorinated biphenyls

the four measured PCBs were inversely associated with T4 level, although the association was not statistically significant for PCB 180. Using multiple logistic regression, the ORs for each increase in one unit of ln PCB concentration (ng/g lipid), comparing the lowest 20% with the highest 80% of T4 level, adjusted for baby’s sex, gestational age, maternal age, maternal race, maternal pregnancy BMI, smoking status and time (days) since birth, were as follows: PCB 138: 5.30 (95% CI 1.73–16.21); PCB 153: 3.40 (95% CI 1.31–8.83); and PCB 180: 1.89 (95% CI 0.83–4.30). The only DL-PCB congener measured in this study, PCB 118, was also inversely associated with T4 level, with OR 4.53 (95% CI 1.53–13.4). Primiparous women in the Uppsala region of Sweden who were recruited between 1996 and 1999 (POPUP cohort) agreed to donate a serum sample in late pregnancy (32–34 weeks), and breast milk was obtained from the women 3 weeks after delivery. Newborns were sampled for blood at 3 weeks and 3 months after birth. Thyroid hormones were measured in serum, and concentrations of 10 PCB congeners were measured in serum and milk. The analysis was carried out by means of linear regression using log-transformed concentrations of TSH (mU/L), total T3 (nmol/L) or free T4 (pmol/L) with log-transformed concentrations (ng/g lipid) of three groups of PCBs: lower chlorinated PCBs (PCBs 28, 52 and 101), di-ortho-substituted PCBs (PCBs 138, 153 and 180) and DL-PCBs (PCBs 105, 118, 156 and 167); concentrations of the DL-PCBs were expressed in TEQs. In simple regression models, only two significant associations were observed. The inverse association between prenatal exposure to lower chlorinated PCBs and total T3 levels at 3 weeks of age became non-significant (P-value = 0.057) after adjustment for the covariate with the strongest influence on the association (birth weight), but further adjustment of the association with alcohol consumption, age of the mother and sex of the infant made the association significant again (P-value = 0.048) (adjusted coefficients with 95% CI not provided). In contrast, the positive association between TSH levels at 3 months of age and prenatal di-ortho-substituted PCB exposure became non-significant (P-value = 0.093) after adjustment of the association for sex of the infant, which was the potential confounder with the strongest influence on the association (adjusted coefficients with 95% CI not provided) (Darnerud et al., 2010). In this study, no association was reported between thyroid hormone levels and the sum of the concentrations of the four DL-PCB congeners analysed. (b) Diabetes

In the polybrominated biphenyls (PBBs) cohort in Michigan, USA, who were exposed to PBBs via their food due to accidental contamination of cattle feed in the 1970s, the incidence of diabetes assessed during a 25-year follow-up of more than 1000 participants was associated with concentrations of PCBs measured as 141

Safety evaluation of certain food additives and contaminants Eightieth JECFA

Aroclor 1016, 1254 or 1260 in serum collected at baseline (Vasiliu et al., 2006). In a cohort of Great Lakes sport fish consumers (Wisconsin, USA) formed in 1990, in a cross-sectional analysis of data from 2004 to 2005, the concentration of total PCBs (the sum of the concentrations of 19 congeners, including the DLPCB, PCB 118) in serum was not associated with the prevalence of self-reported diagnosis of diabetes (Turyk et al., 2009a). In the same cohort, concentrations of PCBs were assessed prospectively in patients free of diabetes in 1994–1995 and followed through 2005. The incidence of diabetes was not associated with the baseline concentrations of a sum of 18 PCB congeners (including PCB 118) in serum (Turyk et al., 2009b). In adult subjects from three regions from eastern Slovakia, the prevalence of diabetes (from level of fasting plasma glucose, oral glucose tolerance test or diabetes diagnosed and treated by a physician) was significantly associated with increasing concentrations of the sum of 17 PCB congeners (Ukropec et al., 2010). In these studies, PCB exposure referred to a mixture of PCBs, including some DL-PCB congeners, and results for individual congeners were not reported. Therefore, these studies do not provide any relevant information on the potential association between diabetes and exposure to NDLPCBs. (i) Cross-sectional analyses

In a cohort of professional fishermen and their wives (aged 49–77 years) from the Swedish east coast, the participants were asked if they had diabetes and, if so, since which year; moreover, they were asked if they were taking oral antidiabetic drugs or insulin or were on a diet. In a cross-sectional analysis, the prevalence of diabetes, adjusted for potential confounders, was significantly associated with increased concentration of PCB 153 in serum (Rylander, Rignell-Hydbom & Hagmar, 2005). In an analysis of the previous cohort, there was also a significant association between higher concentrations of PCB 153 in serum and prevalence of type 2 diabetes among women (Rignell-Hydbom, Rylander & Hagmar, 2007). In the adult population of the USA from the NHANES 1999–2002, the participants were considered to have diabetes if their fasting plasma glucose was ≥126 mg/dL, their non-fasting plasma glucose was ≥200 mg/dL or they reported a history of physician-diagnosed diabetes. Compared with subjects with concentrations of PCB 153 in serum below the LOD, after adjustment for age, sex, race and ethnicity, poverty income ratio, BMI and waist circumference, diabetes prevalence was strongly positively associated with lipid-adjusted concentrations of PCB 153 in serum (Lee et al., 2006). In the NHANES 2002–2004, a positive finding of diabetes (type 2 diabetes) for a participant was determined if glycosylated haemoglobin was ≥6.5% or if the participant answered “yes” to the question, “Have you ever been told by a doctor or health professional that you

142

WHO Food Additives Series No. 71-S1, 2016

Supplement 1: Non-dioxin-like polychlorinated biphenyls

have diabetes or sugar diabetes?”. A positive association of prevalence of diabetes with concentration of PCB 153, as well as with concentrations of the DL-PCBs, PCB 118 and PCB 126, was observed (Hofe et al., 2014). In a cross-sectional study of an adult (≥30 years) Native American (Mohawk) population (New York, USA), fasting serum samples were analysed for 101 PCB congeners. Participants who had fasting glucose values above 125 mg/dL or who were taking antidiabetic medication were defined as persons with diabetes. The prevalence of diabetes was associated with higher concentrations of PCB 153 and PCB 74 in serum; however, this association became non-significant after adjustment for concentrations of other analytes (DDE, HCB and mirex) in serum (Codru et al., 2007). From 1999 to 2002, the Greenland population study was carried out among adult Inuit living in Greenland. Thirteen PCB congeners were measured in the serum of participants. Participants were asked whether a physician had ever told them that they had diabetes, in which case a fasting venous plasma glucose measurement was taken. All other participants had an oral glucose tolerance test. No significant associations were observed between the prevalence of diabetes or impaired glucose tolerance and concentrations of NDL-PCBs (sum of PCBs 28, 52, 99, 101, 128, 138, 153, 163, 170 and 180) in serum (Jørgensen, Borch-Johnsen & Bjerregaard, 2008). Concentrations of eight PCB congeners were measured in serum from adults (≥15 years) from a First Nations community in northern Ontario, Canada. A significant association was observed between the prevalence of self-reported diabetes and concentrations of PCB 153 and PCB 74 (Philibert, Schwartz & Mergler, 2009). In a study performed among the participants, aged 40–64 years, in a Control Obesity Program in Saku, Japan, concentrations of 13 PCB congeners were measured in blood. Subjects with a fasting plasma glucose concentration of ≥126 mg/dL, a glycated haemoglobin (HbA1c) level of ≥6.5%, a prescription for hypoglycaemic medicine or a history of physician-diagnosed diabetes were defined as having diabetes. A significant association with prevalence of diabetes was observed for higher concentrations of PCB 180, but no association was found for other NDL-PCB congeners (PCB 138, 153 or 180) (Tanaka et al., 2011). In a study of previous employees in a capacitor manufacturing plant in Illinois, USA, a cross-sectional analysis was carried out with data collected in 1996 on postmenopausal women. Concentrations of 38 PCB congeners were measured in blood, and diabetes was defined as an affirmative answer to the question, “Have you ever been diagnosed by a doctor as having diabetes mellitus or high blood sugar?”. A significant increase in prevalence of diabetes was associated with higher concentrations of NDL-PCBs (sum of 31 congeners) and 143

Safety evaluation of certain food additives and contaminants Eightieth JECFA

WHO Food Additives Series No. 71-S1, 2016 144

estrogenic PCBs (sum of PCBs 52, 99, 101, 110 and 153) in serum (Persky et al., 2011). A cross-sectional study was carried out in elderly participants (aged 57– 70 years) from the Helsinki Birth Cohort Study in Finland, which represents a general adult urban Finnish population. The diagnosis of diabetes was based on an oral glucose tolerance test and the WHO (1999) criteria for glucose intolerance. A positive but non-significant association was observed between the prevalence of diabetes and higher concentrations of PCB 153 (Airaksinen et al., 2011). Volunteers from a cross-sectional study of randomly selected adults (≥18 years) from the Anniston community in Alabama, USA, completed a health survey and underwent measurements of fasting glucose and 35 PCB congeners in serum. Diabetes was defined as self-report of physician-diagnosed diabetes or a fasting glucose level above 125 mg/dL. No significant association was found between the prevalence of diabetes and the concentrations of two groups of NDLPCBs in serum: estrogenic congeners (PCBs 44, 49, 66, 74, 99, 110 and 128) and di‑ortho-substituted, tri‑ortho-substituted and tetra-ortho-substituted PCBs (20 congeners, including PCBs 52, 101, 138, 153 and 180) (Silverstone et al., 2012). In a representative sample of the general adult population of Catalonia, Spain, aged 18–74 years, concentrations of PCBs 28, 52, 101, 118, 128, 153 and 180 were analysed in serum from blood samples obtained during the health examination of the Catalan Health Survey. Participants were considered to have diabetes if their fasting plasma glucose level was ≥126 mg/dL, they answered affirmatively to the question, “Do you have or did a doctor tell you that you have diabetes?” or they reported a current use of insulin or antidiabetic medication. The prevalence of diabetes increased in a dose-dependent manner across concentration quartiles of PCBs 138, 153 and 180, as well as the DL-PCB, PCB 118 (Gasull et al., 2012). In a hospital-based cross-sectional study among adults (>16 years) from Granada, Spain, concentrations of PCBs 138, 153 and 180 were measured in serum and adipose tissue samples. The prevalence of diabetes was assessed using both self-reported information and clinical records. In multiple logistic models adjusted for age, sex, BMI and adipose tissue of origin, no significant associations were found between the prevalence of diabetes and the concentrations of PCB 138, PCB 153 or PCB 180 in adipose tissue (Arrebola et al., 2013). (ii) Prospective studies

A case–control study was performed within a well-defined cohort of women aged 50–59 years (the Women’s Health In the Lund Area cohort) from the southern part of Sweden. Several biomarkers, including PCB 153 concentration, were analysed in stored serum samples from blood collected at the baseline

Supplement 1: Non-dioxin-like polychlorinated biphenyls

examination (1995–2000). Women with previously confirmed diabetes or with prevalent diabetes according to plasma glucose level or oral glucose tolerance test assessed at recruitment were excluded from this study. By linkage with the Swedish inpatient and outpatient registers, 410 women from the cohort who had developed type 2 diabetes before 31 December 2006 were identified (cases). One control per case was randomly selected from the cohort and matched for age, calendar year, BMI and the presence or absence of any features of metabolic syndrome. No association was found for the risk of diabetes in relation to PCB 153 exposure; the OR estimated by means of conditional logistic regression (adjusted for the matching variables) was 0.99 (95% CI 0.71–1.40), comparing women with PCB 153 concentrations in serum above 1790 µg/mL (the 75th percentile) with women with PCB 153 concentrations in serum at or below 1790 µg/mL. However, this association tended to increase with the time of follow-up; for instance, when only cases (and their matched controls) diagnosed more than 7 years after the baseline examination were included, the OR was 1.6 (95% CI 0.61–4.0) (Rignell-Hydbom et al., 2009). The risk of type 2 diabetes was investigated within the Coronary Artery Risk Development in Young Adults (CARDIA) cohort, with participants at baseline (1985–1986) in four centres from the USA (Birmingham, Alabama; Chicago, Illinois; Minneapolis, Minnesota; and Oakland, California); followup examinations were completed at years 2, 5, 7, 10, 15 and 20 (2005–2006). Individuals were eligible for the current study if they had had no diagnosis of diabetes at years 0 and 2 and if they had been diagnosed with type 2 diabetes at any subsequent examination. Diabetes was defined as ever having taken antidiabetic medications or ever having had a fasting glucose level of ≥126 mg/ dL at two or more examinations during the 18 years of follow-up. Concentrations of a total of 35 PCB congeners were measured in blood samples collected at baseline. No significant ORs (adjusted for age, sex, race, BMI, and triglyceride and total cholesterol levels at year 2) were observed across quartiles of any of the 12 congeners analysed individually (including PCB 153) or different groups of PCBs (Lee et al., 2010). In the population-based Prospective Investigation of the Vasculature in Uppsala Seniors (PIVUS) study in elderly subjects (70 years of age) from Uppsala, Sweden, concentrations of 14 PCBs were measured in plasma collected at baseline (2000–2004). During 2006–2009, when the subjects turned 75 years of age, reinvestigation of the cohort was performed; using this information, type 2 diabetes was defined as a fasting blood glucose level above 6.2 mmol/L or the use of insulin or oral hypoglycaemic agents. After adjustment for sex, BMI, cigarette smoking, exercise, alcohol consumption, triglyceride level and total cholesterol level, there was a significant trend of increasing risk of type 2 diabetes across quintiles of concentration of PCBs 74, 99, 194 and 206. In contrast, significant 145

Safety evaluation of certain food additives and contaminants Eightieth JECFA

ORs were found when comparing the risk of diabetes in subjects in the fifth quintile with those in the first quintile of PCB 174 and PCB 180 concentrations, as well as for the fourth concentration quartile for PCB 153 (OR 3.4, 95% CI 1.1–10.2), PCB 138 (OR 5.9, 95% CI 1.6–21.7) and PCB 170 (OR 5.0, 95% CI 1.2–20.5); in these cases, non-significantly increased risks were also found in the fifth quintile. No significant ORs were reported for the four DL-PCBs analysed (PCBs 105, 118, 156 and 157) (Lee et al., 2011a). Concentrations of the four most abundant PCB congeners (PCBs 118, 138, 153 and 180) were measured in plasma from blood drawn in 1989–1990 from 1095 women who were free of diabetes and participated in two nested case–control studies in the Nurses’ Health Study, with cases identified through June 2008. Diagnosis of diabetes was based upon self-reported information on symptoms, diagnostic tests and treatment, using criteria of the American Diabetes Association (1997) to confirm the cases. No significant increase in risk of type 2 diabetes was observed across tertiles of any of the congeners analysed. The OR for the third versus the first tertile of lipid-standardized plasma concentration of PCB 153, adjusted for age, smoking status, alcohol intake, physical activity, family history of diabetes and baseline BMI, was 1.36 (95% CI 0.59–3.10). Furthermore, no significant association for any of the four congeners was observed after pooling the results of these two case–control studies with those of six published prospective studies including 842 diabetes cases in total (Wu et al., 2013). (iii) Type 1 diabetes

146

WHO Food Additives Series No. 71-S1, 2016

All the studies reviewed so far (either cross-sectional or prospective studies) dealt with adult or type 2 diabetes. Only one study was identified on the relationship between PCB concentrations and type 1 diabetes. The study was performed as a case–control study within a biobank in Malmö, southern Sweden, and included 150 cases (children who had their diagnosis mostly before 18 years of age) and 150 controls, matched for sex and day of birth. The concentration of PCB 153 was measured in venous blood samples from mothers and umbilical cord blood samples obtained routinely at the hospital maternity unit. Cases were children born from 1970 to 1990 who developed type 1 diabetes before the year 2002. The vast majority (88%) were diagnosed before the age of 18 years and were diagnosed and treated at the Department of Paediatrics at Malmö University Hospital. Patients who developed type 1 diabetes between 18 and 27 years of age were identified through the Diabetes Incidence Study in Sweden registry. When comparing the quartile with the highest concentrations of PCB 153 in maternal serum with the lowest quartile, an OR of 0.64 (95% CI 0.32–1.29) was obtained (Rignell-Hydbom et al., 2010).

Supplement 1: Non-dioxin-like polychlorinated biphenyls

(c) Obesity and metabolic-related conditions (i) Obesity in adults: prospective studies

A two-generation study was carried out in 11 communities of Lake Michigan, USA; prenatal exposure to PCBs was estimated from maternal blood collected at delivery, and anthropometric measurements were taken in the female offspring aged 20–50 years. Prenatal PCB concentrations were not associated with height, weight or BMI in female offspring. However, no details about individual congeners were provided, and therefore no relevant information can be derived on the relationship between NDL-PCBs and BMI (Karmaus et al., 2009). In the CARDIA cohort in four centres from the USA described above, measurements of 22 PCBs in blood samples taken at baseline were compared with BMI and other factors related to insulin resistance after 20 years of followup in subjects free of diabetes. There was a significant (or marginally significant) trend to a decrease in BMI across quartiles of concentration of PCB 194, PCBs 196 + 203 and PCB 209, adjusted for age, sex, race, triglyceride level and total cholesterol level. No associations were observed for PCB 153 or PCB 180 (Lee et al., 2011b). In the population-based PIVUS study in Uppsala, Sweden, described above, concentrations of 14 PCBs measured in plasma collected at baseline from elderly subjects (70 years of age) were compared with waist circumference after a 5-year follow-up. There was a non-significant trend to a decreased waist circumference across quintiles of individual PCB concentrations in men, but there were significant associations in women for PCBs 170, 180, 194 and 106, as well as for two DL-PCB congeners, PCBs 156 and 189. For PCB 153, the OR for the fifth versus the first quintile, adjusted for calorie intake, exercise, cigarette smoking, alcohol consumption, triglyceride level and total cholesterol level, was 0.3 (95% CI 0.1–1.1) (Lee et al., 2012). (ii) Obesity in children: prospective studies

In a population-based birth cohort study (Asthma Multicentre Infants Cohort Study [AMICS]–INMA) from Menorca, Spain, the concentrations of seven PCB congeners (PCBs 28, 52, 101, 118, 138, 153 and 180) in cord blood were compared for those overweight at 6.5 years, defined as a BMI z-score at or above the 85th percentile of the WHO reference values. There was a significant increase in relative risk (RR) of overweight for the third compared with the first tertile of the sum of the concentrations of the seven PCB congeners, adjusted for birth weight, previous parity, maternal pre-pregnancy BMI, maternal education and social class at pregnancy, maternal smoking in pregnancy, maternal age at delivery and breastfeeding. This RR became non-significant when additionally adjusted for concentrations of HCB and DDT/DDE in cord blood. No results for individual 147

Safety evaluation of certain food additives and contaminants Eightieth JECFA

congeners were provided (Valvi et al., 2012). Although it is likely that the highest contribution to the sum of PCBs in this study comes from NDL-PCB congeners, the group includes PCB 118, a DL-PCB congener. Therefore, these results cannot be linked to NDL-PCB exposure. A birth cohort formed in the Faroe Islands was established in 1997– 2000. Prenatal exposure to PCBs was determined from maternal concentrations of PCBs in pregnancy serum (blood obtained at gestational week 34) and milk. Total NDL-PCBs was defined as the sum of the concentrations of PCB 138, PCB 153 and PCB 180. Subsequent BMI was measured in children at 5 and 7 years of age. No associations were observed between PCB concentrations and BMI in boys, 5-year-old girls or 7-year-old girls with normal weight mothers. For 7-yearold girls who had overweight mothers, PCB concentrations were associated with increased BMI (β 2.07, P-value = 0.007) and with an increase in BMI from 5 to 7 years of age (β 1.23, P-value = 0.003). PCB concentration was associated with increased waist circumference in girls with overweight mothers (β 2.48, P-value = 0.001) and normal weight mothers (β 1.25, P-value = 0.04) (Tang-Péronard et al., 2014). (d) Insulin resistance and metabolic syndrome

Cross-sectional analyses were carried out in non-diabetic participants from the NHANES to assess the relationship between metabolic syndrome (Lee et al., 2007a) and insulin resistance (Lee et al., 2007b) and concentrations of five NDL-PCBs (PCBs 138, 153, 170, 180 and 187) and four DL-PCBs in serum. Metabolic syndrome was diagnosed if a subject satisfied at least three of the following criteria: waist circumference >102 cm (men) or >88 cm (women), triglyceride level ≥1.7 mmol/L, high-density lipoprotein (HDL)–cholesterol level <1.1 nmol/L (men) or <1.4 mmol/L (women), average blood pressure ≥130/85 or currently taking hypertension medication, and fasting serum glucose level ≥5.6 mmol/L. The NDL-PCBs showed a significant, inverse, U-shaped association, with adjusted quartile ORs of 1.0, 1.3, 1.8 and 1.0, with a significant quadratic term. Insulin resistance was estimated using the homeostasis model assessment method, calculated as fasting insulin [mU/L] × fasting glucose [mmol/L]/22.5). Adjusted prevalence (homeostasis model assessment insulin resistance above its 90th percentile) was significantly associated with higher concentrations of PCB 170 and PCB 187 in serum; no associations were observed with the concentrations of other PCB congeners (PCBs 138, 153 and 180). In the CARDIA cohort study described above for BMI, the relationship between the concentrations of 22 PCBs at baseline and insulin resistance after 20 years of follow-up was also analysed prospectively in subjects free of diabetes. No significant associations were observed for increased homeostasis model

148

WHO Food Additives Series No. 71-S1, 2016

Supplement 1: Non-dioxin-like polychlorinated biphenyls

assessment insulin resistance across quartiles of concentration of any of the NDL-PCBs measured, including PCBs 138, 153 and 180 (Lee et al., 2011b). 2.3.8 Reproductive and sexual function

In female Yucheng (see section 2.3.2) adolescents aged 13–19 years, there was a higher rate of irregular menstrual cycle, shorter mean duration of bleeding and higher levels of estradiol compared with controls (Yang et al., 2005). However, no significant differences in menstrual cycle characteristics or age at menarche were reported between adult Yucheng women (mean age 39 years) and agematched controls (Yang et al., 2011). In boys born to exposed Yucheng women, the estradiol levels in serum at puberty (>12 years) were higher than in controls, and testosterone levels and FSH activities were lower than in controls (Hsu et al., 2005). In Yucheng women, a significant increase in time to pregnancy and a significantly higher infertility ratio compared with controls were observed (Yang et al., 2008). For Yusho patients (see section 2.3.2) during the first 10-year follow-up, there were higher rates of induced and spontaneous abortions and preterm delivery compared with controls (Tsukimori et al., 2008), and a low sex ratio (proportion of male births) was reported in the second generation of Yusho women (Tsukimori et al., 2012). Levels of free testosterone, FSH, LH and sex hormone binding globulin (SHBG) were not significantly associated with the sum of the concentrations of all PCBs in serum from male Great Lakes fish consumers (Turyk et al., 2006). In 9-year-old girls from New York, USA, breast development was reduced among those with a higher sum of the concentrations of four PCB congeners (PCBs 118, 138, 153 and 180) and low (below median) BMI (Wolff et al., 2008). In offspring of a female cohort in Michigan, USA, there was an increased but non-significant OR for a male birth for higher parental exposure to PCBs (measured as Aroclor 1254) (Terrell et al., 2009). However, secondary sex ratio was not associated with occupational exposure to PCBs in female workers in three electrical capacitor manufacturing plants in the USA (Rocheleau et al., 2011). In a Russian study, higher PCB concentration (sum of 41 congeners) in maternal serum was associated with earlier pubertal onset in 8- to 9-year-old boys followed for 4 years (Humblet et al., 2011). In middle-aged (41–55 years) males from three regions in Slovakia, no significant correlation was observed between concentrations of testosterone and total PCBs (sum of 15 congeners) in serum (Langer et al., 2012). In a birth cohort in Duisburg, Germany, the association between maternal exposure to PCBs (sum of 18 congeners) and behavioural sexual dimorphism was examined using Pre-school Activities Inventory (PSAI), consisting of 24 items grouped into three categories: preferred toys (e.g. guns, dolls), preferred activities (e.g. playing house, fighting) and behavioural characteristics (e.g. enjoys rough and tumble play, likes pretty 149

Safety evaluation of certain food additives and contaminants Eightieth JECFA

things). Exposure to dioxins and PCBs in boys was associated with more feminine behaviour, whereas exposure in girls was associated with less feminine behaviour (Winneke et al., 2014). In all these studies, exposure to PCBs involved complex mixtures of PCB congeners, often unknown, but in most instances including both DL- and NDLPCB congeners; moreover, results for individual congeners were not reported. Therefore, these studies do not provide any relevant information regarding the potential health effects of NDL-PCBs. (a) Male reproductive function (i) Cross-sectional studies

Fishermen aged 27–67 years from the west and east coasts of Sweden provided semen and blood for the analysis of hormones and PCB 153. The proportion of Y and X chromosome–bearing sperm in semen samples was determined by fluorescence in situ hybridization (FISH) analysis. The log-transformed lipidadjusted PCB 153 concentration was significantly (P-value = 0.05) associated with the Y chromosome fraction (β 0.42, 95% CI 0.01–0.83) (Tiido et al., 2005). In a cross-sectional study of men seeking infertility evaluation from the Massachusetts General Hospital, Boston, Massachusetts, USA, FISH for chromosomes X and Y was used to assess sex chromosome disomy in sperm nuclei, and serum was analysed for concentrations of 57 PCB congeners. Significant increases in the rate of XY and total sex chromosome disomy and a decrease in XX disomy were observed across quartiles for the sum of estrogenic PCBs (PCBs 44, 49, 52, 101, 187, 174, 177 and 157 + 201) (McAuliffe et al., 2012). Semen samples from Faroese men were obtained from three cohorts. The UM cohort consisted of young volunteers from the general population who had been born in 1981–1984; the K1S cohort was generated from consecutive births at the three Faroese hospitals in 1986–1987; and the K5P cohort consisted of men whose pregnant partner participated in a study focusing on fertility and environmental factors. Concentrations of PCBs 138, 153 and 180 were measured in the serum from blood collected from 499 male subjects. The Y:X ratio (measured in semen by FISH) was not significantly associated with the concentration of PCBs (sum of log-transformed concentrations of three congeners multiplied by 2) in serum in the total cohort. However, a statistically significant negative association was found in the K5P subcohort (P-value = 0.030) (Kvist et al., 2014). A cross-sectional study involving Inuit from Greenland, Swedish fishermen and men from Warsaw, Poland, and Kharkiv, Ukraine, with a total of 707 adult males aged 18–67 years, was carried out. Semen and blood samples were collected between 2002 and 2004 for the measurement of concentrations of PCB 153 in serum and several indicators or markers of male reproductive function.

150

WHO Food Additives Series No. 71-S1, 2016

Supplement 1: Non-dioxin-like polychlorinated biphenyls

A sperm chromatin structure assay was used to assess sperm DNA/chromatin integrity, and a strong and monotonically increasing DNA fragmentation index was observed with increasing concentrations of PCB 153 in serum among European but not Inuit men, reaching a 60% higher average level in the highestexposure group (Spanò et al., 2005). No significant association was found with sex hormones, but there were differential results across centres: in Kharkiv, statistically significant positive associations were found between concentration of PCB 153 and SHBG and LH levels; in Greenland, there was a positive association with LH level (Giwercman et al., 2006). As for the relationship between PCB 153 concentration in serum and indicators of epididymal and accessory gland function (measured in seminal plasma), a negative association with the activity of neutral α-glucosidase was found in the aggregated cohort, a positive association was observed between PCB 153 and prostate-specific antigen as well as zinc level among Kharkiv men, and a negative association was found between PCB 153 and fructose level in the Swedish cohort (Elzanaty et al., 2006). In relation to sperm quality, sperm motility decreased across all four regions, on average by 3.6% (95% CI 1.7–5.6%) per one unit increase in the log of PCB 153 concentration in blood (ng/g lipid). No associations were found for sperm concentration or the proportion of morphologically normal sperm (Toft et al., 2006). In the Swedish cohort, log-transformed PCB 153 concentration was significantly positively associated with Y chromosome fractions (P-value = 0.04), whereas in the Polish cohort, PCB 153 concentration correlated negatively with the proportion of Y-bearing fraction of spermatozoa (P-value = 0.008) (Tiido et al., 2006). Finally, an article summarizing the main results from several publications on these four populations concluded that (1) an association between high PCB 153 concentrations in serum and low sperm counts was detected within a subgroup of men with short androgen receptor CAG repeat length; (2) a relationship between increased PCB 153 concentrations in serum and decreased sperm motility was seen in all four studied regions, and indications of reduced neutral α-glucosidase activity in seminal plasma point to a post-testicular effect; and (3) in Greenlandic Inuit, impairment of sperm chromatin integrity was related to PCB 153 concentration. Despite these effects, fertility in terms of time taken to conceive was not related to PCB 153 concentration in serum, except in Inuit (Bonde et al., 2008). (ii) Prospective and case–control studies

All men aged 50–80 years who visited a group of five urologists for various conditions from 1997 to 1999 in Kingston, Ontario, Canada, were potentially eligible for this study. Cases were 101 men with a clinical diagnosis of erectile dysfunction at the time of study enrolment; controls comprised all 151

Safety evaluation of certain food additives and contaminants Eightieth JECFA

other participants who were not suspected on a clinical basis to have erectile dysfunction (234 men) and who had been diagnosed with a variety of benign urological conditions. Concentrations of a total of 14 PCB congeners (PCBs 28, 52, 99, 101, 105, 118, 128, 138, 153, 156, 170, 180, 183 and 187) were measured. No associations were found between erectile dysfunction and concentrations of any of the congeners analysed (Polsky et al., 2007). In a birth cohort formed in 1986–1987 in the Faroe Islands, prenatal exposure was determined from maternal concentrations of PCBs in pregnancy serum (blood obtained at gestational week 34) and milk; total NDL-PCB concentration was defined as the sum of the concentrations of PCB 138, PCB 153 and PCB 180. At age 14 years, a total of 438 boys participated in a clinical examination to assess Tanner stage for puberty development (based upon pubic hair and genital development), and a venous blood sample was obtained to measure the concentrations of sex hormones and binding proteins in serum. Regression analysis was carried out with LH, testosterone and SHBG as dependent variables, with adjustment for age, BMI, history of cryptorchidism and time of blood sampling. The free fraction of testosterone, as expressed by the testosterone/SHBG ratio, and LH were inversely related to the concentration of PCB 153 (P-values = 0.016 and 0.023, respectively), whereas SHBG level was positively associated with the concentration of PCB 153 (P-value = 0.003). After additional adjustment for Tanner stage, only the result for SHBG remained statistically significant (β 0.08, P-value = 0.03). As both variables were log transformed, this value corresponds to an increase in SHBG level (nmol/mL) by about 6% for each doubling in the concurrent PCB exposure (Grandjean et al., 2012). A case–control study was carried out within a screening study during 2002–2004 in two hospitals from Nice, France. Diagnosis of cryptorchidism was accepted after at least two concordant examinations by a senior paediatrician before discharge; babies were followed at 3 and 12 months to confirm the diagnosis. In a total of 78 cases and 86 controls, concentrations of seven PCB congeners (the six indicator PCBs and PCB 118) were measured in cord blood and maternal milk. No significant differences were observed between median concentrations (in cord blood or milk) of any of the measured PCBs in cases and controls. In the analysis using logistic regression after adjustment for gestational age, birth weight, maternal BMI before pregnancy, maternal age, parity, season of birth, paternal history of cryptorchidism and city of delivery, the OR for cryptorchidism at birth, comparing above with below sum of the median concentrations of all measured PCBs in maternal milk, was 2.74 (95% CI 1.15–6.53); however, the OR became non-significant for cryptorchidism at 3 months. Moreover, it should be taken into account that total PCBs included six NDL-PCBs plus the DL-PCB congener, PCB 118 (Brucker-Davis et al., 2008).

152

WHO Food Additives Series No. 71-S1, 2016

Supplement 1: Non-dioxin-like polychlorinated biphenyls

The Collaborative Perinatal Project enrolled pregnant women in 12 United States medical centres during 1959–1965. For the present study, thirdtrimester blood samples from mothers and 230 sons with cryptorchidism, 201 with hypospadias and 593 with neither condition were analysed to measure concentrations of 11 PCB congeners in serum (McGlynn et al., 2009b). In the analysis by logistic regression, adjusting for study centre and DDE concentration in serum, no significant ORs were observed between either cryptorchidism or hypospadias and higher concentrations of any of the PCBs measured. A significant increase of 69% in risk of hypospadias was associated with the highest versus the lowest quartile of the concentration of all PCBs. However, it must be kept in mind that the group of 11 measured PCBs included two DL-PCBs (PCBs 15 and 118). A study was carried out on 80 children with hypospadias and 80 healthy controls, recruited during 2005–2007 in two hospitals in Rome, Italy, where children with hypospadias are surgically treated. Cases were defined as children aged 0–24 months who were suffering from any form of hypospadias that required surgical treatment, and controls were defined as healthy male children aged 0–24 months without any congenital defect who were attending the Outpatient Vaccination Service of the same hospital. Data on occupational and dietary exposures to polychlorinated compounds (without further specification) in the perinatal period were collected; maternal exposure to four PCB congeners (PCBs 118, 130, 153 and 180) was also ascertained in blood samples taken 8–12 months after delivery from a subset of cases and controls. No association was observed with any of the measured PCBs (Giordano et al., 2010). It should be noted that exposure was measured several months after the outcome; although concentrations of persistent compounds in serum can be an indicator of past exposure, this study cannot be strictly seen as a prospective study. (b) Female reproductive function

A cross-sectional analysis was carried out among Akwesasne Mohawk Nation girls, 10–16 years of age, resident in the proximity of the St Lawrence River in New York State, USA, and in Ontario and Quebec, Canada. Blood samples and sociodemographic data were collected, and attainment of menses (menarche) was assessed as present or absent at the time of the interview. The 16 PCBs detected in more than 50% of the samples were grouped (Wolff ’s groups). The OR for attainment of menarche was 8.39 (P-value = 0.04) for an increase in one unit in the log scale of the sum of the concentrations of four PCBs in the Wolff ’s group of estrogenic PCBs (PCBs 52, 70, 101 and 187) (Denham et al., 2005). A study was carried out among women born between 1960 and 1963 in the Oakland area, California, USA; these women and their mothers were among 153

Safety evaluation of certain food additives and contaminants Eightieth JECFA

participants in the Child Health and Development Study. PCBs were measured in mothers’ preserved serum samples collected 1–3 days after their daughters’ birth, and time to pregnancy in daughters 28–31 years later was recorded. The fecundability ratio (ratio of probability of pregnancy in each cycle for daughters exposed to higher versus lower PCB concentrations in utero) was estimated by Cox regression; a fecundability ratio above 1 indicates a greater probability of pregnancy per cycle and therefore a shorter time to pregnancy. Increasing concentrations of PCB 99 and PCB 187 were significantly associated with higher infertility (ORs of 0.75 and 0.44, respectively, for 1 standard deviation of concentration in serum), whereas higher concentrations of PCB 138 were associated with higher probability of pregnancy (OR of 1.99 [95% CI 1.45–2.75] for 1 standard deviation of concentration in serum). Significant associations were also found for the DL-PCB congeners, PCBs 105 and 156 (Cohn et al., 2011). Analysis of mixtures of PCBs may be problematic, as components have a complex correlation structure, and, along with limited sample sizes, standard regression strategies are problematic. A novel weighted quantile sum approach indicated that the dominant functionality groups associated with longer time to pregnancy were the dioxin-like, anti-estrogenic group (average weight, 22%). In contrast, in the association with shorter time to pregnancy, the anti-estrogenic group and the phenobarbital inducers group (phenobarbital, CYP1A and CYP2B inducers) played a more important role (60% and 23%, respectively) (Gennings et al., 2013). (c) Effects on reproductive and sexual function assessed in both males and females

A cross-sectional study was carried out in Akwesasne Mohawk adults, aged 18–95 years, resident along the St Lawrence River near the junction of New York State, USA, and Ontario and Quebec, Canada, for at least 5 years. Fasting serum samples were analysed for 101 PCB congeners, as well as testosterone, cholesterol and triglycerides. Testosterone levels in males were significantly and inversely related to concentrations of PCBs 74, 99 and 153, whereas testosterone levels were much lower in females than in males and not significantly related to concentrations of PCBs in serum. A significant inverse association was also found between testosterone level and the sum of the concentrations of DL-PCB congeners (Goncharov et al., 2009). In a birth cohort study initiated in the year 2000, healthy mother–infant pairs were recruited in Duisburg, Germany. Concentrations of PCBs (the six indicator PCBs, PCBs 28, 52, 101, 138, 153 and 180, and 12 DL-PCB congeners) were measured in maternal blood during pregnancy (32 weeks) and in maternal milk. Testosterone and estradiol levels were measured in maternal and cord serum of a subsample of mother–infant pairs. Linear regression analysis was used to describe the association of PCB concentrations in maternal blood or milk

154

WHO Food Additives Series No. 71-S1, 2016

Supplement 1: Non-dioxin-like polychlorinated biphenyls

with sex steroid concentrations after adjustment for age of mother, nationality of mother, BMI of mother before pregnancy, length of gestation, birth weight, alcohol consumption during pregnancy and smoking during pregnancy. There was a significant reduction in testosterone level in cord serum of girls related to the sum of the concentrations of the six indicator PCBs in maternal blood, but not in boys; among girls, the change in geometric mean of hormone concentrations with a doubling of the sum of the concentrations of the six indicator PCBs, expressed as a mean ratio, was 0.76 (95% CI 0.61–0.96) (Cao et al., 2008). In the Child Health and Development Study described above, serum specimens were collected from pregnant women from the San Francisco Bay area, California, USA. Concentrations of a total of 11 PCBs were determined in specimens collected during the second or third trimester of gestation, and the secondary sex ratio, or sex ratio at birth, was evaluated as a function of maternal serum concentrations. There were inverse associations between the risk of a male birth and concentrations of most NDL-PCBs measured, although none of them was significant. The RRs for a male birth for a given increase in PCB concentration from the 10th to the 90th percentile, adjusted for specimen characteristics and an indicator of a prescription for sex steroids, oral contraceptives or corticosteroids, were 0.82, 0.82 and 0.85 for PCB congeners 138, 153 and 180, respectively. There was a significantly decreased risk for a male birth associated with the sum of the concentrations of all measured PCBs (PCB congeners 105, 110, 118, 137, 138, 153, 170, 180 and 187), which includes both NDL- and DL-PCB congeners (Hertz-Picciotto et al., 2008). In a cohort of pregnant women 25–34 years of age and newborns from a general population in central Taiwan, China, concentrations of 27 PCBs were measured in placental tissue. Children were examined at 8 years of age, and bone age, sex hormone concentrations and indicators of reproductive development, including Tanner, breast, genital and armpit stages, were assessed. No significant association was observed between the concentration of indicator NDL-PCBs (sum of PCBs 138, 153 and 180) and any indicators of reproductive development, except for a significant association with lower fundus length of the uterus for in utero exposure to increased concentrations of indicator PCBs in girls; the OR for high versus low (according to the median) PCB concentrations and fundus length above versus below the median was 0.08 (95% CI 0.01–0.83). In the same study, no significant associations were found for dioxin compounds (Su et al., 2012). 2.3.9 Immune function and related outcomes

This section includes a diversity of outcomes, including thymus size, levels of immunoglobulins (Igs) in blood, blood leukocyte counts and other outcomes. 155

Safety evaluation of certain food additives and contaminants Eightieth JECFA

Infections would be included in this section as well, but as all the studies dealing with infections and PCB exposure addressed respiratory infections, they have been reported in section 2.3.11 below, together with other respiratory effects. Second-grade schoolchildren, aged 7–10 years, were recruited in the state of Hesse, Germany. Several immune function biomarkers were measured in blood, together with concentrations of eight PCBs (PCBs 110, 118, 138, 153, 170, 180, 183 and 187) in serum. In a cross-sectional analysis, the sum of PCB concentrations was significantly associated with increased IgM levels (Karmaus et al., 2005). No results for individual congeners were provided. Although it is likely that the highest contribution to the sum of PCB concentrations comes from NDL-PCB congeners, the group includes PCB 118, a DL-PCB congener. Therefore, these results cannot be linked to NDL-PCB exposure. In mothers recruited in 2002–2004, at the time of delivery, from two districts in eastern Slovakia (Michalovce and Svidnik), blood samples were collected at delivery for measurement of the concentrations of 15 PCBs (PCBs 28, 52, 101, 105, 114, 118, 123, 138, 153, 156, 157, 167, 170, 180 and 189). Thymus volume was measured in infants shortly after birth and at ages 6 and 16 months using ultrasonography. Prenatal PCB exposure was associated with a smaller thymic index at birth (Park et al., 2008), whereas PCB concentration was not predictive of 6- or 16-month thymus volume (Jusko et al., 2012). The exposure refers to a mixture of PCBs, including DL- and NDL-PCBs, and therefore these results cannot be linked to NDL-PCB exposure. In this same cohort from Slovakia, in addition to maternal blood, cord blood and blood from infants at 6 months were collected. IgG, IgA and IgM levels were measured in the blood collected at 6 months. PCB concentrations in maternal, cord or 6-month infant blood were not associated with total immunoglobulin levels in serum at 6 months, regardless of the timing of PCB exposure, PCB congener or specific immunoglobulin. Congener-specific results were not reported (Jusko et al., 2011). Concentrations of IgG-specific anti-Haemophilus influenzae type b, tetanus toxoid and diphtheria toxoid were assayed in 6-month infant sera using enzyme-linked immunosorbent assay (ELISA) methods. Overall, there was no evidence of an association between infant antibody concentrations and PCB measures during the prenatal and early postnatal period. Specific results were reported for PCB 138, PCB 153, PCB 170 and PCB 180, with no significant results for any of them (Jusko et al., 2010). Two birth cohorts were formed in the Faroe Islands: in 1994–1995, a cohort of births was established from consecutive spontaneous singleton births at term; and during 1999–2001, an additional birth cohort was formed using similar criteria. Prenatal exposure was determined from concentrations of PCBs in pregnancy serum (blood obtained at gestational week 34) and maternal milk; total NDL-PCB concentration was defined as the sum of the concentrations

156

WHO Food Additives Series No. 71-S1, 2016

Supplement 1: Non-dioxin-like polychlorinated biphenyls

of PCB 138, PCB 153 and PCB 180. Following routine childhood vaccinations against tetanus and diphtheria, 119 children were examined at 18 months and 7 years of age, and their serum samples were analysed for tetanus and diphtheria toxoid antibodies. The antibody response to diphtheria toxoid decreased at 18 months by 24.4% (95% CI 1.63–41.9) for each doubling of the cumulative PCB exposure, but there was no association at age 7 years. The tetanus toxoid antibody response was affected mainly at age 7 years, decreasing by 16.5% (95% CI 1.51–29.3) for each doubling of the prenatal exposure, whereas no association was observed at 18 months (Heilmann et al., 2006). In the subcohort 1999–2001, further assessment of antibodies was performed at age 5 years. The risk for an anti-diphtheria antibody concentration below a clinically protective level (0.1 IU/L) increased by about 3% for a doubling of the concentration of NDL-PCBs in milk or 18-month serum; the ORs were 1.34 (95% CI 1.11–1.61) and 1.30 (95% CI 1.08–1.57), respectively. No significant associations were found for antitetanus toxoid (Heilmann et al., 2010). Moreover, no association was reported between prenatal concentrations of NDL-PCBs and IgE level, allergy or atopic dermatitis at age 7 years (Grandjean et al., 2010). The potential relationship between PCB exposure and immune function markers has also been assessed in adults. A cross-sectional analysis was carried out in a cohort of individuals aged 70 years from Uppsala, Sweden. Levels of protein complement 3 (C3), 3a (C3a), 4 (C4) and C3a/C3 ratio and 16 PCBs were measured in serum from blood samples collected at recruitment. No association was observed between any of the protein complements and the NDL-PCB congeners analysed individually, including PCBs 138, 153, 170 and 180 (Kumar et al., 2014). Plasma PCB concentrations (36 congeners) and immune function natural killer cell cytotoxicity and phytohaemagglutinin-induced T-lymphocyte proliferation were measured at baseline and 1 year later in postmenopausal, overweight women in the Seattle area, Washington, USA. Higher concentrations of moderately and highly chlorinated PCBs (all of them NDL-PCBs) were associated with higher phytohaemagglutinin-induced T-lymphocyte proliferation crosssectionally, but not longitudinally. No association was observed with natural killer cell cytotoxicity (Spector et al., 2014). 2.3.10 Cardiovascular outcomes

Blood pressure, or specifically hypertension, is the cardiovascular outcome or condition most often assessed in relation to PCB exposure. Other cardiovascularrelated risk factors or outcomes analysed are carotid atherosclerosis, left ventricular systolic and diastolic function, and serum lipids. Unfortunately, the potential relationship with PCB exposure is always evaluated by means of a crosssectional approach (the exposure is related to the prevalence, not to the incidence 157

Safety evaluation of certain food additives and contaminants Eightieth JECFA

of the disease). In this design, both the exposure (PCB level) and the outcome (i.e. hypertension) are evaluated at the same time, so there is not a clear time sequence of events; it could be the case that the disease onset took place before the exposure. These studies at most can suggest an association, but they provide only weak support for a potential causal relationship. (a) Blood pressure or prevalence of hypertension

Cross-sectional association between the concentrations of 11 PCB congeners (including six NDL-PCBs) in serum and the prevalence of hypertension was investigated in subjects from the NHANES 1999–2002; hypertension was defined as having systolic blood pressure (SBP) of 140 mmHg (18.7 kPa) and above or diastolic blood pressure (DBP) of 90 mmHg (12.0 kPa) and above (subjects with hypertension treatment were excluded). NDL-PCBs tended towards a positive association among men, with a significant trend across quartiles for the sum of the concentrations of the six PCB congeners (PCBs 99, 138, 153, 170, 180 and 197) and for the concentration of PCB 170; however, none of the ORs for the highest versus the lowest quartile was significant. No association was observed among women (Ha et al., 2009). In the NHANES 1999–2004, serum measurements included 22 PCB congeners, including the six indicator PCBs. In addition to the blood pressure measurement (SBP ≥140 mmHg or DBP ≥90 mmHg), the definition of prevalent hypertension included a previous diagnosis of hypertension by a doctor or being under treatment for hypertension. In a cross-sectional analysis for individual congeners, a positive association was found for PCB 52, PCB 101 and PCB 153, but not for PCB 138, PCB 170 or PCB 180. Correlation and multicollinearity among PCB congeners were evaluated, clustering analyses were performed to determine groups of related congeners and a weighted sum was constructed to represent the relative importance of each congener. Using this approach to equalize different ranges and potencies, PCBs 66, 101, 118, 128 and 187 were significantly associated with hypertension (Yorita Christensen & White, 2011). In Aniston, Alabama, USA, the site of a PCB production plant until 1971, concentrations of 37 PCBs were measured in serum in 2005–2007 in a sample of individuals over 18 years of age, together with other lifestyle and healthrelated data, including the presence of hypertension (SBP ≥140 mmHg or DBP ≥90 mmHg, or being on antihypertensive treatment). Hypertension prevalence was found to be positively associated with the sum of the concentrations of all measured PCBs. No details for individual congeners were reported, and therefore this result cannot be linked to NDL-PCB exposure (Goncharov et al., 2010). However, an analysis of the subsample of these subjects who were not taking antihypertensive medication provided results on specific congeners. Multiple

158

WHO Food Additives Series No. 71-S1, 2016

Supplement 1: Non-dioxin-like polychlorinated biphenyls

linear regression analysis of log-transformed SBP and DBP was used to compare means of blood pressure across tertiles of individual congeners; a significant increase of the mean SBP and DBP was observed across tertiles for PCB 138, PCB 153 and PCB 180 (Goncharov et al., 2011). Concentrations of 15 PCB congeners were measured in plasma of adults (≥18 years) living in 13 Inuit villages in Greenland, together with other healthrelated data gathered by interview and anthropometric and blood pressure measurements. Hypertension was defined as having SBP of 140 mmHg and above or DBP of 90 mmHg and above. No association was found between the prevalence of hypertension and the sum of the concentrations of the 12 NDL-PCB congeners analysed or the concentration of any of the individual congeners, including the six indicator congeners. It must be noted that this is a highly exposed population: for PCB 153, the geometric mean concentration in plasma was 504 ng/g lipid (range 478–530 ng/g lipid) (Valera et al., 2013a). In Inuit from Nunavik, Quebec, Canada, concentrations of 14 PCBs were measured in plasma samples from adults (≥18 years); hypertension was defined as having SBP of 140 mmHg or above, having DBP of 90 mmHg or above or being on antihypertensive medication. The prevalence of hypertension was significantly associated with higher sum of the concentrations of nine NDL-PCBs as well as with higher concentrations of PCBs 101, 138 and 187 (but not PCB 153, 170 or 180) in plasma; in this population, the level of exposure was lower than for the previous one, with a geometric mean concentration of 283.6 ng/g lipid for PCB 153 in serum (Valera et al., 2013b). Blood pressure and several factors related to hypertension were recorded for a representative sample of the adult population, aged ≥18 years, in the Canary Islands, Spain, together with concentrations of 18 PCB congeners in serum samples. In a cross-sectional analysis, the association of hypertension (SBP ≥140 mmHg, DBP ≥90 mmHg or use of antihypertensive medication) in relation to groups of PCBs or individual congeners, adjusted for potential risk factors for hypertension, was investigated. No association was observed between the prevalence of hypertension and the sum of the concentrations of the NDL-PCBs or the concentration of PCB 153 or 180 in serum (Henríquez-Hernández et al., 2014). In the study from Aniston, Alabama, USA, described above, in addition to blood pressure, concentrations of lipids in serum were measured as well. In a cross-sectional analysis, a significant positive association was found between concentrations of PCBs 28, 138, 153, 170 and 180 and concentrations of total lipids, triglycerides and total cholesterol in serum, but not concentrations of HDL- or low-density lipoprotein (LDL)–cholesterol (Aminov et al., 2013).

159

Safety evaluation of certain food additives and contaminants Eightieth JECFA

(b) Other cardiovascular outcomes

In the population-based PIVUS study in seniors 70 years of age from Uppsala, Sweden, the prevalence of carotid artery plaques was determined by ultrasound, and the intima-media thickness (IMT) and grey-scale median of the intimamedia complex (IM-GSM) were measured; concentrations of 16 PCBs in serum were also measured. In a cross-sectional analysis, concentrations of PCB 138, PCB 153, PCB 170 and PCB 180 were significantly associated with higher prevalence of carotid artery plaques, but no association was observed with IMT or IM-GSM. PCB 206 and PCB 209 had significant positive associations with higher prevalence of carotid artery plaques, higher values of IMT and lower values of IM-GSM (Lind et al., 2012). In the same study, cross-sectional analysis was carried out to assess the potential association between PCB concentrations and left ventricular dysfunction. Concentrations of PCBs 138, 153, 170 and 180 were significantly associated with lower ventricular ejection function, whereas no association was found between any of these congeners and the ratio between the Doppler transmitral amplitude of waves E and A (E/A ratio) and the isovolumetric relaxation time (Lind et al., 2012). The only vascular outcome assessed in a prospective study was stroke. Incident ischaemic or haemorrhagic strokes were diagnosed during a 12-year follow-up in the Swedish Mammography Cohort, a population-based prospective cohort of women established in 1987–1990. Estimates of dietary PCB exposure were obtained via a food frequency questionnaire administered at baseline that had been validated against PCB concentrations in serum from women from the cohort. A significant increase in risk for all kinds of stroke was observed with increased dietary PCB exposure (Bergkvist et al., 2014). Unfortunately, in this study, no results for individual congeners or groups of PCBs were provided, and therefore these results cannot be linked to specific NDL-PCB exposures. WHO Food Additives Series No. 71-S1, 2016 160

2.3.11 Respiratory outcomes

Prenatal exposure to PCBs (PCBs 28, 52, 101, 118, 153, 138 and 180) was assessed in cord serum of 405 children from Menorca, Spain, and asthma (current wheezing, persistent wheezing or doctor-diagnosed asthma) was assessed at 4 years of age. No association was found between overall exposure to PCBs and asthma (Sunyer et al., 2005). In a subset of these children, prenatal exposure to PCBs was not associated with inflammatory cytokines (Gascon et al., 2014a). No results for individual congeners were provided. Although it is likely that the highest contribution to the sum of PCBs in this study comes from NDL-PCB congeners, the group includes PCB 118, a DL-PCB congener. Therefore, these results cannot be linked to NDL-PCB exposure.

Supplement 1: Non-dioxin-like polychlorinated biphenyls

In the birth cohort from the Faroe Islands (1994–1995) described above, prenatal exposure was determined from concentrations of NDL-PCBs in maternal serum in gestational week 34 and in maternal milk (sum of PCB 138, PCB 153 and PCB 180). No significant association was reported between prenatal concentrations of NDL-PCBs and asthma at age 7 years (Grandjean et al., 2010). Children from the BraMat, a subcohort of the Norwegian Mother and Child Cohort Study, were followed during the first 3 years of life using annual questionnaires, and blood parameters were examined at 3 years of age. The questionnaire included the occurrence of respiratory disorders, infections, allergy and immune-related outcomes. The maternal dietary exposure to PCBs was estimated using a validated food frequency questionnaire and available data on concentrations of PCBs in Norwegian foods. The occurrence of wheeze during the first year of life was significantly associated with maternal exposure to NDLPCBs; the adjusted OR for a dietary exposure at or above the 80th percentile was 2.79 (95% CI 1.20–6.49) (Stølevik et al., 2011). In the extended follow-up up to 3 years of age, the association between maternal NDL-PCB exposure and wheeze remained (adjusted OR for dietary exposure at or above the 80th percentile was 4.03, 95% CI 1.59–10.2), but there was no association with asthma (Stølevik et al., 2013). Concentrations of PCBs (PCBs 118, 138, 153, 156, 170 and 180) were measured in maternal serum collected at the 30th gestational week in a birth cohort formed in 1988–1989 in Aarhus, Denmark. Risk of asthma in offspring after 20 years of follow-up was obtained from a national registry. Weak (nonsignificant) associations were found between incidence of offspring asthma and maternal exposure to NDL-PCBs (sum of concentrations of PCBs 138, 153, 170 and 180 in serum). The hazard ratio (HR) for the highest (compared with the lowest) tertile of NDL-PCB concentration was 1.30 (95% CI 0.78–2.17) (Hansen et al., 2014). A meta-analysis was carried out in 4422 mothers and children enrolled in nine birth cohort studies from six European countries with available data on PCB 153 concentration in cord serum, as well as parent-reported bronchitis or wheeze during the first 4 years of life. Overall, there was no association between PCB 153 exposure and occurrence of bronchitis or wheeze; the RR for the highest versus the lowest PCB 153 concentrations was 0.95 (95% CI 0.75–1.21). Within the nine studies included in the meta-analysis, there was a significant increase in risk for two cohorts, a significant decrease in risk for one cohort and no significant association for the remaining six cohorts. In contrast, a significant association was observed for the occurrence of bronchitis before age 18 months with PCB 153 concentration measured on a continuous scale; the RR for doubling the concentration of PCB 153 was 1.06 (95% CI 1.01–1.12). However, the association was not significant when comparing the third with the first tertile (RR 1.17, 95% 161

Safety evaluation of certain food additives and contaminants Eightieth JECFA

162

WHO Food Additives Series No. 71-S1, 2016

CI 0.97–1.14). The occurrence of wheeze (either before or after age 18 months) was not associated with PCB 153 concentration in serum (Gascon et al., 2014b). A general population-based birth cohort was established in three Spanish regions (INMA study). Maternal blood was obtained between the seventh and 26th weeks of pregnancy, and concentrations of PCB congeners 28, 118, 138, 153 and 180 in serum were obtained; parental reports on lower respiratory tract infection were obtained when children were 12–14 months old. No association was observed between the concentration of PCBs (sum of the five congeners) and the occurrence of lower respiratory tract infection (Sunyer et al., 2010; Gascon et al., 2012). No results for individual congeners were provided. Although it is likely that the highest contribution to the sum of PCBs comes from NDL-PCB congeners, the group includes PCB 118, a DL-PCB congener. Therefore, these results cannot be linked to NDL-PCB exposure. In the Norwegian cohort BraMat described above, maternal dietary exposure to NDL-PCBs was associated with infections of the upper respiratory tract. The OR for the number of upper respiratory tract infections diagnosed between ages 0 and 3 years was 1.04 (95% CI 1.01–1.08) for each nanogram per kilogram of body weight of daily exposure (Stølevik et al., 2013). In a cohort of newborns from 14 Inuit communities of Nunavik, Quebec, Canada, PCB 153 concentration was measured in umbilical cord plasma; occurrence of respiratory infections from 0 to 5 years of age was assessed by a review of medical history. Exposure to PCB 153 was significantly associated with acute otitis media and upper respiratory tract infection, but not with lower respiratory tract infection. The RRs for fourth quartile PCB 153 concentrations (compared with the first quartile) and acute otitis media and upper respiratory tract infection were, respectively, 1.37 (95% CI 1.20–1.55) and 1.44 (95% CI 1.20– 1.72), whereas for lower respiratory tract infection, the RR was 1.09 (95% CI 0.97–1.24) (Dallaire et al., 2006). Maternal serum during pregnancy (weeks 32–34) and mothers’ milk during the third week after pregnancy from primiparous women from Uppsala, Sweden, were used to assess perinatal exposure to 10 PCB congeners (four DL-PCBs and six NDL-PCBs); when the infant was 3 months old, respiratory infections and other health outcomes were assessed by interview with the mother. No significantly increased risk of respiratory infections was observed in relation to postnatal exposure to NDL-PCBs. Regarding prenatal exposure, there was no association with PCB 153 concentration, but the third tertile of the sum of the concentrations of PCBs 28, 52 and 101, compared with the first tertile, had an adjusted RR of 3.4 (95% CI 1.4–7.8) (Glynn et al., 2008).

Supplement 1: Non-dioxin-like polychlorinated biphenyls

2.3.12 Hepatic effects

In an analysis of a 24-year follow-up among Yucheng patients, mortality from liver cancer disease and cirrhosis was increased for men, but not for women (Tsai et al., 2007). As mentioned in section 2.3.2, these subjects were exposed to PCDFs and to a complex mixture of PCBs, and therefore these effects cannot be specifically linked to NDL-PCBs. Alanine aminotransferase function, an indicator of non-alcoholic fatty liver disease, and concentrations of PCBs in serum were assessed in adults from the NHANES 2003–2004. In a cross-sectional analysis, a significant association was found between increased levels of alanine aminotransferase and higher concentrations of NDL-PCBs (sum of 27 congeners), as well as with concentrations of PCB 138 and PCB 153 (Cave et al., 2010). 2.3.13 Musculoskeletal effects

Among Yucheng patients followed for 24 years, there was a significant increase in risk of dying from diseases of the musculoskeletal system and systemic lupus erythematosus in women, but not in men (Tsai et al., 2007). As already mentioned, these effects cannot be specifically linked to NDL-PCBs. Cross-sectional analyses have addressed the potential relationship between concentrations of some NDL-PCBs and bone mineral density and bone metabolism markers in adult or elderly populations in Sweden and the USA. Forearm bone mineral density and serum markers of osteoblastic (osteocalcin) and osteoclastic (CrossLaps) functions were measured in men and women, median age 59 and 62 years, respectively, from the east (Baltic) coast of Sweden. After correction for age and BMI, no association was observed between serum concentrations of PCB 153 and bone mineral density, osteocalcin or CrossLaps (Wallin et al., 2005). In a subset of women from this population, concentrations of hydroxylated PCB metabolites in serum were measured for some PCB congeners (4-hydroxy-PCB 107, 4-hydroxy-PCB 146, 4-hydroxy-PCB 187); no association was observed between bone mineral density and any of the hydroxyPCB metabolites (Weiss et al., 2006). Bone mineral density was measured in a population 60–81 years of age living near the Baltic coast of Sweden, close to a river contaminated by PCBs and heavy metals (nickel, cadmium). The sum of the concentrations of the three most abundant NDL-PCBs (PCBs 138, 153 and 180) in serum was not associated with low bone mineral density (Hodgson et al., 2008). In adults (≥20 years old) from the NHANES 1999–2004, bone mineral density was not found to be associated with concentrations of PCB 153 or PCB 138 in serum, but PCB concentrations modified the association between bone mineral density and fat mass or lean mass (Cho et al., 2011). 163

Safety evaluation of certain food additives and contaminants Eightieth JECFA

Cross-sectional association of concentrations of five NDL-PCBs (PCBs 138, 153, 170, 180 and 187) in serum with the prevalence of self-reported arthritis was assessed in adults (≥20 years old) from the NHANES 1999–2002. There was a significant association between the sum of the concentrations of the five NDL-PCBs and prevalence of arthritis in women, but no clear association was observed in men. The association among females was stronger for rheumatoid arthritis; compared with the lowest quartile of PCB concentrations in serum, the ORs for the third and fourth quartiles were, respectively, 4.4 and 5.4, both statistically significant, with a significant trend (P-value for trend < 0.01) (Lee, Steffes & Jacobs, 2007). 2.3.14 Endometriosis

Concentrations of 40 PCB congeners were measured in serum from infertile Japanese women examined by laparoscopy and diagnosed as either cases with endometriosis or controls. No association was observed with concentrations of total PCBs in serum (Tsukino et al., 2005). Similarly, concentrations of 62 PCB congeners were measured in serum in women undergoing laparoscopy in Buffalo, New York, USA. No association was observed with concentrations of total PCBs or subgroups of congeners (estrogenic or anti-estrogenic) in serum (Louis et al., 2005). Serum samples from a female cohort in Michigan, USA, were collected at baseline, and PCBs were quantified as Aroclor 1254. The incidence of endometriosis was higher in women with higher concentrations of PCBs in serum, but the association was not significant (Hoffman et al., 2007). In these studies, the exposure referred to a mixture of PCBs, including DL- and NDLPCB congeners, without specific information on individual congeners; therefore, no relevant conclusions concerning the effect of NDL-PCBs can be drawn. Several case–control studies on the relationship between endometriosis and NDL-PCB exposure have been published in India, Italy and the USA. Detectable concentrations of four NDL-PCBs (PCBs 1, 5, 29 and 98) were measured in serum collected from infertile women from southern India who underwent transvaginal ultrasound scan screening followed by laparoscopy; 85 were classified as cases with endometriosis, whereas 135 were classified as controls. The mean concentration of the four PCBs was significantly higher in cases than in controls; ORs were not reported (Reddy et al., 2006). Women living in the Rome, Italy, area and undergoing laparoscopy for suspected endometriosis or other benign gynaecological conditions provided a blood sample before laparoscopy, and concentrations of 11 PCB congeners were measured in serum. Cases were 40 women with histologically confirmed endometriosis; controls were 40 women with benign gynaecological conditions with no evidence of endometriosis. The adjusted OR for women in the third

164

WHO Food Additives Series No. 71-S1, 2016

Supplement 1: Non-dioxin-like polychlorinated biphenyls

tertile of PCB 153 concentration compared with the first tertile was 9.1 (95% CI 1.9–43). An increased risk of endometriosis was also reported for increasing concentrations of PCBs 138 and 180 (Porpora et al., 2006). The period of recruitment was extended up to the inclusion of 80 cases of endometriosis and 78 controls. In this analysis, the adjusted ORs for women in the third tertile of PCB 138, PCB 153 and PCB 180 concentration were, respectively, 3.78 (95% CI 1.60– 8.94), 4.88 (95% CI 2.01–11.0) and 3.05 (95% CI 1.25–7.42) (Porpora et al., 2009). In a subset of these patients, a gene–environment interaction was observed for GSTP1 Ile/Ile and GSTM1 null genotypes, modulating the effect of PCB 153 and PCB 180 (Vichi et al., 2012). Among women 20–45 years of age seeking reproductive assistance in Atlanta, Georgia, USA, 60 women with recently diagnosed endometriosis by biopsy and laparoscopy were defined as cases and were compared with 60 controls; in all of them, concentrations of 13 PCB congeners were measured in serum. No significant differences in median concentrations of PCB 138, PCB 153 or PCB 180 were observed; ORs for NDL-PCB congeners were not reported (Niskar et al., 2009). In a case–control study in Washington State, USA, concentrations of 20 PCB congeners were measured in serum from 251 surgically confirmed endometriosis cases that were newly diagnosed in 1996–2001 and from 538 controls matched for age and reference year. None of the six indicator PCBs (PCBs 28, 52, 138, 153, 170 and 180) was associated with an increased risk of endometriosis. The adjusted OR for a unit increase in the natural log–transformed concentration of PCB 153 was 1.1 (95% CI 0.8–1.8) (Trabert et al., 2010). 2.3.15 Other health effects

The cross-sectional relationship between concentrations of 14 PCB congeners in serum and some molecules sensitive to oxidative stress was assessed in Inuit adults. The PCB concentration in serum was associated with the LDL oxidation (Bélanger et al., 2006). This result refers to the whole group of PCBs; no specific results for NDL-PCBs were reported. Concentrations of 20 PCB congeners and inflammatory markers were assessed in Japanese children 0–4 years of age. In cross-sectional analysis, IL-8 was significantly correlated with PCB congeners 163 + 162, 177 and 180 + 193, whereas COX-2 correlations with individual congener levels were recognized only among control subjects, not among asthmatic subjects. No significant results were reported for PCB 153 (Tsuji et al., 2012). 2.3.16 Summary of epidemiological studies

A number of potential health effects associated with exposure to NDL-PCBs have been identified, including changes in thyroid hormone homeostasis, 165

Safety evaluation of certain food additives and contaminants Eightieth JECFA

166

WHO Food Additives Series No. 71-S1, 2016

neurodevelopmental effects, immunological effects and some types of cancer. Some of the results offer support for the toxicological findings. The results of prospective (Chevrier et al., 2007; Herbtsman et al., 2008; Darnerud et al., 2010) and cross-sectional studies in newborns (Takser et al., 2005; Wang et al., 2005; Herbstman et al., 2008) and children (Álvarez-Pedrerol et al., 2008; Schell et al., 2008) suggest that increasing concentrations of NDL-PCBs are correlated with lower levels of T4 and higher levels of TSH in blood, although there are some inconsistencies between results across studies. The key human studies are summarized in Tables 15 and 16. Maternal and early postnatal exposure to NDL-PCBs in some birth cohorts was also associated with impaired behavioural, cognitive and psychomotor development (Stewart et al., 2005; Park et al., 2010; Forns et al., 2012b; Lynch et al., 2012; Gascon et al., 2013; Tatsuta et al., 2014) and with alteration of VEPs (Saint-Amour et al., 2006). Perinatal exposure to NDL-PCBs in birth cohorts was found to be associated with increased incidence of acute respiratory infections in children (Dallaire et al., 2006; Glynn et al., 2008; Stølevik et al., 2013). Regarding cancer, the recent evaluation by IARC (2015) reported an association between melanoma and PCB exposure, mainly based upon cohort studies of exposed workers in various industries, for whom exposure would be by multiple routes. IARC (2015) also considered studies in the general population with different study designs. Only one population-based case–control study (Gallagher et al., 2011) reported specific results for NDL-PCBs, showing a significantly increased risk of melanoma for a group of 11 NDL-PCBs, as well as for some individual congeners. In this study, a similar increased risk was also observed for two DL-PCB congeners. The association between NDL-PCBs and NHL has also been assessed in several prospective cohorts (L.S. Engel et al., 2007; Bertrand et al., 2010; Laden et al., 2010; Bräuner et al., 2012), but the results were not consistent.

Supplement 1: Non-dioxin-like polychlorinated biphenyls

Table 15 Prospective studies on the association between NDL-PCBs and thyroid hormones in newborns and infants Location and period Salinas Valley, California, USA 1999–2000 (CHAMACOS cohort) Blood/serum concentration Subjects (ng/g lipid) Birth cohort: 285 34 PCBs in pairs of pregnant maternal serum Mexican Ameri- (26th week) can women and PCB 153: newborns GM (95% CI) 5.6 (5.2–6.0) Range 0.3–85.6

Reference Chevrier et al. (2007)

Outcome TSH in serum of children taken shortly after birth (mIU/L) GM (95% CI) 5.7 (5.3–6.1)

Main results % increase in TSH (95% CI) For 10-fold increase in PCB PCB 101: 23% (7–45%) PCB 138: 23% (2–51%) PCB 153: 20% (0–48%) PCB 180: 23% (2–48%) Inducers of CYP2B: 29% (2–58%) (PCBs 99, 153, 180, 183)

Herbstman Johns Hopkins et al. (2008) Hospital, Baltimore, Maryland, USA 2004–2005

Birth cohort: 92 infants born by spontaneous, unassisted delivery

Umbilical cord blood PCBs measured: 74, 99, 118, 138 + 158, 153, 180 PCB 153: Mean (SD) 6.8 (2.1)

Darnerud et Uppsala region, al. (2010) Sweden 1996–1999 (POPUP cohort)

Birth cohort: 160 primiparous women and 150 infants

10 PCBs in maternal blood (32nd–34th weeks) PCB 153: Median (range) 65 (23–158)

Total T4 in blood of children taken at average of 18 days of age (range 5–117 days) Total T4: Mean (SD) 15.16 (3.92) (µg/dL) TSH, total T3, free T4 in children 3 weeks and 3 months after birth

Comments Adjusted for neonatal age at blood collection, gestational age at birth, birth weight and mother’s pre-pregnancy BMI PCBs 118 and 156 (DL-PCBs): no significant association OR (95% CI) comparing Adjusted for sex, the lowest 20% of total gestational age, T4 concentration to the maternal age, rest, for 1-unit ln PCB maternal race, maconcentration: ternal pregnancy PCB 138: 5.30 BMI, smoking and (1.73–16.21) days since birth PCB 153: 3.40 (1.31–8.83) PCB 118 (DL-PCB): PCB 180: 1.89 (0.83–4.30) significantly Estimated lowest 20% of associated total T4: 11.86 µg/dL PCBs 138, 153, 180 Adjusted for birth inversely associated weight, sex, alcohol with total T3 at 3 weeks consumption and (P-value = 0.048) age of mother Non-significant, positive (total T3); adjusted association with TSH at 3 for sex (TSH) months (P-value = 0.093) No significant association for free T4 or for the sum of DL-PCBs

CI: confidence interval; GM: geometric mean; IU: International Units; SD: standard deviation; T3: triiodothyronine; T4: thyroxine; TSH: thyroid stimulating hormone

167

Safety evaluation of certain food additives and contaminants Eightieth JECFA

Table 16 Cross-sectional analyses on the association between NDL-PCBs and thyroid hormones in newborns and children Reference Wang et al. (2005) Location Central Taiwan, China Subjects 118 newborns Exposure measurements 27 PCBs in cord blood Outcome Free T4, T3, TSH, TBG Main results Significant positive correlation of free T4 with sum of PCBs 138, 153 and 180 only among female neonates No associations for any hormone with any PCB Comments No associations between T3, TSH or TBG and any of the PCBs measured

Takser et al. (2005)

Quebec, Canada 92 newborns

14 PCBs in cord blood 5 PCBs in cord blood

Free T4, total T3, TSH Free T4, free T3, TSH

Maervoet et Antwerp, al. (2007) Belgium

198 newborns

Herbstman Johns Hopkins et al. (2008) Hospital, Baltimore, Maryland, USA

289 newborns

6 PCBs in cord blood

Free T4, total T4, TSH

ÁlvarezPedrerol et al. (2008)

Birth cohort 259 children Menorca, Spain aged 4 years

7 PCBs in serum (28, 52, 101, 138, 153, 180)

Free T4, total T3, TSH

WHO Food Additives Series No. 71-S1, 2016

Schell et al. (2008)

USA/Canada

232 Akwesasne Mohawk youth, 13 years old

18 PCBs in serum

Free T4, total T4, total T3, TSH

The 14 PCBs included the six indicator PCBs No associations Significant inverse associations of PCBs 138, between TSH and 170, 180 with free T4 and any of the PCBs free T3 and between PCB measured (118, 128, 153, 170 and 153 and free T4 180) No associations in the No associations whole population. Among between TSH and infants born by spontane- any of the PCBs ous unassisted delivery, measured (74, 99, significant inverse associ- 118, 138 + 158, ations between free T4 and 153 and 180) total T4 and PCBs 138 (+ 158), 153 and 180 Significant inverse asso- No associations ciations between total T3 between free T4 or and PCBs 138 and 153 TSH and any of the PCBs measured Significant association between total T3 and PCB 118 (DL-PCB) No associations Significant positive associations between TSH between total T3 or and PCB 153 and PCBs 138 total T4 and any of the PCBs measured + 163 + 164 Significant inverse associ- Significant positive ations between free T4 and association PCBs 52, 101 + 90, 153 between TSH and and 138 + 163 + 164 PCB 118 (DL-PCB)

T3: triiodothyronine; T4: thyroxine; TBG: thyroxine-binding globulin; TSH: thyroid stimulating hormone

168

3. Analytical methods 3.1

Chemistry

The general structure of PCBs and the numbering system for individual PCB congeners have been described in section 1.2. PCBs are anthropogenic compounds that were produced commercially in considerable amounts between the 1930s and 1970s and were used for a wide range of applications. PCBs can also be unintentionally produced and released from many industrial thermal processes, including waste incineration and metallurgical processes (Nie et al., 2012; Liu et al., 2013). They are chemically resistant to hydrolysis and oxidation, but are subject to photolysis under some conditions (Hutzinger, Safe & Zitko, 1974). Owing to their overall chemical stability and their resistance to biological transformation, PCBs were widely used as heat transfer and hydraulic fluids, solvent extenders, flame retardants, organic diluents, paints and dielectric fluids (Hutzinger, Safe & Zitko, 1974; Jones, 1988). Technical/commercial PCB formulations were marketed worldwide under different names: they were sold as Aroclors in North America and the United Kingdom, Clophens in Germany, Chlorofen in Poland, Delors in Slovakia, Kanechlors and Santotherm in Japan, Phenochlor and Pyralène in France, Sovol mainly in Russia and Fenclor in Italy (Hutzinger, Safe & Zitko, 1974; Noma et al., 2006; Burkhard & Lukasewycz, 2008; Mandalakis et al., 2008). The chlorine content varied between different formulations and in some cases differed between lots (e.g. Aroclor 1254 lot A4 versus lot G4) (Hansen, 1999). Although there are 209 possible PCB congeners, only approximately 130 have been reported in commercial mixtures (Fattore et al., 2008; Elabbas et al., 2013). The physicochemical properties of PCBs vary between individual PCB congeners. PCBs are liquids or solids at room temperature, appear colourless to pale yellow and are odourless (EFSA, 2010). PCBs have relatively low vapour pressures, with volatility decreasing with increasing degrees of chlorination, and have limited solubility in water (Fischer, Wittlinger & Ballschmiter, 1992; Shiu & Ma, 2000). PCBs are thermally stable, persist in the environment and are transported large distances beyond their area of release (Safe, 1994). PCBs are also lipophilic compounds and accumulate in the tissues of living organisms (Safe, 1994; Jankovic et al., 2011; Klincic et al., 2014); they are taken up by humans primarily through the consumption of food, with foods of animal origin being the primary source of human exposure (Henríquez-Hernández et al., 2011). In general, individual laboratories restrict their analysis to a subset of the 209 PCB congeners. A key challenge associated with the analysis of PCBs originates in the decision associated with which congeners should be considered 169

Safety evaluation of certain food additives and contaminants Eightieth JECFA

to be representative of the NDL-PCBs. The analysis frequently involves the separate determination of both DL- and NDL-PCB congeners. Unfortunately, there is no universal approach to the selection of the NDL-PCB congeners to be measured. As the determination of all 209 individual congeners is not routinely performed, owing to the onerous nature of isolating each congener, laboratories have developed their own suite of congeners to represent total PCBs (DL-PCBs + NDL-PCBs) and more recently total NDL-PCBs (Boumphrey et al., 1993; Behnisch et al., 1997; Atuma et al., 1998; Ayotte et al., 2005). Lack of comparability between laboratories reporting total PCB or total NDL-PCB concentrations can thus be challenging, as well as complicating toxicological assessment (Bordajandi et al., 2003). Owing to the high degree of effort required to measure the full suite of PCB congeners, some laboratories use technical mixtures or combinations of them (e.g. Aroclor 1254:1260) for quantification of PCB concentrations in unknown samples (Bhavsar et al., 2007), whereas others have established a subset of individual congeners to represent total PCBs (Jones, 1988; Zuccato et al., 1999; Baars et al., 2004; Bloom et al., 2005). One suite of congeners that has been used is composed of seven marker PCBs (i.e. PCBs 28, 52, 101, 118, 138, 153 and 180), selected because they were identified as representing approximately 50% of the total PCBs detected in food (Arnich et al., 2009). In human plasma samples, it was found that the sum of PCB congeners 138, 153 and 180, multiplied by a factor of 2, resulted in a concentration equal to 50% of the total PCBs in the samples (Morck et al., 2014). In another study, the sum of the concentrations of the seven marker PCBs (PCBs 28, 52, 101, 118, 138, 153 and 180) in eel liver and muscle samples represented 22% and 29% of total PCB concentrations, respectively (Oliveira Ribeiro et al., 2008). The use of this suite of the seven marker congeners seemed to result in variability in concentration relative to the total PCBs being reported. Unfortunately, this subset also includes PCB 118, which has dioxinlike activity and therefore cannot be used for reporting NDL-PCB concentrations (Chewe et al., 1997; Baars et al., 2004; Kim et al., 2004; Zuccato et al., 2008; Arnich et al., 2009; Blanchet-Letrouvé et al., 2014). Other congener subsets have been suggested for use as indicator PCBs because these congeners have been observed in technical mixtures and other sources of PCBs, such as municipal solid waste, emission gases and indoor air (Ishikawa et al., 2007). One example is the subset consisting of PCBs 3, 8, 28, 52, 77, 101, 105, 118, 126, 138, 153, 180, 194, 206 and 209. Again, however, this proposed set of congeners includes congeners with dioxin-like activity. A smaller subset (n = 4) of congeners (PCBs 118, 138, 153 and 180) has also been proposed as the suite for use in tracking body burdens (Axelrad, Goodman & Woodruff, 2009). Ultimately, a subset of congeners established to represent exclusively NDL-PCBs has been developed and includes the six indicator PCBs, PCBs 28, 52, 101, 138, 153 and 180 (Cerveny et al., 2014).

170

WHO Food Additives Series No. 71-S1, 2016

Supplement 1: Non-dioxin-like polychlorinated biphenyls

Current publications increasingly report this subset in the literature (Traag et al., 2006; Blanchet-Letrouvé et al., 2014). These six indicator PCBs have been selected not because they all have similar toxicological actions, but because they represent approximately 50% of total NDL-PCBs measured in analysed samples (Cimenci et al., 2013). Although indicator PCBs are used for reporting concentrations in the scientific literature and in some regulations (e.g. European Union: European Commission, 2011; and China: Ministry of Health of the People’s Republic of China, 2012, 2014; Shao et al., 2014), there are some national guidelines that continue to be based on total PCB concentrations, including those of Japan, the USA and Canada (currently under review) (Japan External Trade Organization, 2011; Health Canada, 2012; USFDA, 2014). 3.2 Description of analytical methods 3.2.1 Introduction

The methodology for the analysis of PCBs has evolved with advances in the technology since the initial research reported in the early 1970s. Some of the early analysis of PCBs was performed using PCB technical mixtures, prior to the availability of standards of the individual congeners, and this practice has continued over the years in some laboratories (Webb & McCall, 1973; Boonyathumanondh et al., 1995). Separation of PCBs was initially achieved using packed columns (Albro, Corbett & Schroeder, 1981). PCB separation improved as the column type changed: from megabore columns to support-coated open tubular columns, with capillary columns being adopted universally (Sissons & Welti, 1971; Frame, 1997). Although multidimensional gas chromatography (GC × GC) has been applied to PCB analysis, multiple columns with different polarities have been used in the separation of congeners and aid in the identification of co-eluting congeners (Duinker & Hillebrand, 1979; Albro et al., 1981; de Vos et al., 2011). The synthesis of individual congeners and their availability for use as analytical standards, coupled with the capillary column use for separation of individual congeners, have allowed laboratories to determine and report individual PCB congeners and total PCB concentrations as a function of the sum of the individual congeners (Mullin et al., 1984). Owing to the thermal stability of PCBs, analyses have been performed primarily using GC for the separation of congeners combined with a variety of detection methods. Nuclear magnetic resonance spectroscopy was also employed in the effort to identify congener-specific information (Sissons & Welti, 1971). Webb & McCall (1973) established the relative chlorine content of individual Aroclor mixtures through the application of a gas chromatograph coupled 171

Safety evaluation of certain food additives and contaminants Eightieth JECFA

to a Coulson conductivity detector, in addition to employing electron capture detection (ECD) and mass spectrometry (Webb & McCall, 1973). All 209 PCB congeners were synthesized and confirmed using nuclear magnetic resonance spectroscopy, and relative elution order for each of the congeners was reported in 1984 (Mullin et al., 1984). Although ECD has been used consistently for the detection and quantification of PCBs in technical mixture formulations and extracts of biotic and abiotic matrices, additional PCB analysis has been performed using flame ionization detection (Schulz, Petrick & Duinker, 1989). Mass spectrometry was used in the original identification of PCB congeners and continues to be used in the analysis of these compounds at present, owing to the selectivity and specificity of this technique (Sissons & Welti, 1971; Loutfy et al., 2008; Masci, Orban & Nevigato, 2015). 3.2.2 Screening tests

WHO Food Additives Series No. 71-S1, 2016 172

In some of the early work, PCB confirmation of quantified samples was obtained by using pre-coated thin-layer chromatography plates (Mes & Davies, 1979). Comprehensive two-dimensional GC was also used for rapid PCB analysis of human serum (Patterson et al., 1996; Liem, 1999). The analysis of coplanar PCBs (DL-PCBs) has been performed using rapid methodologies developed for screening, such as the DR CALUX assay (Hoogenboom et al., 2006a,c). There are few screening tests for the DL-PCBs, because they are present at very low concentrations, and purification of extracts and isolation of these compounds are time consuming and expensive. The analysis of NDL-PCBs generally does not require as extensive a cleanup procedure as for the DL-PCBs or PCDDs/PCDFs; thus, routine screening methods are not generally reported in the literature. ELISAbased detection methods have been reported for PCBs, following standard sample preparation techniques for POP analysis, and the results obtained compared well with concentrations determined using traditional detection methods (Galloway et al., 2002). 3.2.3 Quantitative methods

Quantitative methods for the analysis of PCBs were initially developed using solvent extraction coupled to GC separation of peaks and analysis using ECD. The quantification was completed by comparing the results in unknown samples relative to established technical mixtures (e.g. Aroclor, Kanechlor), and this practice has continued in some laboratories (Bhavsar et al., 2007), although the transition to congener-specific analytical standards has been more frequently the approach taken (Boonyathumanondh et al., 1995; Gill, Schwartz & Wheatley, 1995; Polder et al., 1998; Eljarrat et al., 2001; Bordajandi et al., 2003; Fernandes et al., 2004; Kim et al., 2004; Bloom et al., 2005; Gomara et al., 2005; Costopoulou et al.,

Supplement 1: Non-dioxin-like polychlorinated biphenyls

2006; Jursa et al., 2006; Storelli et al., 2006; Malisch & Dilara, 2007; Esteve-Turrillas et al., 2008; Loutfy et al., 2008; Arnich et al., 2009; Turrio-Baldassarri et al., 2009; Grassi et al., 2010; Jankovic et al., 2011; Cimenci et al., 2013; Julshamn et al., 2013; Blanchet-Letrouvé et al., 2014; Cerveny et al., 2014). The general steps required for the determination of NDL-PCBs in food and other matrices (e.g. environmentally relevant organisms, human tissues and fluids, environmental compartments) are to extract them from the matrix in which they are to be determined, followed by isolation and purification of the prepared extracts, with the effort culminating in analysis. In some cases, samples have been freeze-dried before initiation of sample extraction, which is frequently done with food samples (Fernandes et al., 2004; Cimenci et al., 2013). Prior to the initiation of an extraction of PCBs from a matrix, PCB analogue analytical standards – whether they are stable isotope (13C) analogues of analytes of interest or native (12C) PCB congeners known to be absent from the matrix under investigation – are added to the sample (Dewailly et al., 1991; Boumphrey et al., 1993; Borgå et al., 2005; Weiss, Papke & Bergman, 2005; Ingelido et al., 2007). Known quantities of these standards are added for use as surrogates to aid in the correction for any losses during the sample preparation. These compounds have properties similar to those of NDL-PCBs because they are structural analogues and they behave similarly to the analytes themselves, and any losses of the analogues that may occur at any stage of the sample preparation can be determined and used to correct for the unknown PCB concentrations. Extraction of PCBs generally requires organic solvents combined with some mixing procedure to ensure that the sample is exposed to the maximum solvent possible to enhance partitioning into the solvent from the original matrix. The methods employed include solvent extraction using liquid–liquid partitioning (Weistrand & Noren, 1993; Storelli et al., 2012b) or rotation of homogenized samples with solvent (Bhavsar et al., 2007). Food and other samples can be ground with anhydrous sodium sulfate and other adsorbents (e.g. silica gel, Florisil) in advance of the extraction step to dry the samples and bind any analytes and added to columns with or without prior shaking followed by elution from the column using solvent, which is the method known as matrix solidphase dispersion (Criado et al., 2004; Fernandes et al., 2004; Bordajandi, Abad & Gonzalez, 2008). In the case of samples with very high lipid content (e.g. melted butter, oil) or liquid samples, solvents can be added directly to the sample and allowed to sit to maximize the PCB partitioning into the solvent phase (Criado et al., 2004; Weiss, Papke & Bergman, 2005; Cimenci et al., 2013). Saponification using sodium or potassium hydroxide/ethanol has also been performed for these matrices (Tuinstra, Traag & Keukens, 1980; Chou, Chen & Li, 2004). Similar to matrix solid-phase dispersion, the rationale for homogenization of samples with the solvents is to maximize the ratio of the sample surface area to solvent exposure (Elskus et al., 1994; Rawn et al., 2012). Soxhlet extraction, which uses 173

Safety evaluation of certain food additives and contaminants Eightieth JECFA

repeated cycles of hot solvent, is another effective method for extracting PCBs from samples and is sometimes performed using automated Soxhlet systems (e.g. Soxtec) (Chu, Covaci & Schepens, 2003; Grassi et al., 2010; Carro et al., 2012). Additionally, ultrasonication has been used to extract PCBs from non-food organic material (e.g. graphite, humic material) (Abrha & Raghavan, 2000). Increasingly, the development of automated methods for sample preparation is being undertaken. These methods may involve grinding samples and drying reagents prior to extraction, which is performed using hot, pressurized solvents in repeated cycles using a pressurized solvent extractor or accelerated solvent extraction systems, for example (Esteve-Turrillas et al., 2008; Cimenci et al., 2013). Although supercritical fluid extraction was investigated in lipid-rich samples (e.g. herring), it is not routinely utilized in PCB analysis (van der Velde et al., 1996). In addition to these automated techniques, the application of microwave-assisted extraction has been used for the extraction of PCBs, particularly from sediments and water, although it has been tested using blubber (Camel, 2000; Sparr Eskilsson & Björklund, 2000; Fujita et al., 2009). Filtration of samples is required following microwave-assisted extraction prior to performance of the needed extensive cleanup, owing to the exhaustive nature of this method (Fidalgo-Used, Blanco-González & Sanz-Medel, 2007). The application of newer technologies allows analysts to use lower solvent volumes and decrease the sample preparation time prior to analysis. A critical parameter in the successful extraction of PCBs, including NDL-PCBs, is the selection of appropriate solvents. Although PCBs are nonpolar, hydrophobic compounds, extraction from food, biological samples and abiotic environmental samples is generally accomplished using a mixture of polar and non-polar solvents in variable ratios. Some of the popular choices include acetone:hexane (Chu, Covaci & Schepens, 2003; Gomara et al., 2005) and dichloromethane:hexane (Guruge et al., 2005; Bjermo et al., 2013), although toluene:acetone (Blanchet-Letrouvé et al., 2014), cold light petroleum:acetone (Cirillo et al., 2008), diethyl ether:petroleum ether (Esposito et al., 2014), chloroform:methanol (Klincic et al., 2014), ethyl ether:hexane (Ingelido et al., 2007) and isopropanol:ethyl acetate (Julshamn et al., 2013) are also used for the extraction of PCBs. Initial extraction from some sample types (e.g. blubber) can be accomplished using hexane alone in a Soxhlet apparatus (Kannan et al., 1993). Once the samples have been extracted, isolation and purification of the extracts must be performed to ensure that the compounds of interest are retained and free of interfering co-extractives prior to analysis. Given that the PCBs are lipophilic compounds and found at elevated levels in lipid-rich foods, separation of the NDL-PCBs from the lipid is a critical step in the process of isolating the NDLPCBs from the DL-PCBs and other persistent organic compounds (e.g. PCDDs/ PCDFs). Given the chemical stability of PCBs to acid, treatment of extracts with

174

WHO Food Additives Series No. 71-S1, 2016

Supplement 1: Non-dioxin-like polychlorinated biphenyls

acid to digest the lipid is frequently performed and aids in the removal of coextracted compounds more easily degraded through digestion under these harsh conditions. Acid digestion is frequently achieved using acid-treated silica gel or via washing the extract with excess sulfuric acid, treatment with fuming sulfuric acid or some combination of these approaches (Kim et al., 2004; Borgå et al., 2005; Witczak & Ciereszko, 2006b; Julshamn et al., 2015). Although digestion is usually performed with sulfuric acid, formic acid has also been used in the extraction of human serum as part of the sample preparation for NDL-PCB analysis (Koppen et al., 2002; Jursa et al., 2006). Additionally, size exclusion chromatography or gel permeation chromatography is employed early in the cleanup process (Masuda et al., 2005; Malisch & Dilara, 2007). Compounds are eluted from the gel permeation chromatography columns based on molecular size so that separation can be completed without consideration of polarity, allowing for separation of other nonpolar compounds from the NDL-PCBs. When working with samples containing cholesterol, the application of gel permeation chromatography in advance of treatment with strong acid may be advantageous, because under harsh acidic conditions, the cholesterol may be converted to cholestadiene (Burke et al., 1974; Schüpfer & Gülaçar, 2000; Xiong, Wilson & Pang, 2007). Cholestadiene is difficult to remove from extracts using the usual routine adsorption chromatography approaches (e.g. Florisil, silica gel), and its presence has an impact on analytical results. Without the use of gel permeation chromatography cleanup in advance of acid digestion, separation of cholestadiene from the NDL-PCBs in the extracts can be achieved using silica that has not been acidified, for samples with high levels of cholesterol (e.g. egg yolks) (Rawn et al., 2012). Fractionation of the different groups of PCBs (DL- versus NDL-PCBs) can be achieved using stepwise elution from columns prepared with different adsorbents. The columns may be multilayered or prepared in series and generally include acidified silica, neutral silica gel and anhydrous sodium sulfate combined with a silicate (e.g. aluminosilicate, magnesium silicate, potassium silicate) and carbon. Silica treated with potassium hydroxide may be employed as well (Berggrena et al., 1999; Gomara et al., 2005; Costopoulou et al., 2006; Bordajandi, Abad & Gonzalez, 2008; Konuspayeva et al., 2011; Klincic et al., 2014). Separation of the NDL-PCBs, including the six indicator congeners and mono-ortho-substituted congeners, can be achieved through the combination of adsorbents and elution solvents, with the ratio between the polar and non-polar solvents being a critical parameter to achieve separation (Košcan et al., 1994). The packing of columns can depend on how the sample is prepared and extracted. For instance, if a sample is freeze-dried or ground with a drying reagent, it may be packed directly into the column above the adsorbents used for the isolation of the NDL-PCBs from other co-extractives (Krokos et al., 1997). The multiple layers of adsorbents, including potassium silicate and silica gel, allow for the removal of 175

Safety evaluation of certain food additives and contaminants Eightieth JECFA

numerous co-extractives, whereas the use of carbon allows the separation of the planar DL-PCBs and PCDDs/PCDFs through an appropriate choice of solvents (Krokos et al., 1997). High-performance liquid chromatography has been used in the separation of different PCB congeners (e.g. ortho-substituted versus nonortho-substituted), although this is not routinely reported in the literature (Echols et al., 1997; Pietrogrande et al., 2002). Automated systems for the cleanup and separation of different fractions (e.g. coplanar PCBs + PCDDs/PCDFs; NDL-PCBs + mono-ortho-substituted PCBs) of individual extracts are being used with greater frequency (Eljarrat et al., 2001; Focant, Pirard & De Pauw, 2004; Focant et al., 2006; Uçar et al., 2011). A series of columns containing alumina, Florisil and carbon have been used for cleanup. The NDL-PCBs can be eluted from the alumina column using dichloromethane:hexane (50:50), with elution of the planar molecules (e.g. DLPCBs) from carbon columns using toluene (Kim et al., 2004). Additionally, the application of multiple layered columns (e.g. acid/basic silica) has been used to destroy the lipid in samples, followed by a partly deactivated alumina column for the separation of the planar PCBs + PCDDs/PCDFs from the non-planar PCBs, including the NDL-PCB congeners (L’Homme et al., 2015). Owing to the high number of individual compounds being analysed when PCBs are under investigation and all of them having a similar structure, coelution of some congeners is known to occur using most GC columns available. The choice of column or columns used in GC measurement of PCBs will have an impact on which congeners will co-elute (Fujita et al., 2009). Some researchers have used temperature programming, different film thicknesses and carrier gas flow rates to improve the separation; others have performed sample analysis using more than one column, where each has a different polarity (Frame, 1997; Focant, Sjödin & Patterson, 2004). Employment of chiral columns to separate individual congeners has also been performed (Bordajandi et al., 2005). Additionally, the use of high-performance liquid chromatography with pyrenyl columns coupled to GC for separation of PCB congeners has been reported (Ramos, Hernández & González, 1999). Despite different approaches to overcome the issue of co-eluting congeners, separation of some congeners while performing routine analysis is an ongoing challenge. Detection of the PCBs has been performed routinely using 63Ni ECDs since the first analyses, and ECDs or micro-ECDs continue to be used for routine PCB analyses in some laboratories (Webb & McCall, 1973; Mullin et al., 1984; Schulz, Petrick & Duinker, 1989; Dewailly et al., 1991; Danielsson et al., 2005; Gomara et al., 2005; Bhavsar et al., 2007; Van Leeuwen et al., 2007; Bordajandi, Abad & Gonzalez, 2008; Oliveira Ribeiro et al., 2008; Storelli et al., 2008; Cirillo et al., 2009; Jankovic et al., 2011). Owing to the presence of the chlorines (n = 1–10) on each PCB congener, the detection of PCBs is sensitive; however, ECDs are not

176

WHO Food Additives Series No. 71-S1, 2016

Supplement 1: Non-dioxin-like polychlorinated biphenyls

specific detectors, and the possibility of misidentification of analytes is present. To overcome this issue, ECDs have been used in combination with two GC columns of differing polarity for separation of PCBs, thus providing confirmation of the analyses (Storr-Hansen & Cederberg, 1992; Atuma et al., 1998; Kearney et al., 1999; Danielsson et al., 2005; Polder et al., 2008b; Cerveny et al., 2014). Some laboratories have elected to use both ECD and mass spectrometry for quantification (Criado et al., 2004). As accessibility to mass spectrometers has increased, so has their use in PCB analysis. They are a popular tool, because they can provide structural information about the analytes being measured, rather than simply a response to certain functional groups. In some cases, the application used is low-resolution mass spectrometry (Tanabe et al., 1997; Pietrogrande et al., 2002; Fernandes et al., 2004; Witczak & Ciereszko, 2006a; Storelli et al., 2012a), which may be coupled with large-volume injections to improve LODs (Traag et al., 2006; Witczak & Ciereszko, 2006a) using selected ion monitoring. Chemical ionization has been used for these analyses (Polder et al., 2008a), although electron ionization is used more frequently (Turrio-Baldassarri et al., 2009). High-resolution mass spectrometry is also used with greater frequency (Arnich et al., 2009), in part because the cost of these instruments has decreased and there is greater accessibility to them. Laboratories analysing PCDDs/PCDFs and DL-PCBs require the use of high-resolution mass spectrometric detection for these compounds because of their ultra-low concentrations in food and other tissues; therefore, some authors are reporting their use in the analysis of NDL-PCBs as well (Tsutsumi et al., 2002; Baars et al., 2004; Wingfors et al., 2005; Focant et al., 2006; Noma et al., 2006; Zuccato et al., 2008; Rawn et al., 2012; Blanchet-Letrouvé et al., 2014; Pavuk et al., 2014). In some publications, authors are adopting the technology of coupling gas chromatographs to tandem mass spectrometers for NDL-PCB analysis and may employ large-volume sample injection volumes (Pirard, Focant & De, 2002; Esteve-Turrillas et al., 2008; Cimenci et al., 2013). The use of GC coupled to ion trap detection capable of tandem mass spectrometry has been applied to PCB analysis in food samples. Some ionization inhibition occurred in the more highly contaminated samples analysed (e.g. cod liver); however, this method proved beneficial, owing to its greater selectivity compared with the more routinely used micro-ECD (Gomara et al., 2006). LODs and limits of quantification (LOQs) may vary between methods used for NDL-PCB analysis. The LOQ is generally taken to be a concentration 3-fold higher than the LOD. Numerous factors contribute towards the LOD/LOQ values attainable during analysis. The matrix under investigation, the sample size taken for analysis, whether a sample extract was split for additional analyses and the final volume to which the sample extract is taken all have an impact on the LOD. Extraction efficiency may be better in some matrices relative to others, which 177

Safety evaluation of certain food additives and contaminants Eightieth JECFA

will contribute towards improved LODs, whereas other matrices may have coextracting artefacts leading to elevated LODs. Additionally, the LOD may increase or decrease as a result of the sensitivity of the instrument used for detection. The LODs reported in the data submitted to the Global Environment Monitoring System – Food Contamination Monitoring and Assessment Programme (GEMS/ Food) database in response to the call for NDL-PCB concentration data in food ranged from 0.03 pg/kg (0.000 03 ng/kg), reported in rice, to 0.02 mg/kg (20 000 ng/kg), reported in fish. Individual PCB congener concentrations reported in the literature for food commodities and human tissues generally are below 1 ng/mL or 1 µg/kg (1000 ng/kg) in manuscripts dating from 2000 forward. Frequently, LODs are reported at levels of ≤0.2 µg/kg or <0.5 ng/mL (Chu, Covaci & Schepens, 2003; Fernandes et al., 2004; Costopoulou et al., 2006; Bordajandi, Abad & Gonzalez, 2008; Esteve-Turrillas et al., 2008; Cirillo et al., 2009; Grassi et al., 2010; Jankovic et al., 2011; Cimenci et al., 2013; Fromme et al., 2015). If LOD values are very high, exposure estimates developed using the LOD for the upperbound concentrations may result in artificially elevated exposure levels. 3.2.4 Quality assurance considerations

Owing to the ubiquitous nature of the NDL-PCBs, the confirmation and removal of background levels from analytical results must be performed. Within each set of unknown samples to be analysed, NDL-PCBs should be measured in one or two laboratory blank or reagent blank samples. By establishing the background level in the laboratory where the analyses are performed, corrections can be made so that the final reported concentrations in the unknown samples do not include background concentrations resulting from handling in the laboratory. Additionally, ongoing testing of laboratory quality should be performed through the routine testing of quality assurance samples. These samples may be reference materials with certified concentrations of NDL-PCBs or samples that have been tested over time in the laboratory, so that results may be evaluated on an ongoing basis. The uncertainty of sample analyses should be determined for the analytical results developed by a laboratory. The uncertainty can be established using more than one approach: (1) cumulative uncertainty developed through examination of each step in the sample preparation and analysis protocol and (2) comparison of the results obtained relative to the reference materials tested (Fernandes et al., 2004; Eppe et al., 2014). The use of proficiency testing programmes and participation in interlaboratory studies such as the POPs programme organized by the Norwegian Institute of Public Health (Bremnes, Broadwell & Becher, 2013) are encouraged, to aid in validation of analytical results.

178

WHO Food Additives Series No. 71-S1, 2016

Supplement 1: Non-dioxin-like polychlorinated biphenyls

3.2.5 Reference methods

The long history with the analysis of PCBs, including both DL- and NDL-PCBs, has allowed different groups to establish methods for use as references for the analysis of these compounds. The United States Environmental Protection Agency (USEPA) has established a method for the analysis of PCBs, as both individual congener concentrations and concentrations based on Aroclor (1016, 1221, 1232, 1242, 1248, 1254, 1260) equivalents. The method is recommended for use with solid and aqueous matrices and is available online (EPA method 8082) (USEPA, 1996). EPA method 8082 was subsequently expanded to include analysis of PCBs in tissue samples (EPA method 8082A) (USEPA, 2000). These methods were developed for use with ECD or electrolytic conductivity detection and include five of the six indicator PCB congeners (PCBs 52, 101, 138, 153, 180); PCB 28 is not listed as one of the target compounds for these analyses (USEPA, 1996, 2000). These methods have been used and adapted by many researchers for successful PCB analysis worldwide. More recently, the USEPA developed EPA method 1668 for the analysis of PCB congeners in water, soil, sediment and biosolids, in addition to tissue. It employs the use of high-resolution mass spectrometry (USEPA, 2008). Additionally, other agencies (e.g. AOAC International and the International Organization for Standardization) have developed validated matrixspecific methods for the analysis of PCBs (AOAC International, 1983; ISO, 2008). In contrast, the European Union has developed criteria for the analysis of PCBs (NDL- and DL-PCBs), in addition to the PCDDs/PCDFs, in feed and food. The criteria established for NDL-PCB analysis focused on GC methods employing ECD, low-resolution mass spectrometry (e.g. unit mass resolution mass spectrometry), tandem mass spectrometry and high-resolution mass spectrometry (>8000 resolution) (Malisch, Kotz & Wahl, 2009). The criteria included limitations to the deviation in relative retention time of the analyte in unknown samples when compared with analytical standards, the number of ions to be included for each analyte and limitations to mass range, particularly for the low-resolution methods (Malisch, Kotz & Wahl, 2009). The issues of concern identified for consideration include those parameters required for good quality assurance practices (e.g. validation of methods within the range of interest, recoveries of surrogate standards, consistently having high-quality results when examining quality control samples). The European Union conducted five proficiency tests between 2009 and 2013 using foods of animal origin for testing (Kotz et al., 2014). As part of the discussion on the results, it is noted that the criteria for the methods do not extend to the extraction methods, but rather focus on the extraction efficiency of the methods employed (Kotz et al., 2014). 179

4. Sampling protocols Although there are no established protocols set specifically for the collection and storage of samples for NDL-PCB analysis, the best practices established for other POPs present at ultra-trace levels (e.g. PCDDs/PCDFs, DL-PCBs) should be followed. The collection of samples should be performed using containers that are non-reactive (e.g. glass, aluminium) and have been chemically cleaned or certified to be free of contaminants, and the samples should be collected by experienced individuals. When sampling commercial food products, samples collected must be representative of the lot. Therefore, collection of multiple samples (incremental) from within the lot is recommended, and the multiple samples may be used to form an aggregate sample from which laboratory samples may be analysed. Adoption of a subsampling protocol similar to that identified by the European Union may be beneficial (European Commission, 2014). Prior to the subsampling for laboratory analysis, homogenization of the aggregate sample should be performed. Sample container lids should be lined with a non-reactive material. In some situations, plastic bags may be used for sample collection. There is concern, however, that the sample may be exposed to phthalates and other plasticizers, so the use of plastic bags for long-term storage and throughout the sample preparation phase is discouraged (De Boer, 1999; Muir & Sverko, 2006). Food samples must be stored in a manner that ensures the integrity of the food during transport to the laboratory and through storage (Melis & Zuccato, 2012). Lyophilization of samples is known to aid in removal of moisture prior to storage, which is beneficial when working with fresh food (Fernandes et al., 2004). The use of freeze-drying, however, can result in the loss of the more volatile, lower chlorinated PCB congeners, owing to their greater volatility and aqueous solubility (Berdié & Grimalt, 1998; Juan et al., 1999; De Voogt, Van Der Wielen & Govers, 2000). PCB losses during the freeze-drying process are influenced by the properties of the food being dried (e.g. fat content) and not strictly the properties of the analytes of interest (Witczak & Ciereszko, 2006b). The drying of samples with high fat content (e.g. meat) may result in cells collapsing and clumping, which could limit the efficacy of the solvent in the extraction of analytes from the matrices (Ridgway, Lalljie & Smith, 2007). The use of drying agents (e.g. anhydrous sodium sulfate) to dry and grind food may be more advantageous in some cases. Sample storage space is a restriction placed on most laboratories, and ensuring sufficient space to retain subsamples for archival purposes is advantageous (Gunter, 1997; Wells & Hess, 2000; Muir & Sverko, 2006). Samples retained for short periods pending PCB analysis can be stored at a higher 180

Supplement 1: Non-dioxin-like polychlorinated biphenyls

temperature (e.g. −25 °C) than that used for samples being retained for a long period (>2 years), where the suggested temperature is less than −70 °C (De Boer & Smedes, 1997). The storage of multiple aliquots that are relatively small in size may be more beneficial than storage of one large sample, particularly for frozen samples, because this will limit the number of freeze–thaw cycles to which the sample may be subjected when reanalysis is to be performed (Gunter, 1997; Wells & Hess, 2000).

181

5. Effects of processing In contrast to other contaminants, NDL-PCBs are thermally stable and resistant to degradation. Studies on the impact of processing on PCB concentrations have been largely focused on the cooking techniques used to prepare foods. Techniques that change the fat content will have an impact on PCB concentrations (e.g. PCB concentrations are lowered in skimmed milk, but increased concentrations are found in foods with higher fat content, such as cheese or cream) (Newsome, Davies & Sun, 1998). Although the studies related to the impact of processing on PCB concentrations include both DL-PCBs and NDL-PCBs, the impact on the concentrations is similar for both groups. Ultimately, processing that results in the removal of lipids will lead to a decrease in PCB concentrations in the final food product. Food preparation prior to the cooking stage can lead to a reduction in the concentration of PCBs. In studies using salmon and bluefish, the removal of the skin resulted in a reduction of PCB content compared with fish prepared with the skin remaining (Salama et al., 1998; Bayen et al., 2005). However, a review of published studies to determine the impact of processing of fish on contaminant concentrations resulted in a conclusion that losses observed with different cooking methods are not affected by whether the skin was removed (Wilson et al., 1998). Additionally, the location along the fish from which the salmon steaks were taken was observed to have an impact on PCB concentrations, with those taken closer to the head having higher concentrations than those prepared from the region nearer to the tail (Bayen et al., 2005). The process of cooking food can also have an impact on PCB residue concentrations. Cooking has been reported to lower PCB concentrations in bluefish, although frying and baking resulted in less reduction in the PCB concentrations than observed for other techniques, such as smoking, microwave cooking or charbroiling (Salama et al., 1998). Baking, charbroiling and smoking also resulted in significantly lower PCB residues compared with raw trout samples, although smoking lowered PCB concentrations to a greater extent than the other methods tested (Zabik et al., 1996). The greater losses found in the smoked fish were hypothesized to result from the long, slow smoking process with its higher final temperature, which ultimately leads to greater fat losses (Zabik et al., 1996). In the study focused on salmon steaks, cooking by pan frying, microwave cooking, boiling or baking generally reduced PCB concentrations, although the differences in the PCB content were not significantly lower, regardless of the type of cooking method used (Bayen et al., 2005). A reduction in PCB concentrations was generally observed following cooking of samples of a variety of finfish and non-finfish species, although only relatively small changes in the PCB concentrations were observed, regardless of whether samples were baked, 182

Supplement 1: Non-dioxin-like polychlorinated biphenyls

boiled or fried (Rawn et al., 2013). In a review of the data developed to determine the impact of cooking on contaminant (DDT, PCBs) concentrations, Wilson et al. (1998) found that greater variability in the reduction of the compounds of interest was observed when broiling, frying and boiling, relative to baking or smoking. Additionally, baking was thought to remove more consistent amounts of fat than would occur for other methods (Wilson et al., 1998). Sherer & Price (1993) recommended reporting results of this type based on the mass of PCBs in the final cooked samples relative to the raw products. Although the literature generally is in agreement that a reduction in PCB concentration occurs with cooking, the approaches dealing with the data differ, with some researchers focusing on changes based on whole weight concentrations and others focusing on lipid-adjusted concentrations. Overall, multiple authors have observed PCB losses associated with cooking different fish species (e.g. salmon, bluefish, trout), ranging from approximately 25% to 65% (Salama et al., 1998; Wilson et al., 1998; Bayen et al., 2005). In a study to examine the effect of the smoking process exclusively, it was determined that cold smoking of mackerel and herring fillets resulted in an elevation in PCB concentrations, whereas hot smoking resulted in an overall lowering of PCB concentrations (Witczak & Ciereszko, 2006b). PCB reductions were observed during the first hour of the hot smoking process, although a small increase in these concentrations was found during the later stages of this process (Witczak & Ciereszko, 2006b). This small increase in PCB concentrations at the later stages of the hot smoking process was postulated to have occurred as a result of the co-distillation of PCBs with water vapour. Additionally, studies have been performed to determine the impact of the refining process on PCB concentrations in crude fish oils for use as dietary supplements. Among the different steps in the refining process (neutralization, bleaching and deodorization), deodorization resulted in the greatest reduction in PCB concentrations in fish oils, whereas the other steps did not have a significant effect on the concentrations (Hilbert et al., 1998). Fish meal and fish oil are known to be sources of POPs, including both PCBs and PCDDs/PCDFs. A number of approaches have been investigated to decrease the levels of POPs in fish meal products, including ultraviolet light treatment, enzymatic treatment, pH shift processes and distillation processes (Baron, Børresen & Jacobsen, 2005, 2007; Marmon, Liljelind & Undeland, 2009; Oterhals & Berntssen, 2010). Although ultraviolet A exposure did not lower POP concentrations, ultraviolet B exposure did. Although changes in pH resulted in lower PCDD/ PCDF and DL-PCB concentrations, protein losses were observed in the herring mince samples tested (Baron, Børresen & Jacobsen, 2005; Marmon, Liljelind & Undeland, 2009). Only limited reduction occurred, however, when fish meal was exposed to oxidoreductase enzymes (Baron, Børresen & Jacobsen, 2007). 183

Safety evaluation of certain food additives and contaminants Eightieth JECFA

Short-path distillation of fish oil contributed to lower PCDD/PCDF and DL-PCB concentrations; however, losses of vitamins were detected (Oterhals et al., 2007). The impact of temperature, moisture content and leaching time during fish meal production was investigated to establish optimal conditions to reduce PCDD/ PCDF and DL-PCB concentrations while maintaining fish meal protein quality. Leaching (solid–liquid) extraction with triglyceride used as the extracting solvent reduced POP concentrations and maintained the quality of the final fish meal products (Oterhals & Kvamme, 2013). Although the majority of studies to determine the impact of cooking on POP concentrations have been performed using fish, the impact on meat has also been reported. No consistent decrease in PCDD/PCDF or DL-PCB concentrations in beef, pork and fish products was observed following broiling, compared with concentrations in the raw food. No investigation of the NDLPCB congeners was performed, although it is thought that the broad behaviour of the NDL-PCBs generally can be represented by the DL-PCB congeners in this context (Schecter et al., 1998).

184

WHO Food Additives Series No. 71-S1, 2016

6. Prevention and control Food consumption is well known to be the primary pathway of exposure to PCBs and other POPs (e.g. PCDDs/PCDFs). Exceptions to this may occur for individuals who have been exposed through occupational activities. Within the broad category of food, foods of animal origin are the most significant contributors to human PCB body burdens, owing to the bioaccumulation of these chemicals through the food-chain to the top-level predators (i.e. humans) (Malisch & Kotz, 2014). Given the major source of PCBs and other POPs, the focus of efforts related to preventing exposure is through limitations to exposure along the food-chain, including exposure of food production animals. With the knowledge that fish, meat and dairy product consumption contributes most greatly to PCB exposure, methods of PCB reduction in animals from which the foods are derived are of primary interest (Jelinek, 1985; Malisch & Kotz, 2014). Knowledge on the transfer of contaminants from feed to animal food products is essential for food safety risk assessment. Feeding studies have been carried out with both terrestrial and aquatic farmed animals by exposing them to contaminated diets under controlled conditions, which allowed for the determination of the carry-over rate from the feed into food. These data have been summarized in a review article and in a database of the carry-over rates for contaminants, including PCBs, in various terrestrial farmed animals, including cattle, poultry, pigs, sheep, goats, rabbits and birds (Kan & Meijer, 2007; Leeman, Van Den Berg & Houben, 2007). Geometric mean carry-over rates, calculated as the concentration of PCBs in animal products relative to the PCB concentrations in animal feed, differed by commodity: meat (0.14), fat (3.9), edible offal (0.7), eggs (0.92) and whole milk (0.26) (Leeman, Van Den Berg & Houben, 2007). Berntssen et al. (2011) determined the amount of POPs, including the six indicator PCBs, retained in farmed Atlantic salmon fillets relative to the amount consumed via feed and reported that the retention ranged from 42 ± 3% to 50 ± 2% for PCBs 138 and 101, respectively (sum of six indicator PCBs = 46 ± 3%). Additionally, they compared the carry-over rates reported for PCBs in terrestrial farmed animals (median = 0.12, maximum = 0.36) with those determined for DL-PCBs in the fish fillet (0.79) and found that the carry-over rate for salmon fillets was approximately 2–7 times higher than that for terrestrial meat products, depending on whether the median or maximum value was compared (Berntssen et al., 2011). In the above studies, information on carry-over rates specifically for NDL-PCBs, however, was insufficient, as few data sets distinguished NDL-PCBs from the sum of PCBs in these studies. Transfer of DL- and NDL-PCBs from feed to animal food products has been reported in other studies in the literature. The feed to milk transfer of PCBs has been assessed using lactating goats and cows (Thomas, Sweetman 185

Safety evaluation of certain food additives and contaminants Eightieth JECFA

& Jones, 1999a,b; Kerst et al., 2004; Costera et al., 2006; Ounnas et al., 2010). Carry-over percentages of the NDL-PCBs have been determined in lactating goats fed with contaminated hay and in cows reared with naturally contaminated or background-level pastures. The carry-over rates of the NDL-PCB congeners, reported as a percentage, into milk were similar for goats and cows: PCBs 138, 153 and 180 showed high carry-over percentages, more than 40% for goats and more than 60% for cows, whereas PCBs 28, 52 and 101 were transferred poorly, less than 30% for goats and less than 10% for cows (Thomas, Sweetman & Jones, 1999a; Costera et al., 2006). Similar transfer patterns for indicator PCBs were observed in the eggs and fat of laying hens (Hoogenboom et al., 2006b) and in the fat of broilers and pigs (Hoogenboom et al., 2004). Following removal from exposure to contaminated feed, concentrations of indicator PCBs in milk and egg fat decreased rapidly to about 50% within 10 days, followed by a gradual elimination (Hoogenboom et al., 2006b; Fournier et al., 2013). Owing to the efficiency of the transfer of NDL-PCBs from feed to food products, it has been stated by multiple authors that the international limits for PCBs in feed may be approached and that these limits may be insufficient to guarantee related food safety (Costera et al., 2006; Hoogenboom et al., 2006b; Ounnas et al., 2010; Fournier et al., 2013). Given the limits established in 2006 for PCDD/PCDF and DL-PCB concentrations in food and feed in the European Union (European Commission, 2006), studies have been undertaken to determine the impact of food production animal environments on PCB and PCDD/PCDF concentrations. PCB and PCDD/ PCDF concentrations in beef exceeded maximum guideline levels when animals were raised on floodplains with elevated PCB and PCDD/PCDF concentrations, similar to animals raised on lands near sources of these contaminants and those raised on land treated with sewage sludge in the 1960s/1970s (Weber et al., 2014). The results indicated that for animals to have concentrations of these contaminants below regulated levels, pasturelands having low POP concentrations are necessary. Poultry is another important source of meat for human consumption. An investigation into the use of activated carbon through the feed as a means of reducing PCDDs/PCDFs and DL-PCBs was performed. Although the carbon treatment did reduce POP concentrations in chicken muscle, abdominal fat and eggs, a reduction in important vitamins (α- and γ-tocopherol) was also observed (Guruge et al., 2012). Although these studies have been focusing on the determination of the impact of selected treatment on the levels of PCDDs/PCDFs and DL-PCBs, the extension to the NDL-PCBs is something to consider. The adherence to good agricultural practices and good animal feeding practices should contribute to the efforts to reduce POPs (e.g. PCBs) in food for human consumption (FAO/WHO, 2006; Mansour, 2011). PCB contamination of animal housing and/or buildings (e.g. silos) near animal pastures contributes to

186

WHO Food Additives Series No. 71-S1, 2016

Supplement 1: Non-dioxin-like polychlorinated biphenyls

animal exposure levels, as does the pasturing of animals on lands contaminated with PCBs during periods when these lands were used for other activities (e.g. railway lines) (Weber et al., 2014). Potential PCB exposure of grazing animals necessitates the need to identify contaminated pastures to ensure that they are not used for this purpose. PCB contamination can be further reduced by establishing and adhering to soil guideline values for agricultural purposes, performing diligent monitoring programmes to confirm compliance and establishing critical control points for the feed manufacturing process where PCBs can be reduced (FAO/WHO, 2006). Additionally, farming practices should include plans for isolation, among other procedures, if contamination is detected (Mansour, 2011).

187

7. Levels and patterns of contamination of food commodities Thirty countries provided data to the Committee for its review of NDL-PCBs. Among the submissions from countries responding to the call, 23 were from Europe through EFSA. Six other contributing countries were from the western Pacific region, and one was from the Americas (Table 17). Those data that were not in an acceptable format for evaluation by the Committee were removed from the data set. Among the submitted data, the vast majority represented commercialtype food, with only two records of human milk reported (for PCBs 28 and 101). In general, countries reported the NDL-PCB occurrence data as individual congeners. The majority of countries restricted the submission of data to the six indicator PCB congeners (PCBs 28, 52, 101, 138, 153 and 180) for each food category. The focus on the six indicator PCB congeners is consistent with the regulations pertaining to NDL-PCBs in foodstuffs within the European Union. Maximum European Union limits are established for the sum of the six indicator PCB congeners in meat and meat products (40 ng/g fat), liver of terrestrial animals and derived products (40 ng/g fat), fish liver and derived products (200 ng/g ww), fish and fishery products (75 ng/g ww), wild-caught freshwater fish (125 ng/g ww), wild-caught eel (300 ng/g ww), marine oils (200 ng/g fat), raw milk and dairy products (40 ng/g fat), hen eggs and egg products (40 ng/g fat), fat from bovine animals, sheep, poultry and pigs (40 ng/g fat), mixed animal fats (40 ng/g fat), vegetable oils and fats (40 ng/g fat) and foods for infants and young children (1 ng/g ww) (European Commission, 2011). The European Commission had also recommended that member states monitor concentrations of PCDDs/ PCDFs and PCBs in food and feed, and the results of the monitoring activity for the NDL-PCBs have been reported by EFSA (2010, 2012). Recommendations as to the annual minimum number and type of food samples to be analysed for PCDDs/PCDFs, DL-PCBs and, if possible, NDL-PCBs were established for member countries in 2004 (European Commission, 2004). Some countries reported additional congeners for foods of high interest (e.g. fish), while maintaining the reporting of the six indicator congeners for all other food types reported. Still other countries reported all NDL-PCB congeners measured. One country supplied the majority of their occurrence data as technical mixture (Aroclor 1254, 1260) equivalents, with indicator congeners reported only in fish samples. Of the 30 reporting countries, only four (Australia, Germany, New Zealand and Slovakia) reported some aggregated results; among them, only New Zealand restricted reporting to aggregated sample results.

188

Supplement 1: Non-dioxin-like polychlorinated biphenyls

Table 17 Summary details of the NDL-PCB data sets submitted to JECFA, by countrya No. of aggregated results reported 42 0 0 0 0 No. of individual results reported 314 978 3 804 2 120 8 694 1 704 1 092 2 688 2 388 7 159 8 010 46 272 2 700 822 2 178 1 554 505 72 1 278 0 6 540 2 838 5 424 1 050 36 468 1 314 206 3 858 7 856

Country Australia Austria Belgium Canada China Mainland Hong Kong SAR Cyprus Czech Republic Estonia Finland France Germany Greece Iceland Ireland Italy Japan Luxembourg Netherlands New Zealand Norway Poland Republic of Korea Romania Singapore Slovakia Slovenia Spain Sweden United Kingdom

No. of records 356 978 3 804 2 120 10 398 8 694 1 704 1 092 2 688 2 388 7 159 8 010 46 311 2 700 822 2 178 1 554 505 72 1 278 2 6 540 2 838 5 424 1 050 36 1 192 1 314 206 3 858 7 856

No. of congeners routinely reported –b 6 indicators 6 indicators 20c 6 indicators

Sample collection years 2005–2008 2007–2012 2002–2013 2005–2006 2005–2014 2010–2011 2009–2013 2005–2013 2003–2007 2002–2010 2003–2010 1995–2010 2000–2010 2003–2005 2003–2006 2001–2003 2012 2002 1999–2009 1998 1999–2013 2006–2010 2012–2013 2006–2007 2013–2014 2000–2010 2004–2012 2004–2012 2000–2013 2001–2006

6 indicators 6 indicators 6 indicators (19 in a subset of samples) 6 indicators (25 congeners in fish) 6 indicators 6 indicators 6 indicators 6 indicators 6 indicators 6 indicators 5 indicatorsd 6 indicators 6 indicators 2 indicators 6 indicators 6 indicators 6 indicators 6 indicators 6 indicators 6 indicators 6 indicators 6 indicators 6 indicators 6 indicators (35 congeners in fish)

0 0 0 0 0 39 0 0 0 0 0 0 0 2 0 0 0 0 0 724 0 0 0 0

SAR: Special Administrative Region a Data from Denmark not included. b Most reported data as Aroclor equivalents (1254, 1260). c Co-elution of PCB 101 with other congeners, removed from submission. d Co-elution of PCB 52 with other congeners, data not submitted

189

Safety evaluation of certain food additives and contaminants Eightieth JECFA

190

WHO Food Additives Series No. 71-S1, 2016

Most countries reported PCB concentrations in fish, milk, meat, eggs, and fats and oils, although some countries reported results for a much wider range of foods, including composite foods (e.g. pizza), fruit, legumes, nuts, snacks, starchy roots and tubers, and vegetables. Because the majority of exposure to PCBs occurs through the consumption of fish, milk, meat, eggs and their products, the summary of the results has focused on these categories. Infant food was considered using food produced exclusive of human milk, as the concentrations of NDL-PCBs were below the LOD in the only human milk results submitted. Each country submitted data using relevant time frames for their work; some included more historical data sets (i.e. ≤ year 2000), whereas others reported data for recent collections only (≥ year 2010). The collection period differential may have contributed to the difference in maximum concentrations observed. This is consistent with the temporal decrease in POP, including NDL-PCB, concentrations in food reported in the literature. In this data set, foods known to have elevated levels of animal fat had higher indicator PCB concentrations relative to other food types. Maximum concentrations in foods from these categories (fish, meat, eggs, milk) were well in excess of 10 000 ng/kg ww (PCB 101: 1 200 000 ng/ kg ww in fish and 240 000 ng/kg ww in meat; PCB 138: 34 820 ng/kg ww in eggs; PCB 153: 16 900 ng/kg ww in milk). The maximum indicator PCB concentration in foods identified for infants and very young children was less than 10 000 ng/ kg ww (PCB 153: 5920 ng/kg ww). Although maximum concentrations were elevated, the central tendency of the data was low, with median values generally below 1 µg/kg (1000 ng/kg) for the indicator congeners. This result is consistent with the contributed data sets having had data with concentrations frequently below the LOD (Table 18) and taken to be zero for summary purposes. A more detailed description of minimum and maximum concentrations of the indicator PCBs by country and food category may be found in Appendix 1, with mean concentrations reported in Appendix 2. The largest number of results submitted for this review was from samples belonging to the fish and fish products category (>7000 records for each indicator congener). The maximum indicator PCB concentration was observed in foods from this category. It was noted that increasing concentrations of the higher chlorinated congeners (PCBs 101, 138, 153 and 180) were observed in the fish products relative to the trichlorinated PCB 28 and tetrachlorinated PCB 52 (Fig. 3), similar to the pattern observed in eggs (Fig. 4). Although there was a large spread in concentrations reported, the median concentrations remained very low for all six congeners for both food groups.

Supplement 1: Non-dioxin-like polychlorinated biphenyls

Table 18 Summary of the number of samples and proportion of detected samples by country and food category for the submitted data for the six indicator PCB congeners No. of data submitted 42 12 156 132 72 12 396 168 30 402 978 1 050 60 36 6 162 996 30 84 90 10 20 40 50 10 130 140 10 10 10 10 252 288 No. of detectionsa 0 12 156 132 72 12 396 168 30 169 218 488 0 1 0 15 417 1 14 90 10 20 40 50 10 130 140 10 10 10 10 178 195 Frequency of detection (%) (six indicator congeners) 0 100 100 100 100 100 100 100 100 42 22 46 0 3 0 9 42 3 17 100 100 100 100 100 100 100 100 100 100 100 100 71 68

Food category (by country) Australia Fish and other seafood (including amphibians, reptiles, snails and insects) Austria Cereals and cereal-based products Eggs and egg products Fish and other seafood (including amphibians, reptiles, snails and insects) Food for infants and small children Herbs, spices and condiments Meat and meat products (including edible offal) Milk and dairy products Products for special nutritional use Belgium Eggs and egg products Fats and oils of animal and vegetable origin (excluding butter) Fish and other seafood (including amphibians, reptiles, snails and insects) Food for infants and small children Fruit and fruit products Legumes and pulses Meat and meat products (including edible offal) Milk and dairy products Other foods Products for special nutritional use Canada Composite food (including frozen products) Eggs and egg products Fats and oils of animal and vegetable origin (excluding butter) Fish and other seafood (including amphibians, reptiles, snails and insects) Food for infants and small children Herbs, spices and condiments Meat and meat products (including edible offal) Milk and dairy products Nuts and oilseeds Snacks and desserts Starchy roots and tubers Sugar and confectionary (including cocoa products) China, mainland Cereals and cereal-based products Eggs and egg products

191

Safety evaluation of certain food additives and contaminants Eightieth JECFA

Table 18 (continued) Frequency of detection (%) (six indicator congeners) 87 77 85 70 79 74 63 0 0 0 0 45 0 2 7 0 0 0 0 8 16 68 100 44 34 100 13 38 46 33 67 43 49 15 8 0

Food category (by country) Fish and other seafood (including amphibians, reptiles, snails and insects) Food for infants and small children Legumes and pulses Meat and meat products (including edible offal) Milk and dairy products Starchy roots and tubers Vegetables and vegetable products (including fungi) China, Hong Kong Special Administrative Region Cereals and cereal-based products Composite foods (including frozen products) Eggs and egg products Fats and oils of animal and vegetable origin (excluding butter) Fish and other seafood (including amphibians, reptiles, snails and insects) Herbs, spices and condiments Meat and meat products (including edible offal) Milk and dairy products Non-alcoholic beverages (excluding milk, fruit and vegetable juice, water and stimulants) Snacks and desserts Starchy roots and tubers Stimulant beverages, dried and diluted excluding cocoa products Sugar and confectionary (including cocoa products) Cyprus Eggs and egg products Fish and other seafood (including amphibians, reptiles, snails and insects) Food for infants and small children Meat and meat products (including edible offal) Milk and dairy products Other foods Czech Republic Alcoholic beverages Cereals and cereal-based products Composite food (including frozen products) Drinking-water (water without any additives except carbon dioxide; includes water ice for consumption) Eggs and egg products Fats and oils of animal and vegetable origin (excluding butter) Fish and other seafood (including amphibians, reptiles, snails and insects) Food for infants and small children Fruit and fruit products Fruit and vegetable juices

No. of data submitted 2 082 822 252 2 274 2 226 252 246 312 240 72 24 456 24 288 144 24 24 24 48 24 300 240 18 204 318 12 24 234 54 12 36 84 276 54 174 6

No. of detectionsa 1 814 637 215 1 597 1 753 186 154 0 0 0 0 205 0 6 10 0 0 0 0 2 48 162 18 90 108 12 3 90 25 4 24 36 136 8 14 0

192

WHO Food Additives Series No. 71-S1, 2016

Supplement 1: Non-dioxin-like polychlorinated biphenyls

Food category (by country) Herbs, spices and condiments Legumes and pulses Meat and meat products (including edible offal) Milk and dairy products Non-alcoholic beverages (excluding milk, fruit and vegetable juice, water and stimulants) Nuts and oilseeds Snacks and desserts Starchy roots and tubers Stimulant beverages, dried and diluted excluding cocoa products Sugar and confectionary (including cocoa products) Vegetables and vegetable products (including fungi) Estonia Eggs and egg products Fats and oils of animal and vegetable origin (excluding butter) Fish and other seafood (including amphibians, reptiles, snails and insects) Meat and meat products (including edible offal) Milk and dairy products Sugar and confectionary (including cocoa products) Finland Composite food (including frozen products) Eggs and egg products Fish and other seafood (including amphibians, reptiles, snails and insects) Food for infants and small children Meat and meat products (including edible offal) Milk and dairy products France Eggs and egg products Fats and oils of animal and vegetable origin (excluding butter) Fish and other seafood (including amphibians, reptiles, snails and insects) Food for infants and small children Fruit and fruit products Legumes and pulses Meat and meat products (including edible offal) Milk and dairy products Germany Alcoholic beverages Cereals and cereal-based products Composite food (including frozen products) Eggs and egg products Fats and oils of animal and vegetable origin (excluding butter) Fish and other seafood (including amphibians, reptiles, snails and insects)

No. of data submitted 18 36 720 282 24 12 36 102 12 216 276 144 6 708 78 90 72 6 60 3 006 48 168 90 690 66 5 220 102 30 18 1 674 210 66 66 30 1 704 102 8 053

No. of detectionsa 4 1 287 112 6 6 8 14 0 15 60 135 0 522 46 46 6 6 60 2 998 48 168 44 510 66 5 030 102 30 18 579 180 0 60 6 978 74 3 326

Frequency of detection (%) (six indicator congeners) 22 3 40 40 25 50 22 14 0 7 22 94 0 74 59 51 8 100 100 100 100 100 49 74 100 96 100 100 100 35 86 0 91 20 57 73 41

193

Safety evaluation of certain food additives and contaminants Eightieth JECFA

Table 18 (continued) Frequency of detection (%) (six indicator congeners) 95 94 0 100 39 96 69 100 0 0 98 59 15 44 100 100 53 86 33 70 83 71 84 92 67 100 100 88 98 99 0 0 100 100 76 0 0

Food category (by country) Food for infants and small children Fruit and fruit products Fruit and vegetable juices Legumes and pulses Meat and meat products (including edible offal) Milk and dairy products Products for special nutritional use Starchy roots and tubers Stimulant beverages, dried and diluted excluding cocoa products Sugar and confectionary (including cocoa products) Vegetables and vegetable products (including fungi) Greece Eggs and egg products Fats and oils of animal and vegetable origin (excluding butter) Fish and other seafood (including amphibians, reptiles, snails and insects) Food for infants and small children Fruit and fruit products Meat and meat products (including edible offal) Milk and dairy products Iceland Cereals and cereal-based products Eggs and egg products Fish and other seafood (including amphibians, reptiles, snails and insects) Meat and meat products (including edible offal) Milk and dairy products Products for special nutritional use Starchy roots and tubers Ireland Cereals and cereal-based products Composite food (including frozen products) Eggs and egg products Fats and oils of animal and vegetable origin (excluding butter) Fish and other seafood (including amphibians, reptiles, snails and insects) Fruit and fruit products Herbs, spices and condiments Meat and meat products (including edible offal) Milk and dairy products Products for special nutritional use Starchy roots and tubers Vegetables and vegetable products (including fungi)

No. of data submitted 162 128 126 60 3 534 31 788 42 120 6 78 246 246 894 426 30 36 798 270 6 30 522 96 138 24 6 18 6 630 492 726 18 6 186 312 276 12 36

No. of detectionsa 154 120 0 60 1 371 30 652 29 120 0 0 240 144 136 189 30 36 420 232 2 21 435 68 116 22 4 18 6 557 480 720 0 0 186 312 210 0 0

194

WHO Food Additives Series No. 71-S1, 2016

Supplement 1: Non-dioxin-like polychlorinated biphenyls

Food category (by country) Italy Eggs and egg products Fats and oils of animal and vegetable origin (excluding butter) Fish and other seafood (including amphibians, reptiles, snails and insects) Meat and meat products (including edible offal) Milk and dairy products Sugar and confectionary (including cocoa products) Japan Fish and other seafood (including amphibians, reptiles, snails and insects) Luxembourg Fish and other seafood (including amphibians, reptiles, snails and insects) Netherlands Eggs and egg products Fats and oils of animal and vegetable origin (excluding butter) Fish and other seafood (including amphibians, reptiles, snails and insects) Meat and meat products (including edible offal) Milk and dairy products Snacks and desserts New Zealand Milk and dairy products Norway Cereals and cereal-based products Eggs and egg products Fats and oils of animal and vegetable origin (excluding butter) Fish and other seafood (including amphibians, reptiles, snails and insects) Fruit and fruit products Legumes and pulses Meat and meat products (including edible offal) Milk and dairy products Other foods Snacks and desserts Vegetables and vegetable products (including fungi) Poland Composite food (including frozen products) Eggs and egg products Fish and other seafood (including amphibians, reptiles, snails and insects) Meat and meat products (including edible offal) Milk and dairy products Sugar and confectionary (including cocoa products) Republic of Korea Fish and other seafood (including amphibians, reptiles, snails and insects)

No. of data submitted 162 72 732 342 174 72 505 72 12 24 1 158 36 36 12 2 24 114 114 5 658 12 6 306 288 6 6 6 6 390 1 608 444 360 30 5 424

No. of detectionsa 6 0 576 12 0 0 499 72 12 19 1 158 36 30 12 0 24 89 63 5 487 11 0 208 245 6 6 6 0 299 1 582 423 323 5 2 930

Frequency of detection (%) (six indicator congeners) 4 0 79 4 0 0 99 100 100 79 100 100 83 100 0 100 78 55 97 92 0 68 85 100 100 100 0 77 98 95 90 17 54

195

Safety evaluation of certain food additives and contaminants Eightieth JECFA

Table 18 (continued) Frequency of detection (%) (six indicator congeners) 0 0 0 0 0 0 0 0 0 0 63 24 28 33 44 60 45 28 23 33 77 52 8 100 17 0 0 0 100 0 60 100 89 69

Food category (by country) Romania Cereals and cereal-based products Composite food (including frozen products) Fats and oils of animal and vegetable origin (excluding butter) Fish and other seafood (including amphibians, reptiles, snails and insects) Meat and meat products (including edible offal) Milk and dairy products Other foods Stimulant beverages, dried and diluted excluding cocoa products Sugar and confectionary (including cocoa products) Singapore Food for infants and small children Slovakia Eggs and egg products Fats and oils of animal and vegetable origin (excluding butter) Fish and other seafood (including amphibians, reptiles, snails and insects) Meat and meat products (including edible offal) Milk and dairy products Slovenia Eggs and egg products Fats and oils of animal and vegetable origin (excluding butter) Fish and other seafood (including amphibians, reptiles, snails and insects) Food for infants and small children Fruit and fruit products Meat and meat products (including edible offal) Milk and dairy products Nuts and oilseeds Products for special nutritional use Sugar and confectionary (including cocoa products) Vegetables and vegetable products (including fungi) Spain Eggs and egg products Fats and oils of animal and vegetable origin (excluding butter) Fish and other seafood (including amphibians, reptiles, snails and insects) Milk and dairy products Sweden Cereals and cereal-based products Composite food (including frozen products) Eggs and egg products Fats and oils of animal and vegetable origin (excluding butter)

No. of data submitted 6 12 12 24 708 198 18 6 66 36 111 25 278 492 286 30 114 798 168 12 30 42 48 30 30 12 95 50 6 55 120 24 474 558

No. of detectionsa 0 0 0 0 0 0 0 0 0 0 70 6 78 161 127 18 51 223 38 4 23 22 4 30 5 0 0 0 6 0 72 24 424 385

196

WHO Food Additives Series No. 71-S1, 2016

Supplement 1: Non-dioxin-like polychlorinated biphenyls

Food category (by country) Fish and other seafood (including amphibians, reptiles, snails and insects) Food for infants and small children Fruit and fruit products Meat and meat products (including edible offal) Milk and dairy products Products for special nutritional use Starchy roots and tubers Vegetables and vegetable products (including fungi) United Kingdom Cereals and cereal-based products Composite food (including frozen products) Eggs and egg products Fats and oils of animal and vegetable origin (excluding butter) Fish and other seafood (including amphibians, reptiles, snails and insects) Fruit and fruit products Meat and meat products (including edible offal) Milk and dairy products Nuts and oilseeds Products for special nutritional use Starchy roots and tubers Sugar and confectionary (including cocoa products) Vegetables and vegetable products (including fungi) a

No. of data submitted 1 818 144 42 168 372 66 24 48 30 978 360 114 1 704 30 1 782 396 6 30 6 6 84

No. of detectionsa 1 802 118 42 125 308 24 24 48 22 848 293 73 1 615 18 1 149 297 4 17 5 6 62

Frequency of detection (%) (six indicator congeners) 99 82 100 74 83 36 100 100 73 87 81 64 95 60 64 75 67 57 83 100 74

Detection frequency was impacted by the LODs attained with the methods used by individual countries.

Fig. 3 Indicator PCB congener concentrations in fish and fish products

197

Safety evaluation of certain food additives and contaminants Eightieth JECFA

Fig. 4 Indicator PCB congener concentrations in egg and egg products

Indicator congener concentrations were more consistently distributed in meat and meat products, and no clear increase in concentration was associated with degree of chlorination (Fig. 5). Although similar to the fish products, the median values for the indicator congeners in the meat and meat products were very low relative to the maximum concentrations. Variability in the concentration patterns was observed in milk and milk products, and maximum values were more than an order of magnitude lower than those observed in fish and meat products (Fig. 6). Fig. 5 Indicator PCB congener concentrations in meat and meat products WHO Food Additives Series No. 71-S1, 2016 198

Supplement 1: Non-dioxin-like polychlorinated biphenyls

Fig. 6 Indicator PCB congener concentrations in milk and milk products

199

8. Dietary exposure assessment 8.1 Introduction and background

An evaluation of the dietary exposure to NDL-PCBs was undertaken by the Committee. The focus of this evaluation was the six indicator PCBs – namely, PCBs 28, 52, 101, 138, 153 and 180. PCB 128 was also included where relevant and possible. This selection of PCBs for inclusion in the dietary exposure assessment was based on the availability of toxicological and exposure data. The six indicator PCBs make up around 50% of total NDL-PCBs in food (EFSA, 2005). For both the occurrence data submitted to the GEMS/ Food database and the available estimates of dietary exposure, a range of NDLPCBs was included; however, the majority of this information was for the six indicator PCBs. This may be as a result of the setting of maximum levels by the European Commission (2011) for the six indicator PCBs, along with required monitoring, as much of the available information was from European countries. The information on the NDL-PCBs other than the six indicator congeners was much less substantial. As outlined in the section above on occurrence, the predominant foods containing NDL-PCBs are those of animal origin, such as fish and seafood, eggs, milk and dairy products, meat and offal. The lowest concentrations are in foods of plant origin (EFSA, 2010). Owing to their lipophilic nature, the NDL-PCB residues are usually found in the fatty portion of foods. It was therefore important to evaluate estimates of dietary exposure in which these types of foods were included. 8.2 Methods 8.2.1 Overview of the methods for the NDL-PCBs exposure assessment

Both national and international assessments of dietary exposure were reviewed by the Committee. These included those sourced from the literature, information provided to JECFA and estimates calculated by the Committee. Based on the outcomes of the toxicological assessment by the Committee, it was determined that only chronic dietary exposure assessments were required for the evaluation. For NDL-PCBs, there are no TEFs for use in the assessment of dietary exposure to adjust for varying toxicities of different congeners when the sum of a number of congeners is used for the assessment. This is unlike the process used for DL-PCBs (WHO, 1998; van den Berg et al., 2006). Estimates of dietary exposure to NDL-PCBs have been reported for whole populations, as well as adults and children separately, and some “lifespan” estimates have been reported. Therefore, a review of all of these groups was 200

Supplement 1: Non-dioxin-like polychlorinated biphenyls

included in the evaluation. PCBs are transferred from mother to child through breastfeeding; therefore infants, both breastfed and formula fed, were also considered, where possible, as a specific subgroup. Populations with potentially higher exposures were also evaluated where information was available. The occurrence data indicated that animal foods, particularly the lipid portions, have the highest concentrations of NDL-PCBs. Many of the dietary exposure assessments included a focus on animal foods, but often were not limited to these food groups. Where the estimates of dietary exposure were focused just on animal and animal-based foods, the results were still included in this evaluation, as they were considered to include the key contributors to dietary exposure. Less emphasis was placed on estimates of dietary exposure that did not include animal-based foods; however, this was assessed on a case-by-case basis, because for some countries, non-animal foods may make up a larger portion of the diet (e.g. high vegetable consumption in China; Xing, Wu & Wong, 2010), in which case they would be important inclusions in a dietary exposure assessment. A literature review was undertaken to source published papers with estimates of dietary exposure to NDL-PCBs in any combination. The EFSA (2005) review of NDL-PCBs included dietary exposure estimates from a number of European countries; therefore, the EFSA report was used to obtain key estimates of relevance to this assessment as a starting point. The literature search aimed to source any later published papers from Europe or papers from non-European countries. The countries for which estimates of dietary exposure obtained from the literature included the six indicator PCBs included Austria, Belgium, China (including Hong Kong Special Administrative Region), Egypt, France, Germany, Italy, the Netherlands, Norway, Serbia, Slovakia, Spain and Turkey. Many of the estimates of dietary exposure reviewed from the literature included different combinations of PCBs. For this evaluation, estimates covering the sum of the six indicator PCBs were included. Also included were estimates based on the sum of seven PCBs (i.e. the six indicator PCBs plus PCB 118, a DL-PCB), with a conversion factor of 0.85 applied to express the exposure in the equivalent form of the six indicator PCBs (EFSA, 2005). A number of national and international dietary exposure estimates were also calculated specifically for this evaluation. These were conducted using concentration data submitted via the GEMS/Food concentration database. For the national estimates of dietary exposure, concentration data from individual countries were summarized and combined with country-specific consumption data from the FAO/WHO Chronic Individual Food Consumption database – summary statistics (CIFOCOss). More details of how this was done are provided below. For international estimates of dietary exposure, the concentration data were pooled for countries according to the 17 GEMS/Food cluster diets, then summarized and combined with the consumption data for each cluster to estimate 201

Safety evaluation of certain food additives and contaminants Eightieth JECFA

dietary exposure. The specific methods used to summarize the GEMS/Food data and match them with the relevant consumption data are described further below. The aim of the international assessment is to provide information on clusters of countries that may have higher dietary exposures compared with other clusters of countries and to evaluate key contributing foods to dietary exposure in different regions. It also provides useful information about possible levels of dietary exposure for countries for which no national data exist. Calculations of national and international dietary exposures were done deterministically using mean concentrations multiplied by mean food consumption per person. A common approach of multiplying mean exposures by a factor to obtain an approximate high-percentile exposure (WHO, 1985; FAO/WHO, 2009) was used and is outlined further below. The national and international estimates of dietary exposure, although both based on GEMS/Food concentration data, are not directly comparable because of the different bases for the consumption data. Where national estimates of dietary exposure were calculated by the Committee, this was done both for the sum of the six indicator PCBs as well as for each of these six indicator PCBs individually. The estimates for the individual congeners were used to estimate body burdens of individual congeners. Individual values were needed, given the different ways in which the congeners behave in the body, including their metabolism and differing half-lives. The main foods contributing to dietary exposures were also determined. These may have been reported in studies in the literature. The majority (around 90%) of total exposure to NDL-PCBs comes from food (EFSA, 2005). Other routes of exposure, such as dermal and inhalation routes, were assessed where information was available. Assessing the impact of potential maximum levels in foods on dietary exposure was not required for this assessment. WHO Food Additives Series No. 71-S1, 2016 202

assessment

8.2.2 How the GEMS/Food concentration data were used for the dietary exposure

Concentration data for NDL-PCBs from the GEMS/Food database available for use in the dietary exposure assessment were from Australia, Austria, Belgium, Canada, China (including Hong Kong Special Administrative Region), Cyprus, Czech Republic, Estonia, Finland, France, Germany, Greece, Iceland, Ireland, Italy, Japan, Luxembourg, the Netherlands, New Zealand, Norway, Poland, Republic of Korea, Romania, Singapore, Slovakia, Slovenia, Spain, Sweden and the United Kingdom. The national data sets varied widely in the range of congeners reported. Many included only the six indicator PCBs. A range of food groups was included, ranging from only one food for a country to just animal-based foods to a broad range of foods. Some data were derived using analytical methods with a

Supplement 1: Non-dioxin-like polychlorinated biphenyls

greater sensitivity than others (i.e. had lower LODs). These data were used by the Committee to calculate estimates of dietary exposure, with edits and exclusions made where required for the assessment as outlined below. There were over 125 000 data points for NDL-PCBs in the data set; around 90% of these were for the six indicator PCB congeners (with an even proportion for each of the six congeners). The remainder of the data set was made up of results for 36 other congeners and mixtures. For the dietary exposure assessment, data for total PCBs, total NDL-PCBs and PCB mixtures were excluded. Data for individual congeners were retained, allowing them to be combined and summarized as required for the dietary exposure assessment. Individual data for the six indicator PCB congeners and PCB 128 were used for this evaluation. The majority (around 85%) of analytical results were from European countries, and the rest were from the non-European countries noted above. The majority of data points related to “individual” samples of food. In the majority of cases, it was not specifically stated if these were single purchase units or composite samples. For some data points, it was noted that the result was for a composite (e.g. composite of three samples). Some of the data were provided as aggregates or summaries of many individual results, where only summary statistics from all analytical results were provided (i.e. number of samples, means and percentiles). Aggregated data were not used in the national assessments where those countries had also provided individual (non-aggregated) data or where the data did not relate to a food or PCB congener of relevance. Aggregated data were not used in the international assessments, as they could not be pooled with the individual data points from other countries for deriving summary concentrations for each cluster. The data needed to be manipulated in order to make them consistent and in a usable format for summarizing and using in the dietary exposure assessment. This is outlined below. All data points were converted to the same concentration units (ng/kg). Data were available for the years 1995–2014. No data were excluded because they were gathered at earlier times. This is despite evidence that concentrations have decreased over time. However, data from earlier time periods have still contributed to the body burden of lifetime exposure for older persons and are relevant for lifetime or population-based estimates of dietary exposure. Analytical results below the LOD (non-detects or left-censored data) made up around 24% of the data points. However, there was considerable variation in the LODs reported by different countries. To enable concentration data to be summarized, a numerical value was required to be assigned to the non-detects. For the dietary exposure calculations, two scenarios were defined, one where non-detects were assigned a zero concentration (lower bound) and one where they were assigned the value of the LOQ (upper bound). The actual 203

Safety evaluation of certain food additives and contaminants Eightieth JECFA

concentration will be somewhere between the lower bound and the upper bound. Some samples had high LOQs (up to around 80 µg/kg). Substituting the value of the LOQ for non-detects could lead to a very conservative estimate of the upper bound of dietary exposure. It was determined by the Committee that analyses with LODs of 1 µg/kg (equivalent to an LOQ of around 3.5 µg/kg) or less represented more reliable results. Therefore, dietary exposures were estimated after eliminating samples with LOQs above this level. A small number of samples (0.2%) had a numerical concentration provided that was between the LOD and LOQ (i.e. a “trace” result). These concentrations were accepted and used in relevant calculations. Where estimates of dietary exposure were calculated by the Committee, the lower-bound scenario (where a concentration of zero was assigned to any result below the LOD) was used to determine the major contributors in order to not exaggerate the contribution of food groups that had non-detectable levels of NDL-PCBs. Where concentrations were expressed on a lipid basis (45% of samples), concentrations were converted to a fresh weight basis to make them consistent with the food consumption data. The conversions used food composition data on the proportion of fat in the food from a range of data sources (e.g. as provided in the GEMS/Food database for the specific food, United States Department of Agriculture, Food Standards Australia New Zealand, European Food Information Resource Network). Some concentrations were expressed on a dry weight basis (e.g. shrimp, milk powder) and therefore were converted to a fresh weight basis to match the food consumption data. This conversion used a factor derived from the difference in moisture content between the dried and fresh versions of the food. The moisture content was derived from food composition tables. The majority of results were provided for foods analysed in their raw state (26%) or unknown (72%), with around 2% of results relating to cooked foods. No conversion of concentrations from a cooked to raw state was made in order to make the samples match the form for the majority of samples in the data set, because the small number of data points would have had a very minor impact on the estimated dietary exposures. Concentration data were matched to consumption data for the raw commodity in the majority of cases. Animal foods, such as meat, eggs and seafoods, are usually consumed cooked; therefore, the effect of cooking on NDL-PCB concentrations needs to be taken into account when interpreting the dietary exposure estimates. Where unique sample identifiers were provided for each data point, the sum of the six indicator congeners for each individual sample was determined before deriving mean concentrations per food group. Therefore, if concentration data for one of the six congeners were missing for a sample, this sample was

204

WHO Food Additives Series No. 71-S1, 2016

Supplement 1: Non-dioxin-like polychlorinated biphenyls

excluded from the data used to derive the mean for the food group in question. In addition, where the assessment excluded samples with LOQs above 3.5 µg/kg, the whole sample was excluded if one of the six congeners fell into this category. This method of summing the six indicator PCB concentrations by sample unique identifier was used for 25 of the 31 countries that submitted data. However, some data sets did not provide unique identifiers for samples (Canada, Japan, Republic of Korea and Singapore). In these cases, the mean concentration for each food group for each of the six indicator congeners was determined separately in the first instance, then the means were summed for the six congeners. One limitation of this approach is that there may have been a different number of samples for each food for each of the six congeners. Concentration data for foods for infants were also included in the data set. There were two rows of data for human milk that were aggregate data from New Zealand for two of the indicator PCBs (PCBs 28 and 101). For infant formula, data were submitted for both powder and liquid formulas. Where the PCBs analysed were mixtures, these data points were excluded (eight data points for Australia). There were 80 data points from Canada for a range of congeners in infant formula; however, there were only data for five of the six indicator congeners. Concentration data on infant formula were available from China (n = 822 data points), Czech Republic (n = 12 data points), France (n = 24 data points), Sweden (n = 30 data points) and Singapore (n = 35 data points). For follow-on formula, there were a small number of data points (n = 12) from France. There were around 900 data points for foods for infants (cereal, vegetable, dairy or meat based, etc.). Most results in the concentration data set were for the edible portion of the food only (99%), with the remainder relating to the edible plus inedible portion. Based on further analysis of these data, no conversions were made of concentrations given for edible plus inedible to be representative of just the edible portions. This was for a number of reasons, including that “edible plus inedible” was used mostly for seafoods, the data were for foods that included shells, etc., such as oysters, and were unlikely to be analysed with the shell, and the small number of results would have had a very minor impact on the estimated dietary exposures. A small number of data points (2%) were indicated to be from “targeted” sampling. The remainder were “random” samples or unknown. There could be many different reasons for targeting samples, including investigation of a contamination incident or targeting analysis of specific foods in which the food chemical in question is known to occur. The database does not specify the reason. Results from targeted sampling were not excluded from further analyses. The small number of data points would have had a very minor impact on the estimated dietary exposures. 205

Safety evaluation of certain food additives and contaminants Eightieth JECFA

No weightings were applied to the concentrations, as there was insufficient information about the available data to allow this to be done. There were some outliers in the concentration data sets. Outliers will have the effect of skewing mean concentrations upwards, and therefore dietary exposure estimates will be inflated. Some outliers were in the data sets for countries for which a dietary exposure could not be calculated and therefore did not influence the evaluation or conclusions. The treatment and effect of the outliers on the dietary exposure are discussed in the text where relevant to the assessment. 8.2.3 National estimates of dietary exposure using the FAO/WHO Chronic Individual Food Consumption database – summary statistics (CIFOCOss)

The CIFOCOss data were used for the national estimates of dietary exposure calculated by the Committee. The diets were for individual countries and are based on national food consumption survey data from individuals with 2 or more days of data (for further details, see http://www.who.int/foodsafety/databases/ en/). The consumption data were combined with the GEMS/Food concentration data to estimate national dietary exposure to NDL-PCBs, where both data sets were available for the country. Consumption data were available for 27 countries; however, not all of these countries had submitted concentration data to the GEMS/Food database. National estimates of dietary exposure were derived for Belgium, China (excluding Hong Kong Special Administrative Region), Cyprus, Czech Republic, Finland, France, Germany, Greece, Ireland, Italy, Japan, the Netherlands, Republic of Korea, Spain, Sweden and the United Kingdom. National estimates were not conducted where either no CIFOCOss data were available (Austria, Canada, Estonia, Iceland, Luxembourg, New Zealand, Norway, Poland, Romania, Singapore, Slovakia and Slovenia) or the concentration data for the country were not suitable, sufficient or in a usable format (e.g. Australia only had NDL-PCB mixture data, New Zealand only had two aggregate data points for human milk, too few samples for Spain). Food consumption data in the CIFOCOss database have been classified using a combination of raw commodity classification codes and classifications for some processed foods. Each food in the GEMS/Food concentration database was assigned to a food in the most specific level of the CIFOCOss classification structure (Level 3) to enable them to be summarized and mean concentrations to be derived for each classification for use in the dietary exposure calculations. Some foods in the CIFOCOss data set could have more than one classification code, which could include raw commodity codes and/or processed food classifications. For example, the possible classification codes that could be used for consumption data for butter are 2.2.1 as a processed commodity, ML0100 Mammalian fat or

206

WHO Food Additives Series No. 71-S1, 2016

Supplement 1: Non-dioxin-like polychlorinated biphenyls

ML0812 Cattle fat. Therefore, all possible codes were evaluated for the relevant country to determine where the consumption data were reported, and then the concentration data were assigned that same code. The major foods in which NDL-PCBs are detected are foods of animal origin at the raw commodity level. Many of these foods, such as eggs, milk and meat, are used as ingredients in mixed dishes. It was originally requested that food consumption data for mixed dishes be broken down by recipes for the purposes of the CIFOCOss data set so consumption of the individual ingredients would be captured under each raw commodity classification. However, in the consumption data set, some data related to composite dishes were included. The mean consumption amount for all respondents (consumers and non-consumers) in grams per kilogram of body weight per day for each food was used to allow the exposure to be summed across food groups in order to estimate exposure from the whole diet, or from as much of the diet as captured by the concentration data. By using the consumption data on a gram per kilogram of body weight basis, the relevant body weights from each country were implicitly taken into account. As the high-percentile consumption data provided in the CIFOCOss data set could not be used to sum exposures across food groups, a factor of 2 was applied to the mean exposure estimate to approximate potential high-percentile exposures. This factor approximates the 90th percentile. In some cases, there may have been concentration data available for a food, but no consumption data – for example, for specific types of seafood for toddlers or young children. Food consumption data for a number of different age groups were represented in the CIFOCOss data set. These included population subgroups such as toddlers, children, adolescents, adults, elderly and very elderly. Some countries provided specific age ranges for their data; others from Europe follow the EFSA-defined age ranges for the population group descriptors (EFSA, 2011). Therefore, dietary exposure was estimated for a range of ages in the population where the data existed for the countries with concentration data available. The data were not split by sex. 8.2.4 International estimates of dietary exposure using the GEMS/Food cluster diets

The GEMS/Food cluster diets were used for the international estimates of dietary exposure calculated by the Committee. GEMS/Food supranational model diets, called cluster diets, are based on food balance sheet data from individual countries and are for mean amounts of food available for consumption per capita. The diets include consumption amounts of raw or minimally processed foods. They do not include consumption data for any population subgroups by either age or sex; 207

Safety evaluation of certain food additives and contaminants Eightieth JECFA

therefore, they do not allow for separate assessments for groups such as children or infants. Further information on the cluster diets can be found at http://www. who.int/foodsafety/databases/en/. Lower- and upper-bound estimates of international dietary exposure were derived deterministically using lower-bound mean and upper-bound mean concentrations of the six indicator PCBs for relevant foods in each cluster food group, where available. Consumption data for the clusters were not available on a body weight basis; therefore, estimates of dietary exposure were divided by a standard body weight of 60 kg to enable them to be expressed per kilogram of body weight. There are 17 GEMS/Food cluster diets. The individual data points from the GEMS/Food concentration data were pooled from the countries that were in the same cluster to determine mean concentrations per food group for each cluster. Concentration data were available for only seven of the 17 clusters from 27 individual countries. The seven clusters (and concentration data sources) were cluster 6 (Greece), cluster 7 (Finland, France, Iceland, Luxembourg, Norway, United Kingdom), cluster 8 (Austria, Germany, Poland, Spain), cluster 9 (China, including Hong Kong Special Administrative Region), cluster 10 (Canada, Cyprus, Estonia, Italy, Japan, Republic of Korea), cluster 11 (Belgium, the Netherlands) and cluster 15 (Czech Republic, Ireland, Romania, Slovakia, Slovenia, Sweden). There were no data from the other clusters, which covered mostly African and South American countries. Data from some countries were excluded from the data set if they were aggregate results, did not contain concentrations for the six indicator PCBs or were not of sufficient quality. In total, there were over 124 000 data points that could be assigned to relevant clusters and the cluster food groups. There were 2700 data points for cluster 6, 30 459 for cluster 7, 50 294 for cluster 8, 10 392 for cluster 9, 13 083 for cluster 10, 5058 for cluster 11 and 12 096 for cluster 15. Where all of the countries in the one cluster had data for the six indicator PCBs, as well as sample unique identifiers for each sample, the sum of the six congeners was calculated per sample first, then the mean concentration for the food group for that cluster was derived (clusters 6, 7, 8, 11 and 15). Where all countries in the cluster had data for the six indicator PCBs, but did not have sample unique identifiers for all countries in the cluster, the mean concentration for each individual indicator congener for a food group from all country data for that cluster was derived, then the sum of the six congeners was calculated (clusters 9 and 10). Where there were no data points for a cluster (clusters 1, 2, 3, 4, 5, 12, 13, 14, 16 and 17), the data from all clusters were pooled to obtain mean concentration data for the dietary exposure calculation for these clusters (n = 124 082 data points). This is based on the assumption that there is a global food

208

WHO Food Additives Series No. 71-S1, 2016

Supplement 1: Non-dioxin-like polychlorinated biphenyls

supply with many foods traded between countries, and concentration data from one area of the world are representative of the concentrations in foods consumed in other areas of the world. 8.3 Estimates of dietary exposure 8.3.1 National estimates of dietary exposure

National estimates of dietary exposure to NDL-PCBs were from two sources – firstly, the literature, and secondly, the estimates calculated by the Committee. EFSA (2005) derived national estimates of dietary exposure to NDLPCBs for some European countries and concluded that average daily exposures to total NDL-PCBs for general adult populations were in the range of 10–45 ng/ kg bw per day, depending on the country, and for children, 27–50 ng/kg bw per day. They also concluded that high meat eaters had an exposure around 20 ng/kg bw per day, and high fish eaters, 35 ng/kg bw per day. EFSA (2012) more recently estimated dietary exposure to NDL-PCBs across 17 member states for various population groups, based on the most recent monitoring data at that time. The mean dietary exposure to the sum of the six indicator PCBs was estimated to be between 4.3 and 25.7 ng/kg bw per day, depending on the country and population subgroup. This was very similar to the conclusions of the 2005 report, where mean estimates of exposure to the sum of the six indicator PCBs were summarized as 5–23 ng/kg bw per day (50% of the total NDL-PCB exposure of 10–45 ng/kg bw per day). Dietary exposures at the 95th percentile were estimated as 7.8–53.7 ng/kg bw per day for the sum of the six indicator PCBs. A literature review identified a number of published papers with national estimates of dietary exposure for a range of NDL-PCBs in varying combinations, including individual congeners and the sum of a number of congeners (e.g. three, six, seven or 23 congeners). Some estimates were for usual long-term exposures (based on statistical adjustment methods), most were deterministic calculations, some used individual dietary records and others used duplicate diet approaches. In relation to food consumption data, a range of sources was used, including national consumption surveys, 7-day surveys, market baskets, duplicate diet studies or food balance sheet data. Many of the assessments focused on foods of animal origin; however, some also included a broader range of foods. Some estimates based on the sum of seven PCBs have been included here, with a conversion factor of 0.85 applied to express the exposure to the equivalent amount of the sum of the six indicator congeners (EFSA, 2005). National dietary exposures were also estimated by the Committee where GEMS/Food concentration data and consumption data from CIFOCOss were 209

Safety evaluation of certain food additives and contaminants Eightieth JECFA

available. Food consumption data were provided according to both the CIFOCOss classification system, which for most food groups is based on the Codex raw commodity classification system, as well as the classification system in the Codex General Standard for Food Additives, for some processed commodities. Not all foods that had consumption data had associated NDL-PCB concentration data, and not all concentration data had corresponding food consumption information (e.g. some seafoods for young children). Estimates of national dietary exposure focusing on the sum of the six indicator PCBs, either from the literature or calculated by the Committee, were from a range of countries, including Belgium, China, Cyprus, Czech Republic, Denmark, Egypt, Finland, France, Germany, Greece, Ireland, Italy, the Netherlands, Norway, Republic of Korea, Serbia, Slovakia, Spain, Sweden and the United Kingdom. A summary of the national estimates of dietary exposure for the sum of the six indicator PCBs can be found in Table 19 for the published dietary exposure estimates and Table 20 for the national estimates calculated by the Committee. Further details can be found in Appendix 3. Overall, mean national estimates of dietary exposure to the sum of the six indicator PCBs from the literature for adults ranged between 1 and 18 ng/kg bw per day (lower bound to upper bound), with high-percentile exposure ranging between 8 and 45 ng/kg bw per day. For children, the dietary exposure estimates were slightly higher, with a mean exposure of 3–24 ng/kg bw per day and a highpercentile exposure of 12–87 ng/kg bw per day. Lifelong averaged exposures estimated by one country (the Netherlands) were within this range, being 6 ng/ kg bw per day at the mean and 10–12 ng/kg bw per day for a high consumer. The estimates of national dietary exposures calculated by the Committee covered a slightly broader range than the literature values. Mean exposures for adults ranged from <1 to 25 ng/kg bw per day, and high exposures ranged from <1 to 51 ng/kg bw per day. For children (also including adolescents), mean exposures ranged from <1 to 82 ng/kg bw per day, and high exposures from 1 to 163 ng/kg bw per day. Some estimates of dietary exposure made by the Committee were lower than those from other sources. This is due to the different consumption and concentration data sets used, along with the different methodologies used to estimate the exposures. Even where the same concentration data sets were used (e.g. for estimates for some European countries), the concentrations were assigned to different food groups, and the classifications assigned to foods in the data submitted to GEMS/Food may not have had a direct match in the CIFOCOss consumption database, which will lead to different estimates. Better matching between and use of the two input data sets could be done if there were more metadata for both the concentration and consumption data sets and more detail in the data descriptors that would allow better interpretation of the data.

210

WHO Food Additives Series No. 71-S1, 2016

Table 19 Summary of national estimates of dietary exposure to sum of the six indicator PCBs from the literature Population groups (age range Estimated dietary exposure in years) (ng/kg bw per day) Major contributors Reference Fish and fish products (23–27%), Mihats et al. Mean Children (6–15) cheese (18–23%), butter (2015) Children = 3.4 Adult males (19–65) (19–22%), meat (15–23%), milk Adult males = 2.6 Adult females (19–65) (11% children, 5–6% adults), Adult females = 3.2 eggs (4–7%) Mean LB = 5.3 MB = 5.7 UB = 6.1 Median LB = 4.9 P99 LB = 16.1 2007: Mean = 1.88 (difference 2007: cereals 40%, aquatic 19%, of 7 across provinces) vegetables 14%, meat 13% Data submitted to JECFA Dairy, then fish (no proportions given) Cimenci et al. (2013)

Country Austria

Food concentration data used Range of animal foods (meat, poultry, game, offal, fish and fish products, milk and dairy, eggs and egg products) (n = 157 foods). Analysed 2006–2011. No NDs.

Consumption data used National food consumption data. Men and women a 24 h recall. Children a 3-day food record.

Belgium

Game meat, processed meat, liver and products, fish, crustaceans, molluscs, fish salads, fish products, vegetable oil, cakes/pies/ pastries. Analysed 2008. LB, MB, UB concentrations.

2004 Belgian food consumption Adults (15+) survey, 15+ years, 2 × 24 h recall, n = 3 083. C-SIDE, software for intake distribution estimation, used to determine usual intake.

China, mainland

Meat, eggs, milk, cereals, legumes, nuts, tubers, vegetables. Adults

Adult males (18–45) Total diet study 2007 and 2011. 3-day household weighed record. 12 provinces Mean consumption 8 foods derived.

China, Hong Kong Special Administrative Region Whole population

2005–2007 population-based consumption survey.

2011: Mean = 0.67 (difference of 26 across provinces) Mean 2.83 P95 9.48

Xu & Cai (2015)

Supplement 1: Non-dioxin-like polychlorinated biphenyls

Egypt

Range of food groups (pork, ruminant meat, poultry, fish, other seafood, dairy products, eggs and fat/oils). Analysed 2008–2010. Foods of animal origin: meat (beef, chicken), dairy (butter, cream, cheese), seafood (fish, crab, bivalves). Sampled in 2003. LB used (ND = 0).

Food balance sheet data for animal foods. Mean 60 kg body weight.

Mean = 4.36

2011: aquatic 48%, vegetables 17%, meat 13%, cereals 11% Fish 60.3%, other seafood 9.0%, poultry 8.6%, pork 7.5%, eggs 4.7%, ruminants 3.9%, fats and oils (3.8%) Fish 47%, chicken 40%

Loutfy et al. (2008)

211

212 Population groups (age range Estimated dietary exposure in years) (ng/kg bw per day) Major contributors Fish products (children 38%, Mean (sum of 6): Children 3–14 (n = 1 018) females 19–44 years 42%, other Children 3–14 years mean Females 19–44 (n = 389) adults 47%) 12.9, P95 27.3 All other adults 15+ (n = 1 085) Meat products (children 24%, Females 19–44 years mean females 19–44 years 22%, other 7.6, P95 16 adults 21%) All other adults 15+ years Dairy products (children 21%, mean 7.7, P95 16 females 19–44 years 20%, other adults 18%) Reference Arnich et al. (2009) Fruit and vegetables (children 9%, females 19–44 years 8%, other adults 7%) Eggs (children 6%, females 19–44 years 5%, other adults 4%) Adults: Mean = 2.71 P95 = 7.90 Children and teenagers: Mean = 3.77 Munich adults (19–58) n = 20 (10 females, 10 males) P95 = 11.7 MB Mean = 5.6 Median = 5.6 P95 = 8.9 Safety evaluation of certain food additives and contaminants Eightieth JECFA Seafood (children 2%, females 19–44 years 2%, other adults 3%) Adults: fish 59%, butter 7%, meat Sirot et al. (2012) 7%, cheese 6% Children/teenagers: fish 48%, butter 9%, meat 8%, milk 6%, other dairy/yoghurts 6%, cheese 6% Not provided Fromme et al. (2007, 2009b)

WHO Food Additives Series No. 71-S1, 2016

Table 19 (continued)

Country France

Food concentration data used Consumption data used 1 665 samples over 4 years of 1998–1999. 7 days. 3+ years. monitoring and other surveys. Mean concentrations. No NDs in results.

France

Range of foods from Total Diet Study, prepared ready for consumption.

National individual food Adults (18–79) consumption data (INCA2). 7-day Children and teenagers (3–17) food record.

Germany

Range of foods.

Integrated exposure assessment survey (INES). Duplicate diet for 7 days.

Country Germany

Consumption data used National Health Survey 2005– 2006 (NVZ-II) study of German consumption. Diet histories. Toddlers and young children (0.5–6) excluding breastfed Children (7–12) Adults (13–94) Adults: mean 10.9, median 11.0, P90 20.0, P95 23.8 For P95 exposed: Toddlers: fish 59%, milk 24%, vegetable oil 13% Children: fish 60%, milk 23%, vegetable oil 11% Adults: fish 42%, milk 24%, vegetable oil 16%, pork 6% Children: mean 16.1, P90 28.7, Children: fish 38%, milk 28%, vegetable oil 15%, pork 7% P95 33.8 Toddlers: milk 38%, fish 36%, Toddlers and young children: mean 24.6, P90 49.3, P95 60.0 vegetable oil 12% Fattore et al. (2008)

Population groups (age range in years) General public (14–80); n = 20 000 persons

Estimated dietary exposure (ng/kg bw per day) Major contributors Reference Mean = 10.86 Fish products 36%, dairy products Schwarz et al. 30%, vegetable products 7%, (2010, 2014) High consumer method = meat products 7% 22.61

Italy

Food concentration data used From Germany and other sources (if German data were not available). Range of foods. n = 545 individual food items. Cereals and products, fruits and vegetables, eggs, fats and oils, fish and seafood, meat and products, offal, milk and dairy. Post-1997 data.

3- to 7-day diary national consumption survey individuals 1994–1996, 0.5–94 years of age, n = 1940.

Italy 2, 10, 40, all

Range of foods, whole diet. Mid-1990s.

Duplicate diet, 3 non-consecutive Mean age (±SD) 38 ± 13 days, n = 20.

Mean = 18 P95 = 39 Σ6 congeners (derived by Σ7*0.85) 2 years: mean 11.5, median 10.3, P95 21.8 10 years: mean 7.8, median 5.8, P95 12.5 40 years: mean 4.9, median 4.4, P95 8.1 Lifelong averaged median: 4.8 P95 lifelong: 10.1 (based on LB concentrations)

Adults: fish 68%, milk 12%, vegetable oil 12% Not provided Meat 27%, fish 26%, industrial oils and fats 18%, dairy products 17%, eggs 5%, vegetable products 7%

Zuccato et al. (1999) Bakker et al. (2003); Baars et al. (2004); EFSA (2005)

Supplement 1: Non-dioxin-like polychlorinated biphenyls

Netherlands

Range of food groups (including meat, seafood, oils, nuts, vegetables) via analysis and recipes. Survey 1998–1999.

3rd Dutch National Food Consumption Survey 1997–1998, n = 6 250, 2-day dietary record. Usual intake via STEM.

213

214 Population groups (age range in years) 21–80; n = 184 (73 in representative group, 111 in high consuming group) Reference Kvalem et al. (2009) Per capita UB mean sum 6 = 3.4 Major contributors Semi-oily and oily fish For fish and seafood PCB 153, representative consumers 68%, high consumers 63% Wholegrain wheat flour (79%), but that is of three food groups: flour, oil, sugar Skrbic & DurisicMladenovic (2007) Estimated dietary exposure (ng/kg bw per day) Median 5.2 whole group LB mean: representative consumers 4.3, high consumers 6.4 LB mean sum 6 = 2.9 Average inhabitant Children 4–6, 7–11, 12–15 70 kg adult 4–6 years: mean 30, P95 87 7–11 years: mean 20, P95 56 12–15 years: mean 17, P95 47 Mean calculated from raw data Fish and seafood 91%, milk 2% in the report: 15 Llobet et al. (2008) Note: underestimates, because no animal products included; article shows meat, milk and fish make up 36% of market basket. Fats (47%), eggs and egg Adults: mean 17, P95 45 products (28%) Children: Salgovicová & Pavlovicová (2007) Safety evaluation of certain food additives and contaminants Eightieth JECFA Butter only: mean 0.059 Milk, dairy and butter: mean 0.183 Not provided Uçar et al. (2011)

WHO Food Additives Series No. 71-S1, 2016

Table 19 (continued)

Country Norway

Consumption data used Semiquantitative food frequency questionnaire.

Serbia

Food concentration data used Dairy, eggs, meat, vegetable oils, lean fish, oily fish, fish liver, shellfish, fish oils, seagull eggs, other. Wholegrain wheat flour, wheat bran, sunflower oil, sugar beet, molasses, sugar. 2002 analysis. LB and UB. Nationally analysed foods only (no imports included).

Mean daily per capita consumption of foods analysed, plus proportion of mixed foods where they were ingredients (21% of Serbian market basket). Mean body weight of 60 kg assumed.

Slovakia

Animal origin. 48 groups.

National consumption data, food balance sheets.

Spain

Turkey

Range of foods (meat, fish, milk, 2002–2003 Catalan Nutrition Sur- Adults dairy, cereals, pulses, vegetables, vey (Serra-Majem et al., 2003). Mean body weight = 70 kg tubers, fruits, fats and oils, bakery products). 2002–2003 Catalan Nutrition Survey. Milk, dairy products and butter. Food consumption data from Turkish population various regions in Turkey. No further details provided.

INCA2: Etude individuelle nationale des consommations alimentaires 2; LB: lower bound; MB: middle bound; ND: non-detects; P90: 90th percentile; P95: 95th percentile; P99: 99th percentile; SD: standard deviation; STEM: STatistical Exposure Model; UB: upper bound

Table 20 Summary of the national estimates of dietary exposure for the sum of the six indicator PCBs from the calculations conducted by the Committee at mean and high percentilea for various population groupsb

Country Belgium China Cyprus Czech Republic Finland

Infants High Mean exposure – – – – – – – – – –

Toddlers High Mean exposure 16–82 33–163 – – – – – – 11–11 21–21

Very elderly High Mean exposure 7–25 13–51 – – – – – – – –

General population High Mean exposure – – 0.5–0.5 1.0–1.0 – – – – – –

France Germanyc

– –

– –

– 9.3–10.8

1.0–1.2 3.8–4.2 – – 0.1–0.8

2.0–2.3 7.6–8.4 – – 0.2–1.5

– – – – –

– – – – –

Greece Ireland Italy 3.5–3.5 0.2–0.2 6.7–13.7 – – – 3.4–7.5 6.7–15.0 5.3–7.1 1.4–1.5 6.9–7.0 0.3–0.4 – – – – 2.6–2.6 – 5.2–5.2 – – – – –

Netherlands Republic of Korea Sweden 13.3– 27.4 –

– – 0.003– 7.4 – –

– – 0.007– 14.8 – –

– – 0.02–3.5

– 18.5– 21.6 – – 0.04–7.0 – – – –

Lower bound–upper bound dietary exposure (ng/kg bw per day) Children Adolescents Adults Elderly High High High High Mean exposure Mean exposure Mean exposure Mean exposure 12–63 25–125 11–30 22–60 8–24 15–47 6–23 12–46 1.0–1.0 1.9–2.0 – – – – – – – – 4–17 7–35 – – – – 0.7–5.2 1.4–10.4 0.6–3.5 1.1–6.9 0.3–1.8 0.7–3.6 – – 1.4–1.4 2.9–2.9 – – 0.8–0.8 1.6–1.6 1.7–1.7 3.5–3.5 DIPP DIPP 9.5–9.5 19–19 STRIP STRIP 1.3–1.7 2.5–3.3 0.6–0.8 1.2–1.6 0.7–0.9 1.5–1.8 1.0–1.1 1.9–2.3 5.9–6.2 11.9– 2.6–2.8 5.2–5.5 3.4–3.6 6.7–7.2 3.8–4.1 7.5–8.2 12.4 10–14 21–27 – – – – – – – – – – 0.8–1.2 1.6–2.4 – – 0.6–2.7 1.2–5.3 0.3–1.1 0.7–2.2 0.3–1.0 0.6–2.0 0.2–0.9 0.4–1.8 – – – – – – – – – 0.5–0.6 – –

3.7–3.7 –

7.4–7.5 –

– 0.9–1.1 – –

–

–

–

–

Supplement 1: Non-dioxin-like polychlorinated biphenyls

United Kingdom

–

–

–

–

10.5– 14.1 2.8–2.9

DIPP: Finnish Type I Diabetes Prediction and Prevention Nutrition Study; STRIP: Turku Coronary Risk Factor Intervention Project for Children a Notes: high exposure = mean exposure × 2, which approximates the 90th percentile. b Where specified based on the nutrition survey data, infants are <1 year, toddlers >1 to <3 years, children >3 to <10 years, adolescents >10 to <18 years, adults >18 to <65 years, elderly >65 to <75, very elderly >75 years. c Three surveys used for toddlers and children, all very similar, so only showing results based on the most recent 2008 consumption survey.

215

Safety evaluation of certain food additives and contaminants Eightieth JECFA

Overall, the main contributor to dietary exposure to the sum of the six indicator PCBs was fish and seafood for a large number of the countries assessed. Meat and meat products along with dairy products (mainly cheese, butter and/or milk) were also common contributors, although usually to a lesser degree and for fewer countries, compared with fish and seafood. This was for both the assessments in the literature and those estimated by the Committee. Information on food contributors from the various estimates of dietary exposure was presented in different levels of detail, from specific foods to food groups. Owing to the variable nature of concentrations and consumption patterns within and between countries, the contribution reported from different foods is highly variable. These factors, combined with the different methods used to estimate the dietary exposure and major contributors, make comparison of contributors within and between countries difficult. From both the reports in the literature and the dietary exposures estimated by the Committee, contributions for fish and seafood ranged from 2% to 100%, depending on the country and population group. For meat and meat products, contributions ranged from <1% to 100%, depending on the country and population group. For dairy products, cheese contributed 5–60%, butter contributed 2–54% and milk contributed <1–79%. For children, milk and dairy products made a greater contribution to dietary NDLPCB exposure, usually double the contribution compared with that for adults where both groups were assessed. (a) Austria

WHO Food Additives Series No. 71-S1, 2016 216

Dietary exposure to the sum of the six indicator PCBs was estimated for Austria from a range of animal-based foods for adults and children (Mihats et al., 2015). The estimated mean exposure was 3.4 ng/kg bw per day for children, 3.2 ng/kg bw per day for women and 2.6 ng/kg bw per day for men. The results for children were only slightly higher than those for adults. Although this is not the usual trend, children usually have higher exposures than adults on a body weight basis; this may be explained by the fact that there were 3 days of consumption data for children and only 1 day of data for adults. The major contributor to exposure was fish and fish products for all three population groups (23–27%). Children had double the contribution from milk (11%) compared with adults (5–6%) due to their higher consumption of this food. (b) Belgium

Estimated dietary exposures to the sum of the six indicator PCBs for Belgian adults (15 years and above) (Cimenci et al., 2013) ranged between 5.3 and 6.1 ng/ kg bw per day (lower bound [LB] to upper bound [UB] mean, respectively). High consumers (99th percentile) had a LB exposure of 16.1 ng/kg bw per day. The

Supplement 1: Non-dioxin-like polychlorinated biphenyls

main contributors to dietary exposure were dairy (as a whole group) followed by fish. EFSA (2012) also estimated dietary exposures for Belgium for a range of population subgroups using monitoring data. Mean dietary exposures (LB mean – UB mean) for toddlers were 10.8–11.8 ng/kg bw per day, for children, 9.3– 10.2 ng/kg bw per day, for adolescents, 3.7–4.5 ng/kg bw per day, and for adults and the elderly, 4.6–6.6 ng/kg bw per day. Ninety-fifth percentile exposures for toddlers were 20.0–23.5 ng/kg bw per day, for children, 24.4–26.9 ng/kg bw per day, for adolescents, 12.4–13.1 ng/kg bw per day, and for adults and the elderly, 14.7–17.4 ng/kg bw per day. For toddlers, milk contributed 44%, meat 25% and fish 21%. The same foods contributed for children (milk 36%, fish 30%, meat 24%). By adolescence, meat was the main contributor (37%), then fish (31%) and milk (20%). For adults and the elderly, fish contributed 43–54%, meat 21–29% and milk 11–16%. Estimates of exposure for Belgium from Cimenci et al. (2013) and EFSA (2012) were similar. The estimates conducted by the Committee were similar for the LB exposures, although the UB estimates were higher (up to 82 ng/kg bw per day for children and 25 ng/kg bw per day for adults at the mean). Despite the same consumption data set being used by EFSA (2012) and the Committee, the concentration data set was different, and foods may have been assigned differently to food groups. LOQs in the GEMS/Food concentration database ranged from 0.05 up to 3.3 µg/kg (fresh weight basis, when results with high LOQs over 3.5 µg/kg were excluded), whereas in the EFSA (2012) monitoring data, LOQs were <0.001–1.5 µg/kg on a fresh weight basis, which would have a large impact on the UB estimates, especially when 65% of the GEMS/Food data had non-detect values. The main contributors to dietary exposure, based on the calculations conducted by the Committee, were milk (60–65% children, 30–45% adults), cheese (10–36% children, 18–20% adults) and eggs for adults (15%) (see Appendix 3). (c) China

The estimated dietary exposures for China that were calculated by the Committee were very low, at a mean of 1.0 ng/kg bw per day for children and 0.5 ng/kg bw per day for the general population. The concentration data set for China included a broad range of foods, including a range of animal foods. The low exposure is a result of the very low LOD used for the analysis of the foods. The main contributors to dietary exposure for children were freshwater fish (26%), marine fish (13%), pork (13%), chicken eggs (12%), rice (10%) and wheat (10%). For the general population, the main contributors were freshwater fish (27%), marine fish (17%), pork (12%), rice (10%) and wheat (10%). 217

Safety evaluation of certain food additives and contaminants Eightieth JECFA

For the Hong Kong Special Administrative Region of China, there was one study from the literature that estimated dietary exposure to the sum of the six indicator PCBs (Xu & Cai, 2015). Dietary exposures were based on the concentrations analysed in a range of eight groups of animal-based foods. Mean dietary exposure was estimated as 2.83 ng/kg bw per day and 95th percentile exposure as 9.48 ng/kg bw per day for the Hong Kong Special Administrative Region adult population. The major contributors to dietary exposure were fish (60.3%), other seafood (9.0%), poultry (8.6%) and pork (7.5%). (d) Cyprus

WHO Food Additives Series No. 71-S1, 2016 218

The dietary exposure to NDL-PCBs for Cyprus was estimated by the Committee. Consumption data were available only for adolescents, and dietary exposures (LB–UB) were 4–17 ng/kg bw per day at the mean and 7–35 ng/kg bw per day for a high exposure. The main contributors to dietary exposure were milk (44%), chicken (36%) and pork (17%). EFSA (2012) also estimated dietary exposure for adolescents in Cyprus based on monitoring data. Mean exposures (LB–UB) were 4.7–5.9 ng/kg bw per day, and 95th percentile exposures were 15.9–17.1 ng/kg bw per day. The major contributors to dietary exposure were fish (45%), meat (32%) and milk (20%). The exposure estimates from the Committee and EFSA (2012) were based on the same consumption data, but different concentration data sets made up of different foods. There was a limited range of fish data in the GEMS/ Food database that matched with consumption data; therefore, the Committee’s estimates did not show fish as a major contributor. The estimated exposures were similar for LB means; however, the Committee’s UB estimates were higher. This could be attributed to the different data sets and the more stringent criteria used by EFSA (2012) to exclude data (i.e. results removed included those above LOQ 2 µg/kg, where there was a high per cent difference between LB and UB results and outliers). (e) Czech Republic

Dietary exposure to NDL-PCBs for the Czech Republic was estimated by the Committee for children, adolescents and adults. Mean dietary exposures (LB– UB) for children were 0.7–5.2 ng/kg bw per day, for adolescents, 0.6–3.5 ng/kg bw per day, and for adults, 0.3–1.8 ng/kg bw per day. High-percentile exposures for children were 1.4–10.4 ng/kg bw per day, for adolescents, 1.1–6.9 ng/kg bw per day, and for adults, 0.7–3.6 ng/kg bw per day. The main contributors to dietary exposure were butter (27–35%), pork (19–28%), eggs (18–20%) and unprocessed meat not further specified (8–12%).

Supplement 1: Non-dioxin-like polychlorinated biphenyls

EFSA (2012) also estimated dietary exposure to the sum of the six indicator PCBs. Mean exposures (LB–UB) were, for children, 12.4–15.5 ng/kg bw per day, for adolescents, 9.3–11.0 ng/kg bw per day, and for adults, 6.2–7.4 ng/kg bw per day. Ninety-fifth percentile exposures were estimated for children as 41.7–44.7 ng/kg bw per day, for adolescents, 29.2–30.8 ng/kg bw per day, and for adults, 19.7–21.3 ng/kg bw per day. The main contributors to dietary exposure were meat (39–46%), fish (33–34%) and milk (11–15%). (f) Denmark

EFSA (2012) estimated dietary exposure to the sum of the six indicator PCBs for children, adolescents, adults, the elderly and the very elderly. For children and adolescents, mean exposures (LB–UB) ranged between 6.0 and 12.8 ng/kg bw per day, and 95th percentile exposures ranged between 11.8 and 26.7 ng/kg bw per day. For all adult and elderly population groups, mean exposures ranged between 6.2 and 10.3 ng/kg bw per day, and 95th percentile exposures ranged between 10.8 and 36.5 ng/kg bw per day. The main contributors to dietary exposure for children and adolescents were meat (36–40%), fish (25–27%), milk (19–20%) and fats (14–16%); for adults and the elderly, the main contributors were fish (36–54%), meat (21–34%), milk (11–15%) and fats (12–14%). (g) Egypt

One estimate of dietary exposure for the sum of the six indicator PCBs for Egypt was reported in the literature (Loutfy et al., 2008). Mean exposure based on food balance sheet consumption data using concentrations from meat, seafood and dairy (excluding milk) was 4.36 ng/kg bw per day. Although the use of food balance sheet consumption data usually leads to an overestimate of dietary exposure, this estimate does not include other food groups known to contribute to NDL-PCB exposures, such as eggs and milk; therefore, actual exposures may be higher than those estimated. (h) Finland

Dietary exposures for a range of population subgroups were estimated by the Committee for Finland for the sum of the six indicator PCBs. Toddlers had the highest mean exposure, at 11 ng/kg bw per day. Mean exposure for adults was 0.8 ng/kg bw per day, and for the elderly, 1.7 ng/kg bw per day. For children, there were two consumption surveys; using the Finnish Type I Diabetes Prediction and Prevention Nutrition Study (DIPP) survey, the mean exposure was 1.4 ng/kg bw per day, and using the Turku Coronary Risk Factor Intervention Project for Children (STRIP) survey, it was 9.5 ng/kg bw per day. There was a large difference between the exposure estimates, which was the result of a high concentration 219

Safety evaluation of certain food additives and contaminants Eightieth JECFA

in a fish-based meal for which one of the surveys had a consumption value and the other did not. For the STRIP survey, 81% of the exposure was from fishbased meals, whereas for the DIPP survey, 54% of the exposure was from cheese and 30% from fish. High contributors to dietary exposure for adults were cheese (59%) and fish and seafood (40%), and for the elderly, fish and seafood (82%) and cheese (18%). Ready-to-eat meals containing meat/fish were the main contributor (92%) for toddlers. EFSA (2012) also estimated dietary exposures for Finland. Exposures (LB–UB) for toddlers were 13.6–15.9 ng/kg bw per day (mean) and 41.0–46.1 ng/kg bw per day (95th percentile); for children in the DIPP survey, 13.5–15.6 ng/kg bw per day (mean) and 33.8–37.0 ng/kg bw per day (95th percentile); for children in the STRIP survey, 7.4–8.2 ng/kg bw per day (mean) and 19.9–21.7 ng/ kg bw per day (95th percentile); for adults, 6.1–6.9 ng/kg bw per day (mean) and 18.6–19.7 ng/kg bw per day (95th percentile); and for the elderly, 7.5–8.1 ng/kg bw per day (mean) and 23.0–23.2 ng/kg bw per day (95th percentile). The food groups contributing to dietary exposure for toddlers were fish (40%), meat (33%) and milk (17%); for children, the two surveys produced similar contributions for fish (36–38%), meat (34–35%) and milk (18%). For adults and the elderly, the main contributor was fish (49% and 65%, respectively). (i) France

220

WHO Food Additives Series No. 71-S1, 2016

For France, there were two estimates from the literature. Arnich et al. (2009) estimated dietary exposures to the sum of the six indicator PCBs to be 12.9 ng/kg bw per day (mean) and 27.3 ng/kg bw per day (95th percentile) for children, 7.6 ng/kg bw per day (mean) and 16 ng/kg bw per day (95th percentile) for females aged 19–44 years, and 7.7 ng/kg bw per day (mean) and 16 ng/kg bw per day (95th percentile) for other adults. Fish products were the highest contributor to dietary exposure for all population groups (38–47%), followed by meat products (21–24%) and dairy products (18–21%). Sirot et al. (2012), from the French Total Diet Study, estimated dietary exposures for children and teenagers to be 3.77 ng/kg bw per day (mean) and 11.7 ng/kg bw per day (95th percentile) and for adults, 2.71 ng/kg bw per day (mean) and 7.90 ng/kg bw per day (95th percentile). The main contributor to dietary exposure was fish (48% for children and teenagers and 59% for adults). The dietary exposures estimated by the Committee for France were much lower, ranging from 0.6 to 3.3 ng/kg bw per day for children and adolescents, including both LB and UB and mean and high percentile, and ranging from 0.7 to 2.3 for adults and the elderly for the same range. The main contributors to dietary exposure for all population groups were butter (28–54%) and fish (34–69%).

Supplement 1: Non-dioxin-like polychlorinated biphenyls

EFSA (2012) also estimated dietary exposures for the sum of the six indicator PCBs (LB–UB) for children as 14.7–17.1 ng/kg bw per day (mean) and 31.6–33.6 ng/kg bw per day (95th percentile), for adolescents, 7.4–8.8 ng/ kg bw per day (mean) and 16.7–18.2 ng/kg bw per day (95th percentile), and for adults/elderly/very elderly, 6.7–8.4 ng/kg bw per day (mean) and 14.3–18.1 ng/ kg bw per day (95th percentile). The main contributors to dietary exposure for all population subgroups were fish (39–51%) followed by meat (28–40%). (j) Germany

From the literature, estimated dietary exposures for Germany (Fromme et al., 2007, 2009b) were 5.6 ng/kg bw per day (mean) and 8.9 ng/kg bw per day (95th percentile) for adults. Slightly higher estimates were generated from another study (Schwarz et al., 2010, 2014), where the mean for the general population was 10.9 ng/kg bw per day and for the high consumer, 22.6 ng/kg bw per day. Estimates of dietary exposure derived by the Committee were in a similar range, with means (LB–UB) for toddlers and children of 5.9–10.8 ng/kg bw per day, for adolescents, 2.6–2.8 ng/kg bw per day, and for adults/elderly/very elderly, 3.4–4.2 ng/kg bw per day. Estimates for high consuming adults were slightly lower than the published estimates, at 6.7–8.4 ng/kg bw per day. The major contributors to dietary exposure for adults and the elderly were fish (21–33%), then butter (14–15%). For children, meat was the highest contributor (29%); for toddlers, it was infant foods (23%). EFSA (2012) estimated dietary exposures for German population subgroups, with mean exposures (LB–UB) for children and toddlers of 8.8–11.8 ng/kg bw per day, for adolescents, 3.3–4.3 ng/kg bw per day, and for adults and the elderly, 4.3–6.2 ng/kg bw per day. Ninety-fifth percentile exposures (LB–UB) were, for toddlers and children, 19.2–23.4 ng/kg bw per day, for adolescents, 8.5–11.3 ng/kg bw per day, and for adults and the elderly, 14.4–18.6 ng/kg bw per day. The major contributors to dietary exposure were milk for toddlers (28%), meat for children and adolescents (35–50%) and both meat (27–39%) and fish (37–51%) for adults and the elderly. (k) Greece

Dietary exposures for Greek children (not including toddlers and infants) were estimated by the Committee. Mean exposures (LB–UB) were estimated as 10–14 ng/kg bw per day, with high consumer exposures at 21–27 ng/kg bw per day. Milk contributed 80% to the exposure, eggs 10% and fish 6%. EFSA (2012) also estimated dietary exposures for children. Mean exposures (LB–UB) were 8.4–9.6 ng/kg bw per day, and 95th percentile exposures (LB–UB) were 22.7–24.1 ng/kg bw per day. These estimates were very similar 221

Safety evaluation of certain food additives and contaminants Eightieth JECFA

to those estimated by the Committee. Major contributors to dietary exposure were milk (41%), fish (30%) and meat (23%). The contribution from fish was higher from EFSA (2012), whereas there was a limited range of fish and seafood included in the estimates by the Committee. (l) Ireland

Dietary exposures to the sum of the six indicator PCBs for Ireland were estimated by the Committee for adults only. Dietary exposures (LB–UB) were very low, with a mean exposure of 0.8–1.2 ng/kg bw per day and a high exposure of 1.6– 2.4 ng/kg bw per day. Major contributors were dietary supplements (32%), fish (22%), eggs (14%), butter (13%) and cheese (8%). Dietary supplements are a main contributor as a result of a mean concentration of around 24 000 ng/kg for this group, which included fish oil supplements. However, this food category may include consumption of supplemental beverages, rather than only dietary supplements, and therefore may be an overestimation of consumption of dietary supplements. There was no meat in the concentration data set; therefore, these estimates are likely to be an underestimate of exposure. EFSA (2012) also estimated dietary exposures for Irish adults. Mean exposures (LB–UB) were 4.8–6.7 ng/kg bw per day, and 95th percentile exposures (LB–UB) were 10.9–13.5 ng/kg bw per day. Major contributors were meat (49%), fish (25%) and milk (15%). (m) Italy

From the literature, one estimate of dietary exposure for adults was reported (n = 20 from a duplicate diet study), with a mean of 18 ng/kg bw per day and a 95th percentile of 39 ng/kg bw per day (Zuccato et al., 1999). Another estimate reported for adults (based on national consumption data) was a mean exposure of 10.9 ng/kg bw per day and a 95th percentile exposure of 23.8 ng/kg bw per day (Fattore et al., 2008). From this second study, exposures for toddlers and young children were 24.6 ng/kg bw per day (mean) and 60.0 ng/kg bw per day (95th percentile), and for children (7–12 years), 16.1 ng/kg bw per day (mean) and 33.8 ng/kg bw per day (95th percentile). The main contributors to dietary exposure for toddlers and young children were milk (38%) and fish (36%). For children (7–12 years), the main contributors were fish (38%) and milk (28%). For adults, the main contributors were fish (42%) and milk (24%). Estimates of dietary exposure for infants were calculated by the Committee. The estimated dietary exposure to the sum of the six indicator congeners was very low (mean LB–UB: 0.003–7.4 ng/kg bw per day). The GEMS/Food data set used for this assessment included concentration data for a range of animal-based foods, but the total exposure could be attributed to beef

222

WHO Food Additives Series No. 71-S1, 2016

Supplement 1: Non-dioxin-like polychlorinated biphenyls

consumption (LB scenario used) for infants and toddlers. NDL-PCBs were not detected in milk samples. There was a concentration assigned to milk at the UB resulting in a mean dietary exposure for infants of 7.4 ng/kg bw per day. Actual dietary exposures for infants are likely to fall somewhere between the LB and UB estimates, as they are not likely to consume all of their milk with a true zero concentration of NDL-PCBs, as is assumed in the LB scenario. For toddlers, mean exposures were 0.02–3.5 ng/kg bw per day, children, 0.6–2.7 ng/kg bw per day, adolescents, 0.3–1.1 ng/kg bw per day, and adults/elderly/very elderly, 0.1–1.0 ng/kg bw per day. The main contributor for adolescents and adults/elderly/very elderly was fish (≥92%). EFSA (2012) also estimated dietary exposures for a range of population groups. Estimated dietary exposures (LB–UB) for infants were 8.3–8.5 ng/ kg bw per day (mean) and 31.7–35.4 ng/kg bw per day (95th percentile), for toddlers, 16.9–19.2 ng/kg bw per day (mean) and 36.6–39.8 ng/kg bw per day (95th percentile), for children, 16.1–18.8 ng/kg bw per day (mean) and 45.4–47.2 ng/kg bw per day (95th percentile), for adolescents, 9.0–10.3 ng/kg bw per day (mean) and 36.4–36.4 ng/kg bw per day (95th percentile), and for adults and the elderly, 4.7–8.7 ng/kg bw per day (mean) and 13.2–33.0 ng/kg bw per day (95th percentile). The main contributors to dietary exposure were infant foods (39%) for infants, fish (37%) and milk (32%) for toddlers, and fish (51–58%) for children, adolescents, adults and the elderly. (n) Netherlands

From the literature, the one estimate of dietary exposure was determined for the sum of seven PCB congeners (six indicator PCBs and PCB 118) (Bakker et al., 2003; Baars et al., 2004; EFSA, 2005). These were converted to the equivalent amount of the sum of the six indicator congeners by Σ7*0.85. Mean and 95th percentile exposures were, respectively, for 2-year-olds, 11.5 and 21.8 ng/kg bw per day, for children 10 years of age, 7.8 and 12.5 ng/kg bw per day, and for adults 40 years of age, 4.9 and 8.1 ng/kg bw per day; lifelong averaged median and 95th percentile exposures were 4.8 and 10.1 ng/kg bw per day, respectively. Estimates of dietary exposure derived by the Committee were mean and high exposures, respectively, for toddlers of 3.7 and 7.5 ng/kg bw per day, for children, 3.5 and 7.0 ng/kg bw per day, and for adults, 2.6 and 5.2 ng/kg bw per day. The main contributors to dietary exposure were cheese for all population groups (23–49%), the highest being for toddlers. Fish was also a contributor for adults (16%). EFSA (2012) estimated dietary exposures (LB–UB) to be, for toddlers, 8.3–8.8 ng/kg bw per day (mean) and 18.2–20.5 ng/kg bw per day (95th percentile), for children, 7.3–7.8 ng/kg bw per day (mean) and 16.6–16.7 ng/kg 223

Safety evaluation of certain food additives and contaminants Eightieth JECFA

bw per day (95th percentile), and for adults, 3.8–4.5 ng/kg bw per day (mean) and 8.1–9.5 ng/kg bw per day (95th percentile). The main contributor to dietary exposure was meat for toddlers and children (34%) and adults (43%). (o) Norway

From the literature, Kvalem et al. (2009) estimated LB mean dietary exposure to be 4.3 ng/kg bw per day for a representative group in the population and 6.4 ng/ kg bw per day for high consumers; a median dietary exposure of 5.2 ng/kg bw per day was estimated for the whole group. Semi-oily and oily fish were the key contributors to dietary exposure. (p) Republic of Korea

WHO Food Additives Series No. 71-S1, 2016 224

The estimated dietary exposures for the Republic of Korea that were calculated by the Committee were very low: 0.2 ng/kg bw per day at the mean for children and 0.5–0.6 ng/kg bw per day at the mean for the general population. Concentration data in the GEMS/Food database for the Republic of Korea were available only for fish and seafood. The consumption data for children included consumption only for “not specified marine fish”, with the remaining foods being non-seafoods. However, there was no marine fish in the concentration database assigned to this “not specified” code. Therefore, all marine fish samples for the Republic of Korea were pooled to derive a concentration to use to determine an estimate of dietary exposure. As a result, 100% of the dietary exposure for children was from marine fish. There were consumption data for the general population for meat, eggs, milk and other animal foods known to contain NDL-PCBs; however, there were no concentration data for these foods. The main contributors to dietary exposure for adults were squid (40%), mackerel (24%), tuna (9%), pollack, crabs and prawns (all 7%) and clams (5%). The limited range of foods with concentration data for the Republic of Korea means that the estimated dietary exposures are likely to be underestimates for both population groups assessed, despite fish usually being a key contributor to NDL-PCB exposure. (q) Serbia

Estimated dietary exposures to the six indicator PCBs from grain foods, oil, and sugar and sugar products in Serbia were calculated (Skrbic & Durisic-Mladenovic, 2007). Mean exposures (LB–UB) were 2.9–3.4 ng/kg bw per day. Wholegrain wheat flour contributed 79% to dietary exposure from the limited range of foods. This will likely be an underestimate of actual dietary exposures, as no animal-

Supplement 1: Non-dioxin-like polychlorinated biphenyls

based foods were included in the assessment, despite per capita consumption data being used, which usually tend to produce an overestimate of exposure. (r) Slovakia

From the literature, one assessment based on foods of animal origin (Salgovicová & Pavlovicová, 2007) had estimated dietary exposures for adults of 17 ng/kg bw per day (mean) and 45 ng/kg bw per day (95th percentile). Children 4–6 years of age had a mean exposure of 30 ng/kg bw per day and a 95th percentile exposure of 87 ng/kg bw per day, children 7–11 years had a mean exposure of 20 ng/kg bw per day and a 95th percentile exposure of 56 ng/kg bw per day, and children 12–15 years had a mean exposure of 17 ng/kg bw per day and a 95th percentile exposure of 47 ng/kg bw per day. The main contributors to dietary exposure were fats (47%) and eggs and egg products (28%). (s) Spain

Based on concentration data from individual congeners and consumption data provided in Llobet et al. (2008), the Committee was able to estimate dietary exposure for the sum of the six indicator PCBs as 15 ng/kg bw per day. The main contributor to dietary exposure was fish and seafood, at 91%. EFSA (2012) estimated dietary exposures for a range of population subgroups (LB–UB) for toddlers as 18.9–25.7 ng/kg bw per day (mean) and 52.7–52.7 ng/kg bw per day (95th percentile), for children (two surveys), 13.7– 18.5 ng/kg bw per day (mean) and 38.8–42.1 ng/kg bw per day (95th percentile), for adolescents, 8.3–10.3 ng/kg bw per day (mean) and 22.5–25.1 ng/kg bw per day (95th percentile), and for adults (from two different surveys), 8.0–11.5 ng/kg bw per day (mean) and 21.0–26.7 ng/kg bw per day (95th percentile). The main contributors to dietary exposure were meat (36–47%) for toddlers, children and adolescents, fish (40–43%) for children and adolescents and fish (50–52%) for adults. (t) Sweden

The Committee estimated dietary exposures for a range of population subgroups. Mean and high consumer exposures (LB–UB) were, respectively, for children, 6.7– 13.7 ng/kg bw per day and 13.3–27.4 ng/kg bw per day, for adolescents, 3.4–7.5 ng/ kg bw per day and 6.7–15.0 ng/kg bw per day, and for adults, 5.3–7.1 ng/kg bw per day and 10.5–14.1 ng/kg bw per day. The main contributors to dietary exposure were fish-based meals (adults 72%, adolescents and children 43%), milk (adults and adolescents 39%, children 33%) and fish roe (children 49% and adults 7%). Estimated dietary exposures by EFSA (2012) (LB–UB) were, for children, 9.2–10.3 ng/kg bw per day (mean) and 20.9–22.7 ng/kg bw per day 225

Safety evaluation of certain food additives and contaminants Eightieth JECFA

(95th percentile), for adolescents, 5.6–6.4 ng/kg bw per day (mean) and 14.2– 15.1 ng/kg bw per day (95th percentile), and for adults, 5.7–6.0 ng/kg bw per day (mean) and 12.8–13.1 ng/kg bw per day (95th percentile). The main contributors to dietary exposure for children were fish (35%) and meat (35%), for adolescents, meat (37%) and fish (38%), and for adults, fish (52%). (u) Turkey

Dietary exposure to the sum of the six indicator PCBs for the Turkish population was estimated from milk and dairy products, including butter, based on concentrations in butter and extrapolation to other dairy foods on a fat per cent basis (Uçar et al., 2011). Details of the consumption data used and exact population covered were not provided in the paper. Mean dietary exposure was estimated as 0.183 ng/kg bw per day. This will, however, likely be an underestimate, as it did not include other animal-based foods known to contain NDL-PCBs. (v) United Kingdom

WHO Food Additives Series No. 71-S1, 2016 226

Estimated dietary exposures for adults in the United Kingdom were estimated using the CIFOCOss consumption data and GEMS/Food concentration data. Estimates of dietary exposure for the sum of the six indicator PCBs (LB–UB) were low for the mean, at 1.4–1.5 ng/kg bw per day, as well as for a high-percentile exposure, at 2.8–2.9 ng/kg bw per day. The highest contributor to dietary exposure was dietary supplements (e.g. fish oil supplements: 50%). Fish oils have a very high concentration of NDL-PCBs (mean around 63 000 ng/kg). The consumption category used (13.6; dietary supplements/food supplements) had a low reported level of consumption (0.01 g/kg bw per person per day). However, this food category may include consumption of supplemental beverages, rather than only dietary supplements, and may be an overestimation of consumption of dietary supplements. Other high contributors to dietary exposure were eggs (32%), cheese (8%) and fishes and aquatic animals not elsewhere specified (7%). There were no concentration data for milk in the data set, which may also mean that the exposures were underestimated. EFSA (2012) also estimated dietary exposures to the sum of the six indicator PCBs for adults. Exposures (LB–UB) were 4.1–5.3 ng/kg bw per day (mean) and 9.8–11.7 ng/kg bw per day (95th percentile). Major contributors to dietary exposure were fish (53%), meat (23%) and milk (15%). 8.3.2 National estimates of dietary exposure for individual indicator PCBs (a) Estimated dietary exposure to the six indicator PCBs individually

Dietary exposure based on national data was also estimated by the Committee for the six indicator PCBs individually using the same methodology as outlined for

Supplement 1: Non-dioxin-like polychlorinated biphenyls

the assessments in section 8.3.1. This was for use in the modelling of body burden from dietary exposure (section 9). The only methodological difference was that no unique identifiers for the samples were needed for summing concentrations for a number of congeners. Therefore, there may have been a different number of samples included in this analysis compared with the national estimates for the sum of the six congeners. It was also possible to assess more countries. For example, Japan had concentration data for five of the six indicator PCBs, so although dietary exposures could not be estimated for the sum of the six indicator PCBs, they could be estimated for the five congeners separately. A summary of these estimates is shown in Table 21. Full results are shown in Appendix 4 for each country and population group assessed. Some countries had very low estimates for the minimum exposures across all countries. This was often due to there being a small number of foods that were sampled (e.g. a specific type of fish) for which there were no consumption data for a country. Only one paper was identified in the literature that included an estimate of dietary exposure for single congeners. That was for PCB 153 for Sweden (Darnerud et al., 2006), where the mean per capita exposure from a range of foods was 139 ng/day or 1.9 ng/kg bw per day, which is within the range of exposures estimated by the Committee. (b) PCB 128

Given that there were some toxicological data available for PCB 128, estimated dietary exposures were also of relevance for the Committee’s evaluation of this congener; however, the data were limited. Out of the total GEMS/Food concentration database (n = >125 000 data points), there were only 387 data points for PCB 128 from three countries (Estonia, n = 68; Finland, n = 199; and the United Kingdom, n = 120). There were CIFOCOss consumption data only for Finland and the United Kingdom. The concentration data for these two countries for this congener were only from fish (both countries) and other seafood (United Kingdom only). This was despite both of these countries having concentration data for a broad range of foods for the six indicator PCBs. The estimates of dietary exposure for Finland at the mean were, for children, 0.000 02 ng/kg bw per day and, for adults, up to 0.000 06 ng/kg bw per day (UB); for the high percentile, estimates of dietary exposure were up to 0.000 04 ng/kg bw per day (UB) for children and up to 0.000 12 ng/kg bw per day (UB) for adults. For the United Kingdom, for adults, estimated dietary exposures at the mean were 0.000 005 ng/ kg bw per day and at the high percentile, 0.000 01 ng/kg bw per day. These results are based on very limited data and therefore will be an underestimate of actual dietary exposures to this congener. 227

Safety evaluation of certain food additives and contaminants Eightieth JECFA

Table 21 Summary of estimated dietary exposures for individual indicator PCBs showing minimum and maximum for adults and children from all countries assessed Dietary exposure Population group measure Children (infant to Minimum adolescent) Maximum Adults Minimum Maximum Children (infant to Minimum adolescent) Maximum Adults Minimum Maximum Children (infant to Minimum adolescent) Maximum Adults Minimum Maximum Children (infant to Minimum adolescent) Maximum Adults Minimum Maximum Children (infant to Minimum adolescent) Maximum Adults Minimum Maximum Children (infant to Minimum adolescent) Maximum Adults Minimum Maximum Dietary exposure (ng/kg bw per day) Mean LB 0.002 1.2 0.005 0.3 0.000 7 2.4 0.01 1.2 0.000 3 3.0 0.01 0.7 0.000 6 4.3 0.03 1.7 0.001 6.5 0.03 3.0 0.001 2.0 0.01 0.7 UB 0.03 17.5 0.02 4.7 0.03 17.2 0.04 4.6 0.03 17.2 0.05 4.8 0.04 17.5 0.1 4.7 0.04 15.7 0.1 5.0 0.03 17.3 0.07 4.7 High percentile LB UB 0.000 4 0.05 2.5 35.1 0.01 0.05 2.0 9.5 0.001 0.05 4.8 34.3 0.02 0.08 2.3 9.2 0.001 0.07 6.0 34.4 0.03 0.1 2.0 9.5 0.001 0.08 8.6 35.0 0.06 0.3 3.5 9.3 0.002 0.08 13.1 31.5 0.06 0.3 6.0 9.9 0.002 0.05 4.1 34.5 0.03 0.1 1.5 9.4

PCB 28

52

101

138

153

180

LB: lower bound; UB: upper bound

WHO Food Additives Series No. 71-S1, 2016 228

Given that the estimated dietary exposures were limited and not likely to be a realistic estimate of dietary exposure, an alternative approach was used whereby the proportion that the concentration of PCB 128 represents of the six indicator PCBs considered individually was determined on the basis of data from GEMS/Food (the average was 16%). This was then applied to the average of the upper end of the range of all indicator PCB congeners for both mean and highpercentile exposures for adults from any of the individual indicator congeners, as a way of representing approximate PCB 128 exposure before estimating the MOE.

Supplement 1: Non-dioxin-like polychlorinated biphenyls

8.3.3 International estimates of dietary exposure

International dietary exposures to the sum of the six indicator PCBs were estimated using consumption data from the 17 GEMS/Food cluster diets and occurrence data for NDL-PCBs from the GEMS/Food concentration database, based on the methodology described above. The consumption amount of each food group by cluster can be found in Appendix 5, along with the LB and UB concentrations for each food group, estimated dietary exposure from each food (for both the LB and UB scenarios) and the total mean dietary exposures from all foods. The contribution of each food group to the total exposure is shown in Table 22. The contributions were based on the LB scenario only so as to not overestimate the likely contribution of food groups for which there were not detectable concentrations of NDLPCBs. Concentrations varied widely between and within regions for the range of foods for which data were available. The highest concentrations were for fish up to LB mean concentrations of 82 000 ng/kg for clusters using data for all countries and around 100 000 ng/kg (cluster 7) and 200 000 ng/kg (cluster 11) for clusters with data from their own countries. Other food groups with high concentrations included marine animal fat, with a LB mean of around 20 000 ng/ kg, milk fats, around 11 000 ng/kg, animal fats, up to around 60 000 ng/kg, and “foods out of classifying” (e.g. mixed dishes or foods that did not fit specifically within a specific food group), at around 13 000 ng/kg. Relatively lower levels (usually under 1000 ng/kg) were mainly found in fruits and vegetables, where data were available. Across all of the clusters, there were no data for the food groups egg products and processed eggs, hops, roots and tubers raw or boiled, and root vegetables not elsewhere specified (nes) (there were data for roots and tubers nes). The majority of the food groups in the cluster diets are at the raw commodity level. There is a small number of minimally processed food categories in the cluster diets where the concentrations for some processed foods were captured. However, concentrations for many of the mixed dishes or highly processed foods were captured under Level 2 code 190 Out of Classifying. Estimated mean dietary exposures to the sum of the six indicator PCBs are summarized in Table 23. The range between the clusters for the LB scenario was from 1 ng/kg bw per day (cluster 9) to 60 ng/kg bw per day (cluster 11). For UB mean exposures, the range between clusters was from 2 ng/kg bw per day (cluster 9) to 83 ng/kg bw per day (cluster 11). Cluster 9 has the lowest estimate of dietary exposure (LB–UB mean: 1–2 ng/kg bw per day). The cluster is made up of Asian countries, and the concentration data came from China (including Hong Kong Special Administrative Region). 229

230 2 0.2 0.0 0.0 0.3 1.1 0.0 2.0 1.2 1.4 0.4 4.5 1.2 21.8 0.9 1.9 0.4 0.0 5.7 0.9 2.4 3.2 6.7 0.3 0.4 0.1 32.1 0.1 1.2 0.5 1.9 0.6 0.1 2.2 0.7 8.5 5.3 0.1 – 5.6 22.9 – 37.7 0.0 0.1 4.1 0.2 1.3 0.0 0.5 0.8 0.1 4.5 0.4 4.1 0.9 0.8 1.8 0.2 5.4 0.1 11.2 0.2 1.2 – – 5.1 0.3 35.2 4.2 2.0 4.0 – 7.2 1.3 0.6 1.0 3.4 16.2 – – 11.6 13.9 7.1 8.5 0.2 – – – 1.4 0.1 6.7 0.2 – 0.4 – 9.5 0.4 6.1 1.2 0.3 0.3 4.3 12.0 2.9 1.2 0.4 0.0 8.4 1.6 0.6 1.4 0.8 2.8 1.6 – – 6.8 9.0 1.9 0.0 11.8 – – 12.1 – – 1.0 1.0 0.3 0.4 1.5 5.9 3.1 0.7 0.6 0.8 0.4 1.5 3.7 0.9 0.3 0.2 6.5 0.3 9.1 0.2 1.5 0.1 3.7 1.8 0.2 0.0 0.5 0.0 0.2 0.0 – – – – 1.3 0.0 3.7 1.2 – – – – 0.3 0.0 0.2 0.0 0.1 0.0 0.6 0.4 0.1 0.0 0.5 0.0 0.3 0.0 0.1 0.0 7.9 0.2 0.0 0.0 1.2 0.5 2.3 3.3 0.7 0.4 7.9 0.4 1.9 0.3 0.2 9.8 0.6 0.0 2.2 1.1 0.1 0.2 6.0 0.3 7.6 0.0 1.5 0.2 8.0 0.8 3 0.0 0.1 0.2 0.0 4 0.1 0.1 0.1 0.1 5 0.1 0.2 0.1 0.2 6 – – – – % contribution in following clusters: 7 8 9 10 11 – 1.1 – – – – – 1.4 – – – – – – – 2.2 0.2 3.4 – – 12 0.0 0.1 0.0 0.1 13 0.0 0.4 0.4 0.1 14 0.0 0.0 0.0 0.0 15 0.1 0.0 0.1 0.2 16 0.0 0.9 1.5 0.0 17 0.0 0.0 0.0 – 0.0 0.0 0.2 0.2 0.2 0.1 0.7 0.7 0.3 0.0 0.4 0.2 – 0.4 Safety evaluation of certain food additives and contaminants Eightieth JECFA

WHO Food Additives Series No. 71-S1, 2016

Table 22 Per cent contribution (from the LB scenario) of food groups (only where >1% for any one cluster or more) to the dietary exposure to the sum of the six indicator PCBs from GEMS/Food cluster diets

Food code 12 20 21 40

1 0.1 0.1 0.1 0.1

42 43

0.4 0.0

60 70

2.8 1.7

80 81 84 85 90 91 100

1.8 0.3 6.0 0.1 15.7 0.3 1.1

101

0.3

104

0.5

110

Food name Pome fruits, fresh Pulses (dry, prepared and composites) Oilseed (incl. flour) Brassica (cole or cabbage) vegetables, head cabbages, flowerhead cabbages Fruiting vegetables, cucurbits Fruiting vegetables (other than cucurbits) and mushrooms Cereal grains & flours Sugars, honey, candies (excl. chocolate), sweeteners and molasses Milk fats Mammalian fats (no milk fat) Plant origin fat Animal or vegetable fat, nes Milks (no other ingredients) Dairy products (incl. whey, excl. milk fats) Mammalian (not marine) meat, unprocessed (incl. home-cooked) Poultry (incl. pigeon) meat, unprocessed (incl. home-cooked) Meat and offal, nes (incl. reptiles and amphibians), unprocessed (incl. home-cooked) Eggs

2.4

Food code 120 121 122 123 2 1.5 3.5 0.1 0.0 44.6 5.1 1.5 4.7 0.7 3.5 3.7 1.5 0.4 – – 7.0 21.2 – 0.2 – – – 0.0 – – 3.1 10.8 3.9 0.8 1.4 0.2 0.0 20.9 62.0 67.3 56.1 12.0 27.4 21.0 6.3 3.2 68.4 60.0 57.6 83.9 35.2 26.7 2.3 1.1 89.5 2.5 2.0 3 6.2 9.4 0.0 0.0 4 1.7 4.3 0.1 0.0 5 6.7 5.0 0.1 0.1 6 4.7 1.0 – 1.0 12 1.4 5.8 0.3 0.3 13 5.0 2.1 0.0 0.0 14 0.9 2.9 0.0 0.0 15 12.5 0.6 – 0.5 16 37.5 0.7 0.0 0.0 17 0.0 2.1 0.0 0.1

1 1.9 2.1 0.0 0.0

% contribution in following clusters: 7 8 9 10 11 3.6 0.3 32.0 4.2 0.4 0.8 0.1 1.4 0.2 0.1 0.1 0.0 2.7 0.4 0.0 0.2 0.0 20.6 2.2 0.1

124

60.3

127 190

Food name Freshwater fish, unprocessed (incl. home-cooked) Marine fish, unprocessed (incl. home-cooked) Crustaceans, unprocessed (incl. home-cooked) Molluscs and cephalopods, unprocessed (incl. home-cooked) Other fishes and aquatic animals, unprocessed (incl. home-cooked) Processed aquatic animals Out of classifying

0.7 0.5

excl.: excluding; incl.: including; LB: lower bound; nes: not elsewhere specified; –: zero contribution, due to either no consumption or no concentration; 0.0: very small contribution

Supplement 1: Non-dioxin-like polychlorinated biphenyls

231

Safety evaluation of certain food additives and contaminants Eightieth JECFA

Table 23 Summary of estimated dietary exposures to the sum of the six indicator PCBs for the GEMS/ Food cluster diets Cluster number 1 2 3 4 5 6 7 8c 9 10 11 12 13 14 15 16 17 a

Mean exposure (ng/kg bw per day)b LB UB 9 20 11 23 9 13 17 28 8 18 3 3 17 19 37 39 1 2 5 46 60 83 16 31 6 11 39 50 5 11 4 9 52 61

Source of concentration data All countries All countries All countries All countries All countries Countries from own cluster Countries from own cluster Countries from own cluster Countries from own cluster Countries from own cluster Countries from own cluster All countries All countries All countries Countries from own cluster All countries All countries

LB: lower bound; UB: upper bound a Countries in each cluster are given in Appendix 6. b Body weight = 60 kg. c Excluding outlier for plant fats and oils.

The concentrations for the sum of the six indicator PCBs were low across all foods, but they were detectable because of the extremely low LODs used. The low LODs used also meant that a low number was substituted for the UB scenario for not detected results, so UB estimates of exposure would not be overestimated. The samples included a wide range of foods, including fish and seafood, meat and offal, milk, eggs, fruit and vegetables. Cluster 6 also had a low estimate of dietary exposure, at 3 ng/kg bw per day at both the LB and UB. The concentration data for this cluster were only from Greece. There was a wide representation of animal foods in the samples, including meat and meat products, milk and dairy products, eggs, fish and seafood, as well as plant-based fats. Therefore, this estimate should not be underestimated to a large degree. Apart from higher concentrations in “animal and plant fats nes”, all food groups had mean concentrations under 5200 ng/kg. Cluster 11 had the highest LB estimate of dietary exposure of any cluster (60 ng/kg bw per day). This cluster is made up of European countries. The

232

WHO Food Additives Series No. 71-S1, 2016

Supplement 1: Non-dioxin-like polychlorinated biphenyls

concentration data were from Belgium and the Netherlands. Neither of these countries contributed concentration data with extremely high LODs; however, Belgium did have a higher proportion of data that were not detected. This cluster had around 70% of its dietary exposure from the “other fish” food group, for which it has a consumption level around midway between the highest and lowest clusters for this food. The mean concentration for this food group was the highest for any cluster, at around 200 000 ng/kg. The food samples that contributed to this were some very high concentrations of NDL-PCBs in eels from the Netherlands. EFSA (2012) also noted the high concentration of NDL-PCBs in eel. This cluster also had 12% of the exposure coming from milk. This food was the highest contributor in the national estimates of exposure for Belgium owing to a high concentration in milk relative to the concentrations in milk in other countries. Table 22 shows the details of the food group contributions to estimated dietary exposure to the sum of the six indicator PCBs based on the LB scenario. The main food groups contributing to estimated dietary exposure for the majority of clusters were the fish and seafood groups. This was due to either their high NDL-PCB concentrations or high consumption. Contributions from “other fishes and aquatic animals” ranged from 3% in cluster 10, 6% in cluster 9, 12% in cluster 6, 21% in cluster 8, 27% in clusters 7 and 16, 35% in cluster 15 and 45% in cluster 2 to around 60% or above for the remaining clusters, with the highest contribution at 90% for cluster 17. The clusters with the highest consumption of fish had the highest contribution. Cluster 17 had the highest consumption by far compared with the other clusters. Freshwater fish were also a contributor for a number of clusters: 6% in cluster 3, 7% in cluster 5, 32% in cluster 9, 13% in cluster 15 and 38% in cluster 16. Marine fish contributed to a number of clusters: 9% in cluster 3, 6% in cluster 12 and 5% in cluster 5. Processed aquatic animals contributed 7% for cluster 7 and 5% for clusters 2 and 3. Molluscs were a high contributor for cluster 9 (21%). There were also contributions to dietary exposure (5% or greater) for a number of clusters for milk, eggs, and plant fats and oils, and a small number of clusters had marked contributions from meat or poultry, sugars/chocolates and cereal grains. Milks contributed 35% for cluster 6, 22% for cluster 2, between 10% and 20% for clusters 1, 5, 8 and 11, and between 5% and 10% for clusters 10, 13, 15 and 16. Eggs contributed 38% for cluster 10, 32% for cluster 6, 10% for cluster 15, 8% for cluster 11 and 6% for clusters 2 and 9. Meat and offal not further specified contributed 8% for clusters 8, 14 and 16 and 7% for cluster 3. Poultry contributed for only one cluster (23% for cluster 10). Mammalian meat contributed 9% for cluster 8, 7% for cluster 9 and 6% for cluster 10. For fats, plant fats contributed 5–7% for clusters 1, 2, 4, 5, 6, 7, 13 and 16; milk fats contributed 16% for cluster 8; and animal or plant fats not further specified contributed 6% for cluster 12 and 12% for cluster 8. Sugars/honey/candies (excluding chocolate) 233

Safety evaluation of certain food additives and contaminants Eightieth JECFA

contributed 9% for cluster 7 and 12% for cluster 10. Cereal grains contributed 7% for cluster 7 and 12% for cluster 9. In most cases, the high contribution was for the clusters with the higher consumption amounts of the food group across the clusters; however, this was not always the case. The dietary exposure estimates are likely to be an overestimate, as the food balance sheet data on which they are based are amounts of food “available for consumption”, and waste is not taken into account. The results do provide useful comparative information about the areas of the world that may have potentially higher dietary exposures. For the clusters that employed the same concentration data (clusters 1, 2, 3, 4, 5, 12, 13, 14, 16 and 17; pooled data from all countries), there were higher exposures for cluster 14, which is made up primarily of Pacific Island countries and has a high consumption of marine fish and other fish, both categories that had high concentrations. However, when making comparisons between clusters, it must be considered that for many of the clusters, the whole data set of concentration data was combined and used for the calculations. Therefore, for these clusters, differences in dietary exposure will solely reflect differences in food consumption patterns. Therefore, a limitation of this methodology is that comparison across clusters is not on the same basis. 8.3.4 Dietary exposures for infants

WHO Food Additives Series No. 71-S1, 2016 234

PCBs are transferred from mother to child through breastfeeding. Concentrations of PCBs in breast milk can depend on a number of factors, including the maternal place of residence, diet and the number of previous children. Occurrence data also show detectable concentrations of NDL-PCBs in infant formula. Exposure of an infant to PCBs is an important contributor to the total body burden. Therefore, infants were included in the evaluation of dietary exposure to the six indicator PCBs where possible. Estimated dietary exposures from both breast milk and infant formula were relevant to the evaluation. There were approximately 900 data points for NDL-PCBs in the GEMS/ Food concentration data set for foods for infants (cereal-, vegetable-, dairy- or meat-based foods, etc.). However, national consumption data were not available for infants other than for Italy, which had no concentration data for infant foods; therefore, no estimates of dietary exposure could be calculated for infants consuming a diet that includes milk and solid foods. (a) Estimated dietary exposure for breastfed infants

Breast milk has many beneficial aspects, including nutritional and immunological benefits, despite containing environmental contaminants, such as NDL-PCBs. However, based on present knowledge, the benefits of breastfeeding (WHO, 2015)

Supplement 1: Non-dioxin-like polychlorinated biphenyls

are considered to outweigh the possible disadvantages that may be associated with the presence of NDL-PCBs in breast milk. Concentrations of PCBs in human milk are highly variable both between and within countries. European countries have reported higher concentrations of NDL-PCBs in breast milk than in other parts of the world. Liem, Fürst & Rappe (2000) included a summary of concentrations of the sum of the six indicator PCBs in breast milk from a number of countries across the world. These ranged between 19 and 1068 ng/g in the fat for LB mean concentrations. Highest concentrations were found in the Czech Republic (1068 ng/g fat) and Slovakia (1015 ng/g fat), with all other countries having concentrations of 682 ng/g fat or less. The biomonitoring section of this report (section 2.3.1) noted the WHO worldwide measurement campaigns to determine the exposure of infants to the 12 DL-PCBs and the six indicator PCBs over time, where five rounds of monitoring have been conducted to date (UNEP/WHO, 2013). For the sum of the six indicator PCBs, the median concentration in breast milk over the 12 years of the survey was between 10.8 and 30.7 ng/g lipid, with maximum concentrations between 37.1 and 65.8 ng/g lipid. Data on human milk were submitted to the GEMS/Food concentration database by New Zealand. These were aggregate data for only two of the indicator PCBs, where the mean concentration for PCB 28 was 90 ng/kg (fresh weight) (from 50 samples) and the UB mean concentration for PCB 101 was 10 ng/kg (fresh weight) (from 15 samples). Therefore, it was not possible to use these data in this evaluation for calculating estimates of dietary exposure, as the focus was on the sum of the six indicator PCBs. Estimating dietary exposures to any chemical in breast milk is difficult to do because of variation in the duration of breastfeeding, the number of children fed by the mother, the duration of each single feed, the volume consumed, the variable fat content in the milk, the age of the child, etc. For breastfed infants, the most recent WHO study of PCB exposure reported a mean dietary exposure of about 1600 ng/kg bw per day (range of 230– 7300 ng/kg bw per day) for the sum of the six indicator PCBs (IARC, 2015). Based on data from 58 human milk pools from 18 European countries, it has been estimated that the dietary exposure to the sum of the six indicator PCBs for exclusively breastfed infants is a median exposure of 984 ng/kg bw per day or a mean exposure of 1177 ng/kg bw per day, assuming a milk intake of 800 mL/ day with a fat content of 3.5% and a body weight of 5 kg. A maximum exposure of 5651 ng/kg bw per day was estimated (EFSA, 2005). These values are within the range reported by IARC (2015). This is significantly higher than estimates of dietary exposure for the rest of the population on a body weight basis. In a study from Slovakia, a median dietary exposure to the sum of the six indicator PCBs for fully breastfed infants of 1608 ng/kg bw per day was reported 235

Safety evaluation of certain food additives and contaminants Eightieth JECFA

(Drobna et al., 2011). Another study from Sweden estimated that the mean exposures to the sum of five NDL-PCBs (PCBs 28, 138, 153, 170 and 180) in 1-, 3- and 6-month-old infants were 418, 294 and 165 ng/kg bw per day, respectively. This showed a clear trend in the reduction of NDL-PCB exposure over the breastfeeding period (Bergkvist et al., 2010). (b) Estimated dietary exposure for fully formula-fed infants

WHO Food Additives Series No. 71-S1, 2016 236

Some data were submitted to the GEMS/Food concentration data set on NDLPCB concentrations in infant formula. A summary of these data is shown in Table 24. Concentrations were provided for infant formula (0–6 months) in both liquid and powdered forms and follow-on formula (usually 6 months and above) in liquid form. No estimates of dietary exposure for formula-fed infants were derived using these data for a number of reasons, including that there were only a small number of countries with concentration data, that the majority of those had only a small number of samples and that some surveys had very high LODs or included only some of the six indicator PCBs. There have been some estimates of dietary exposure of infants to NDLPCBs from infant formula reported in the literature. One study from Spain (Lorán et al., 2009) estimated dietary exposure for 3-month-olds to the sum of the six indicator PCBs (estimated from Σ7*0.85) for boys of 6.9–9.2 ng/kg bw per day (LB–middle bound [MB]), and the equivalent for girls of 7.5–10.0 ng/kg bw per day, based on consumption of 841 mL formula and concentrations of 51 ng/ kg (LB) and 68 ng/kg (MB). A study from a PCB-contaminated area in Slovakia (Drobna et al., 2011) provided a comparison between dietary exposure to total summed PCBs over 10 months for exclusively breastfed infants compared with exclusively formulafed infants. The results provided equated to a dietary exposure for the breastfed infants of 1286 ng/kg bw per day, compared with 2.8 ng/kg bw per day for formula-fed infants in the same area. From the Hong Kong Special Administrative Region of China (Xu & Cai, 2015), concentrations in infant formula samples were reported as 0.065–1.15 ng/g fat. For exclusively formula-fed infants, the estimated dietary exposures based on the analysed infant formula concentrations ranged from 2.42 ng/kg bw per day during the first month of life to 1.70 ng/kg bw per day in the sixth month of life. These values were reported as being 1 order of magnitude lower than exposures for breastfed infants. 8.3.5 Dietary exposures for specific population subgroups

The occurrence data for NDL-PCBs show high concentrations in animalbased foods, particularly fish and seafood. Therefore, populations with high

Supplement 1: Non-dioxin-like polychlorinated biphenyls

Table 24 Summary of infant formula data submitted to the GEMS/Food concentration database Country Canada China PCBs Food 20 congeners; however, Infant formula liquid only 5 of the 6 indicators 6 indicators Infant formula powder Number of data points 80 (no sample unique identifiers provided) 822 (137 samples) Concentration data LOQs up to 0.13 ng/kg. 23% ND. LOQs 0.9 ng/kg or less. Range of detected values 0.17–153 ng/kg for individual congeners. 75% ND. LOQs up to 33 ng/kg. Range of detected values 0.10–0.14 ng/kg based on individual congeners. 0% ND. Range 0.002–0.05 ng/kg based on individual congeners. 0% ND. Range 0.000 6–0.011 ng/kg. 0% ND. Range of values for individual congeners 0.07–0.3 ng/kg. All ND. High LOQs for all of the samples, ranging from 35 to 75 µg/kg, depending on the congener. 40% ND. LOQ 0.14 ng/kg or less. Range of detected values 0.001–0.03 ng/kg.

Czech Republic France France Greece Singapore

6 indicators

Infant formula powder

12 (2 samples)

6 indicators 6 indicators 6 indicators 6 indicators

Infant formula powder

24 (4 samples)

Follow-on formula liquid 6 (1 sample) Follow-on formula liquid 6 (1 sample) Infant formula powder 35 (5 samples/different products) 30 (5 samples)

Sweden

6 indicators

Infant formula powder

LOQ: limit of quantification; ND: not detected

consumption of fish and seafood were also assessed, where information was available. EFSA (2005) provided a good summary of exposures for population subgroups with higher exposures than for the general population. These included Baltic fishermen who consumed fish more frequently and in greater quantities compared with the general population in the same region. Fish from the Baltic Sea also contain higher PCB concentrations compared with other regions of the world. Dietary exposure to the sum of the six indicator PCBs by fishermen from the Swedish coasts were approximately 5- to 10-fold higher than the corresponding exposure from fish estimated for the average male adult population in Sweden (e.g. east coast fishermen had an exposure up to 41.5 ng/kg bw per day, compared with 5.3 ng/kg bw per day for males in the general population). Residents of the Faroe Islands have a high consumption of seafood and a higher dietary exposure to PCBs. A major contributor to the exposure is pilot whale blubber, which contains an average PCB concentration of 20 000 µg/kg. Residents living in areas where PCBs had been previously produced, where waste may have contaminated locally grown and consumed agricultural crops and livestock, can have higher dietary exposures. Concentrations of three of the indicator PCBs (PCBs 138, 153 and 180) in blood from adults and children 237

Safety evaluation of certain food additives and contaminants Eightieth JECFA

in the Michalovce district of Slovakia were 2–2.5 times higher than those in residents from a control area. A study in an area that recycles transformers showed that mean exposures to 37 PCBs in lactating women were 92.8 ng/kg bw per day, compared with a reference site with exposures of 7.3 ng/kg bw per day (Xing, Wu & Wong, 2010). Residents consuming fish and produce from the locally contaminated area had higher dietary exposures as a result. The information on higher dietary exposures in these population subgroups is consistent with the higher blood/serum concentrations in these subgroups, as shown in the biomonitoring data evaluated by the Committee (see section 2.3.1). 8.4 Other routes of exposure

238

WHO Food Additives Series No. 71-S1, 2016

The major source of exposure to PCBs, including NDL-PCBs, is via the diet, with more than 90% coming from food (IPCS, 1993; EFSA, 2005; Loutfy et al., 2008). Dermal and inhalation exposure may also occur, usually making a small contribution to exposure. Occupational exposure can be an important source. IARC (2015) undertook a large review of environmental sources of contamination and routes of exposure to PCBs, including inhalation and dermal routes. Dermal absorption of PCBs may occur via the occupational setting or other means, such as contact with contaminated sediments or soil or topical applications to the skin. Estimates of the concentrations in air, dust and soil were provided by IARC (2015) and EFSA (2005). Indoor air can contain PCBs where they are released into the environment as a result of being incorporated into building materials, such as sealants and paints. Residents living in contaminated buildings have higher PCB concentrations in blood compared with residents living in control buildings, as do pupils and teachers in contaminated schools compared with controls (EFSA, 2005). In one study where the sum of the six indicator PCBs was investigated, median concentrations in plasma were 3.4 times higher in residents living in buildings containing PCBs than in residents not exposed from their dwellings (Meyer et al., 2013). The differences were usually observed for the lower chlorinated congeners (PCBs 28, 52 and 101). Inhalation may account for 4–63% (median 15%) of overall exposure in humans (IARC, 2015). IARC (2015) reviewed a large number of studies of concentrations of total and individual PCB congeners, covering both DL- and NDL-PCBs, in blood. These clearly showed higher concentrations in those who are occupationally exposed compared with controls. Whereas occupational exposure many years ago would have occurred during the manufacturing process and capacitor and

Supplement 1: Non-dioxin-like polychlorinated biphenyls

transformer repair, more recent exposures occur from waste incineration, fires and waste recycling (IARC, 2015). EFSA (2005) calculated the relative contribution to exposure from air and soil compared with food. Exposure to PCBs (not specifically the NDLPCBs in this case) via inhalation was estimated to be 0.05–0.5 ng/kg bw per day for children and 0.03–0.3 ng/kg bw per day for adults. Exposures from soil or dust would be small, at around 0.06–0.6 ng/kg bw per day for children. Dermal absorption was estimated to be around 5 pg/kg bw per day for children and 0.76 pg/kg bw per day for adults. From these values, ESFA (2005) concluded that exposure from these sources was around 3–4 orders of magnitude lower than exposure from food. Ingestion of contaminated soil or dust can be a minor source of exposure for children. The information on higher exposures through the skin and inhalation in population subgroups living or working in contaminated sites or buildings is consistent with the higher blood/serum concentrations in these subgroups, shown in the biomonitoring section of this monograph (section 2.3.1). 8.5 Temporal trends in dietary exposure

A number of papers and reviews have indicated that dietary exposure to PCBs is decreasing over time. Concentrations in foods have been demonstrated to be dropping over time. In Belgium, Cimenci et al. (2013) showed decreasing concentrations in milk and dairy products and eggs and egg products. The Chinese Total Diet Study (Y. Wu, personal communication, 2015) demonstrated an apparent approximate halving of the sum of the concentrations of the six indicator PCBs in milk, from a mean of 11.7 µg/kg in 2007 (12 provinces, n = 1237 milk samples, concentration range 2.4–28.8 ng/g lipid weight) to a mean of 6.6 µg/kg in 2011 (16 provinces, n = 1760 milk samples, concentration range 2.3–19.0 ng/g lipid weight). EFSA (2012) conducted a time trend analysis where three groups of foods (eggs, fish, and milk and dairy products) showed a decreasing trend in median concentrations for the sum of the six indicator PCBs over time. This was statistically significant for milk and dairy products, with a 64% reduction in 15 years. Dietary exposures have been demonstrated to be dropping in a number of countries. In Italy, in 2008, dietary exposures to NDL-PCBs were estimated to be 30% lower than found in a corresponding study in the mid-1990s (Fattore et al., 2008). In the Netherlands, exposures to the sum of the six indicator PCBs have been decreasing over time, although this trend has been slower in more recent years, with a flattening out over the preceding decade. Exposures declined from around 80 ng/kg bw per day between 1975 and 1980 to 39 ng/kg bw per day in 1984–1985 to 10 ng/kg bw per day in 1994 (Bakker et al., 2003; Baars et 239

Safety evaluation of certain food additives and contaminants Eightieth JECFA

WHO Food Additives Series No. 71-S1, 2016 240

al., 2004). Mean exposures to the sum of the six indicator PCBs for adults in Slovakia decreased from 26 ng/kg bw per day in 1994 to 17 ng/kg bw per day in 2004 (Salgovicová & Pavlovicová, 2007). The National Monitoring Program in the Czech Republic showed a decreasing trend in exposures from 1994 to 2003 (EFSA, 2005). Studies in Sweden have also indicated a decrease in exposure over time; 28 PCBs, including DL- and NDL-PCBs, had mean MB concentrations of 8.3 ng/kg bw per day in 1999 and 4.9 ng/kg bw per day in 2011. This was proposed to be due to the drop in concentrations in foods over time and more specific analytical methods (Törnkvist et al., 2011). In France, dietary exposures to total PCBs decreased 3-fold between 2005 and 2012 (Sirot et al., 2012). EFSA (2012) also showed a statistically significant drop in dietary exposures over time (2002–2010) across all population subgroups and all countries. Liem, Fürst & Rappe (2000) noted that countries that started to implement measures to reduce dioxin emissions in the late 1980s, such as the Netherlands, the United Kingdom and Germany, showed a decrease in PCB concentrations in food and therefore a lower dietary exposure, by almost a factor of 2, within the preceding 7 years. Concentrations in human milk have also been dropping over time (IARC, 2015). Therefore, the dietary exposures for breastfed infants will also decrease, as has been reported by a number of European countries. In the Federal Republic of Germany before reunification, for example, PCB exposures decreased by 80% between 1984 and 2003 (EFSA, 2005). In Sweden, a significant decrease was estimated for mean exposures of breastfed infants between 2000 and 2006, by a factor of 1.5 (Bergkvist et al., 2010). Reductions in concentrations in foods and therefore estimated dietary exposures over time could be due to a number of reasons, including reduced concentrations of PCBs over time, better analytical methods that report lower concentrations in foods (this will reduce the level of uncertainty and decrease UB estimates of exposure) and risk management measures, such as the implementation of maximum levels in Europe. The evidence supporting a decrease in dietary exposure to PCBs over time is consistent with information summarized in the biomonitoring section of this monograph (section 2.3.1).

estimates

8.6 Limitations and uncertainties associated with the exposure

A number of limitations and uncertainties in the dietary exposure estimates exist for NDL-PCBs. These include the following:

Supplement 1: Non-dioxin-like polychlorinated biphenyls

■■ Some LOQs in the GEMS/Food database were very high (up to 83 µg/kg). A reliable sensitive method of analysis for NDL-PCBs was determined to be one with an LOQ of 3.5 µg/kg or less. Leaving all data points in the data set can lead to a higher UB estimate of the dietary exposure and a wider LB–UB range. Eliminating results with LOQs over the specified cut-off of 3.5 µg/kg will affect the dietary exposures estimated for countries that have high LOQs, but may introduce some bias into the data set by eliminating parts of data sets. This variation in concentration data was one of the largest areas of uncertainty for the estimates of dietary exposure conducted by the Committee. ■■ Not all data points in the GEMS/Food database had unique identifiers for each sample. This made it difficult to sum the six indicator PCB congeners for each individual sample before deriving mean concentrations per food group, which is the ideal approach to summarizing the data for the congeners of interest. Therefore, in some instances, the mean concentration across the food group was obtained for each congener separately, which may or may not have been based on the same number of samples, and the sum of the mean concentrations was calculated from there. ■■ Concentration data submitted to the GEMS/Food database in some cases were only for a limited range of foods. Some of the estimates of dietary exposure in the literature were also based on only a limited range of foods. These estimates would therefore tend to be an underestimate of dietary exposure. ■■ Not all countries that provided concentration data to the GEMS/ Food contaminants database provided country-specific consumption data for inclusion in the FAO/WHO CIFOCOss database. Therefore, specific country exposure assessments could not be undertaken for all countries providing concentration data. ■■ Concentration data for NDL-PCBs in the GEMS/Food contaminants database were available only from countries from seven out of the 17 clusters, meaning that concentration data for all submitting countries were used for dietary exposure estimation for clusters without specific concentration data. This limits the comparability of results across clusters and the robustness of the conclusions drawn from the assessment. For clusters using global concentration data, differences in dietary exposure estimates for NDL-PCBs will reflect differences in food consumption patterns only. ■■ Concentration data submitted to the GEMS/Food database were primarily from European countries (90%). Therefore, the majority of 241

Safety evaluation of certain food additives and contaminants Eightieth JECFA

WHO Food Additives Series No. 71-S1, 2016

the dietary exposures estimated by the Committee were for European countries. From the literature, the estimates were also predominantly from Europe, with a small number from Asian countries. Information from a broader range of countries around the world is needed in order to provide a more globally representative assessment of dietary exposure. ■■ Food groups in the CIFOCOss consumption data set and the GEMS/ Food concentration data may have been given different classification codes. Whereas codes for the concentration data were reassigned if the original code was not appropriate, there was often still some mismatch between the two data sets. This made it difficult to match the two sets of data for the exposure calculations, and some assumptions had to be made about what concentration data were matched with what consumption data. For example, the concentration data set may have had data for “liquid milk” with no indication of origin; however, the only consumption value in the data set was “cow’s milk”, and therefore these may have been matched together in order to include an estimate of exposure and contribution from this food. This often led to lower national exposures being estimated by the Committee compared with other sources, and having better metadata for both the concentration and consumption data sets would enable better matching between and use of the two input data sets. ■■ Concentrations of NDL-PCBs have been shown to be decreasing over time. Therefore, older concentration data from the GEMS/Food contaminants database may be less representative for contemporary dietary exposure estimates, particularly estimates for children who are unlikely to have consumed foods containing concentrations of NDLPCBs represented in the older concentration data. Inclusion of these older concentration data will lead to a more conservative (higher) estimate of dietary exposures for children. ■■ As outlined in section 5 on effects of processing, there are some decreases in NDL-PCB concentrations due to cooking of food, depending on the cooking method, but particularly where the cooking method results in loss of fat. Therefore, estimated dietary exposures, where not based on foods prepared ready for consumption, may be slight overestimates. However, it is not expected that this would change the general range of estimated exposures to NDL-PCBs in this evaluation or change the conclusions based on the estimates provided. Despite these limitations, the estimates of dietary exposure were generally within the same ranges across all estimates of dietary exposure assessed, particularly for LB means.

242

9. Modelling of body burden from dietary exposure PCBs, including NDL-PCBs, are known to be slowly eliminated from the body, as shown by the very long half-lives of some congeners (Ritter et al., 2011). As a consequence, recurrent exposure results in bioaccumulation – that is, increasing body burden over time. For this reason, the accumulated amount in the body (body burden), rather than the daily exposure, is typically considered a more relevant exposure end-point to be used in risk assessment for these chemicals (Van Leeuwen & Younes, 2000; USEPA, 2010; Annex 1, references 154 and 155). The objective of this section is to estimate the body burden corresponding to the dietary exposure as assessed in section 8. Body burden estimation becomes particularly relevant if tolerable daily intakes (TDIs) are set, but not if an MOE approach is taken. 9.1 Dynamic modelling of exposure from food

In order to estimate the body burden corresponding to the dietary exposure assessed in section 8, the Kinetic Dietary Exposure Model was used (Verger, Tressou & Clémençon, 2007). This model describes the dynamic evolution of internal exposure over time using a dynamic exposure process (mathematically detailed in Bertail, Clémençon & Tressou, 2010), which is determined by the accumulation due to recurrent dietary exposures and by the elimination process that occurs over time. Such a model was previously used to estimate the body burden of PCDDs, PCDFs and DL-PCBs in France and validated with biomonitoring data for these chemicals (Béchaux et al., 2014). Body burdens for each congener were estimated in child and adult populations in the countries from which dietary exposure data were available (China, Czech Republic, Finland, France, Germany, Ireland, Italy, Japan, the Netherlands, the Republic of Korea, Sweden and the United Kingdom). No modelled body burdens for PCB 128 could be developed owing to the lack of half-life information. Between exposures, the change in the body burden is described by a simple one-compartment, first-order toxicokinetic model that is commonly used for chemicals with long half-lives, such as PCBs (Lorber, 2008; Amzal et al., 2009; Fromme et al., 2009a). Consequently, the kinetics of elimination is described with two parameters – namely, half-life and absorption rate. A more complex kinetic model is available for PCB 153 (Redding, 2010). However, in order to retain consistency across the different congeners, it was decided to implement the same single-compartment model for each congener.

243

Safety evaluation of certain food additives and contaminants Eightieth JECFA

9.1.1 Model parameters (a) Absorption

Absorption of PCBs following dietary exposure is congener specific and varies according to age and exposure level. Indeed, Moser & McLachlan (2001) showed that absorption varies with dietary exposure in the range 90–100%. Based on the difference between the amount ingested and the unabsorbed PCB fraction excreted in the faeces, McLachlan (1993) estimated the absorption of NDL-PCBs in a breastfed child to be 96–98% for the main congeners present in the mother’s milk. Alcock et al. (2000) also calculated absorption of around 80% for adults and 95% for infants for PCB 101. Finally, in rats, lower chlorinated congeners with six or fewer chlorine atoms have greater than 90% absorption, and higher chlorinated congeners have about 75% absorption (Albro & Fishbein, 1972; ATSDR, 2000; EFSA, 2005). Consequently, two scenarios were envisioned for the model: ■■ Scenario 1 (base case): A conservative scenario was defined by setting absorption to 100%. ■■ Scenario 2 (sensitivity analysis): A more realistic scenario was defined by setting absorption potentially varying from 75% to 100%. (b) Half-life

WHO Food Additives Series No. 71-S1, 2016

As highlighted in Table 2 and discussed in section 2.1.1, the published estimates of PCB half-lives in humans vary considerably. Ritter et al. (2011) proposed intrinsic elimination half-lives for the different congeners. Such half-lives differ from most empirical estimates of human elimination kinetics for persistent chemicals (i.e. apparent elimination half-lives) that represent the aggregated effect of intrinsic elimination, ongoing exposure and changes in body weight. The approach taken by Ritter et al. (2011) is therefore more accurate and was used to define the base case scenario. As a consequence: ■■ Scenario 1 (base case) is defined with the use of half-lives as estimated from Ritter et al. (2011) where available (i.e. for PCBs 28, 58, 138, 153 and 180). For PCB 101, the single study proposing a human half-life is used (Wolff, Fischbein & Selikoff, 1992). However, this half-life has been calculated jointly for both PCB 99 and PCB 101. ■■ Scenario 2 (sensitivity analysis) is defined to reflect uncertainty and variability in estimated absorption and half-lives across the published studies. A normal distribution of half-lives is considered to describe the uncertainty and variability for each congener. The distribution mean is set to the half-life mean as provided by the literature (Table

244

Supplement 1: Non-dioxin-like polychlorinated biphenyls

2), weighted by the respective study sample sizes. The distribution standard deviation is set to (maximum − mean)/8 (i.e. the z-score between the minimum and the maximum of a normal distribution). (c) Dietary exposure

The dietary exposures (in ng/kg bw per day) estimated in section 8 are used for each country and population subgroup for which there were consumption data. 9.2 Estimates of body burden

With scenario 1, described in Table 25, body burdens (in ng/kg bw) associated with dietary exposure (mean and high consumers) are calculated for each of the six indicator PCBs for children and adults and are shown in Tables 26–32. With scenario 2, described in Table 33, body burdens associated with dietary exposure are calculated for each of the six indicator PCBs for adults who are average consumers and are shown in Tables 34–35. The body burden of children who are average consumers is estimated assuming a dietary exposure equal to the average dietary exposure every day. The body burden of children who are high consumers is estimated assuming a dietary exposure equal to the high consumer dietary exposure every day. The body burden of adults who are average consumers is estimated assuming a dietary exposure equal to the dietary exposure for children who are average consumers every day until 18 years and then the dietary exposure for adults who are average consumers every day until 50 years. The body burden of adults who are high consumers is estimated assuming a dietary exposure equal to the dietary exposure for children who are high consumers every day until the age of 18 years and then to the dietary exposure for adults who are high consumers every day until the age of 50 years. Table 25 Values for the parameters of the model to estimate body burdens with scenario 1 PCB congener 28 52 101 138 153 180 Absorption (%) 100 100 100 100 100 100 Half-life (years) 5.6a 2.6a 5.7b 10.8a 14.4a 11.5a Exposure (ng/kg bw per day) From Table 21 (mean and high consumer, LB and UB) From Table 21 (mean and high consumer, LB and UB) From Table 21 (mean and high consumer, LB and UB) From Table 21 (mean and high consumer, LB and UB) From Table 21 (mean and high consumer, LB and UB) From Table 21 (mean and high consumer, LB and UB)

LB: lower bound; UB: upper bound a Ritter et al. (2011). b Wolff, Fischbein & Selikoff (1992) (co-elution between PCB 99 and PCB 101).

245

Safety evaluation of certain food additives and contaminants Eightieth JECFA

Table 26 PCB 28: estimated body burdens with scenario 1 Country Belgium China Czech Republic Finland France Greece Italy Japan Netherlands Republic of Korea Sweden Belgium China Czech Republic France Germany Ireland Italy Japan Netherlands Republic of Korea Sweden United Kingdom Population group Other children Children Other children Other children Other children Other children Other children Children Other children Children Other children Adults General population Adults Adults Adults Adults Adults General population Adults General population Adults Adults M_LB_28 589 676 292 563 89 1 087 76 71 347 76 324 633 386 168 68 736 109 41 53 286 297 224 32 Body burden (ng/kg bw) M_UB_28 H_LB_28 36 825 1 179 676 1 353 2 231 584 563 1 129 271 6 710 2 737 2 174 968 153 71 142 384 695 89 150 5 597 647 13 224 1 266 386 771 877 336 156 4 347 933 1 472 317 218 377 85 53 103 294 574 353 597 1 698 447 68 65 H_UB_28 73 652 1 355 4 463 1 129 27 546 5 473 1 937 142 768 176 11 197 26 449 771 1 757 13 725 1 869 633 751 103 586 706 3 397 135

H: high consumer; LB: lower bound; M: mean; UB: upper bound

WHO Food Additives Series No. 71-S1, 2016

Table 27 PCB 52: estimated body burdens with scenario 1 Country Belgium China Czech Republic Finland France Greece Italy Netherlands Republic of Korea Sweden Population group Children Children Children Children Children Children Children Children Children Children M_LB_52 2 215 141 58 87 45 2 100 95 152 33 232 Body burden (ng/kg bw) M_UB_52 H_LB_52 18 627 4 431 143 283 1 080 117 87 174 139 3 464 2 776 4 200 557 182 171 306 35 65 2 251 466 H_UB_52 37 255 285 2 161 174 14 219 5 553 1 104 344 69 4 500

246

Supplement 1: Non-dioxin-like polychlorinated biphenyls

Country Belgium China Czech Republic France Germany Ireland Italy Netherlands Republic of Korea Sweden United Kingdom

Population group Adults General population Adults Adults Adults Adults Adults Adults General population Adults Adults

M_LB_52 1 579 79 32 33 260 60 41 138 95 214 67

Body burden (ng/kg bw) M_UB_52 H_LB_52 6 099 3 160 79 159 367 63 73 2 023 430 519 158 121 205 90 142 278 104 189 726 426 88 133

H_UB_52 12 198 160 733 6 387 860 315 401 286 207 1 452 175

H: high consumer; LB: lower bound; M: mean; UB: upper bound

Table 28 PCB 101: estimated body burdens with scenario 1 Country Belgium China Czech Republic Finland France Greece Italy Japan Netherlands Republic of Korea Sweden Belgium China Czech Republic France Germany Ireland Italy Japan Netherlands Republic of Korea Sweden United Kingdom Population group Children Children Children Children Children Children Children Children Children Children Children Adults General population Adults Adults Adults Adults Adults General population Adults General population Adults Adults M_LB_101 3 161 341 229 504 224 7 938 232 416 331 75 1 575 1 962 216 165 186 1 024 306 132 306 437 237 1 893 204 Body burden (ng/kg bw) M_UB_101 H_LB_101 36 737 6 323 344 682 2 164 458 506 1 008 413 6 797 8 935 15 876 1 136 464 416 832 410 658 88 152 8 002 3 151 13 266 3 921 216 428 881 332 279 4 424 1 147 2 052 509 614 470 261 306 611 470 875 282 476 3 723 3 783 240 407 H_UB_101 73 476 685 4 329 1 013 27 903 17 867 2 274 832 824 176 16 004 26 531 431 1 761 13 966 2 297 1 021 940 611 943 563 7 449 476

H: high consumer; LB: lower bound; M: mean; UB: upper bound

247

Safety evaluation of certain food additives and contaminants Eightieth JECFA

Table 29 PCB 138: estimated body burdens with scenario 1 Country Belgium China Czech Republic Finland France Greece Italy Japan Netherlands Republic of Korea Sweden Belgium China Czech Republic France Germany Ireland Italy Japan Netherlands Republic of Korea Sweden United Kingdom Population group Children Children Children Children Children Children Children Children Children Children Children Adults General population Adults Adults Adults Adults Adults General population Adults General population Adults Adults M_LB_138 12 291 764 397 1 788 1 500 16 768 705 577 4 114 136 8 431 9 267 562 251 1 135 4 776 1 015 497 529 4 115 431 9 431 1 812 Body burden (ng/kg bw) M_UB_138 H_LB_138 55 157 24 583 764 1 527 3 346 791 1 788 3 576 1 773 9 934 18 287 33 539 2 026 1 410 577 1 153 4 114 8 228 148 269 9 930 16 865 23 998 18 535 562 1 124 1 588 502 1 299 8 061 4 868 9 551 1 370 2 030 1 113 988 529 1 053 4 126 8 236 486 857 9 999 18 857 1 828 3 619 H_UB_138 110 310 1 531 6 689 3 576 40 781 36 578 4 056 1 153 8 228 296 19 861 48 001 1 130 3 182 25 450 9 737 2 740 2 227 1 053 8 258 977 20 003 3 657

H: high consumer; LB: lower bound; M: mean; UB: upper bound

WHO Food Additives Series No. 71-S1, 2016

Table 30 PCB 153: estimated body burdens with scenario 1 Country Belgium China Czech Republic Finland France Greece Italy Japan Netherlands Republic of Korea Sweden Population group Children Children Children Children Children Children Children Children Children Children Children M_LB_153 21 660 927 1 143 2 575 2 544 2 307 809 1 116 6 258 167 11 918 Body burden (ng/kg bw) M_UB_153 H_LB_153 53 173 43 321 927 1 854 4 570 2 290 2 575 5 146 2 839 11 206 5 058 4 614 2 298 1 617 1 116 2 237 6 258 12 516 185 334 14 563 23 831 H_UB_153 106 347 1 859 9 141 5 146 46 002 10 112 4 601 2 237 12 516 365 29 127

248

Supplement 1: Non-dioxin-like polychlorinated biphenyls

Country Belgium China Czech Republic France Germany Ireland Italy Japan Netherlands Republic of Korea Sweden United Kingdom

Population group Adults General population Adults Adults Adults Adults Adults General population Adults General population Adults Adults

M_LB_153 20 789 807 793 2 318 7 631 1 808 635 1 145 7 162 580 13 758 2 725

Body burden (ng/kg bw) M_UB_153 H_LB_153 33 064 41 579 807 1 608 2 498 1 587 2 518 10 191 7 900 15 269 2 256 3 616 1 414 1 270 1 145 2 298 7 162 14 317 662 1 152 14 862 27 524 2 753 5 451

H_UB_153 66 129 1 615 4 989 32 174 15 801 4 513 2 829 2 298 14 317 1 325 29 718 5 499

H: high consumer; LB: lower bound; M: mean; UB: upper bound

Table 31 PCB 180: estimated body burdens with scenario 1 Country Belgium China Czech Republic Finland France Greece Italy Japan Netherlands Republic of Korea Sweden Belgium China Czech Republic France Germany Ireland Italy Japan Netherlands Republic of Korea Sweden United Kingdom Population group Children Children Children Children Children Children Children Children Children Children Children Adults General population Adults Adults Adults Adults Adults General population Adults General population Adults Adults M_LB_180 6 419 313 421 914 778 2 915 289 285 2 434 76 3 584 4 313 248 265 610 3 265 680 207 265 2 528 346 3 927 553 Body burden (ng/kg bw) M_UB_180 H_LB_180 55 421 12 842 313 621 3 436 846 914 1 824 1 050 1 548 4 984 5 830 1 652 581 285 569 2 434 4 867 100 156 4 210 7 169 25 828 8 627 248 495 1 670 536 783 1 209 3 778 6 530 1 054 1 359 858 409 265 536 2 534 5 062 449 697 4 169 7 855 610 1 106 H_UB_180 110 842 629 6 872 1 824 6 940 9 967 3 304 569 4 867 204 8 420 51 656 495 3 340 4 342 7 555 2 108 1 716 536 5 068 904 8 344 1 227

H: high consumer; LB: lower bound; M: mean; UB: upper bound

249

Safety evaluation of certain food additives and contaminants Eightieth JECFA

Table 32 Sum of the six indicator PCBs: estimated body burdens with scenario 1 Country Belgium China Czech Republic Finland France Greece Italy Netherlands Republic of Korea Sweden Belgium China Czech Republic France Germany Ireland Italy Netherlands Republic of Korea Sweden United Kingdom Population group Children Children Children Children Children Children Children Children Children Children Adults General population Adults Adults Adults Adults Adults Adults General population Adults Adults M_LB_Sum 46 337 3 162 2 541 6 431 5 180 33 114 2 206 13 635 563 26 064 38 544 2 298 1 673 4 350 17 692 3 977 1 553 14 667 1 985 29 447 5 393 Body burden (ng/kg bw) M_UB_Sum H_LB_Sum 255 939 92 678 3 166 6 321 16 829 5 086 6 434 12 857 6 485 39 659 42 777 66 232 8 637 4 407 13 771 27 270 646 1 126 44 553 52 129 115 479 77 086 2 298 4 586 7 881 3 356 5 108 30 256 19 057 35 393 5 664 7 957 4 438 3 103 14 729 29 342 2 336 3 969 35 178 58 892 5 587 10 780 H_UB_Sum 511 882 6 345 33 655 12 862 163 391 85 550 17 275 27 546 1 287 89 108 230 964 4 602 15 762 96 044 38 119 11 329 8 864 29 458 4 682 70 363 11 170

H: high consumer; LB: lower bound; M: mean; UB: upper bound

Table 33 Values for the parameters of the model to estimate body burdens with scenario 2 WHO Food Additives Series No. 71-S1, 2016 PCB congener Absorption Half-life (years) 28 Sampled from uniform (75–100%) Sampled from normal (3.5 ; 0.5) 52 101 138 153 180 Sampled from uniform (75–100%) Sampled from normal (3.8 ; 0.4) Sampled from uniform (75–100%) Sampled from normal (5.7 ; 2) Sampled from uniform (75–100%) Sampled from normal (10.7 ; 3.5) Sampled from uniform (75–100%) Sampled from normal (14.3 ; 5.5) Sampled from uniform (75–100%) Sampled from normal (7.4 ; 1.5) Exposure (ng/kg bw per day) From Table 21 (mean consumer, sampled between LB and UB) From Table 21 (mean consumer, sampled between LB and UB) From Table 21 (mean consumer, sampled between LB and UB) From Table 21 (mean consumer, sampled between LB and UB) From Table 21 (mean consumer, sampled between LB and UB) From Table 21 (mean consumer, sampled between LB and UB)

LB: lower bound; UB: upper bound a Ritter et al. (2011). b Wolff, Fischbein & Selikoff (1992) (co-elution between PCB 99 and PCB 101).

250

Supplement 1: Non-dioxin-like polychlorinated biphenyls

Table 34 PCBs 28, 52 and 101: estimated body burdens with scenario 2 PCB 28 P5 544 990 590 2 382 1 240 625 322 164 1 018 216 97 266 Body burden (ng/kg bw) PCB 52 PCB 101 Median P5 P95 Median P5 5 822 3 075 9 793 7 336 1 908 4 671 2 533 7 808 5 308 1 821 4 498 2 077 8 007 6 036 1 828 12 637 16 278 4 583 180 100 2 482 505 216 727 3 845 5 591 2 165 152 84 774 96 50 208 24 034 31 283 8 123 231 128 4 499 874 528 1 431 15 474 19 313 5 668 327 183 3 313 752 391 946 3 319 5 780 1 367 140 79 679 205 135 186

Country Belgium China Czech Republic France Germany Ireland Italy Japan Netherlands Republic of Korea Sweden United Kingdom

Population group Adults General population Adults Adults Adults Adults Adults General population Adults General population Adults Adults

Median 4 048 4 108 3 710 11 546 14 823 3 783 421 214 2 692 476 260 858

P95 8 512 7 104 7 015 21 382 29 432 7 204 539 275 4 734 937 498 1 445

P95 15 975 13 062 13 045 39 306 47 271 13 484 497 278 7 417 1 662 881 2 534

P5: 5th percentile; P95: 95th percentile

Table 35 PCBs 138, 153 and 180: estimated body burdens with scenario 2 PCB 138 P5 8 771 5 083 4 367 12 409 13 471 5 928 428 225 3 851 383 186 417 Body burden (ng/kg bw) PCB 153 PCB 180 Median P5 P95 Median P5 31 872 12 838 47 274 9 709 4 484 23 651 9 535 34 152 7 874 3 050 20 941 8 306 31 050 7 688 2 630 49 028 64 046 21 627 1 312 728 13 171 2 465 1 351 3 905 23 030 24 925 10 403 570 316 4 631 1 095 627 1 208 90 673 104 586 32 897 1 755 973 24 049 4 792 2 396 7 358 20 469 27 781 8 125 264 145 4 982 1 149 565 1 488 7 000 7 983 2 838 175 97 1 764 433 229 549

Country Belgium China Czech Republic France Germany Ireland Italy Japan Netherlands Republic of Korea Sweden United Kingdom

Population group Adults General population Adults Adults Adults Adults Adults General population Adults General population Adults Adults

Median 16 579 12 748 11 956 33 707 47 259 13 250 901 473 8 292 1 200 790 1 932

P95 30 440 22 192 21 408 70 877 85 140 23 731 1 274 670 18 305 2 731 1 505 4 627

P95 20 054 16 838 15 322 49 988 62 036 16 670 355 195 9 178 2 203 1 128 3 067

P5: 5th percentile; P95: 95th percentile

251

Safety evaluation of certain food additives and contaminants Eightieth JECFA

9.3 Contribution of congeners to the total body burden

The contribution of the different congeners to the total body burden (sum of the six indicator PCBs) varies from one country to another. However, on average, PCB 153 is the main contributor (41%), followed by PCB 138 (28%), PCB 180 (14%), PCB 101 (8%), PCB 28 (6%) and PCB 52 (3%). These contributions are explained by both the dietary exposure and the kinetics of the congeners. Indeed, regarding dietary exposure only, PCB 153 is also the main contributor (33%), followed by PCB 138 (26%), PCB 180 (12%), PCB 101 (12%), PCB 28 (9%) and PCB 52 (7%). The contributions of PCBs 153, 138 and 180 are higher when considering the body burden, as they have the longest half-lives and thus are accumulated more in the human body. 9.4 Comparison with biomonitoring data

The comparison of the body burden estimated from dietary exposure with biomonitoring data cannot be conducted directly. Indeed, because of the accumulation of PCBs throughout life, the body burden measured through biomonitoring studies reflects not only the current exposure, but also previous exposure, including the past few decades. As was highlighted previously in sections 2 and 8, exposure to PCBs has varied substantially over past decades and is known to have been much higher in the 1970s. Consequently, for some countries in which exposure has varied a lot in the past, the actual measured body burden is higher than the body burden estimated in this work, especially for the congeners with the longest half-lives. However, a crude comparison with biomonitoring data shows that the body burdens estimated from the model are of the same order of magnitude as the body burdens estimated from biomonitoring data. As PCBs accumulate in fat, human milk is often used as an indicator of steady-state body burden. Based on the median total concentration of about 240 ng/g fat for all NDL-PCBs measured in human milk sampled in European countries and assuming 20% fat content in the human body, a median human body burden of about 50 000 ng/ kg bw was estimated (EFSA, 2005). A comparison of estimated body burdens for adults (Tables 26–32) with concentrations in human milk (Table 13) shows that estimation of body burden from current dietary exposure is in the same order of magnitude, although a little lower for some countries, as expected (Table 32: Belgium: 38 544–230 964 ng/kg bw; Czech Republic: 1673–15 762 ng/kg bw; France: 4350–96 044 ng/kg bw; Germany: 17 692–38 119 ng/kg bw; Ireland: 3977– 11 329 ng/kg bw; Italy: 1553–8864 ng/kg bw; Netherlands: 14 667–29 458 ng/kg bw; Sweden: 29 447–70 363 ng/kg bw; United Kingdom: 5393–11 170 ng/kg bw). As the concentration in adipose tissue is also known to be a good indicator of steady-state body burden, Table 12 from section 2 can be used for comparison,

252

WHO Food Additives Series No. 71-S1, 2016

Supplement 1: Non-dioxin-like polychlorinated biphenyls

congener by congener. For PCB 28, still assuming 20% fat content in the human body, these results provide body burdens of around 260 ng/kg bw for Belgium, 2580 ng/ kg bw for China and 400 ng/kg bw for the Czech Republic. This is in the same order of magnitude as the results for PCB 28 from Table 26, except for China, where the body burden estimated from current dietary exposure is lower. Regarding PCB 52, concentrations in adipose tissue provide a body burden of 120 ng/kg bw for China and 320 ng/kg bw for the Czech Republic, which are in the same order of magnitude as the predictions from dietary exposure. The same results can be expected for PCB 101 for Belgium, where the body burden estimated from dietary exposure seems to be higher than the one estimated from adipose tissue concentrations. Finally, for PCBs 138, 153 and 180, the comparison highlights that the body burdens estimated from current dietary exposure are much lower for China and the Czech Republic than the body burdens estimated from concentrations in adipose tissue (10 400 ng/ kg bw versus 532–1130 ng/kg bw for China and 24 000 ng/kg bw versus 793–4989 ng/kg bw for the Czech Republic), whereas the two estimates are closer for Belgium (42 000 ng/kg bw versus 9267–48 000 ng/kg bw). Overall, the comparison with biomonitoring data shows some differences that can be mainly explained by the trend in exposure to PCBs over the last decades and also by the uncertainty around the dietary exposure estimations and the kinetic model used, especially around the half-life. 9.5 Uncertainty around the estimated body burdens (scenario 2)

With scenario 2, three sources of uncertainty were taken into account in the estimation of the body burdens for the mean consumer: (1) the uncertainty around the absorption of PCBs, (2) the uncertainty around the half-life of PCBs and (3) the uncertainty around the true level of exposure due to data below the LODs. Tables 34 and 35 address that by taking into account this uncertainty; the confidence interval around the estimate of the body burdens is large, but still within a reasonable order of magnitude. This result is even more obvious for PCBs 138 and 153, and this can be explained by the great uncertainty around the half-life, which was considered with scenario 2. Indeed, as shown in Table 2 discussed in section 2, the half-lives provided by the literature vary a lot from one study to another and lead to a significant uncertainty. In this modelling, the interindividual variability in kinetics and especially in half-lives was not taken into account, as data on these apects were not available in the literature. However, as a large distribution was used in scenario 2 for half-life and absorption, it seems reasonable to assume that the interindividual variability would be covered by such a distribution.

253

10. Dose–response analysis Following a review of both the experimental and human data on NDL-PCBs, the Committee noted that there are a large number of in vivo and in vitro studies available covering several of the indicator PCBs with respect to possible hepatotoxicity, thyroid toxicity, and neurodevelopmental or neurotoxic effects. However, there is a general lack of in vivo toxicity data for two of the indicator PCBs (PCB 101 and PCB 138). Given that some of the experimental toxicological data on NDL-PCBs indicate some common targets for toxicity, the Committee considered whether a group evaluation could be undertaken. Noting the lack of data on PCB 101 and PCB 138, the latter being one of the major contributors to dietary exposure to the six indicator PCBs, and that there is not enough information on relative potencies for each congener with respect to receptor interactions and the downstream consequences, it was decided not to undertake a group evaluation. The Committee further concluded that none of the available shortterm toxicity studies on four of the indicator PCBs and PCB 128 was suitable for the derivation of health-based guidance values (e.g. provisional tolerable monthly intakes) or assessment of their relative potency compared with a reference compound, such as PCB 153. Therefore, a comparative approach using the minimal effect doses from the available studies was developed in order to estimate MOEs to provide guidance on human health risk. The available rat toxicological data on individual congeners showed that minimal changes in liver and thyroid histopathology were evident from the lowest doses tested of 2.8–7 µg/kg bw per day and were similar across the shortterm and long-term studies of toxicity. Bearing in mind that, with the exception of PCB 153, the available studies on individual NDL-PCB congeners were of relatively short duration (28 or 90 days), the Committee decided to take the lower end of the range of administered (external) doses used for each congener at which these minimal changes occurred as a conservative point of departure for estimating MOEs. Given the major difference in dosing regimens between the 28-day and 90-day studies and the bioaccumulative nature of PCBs, MOEs were estimated on the basis of both external dose and internal dose. Table 36 shows the ranges of administered doses in rats and the corresponding measured internal doses (adipose tissue concentrations). Fig. 7 shows the same information in graphical form for easier comparison. The internal dose MOEs based on amounts present in adipose tissue were considered the most appropriate comparison, particularly because they also eliminate interspecies differences in toxicokinetics.

254

Supplement 1: Non-dioxin-like polychlorinated biphenyls

Table 36 NDL-PCB concentrations in fat tissue in critical studies in rats PCB congener Reference Study duration Males PCB 28 Chu et al. (1996a) 90 days Daily dose or total dose administered (units) Daily dose (µg/kg bw per day) 0 2.8 36 359 3 783 Daily dose (µg/kg bw per day) 0 4.2 42 425 4 210 Daily dose (µg/kg bw per day) 0 3.6 34 346 3 534 Total dose (mg/kg bw) 0 3 10 30 100 300 1 000 1 700 Daily dose (µg/kg bw per day) 0 2.9 37 365 3 956 Daily dose (µg/kg bw per day) 0 4.5 45 441 4 397 Daily dose (µg/kg bw per day) 0 4.2 42 428 4 125 Concentration of PCB in lipid (µg/g)

0.32 0.74 4.6 48 531 0.02 0.67 8.12 69.8 531 0.03 1.2 11.6 89 866 0.26 15.7 36.4 122 372 1 320 3 920 7 390

PCB 128 Lecavalier et al. (1997) 90 days

PCB 153 Chu et al. (1996b) 90 days

PCB 180 Viluksela et al. (2014) 28 days

Females PCB 28 Chu et al. (1996a) 90 days

0.47 0.49 4 40 563 0.16 0.97 12.8 125 1 006 0.16 2.5 36 157 1 840

PCB 128 Lecavalier et al. (1997) 90 days

PCB 153 Chu et al. (1996b) 90 days

255

Safety evaluation of certain food additives and contaminants Eightieth JECFA

Table 36 (continued) PCB congener Reference Study duration PCB 153 NTP (2006a) 2 years

PCB 180 Viluksela et al. (2014) 28 days

Daily dose or total dose administered (units) Daily dose (µg/kg bw per day) 0 10a 100a 300a 1 000a 3 000a Total dose (mg/kg bw) 0 3 10 30 100 300 1 000 1 700

Concentration of PCB in lipid (µg/g) 0.436 20.1 158 519 1 557 4 292 0.29 15.5 42.9 165 435 1 380 4 930 11 300

a

These doses were given on 5 days/week. If adjusted for 7 days/week dosing, the doses are equivalent to approximately 0, 7, 71, 214, 714 and 2143 µg/kg bw per day, respectively.

Fig. 7 Comparison of adipose tissue concentrations of PCBs in critical studies on male rats and indicative concentrations of the same congeners in human fat 0 2.8 36 359 3783 µg/kg bw/d

0

3.6

34

346

3534 µg/kg bw/d

WHO Food Additives Series No. 71-S1, 2016

0

4.2

42

425

4210 µg/kg bw/d

total dose mg/kg bw

0

3

10

30

100

300

1000 1700

0.01

0.1

1

10

100

1000

mg/kg fat tissue

256

Supplement 1: Non-dioxin-like polychlorinated biphenyls

Legend Figures in boxes represent administered doses in each study. Symbols represent the corresponding concentrations of PCB congeners in adipose tissue. PCB 28, Chu et al. (1996a), 90 days PCB 153, Chu et al. (1996b), 90 days PCB 128, Lecavalier et al. (1997), 90 days PCB 180, Viluksela et al. (2014), 28 days Average concentration in human fat (not available for PCB 128 )

257

11. Comments 11.1 Toxicokinetics and mode of action

The main determinants of the fate and behaviour of PCB congeners in the body are their lipid solubility and their rate of metabolism. In general, PCBs are lipid soluble and are well absorbed from the gastrointestinal tract in mammalian species. They are rapidly distributed to all body compartments, especially the liver and muscle. The highest amounts of PCBs are usually found in the liver, fat, skin and breast milk. Rates of metabolism of PCBs vary greatly across species and also vary with the number and position of the chlorine atoms in the different congeners. In all species studied, PCB congeners with adjacent unsubstituted carbon atoms in the meta and para positions are more readily metabolized, whereas congeners without such adjacent unsubstituted carbon atoms are generally metabolized and cleared very slowly. PCBs with higher numbers of chlorine atoms are generally metabolized more slowly than those with lower numbers of chlorine atoms. Some PCBs with few chlorine atoms have apparent half-lives in blood as short as a week in experimental animals. However, many higher chlorinated PCB congeners have half-lives in humans that are much longer and therefore accumulate in the body. Depending on the species, half-lives vary from several months (e.g. rat) or a year or more (e.g. monkey) to over a decade (humans). The long half-lives in humans compared with rodents have been attributed to the poor metabolism of these compounds, and this, combined with the high lipid solubility of PCBs, results in high PCB content in the adipose tissue of humans. In terms of PCB metabolites, the hydroxy metabolites tend to be polar and more readily excreted, whereas the methyl sulfone metabolites are lipophilic and can be retained in adipose tissue. PCBs are potent inducers of both phase I and phase II enzymes in the liver, and there are several routes of metabolism for PCBs. Biotransformation involves initial phase I oxidation by cytochrome P450 enzymes. PCBs can be oxidized across the aromatic ring to one or more unstable intermediate arene oxides, which can spontaneously rearrange to produce hydroxy metabolites. PCBs that oxidize to more stable arene oxides are subsequently reduced to dihydroxy metabolites. Dihydroxy metabolites can then be dehydrogenated to form catechols, which are in equilibrium with their oxidized forms, the corresponding hydroquinones and quinones. In these metabolic processes, dechlorination and shift of chlorine atoms may also occur. Thus, lower chlorinated NDL-PCB congeners can be metabolized to reactive intermediates, such as epoxides, quinones and reactive oxygen species, that can form adducts with macromolecules, such as proteins, DNA, RNA and lipids. PCB congeners lacking adjacent unsubstituted hydrogen atoms do not easily form arene oxides but can be metabolized by an alternative 258

Supplement 1: Non-dioxin-like polychlorinated biphenyls

pathway of direct insertion of a hydroxyl group to form a monohydroxy metabolite. Hydroxy metabolites are excreted as such, or the lower chlorinated PCBs can be conjugated with glucuronide or sulfate by the phase II enzymes UGT and SULT. Another route of metabolism for PCBs, which involves GST, is the formation of methyl sulfones. Methyl sulfones are the final product of the most rapidly metabolized PCBs. Approximately 40 of the hundreds of potential hydroxy metabolites have been identified in human blood. Of these, only five persist in the blood, and these are hydroxy metabolites that are substituted in the para position, with chlorine atoms on each side of the hydroxyl group. In blood, they are bound to transthyretin, which normally binds T4. Concentrations of hydroxy metabolites in human blood are approximately 5–10 times lower than those of the most persistent parent PCB congeners. Methyl sulfones undergo enterohepatic recirculation, and this, together with their lipophilicity, may account for some of the long retention times for methyl sulfone metabolites. Fifty or more methyl sulfone metabolites have been detected in human blood, but their concentrations in blood are low (generally less than 1% of the concentration of the most persistent PCB congeners), and much lower than those of the hydroxy metabolites. For both hydroxy and methyl sulfone metabolites, it may be more relevant to assess the effects of those metabolites with the highest retention potential than to assess the effects of the parent congener. PCBs can interact with several cellular receptors, including CAR, PXR and AhR. The induction profile for these receptors differs between NDL-PCBs and DL-PCBs: NDL-PCBs most typically activate CAR and PXR, whereas DLPCBs induce a pronounced activation of AhR, but not of CAR or PXR. Activation of either CAR and PXR or AhR results in different cytochrome P450 enzyme induction profiles. PXR and CAR activation induces CYP3A and CYP2B isoforms, respectively, whereas AhR activation induces CYP1A1, CYP1A2 and CYP1B1 isoforms. These differing cytochrome P450 induction profiles have traditionally been used to differentiate between NDL-PCB and DL-PCB congeners for toxicological purposes. Interactions with CAR and PXR are crucial in the biotransformation and elimination of NDL-PCBs, because they can induce the relevant cytochrome P450 enzymes that metabolize NDL-PCBs. The activation of CAR and PXR also has potential toxicological implications, as these receptors play a significant role in the metabolism of endogenous molecules, such as hormones and vitamins. For example, induction of cytochrome P450 and conjugation enzymes by NDL-PCBs can influence hormonal homeostasis, as demonstrated in animal experiments for thyroid and steroid hormones, corticosteroids and retinoids. Recent studies have also indicated that CAR and PXR play an important role in the development of diabetes and inflammation. 259

Safety evaluation of certain food additives and contaminants Eightieth JECFA

NDL-PCBs also induce conjugation enzymes, such as UGTs, SULTs and GSTs. All these enzymes play important roles in the metabolism of NDL-PCBs, which is not only an important detoxification process, but also a route for the formation of transient reactive intermediates and more persistent hydroxy and methyl sulfone metabolites that are toxicologically relevant. Hydroxy metabolites can be agonists or antagonists for estrogen receptors, can interfere with thyroid hormone homeostasis and have neurotoxic potential. Methyl sulfone metabolites also interfere with thyroid hormone homeostasis and have been shown in vitro to have anti-estrogenic activity and to act as antagonists for glucocorticoid receptors. Based on the above mechanisms, a sustained exposure to NDL-PCBs may have toxicological implications. In addition, there is cross-talk between PXR or CAR and other nuclear receptors, but the full scope of these interactions has not yet been fully elucidated. NDL-PCBs also activate RyR, which plays a crucial role in calcium signalling and in the decrease of brain dopamine levels. These mechanisms are thought to be major pathways leading to the observed neurobehavioural toxicity of NDL-PCBs in experimental animals. In general, the interactions of NDL-PCBs with these receptors and the enzyme activation reported in animal studies are considered to have human relevance. 11.2 Toxicological data 11.2.1 Acute toxicity and short-term studies of toxicity

There is no information on the acute toxicity of individual NDL-PCB congeners. The available data are mainly on rats and include 28-day studies on PCB 52 and PCB 180 and 90-day studies on PCB 28, PCB 128 and PCB 153. There are no short-term studies on two of the indicator PCBs, PCB 101 and PCB 138. In the two 28-day studies, the doses were expressed as total doses administered by oral gavage over the entire study period; the total dose comprised four (PCB 52) or six (PCB 180) higher daily loading doses during week 1, followed by three lower maintenance doses given 3 times per week during weeks 2–4. The total doses over the entire study period ranged from 3 to 3000 mg/kg bw for PCB 52 and from 3 to 1700 mg/kg bw for PCB 180. In the 90-day studies, fixed concentrations were administered in the diet and expressed as daily doses, which were similar for all three congeners, ranging from approximately 0.003 to 4.4 mg/kg bw per day. The main effects observed in these repeated-dose studies were on liver and thyroid and were fairly consistent across studies. Body weight was not affected in any of the 90-day studies. In the liver, one of the most sensitive responses to exposure to NDLPCBs is induction of phase I and phase II enzymes. In the short-term studies, this was reflected in structural changes, such as hepatocyte hypertrophy, cellular

260

WHO Food Additives Series No. 71-S1, 2016

Supplement 1: Non-dioxin-like polychlorinated biphenyls

vacuolation and alterations in cytoplasm density and homogeneity, which are well-recognized adaptive signs of increased liver activity following exposure to xenobiotics. The pattern of effects was similar for the five congeners tested, with effects being observed at most doses in both the 28-day and 90-day studies, usually from the lowest dose tested, and with males tending to be more sensitive than females. In many instances, there was little or no dose–response relationship in either incidence or severity of these minimal changes over the entire dose range covering 3 orders of magnitude (whether expressed as administered dose or as adipose tissue concentration at the end of dosing). In the 28-day studies, liver weight was increased at the highest total dose of 3000 mg/kg bw for PCB 52 and with a dose–response relationship for PCB 180 at doses of 300 mg/kg bw and higher. In the 90-day studies, liver weight was increased at the highest dose of 4.4 mg/kg bw per day for PCB 128 and at the highest dose of 4.1 mg/kg bw per day for PCB 153. Studies on individual NDL-PCB congeners have shown effects on thyroid histology and/or circulating thyroid hormone concentrations in adults. Effects on thyroid histology were seen in all the 28-day and 90-day studies from the lowest doses tested. The effects included reductions in the size of large follicles, collapsed follicles, increases in epithelial height and cytoplasmic vacuolation. Blood thyroid hormone levels were measured in the two 28-day studies, with reductions in T4 levels at and above 300 mg/kg bw total dose for PCB 52 and doserelated reductions at and above 100 mg/kg bw total dose for PCB 180. Effects on the thyroid are potentially important, particularly because of the sensitivity of the developing brain to reductions in maternal and early postnatal thyroid hormone levels. It should be noted that both DL-PCBs and NDL-PCBs (and their hydroxy metabolites) can have effects on the thyroid and/or circulating thyroid hormone levels. In rodents, commercial mixtures of PCBs reduce circulating total T4 and free T4, but have little or no effect on total or free T3 or on TSH. Studies on NDL-PCBs also show effects on T3, but these are usually less marked than those on T4. The precise mechanisms underlying these changes and their relative contributions to NDL-PCB-induced thyroid effects are not yet clear. 11.2.2 Long-term studies of toxicity and carcinogenicity

The only NDL-PCB congener studied for long-term toxicity and carcinogenicity is PCB 153 (NTP, 2006a). Analytical checks of the test material for purity showed no contamination with DL-PCBs and only minor contamination with PCB 101 (0.21%) and PCB 180 (0.002%). Male animals were not used. Female rats were administered PCB 153 by oral gavage at 0, 10, 100, 300, 1000 or 3000 µg/kg bw per day, 5 days/week, for up to 105 weeks (equivalent to 0, 7, 70, 200, 700 or 2000 µg/ kg bw per day when adjusted for 5 days/week dosing). These doses resulted in a 261

Safety evaluation of certain food additives and contaminants Eightieth JECFA

WHO Food Additives Series No. 71-S1, 2016 262

linear increase in concentrations in fat; at the end of the study, concentrations were approximately 440, 20 000, 160 000, 520 000, 1 600 000 and 4 300 000 ng/g lipid for the 0, 10, 100, 300, 1000 and 3000 µg/kg bw per day dose groups, respectively. There was equivocal evidence of carcinogenic activity of PCB 153, based on the occurrence of a small number of cholangiomas of the liver in two animals in each of the two highest dose groups. The study authors considered that the occurrence of bile duct hyperplasia at doses of 300 µg/kg bw per day and above could have contributed to cholangioma formation, and so the tumours may have been treatment related. The Committee noted that “bile duct hyperplasia” might better be described as atypical tubular epithelial cell hyperplasia. It is notable that in this study, there was no increase in hepatic cell proliferation or any increases in hepatocellular adenomas or carcinomas. This is despite dose-related increases in hepatocyte hypertrophy, which were seen from the lowest dose of 10 μg/kg bw per day, equivalent to 7 μg/kg bw per day when adjusted for 5 days/week dosing, and which became statistically significant at doses of 300 μg/kg bw per day and above during the first year of the study and at all dose levels by the end of the study. There were also statistically significant increases in absolute and/or relative liver weights at 1000 and 3000 μg/kg bw per day at weeks 14 and 31 and at doses of 100 μg/kg bw per day and above at week 53 of the study; there was also evidence of diffuse fatty change in the liver at doses of 300 μg/kg bw per day and above at the end of the study. Concerning the thyroid, there were no increases in thyroid tumours, despite statistically significant reductions in serum thyroid hormone concentrations (total T4, free T4, free T3) in the 3000 µg/kg bw per day group during the first year of the study and a significant increase in follicular cell hypertrophy in the mid-dose group (300 μg/kg bw per day) and highest-dose group (3000 μg/kg bw per day) at the end of the study. There were no effects on thyroid weight. The observations from this long-term study support the view that the liver and thyroid changes observed in the short-term studies on PCB 153 and the four other NDL-PCBs, at lower dose levels than those used in this study, are unlikely to lead to major pathological changes over the long term. NDL-PCBs may have weak tumour promotion effects in the liver, based on studies in rodents using DEN as an inducer. 11.2.3 Genotoxicity

Genotoxicity studies on individual NDL-PCB congeners have produced both positive and negative results. Some in vitro tests on PCB 3/PCB 3 metabolites, PCB 52/PCB 52 metabolites, PCB 101, PCB 138 and PCB 153 were positive for genotoxicity. In vivo studies have been conducted only on PCB 52 and PCB 153, and these were negative. The positive results in vitro may be due to the formation

Supplement 1: Non-dioxin-like polychlorinated biphenyls

of reactive intermediates and induction of oxidative stress. It seems likely that some NDL-PCBs may be indirect-acting genotoxicants. 11.2.4 Reproductive and developmental toxicity

There are no oral reproductive toxicity studies on individual NDL-PCB congeners. In developmental toxicity studies on individual NDL-PCB congeners (PCB 28, PCB 153 and PCB 180) in rodents, reduced birth weight and increases in offspring liver weight were seen only at high doses of 32 mg/kg bw per day and above. The majority of developmental toxicity studies have focused on neurodevelopmental outcomes. Rodent studies have shown that prenatal and/ or postnatal exposures to NDL-PCBs cause effects on end-points such as spontaneous (locomotor) activity, habituation capability, spatial learning and anxiety-like behaviour at maternal doses ranging from 0.2 to 1000 mg/kg bw per day. A limitation of many of the neurobehavioural studies is that they used only one or two dose levels and effects were seen at the lowest dose or at the only dose tested, which does not allow derivation of a NOAEL. The Committee noted that the administered doses used in the developmental neurobehavioural studies were considerably higher than those used in the short- and long-term studies of toxicity and that the lowest effect level in the developmental neurobehavioural studies was at least 2 orders of magnitude higher than the lowest doses that induced minimal changes in liver and thyroid in the short-term studies of toxicity. Internal dose data (concentrations in fat) are not available for the developmental toxicity studies. In vitro studies also indicate a potential role of hydroxylated metabolites of NDL-PCBs in neurodevelopmental mechanisms of toxicity, but support from in vivo studies is very limited. These results from developmental neurobehavioural studies in rodents indicate similar patterns of effects, albeit with congenerspecific differences. Based on in vivo studies, a distinct mechanism of action for the neurodevelopmental effects of NDL-PCBs cannot be established. However, results from in vitro or ex vivo studies using neuronal cells indicate mechanistic pathways that involve disruption of intracellular calcium or thyroid hormone homeostasis. Based on available data, the Committee concluded that neurodevelopmental outcomes are not the most sensitive end-points for the toxicity of NDL-PCBs in rodents. 11.2.5 Immunological studies

In vitro and in vivo studies using doses ranging from 0.1 mg/kg bw per day up to several hundred milligrams per kilogram of body weight per day have shown that NDL-PCBs can exert immunological effects. In these studies, a comparison was often made between TCDD and an NDL-PCB, such as PCB 153. Results from in 263

Safety evaluation of certain food additives and contaminants Eightieth JECFA

vivo studies in mice indicate that the mechanism of action for immunological effects differs between dioxin-like and non-dioxin-like compounds. For example, PCB 153 significantly enhanced splenic PFC responses to SRBCs injected into mice, whereas TCDD induced an opposite response. Moreover, in studies of the effects of co-administration of TCDD and PCB 153, it has been shown that PCB 153 can counteract the AhR-mediated effects of TCDD on PFC responses. The results of repeated-dose studies in rodents suggest that PCB 153 can induce a proinflammatory response in various tissues (e.g. liver, spleen, lungs and uterus), which is not seen with TCDD. From the doses used in the immunological studies, such effects seem unlikely to be the most sensitive end-points for NDL-PCBs, but the data relate mostly to PCB 153. 11.3 Observations in humans 11.3.1 Biomonitoring and modelling of body burden

The most commonly used biomarkers of PCB exposure in humans are PCB concentrations in adipose tissue, serum, plasma and milk. These mainly reflect exposure from the diet. There is a strong correlation between serum and adipose tissue concentrations, when expressed on a lipid basis, and both are widely regarded as useful biomarkers of PCB body burden. In general, concentrations in blood lipids reflect more recent exposures, as well as the full spectrum of PCB congeners to which a person has been exposed, whereas the pattern of PCB congeners in adipose tissue reflects long-term exposure and, to a lesser degree, the extent to which a particular congener is metabolized. Concentrations in human milk largely reflect the pattern and amounts of PCB congeners present in maternal adipose tissue and blood, when expressed on a lipid basis. Numerous publications confirm that PCB 138, PCB 153 and PCB 180 are the most consistently detected and quantitatively dominant PCB congeners found in human blood and tissues, accounting for 65–80% of total PCBs in human serum. If only one congener is to be used as a marker of total PCB exposure, then PCB 153 is a good choice, because it is very stable and often the most abundant congener. PCB 153 has been shown to have a high correlation with the total amount of PCBs in human milk, plasma and serum. However, if a more complete profile of congeners is considered, the correlations are lower, and either total PCBs or PCB 153 as a marker of the total could be misleading indicators of the differential exposure to other individual or groups of congeners of toxicological significance (ATSDR, 2000). The total body burden of PCBs and their metabolites generally increases with age, and this is reflected in blood and adipose tissue biomarker concentrations. With the phasing out of the production of PCBs since the 1980s, results of studies conducted since that time show clear trends for decreasing concentrations of

264

WHO Food Additives Series No. 71-S1, 2016

Supplement 1: Non-dioxin-like polychlorinated biphenyls

PCBs in blood and human milk: in Europe, mean concentrations of PCB 138, PCB 153 and PCB 180 in blood appear to have decreased by approximately 80% in 20 years; in WHO surveys on human milk, concentrations of the six indicator PCBs have steadily decreased by approximately 10-fold over the decade 2000– 2010. Biomonitoring results from numerous countries over the preceding decade illustrate a wide range of concentrations for individual congeners in human serum, adipose tissue and milk. The results are summarized in Table 37, together with equivalent and modelled body burdens. The concentrations of hydroxy metabolites of PCBs in human serum are not shown, but are in a similar range to the concentrations of many parent PCB congeners, except for those PCBs that are the most prevalent or persistent. Human milk, serum and adipose tissue are all considered to be relevant matrices for assessment of body burdens of PCBs. Human milk has been recognized by WHO as the preferred matrix for monitoring levels of environmental contaminants. In Table 37, ranges of equivalent body burdens were derived using the available range of mean PCB concentrations reported in human milk, as the Committee considered that the milk values were more representative, reflecting both adult (maternal) and infant exposures. The Kinetic Dietary Exposure Model (Verger, Tressou & Clémençon, 2007) was used to simulate body burdens in adult populations in the countries from which dietary exposure data (see below) were available (China, Czech Republic, Finland, France, Germany, Ireland, Italy, Japan, the Netherlands, the Republic of Korea, Sweden and the United Kingdom). Based on UB mean dietary exposures, body burdens were also modelled for each congener and each country. Ranges of modelled body burden estimates across countries are reported in Table 37. No modelled body burdens for PCB 128 could be developed owing to the lack of half-life information. The human body burden for each congener was predicted using a onecompartment model, an assumption of 20% body fat and information on dietary exposure (see below). Half-life estimates were drawn from Ritter et al. (2011), and sensitivity analyses were performed to take into account the large variability in half-life data across other publications. The modelling predicted body burdens of the same order of magnitude as those reported in human biomonitoring data. 11.3.2 Epidemiology

Humans are exposed to complex mixtures of PCBs. The epidemiological literature covers studies that have analysed outcomes according to both exposure to complex mixtures of PCBs and exposure to specific marker NDL-PCBs. The studies included a broad range of potential outcomes, including growth and 265

Safety evaluation of certain food additives and contaminants Eightieth JECFA

Table 37 Summary of human biomonitoring results on selected NDL-PCBs from studies published in the last decade and body burden estimates Range of mean concentrations (ng/g lipid) NDL-PCB congener 28 52 101 128e 138 153 180 Seruma 1–14 0.2–26 0.1–5.2 NA 3.6–186 11–423 6–374 Adipose tissue 0.7–39 0.3–72 0.2–58 NA 3.6–181 0.9–310 1.7–245 Milkb 0.6–7.8 0.2–15.8 0.3–8.3 0.2–4.0 1.6–98 3.4–130 1.6–96 Range of equiva- Range of modelled lent body burdenc body burdend (µg/kg bw) across (µg/kg bw) across studies countries 0.12–1.6 0.05–1.7 0.04–3.6 0.06–0.7 0.06–1.7 0.14–3.7 0.04–0.8 NA 0.32–20.1 0.49–10.0 0.68–26.7 0.66–14.9 0.32–19.7 0.25–4.2

NA: not available a The majority of serum samples were at the low end of the range. b Excludes results from the upper end of the range in Czech Republic, as the data set included results from an industrial area with a former PCB-based paint production plant. c Body burdens based on reported range of concentrations in human milk; 65 kg bw adult; 20% lipid. d Model-based estimates using upper-bound mean dietary exposure in China, Czech Republic, Finland, France, Germany, Ireland, Italy, Japan, the Netherlands, the Republic of Korea, Sweden and the United Kingdom. e Based on limited human milk analysis (Koopman-Esseboom et al., 1994; Todaka et al., 2010; Ryan & Rawn, 2014).

development, neurodevelopment and neurobehaviour in childhood, neurotoxic effects in adults, cancer, endocrine and metabolic effects (e.g. on thyroid hormone homeostasis, diabetes, obesity, insulin resistance and metabolic syndrome), reproductive effects in males and females, immunological effects and infections, respiratory diseases, cardiovascular diseases, hepatic effects, musculoskeletal effects and endometriosis. However, methodological issues and some study design features must be considered for a proper interpretation of human studies. Of particular concern is the extensive use of cross-sectional studies reporting associations between exposure and outcome, in which the exposure measurements are taken at the same time as the outcome is ascertained. A cross-sectional estimate of body burden may not reflect the exposure during the time period critical for the development of a particular outcome. Exceptions are in cases when the exposure and response are known to occur during a defined short period (e.g. prenatal exposure measured in cord blood in relation to effects in newborns). Other than this, cross-sectional studies are of little value. In case–control studies, major drawbacks arise from the fact that measurement may be affected by the disease and treatments, as well as the potential for selection and information bias in hospital-based studies. The degree of control of relevant confounders is highly variable across studies, including exposure to other contaminants, and the potential for confounding by factors not explicitly considered in the analysis cannot be ruled out. Finally, there is always co-exposure to dioxin-like congeners;

266

WHO Food Additives Series No. 71-S1, 2016

Supplement 1: Non-dioxin-like polychlorinated biphenyls

given the strong collinearity between exposure to DL-PCBs and NDL-PCBs, this makes it very difficult to make a valid estimate of the independent effect of NDLPCBs. In spite of all the limitations, some well-designed and well-conducted studies have identified potential health effects associated with exposure to NDLPCBs, including changes in thyroid hormone homeostasis, neurodevelopmental effects, immunological effects and some types of cancer. Some of the results offer support for the toxicological findings, especially regarding thyroid effects, identified as potentially relevant for NDL-PCBs in animal studies. The results of prospective (Chevrier et al., 2007; Herbtsman et al., 2008; Darnerud et al., 2010) and cross-sectional studies in newborns (Takser et al., 2005; Wang et al., 2005; Herbstman et al., 2008) and children (Álvarez-Pedrerol et al., 2008; Schell et al., 2008) suggest that increasing concentrations of NDL-PCBs are correlated with lower levels of T4 and higher levels of TSH in the blood, although there are some inconsistencies between results across studies. Perinatal exposure to NDL-PCBs in birth cohorts was found to be associated with increased incidence of acute respiratory infections in children (Dallaire et al., 2006; Glynn et al., 2008; Stølevik et al., 2013). Maternal and early postnatal exposure to NDL-PCBs in some birth cohorts was associated with impaired behavioural, cognitive and psychomotor development (Stewart et al., 2005; Park et al., 2010; Forns et al., 2012b; Lynch et al., 2012; Gascon et al., 2013; Tatsuta et al., 2014) and with alteration of VEPs (Saint-Amour et al., 2006). Regarding cancer, the recent evaluation by IARC (2015) reported an association between melanoma and PCB exposure, mainly based upon cohort studies of exposed workers in various industries, for whom exposure would be by multiple routes. IARC (2015) also considered studies in the general population with different study designs. Only one population-based case–control study (Gallagher et al., 2011) reported specific results for NDL-PCBs, showing a significantly increased risk of melanoma for a group of 11 NDL-PCBs, as well as for some individual congeners. In this study, a similar increased risk was also observed for two DL-PCB congeners. The association between NDL-PCBs and NHL has also been assessed in several prospective cohorts (Engel et al., 2007; Bertrand et al., 2010; Laden et al., 2010; Bräuner et al., 2012), but the results were not consistent. 11.4 Analytical methods

The methodologies used for the analysis of NDL-PCBs are largely similar regardless of the laboratory performing the analysis. Although some agencies (e.g. AOAC International, International Organization for Standardization) have developed validated matrix-specific methods of analysis that may be followed for the analysis 267

Safety evaluation of certain food additives and contaminants Eightieth JECFA

of NDL-PCBs and DL-PCBs, others have developed a set of performance criteria to ensure that laboratories develop data of acceptable quality. The use of automated solid-phase extraction systems for the extraction and/or cleanup of samples has increased, which has improved efficiency. GC coupled to 63Ni ECDs and mass spectrometers (including ion trap, low-resolution, high-resolution and tandem mass spectrometers) have been used in the analysis of NDL-PCBs. The availability and use of stable isotope internal standards and certified reference materials lead to improved accuracy of analytical results. The use of analytical methods with satisfactory performance characteristics, as well as methods subjected to interlaboratory comparison studies, should be considered, as appropriate, to ensure that the data submitted for evaluation are of adequate quality. 11.5 Sampling protocols

Although there are no established protocols set specifically for the collection and storage of samples for NDL-PCB analysis, best practices have been established for other POPs present at ultra-trace levels (e.g. PCDDs/PCDFs, DL-PCBs). These practices include the collection of samples using containers that are non-reactive (e.g. glass, aluminium) and that have been chemically cleaned or certified to be free of contaminants. 11.6 Effects of processing

WHO Food Additives Series No. 71-S1, 2016 268

NDL-PCBs are thermally stable and resistant to degradation. Studies on the impact of processing in relation to PCB concentrations have been largely focused on the cooking techniques used to prepare foods and techniques that change the fat content (e.g. PCB levels are lowered in skimmed milk, but increased concentrations are found in foods with higher fat content, such as cheese or cream). Although the studies related to the impact of processing on PCB concentrations include both DL-PCBs and NDL-PCBs, the impact on the concentrations is similar for both groups. Ultimately, processing that results in the removal of lipids will lead to a decrease in PCB concentrations in the final food product. 11.7 Prevention and control

The focus of efforts related to preventing exposure to POPs, including NDLPCBs, is on limiting contamination of the food-chain, including exposure of food-producing animals to PCBs. With the knowledge that fish, meat and dairy product consumption makes the most significant contribution to human PCB exposure, methods of PCB reduction in animals from which these foods are derived are of primary interest. Transfer of DL-PCBs and NDL-PCBs from

Supplement 1: Non-dioxin-like polychlorinated biphenyls

feed to animal-based food products (e.g. milk) occurs; transfer of PCB 138, PCB 153 and PCB 180 is greater than that observed for PCB 28, PCB 52 and PCB 101. Adherence to good agricultural practices and good animal feeding practices will contribute to the efforts to reduce PCB concentrations in food for human consumption. PCB contamination of animal housing and/or buildings (e.g. silos) near animal pastures may contribute to animal exposure levels, as does the pasturing of animals on lands contaminated with PCBs. It is therefore important to identify contaminated pastures to ensure that they are not used for grazing. PCB contamination can be further reduced by establishing and adhering to soil guidelines for agricultural purposes, performing diligent monitoring programmes to confirm compliance and establishing critical control points for the feed manufacturing process where PCB concentrations can be reduced. Additionally, farming practices should include plans for isolation, among other procedures, if PCB contamination is detected. 11.8 Levels and patterns of food contamination

Thirty countries (Australia, Austria, Belgium, Canada, China, Cyprus, Czech Republic, Denmark, Estonia, Finland, France, Germany, Greece, Iceland, Ireland, Italy, Japan, Luxembourg, the Netherlands, New Zealand, Norway, Poland, the Republic of Korea, Romania, Singapore, Slovakia, Slovenia, Spain, Sweden and the United Kingdom) provided data to the Committee for its review of NDLPCBs. The submissions from Europe were received through EFSA. Those data that were not in an acceptable format for evaluation by the Committee were removed from the data set. Submitted data included results from sample collection periods extending from 1995 to 2014. Most (90.5%) of the data were submitted from Europe; the Pacific region contributed 8.5% of the submissions, and 1% of the data were provided from North America. The food categories having elevated NDL-PCB concentrations were fish/ fish products, meat/meat products, egg/egg products and milk/milk products. Concentrations of the six indicator PCBs in these food categories varied widely (Table 38). Few occurrence data were submitted for PCB 128, and lower concentrations of this congener were reported relative to the indicator congeners (Table 38). It was noted that concentrations of the higher chlorinated indicator congeners (PCB 101, PCB 138, PCB 153 and PCB 180) were higher than those of the trichlorinated PCB 28 and tetrachlorinated PCB 52, particularly in fish/ fish products. The maximum concentration reported in fish/fish products exceeded 1 000 000 ng/kg expressed on a wet weight basis, whereas the median concentrations remained very low for all six congeners (PCB 28: 0.73 ng/kg ww; PCB 52: 2.6 ng/kg ww; PCB 101: 8.9 ng/kg ww; PCB 138: 30 ng/kg ww; PCB 153: 50 ng/kg ww; PCB 180: 7.5 ng/kg ww). 269

Safety evaluation of certain food additives and contaminants Eightieth JECFA

Table 38 Concentrations of the six indicator PCBs and PCB 128 in specific foods Food category PCB 28 Eggs Fish Meat Milk PCB 52 Eggs Fish Meat Milk PCB 101 Eggs Fish Meat Milk PCB 138 Eggs Fish Meat Milk PCB 153 Eggs Fish Meat Milk PCB 180 Eggs Fish Meat Milk PCB 128 Eggs Fish Meat Milk n 1 086 7 146 2 488 6 510 1 069 7 045 2 506 6 513 1 085 7 137 2 463 6 481 1 090 7 144 2 521 6 531 1 092 7 147 2 521 6 534 1 092 7 145 2 517 6 528 2 356 13 15 n < LOD 392 2 119 1 473 736 484 1 875 1 495 716 508 1 762 1 478 741 312 1 543 1 013 524 294 1 463 993 429 310 1 777 1 074 571 1 2 4 3 Concentration range (ng/kg ww) ND–6 800 ND–103 000 ND–16 100 ND–5 250 ND–6 710 ND–610 000 ND–9 440 ND–4 600 ND–8 410 ND–1 200 000 ND–240 000 ND–1 850 ND–34 800 ND–482 000 ND–12 900 ND–5 740 ND–31 000 ND–812 000 ND–17 300 ND–16 900 ND–34 700 ND–280 000 ND–198 000 ND–4 210 ND–0.002 ND–6.63 ND–0.017 ND–1.29

WHO Food Additives Series No. 71-S1, 2016 270

LOD: limit of detection; n: number of samples; ND: not detected; ww: wet weight

11.9 Dietary exposure assessment

Estimates of dietary exposure were evaluated by the Committee, focusing on the six indicator PCBs, singly and in combination. Only chronic dietary exposure assessments were included in the evaluation. Dietary exposure estimates from the literature were reviewed. Both national and international estimates of dietary exposure were made by the Committee based on consumption and concentration data available from the GEMS/Food database. The concentration data submitted

Supplement 1: Non-dioxin-like polychlorinated biphenyls

(as shown at a major food group level in Table 38 above) were summarized by specific food type (e.g. for herring or cow’s milk) for use in the exposure calculations conducted by the Committee. Estimates of dietary exposure for the six indicator PCBs were also calculated individually for countries that provided concentration and consumption data, for both body burden modelling and risk characterization purposes. Exposures to PCB 128 were also estimated, as relevant toxicity data were available for this congener for risk characterization purposes. 11.9.1 National estimates of dietary exposure

Estimated national dietary exposures for the sum of the six indicator PCBs are shown in Table 39. National estimates of dietary exposure were also calculated for each of the six indicator PCBs individually. Exposures were highly variable, depending on the congener, concentration data, country and population subgroup assessed. As a result of a high proportion of “non-detects” (concentrations below the LOQ) across a wide range of food groups, high LOQs and similar LOQs across congeners, the exposures based on data from one country were excluded from the evaluation. The estimates are summarized in Table 40 for the remaining countries. Mean and high-percentile exposures, including LB and UB estimates, across all countries and population groups and individual congeners ranged between <1 and 4.7 ng/kg bw per day at the mean and 9.4 ng/kg bw per day at the high percentile. For PCB 128, dietary exposure estimates were available for only two countries (Finland and the United Kingdom), based on concentration data for limited food commodities. For this reason, the estimated dietary exposures were not used to determine the body burden modelled from external dose for PCB 128. An alternative approach was used, whereby the proportion that the concentration of PCB 128 represents of the six indicator PCBs considered individually was determined on the basis of data from GEMS/Food (the average was 16%). To represent approximate PCB 128 exposure, 16% of the average of the upper-bound exposures for each individual indicator PCB for both mean and high percentile for adults was used (mean 0.2 ng/kg bw per day, high percentile 0.4 ng/kg bw per day). Lower estimates of dietary exposure to the six indicator PCBs were determined where methods with lower LODs were used for the food analysis and low concentrations in foods were determined. This highlights the importance of using specific analytical techniques for this group of contaminants. Higher estimates of dietary exposure, particularly at the UB, were strongly influenced by the sensitivity of the analytical method and therefore the concentration assigned to “non-detects” for UB scenarios. 271

Safety evaluation of certain food additives and contaminants Eightieth JECFA

Table 39 Overall range of estimated national dietary exposures to the sum of the six indicator PCBs Source Literature Estimated by the Committee Population group Childrenc Adults Childrenc Adults LB–UB dietary exposuresa (ng/kg bw per day) Mean exposure High-percentile exposureb 3–24 12–87 1–18 8–45 <1–82 1–163 <1–25 <1–51

LB: lower bound; LOD: limit of detection; LOQ: limit of quantification; UB: upper bound a LB where “non-detects” were assigned a zero concentration; UB where “non-detects” were assigned either the LOD or LOQ. b Usually 90th or 95th percentile reported. The estimates calculated by the Committee were 90th percentiles. c Includes infants consuming a mixed diet; excludes infants solely breastfed or formula fed.

Table 40 Summary of the range of estimated dietary exposures for individual indicator PCBs for adults and childrena from all countries assessed NDL-PCB 28 52 101 138 153 180 Population groupb Children (infants to adolescents) Adults Children (infants to adolescents) Adults Children (infants to adolescents) Adults Children (infants to adolescents) Adults Children (infants to adolescents) Adults Children (infants to adolescents) Adults Estimated dietary exposure (ng/kg bw per day) Mean (LB–UB) High percentile (LB–UB) <1–2.8 <1–5.7 <1 –<1 <1–1.1 <1–2.6 <1–5.3 <1–<1 <1–1.1 <1–3.4 <1–6.7 <1–1.2 <1–2.5 <1–4.7 <1–9.4 <1–1.8 <1–3.7 <1–3.5 <1–7.1 <1–2.2 <1–4.3 <1–2.7 <1–5.4 <1–<1 <1–1.5

WHO Food Additives Series No. 71-S1, 2016 272

a

b

LB: lower bound; UB: upper bound Includes infants consuming a mixed diet; does not include infants exclusively breastfed or formula fed. Where defined for the consumption survey, infants are <1 year, toddlers 1 to <3 years, children 3 to <10 years, adolescents 10–<18 years, adults 18+ years.

Despite variations in methodologies used to estimate dietary exposure, the majority of national estimates of dietary exposure to the NDL-PCBs assessed were in the same ranges, with slightly more variation in the estimates of UB levels of dietary exposure than in the LB estimates. The main contributor to dietary exposure to the sum of the six indicator PCBs at the national level was fish and seafood, followed by meat or meat products. Milk and dairy products also contributed for some populations and population subgroups, particularly children. The differences in the main contributors

Supplement 1: Non-dioxin-like polychlorinated biphenyls

between countries depended on the importance of the food in the countries’ diets as well as the concentration of the NDL-PCBs in the food used in the estimate. 11.9.2 International estimates of dietary exposure

Estimates of international mean dietary exposure per capita for the sum of the six indicator PCBs were calculated by the Committee using concentration data submitted to the GEMS/Food contaminants database and consumption data from the GEMS/Food cluster diets. The range of estimated dietary exposure to the sum of the six indicator PCBs between clusters for the LB scenario was 1–60 ng/kg bw per day. For UB exposures, the range between clusters was 2–83 ng/kg bw per day. Fish was a major contributor to dietary exposure across the majority of the clusters, contributing up to about 90% of dietary exposure for one cluster. This was due to the higher concentration of NDL-PCBs in fish and/or the higher consumption of fish for the cluster. 11.9.3 Dietary exposure of infants

Mean dietary exposure of breastfed infants to the sum of the six indicator PCBs was reported for 11 European countries by EFSA (2005) to be 1200 ng/kg bw per day. The most recent review reported a mean dietary exposure for breastfed infants of 1600 ng/kg bw per day, with a wide range of means from around 200 to 7000 ng/kg bw per day (IARC, 2015). Estimated dietary exposures of breastfed infants to NDL-PCBs are up to 2 orders of magnitude higher than those for the rest of the population. Biomonitoring data indicate that breastfed infants have higher body burdens of NDL-PCBs compared with formula-fed infants. 11.9.4 Contribution of individual congeners to total exposures from all sources

For the sum of the six indicator PCBs, the contribution of each of the individual congeners differs between countries and population subgroups. However, for both dietary exposure and body burden estimates (which also take into consideration kinetics and half-lives), the main contributor is PCB 153 (41%), followed by PCB 138 (28%), PCB 180 (14%), PCB 101 (8%) and PCB 28 (6%), with the lowest contribution from PCB 52 (3%). With the exception of those individuals with high occupational exposure, total exposures from all sources are likely to be only slightly higher than those predicted from the diet alone. Dietary exposures to NDL-PCBs have been decreasing over time, as indicated in studies from a number of countries, owing primarily to the phasing out of the manufacture and use of PCBs.

273

Safety evaluation of certain food additives and contaminants Eightieth JECFA

11.10 Estimation of margins of exposure

For non-genotoxic substances, the Committee would normally develop healthbased guidance values using the most sensitive adverse effect in the most sensitive species as a point of departure. The Committee therefore considered whether the toxicological information available for the six indicator PCBs (PCB 28, PCB 52, PCB 101, PCB 138, PCB 153, PCB 180) and PCB 128 was sufficient and appropriate for such an approach. It was noted that there are a large number of in vivo and in vitro studies available with respect to possible hepatotoxicity, thyroid toxicity, and neurodevelopmental or neurotoxic effects, but that there is a general lack of in vivo toxicity data for two of the indicator PCBs (PCB 101 and PCB 138). In addition, there is not enough information on relative potencies for each congener with respect to receptor interactions and the downstream consequences. The Committee considered whether it would be possible to undertake a group evaluation for NDL-PCBs using the available information for the indicator congeners. Such an approach should be based on internal rather than external dose because of the species differences in half-lives for these congeners. Owing to the lack of relevant toxicological data for two congeners (PCB 101 and PCB 138), the Committee decided not to undertake such a group evaluation. The Committee further concluded that none of the available short-term toxicity studies on four of the six indicator PCBs and PCB 128 were suitable for the derivation of health-based guidance values (e.g. provisional tolerable monthly intakes) or assessment of their relative potency compared with a reference compound, such as PCB 153. This conclusion was based on the lack of clear dose–response relationships, doubts about the toxicological relevance of the observed minimal effects on the liver and thyroid and the limited experimental time periods of most of the studies (28 or 90 days). The Committee considered whether the 2-year NTP study on PCB 153 might form a basis for deriving a BMDL that could be used as a point of departure for a health-based guidance value for that congener. From a human exposure and risk assessment point of view, PCB 153 is highly relevant, as it represents up to 40% of the six indicator PCBs that are present in the diet or in human milk. The most sensitive end-point in the 2-year study was hepatocyte hypertrophy, observed at and above 7 μg/kg bw per day. After critical evaluation of these results, the Committee concluded that the hepatocyte hypertrophy should not be modelled to derive a BMDL, as the end-point may not be toxicologically relevant. Hence, data from the longterm NTP study with PCB 153 were not considered suitable to derive a healthbased guidance value. Therefore, a comparative approach using the minimal effect doses from the available studies was developed in order to estimate MOEs to provide guidance on human health risk.

274

WHO Food Additives Series No. 71-S1, 2016

Supplement 1: Non-dioxin-like polychlorinated biphenyls

The available rat toxicological data on individual congeners showed that minimal changes in liver and thyroid histopathology were evident from the lowest doses tested of 2.8–7 µg/kg bw per day in the 90-day studies (PCB 28, PCB 128, PCB 153) and 3 mg/kg bw total dose in the 28-day studies (PCB 52, PCB 180) and were similar across the short-term and long-term toxicity studies. Bearing in mind that, with the exception of PCB 153, the available studies on individual NDL-PCB congeners were of relatively short duration (28 or 90 days), the Committee decided to take the lower end of the range of test doses used for each congener at which these minimal changes occurred as a conservative point of departure for estimating MOEs. Given the major difference in dosing regimens between the 28-day and 90-day studies and the bioaccumulative nature of PCBs, MOEs have been estimated on the basis of both external dose and internal dose. The internal dose MOEs based on amounts present in fat are considered the most appropriate comparison, particularly because they also eliminate interspecies differences in toxicokinetics. The MOEs obtained for adults are shown in Table 41. The MOE comparisons for internal dose are based on the range of reported mean values for human milk expressed on a fat basis, which should reflect both fetal and adult (maternal) body burdens. The biomonitoring data on concentrations in human adipose tissue were not used because they represent far fewer, non-random samples derived from postmortem tissues from non-homogenous populations.

275

Safety evaluation of certain food additives and contaminants Eightieth JECFA

Table 41 Estimated MOEs for adults from repeated-dose studies on individual NDL-PCB congeners in rats NDL-PCB congener Minimal effect Study duration dose expressed as Reference external dose Mode of administration (µg/kg bw per day) 28 2.8 90 days Chu et al. (1996a) Diet 52 3.0d 28 days Unpublished data provided to WHO by ATHON project study authors Gavage 101 NA 128 4.2 90 days Lecavalier et al. (1997) Diet 138 NA 153 7e 90 days Chu et al. (1996b) Diet 180 3.0d 28 days Viluksela et al. (2014) Gavage Minimal effect dose expressed as body burdena (mg/kg bw) 0.07 External dose MOEb 2 500–5 600 Body burden MOE Body burden MOE (based on human (modelled from milk)c external dose)c 44–580 41–1 400

NA

97 000–210 000

NA

NA

NA 0.07

NA 11 000–21 000

NA 88–1 700

NA NA

NA 2.0

NA 1 600–3 100

NA 75–2 900

NA 130–3 000

1.6

71 000–150 000

81–5 000

380–6 400

276

WHO Food Additives Series No. 71-S1, 2016

MOE: margin of exposure; NA: not available; UB: upper bound a Body burden based on reported concentration of NDL-PCB congener in adipose tissue; 350 g rat with 10% lipid. b For MOEs expressed as a range, the lower end of the range relates to UB adult high-percentile exposure, and the higher end of the range relates to UB adult mean exposure (see Table 40). The dietary exposure estimate for PCB 128 is based on it making, on average, a contribution equal to 16% of the total exposure to the six indicator PCBs. c See Table 37. d Total dose (mg/kg bw) administered over the whole study. e Dose adjusted to 7 µg/kg bw per day from 10 µg/kg bw per day to take account of 5 days/week dosing.

12. Evaluation The body burden MOEs for adults derived from the range of reported mean human milk concentrations range from 4.5 to 5000. The Committee noted that for some of the NDL-PCBs, these body burden estimates were developed from experimental studies with a less than chronic duration of exposure. For PCB 153, the lower end of the body burden MOE range based on human milk was 4.5 when derived from the short-term study, whereas in the long-term study it was 75, which is a 16-fold difference. The lower end of the range of body burden MOEs modelled from external dose also gave a 16-fold difference between longand short-term studies on PCB 153. Thus, in the long-term study, similar hepatic effects to those seen in the short-term study were observed only at an internal dose that was substantially higher. Use of the MOE value from the long-term study on PCB 153 would give lower-end MOEs for all congeners in the range 44–88 for adults. MOEs for breastfed infants, which may have a body burden up to 2-fold higher than that for adults, would be approximately half of the adult values. The MOEs for children would be expected to be intermediate between those for adults and those for breastfed infants, owing to the initial contribution from breastfeeding and the subsequent lower dietary contribution compared with human milk. As the MOEs are based on minimal effect doses, they were considered to be adequate and to give some assurance that dietary exposures to NDL-PCBs are unlikely to be of health concern for adults and children, based on the available data. For breastfed infants, the MOEs would be expected to be lower. However, based on present knowledge, the benefits of breastfeeding (WHO, 2015) are considered to outweigh the possible disadvantages that may be associated with the presence of NDL-PCBs in breast milk. The Committee recognized that there are similarities in some of the reported effects for NDL-PCBs and that, ideally, risk estimates for combined exposure are desirable. The Committee concluded that this cannot be done on the basis of currently available data, but noted that the points of departure selected for derivation of the MOEs were particularly conservative, as they were based on effects on liver and thyroid that were not of clear toxicological significance, the changes were minimal and the lowest doses at which they were seen were used for the points of departure, combined with UB estimates of body burden. 12.1 Recommendations

A more complete toxicological database, including mechanistic studies on both parent congeners and hydroxy metabolites, would have allowed a more definitive assessment. The Committee recommended that further toxicological studies 277

Safety evaluation of certain food additives and contaminants Eightieth JECFA

should be done, particularly in vivo studies for those congeners, such as PCB 101 and PCB 138, that contribute significantly to dietary exposure and human body burden. There were some limitations with the exposure assessment where there was a limited range of countries and/or foods for which concentration data were provided. This meant that some assumptions needed to be made for estimating dietary exposures for national and international assessments. To ensure better estimates of dietary exposure with a lower degree of uncertainty, the Committee recommends the following: ■■ As the risk characterization is driven mainly by dietary exposure estimates from European countries, information from a broader range of countries would be desirable to provide a more globally representative conclusion. ■■ More countries should submit concentration data to the GEMS/Food contaminants database (with as specific details as possible, including unique sample identifiers, specific food details, lower LODs, form of the food, etc.). ■■ National food consumption data should also be submitted by more countries to CIFOCOss to allow a broader range of country-specific exposure assessments to be undertaken in the future. ■■ For analytical surveys on NDL-PCBs, it is important to ensure the generation of occurrence data for congeners beyond the six indicator PCBs for a broad range of foods and to include all key congeners and sources of dietary exposure to NDL-PCBs. ■■ More concentration data in infant formula as well as data from a broader range of countries would assist in determining more reliable estimates of dietary exposure for formula-fed infants. WHO Food Additives Series No. 71-S1, 2016 278

13. References Abdelouahab N, Mergler D, Takser L, Vanier C, St-Jean M, Baldwin M et al. (2008). Gender differences in the effects of organochlorines, mercury, and lead on thyroid hormone levels in lakeside communities of Quebec (Canada). Environ Res. 107:380–92. Abraham K, Hille A, Ende M, Helge H (1994). Intake and fecal excretion of PCDDs, PCDFs, HCB and PCBs (138, 153, 180) in a breast-fed and a formula-fed infant. Chemosphere. 29:2279–86. Abrha Y, Raghavan D (2000). Polychlorinated biphenyl (PCB) recovery from spiked organic matrix using accelerated solvent extraction (ASE) and Soxhlet extraction. J Hazard Mater. 80:147–57. Agudo A, Goñi F, Etxeandia A, Vives A, Millán E, López R et al. (2009). Polychlorinated biphenyls in Spanish adults: determinants of serum concentrations. Environ Res. 109:620–8. Ahlborg UG, Hanberg A, Kenne K (1992). Risk assessment of polychlorinated biphenyls (PCBs). Stockholm: Karolinska Institutet, Institute of Environmental Medicine; Nordic Council of Ministers (Nord 1992:26; http://ki.se/sites/default/files/nordpcb-92.pdf, accessed 3 January 2016). Ahlborg UG, Lipworth L, Titusernstoff L, Hsieh CC, Hanberg A, Baron J et al. (1995). Organochlorine compounds in relation to breast cancer, endometrial cancer, and endometriosis – an assessment of the biological and epidemiological evidence. Crit Rev Toxicol. 25:463–531. Ahrens W, Mambetova C, Bourdon-Raverdy N, Llopis-González A, Guénel P, Hardell L et al. (2007). Occupational exposure to endocrine-disrupting compounds and biliary tract cancer among men. Scand J Work Environ Health. 33:387–96. Airaksinen R, Rantakokko P, Eriksson JG, Blomstedt P, Kajantie E, Kiviranta H (2011). Association between type 2 diabetes and exposure to persistent organic pollutants. Diabetes Care. 34:1972–9. Al-Anati L, Högberg J, Stenius U (2009). Non-dioxin-like-PCBs phosphorylate Mdm2 at Ser166 and attenuate the p53 response in HepG2 cells. Chem Biol Interact. 182:191–8. Albro PW, Fishbein L (1972). Intestinal absorption of polychlorinated biphenyls in rats. Bull Environ Contam Toxicol. 8:26–31. Albro PW, Corbett JT, Schroeder JL (1981). Quantitative characterization of polychlorinated biphenyl mixtures (Aroclors® 1248, 1254 and 1269) by gas chromatography using capillary columns. J Chromatogr A. 205:103–11. Alcock RE, Sweetman AJ, Juan CY, Jones KC (2000). A generic model of human lifetime exposure to persistent organic contaminants: development and application to PCB-101. Environ Pollut. 110:253– 65. Al-Salman F, Plant N (2012). Non-coplanar polychlorinated biphenyls (PCBs) are direct agonists for the human pregnane-X receptor and constitutive androstane receptor, and activate target gene expression in a tissue-specific manner. Toxicol Appl Pharmacol. 263:7–13. Álvarez-Pedrerol M, Ribas-Fitó N, Torrent M, Carrizo D, Grimalt JO, Sunyer J (2008). Effects of PCBs, p,p′-DDT, p,p′-DDE, HCB and beta-HCH on thyroid function in preschool children. Occup Environ Med. 65:452–7. AMAP (2009). AMAP assessment 2009: human health in the Arctic. Oslo: Arctic Monitoring and Assessment Programme (http://www.amap.no/documents/doc/amap-assessment-2009-humanhealth-in-the-arctic/98, accessed 1 January 2016). 279

Safety evaluation of certain food additives and contaminants Eightieth JECFA

American Diabetes Association (1997). Report of the Expert Committee on the Diagnosis and Classification of Diabetes Mellitus. Diabetes Care. 20:1183–97. Aminov Z, Haase RF, Pavuk M, Carpenter DO; Anniston Environmental Health Research Consortium (2013). Analysis of the effects of exposure to polychlorinated biphenyls and chlorinated pesticides on serum lipid levels in residents of Anniston, Alabama. Environ Health. 12:108. Amodio E, Turci R, Massenti MF, Di Gaudio F, Minoia C, Vitale F et al. (2012). Serum concentrations of persistent organic pollutants (POPs) in the inhabitants of a Sicilian city. Chemosphere. 89:970–4. Amzal B, Julin B, Vahter M, Wolk A, Johanson G, Akesson A (2009). Population toxicokinetic modeling of cadmium for health risk assessment. Environ Health Perspect. 117:1293–301. Angus WG, Mousa MA, Vargas VM, Quensen JF, Boyd SA, Contreras ML (1997). Inhibition of L-aromatic amino acid decarboxylase by polychlorinated biphenyls. Neurotoxicology. 18:857–67. Antunes-Fernandes EC, Bovee TF, Daamen FE, Helsdingen RJ, van den Berg M, van Duursen MB (2011). Some OH-PCBs are more potent inhibitors of aromatase activity and (anti-) glucocorticoids than nondioxin like (NDL)-PCBs and MeSO2-PCBs. Toxicol Lett. 206:158–65. AOAC International (1983). AOAC Official Method 983.21. Organochlorine pesticide and polychlorinated biphenyl residues in fish. J Assoc Off Anal Chem. 66:969. Aoki Y, Satoh K, Sato K, Suzuki KT (1992). Induction of glutathione S-transferase P-form in primary cultured rat liver parenchymal cells by co-planar polychlorinated biphenyl congeners. Biochem J. 281 (Pt 2):539–43. Apostoli P, Magoni M, Bergonzi R, Carasi S, Indelicato A, Scarcella C et al. (2005). Assessment of reference values for polychlorinated biphenyl concentration in human blood. Chemosphere. 61:413–21. Ariyoshi N, Oguri K, Koga N, Yoshimura H, Funae Y (1995). Metabolism of highly persistent PCB congener, 2,4,5,2′,4′,5′-hexachlorobiphenyl, by human CYP2B6. Biochem Biophys Res Commun. 212:455–60. Ariyoshi N, Iwasaki M, Kato H, Tsusaki S, Hamamura M, Ichiki T et al. (1998). Highly toxic coplanar PCB126 reduces liver peroxisomal enzyme activities in rats. Environ Toxicol Pharmacol. 5:219–25. Arnich N, Tard A, Leblanc J-C, Le Bizec B, Narbonne JF, Maximilien R (2009). Dietary intake of nondioxin-like PCBs (NDL-PCBs) in France, impact of maximum levels in some foodstuffs. Reg Toxicol Pharmacol. 54:287–93. Aronson KJ, Wilson JW, Hamel M, Diarsvitri W, Fan W, Woolcott C et al. (2010). Plasma organochlorine levels and prostate cancer risk. J Expo Sci Environ Epidemiol. 20:434–45. Arrebola JP, Fernandez MF, Porta M, Rosell J, de la Ossa RM, Olea N et al. (2010). Multivariate models to predict human adipose tissue PCB concentrations in southern Spain. Environ Int. 36:705–13. Arrebola JP, Cuellar M, Claure E, Quevedo M, Antelo SR, Mutch E et al. (2012). Concentrations of organochlorine pesticides and polychlorinated biphenyls in human serum and adipose tissue from Bolivia. Environ Res. 112:40–7. Arrebola JP, Pumarega J, Gasull M, Fernandez MF, Martin-Olmedo P, Molina-Molina JM et al. (2013). Adipose tissue concentrations of persistent organic pollutants and prevalence of type 2 diabetes in adults from southern Spain. Environ Res. 122:31–7.

280

WHO Food Additives Series No. 71-S1, 2016

Supplement 1: Non-dioxin-like polychlorinated biphenyls

Artacho-Cordón F, Belhassen H, Arrebola JP, Ghali R, Amira D, Jiménez-Díaz I et al. (2015). Serum levels of persistent organic pollutants and predictors of exposure in Tunisian women. Sci Total Environ. 511:530–4. ATSDR (2000). Toxicological profile for polychlorinated biphenyls (PCBs). Atlanta (GA): United States Department of Health and Human Services, Public Health Service, Agency for Toxic Substances and Disease Registry (http://www.atsdr.cdc.gov/toxprofiles/tp17.pdf, accessed 1 January 2016). ATSDR (2011). Addendum to the 2000 toxicological profile for polychlorinated biphenyls. Atlanta (GA): United States Department of Health and Human Services, Public Health Service, Agency for Toxic Substances and Disease Registry (http://www.atsdr.cdc.gov/toxprofiles/pcbs_addendum.pdf, accessed 1 January 2016). Atuma SS, Hansson L, Johnsson H, Slorach S, De Wit CA, Lindstrom G (1998). Organochlorine pesticides, polychlorinated biphenyls and dioxins in human milk from Swedish mothers. Food Addit Contam. 15:142–50. Axelrad DA, Goodman S, Woodruff TJ (2009). PCB body burdens in US women of childbearing age 2001–2002: an evaluation of alternate summary metrics of NHANES data. Environ Res. 109:368–78. Aylward LL, Collins JJ, Bodner KM, Wilken M, Bodnar CM (2014). ‘‘Intrinsic’’ elimination rate and dietary intake estimates for selected indicator PCBs: toxicokinetic modeling using serial sampling data in US subjects, 2005–2010. Chemosphere. 110:48–52. Ayotte P, Dewailly E, Lambert GH, Perkins SL, Poon R, Feeley M et al. (2005). Biomarker measurements in a coastal fish-eating population environmentally exposed to organochlorines. Environ Health Perspect. 113:1318–24. Baars AJ, Bakker MI, Baumann RA, Boon PE, Freijer JI, Hoogenboom LAP et al. (2004). Dioxins, dioxinlike PCBs and non-dioxin-like PCBs in foodstuffs: occurrence and dietary intake in the Netherlands. Toxicol Lett. 151:51–61. Bachelet D, Truong T, Verner MA, Arveux P, Kerbrat P, Charlier C et al. (2011). Determinants of serum concentrations of 1,1-dichloro-2,2-bis(p-chlorophenyl)ethylene and polychlorinated biphenyls among French women in the CECILE study. Environ Res. 111:861–70. Bakker MI, Baars AJ, Baumann RA, Boon PE, Hoogerbrugge R (2003). Indicator PCBs in foodstuffs: occurrence and dietary intake in The Netherlands at the end of the 20th century. Bilthoven: National Institute for Public Health and the Environment (RIVM report 639102025/2003; RIKILT report 2003.014; http://www.rivm.nl/dsresource?objectid=rivmp:17768&type=org&disposition=inline&ns_nc=1, accessed 1 January 2016). Ballschmiter K, Zell M (1980). Analysis of polychlorinated biphenyls (PCB) by glass capillary gas chromatography – Composition of technical Aroclor- and Clophen-PCB mixtures. Z Anal Chem. 302:20–31. Ballschmiter K, Schäfer W, Buchert H (1987). Isomer-specific identification of PCB congeners in technical mixtures and environmental samples by HRGC-ECD and HRGC-MSD. Z Anal Chem. 326:253–7. Bandiera SM (2013). Cytochrome P450 enzymes as biomarkers of PCB exposure and modulators of toxicity. In: Robertson LW, Hansen LG, editors. PCBs: recent advances in environmental toxicology and health effects. Lexington (KY): University Press of Kentucky; 185–92. Baron CP, Børresen T, Jacobsen C (2005). UV treatment of fishmeal: a method to remove dioxins? J Agric Food Chem. 53:7091–7. 281

Safety evaluation of certain food additives and contaminants Eightieth JECFA

Baron CP, Børresen T, Jacobsen C (2007). Comparison of methods to reduce dioxin and polychlorinated biphenyls contents in fishmeal: extraction and enzymatic treatments. J Agric Food Chem. 55:1620–6. Barr DB, Weihe P, Davis MD, Needham LL, Grandjean P (2006). Serum polychlorinated biphenyl and organochlorine insecticide concentrations in a Faroese birth cohort. Chemosphere. 62:1167–82. Bates MN, Buckland SJ, Garrett N, Ellis H, Needham LL, Patterson DG Jr et al. (2004). Persistent organochlorines in the serum of the non-occupationally exposed New Zealand population. Chemosphere. 54:1431–43. Bayen S, Barlow P, Lee HK, Obbard JP (2005). Effect of cooking on the loss of persistent organic pollutants from salmon. J Toxicol Environ Health A. 68:253–65. Béchaux C, Zeilmaker M, Merlo M, Bokkers B, Crépet A (2014). An integrative risk assessment approach for persistent chemicals: a case study on dioxins, furans and dioxin-like PCBs in France. Regul Toxicol Pharmacol. 70:261–9. Becker K, Kaus S, Krause C, Lepom P, Schulz C, Seiwert M et al. (2002). German Environmental Survey 1998 (GerES III): environmental pollutants in blood of the German population. Int J Hyg Environ Health. 205:297–308. Becker K, Müssig-Zufika M, Conrad A, Lüdecke A, Schulz C, Seiwert M et al. (2008). German Environmental Survey for Children 2003/06 – GerES IV – Human biomonitoring. Levels of selected substances in blood and urine of children in Germany. Dessau-Roßlau: Federal Environment Agency (Umweltbundesamt) (http://www.umweltbundesamt.de/sites/default/files/medien/publikation/ long/3355.pdf, accessed 1 January 2016). Behnisch P, Engelhart A, Apfelbach R, Hagenmaier H (1997). Occurrence of non-ortho, mono-ortho and di-ortho substituted PCB congeners in polecats, stone martens and badgers from the state of BadenWurttemberg, Germany. Chemosphere. 34:2293–300. Behrens T, Kaerlev L, Cree I, Lutz JM, Afonso N, Eriksson M et al. (2010). Hormonal exposures and the risk of uveal melanoma. Cancer Causes Control. 21:1625–34. Bélanger MC, Dewailly E, Berthiaume L, Noël M, Bergeron J, Mirault ME et al. (2006). Dietary contaminants and oxidative stress in Inuit of Nunavik. Metabolism. 55:989–95. Bemis JC, Seegal RF (2004). PCB-induced inhibition of the vesicular monoamine transporter predicts reductions in synaptosomal dopamine content. Toxicol Sci. 80:288–95. Ben Hassine S, Hammami B, Ben Ameur W, El Megdiche Y, Barhoumi B, El Abidi R et al. (2014). Concentrations of organochlorine pesticides and polychlorinated biphenyls in human serum and their relation with age, gender, and BMI for the general population of Bizerte, Tunisia. Environ Sci Pollut Res Int. 21:6303–13. Berdié L, Grimalt JO (1998). Assessment of the sample handling procedures in a labor-saving method for the analysis of organochlorine compounds in a large number of fish samples. J Chromatogr A. 823:373–80. Berggrena P, Ishaq R, Zebuhr Y, Naf C, Bandh C, Broman D (1999). Patterns and levels of organochlorines (DDTs, PCBs, non-ortho PCBs and PCDD/Fs) in male harbour porpoises (Phocoena phocoena) from the Baltic Sea, the Kattegat-Skagerrak seas and the west coast of Norway. Mar Pollut Bull. 38:1070–84. Berghuis SA, Soechitram SD, Hitzert MM, Sauer PJ, Bos AF (2013). Prenatal exposure to polychlorinated biphenyls and their hydroxylated metabolites is associated with motor development of three-monthold infants. Neurotoxicology. 38:124–30.

282

WHO Food Additives Series No. 71-S1, 2016

Supplement 1: Non-dioxin-like polychlorinated biphenyls

Bergkvist C, Lignell S, Sand S, Aune M, Persson M, Håkansson H et al. (2010). A probabilistic approach for estimating infant exposure to environmental pollutants in human breast milk. J Environ Monit. 12:1029–36. Bergkvist C, Akesson A, Glynn A, Michaëlsson K, Rantakokko P, Kiviranta H et al. (2012). Validation of questionnaire-based long-term dietary exposure to polychlorinated biphenyls using biomarkers. Mol Nutr Food Res. 56:1748–54. Bergkvist C, Kippler M, Larsson SC, Berglund M, Glynn A, Wolk A et al. (2014). Dietary exposure to polychlorinated biphenyls is associated with increased risk of stroke in women. J Intern Med. 276:248– 59. Bergonzi R, De Palma G, Specchia C, Dinolfo M, Tomasi C, Frusca T et al. (2011). Persistent organochlorine compounds in fetal and maternal tissues: evaluation of their potential influence on several indicators of fetal growth and health. Sci Total Environ. 409:2888–93. Berntssen MHG, Maage A, Julshamn K, Oeye BE, Lundebye A (2011). Carry-over of dietary organochlorine pesticides, PCDD/Fs, PCBs, and brominated flame retardants to Atlantic salmon (Salmo salar L) fillets. Chemosphere. 83:95–103. Bertail P, Clémençon S, Tressou J (2010). Statistical analysis of a dynamic model for dietary contaminant exposure. J Biol Dyn. 4:212–34. Bertrand KA, Spiegelman D, Aster JC, Altshul LM, Korrick SA, Rodig SJ et al. (2010). Plasma organochlorine levels and risk of non-Hodgkin lymphoma in a cohort of men. Epidemiology. 21:172– 80. Bezdecny SA, Roth RA, Ganey PE (2005). Effects of 2,2′,4,4′-tetrachlorobiphenyl on granulocytic HL-60 cell function and expression of cyclooxygenase-2. Toxicol Sci. 84:328–34. Bhavsar SP, Jackson DA, Hayton A, Reiner EJ, Chen T, Bodnar J (2007). Are PCB levels in fish from the Canadian Great Lakes still declining? J Great Lakes Res. 33:592–605. Biegel L, Harris M, Davis D, Rosengren R, Safe L, Safe S (1989). 2,2′,4,4′,5,5′-hexachlorobiphenyl as a 2,3,7,8-tetrachlorodibenzo-p-dioxin antagonist in C57BL/6J mice. Toxicol Appl Pharmacol. 97:561–71. Bjermo H, Darnerud PO, Lignell S, Pearson M, Rantakokko P, Nalsen C et al. (2013). Fish intake and breastfeeding time are associated with serum concentrations of organochlorines in a Swedish population. Environ Int. 51:88–96. Blanchet-Letrouvé I, Zalouk-Vergnoux A, Venisseau A, Couderc M, Le Bizec B, Elie P et al. (2014). Dioxinlike, non-dioxin like PCB and PCDD/F contamination in European eel (Anguilla anguilla) from the Loire estuarine continuum: spatial and biological variabilities. Sci Total Environ. 472:562–71. Bloom MS, Vena JE, Swanson MK, Moysich KB, Olson JR (2005). Profiles of ortho-polychlorinated biphenyl congeners, dichlorodiphenyldichloroethylene, hexachlorobenzene, and Mirex among male Lake Ontario sportfish consumers: The New York State Angler Cohort Study. Environ Res. 97:178–94. Boix J, Cauli O, Felipo V (2010). Developmental exposure to polychlorinated biphenyls 52, 138 or 180 affects differentially learning or motor coordination in adult rats. Mechanisms involved. Neuroscience. 167:994–1003. Boix J, Cauli O, Leslie H, Felipo V (2011). Differential long-term effects of developmental exposure to polychlorinated biphenyls 52, 138 or 180 on motor activity and neurotransmission. Gender dependence and mechanisms involved. Neurochem Int. 58:69–77. 283

Safety evaluation of certain food additives and contaminants Eightieth JECFA

Bonde JP, Toft G, Rylander L, Rignell-Hydbom A, Giwercman A, Spano M et al. (2008). Fertility and markers of male reproductive function in Inuit and European populations spanning large contrasts in blood levels of persistent organochlorines. Environ Health Perspect. 116:269–77. Bonefeld-Jorgensen EC, Long M, Bossi R, Ayotte P, Asmund G, Krüger T et al. (2011). Perfluorinated compounds are related to breast cancer risk in Greenlandic Inuit: a case control study. Environ Health. 10:88. Boonyathumanondh R, Watanabe S, Laovakul W, Tabucanon M (1995). Development of a quantification methodology for polychlorinated biphenyls by using Kanechlor products as the secondary reference standard. Fresenius J Anal Chem. 352:261–7. Bordajandi LR, Abad E, Gonzalez MJ (2008). Occurrence of PCBs, PCDD/Fs, PBDEs and DDTs in Spanish breast milk: enantiomeric fraction of chiral PCBs. Chemosphere. 70:567–75. Bordajandi LR, Gomez G, Fernandez MA, Abad E, Rivera J, Gonzalez MJ (2003). Study on PCBs, PCDD/Fs, organochlorine pesticides, heavy metals and arsenic content in freshwater fish species from the River Turia (Spain). Chemosphere. 53:163–71. Bordajandi LR, Korytá P, de Boer J, González MJ (2005). Enantiomeric separation of chiral polychlorinated biphenyls on β-cyclodextrin capillary columns by means of heart-cut multidimensional gas chromatography and comprehensive two-dimensional gas chromatography. Application to food samples. J Sep Sci. 28:163–71. Borgå K, Wolkers H, Skaare JU, Hop H, Muir DCG, Gabrielsen GW (2005). Bioaccumulation of PCBs in Arctic seabirds: influence of dietary exposure and congener biotransformation. Environ Pollut. 134:397–409. Bouchard MF, Oulhote Y, Sagiv SK, Saint-Amour D, Weuve J (2014). Polychlorinated biphenyl exposures and cognition in older U.S. adults: NHANES (1999–2002). Environ Health Perspect. 122:73–8. Boumphrey RS, Harrad SJ, Jones JC, Osborn D (1993). Polychlorinated biphenyl congener patterns in tissues from a selection of British birds. Arch Environ Contam Toxicol. 25:346–52. Braun JM, Kalkbrenner AE, Just AC, Yolton K, Calafat AM, Sjödin A et al. (2014). Gestational exposure to endocrine-disrupting chemicals and reciprocal social, repetitive, and stereotypic behaviors in 4- and 5-year-old children: the HOME study. Environ Health Perspect. 122:513–20. WHO Food Additives Series No. 71-S1, 2016 Bräuner EV, Raaschou-Nielsen O, Gaudreau E, LeBlanc A, Tjønneland A, Overvad K et al. (2011). Predictors of polychlorinated biphenyl concentrations in adipose tissue in a general Danish population. Environ Sci Technol. 45:679–85. Bräuner EV, Sørensen M, Gaudreau E, LeBlanc A, Eriksen KT, Tjønneland A et al. (2012). A prospective study of organochlorines in adipose tissue and risk of non Hodgkin lymphoma. Environ Health Perspect. 120:105. Bräuner EV, Loft S, Wellejus A, Autrup H, Tjønneland A, Raaschou-Nielsen O (2014). Adipose tissue PCB levels and CYP1B1 and COMT genotypes in relation to breast cancer risk in postmenopausal Danish women. Int J Environ Health Res. 24:256–68. Bremnes NB, Broadwell SL, Becher G (2013). Interlaboratory comparison on POPs in food 2013. Fourteenth round of an interlaboratory study. Nydalen: Norwegian Institute of Public Health (http:// www.fhi.no/dokumenter/feda1da435.pdf, accessed 1 January 2016).

284

Supplement 1: Non-dioxin-like polychlorinated biphenyls

Brouwer A, Morse DC, Lans MC, Schuur AG, Murk AJ, Klasson-Wehler E et al. (1998). Interactions of persistent environmental organohalogens with the thyroid hormone system: mechanisms and possible consequences for animal and human health. Toxicol Ind Health. 14:59–84. Brown AP, Ganey PE (1995). Neutrophil degranulation and superoxide production induced by polychlorinated biphenyls are calcium dependent. Toxicol Appl Pharmacol. 131:198–205. Brown AP, Olivero-Verbel J, Holdan WL, Ganey PE (1998). Neutrophil activation by polychlorinated biphenyls: structure–activity relationship. Toxicol Sci. 46:308–16. Brown JF, Lawton RW (1984). Polychlorinated biphenyl (PCB) partitioning between adipose tissue and serum. Bull Environ Contam Toxicol. 33:277–80. Brown JF, Lawton RW, Ross MR, Feingold J, Wagner RE, Hamilton SB (1989). Persistence of PCB congeners in capacitor workers and Yusho patients. Chemosphere. 19:829–34. Brucker-Davis F, Wagner-Mahler K, Delattre I, Ducot B, Ferrari P, Bongain A et al. (2008). Cryptorchidism at birth in Nice area (France) is associated with higher prenatal exposure to PCBs and DDE, as assessed by colostrum concentrations. Hum Reprod. 23:1708–18. Bu Q, MacLeod M, Wong F, Toms L-ML, Mueller JF, Yu G (2015). Historical intake and elimination of polychlorinated biphenyls and organochlorine pesticides by the Australian population reconstructed from biomonitoring data. Environ Int. 74:82–8. Bühler F, Schmid P, Schlatter CH (1988). Kinetics of PCB elimination in man. Chemosphere. 17:1717–26. Burke RW, Diamondstone BI, Velapoldi RA, Menis O (1974). Mechanisms of the Liebermann Burchard and Zak color reactions for cholesterol. Clin Chem. 20:794–801. Burkhard LP, Lukasewycz MT (2008). Toxicity equivalency values for polychlorinated biphenyl mixtures. Environ Toxicol Chem. 27:529–34. Caironi M, Olivari L, Sampietro G, Mandelli G, Mosconi G (2005). [Preliminary results of a mortality study in 471 ex-exposed workers to PCBs.] G Ital Med Lav Ergon. 27:279–81 (in Italian). Camel V (2000). Microwave-assisted solvent extraction of environmental samples. TrAC Trends Anal Chem. 19:229–48. Cao LL, Yan CH, Yu XD, Tian Y, Zhao L, Liu JX et al. (2011). Relationship between serum concentrations of polychlorinated biphenyls and organochlorine pesticides and dietary habits of pregnant women in Shanghai. Sci Total Environ. 409:2997–3002. Cao Y, Winneke G, Wilhelm M, Wittsiepe J, Lemm F, Fürst P et al. (2008). Environmental exposure to dioxins and polychlorinated biphenyls reduce levels of gonadal hormones in newborns: results from the Duisburg cohort study. Int J Hyg Environ Health. 211:30–9. Capen CC (1997). Mechanistic data and risk assessment of selected toxic end points of the thyroid gland. Toxicol Pathol. 25:39–48. Carro N, Garcia I, Ignacio M, Mouteira A (2012). Optimization of Soxtec extraction for the determination of polychlorinated biphenyls (PCBs) in mussel and comparison with Soxhlet extraction, accelerated solvent extraction, and microwave assisted extraction. Anal Lett. 45:2161–75. Castoldi AF, Blandini F, Randine G, Samuele A, Manzo L, Coccini T (2006). Brain monoaminergic neurotransmission parameters in weanling rats after perinatal exposure to methylmercury and 2,2′,4,4′,5,5′-hexachlorobiphenyl (PCB153). Brain Res. 1112:91–8. 285

Safety evaluation of certain food additives and contaminants Eightieth JECFA

Cave M, Appana S, Patel M, Falkner KC, McClain CJ, Brock G (2010). Polychlorinated biphenyls, lead, and mercury are associated with liver disease in American adults: NHANES 2003–2004. Environ Health Perspect. 118:1735–42. CDC (2005). Third national report on human exposure to environmental chemicals. Atlanta (GA): United States Department of Health and Human Services, Centers for Disease Control and Prevention (http://stacks.cdc.gov/view/cdc/21811, accessed 1 January 2016). CDC (2009). Fourth national report on human exposure to environmental chemicals. Atlanta (GA): United States Department of Health and Human Services, Centers for Disease Control and Prevention (http://www.cdc.gov/exposurereport/pdf/fourthreport.pdf, accessed 1 January 2016). CDC (2013). Biomonitoring summaries: non-dioxin-like polychlorinated biphenyls. Atlanta (GA): United States Department of Health and Human Services, Centers for Disease Control and Prevention (http://www.cdc.gov/biomonitoring/NDL-PCBs_BiomonitoringSummary.html, accessed 1 January 2016). CDC (2015). Fourth national report on human exposure to environmental chemicals, updated tables, February 2015. Atlanta (GA): United States Department of Health and Human Services, Centers for Disease Control and Prevention (http://www.cdc.gov/biomonitoring/pdf/FourthReport_ UpdatedTables_Feb2015.pdf, accessed 1 January 2016). Cebecauer L, Radikova Z, Rovensky J, Koska J, Imrich R, Ksinantova L et al. (2009). Increased prevalence and coincidence of antinuclear and antithyroid antibodies in the population exposed to high levels of polychlorinated pollutants cocktail. Endocr Regul. 43:75–81. Cerná M, Malý M, Grabic R, Batáriová A, Smíd J, Benes B (2008). Serum concentrations of indicator PCB congeners in the Czech adult population. Chemosphere. 72:1124–31. Cerná M, Bencko V, Brabec M, Smíd J, Krsková A, Jech L (2010). Exposure assessment of breast-fed infants in the Czech Republic to indicator PCBs and selected chlorinated pesticides: area-related differences. Chemosphere. 78:60–8. Cerná M, Krsková A, Cejchanová M, Spĕváčková V (2012). Human biomonitoring in the Czech Republic: an overview. Int J Hyg Environ Health. 215:109–19. Cerveny D, Zlabek V, Velisek J, Turek J, Grabic R, Grabicova K et al. (2014). Contamination of fish in important fishing grounds of the Czech Republic. Ecotoxicol Environ Saf. 109:101–9. WHO Food Additives Series No. 71-S1, 2016 Chai X, Zeng S, Xie W (2013). Nuclear receptors PXR and CAR: implications for drug metabolism regulation, pharmacogenomics and beyond. Expert Opin Drug Metab Toxicol. 9:253–66. Chauhan KR, Kodavanti PR, McKinney JD (2000). Assessing the role of ortho-substitution on polychlorinated biphenyl binding to transthyretin, a thyroxine transport protein. Toxicol Appl Pharmacol. 162:10–21. Chen PH, Luo ML, Wong CK, Chen CJ (1982). Comparative rates of elimination of some individual polychlorinated biphenyls from the blood of PCB-poisoned patients in Taiwan. Food Chem Toxicol. 20:417–25. Chevrier J, Eskenazi B, Bradman A, Fenster L, Barr DB (2007). Associations between prenatal exposure to polychlorinated biphenyls and neonatal thyroid-stimulating hormone levels in a Mexican-American population, Salinas Valley, California. Environ Health Perspect. 115:1490–6.

286

Supplement 1: Non-dioxin-like polychlorinated biphenyls

Chevrier J, Eskenazi B, Holland N, Bradman A, Barr DB (2008). Effects of exposure to polychlorinated biphenyls and organochlorine pesticides on thyroid function during pregnancy. Am J Epidemiol. 168:298–310. Chewe D, Creaser CS, Foxall CD, Lovett AA (1997). Validation of a congener specific method for ortho and non-ortho substituted polychlorinated biphenyls in fruit and vegetable samples. Chemosphere. 35:1399–407. Chia VM, Li Y, Quraishi SM, Graubard BI, Figueroa JD, Weber JP et al. (2010). Effect modification of endocrine disruptors and testicular germ cell tumour risk by hormone-metabolizing genes. Int J Androl. 33:588–96. Cho MR, Shin JY, Hwang JH, Jacobs DR Jr, Kim SY, Lee DH (2011). Associations of fat mass and lean mass with bone mineral density differ by levels of persistent organic pollutants: National Health and Nutrition Examination Survey 1999–2004. Chemosphere. 82:1268–76. Choksi NY, Kodavanti PR, Tilson HA, Booth RA (1997). Effects of polychlorinated biphenyls (PCBs) on brain tyrosine hydroxylase activity and dopamine synthesis in rats. Fundam Appl Toxicol. 39:76–80. Chou CC, Chen YN, Li CS (2004). Congener-specific polychlorinated biphenyls in cetaceans from Taiwan waters. Arch Environ Contam Toxicol. 47:551–60. Chovancová J, Drobná B, Fabišiková A, Conka K, Wimmerová S, Pavuk M (2014). Polychlorinated biphenyls and selected organochlorine pesticides in serum of Slovak population from industrial and non-industrial areas. Environ Monit Assess. 186:7643–53. Chu I, Villeneuve DC, Yagminas A, Lecavalier P, Poon R, Håkansson H et al. (1996a). Toxicity of 2,4,4′-trichlorobiphenyl in rats following 90-day dietary exposure. J Toxicol Environ Health. 49:301– 18. Chu I, Villeneuve DC, Yagminas A, Lecavalier P, Poon R, Feeley M et al. (1996b). Toxicity of 2,2′,4,4′,5,5′-hexachlorobiphenyl in rats: effects following 90-day oral exposure. J Appl Toxicol. 16:121–8. Chu S, Covaci A, Schepens P (2003). Levels and chiral signatures of persistent organochlorine pollutants in human tissues from Belgium. Environ Res. 93:167–76. Cimenci O, Vandevijvere S, Goscinny S, Van Den Bergh M, Hanot V, Vinkx C et al. (2013). Dietary exposure of the Belgian adult population to non-dioxin-like PCBs. Food Chem Toxicol. 59:670–9. Cirillo T, Viscardi V, Farina A, Fasano E, Del Prete U, Amodio Cocchieri R (2008). Occurrence of polychlorobiphenyls in buffalo mozzarella cheese from Campania (Italy). J Prev Med Hyg. 49:136–41. Cirillo T, Viscardi V, Fasano E, Farina A, Amodio-Cocchieri R (2009). Polychlorinated biphenyls, organochlorine pesticides, and polycyclic aromatic hydrocarbons in wild, farmed, and frozen marine seafood marketed in Campania, Italy. J Food Prot. 72:1677–85. Coburn CG, Currás-Collazo MC, Kodavanti PR (2007). Polybrominated diphenyl ethers and orthosubstituted polychlorinated biphenyls as neuroendocrine disruptors of vasopressin release: effects during physiological activation in vitro and structure–activity relationships. Toxicol Sci. 98:178–86. Coburn CG, Gillard ER, Currás-Collazo MC (2005). Dietary exposure to Aroclor 1254 alters central and peripheral vasopressin release in response to dehydration in the rat. Toxicol Sci. 84:149–56.

287

Safety evaluation of certain food additives and contaminants Eightieth JECFA

Coburn CG, Watson-Siriboe A, Hou B, Cheetham C, Gillard ER, Lin L et al. (2015). Permanently compromised NADPH-diaphorase activity within the osmotically activated supraoptic nucleus after in utero but not adult exposure to Aroclor 1254. Neurotoxicology. 47:37–46. Cocchi D, Tulipano G, Colciago A, Sibilia V, Pagani F, Vigano D et al. (2009). Chronic treatment with polychlorinated biphenyls (PCB) during pregnancy and lactation in the rat. Part 1: Effects on somatic growth, growth hormone–axis activity and bone mass in the offspring. Toxicol Appl Pharmacol. 237:127–36. Coccini T, Randine G, Castoldi AF, Grandjean P, Ostendorp G, Heinzow B et al. (2006). Effects of developmental co-exposure to methylmercury and 2,2′,4,4′,5,5′-hexachlorobiphenyl (PCB153) on cholinergic muscarinic receptors in rat brain. Neurotoxicology. 27:468–77. Coccini T, Roda E, Castoldi AF, Goldoni M, Poli D, Bernocchi G et al. (2007). Perinatal co-exposure to methylmercury and PCB153 or PCB126 in rats alters the cerebral cholinergic muscarinic receptors at weaning and puberty. Toxicology. 238:34–48. Coccini T, Manzo L, Debes F, Steuerwald U, Weihe P, Grandjean P (2009). No changes in lymphocyte muscarinic receptors and platelet monoamine oxidase-B examined as surrogate central nervous system biomarkers in a Faroese children cohort prenatally exposed to methylmercury and polychlorinated biphenyls. Biomarkers. 14:67–76. Coccini T, Roda E, Castoldi AF, Poli D, Goldoni M, Vettori MV et al. (2011). Developmental exposure to methylmercury and 2,2′,4,4′,5,5′-hexachlorobiphenyl (PCB153) affects cerebral dopamine D1-like and D2-like receptors of weanling and pubertal rats. Arch Toxicol. 85:1281–94. Cocco P, Brennan P, Ibba A, de Sanjosé Llongueras S, Maynadié M, Nieters A et al. (2008). Plasma polychlorobiphenyl and organochlorine pesticide level and risk of major lymphoma subtypes. Occup Environ Med. 65:132–40. Codru N, Schymura MJ, Negoita S, Rej R, Carpenter DO; Akwesasne Task Force on the Environment (2007). Diabetes in relation to serum levels of polychlorinated biphenyls and chlorinated pesticides in adult Native Americans. Environ Health Perspect. 115:1442–7. Cohn BA, Cirillo PM, Sholtz RI, Ferrara A, Park JS, Schwingl PJ (2011). Polychlorinated biphenyl (PCB) exposure in mothers and time to pregnancy in daughters. Reprod Toxicol. 31:290–6. WHO Food Additives Series No. 71-S1, 2016 Cohn BA, Terry MB, Plumb M, Cirillo PM (2012). Exposure to polychlorinated biphenyl (PCB) congeners measured shortly after giving birth and subsequent risk of maternal breast cancer before age 50. Breast Cancer Res Treat. 136:267–75. Çok I, Donmez MK, Uner M, Demirkaya E, Henkelmann B, Shen H et al. (2009). Polychlorinated dibenzop-dioxins, dibenzofurans and polychlorinated biphenyls levels in human breast milk from different regions of Turkey. Chemosphere. 76:1563–71. Çok I, Durmaz TC, Durmaz E, Satiroglu MH, Kabukcu C (2010). Determination of organochlorine pesticide and polychlorinated biphenyl levels in adipose tissue of infertile men. Environ Monit Assess.162:301–9. Çok I, Yelken C, Durmaz E, Uner M, Sever B, Satir F (2011). Polychlorinated biphenyl and organochlorine pesticide levels in human breast milk from the Mediterranean city Antalya, Turkey. Bull Environ Contam Toxicol. 86:423–7.

288

Supplement 1: Non-dioxin-like polychlorinated biphenyls

Çok I, Mazmanci B, Mazmanci MA, Turgut C, Henkelmann B, Schramm KW (2012). Analysis of human milk to assess exposure to PAHs, PCBs and organochlorine pesticides in the vicinity Mediterranean city Mersin, Turkey. Environ Int. 40:63–9. Colciago A, Casati L, Mornati O, Vergoni AV, Santagostino A, Celotti F et al. (2009). Chronic treatment with polychlorinated biphenyls (PCB) during pregnancy and lactation in the rat. Part 2: Effects on reproductive parameters, on sex behavior, on memory retention and on hypothalamic expression of aromatase and 5α-reductases in the offspring. Toxicol Appl Pharmacol. 239:46–54. Colt JS, Rothman N, Severson RK, Hartge P, Cerhan JR, Chatterjee N et al. (2009). Organochlorine exposure, immune gene variation, and risk of non-Hodgkin lymphoma. Blood. 113:1899–905. Connor K, Safe S, Jefcoate CR, Larsen M (1995). Structure-dependent induction of CYP2B by polychlorinated biphenyl congeners in female Sprague-Dawley rats. Biochem Pharmacol. 50:1913–20. Correia Carreira S, Cartwright L, Mathiesen L, Knudsen LE, Saunders M (2011). Studying placental transfer of highly purified non-dioxin-like PCBs in two models of the placental barrier. Placenta. 32:283–91. Costabeber I, Emanuelli T (2003). Influence of alimentary habits, age and occupation on polychlorinated biphenyl levels in adipose tissue. Food Chem Toxicol. 41:73–80. Costera A, Feidt C, Marchand P, Le Bizec B, Rychen G (2006). PCDD/F and PCB transfer to milk in goats exposed to a long-term intake of contaminated hay. Chemosphere. 64:650–7. Costopoulou D, Vassiliadou I, Papadopoulos A, Makropoulos V, Leondiadis L (2006). Levels of dioxins, furans and PCBs in human serum and milk of people living in Greece. Chemosphere. 65:1462–9. Covaci A, de Boer J, Ryan JJ, Voorspoels S, Schepens P (2002). Distribution of organobrominated and organochlorinated contaminants in Belgian human adipose tissue. Environ Res. 88:210–8. Covaci A, Voorspoels S, Roosens L, Jacobs W, Blust R, Neels H (2008). Polybrominated diphenyl ethers (PBDEs) and polychlorinated biphenyls (PCBs) in human liver and adipose tissue samples from Belgium. Chemosphere. 73:170–5. Craft ES, DeVito MJ, Crofton KM (2002). Comparative responsiveness of hypothyroxinemia and hepatic enzyme induction in Long-Evans rats versus C57BL/6J mice exposed to TCDD-like and phenobarbitallike polychlorinated biphenyl congeners. Toxicol Sci. 68:372–80. Criado MR, Fernandez DH, Pereiro IR, Torrijos RC (2004). Application of matrix solid-phase dispersion to the determination of polychlorinated biphenyls in fat by gas chromatography with electron-capture and mass spectrometric detection. J Chromatogr A. 1056:187–94. Croes K, Colles A, Koppen G, Govarts E, Bruckers L, Van de Mieroop E et al. (2012). Persistent organic pollutants (POPs) in human milk: a biomonitoring study in rural areas of Flanders (Belgium). Chemosphere. 89:988–94. Crofton KM, Rice DC (1999). Low-frequency hearing loss following perinatal exposure to 3,3′,4,4′,5-pentachlorobiphenyl (PCB 126) in rats. Neurotoxicol Teratol. 21:299–301. Crofton KM, Ding D, Padich R, Taylor M, Henderson D (2000). Hearing loss following exposure during development to polychlorinated biphenyls: a cochlear site of action. Hear Res. 144:196–204. Dahl P, Lindström G, Wiberg K, Rappe C (1995). Absorption of polychlorinated biphenyls, dibenzo-pdioxins and dibenzofurans by breast-fed infants. Chemosphere. 30:2297–306.

289

Safety evaluation of certain food additives and contaminants Eightieth JECFA

Daidoji T, Gozu K, Iwano H, Inoue H, Yokota H (2005). UDP-glucuronosyltransferase isoforms catalyzing glucuronidation of hydroxy-polychlorinated biphenyls in rat. Drug Metab Dispos. 33:1466–76. Dallaire F, Dewailly E, Vézina C, Muckle G, Weber JP, Bruneau S et al. (2006). Effect of prenatal exposure to polychlorinated biphenyls on incidence of acute respiratory infections in preschool Inuit children. Environ Health Perspect. 114:1301–5. Dallaire R, Dewailly E, Pereg D, Dery S, Ayotte P (2009). Thyroid function and plasma concentrations of polyhalogenated compounds in Inuit adults. Environ Health Perspect. 117:1380–6. Dallaire R, Dewailly E, Ayotte P, Forget-Dubois N, Jacobson SW, Jacobson JL et al. (2014). Growth in Inuit children exposed to polychlorinated biphenyls and lead during fetal development and childhood. Environ Res. 134:17–23. Danielsson C, Wiberg K, Korytar P, Bergek S, Brinkman UAT, Haglund P (2005). Trace analysis of polychlorinated dibenzo-p-dioxins, dibenzofurans and WHO polychlorinated biphenyls in food using comprehensive two-dimensional gas chromatography with electron-capture detection. J Chromatogr A. 1086:61–70. Darnerud PO, Atuma S, Aune M, Bjerselius R, Glynn A, Petersson Grawe K et al. (2006). Dietary intake estimations of organohalogen contaminants (dioxins, PCB, PBDE and chlorinated pesticides, e.g. DDT) based on Swedish market basket data. Food Chem Toxicol. 44:1597–606. Darnerud PO, Lignell S, Glynn A, Aune M, Törnkvist A, Stridsberg M (2010). POP levels in breast milk and maternal serum and thyroid hormone levels in mother–child pairs from Uppsala, Sweden. Environ Int. 36:180–7. Dean CE Jr, Benjamin SA, Chubb LS, Tessari JD, Keefe TJ (2002). Nonadditive hepatic tumor promoting effects by a mixture of two structurally different polychlorinated biphenyls in female rat livers. Toxicol Sci. 66:54–61. De Boer J (1999). Capillary gas chromatography for the determination of halogenated microcontaminants. J Chromatogr A. 843:179–98. De Boer J, Smedes F (1997). Effects of storage conditions of biological materials on the contents of organochlorine compounds and mercury. Mar Pollut Bull. 35:93–108. DeCaprio AP, Johnson GW, Tarbell AM, Carpenter DO, Chiarenzelli JR, Morse GS et al. (2005). Polychlorinated biphenyl (PCB) exposure assessment by multivariate statistical analysis of serum congener profiles in adult Native American population. Environ Res. 98:284–302. Denham M, Schell LM, Deane G, Gallo MV, Ravenscroft J, DeCaprio AP; Akwesasne Task Force on the Environment (2005). Relationship of lead, mercury, mirex, dichlorodiphenyldichloroethylene, hexachlorobenzene, and polychlorinated biphenyls to timing of menarche among Akwesasne Mohawk girls. Pediatrics. 115(2):e127–34. De Roos AJ, Hartge P, Lubin JH, Colt JS, Davis S, Cerhan JR et al. (2005). Persistent organochlorine chemicals in plasma and risk of non-Hodgkin’s lymphoma. Cancer Res. 65:11214–26. De Saeger S, Sergeant H, Piette M, Bruneel N, Van de Voorde W, Van Peteghem C (2005). Monitoring of polychlorinated biphenyls in Belgian human adipose tissue samples. Chemosphere. 58:953–60. Després C, Beuter A, Richer F, Poitras K, Veilleux A, Ayotte P et al. (2005). Neuromotor functions in Inuit preschool children exposed to Pb, PCBs, and Hg. Neurotoxicol Teratol. 27:245–57.

290

WHO Food Additives Series No. 71-S1, 2016

Supplement 1: Non-dioxin-like polychlorinated biphenyls

De Voogt P, Van Der Wielen FWM, Govers HAJ (2000). Freeze-drying brings about errors in polychlorinated biphenyl recovery calculations. Trends Anal Chem. 19:292–9. de Vos J, Dixon R, Vermeulen G, Gorst-Allman P, Cochran J, Rohwer E et al. (2011) Comprehensive twodimensional gas chromatography time of flight mass spectrometry (GC×GC-TOFMS) for environmental forensic investigations in developing countries. Chemosphere. 82:1230–9. Dewailly E, Tremblay-Rousseau H, Carrier G, Groulx S, Gingras S, Boggess K et al. (1991). PCDDs, PCDFs and PCBs in human milk of women exposed to a PCB fire and of women from the general population of the province of Québec – Canada. Chemosphere. 23:1831–5. Dewailly E, Mulvad G, Pedersen HS, Ayotte P, Demers A, Weber JP et al. (1999). Concentration of organochlorines in human brain, liver, and adipose tissue autopsy samples from Greenland. Environ Health Perspect. 107:823–8. Dirtu AC, Jaspers VL, Cernat R, Neels H, Covaci A (2010). Distribution of PCBs, their hydroxylated metabolites, and other phenolic contaminants in human serum from two European countries. Environ Sci Technol. 44:2876–83. Dragnev KH, Nims RW, Fox SD, Lindahl R, Lubet RA (1995). Relative potencies of induction of hepatic drug-metabolizing enzyme genes by individual PCB congeners. Toxicol Appl Pharmacol. 132:334–42. Drobna B, Fabisikova A, Conka K, Chovancova J, Domotorova M, Wimmerova S et al. (2011). Differences between dioxin-like PCB, non-dioxin-like PCB, polychlorinated dibenzo-p-dioxin and dibenzofuran intake from human milk and infant formula by infants in the Michalovce district (Slovakia). J Food Nutr Res. 50(2):106–17. Duarte-Davidson R, Jones KC (1994). Polychlorinated biphenyls (PCBs) in the UK population: estimated intake, exposure and body burden. Sci Total Environ. 151:131–52. Duinker JC, Hillebrand MTJ (1979). Mobilization of organochlorines from female lipid tissue and transplacental transfer to fetus in a harbour porpoise (Phocoena phocoena) in a contaminated area. Bull Environ Contam Toxicol. 23:728–32. Echols K, Gale R, Tillitt D, Schwartz T, O’Laughlin J (1997). An automated HPLC method for the fractionation of polychlorinated biphenyls, polychlorinated dibenzo-p-dioxins, and polychlorinated dibenzofurans in fish tissue on a porous graphitic carbon column. Environ Toxicol Chem. 16:1590–7. EFSA (2005). Opinion of the Scientific Panel on Contaminants in the Food Chain on a request from the Commission related to the presence of non dioxin-like polychlorinated biphenyls (PCB) in feed and food. EFSA [European Food Safety Authority] J. 284:1–137 (http://www.efsa.europa.eu/sites/default/ files/scientific_output/files/main_documents/284.pdf, accessed 1 January 2016). EFSA (2010). Results of the monitoring of non dioxin-like PCBs in food and feed [scientific report of EFSA]. EFSA [European Food Safety Authority] J. 8(7):1701 (http://www.efsa.europa.eu/sites/default/ files/scientific_output/files/main_documents/1701.pdf, accessed 1 January 2016). EFSA (2011). Use of the EFSA Comprehensive European Food Consumption Database in exposure assessment [guidance of EFSA]. EFSA [European Food Safety Authority] J. 9(3):2097 (http://www.efsa. europa.eu/sites/default/files/scientific_output/files/main_documents/2097.pdf, accessed 1 January 2016). EFSA (2012) Update of the monitoring of levels of dioxins and PCBs in food and feed [scientific report of EFSA]. EFSA [European Food Safety Authority] J. 10(7):2832 (http://www.efsa.europa.eu/sites/ default/files/scientific_output/files/main_documents/2832.pdf, accessed 1 January 2016). 291

Safety evaluation of certain food additives and contaminants Eightieth JECFA

Eguchi A, Nomiyama K, Devanathan G, Subramanian A, Bulbule KA, Parthasarathy P et al. (2012). Different profiles of anthropogenic and naturally produced organohalogen compounds in serum from residents living near a coastal area and e-waste recycling workers in India. Environ Int. 47:8–16. Ekuase EJ, Liu Y, Lehmler HJ, Robertson LW, Duffel MW (2011). Structure–activity relationships for hydroxylated polychlorinated biphenyls as inhibitors of the sulfation of dehydroepiandrosterone catalyzed by human hydroxysteroid sulfotransferase SULT2A1. Chem Res Toxicol. 24:1720–28. Ekuase EJ, Lehmler HJ, Robertson LW, Duffel MW (2014). Binding interactions of hydroxylated polychlorinated biphenyls (OHPCBs) with human hydroxysteroid sulfotransferase hSULT2A1. Chem Biol Interact. 212:56–64. Elabbas LE , Westerholm E, Roos R, Halldin K, Korkalainen M, Viluksela M et al. (2013). Non dioxin-like polychlorinated biphenyls: exposure and health hazards. In: Rose M, Fernandes A, editors. Persistent organic pollutants and toxic metals in foods. Chapter 10. Cambridge: Woodhead Publishing Ltd; 215– 60. Eljarrat E, Saulo J, Monjonell A, Caixach J, Rivera J (2001). Evaluation of an automated clean-up system for the isotope-dilution high-resolution mass spectrometric analysis of PCB, PCDD, and PCDF in food. Fresenius J Anal Chem. 371:983–8. Elnar AA, Diesel B, Desor F, Feidt C, Bouayed J, Kiemer AK et al. (2012). Neurodevelopmental and behavioral toxicity via lactational exposure to the sum of six indicator non-dioxin-like-polychlorinated biphenyls (∑6 NDL-PCBs) in mice. Toxicology. 299:44–54. Elskus AA, Stegeman JJ, Gooch JW, Black DE, Pruell RJ (1994). Polychlorinated biphenyl congener distributions in winter flounder as related to gender, spawning site, and congener metabolism. Environ Sci Technol. 28:401–7. Elzanaty S, Rignell-Hydbom A, Jönsson BA, Pedersen HS, Ludwicki JK, Shevets M et al. (2006). Association between exposure to persistent organohalogen pollutants and epididymal and accessory sex gland function: multicentre study in Inuit and European populations. Reprod Toxicol. 22:765–73. Engel LS, Laden F, Andersen A, Strickland PT, Blair A, Needham LL et al. (2007). Polychlorinated biphenyl levels in peripheral blood and non-Hodgkin’s lymphoma: a report from three cohorts. Cancer Res. 67:5545–52. WHO Food Additives Series No. 71-S1, 2016 Engel SM, Berkowitz GS, Barr DB, Teitelbaum SL, Siskind J, Meisel SJ et al. (2007). Prenatal organophosphate metabolite and organochlorine levels and performance on the Brazelton Neonatal Behavioral Assessment Scale in a multiethnic pregnancy cohort. Am J Epidemiol. 165:1397–404. Eppe G, Diletti G, Fernandes A, Haedrich J, Hart J, Hove H et al. (2014). Measurement uncertainty for persistent organic pollutants by isotope-dilution mass spectrometry. Organohalogen Compounds. 76:1486–8. Erickson MD (1986). Analytical chemistry of PCBs. Stoneham (MA): Butterworth Publishers. Eriksson P, Fredriksson A (1996). Developmental neurotoxicity of four ortho-substituted polychlorinated biphenyls in the neonatal mouse. Environ Toxicol Pharmacol. 1:155–65. Esposito M, Serpe FP, Diletti G, Messina G, Scortichini G, La Rocca C et al. (2014). Serum levels of polychlorinated dibenzo-p-dioxins, polychlorinated dibenzofurans and polychlorinated biphenyls in a population living in the Naples area, southern Italy. Chemosphere. 94:62–9.

292

Supplement 1: Non-dioxin-like polychlorinated biphenyls

Esteve-Turrillas FA, Caupos E, Llorca I, Pastor A, De La Guardia M (2008). Optimization of large-volume injection for the determination of polychlorinated biphenyls in children’s fast-food menus by lowresolution mass spectrometry. J Agric Food Chem. 56:1797–803. Eum SY, Lee YW, Hennig B, Toborek M (2004). VEGF regulates PCB 104–mediated stimulation of permeability and transmigration of breast cancer cells in human microvascular endothelial cells. Exp Cell Res. 296:231–44. European Commission (1999). EU Commission Decision 1999/788/EC of 3 December 1999 on protective measures with regard to contamination by dioxins of certain products of porcine and poultry origin intended for human or animal consumption. Off J Eur Union. L310:62–70. European Commission (2004). Commission recommendation of 11 October 2004 on the monitoring of background levels of dioxins and dioxin-like PCBs in foodstuffs. Off J Eur Union. L321:45–51. European Commission (2006). Commission Regulation (EC) No 1881/2006 of 19 December 2006 setting maximum levels for certain contaminants in foodstuffs. Off J Eur Union. L364:5–24. European Commission (2011). Commission Regulation (EU) No 1259/2011 of 2 December 2011 amending Regulation (EC) No 1881/2006 as regards maximum levels for dioxins, dioxin-like PCBs and non dioxin-like PCBs in foodstuffs. Off J Eur Union. L320:18–23. European Commission (2014). Commission Regulation (EU) No 589/2014 of 2 June 2014 laying down methods of sampling and analysis for the control of levels of dioxins, dioxin-like PCBs and non-dioxinlike PCBs in certain foodstuffs and repealing Regulation (EU) No 252/2012. Off J Eur Union. L164:18– 40. Fangstrom B, Hovander L, Bignert A, Athanassiadis I, Linderholm L, Grandjean P et al. (2005). Concentrations of polybrominated diphenyl ethers, polychlorinated biphenyls, and polychlorobiphenylols in serum from pregnant Faroese women and their children 7 years later. Environ Sci Technol. 39:9457–63. Fantini F, Porta D, Fano V, De Felip E, Senofonte O, Abballe A et al. (2012). [Epidemiologic studies on the health status of the population living in the Sacco River Valley.] Epidemiol Prev. 36:44–52 (in Italian). FAO/WHO (2006). Code of practice for the prevention and reduction of dioxin and dioxin-like PCB contamination in foods and feeds. Rome: Food and Agriculture Organization of the United Nations; and Geneva: World Health Organization; Joint FAO/WHO Food Standards Programme, Codex Alimentarius Commission (CAC/RCP 62-2006; http://www.codexalimentarius.org/standards/list-of-standards/en/, accessed 6 April 2015). FAO/WHO (2009). Principles and methods for the risk assessment of chemicals in food. A joint publication of the Food and Agriculture Organization of the United Nations and the World Health Organization. Geneva: World Health Organization (Environmental Health Criteria 240; http://www. inchem.org/documents/ehc/ehc/ehc240_index.htm, accessed 2 January 2015). Fattore E, Fanelli R, Dellatte E, Turrini A, Domenico AD (2008). Assessment of the dietary exposure to non-dioxin-like PCBs of the Italian general population. Chemosphere. 73:S278–83. Feeley MM, Jordan SA (1998). Dietary and tissue residue analysis and contaminant intake estimations in rats consuming diets composed of Great Lakes salmon: a multigeneration study. Regul Toxicol Pharmacol. 27:S8–17. Fernandes A, White S, D’Silva K, Rose M (2004). Simultaneous determination of PCDDs, PCDFs, PCBs and PBDEs in food. Talanta. 63:1147–55. 293

Safety evaluation of certain food additives and contaminants Eightieth JECFA

Fernandez MF, Kiviranta H, Molina-Molina JM, Laine O, Lopez-Espinosa MJ, Vartiainen T (2008). Polychlorinated biphenyls (PCBs) and hydroxy-PCBs in adipose tissue of women in southeast Spain. Chemosphere. 71:1196–205. Fidalgo-Used N, Blanco-González E, Sanz-Medel A (2007). Sample handling strategies for the determination of persistent trace organic contaminants from biota samples. Anal Chim Acta. 590:1– 16. Fischer RC, Wittlinger R, Ballschmiter K (1992). Retention-index based vapor pressure estimation for polychlorobiphenyl (PCB) by gas chromatography. Fresenius J Anal Chem. 342:421–5. Fitzgerald EF, Belanger EE, Gomez MI, Cayo M, McCaffrey RJ, Seegal RF et al. (2008). Polychlorinated biphenyl exposure and neuropsychological status among older residents of upper Hudson River communities. Environ Health Perspect. 116:209–15. Flor S, Ludewig G (2010). Polyploidy-induction by dihydroxylated monochlorobiphenyls: structure– activity-relationships. Environ Int. 36:962–9. Focant JF, Pirard C, De Pauw E (2004). Automated sample preparation-fractionation for the measurement of dioxins and related compounds in biological matrices: a review. Talanta. 63:1101–13. Focant JF, Sjödin A, Patterson DG Jr (2004). Improved separation of the 209 polychlorinated biphenyl congeners using comprehensive two-dimensional gas chromatography–time-of-flight mass spectrometry. J Chromatogr A. 1040:227–38. Focant JF, Eppe G, Massart A, Scholl G, Pirard C, De Pauw E (2006). High-throughput biomonitoring of dioxins and polychlorinated biphenyls at the sub-picogram level in human serum. J Chromatogr A. 1130:97–107. Focant JF, Fréry N, Bidondo ML, Eppe G, Scholl G, Saoudi A et al. (2013). Levels of polychlorinated dibenzo-p-dioxins, polychlorinated dibenzofurans and polychlorinated biphenyls in human milk from different regions of France. Sci Total Environ. 452–453:155–62. Forns J, Lertxundi N, Aranbarri A, Murcia M, Gascon M, Martinez D et al. (2012a). Prenatal exposure to organochlorine compounds and neuropsychological development up to two years of life. Environ Int. 45:72–7. Forns J, Torrent M, Garcia-Esteban R, Grellier J, Gascon M, Julvez J et al. (2012b). Prenatal exposure to polychlorinated biphenyls and child neuropsychological development in 4-year-olds: an analysis per congener and specific cognitive domain. Sci Total Environ. 432:338–43. Fournier A, Rychen G, Marchand P, Toussaint H, Le Bizec B, Feidt C (2013). Polychlorinated biphenyl (PCB) decontamination kinetics in lactating goats (Capra hircus) following a contaminated corn silage exposure. J Agric Food Chem. 61:7156–64. Frame GM (1997). A collaborative study of 209 PCB congeners and 6 Aroclors on 20 different HRGC columns: 1. Retention and coelution database. Fresenius J Anal Chem. 357:701–13. Fréry N, Saoudi A, Garnier R, Bidondo ML, Zeghnoun A, Castelbon K (2012). PCB human biomonitoring in general population in France: ENNS study. Saint-Maurice: French Institute for Public Health Surveillance (http://opac.invs.sante.fr/doc_num.php?explnum_id=8220, accessed 3 January 2016). Fritschi L, Benke G, Hughes AM, Kricker A, Vajdic CM, Grulich A et al. (2005). Risk of non-Hodgkin lymphoma associated with occupational exposure to solvents, metals, organic dusts and PCBs (Australia). Cancer Causes Control. 16:599–607.

294

WHO Food Additives Series No. 71-S1, 2016

Supplement 1: Non-dioxin-like polychlorinated biphenyls

Fromme H, Albrecht M, Angerer J, Drexler H, Gruber L, Schlummer M et al. (2007). Integrated Exposure Assessment Survey (INES): exposure to persistent and bioaccumulative chemicals in Bavaria, Germany. Int J Hyg Environ Health. 210:345–49. Fromme H, Korner W, Shahin N, Wanner A, Albrecht M, Boehmer S et al. (2009a). Human exposure to polybrominated diphenyl ethers (PBDE), as evidenced by data from a duplicate diet study, indoor air, house dust, and biomonitoring in Germany. Environ Int. 35:1125–35. Fromme H, Shahin N, Boehmer S, Albrecht M, Parlar H, Liebl B et al. (2009b). Dietary intake of nondioxin-like polychlorinated biphenyls (PCB) in Bavaria, Germany. Results from the Integrated Exposure Assessment Survey. Gesundheitswesen. 71:275–80. Fromme H, Albrecht M, Appel M, Hilger B, Völkel W, Liebl B et al. (2015). PCBs, PCDD/Fs, and PBDEs in blood samples of a rural population in south Germany. Int J Hyg Environ Health. 218:41–6. Fujita H, Honda K, Hamada N, Yasunaga G, Fujise Y (2009). Validation of high-throughput measurement system with microwave-assisted extraction, fully automated sample preparation device, and gas chromatography–electron capture detector for determination of polychlorinated biphenyls in whale blubber. Chemosphere. 74:1069–78. Fürst P (2006). Dioxins, polychlorinated biphenyls and other organohalogen compounds in human milk. Levels, correlations, trends and exposure through breastfeeding. Mol Nutr Food Res. 50:922–33. Gahrs M, Roos R, Andersson PL, Schrenk D (2013). Role of the nuclear xenobiotic receptors CAR and PXR in induction of cytochromes P450 by non-dioxinlike polychlorinated biphenyls in cultured rat hepatocytes. Toxicol Appl Pharmacol. 272:77–85. Gallagher RP, Macarthur AC, Lee TK, Weber JP, Leblanc A, Mark Elwood J et al. (2011). Plasma levels of polychlorinated biphenyls and risk of cutaneous malignant melanoma: a preliminary study. Int J Cancer. 128:1872–80. Galloway TS, Sanger RC, Smith KL, Fillmann G, Readman JW, Ford TE et al. (2002). Rapid assessment of marine pollution using multiple biomarkers and chemical immunoassays. Environ Sci Technol. 36:2219–26. Ganey PE, Sirois JE, Denison M, Robinson JP, Roth RA (1993). Neutrophil function after exposure to polychlorinated biphenyls in vitro. Environ Health Perspect. 101:430–4. Gardner-Stephen D, Heydel JM, Goyal A, Lu Y, Xie W, Lindblom T et al. (2004). Human PXR variants and their differential effects on the regulation of human UDP-glucuronosyltransferase gene expression. Drug Metab Dispos. 32: 340–7. Gascon M, Vrijheid M, Martínez D, Ballester F, Basterrechea M, Blarduni E et al. (2012). Pre-natal exposure to dichlorodiphenyldichloroethylene and infant lower respiratory tract infections and wheeze. Eur Respir J. 39:1188–96. Gascon M, Verner MA, Guxens M, Grimalt JO, Forns J, Ibarluzea J et al. (2013). Evaluating the neurotoxic effects of lactational exposure to persistent organic pollutants (POPs) in Spanish children. Neurotoxicology. 34:9–15. Gascon M, Sunyer J, Martínez D, Guerra S, Lavi I, Torrent M et al. (2014a). Persistent organic pollutants and children’s respiratory health: the role of cytokines and inflammatory biomarkers. Environ Int. 69:133–40. Gascon M, Sunyer J, Casas M, Martínez D, Ballester F, Basterrechea M et al. (2014b). Prenatal exposure to DDE and PCB 153 and respiratory health in early childhood: a meta-analysis. Epidemiology. 25:544–53. 295

Safety evaluation of certain food additives and contaminants Eightieth JECFA

Gaspar-Ramírez O, Pérez-Vázquez FJ, Pruneda-Álvarez LG, Orta-García ST, González-Amaro R, PérezMaldonado IN (2012). Effects of polychlorinated biphenyls 118 and 153 on Th1/Th2 cells differentiation. Immunopharm Immunotoxicol. 34:627–32. Gasull M, Pumarega J, Téllez-Plaza M, Castell C, Tresserras R, Lee DH et al. (2012). Blood concentrations of persistent organic pollutants and prediabetes and diabetes in the general population of Catalonia. Environ Sci Technol. 46:7799–810. Gatto NM, Longnecker MP, Press MF, Sullivan-Halley J, McKean-Cowdin R, Bernstein L (2007). Serum organochlorines and breast cancer: a case–control study among African-American women. Cancer Causes Control. 18:29–39. Gennings C, Carrico C, Factor-Litvak P, Krigbaum N, Cirillo PM, Cohn BA (2013). A cohort study evaluation of maternal PCB exposure related to time to pregnancy in daughters. Environ Health. 12:66. Gill US, Schwartz HM, Wheatley B (1995). Congener specific analysis of polychlorinated biphenyls (PCBs) in serum using GC/MSD. Chemosphere. 30:1969–77. Giordano F, Abballe A, De Felip E, di Domenico A, Ferro F, Grammatico P et al. (2010). Maternal exposures to endocrine disrupting chemicals and hypospadias in offspring. Birth Defects Res A Clin Mol Teratol. 88:241–50. Givens ML, Small CM, Terrell ML, Cameron LL, Michels Blanck H, Tolbert PE et al. (2007). Maternal exposure to polybrominated and polychlorinated biphenyls: infant birth weight and gestational age. Chemosphere. 69:1295–304. Giwercman AH, Rignell-Hydbom A, Toft G, Rylander L, Hagmar L, Lindh C et al. (2006). Reproductive hormone levels in men exposed to persistent organohalogen pollutants: a study of Inuit and three European cohorts. Environ Health Perspect. 114:1348–53. Glauert HP, Tharappel JC, Banerjee S, Chan LS, Kania-Korwel I, Lehmler H-J et al. (2008). Inhibition of the promotion of hepatocarcinogenesis by 2,2′,4,4′,5,5′-hexachlorobiphenyl (PCB-153) by the deletion of the p50 subunit of NF-κB in mice. Toxicol Appl Pharmacol. 232:302–8. Glynn A, Aune M, Darnerud PO, Cnattingius S, Bjerselius R, Becker W et al. (2007). Determinants of serum concentrations of organochlorine compounds in Swedish pregnant women: a cross-sectional study. Environ Health. 6:2. WHO Food Additives Series No. 71-S1, 2016 Glynn A, Thuvander A, Aune M, Johannisson A, Darnerud PO, Ronquist G et al. (2008). Immune cell counts and risks of respiratory infections among infants exposed pre- and postnatally to organochlorine compounds: a prospective study. Environ Health. 7:62. Glynn A, Lignell S, Darnerud PO, Aune M, Halldin Ankarberg E, Bergdahl IA et al. (2011). Regional differences in levels of chlorinated and brominated pollutants in mother’s milk from primiparous women in Sweden. Environ Int. 37:71–9. Glynn AW, Wolk A, Aune M, Atuma S, Zettermark S, Maehle-Schmid M et al. (2000). Serum concentrations of organochlorine compounds in men: a search for markers of exposure. Sci Total Environ. 263:197–208. Glynn AW, Atuma S, Aune M, Darnerud PO, Cnattingius S (2001). Polychlorinated biphenyl congeners as markers of toxic equivalents of polychlorinated biphenyls, dibenzo-p-dioxins and dibenzofurans in breast milk. Environ Res. 86:217–28. Glynn AW, Granath F, Aune M, Atuma S, Darnerud PO, Bjerselius R et al. (2003). Organochlorines in Swedish women: determinants of serum concentrations. Environ Health Perspect. 111:349–55.

296

Supplement 1: Non-dioxin-like polychlorinated biphenyls

Goldey ES, Crofton KM (1998). Thyroxine replacement attenuates hypothyroxinemia, hearing loss, and motor deficits following developmental exposure to Aroclor 1254 in rats. Toxicol Sci. 45:94–105. Goldey ES, Kehn LS, Lau C, Rehnberg GL, Crofton KM (1995). Developmental exposure to polychlorinated biphenyls (Aroclor 1254) reduces circulating thyroid hormone concentrations and causes hearing deficits in rats. Toxicol Appl Pharmacol. 135:77–88. Gomara B, Bordajandi LR, Fernandez MA, Herrero L, Abad E, Abalos M et al. (2005). Levels and trends of polychlorinated dibenzo-p-dioxins/furans (PCDD/Fs) and dioxin-like polychlorinated biphenyls (PCBs) in Spanish commercial fish and shellfish products, 1995–2003. J Agric Food Chem. 53:8406–13. Gomara B, Fernandez MA, Gonzalez M, Ramos L (2006). Feasibility of gas chromatography–ion trap tandem mass spectrometry for the determination of polychlorinated biphenyls in food. J Sep Sci. 29:123–30. Goncharov A, Rej R, Negoita S, Schymura M, Santiago-Rivera A, Morse G et al. (2009). Lower serum testosterone associated with elevated polychlorinated biphenyl concentrations in Native American men. Environ Health Perspect. 117:1454–60. Goncharov A, Bloom M, Pavuk M, Birman I, Carpenter DO (2010). Blood pressure and hypertension in relation to levels of serum polychlorinated biphenyls in residents of Anniston, Alabama. J Hypertens. 28:2053–60. Goncharov A, Pavuk M, Foushee HR, Carpenter DO (2011). Blood pressure in relation to concentrations of PCB congeners and chlorinated pesticides. Environ Health Perspect. 119:319–25. Govarts E, Nieuwenhuijsen M, Schoeters G, Ballester F, Bloemen K, de Boer M et al. (2012). Birth weight and prenatal exposure to polychlorinated biphenyls (PCBs) and dichlorodiphenyldichloroethylene (DDE): a meta-analysis within 12 European birth cohorts. Environ Health Perspect. 120:162–70. Govarts E, Casas M, Schoeters G, Eggesbø M, Valvi D, Nieuwenhuijsen M et al. (2014). Prenatal PCB-153 exposure and decreased birth weight: the role of gestational weight gain. Environ Health Perspect. 122:A89. Gralewicz S, Wiaderna D, Lutz P, Sitarek K (2009). Neurobehavioural functions in adult progeny of rat mothers exposed to methylmercury or 2,2′,4,4′,5,5′-hexachlorobiphenyl (PCB 153) alone or their combination during gestation and lactation. Int J Occup Med Environ Health. 22:277–91. Grandjean P, Poulsen LK, Heilmann C, Steuerwald U, Weihe P (2010). Allergy and sensitization during childhood associated with prenatal and lactational exposure to marine pollutants. Environ Health Perspect. 118:1429–33. Grandjean P, Andersen EW, Budtz-Jørgensen E, Nielsen F, Mølbak K, Weihe P et al. (2012). Serum vaccine antibody concentrations in children exposed to perfluorinated compounds. JAMA. 307:391–7. Grassi P, Fattore E, Generoso C, Fanelli R, Arvati M, Zuccato E (2010). Polychlorobiphenyls (PCBs), polychlorinated dibenzo-p-dioxins (PCDDs) and dibenzofurans (PCDFs) in fruit and vegetables from an industrial area in northern Italy. Chemosphere. 79:292–8. Gray KA, Klebanoff MA, Brock JW, Zhou H, Darden R, Needham L et al. (2005). In utero exposure to background levels of polychlorinated biphenyls and cognitive functioning among school-age children. Am J Epidemiol. 162:17–26. Grimalt JO, Carrizo D, Garí M, Font-Ribera L, Ribas-Fito N, Torrent M et al. (2010). An evaluation of the sexual differences in the accumulation of organochlorine compounds in children at birth and at the age of 4 years. Environ Res. 110:244–50. 297

Safety evaluation of certain food additives and contaminants Eightieth JECFA

Grimm FA, Hu D, Kania-Korwel I, Lehmler HJ, Ludewig G, Hornbuckle KC et al. (2015). Metabolism and metabolites of polychlorinated biphenyls. Crit Rev Toxicol. 45:245–72. Gunter EW (1997). Biological and environmental specimen banking at the Centers for Disease Control and Prevention. Chemosphere. 34:1945–53. Guruge KS, Seike N, Yamanaka N, Miyazaki S (2005). Polychlorinated dibenzo-p-dioxins, dibenzofurans, and biphenyls in domestic animal food stuff and their fat. Chemosphere. 58:883–9. Guruge KS, Iwakiri R, Yamanaka N, Tanimura N, Honda K (2012). Reduction of dioxins contamination in chicken egg and tissues by activated carbon as a feed additive. Organohalogen Compounds. 74:881–4. Ha MH, Lee DH, Son HK, Park SK, Jacobs DR Jr (2009). Association between serum concentrations of persistent organic pollutants and prevalence of newly diagnosed hypertension: results from the National Health and Nutrition Examination Survey 1999–2002. J Hum Hypertens. 23:274–86. Haase RF, McCaffrey RJ, Santiago-Rivera AL, Morse GS, Tarbell A (2009). Evidence of an age-related threshold effect of polychlorinated biphenyls (PCBs) on neuropsychological functioning in a Native American population. Environ Res. 109:73–85. Haddock RL, Badowski G, Bordallo R (2011). Cancer mortality following polychlorinated biphenyl (PCB) contamination of a Guam village. Hawaii Med J. 70:40–2. Hagmar L, Wallin E, Vessby B, Jönsson BA, Bergman A, Rylander L (2006). Intra-individual variations and time trends 1991–2001 in human serum levels of PCB, DDE and hexachlorobenzene. Chemosphere. 64:1507–13. Halldorsson TI, Thorsdottir I, Meltzer HM, Nielsen F, Olsen SF (2008). Linking exposure to polychlorinated biphenyls with fatty fish consumption and reduced fetal growth among Danish pregnant women: a cause for concern? Am J Epidemiol. 168:958–65. Hamers T, Kamstra JH, Cenijn PH, Pencikova K, Palkova L, Simeckova P et al. (2011). In vitro toxicity profiling of ultrapure non-dioxin-like polychlorinated biphenyl congeners and their relative toxic contribution to PCB mixtures in humans. Toxicol Sci. 121:88–100. Hansen L (1999). The ortho side of PCBs: occurrence and disposition. Norwell (MA): Kluwer Academic Publishers. WHO Food Additives Series No. 71-S1, 2016 Hansen S, Strøm M, Olsen SF, Maslova E, Rantakokko P, Kiviranta H et al. (2014). Maternal concentrations of persistent organochlorine pollutants and the risk of asthma in offspring: results from a prospective cohort with 20 years of follow-up. Environ Health Perspect. 122:93–9. Hardell L, Andersson SO, Carlberg M, Bohr L, van Bavel B, Lindström G et al. (2006a). Adipose tissue concentrations of persistent organic pollutants and the risk of prostate cancer. J Occup Environ Med. 48:700–7. Hardell L, Bavel B, Lindström G, Eriksson M, Carlberg M (2006b). In utero exposure to persistent organic pollutants in relation to testicular cancer risk. Int J Androl. 29:228–34. Hardell K, Carlberg M, Hardell L, Björnfoth H, Ericson Jogsten I, Eriksson M et al. (2009). Concentrations of organohalogen compounds and titres of antibodies to Epstein-Barr virus antigens and the risk for non-Hodgkin lymphoma. Oncol Rep. 21:1567–76. Harper N, Howie L, Connor K, Dickerson R, Safe S (1993). Immunosuppressive effects of highly chlorinated biphenyls and diphenyl ethers on T-cell dependent and independent antigens in mice. Toxicology. 85:123–35.

298

Supplement 1: Non-dioxin-like polychlorinated biphenyls

Health Canada (2012). Canadian standards (maximum levels) for various chemical contaminants in foods. Ottawa (ON): Health Canada (http://www.hc-sc.gc.ca/fn-an/securit/chem-chim/contaminantsguidelines-directives-eng.php, accessed 3 January 2016). Heilmann C, Grandjean P, Weihe P, Nielsen F, Budtz-Jørgensen E (2006). Reduced antibody responses to vaccinations in children exposed to polychlorinated biphenyls. PLoS Med. 3:e311. Heilmann C, Budtz-Jørgensen E, Nielsen F, Heinzow B, Weihe P, Grandjean P (2010). Serum concentrations of antibodies against vaccine toxoids in children exposed perinatally to immunotoxicants. Environ Health Perspect. 118:1434–8. Helmfrid I, Berglund M, Löfman O, Wingren G (2012). Health effects and exposure to polychlorinated biphenyls (PCBs) and metals in a contaminated community. Environ Int. 44:53–8. Heneweer M, van den Berg M, de Geest MC, de Jong PC, Bergman A, Sanderson JT (2005). Inhibition of aromatase activity by methyl sulfonyl PCB metabolites in primary culture of human mammary fibroblasts. Toxicol Appl Pharmacol. 202: 50–8. Henríquez-Hernández LA, Luzardo OP, Almeida-González M, Alvarez-León EE, Serra-Majem L, Zumbado M et al. (2011). Background levels of polychlorinated biphenyls in the population of the Canary Islands (Spain). Environ Res. 111:10–16. Henríquez-Hernández LA, Luzardo OP, Zumbado M, Camacho M, Serra-Majem L, Álvarez-León EE et al. (2014). Blood pressure in relation to contamination by polychlorobiphenyls and organochlorine pesticides: results from a population-based study in the Canary Islands (Spain). Environ Res. 135:48– 54. Herbstman JB, Sjödin A, Apelberg BJ, Witter FR, Halden RU, Patterson DG et al. (2008). Birth delivery mode modifies the associations between prenatal polychlorinated biphenyl (PCB) and polybrominated diphenyl ether (PBDE) and neonatal thyroid hormone levels. Environ Health Perspect. 116:1376–82. Hertz-Picciotto I, Charles MJ, James RA, Keller JA, Willman E, Teplin S (2005). In utero polychlorinated biphenyl exposures in relation to fetal and early childhood growth. Epidemiology. 16:648–56. Hertz-Picciotto I, Jusko TA, Willman EJ, Baker RJ, Keller JA, Teplin SW et al. (2008). A cohort study of in utero polychlorinated biphenyl (PCB) exposures in relation to secondary sex ratio. Environ Health. 7:37. Heudorf U, Angerer J, Drexler H (2002). Polychlorinated biphenyls in the blood plasma: current exposure of the population in Germany. Rev Environ Health. 17:123–34. Hilbert G, Lillemark L, Balchen S, Højskov CS (1998). Reduction of organochlorine contaminants from fish oil during refining. Chemosphere. 37:1241–52. Hirai T, Fujimine Y, Watanabe S, Nakano T (2005). Congener-specific analysis of polychlorinated biphenyl in human blood from Japanese. Environ Geochem Health. 27:65–73. Hodgson S, Thomas L, Fattore E, Lind PM, Alfven T, Hellström L et al. (2008). Bone mineral density changes in relation to environmental PCB exposure. Environ Health Perspect. 116:1162–6. Hofe CR, Feng L, Zephyr D, Stromberg AJ, Hennig B, Gaetke LM (2014). Fruit and vegetable intake, as reflected by serum carotenoid concentrations, predicts reduced probability of polychlorinated biphenyl–associated risk for type 2 diabetes: National Health and Nutrition Examination Survey 2003–2004. Nutr Res. 34:285–93. Hoffman CS, Small CM, Blanck HM, Tolbert P, Rubin C, Marcus M (2007). Endometriosis among women exposed to polybrominated biphenyls. Ann Epidemiol. 17:503–10. 299

Safety evaluation of certain food additives and contaminants Eightieth JECFA

Holene E, Nafstad I, Skaare JU, Sagvolden T (1998). Behavioural hyperactivity in rats following postnatal exposure to sub-toxic doses of polychlorinated biphenyl congeners 153 and 126. Behav Brain Res. 94: 213–24. Honma T, Suda M, Miyagawa M, Wang RS, Kobayashi K, Sekiguchi S (2009). Alteration of brain neurotransmitters in female rat offspring induced by prenatal administration of 16 and 64 mg/kg of 2,2′,4,4′,5,5′-hexachlorobiphenyl (PCB153). Ind Health. 47:11–21. Hoogenboom LAP, Kan CA, Bovee TFH, Van Der Weg G, Onstenk C, Traag WA (2004). Residues of dioxins and PCBs in fat of growing pigs and broilers fed contaminated feed. Chemosphere. 57:35–42. Hoogenboom L, Traag W, Bovee T, Goeyens L, Carbonnelle S, Van Loco J et al. (2006a). The CALUX bioassay: current status of its application to screening food and feed. Trends Anal Chem. 25:410–20. Hoogenboom LAP, Kan CA, Zeilmaker MJ, Van Eijkeren J, Traag WA (2006b). Carry-over of dioxins and PCBs from feed and soil to eggs at low contamination levels – Influence of mycotoxin binders on the carry-over from feed to eggs. Food Addit Contam. 23:518–27. Hoogenboom R, Bovee T, Traag W, Hoogerbrugge R, Baumann B, Portier L et al. (2006c). The use of the DR CALUX bioassay and indicator polychlorinated biphenyls for screening of elevated levels of dioxins and dioxin-like polychlorinated biphenyls in eel. Mol Nutr Food Res. 50:945–57. Hooper K, She J, Sharp M, Chow J, Jewell N, Gephart R et al. (2007). Depuration of polybrominated diphenyl ethers (PBDEs) and polychlorinated biphenyls (PCBs) in breast milk from California first-time mothers (primiparae). Environ Health Perspect. 115:1271–5. Hopf NB, Waters MA, Ruder AM, Prince MM (2010). Development of a retrospective job exposure matrix for PCB-exposed workers in capacitor manufacturing. J Occup Health. 52:199–208. Hsia MT, Lin FS, Allen JR (1978). Comparative mutagenicity and toxic effects of 2,5,2′,5′tetrachlorobiphenyl and its metabolites in bacterial and mammalian test systems. Res Commun Chem Pathol Pharmacol. 21:485–96. Hsu PC, Lai TJ, Guo NW, Lambert GH, Guo YL (2005). Serum hormones in boys prenatally exposed to polychlorinated biphenyls and dibenzofurans. J Toxicol Environ Health A. 68:1447–56. Hsu P-C, Pan M-H, Li L-A, Chen C-J, Tsai S-S, Guo Y-L (2007). Exposure in utero to 2,2′,3,3′,4,6′hexachlorobiphenyl (PCB 132) impairs sperm function and alters testicular apoptosis-related gene expression in rat offspring. Toxicol Appl Pharmacol. 221:68–75. Huetos O, Bartolomé M, Aragonés N, Cervantes-Amat M, Esteban M, Ruiz-Moraga M et al. (2014). Serum PCB levels in a representative sample of the Spanish adult population: the BIOAMBIENT.ES project. Sci Total Environ. 493:834–44. Humblet O, Williams PL, Korrick SA, Sergeyev O, Emond C, Birnbaum LS et al. (2011). Dioxin and polychlorinated biphenyl concentrations in mother’s serum and the timing of pubertal onset in sons. Epidemiology. 22:827–35. Hussain RJ, Gyori J, DeCaprio AP, Carpenter DO (2000). In vivo and in vitro exposure to PCB 153 reduces long-term potentiation. Environ Health Perspect. 108:827–31. Hutzinger O, Safe S, Zitko V (1974). The chemistry of PCB’s. Cleveland (OH): CRC Press. IARC (2015). Polychlorinated biphenyls and polybrominated biphenyls. Lyon: International Agency for Research on Cancer (IARC Monographs on the Evaluation of Carcinogenic Risks to Humans, Volume 107).

300

WHO Food Additives Series No. 71-S1, 2016

Supplement 1: Non-dioxin-like polychlorinated biphenyls

Ibarluzea J, Alvarez-Pedrerol M, Guxens M, Marina LS, Basterrechea M, Lertxundi A et al. (2011). Sociodemographic, reproductive and dietary predictors of organochlorine compounds levels in pregnant women in Spain. Chemosphere. 82:114–20. Ibhazehiebo K, Iwasaki T, Okano-Uchida T, Shimokawa N, Ishizaki Y, Koibuchi N (2011). Suppression of thyroid hormone receptor–mediated transcription and disruption of thyroid hormone–induced cerebellar morphogenesis by the polybrominated biphenyl mixture, BP-6. Neurotoxicology. 32:400–9. IEHR (1991). Reassessment of liver findings in five PCB studies in rats. Washingon (DC): Institute for Evaluating Health Risks. Ingelido AM, Ballard T, Dellatte E, di Domenico A, Ferri F, Fulgenzi AR et al. (2007). Polychlorinated biphenyls (PCBs) and polybrominated diphenyl ethers (PBDEs) in milk from Italian women living in Rome and Venice. Chemosphere. 67:S301–6. IPCS (1978). Polychlorinated biphenyls and terphenyls. Geneva: World Health Organization, International Programme on Chemical Safety (Environmental Health Criteria 2; http://www.inchem. org/documents/ehc/ehc/ehc002.htm, accessed 3 January 2016). IPCS (1993). Polychlorinated biphenyls and terphenyls, second edition. Geneva: World Health Organization, International Programme on Chemical Safety (Environmental Health Criteria 140; http://www.inchem.org/documents/ehc/ehc/ehc140.htm, accessed 3 January 2016). Ishikawa Y, Noma Y, Mori Y, Sakai S (2007). Congener profiles of PCB and a proposed new set of indicator congeners. Chemosphere. 67:1838–51. ISO (2008). ISO 8260:2008 (IDF 130: 2008): Milk and milk products – Determination of organochlorine pesticides and polychlorobiphenyls – method using capillary gas–liquid chromatography with electron capture detection. Geneva: International Organization for Standardization. Itoh H, Iwasaki M, Hanaoka T, Kasuga Y, Yokoyama S, Onuma H et al. (2009). Serum organochlorines and breast cancer risk in Japanese women: a case–control study. Cancer Causes Control. 20:567–80. Jackson LW, Lynch CD, Kostyniak PJ, McGuinness BM, Louis GM (2010). Prenatal and postnatal exposure to polychlorinated biphenyls and child size at 24 months of age. Reprod Toxicol. 29:25–31. James MO (2013). Polychlorinated biphenyls: metabolism and metabolites. In: Robertson LW, Hansen LG, editors. PCBs: recent advances in environmental toxicology and health effects. Lexington (KY): University Press of Kentucky; 35–46. Jan J, Sovcikova E, Kocan A, Wsolova L, Trnovec T (2007). Developmental dental defects in children exposed to PCBs in eastern Slovakia. Chemosphere. 67:S350–4. Jankovic S, Curcic M, Radicevic T, Stefanovic S, Lenhardt M, Durgo K et al. (2011). Non-dioxin-like PCBs in ten different fish species from the Danube River in Serbia. Environ Monit Assess. 181:153–63. Japan External Trade Organization (2011). Specifications and standards for foods, food additives, etc. under the Food Sanitation Act (Abstract) 2010 (https://www.jetro.go.jp/en/reports/regulations/pdf/ foodext2010e.pdf, accessed 18 March 2015). Jaraczewska K, Lulek J, Covaci A, Voorspoels S, Kaluba-Skotarczak A, Drews K et al. (2006). Distribution of polychlorinated biphenyls, organochlorine pesticides and polybrominated diphenyl ethers in human umbilical cord serum, maternal serum and milk from Wielkopolska region, Poland. Sci Total Environ. 372:20–31.

301

Safety evaluation of certain food additives and contaminants Eightieth JECFA

Jelinek CF (1985). Control of chemical contaminants in foods: past, present, and future. J Assoc Off Anal Chem. 68:1063–8. Jeong YC, Walker NJ, Burgin DE, Kissling G, Gupta M, Kupper L et al. (2008). Accumulation of M1dG DNA adducts after chronic exposure to PCBs, but not from acute exposure to polychlorinated aromatic hydrocarbons. Free Radic Biol Med. 45:585–91. Johansen EB, Knoff M, Fonnum F, Lausund PL, Walaas SI, Wøien G et al. (2011). Postnatal exposure to PCB 153 and PCB 180, but not to PCB 52, produces changes in activity level and stimulus control in outbred male Wistar Kyoto rats. Behav Brain Funct. 7:18. Johansen EB, Fonnum F, Lausund PL, Walaas SI, Baerland NE, Wøien G et al. (2014). Behavioral changes following PCB 153 exposure in the spontaneously hypertensive rat – an animal model of attentiondeficit/hyperactivity disorder. Behav Brain Funct. 10:1. Johansson M, Nilsson S, Lund BO (1998). Interactions between methylsulfonyl PCBs and the glucocorticoid receptor. Environ Health Perspect. 106:769–72. Johansson M, Larsson C, Bergman A, Lund BO (1998). Structure–activity relationship for inhibition of CYP11B1-dependent glucocorticoid synthesis in Y1 cells by aryl methyl sulfones. Pharmacol Toxicol. 83: 225–30. Johnson-Restrepo B, Kannan K, Rapaport DP, Rodan BD (2005). Polybrominated diphenyl ethers and polychlorinated biphenyls in human adipose tissue from New York. Environ Sci Technol. 39:5177–82. Jones KC (1988). Determination of polychlorinated biphenyls in human foodstuffs and tissues: suggestions for a selective congener analytical approach. Sci Total Environ. 68:141–59. Jordan SA, Feeley MM (1999). PCB congener patterns in rats consuming diets containing Great Lakes salmon: analysis of fish, diets, and adipose tissue. Environ Res. 80:S207–12. Jørgensen ME, Borch-Johnsen K, Bjerregaard P (2008). A cross-sectional study of the association between persistent organic pollutants and glucose intolerance among Greenland Inuit. Diabetologia. 51:1416–22. Juan C, Thomas GO, Semple KT, Jones KC (1999). Methods for the analysis of PCBs in human food, faeces and serum. Chemosphere. 39:1467–76. WHO Food Additives Series No. 71-S1, 2016 Julshamn K, Duinker A, Berntssen M, Nilsen BM, Frantzen S, Nedreaas K et al. (2013). A baseline study on levels of polychlorinated dibenzo-p-dioxins, polychlorinated dibenzofurans, non-ortho and monoortho PCBs, non-dioxin-like PCBs and polybrominated diphenyl ethers in Northeast Arctic cod (Gadus morhua) from different parts of the Barents Sea. Mar Pollut Bull. 75:250–8. Julshamn K, Valdersnes S, Duinker A, Nedreaas K, Sundet JH, Maage A (2015). Heavy metals and POPs in red king crab from the Barents Sea. Food Chem. 167:409–17. Jursa S, Chovancova J, Petrik J, Loksa J (2006). Dioxin-like and non-dioxin-like PCBs in human serum of Slovak population. Chemosphere. 64:686–91. Jusko TA, De Roos AJ, Schwartz SM, Lawrence BP, Palkovicova L, Nemessanyi T et al. (2010). A cohort study of developmental polychlorinated biphenyl (PCB) exposure in relation to post-vaccination antibody response at 6-months of age. Environ Res. 110:388–95. Jusko TA, De Roos AJ, Schwartz SM, Lawrence BP, Palkovicova L, Nemessanyi T et al. (2011). Maternal and early postnatal polychlorinated biphenyl exposure in relation to total serum immunoglobulin concentrations in 6-month-old infants. J Immunotoxicol. 8:95–100.

302

Supplement 1: Non-dioxin-like polychlorinated biphenyls

Jusko TA, Sonneborn D, Palkovicova L, Kocan A, Drobna B, Trnovec T et al. (2012). Pre- and postnatal polychlorinated biphenyl concentrations and longitudinal measures of thymus volume in infants. Environ Health Perspect. 120:595–600. Kalantzi OI, Brown FR, Caleffi M, Goth-Goldstein R, Petreas M (2009). Polybrominated diphenyl ethers and polychlorinated biphenyls in human breast adipose samples from Brazil. Environ Int. 35:113–7. Kalantzi OI, Geens T, Covaci A, Siskos PA (2011). Distribution of polybrominated diphenyl ethers (PBDEs) and other persistent organic pollutants in human serum from Greece. Environ Int. 37:349–53. Kan CA, Meijer GAL (2007). The risk of contamination of food with toxic substances present in animal feed. Anim Feed Sci Technol. 133:84–108. Kang JH, Park H, Chang YS, Choi JW (2008). Distribution of organochlorine pesticides (OCPs) and polychlorinated biphenyls (PCBs) in human serum from urban areas in Korea. Chemosphere. 73:1625– 31. Kannan K, Tanabe S, Borrell A, Aguilar A, Focardi S, Tatsukawa R (1993). Isomer-specific analysis and toxic evaluation of polychlorinated biphenyls in striped dolphins affected by an epizootic in the western Mediterranean Sea. Arch Environ Contam Toxicol. 25:227–33. Karmaus W, Brooks KR, Nebe T, Witten J, Obi-Osius N, Kruse H (2005). Immune function biomarkers in children exposed to lead and organochlorine compounds: a cross-sectional study. Environ Health. 4:5. Karmaus W, Osuch JR, Eneli I, Mudd LM, Zhang J, Mikucki D et al. (2009). Maternal levels of dichlorodiphenyl-dichloroethylene (DDE) may increase weight and body mass index in adult female offspring. Occup Environ Med. 66:143–9. Kashima S, Yorifuji T, Tsuda T (2011). Acute non-cancer mortality excess after polychlorinated biphenyls and polychlorinated dibenzofurans mixed exposure from contaminated rice oil: Yusho. Sci Total Environ. 409:3288–94. Kato Y, Haraguchi K, Shibahara T, Yumoto S, Masuda Y, Kimura R (1998). Reduction of thyroid hormone levels by methylsulfonyl metabolites of polychlorinated biphenyl congeners in rats. Arch Toxicol. 72:541–4. Kato Y, Haraguchi K, Shibahara T, Yumoto S, Masuda Y, Kimura R (1999). Reduction of thyroid hormone levels by methylsulfonyl metabolites of tetra- and pentachlorinated biphenyls in male SpragueDawley rats. Toxicol Sci. 48:51–4. Kato Y, Haraguchi K, Shibahara T, Yumoto S, Masuda Y, Kimura R (2000). Reduction of serum thyroxine concentrations by methylsulfonyl metabolites of tetra-, penta- and hexachlorinated biphenyls in male Sprague-Dawley rats. Chemosphere. 40:1233–40. Kato Y, Ikushiro S, Haraguchi K, Yamazaki T, Ito Y, Suzuki H et al. (2004). A possible mechanism for decrease in serum thyroxine level by polychlorinated biphenyls in Wistar and Gunn rats. Toxicol Sci. 81:309–15. Kato Y, Ikushiro S, Takiguchi R, Haraguchi K, Koga N, Uchida S et al. (2007). A novel mechanism for polychlorinated biphenyl–induced decrease in serum thyroxine level in rats. Drug Metab Dispos. 35:1949–55. Kato Y, Haraguchi K, Ito Y, Fujii A, Yamazaki T, Endo T et al. (2010). Polychlorinated biphenyl–mediated decrease in serum thyroxine level in rodents. Drug Metab Dispos. 38:697–704.

303

Safety evaluation of certain food additives and contaminants Eightieth JECFA

Kato Y, Tamaki S, Haraguchi K, Ikushiro S, Sekimoto M, Ohta C et al. (2012). Comparative study on 2,2′,4,5,5′-pentachlorobiphenyl-mediated decrease in serum thyroxine level between C57BL/6 and its transthyretin-deficient mice. Toxicol Appl Pharmacol. 263:323–9. Kearney JP, Cole DC, Ferron LA, Weber J (1999). Blood PCB, p,p′-DDE, and mirex levels in Great Lakes fish and waterfowl consumers in two Ontario communities. Environ Res. 80:S138–49. Kenet T, Froemke RC, Schreiner CE, Pessah IN, Merzenich MM (2007). Perinatal exposure to a noncoplanar polychlorinated biphenyl alters tonotopy, receptive fields, and plasticity in rat primary auditory cortex. Proc Natl Acad Sci USA. 104:7646–51. Kerst M, Waller U, Reifenhauser W, Korner W (2004). Carry-over rates of dioxin-like PCB from grass to cows’ milk. Organohalogen Compounds. 66:2412–5. Kezios KL, Liu X, Cirillio PM, Kalantzi OI, Wang Y, Petreas MX et al. (2012). Prenatal polychlorinated biphenyl exposure is associated with decreased gestational length but not birth weight: archived samples from the Child Health and Development Studies pregnancy cohort. Environ Health. 11:49. Khan MA, Lichtensteiger CA, Faroon O, Mumtaz M, Schaeffer DJ, Hansen LG (2002). The hypothalamo– pituitary–thyroid (HPT) axis: a target of nonpersistent ortho-substituted PCB congeners. Toxicol Sci. 65:52–61. Khanjani N, Sim MR (2007). Maternal contamination with PCBs and reproductive outcomes in an Australian population. J Expo Sci Environ Epidemiol. 17:191–5. Kim M, Kim S, Yun S, Lee M, Cho B, Park J et al. (2004). Comparison of seven indicator PCBs and three coplanar PCBs in beef, pork, and chicken fat. Chemosphere. 54:1533–8. Kiviranta H, Tuomisto JT, Tuomisto J, Tukiainen E, Vartiainen T (2005). Polychlorinated dibenzo-pdioxins, dibenzofurans, and biphenyls in the general population in Finland. Chemosphere. 60:854–69. Klincic D, Romanic SH, Saric MM, Grzunov J, Dukic B (2014). Polychlorinated biphenyls and organochlorine pesticides in human milk samples from two regions in Croatia. Environ Toxicol Pharmacol. 37:543–52. Knerr S, Schrenk D (2006). Carcinogenicity of “non-dioxinlike” polychlorinated biphenyls. Crit Rev Toxicol. 36:663–94. WHO Food Additives Series No. 71-S1, 2016 Knobeloch L, Turyk M, Imm P, Schrank C, Anderson H (2009). Temporal changes in PCB and DDE levels among a cohort of frequent and infrequent consumers of Great Lakes sportfish. Environ Res. 109:66– 72. Kobayashi K, Miyagawa M, Wang RS, Suda M, Sekiguchi S, Honma T (2008). Effects of in utero exposure to 2,2′,4,4′,5,5′-hexachlorobiphenyl (PCB 153) on somatic growth and endocrine status in rat offspring. Congenit Anom (Kyoto). 48:151–7. Kobayashi K, Miyagawa M, Wang RS, Suda M, Sekiguchi S, Honma T (2009). Effects of in utero exposure to 2,2′,4,4′,5,5′-hexachlorobiphenyl on postnatal development and thyroid function in rat offspring. Ind Health. 47:189–97. Kociba RJ, Keyes DG, Beyer JE, Carreon RM, Wade CE, Dittenber DA et al. (1978). Results of a two-year toxicity and oncogenicity study of 2,3,7,8-tetrachlorodibenzo-p-dioxin in rats. Toxicol Appl Pharmacol. 46:279–303. Kodavanti PR, Tilson HA (1997). Structure–activity relationships of potentially neurotoxic PCB congeners in the rat. Neurotoxicology. 18:425–41.

304

Supplement 1: Non-dioxin-like polychlorinated biphenyls

Kodavanti PR, Shin DS, Tilson HA, Harry GJ (1993). Comparative effects of two polychlorinated biphenyl congeners on calcium homeostasis in rat cerebellar granule cells. Toxicol Appl Pharmacol. 123:97–106. Kodavanti PR, Ward TR, McKinney JD, Tilson HA (1995). Increased [3H]phorbol ester binding in rat cerebellar granule cells by polychlorinated biphenyl mixtures and congeners: structure–activity relationships. Toxicol Appl Pharmacol. 130:140–8. Kodavanti PR, Ward TR, McKinney JD, Tilson HA (1996). Inhibition of microsomal and mitochondrial Ca2+-sequestration in rat cerebellum by polychlorinated biphenyl mixtures and congeners. Structure– activity relationships. Arch Toxicol. 70: 150–7. Kodavanti PR, Derr-Yellin EC, Mundy WR, Shafer TJ, Herr DW, Barone S et al. (1998a). Repeated exposure of adult rats to Aroclor 1254 causes brain region–specific changes in intracellular Ca2+ buffering and protein kinase C activity in the absence of changes in tyrosine hydroxylase. Toxicol Appl Pharmacol. 153:186–98. Kodavanti PR, Ward TR, Derr-Yellin EC, Mundy WR, Casey AC, Bush B et al. (1998b). Congener-specific distribution of polychlorinated biphenyls in brain regions, blood, liver, and fat of adult rats following repeated exposure to Aroclor 1254. Toxicol Appl Pharmacol. 153:199–210. Konuspayeva G, Faye B, Pauw ED, Focant J (2011). Levels and trends of PCDD/Fs and PCBs in camel milk (Camelus bactrianus and Camelus dromedarius) from Kazakhstan. Chemosphere. 85:351–60. Koopman-Esseboom C, Huisman M, Weisglas-Kuperus N, Van der Paauw CG, Tuinstra LGMT, Boersma ER et al. (1994). PCB and dioxin levels in plasma and human milk of 418 Dutch women and their infants. Predictive value of PCB congener levels in maternal plasma for fetal and infant’s exposure to PCBs and dioxins. Chemosphere. 28:1721–32. Kopec AK, Burgoon LD, Ibrahim-Aibo D, Mets BD, Tashiro C, Potter D et al. (2010). PCB153-elicited hepatic responses in the immature, ovariectomized C57BL/6 mice: comparative toxicogenomic effects of dioxin and non-dioxin-like ligands. Toxicol Appl Pharmacol. 243:359–71. Kopec AK, D’Souza ML, Mets BD, Burgoon LD, Reese SE, Archer KJ et al. (2011). Nonadditive hepatic gene expression elicited by 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) and 2,2′,4,4′,5,5′-hexachlorobiphenyl (PCB153) co-treatment in C57BL/6 mice. Toxicol Appl Pharmacol. 256:154–67. Koppen G, Covaci A, Van Cleuvenbergen R, Schepens P, Winneke G, Nelen V et al. (2002). Persistent organochlorine pollutants in human serum of 50–65 years old women in the Flanders Environmental and Health Study (FLEHS). Part 1: Concentrations and regional differences. Chemosphere. 48:811–25. Koppen G, Den Hond E, Nelen V, Van De Mieroop E, Bruckers L, Bilau M et al. (2009). Organochlorine and heavy metals in newborns: results from the Flemish Environment and Health Survey (FLEHS 2002–2006). Environ Int. 35:1015–22. Košcan A, Petrík J, Chovancová J, Drobná B (1994). Method for the group separation of non-ortho-, mono-ortho- and multi-ortho-substituted polychlorinated biphenyls and polychlorinated dibenzop-dioxins/polychlorinated dibenzofurans using activated carbon chromatography. J Chromatogr A. 665:139–53. Kotz A, Hadrich J, Wahl K, Djuchin K, Malisch R (2014). Is standardization of extraction methods for determination of PCDD/Fs and PCBs required? Organohalogen Compounds. 76:297–300. Kretschmer XC, Baldwin WS (2005). CAR and PXR: xenosensors of endocrine disrupters? Chem Biol Interact. 155:111–28. 305

Safety evaluation of certain food additives and contaminants Eightieth JECFA

Krogenaes AK, Nafstad I, Skåre JU, Farstad W, Hafne AL (1998). In vitro reproductive toxicity of polychlorinated biphenyl congeners 153 and 126. Reprod Toxicol. 12:575–80. Krogenaes AK, Ropstad E, Gutleb AC, Hårdnes N, Berg V, Dahl E et al. (2014). In utero exposure to environmentally relevant concentrations of PCB 153 and PCB 118 disrupts fetal testis development in sheep. J Toxicol Environ Health A. 77:628–49. Krokos F, Creaser CS, Wright C, Startin JR (1997). Congener-specific method for the determination of ortho and non-ortho polychlorinated biphenyls, polychlorinated dibenzo-p-dioxins and polychlorinated dibenzofurans in foods by carbon-column fractionation and gas chromatography–isotope dilution mass spectrometry. Fresenius J Anal Chem. 357:732–42. Kumar J, Lind PM, Salihovic S, van Bavel B, Ekdahl KN, Nilsson B et al. (2014). Influence of persistent organic pollutants on the complement system in a population-based human sample. Environ Int. 71:94–100. Kunisue T, Tanabe S (2009). Hydroxylated polychlorinated biphenyls (OH-PCBs) in the blood of mammals and birds from Japan: lower chlorinated OH-PCBs and profiles. Chemosphere. 74:950–61. Kunisue T, Takayanagi N, Isobe T, Takahashi S, Nose M, Yamada T (2007). Polybrominated diphenyl ethers and persistent organochlorines in Japanese human adipose tissues. Environ Int. 33:1048–56. Kvalem HE, Knutsen HK, Thomsen C, Haugen M, Stigum H, Brantsaeter AL et al. (2009). Role of dietary patterns for dioxin and PCB exposure. Mol Nutr Food Res. 53:1438–51. Kvist L, Giwercman A, Weihe P, Jensen TK, Grandjean P, Halling J et al. (2014). Exposure to persistent organic pollutants and sperm sex chromosome ratio in men from the Faroe Islands. Environ Int. 73:359–64. Laden F, Bertrand KA, Altshul L, Aster JC, Korrick SA, Sagiv SK (2010). Plasma organochlorine levels and risk of non-Hodgkin lymphoma in the Nurses’ Health Study. Cancer Epidemiol Biomarkers Prev. 19:1381–4. Lamb MR, Taylor S, Liu X, Wolff MS, Borrell L, Matte TD et al. (2006). Prenatal exposure to polychlorinated biphenyls and postnatal growth: a structural analysis. Environ Health Perspect. 114:779–85. Langer P, Kocan A, Tajtakova M, Petrik J, Chovancova J, Drobna B et al. (2005). Human thyroid in the population exposed to high environmental pollution by organochlorinated pollutants for several decades. Endocr Regul. 39:13–20. Langer P, Tajtáková M, Kocan A, Petrík J, Koska J, Ksinantová L et al. (2007a). Thyroid ultrasound volume, structure and function after long-term high exposure of large population to polychlorinated biphenyls, pesticides and dioxin. Chemosphere. 69:118–27. Langer P, Kocan A, Tajtáková M, Rádiková Z, Petrík J, Koska J et al. (2007b). Possible effects of persistent organochlorinated pollutants cocktail on thyroid hormone levels and pituitary–thyroid interrelations. Chemosphere. 70:110–8. Langer P, Kocan A, Drobna B, Susienkova K, Radikova Z, Huckova M et al. (2012). Blood testosterone in middle aged males heavily exposed to endocrine disruptors is decreasing more with HCB and p,p′-DDE related to BMI and lipids, but not with Σ15PCBs. Endocr Regul. 46:51–9. Langeveld WT, Meijer M, Westerink RH (2012). Differential effects of 20 non-dioxin-like PCBs on basal and depolarization-evoked intracellular calcium levels in PC12 cells. Toxicol Sci. 126:487–96.

306

WHO Food Additives Series No. 71-S1, 2016

Supplement 1: Non-dioxin-like polychlorinated biphenyls

Lans MC, Klasson-Wehler E, Willemsen M, Meussen E, Safe S, Brouwer A (1993). Structure-dependent, competitive interaction of hydroxypolychlorobiphenyls, -dibenzo-p-dioxins and dibenzofurans with human transthyretin. Chem Biol Interact. 88:7–21. Lauby-Secretan B, Loomis D, Grosse Y, El Ghissassi F, Bouvard V, Benbrahim-Tallaa L et al. (2013). Carcinogenicity of polychlorinated biphenyls and polybrominated biphenyls. Lancet Oncol. 14:287–8. Lecavalier P, Chu I, Yagminas A, Villeneuve DC, Poon R, Feeley M et al. (1997). Subchronic toxicity of 2,2′,3,3′,4,4′ hexachloro-biphenyl in rats. J Toxicol Environ Health. 51:265–77. Lee DH, Steffes M, Jacobs DR (2007). Positive associations of serum concentration of polychlorinated biphenyls or organochlorine pesticides with self-reported arthritis, especially rheumatoid type, in women. Environ Health Perspect. 115:883–8. Lee DH, Lee IK, Song K, Steffes M, Toscano W, Baker BA et al. (2006). A strong dose–response relation between serum concentrations of persistent organic pollutants and diabetes: results from the National Health and Examination Survey 1999–2002. Diabetes Care. 29:1638–44. Lee DH, Lee IK, Porta M, Steffes M, Jacobs DR Jr (2007a). Relationship between serum concentrations of persistent organic pollutants and the prevalence of metabolic syndrome among non-diabetic adults: results from the National Health and Nutrition Examination Survey 1999–2002. Diabetologia. 50:1841–51. Lee DH, Lee IK, Jin SH, Steffes M, Jacobs DR Jr (2007b). Association between serum concentrations of persistent organic pollutants and insulin resistance among nondiabetic adults: results from the National Health and Nutrition Examination Survey 1999–2002. Diabetes Care. 30:622–8. Lee DH, Steffes MW, Sjödin A, Jones RS, Needham LL, Jacobs DR Jr (2010). Low dose of some persistent organic pollutants predicts type 2 diabetes: a nested case–control study. Environ Health Perspect. 118:1235–42. Lee DH, Lind PM, Jacobs DR Jr, Salihovic S, van Bavel B, Lind L (2011a). Polychlorinated biphenyls and organochlorine pesticides in plasma predict development of type 2 diabetes in the elderly: the Prospective Investigation of the Vasculature in Uppsala Seniors (PIVUS) study. Diabetes Care. 34:1778– 84. Lee DH, Steffes MW, Sjödin A, Jones RS, Needham LL, Jacobs DR Jr (2011b). Low dose organochlorine pesticides and polychlorinated biphenyls predict obesity, dyslipidemia, and insulin resistance among people free of diabetes. PLoS One. 6:e15977. Lee DH, Lind L, Jacobs DR Jr, Salihovic S, van Bavel B, Lind PM (2012). Associations of persistent organic pollutants with abdominal obesity in the elderly: the Prospective Investigation of the Vasculature in Uppsala Seniors (PIVUS) study. Environ Int. 40:170–8. Leeman WR, Van Den Berg KJ, Houben GF (2007). Transfer of chemicals from feed to animal products: the use of transfer factors in risk assessment. Food Addit Contam. 24:1–13. Lehmler HJ, Robertson LW, Garrison AW, Kodavanti PR (2005). Effects of PCB 84 enantiomers on [3H]phorbol ester binding in rat cerebellar granule cells and 45Ca2+-uptake in rat cerebellum. Toxicol Lett. 156:391–400. Lesmana R, Shimokawa N, Takatsuru Y, Iwasaki T, Koibuchi N (2014). Lactational exposure to hydroxylated polychlorinated biphenyl (OH-PCB 106) causes hyperactivity in male rat pups by aberrant increase in dopamine and its receptor. Environ Toxicol. 29:876–83.

307

Safety evaluation of certain food additives and contaminants Eightieth JECFA

Letcher RJ, Klasson-Wehler E, Bergman Å (2000). Methyl sulphone and hydroxylated metabolites of polychlorinated biphenyls. In: Paarsivirta J, editor. Handbook of environmental chemistry: anthropogenic compounds – new types of persistent halogenated compounds, vol. 3, part K. Heidelberg: Springer-Verlag; 315–59. Letcher RJ, Lemmen JG, van der Burg B, Brouwer A, Bergman A, Giesy JP et al. (2002). In vitro antiestrogenic effects of aryl methyl sulfone metabolites of polychlorinated biphenyls and 2,2-bis(4chlorophenyl)-1,1-dichloroethene on 17β-estradiol-induced gene expression in several bioassay systems. Toxicol Sci. 69:362–72. Levin M, Morsey B, DeGuise S (2006). Modulation of the respiratory burst by organochlorine mixtures in marine mammals, humans, and mice. J Toxicol Environ Health A. 70:73–83. Levin M, Morsey B, Mori C, DeGuise S (2004). Specific non-coplanar PCB-mediated modulation of bottlenose dolphin and beluga whale phagocytosis upon in vitro exposure. J Toxicol Environ Health A. 67:1517–35. Levin M, Morsey B, Mori C, Nambiar PR, DeGuise S (2005). PCBs and TCDD, alone and in mixtures, modulate marine mammal but not B6C3F1 mouse leucocyte phagocytosis. J Toxicol Environ Health A. 68:635–56. L’Homme B, Scholl G, Eppe G, Focant J (2015). Validation of a gas chromatography–triple quadrupole mass spectrometry method for confirmatory analysis of dioxins and dioxin-like polychlorobiphenyls in feed following new EU Regulation 709/2014. J Chromatogr A. 1376:149–58. Li LA (2007). Polychlorinated biphenyl exposure and CYP19 gene regulation in testicular and adrenocortical cell lines. Toxicol In Vitro. 21:1087–94. Li MC, Tsai PC, Chen PC, Hsieh CJ, Leon Guo YL, Rogan WJ (2013). Mortality after exposure to polychlorinated biphenyls and dibenzofurans: 30 years after the “Yucheng accident”. Environ Res. 120:71–5. Li QQ, Loganath A, Chong YS, Tan J, Obbard JP (2006). Levels of persistent organic pollutant residues in human adipose and muscle tissues in Singapore. J Toxicol Environ Health A. 69:1927–37. Li Y, Millikan RC, Bell DA, Cui L, Tse CK, Newman B et al. (2005). Polychlorinated biphenyls, cytochrome P450 1A1 (CYP1A1) polymorphisms, and breast cancer risk among African American women and white women in North Carolina: a population-based case–control study. Breast Cancer Res. 7:R12–8. WHO Food Additives Series No. 71-S1, 2016 Liem DAK (1999). Important developments in methods and techniques for the determination of dioxins and PCBs in foodstuffs and human tissues. Trends Anal Chem. 18:499–507. Liem AKD, Fürst P, Rappe C (2000). Exposure of populations to dioxins and related compounds. Food Addit Contam. 17(4):241–59. Lignell S, Aune M, Darnerud PO, Cnattingius S, Glynn A (2009). Persistent organochlorine and organobromine compounds in mother’s milk from Sweden 1996–2006: compound-specific temporal trends. Environ Res. 109(6):760–7. Lignell S, Aune M, Darnerud PO, Hanberg A, Larsson SC, Glynn A (2013). Prenatal exposure to polychlorinated biphenyls (PCBs) and polybrominated diphenyl ethers (PBDEs) may influence birth weight among infants in a Swedish cohort with background exposure: a cross-sectional study. Environ Health. 12:44.

308

Supplement 1: Non-dioxin-like polychlorinated biphenyls

Lilienthal H, Heikkinen P, Andersson PL, van der Ven LT, Viluksela M (2011). Auditory effects of developmental exposure to purity-controlled polychlorinated biphenyls (PCB52 and PCB180) in rats. Toxicol Sci. 122:100–11. Lilienthal H, Heikkinen P, Andersson PL, Viluksela M (2013). Sexually dimorphic behavior after developmental exposure to characterize endocrine-mediated effects of different non-dioxin-like PCBs in rats. Toxicology. 311:52–60. Lilienthal H, Heikkinen P, Andersson PL, van der Ven LT, Viluksela M (2014). Dopamine-dependent behavior in adult rats after perinatal exposure to purity-controlled polychlorinated biphenyl congeners (PCB52 and PCB180). Toxicol Lett. 224:32–9. Lilienthal H, Korkalainen M, Andersson PL, Viluksela M (2015). Developmental exposure to puritycontrolled polychlorinated biphenyl congeners (PCB74 and PCB95) in rats: effects on brainstem auditory evoked potentials and catalepsy. Toxicology. 327:22–31. Lin KC, Guo NW, Tsai PC, Yang CY, Guo YL (2008). Neurocognitive changes among elderly exposed to PCBs/PCDFs in Taiwan. Environ Health Perspect. 116:184–9. Lin KC, Huang PC, Yeh PS, Kuo JR, Ke DS (2010). Comparing mini-mental state examination and attention and digit span in elderly exposed to polychlorinated biphenyls and polychlorinated dibenzofurans. Psychogeriatrics. 10:191–7. Lind PM, van Bavel B, Salihovic S, Lind L (2012). Circulating levels of persistent organic pollutants (POPs) and carotid atherosclerosis in the elderly. Environ Health Perspect. 120:38–43. Lindell B (2012). 146. Polychlorinated biphenyls (PCBs). Gothenburg: University of Gothenburg, Arbete och Hälsa; The Nordic Expert Group for Criteria Documentation of Health Risks from Chemicals (NR 2012;46(1); https://gupea.ub.gu.se/bitstream/2077/28926/1/gupea_2077_28926_1.pdf, accessed 3 January 2016). Link B, Gabrio T, Zoellner I, Piechotowski I, Paepke O, Herrmann T et al. (2005). Biomonitoring of persistent organochlorine pesticides, PCDD/PCDFs and dioxin-like PCBs in blood of children from south west Germany (Baden-Wuerttemberg) from 1993 to 2003. Chemosphere. 58:1185–201. Link B, Gabrio T, Zollner I, Jaroni H, Piechotowski I, Schilling B et al. (2012). Decrease of internal exposure to chlororganic compounds and heavy metals in children in Baden-Württemberg between 1996/1997 and 2008/2009. Int J Hyg Environ Health. 215:196–201. Liu G, Zheng M, Cai M, Nie Z, Zhang B, Liu W et al. (2013). Atmospheric emission of polychlorinated biphenyls from multiple industrial thermal processes. Chemosphere. 90:2453–60. Liu S, Li S, Du Y (2010). Polychlorinated biphenyls (PCBs) enhance metastatic properties of breast cancer cells by activating Rho-associated kinase (ROCK). PLoS One. 5(6):e11272. Liu Y, Apak TI, Lehmler HJ, Robertson LW, Duffel MW (2006). Hydroxylated polychlorinated biphenyls are substrates and inhibitors of human hydroxysteroid sulfotransferase SULT2A1. Chem Res Toxicol. 19:1420–5. Liu Y, Smart JT, Song Y, Lehmler HJ, Robertson LW, Duffel MW (2009). Structure–activity relationships for hydroxylated polychlorinated biphenyls as substrates and inhibitors of rat sulfotransferases and modification of these relationships by changes in thiol status. Drug Metab Dispos. 37:1065–72. Llansola M, Montoliu C, Boix J, Felipo V (2010). Polychlorinated biphenyls PCB 52, PCB 180, and PCB 138 impair the glutamate–nitric oxide–cGMP pathway in cerebellar neurons in culture by different mechanisms. Chem Res Toxicol. 23:813–20. 309

Safety evaluation of certain food additives and contaminants Eightieth JECFA

Llobet JM, Marti-Cid R, Castell V, Domingo JL (2008). Significant decreasing trend in human dietary exposure to PCDD/PCDFs and PCBs in Catalonia, Spain. Toxicol Lett. 178:117–26. Longnecker MP, Ryan JJ, Gladen BC, Schecter AJ (2000). Correlations among human plasma levels of dioxin-like compounds and polychlorinated biphenyls (PCBs) and implications for epidemiologic studies. Arch Environ Health. 55(3):195–200. Longnecker MP, Klebanoff MA, Brock JW, Guo X (2005). Maternal levels of polychlorinated biphenyls in relation to preterm and small-for-gestational-age birth. Epidemiology. 16:641–7. Lopes B, Arrebola JP, Serafim A, Company R, Rosa J, Olea N (2014). Polychlorinated biphenyls (PCBs) and p,p′-dichlorodiphenyldichloroethylene (DDE) concentrations in maternal and umbilical cord serum in a human cohort from south Portugal. Chemosphere. 114:291–302. Lorán S, Conchello P, Bayarri S, Herrera S (2009). Evaluation of daily intake of PCDD/Fs and indicator PCBs in formula-fed Spanish children. Food Addit Contam. 26(10):1421–31. Lorber M (2008). Exposure of Americans to polybrominated diphenyl ethers. J Expo Sci Environ Epidemiol. 18:2–19. Louis GM, Weiner JM, Whitcomb BW, Sperrazza R, Schisterman EF, Lobdell DT et al. (2005). Environmental PCB exposure and risk of endometriosis. Hum Reprod. 20:279–85. Loutfy N, Fuerhacker M, Lesueur C, Gartner M, Ahmed MT, Mentler A (2008). Pesticide and non-dioxinlike polychlorinated biphenyls (NDL-PCBs) residues in foodstuffs from Ismailia city, Egypt. Food Addit Contam B Surveill. 1:32–40. Ludewig G, Robertson LW (2013). Polychlorinated biphenyls (PCBs) as initiating agents in hepatocellular carcinoma. Cancer Lett. 334:46–55. Lundberg R, Lyche JL, Ropstad E, Aleksandersen M, Rönn M, Skaare JU et al. (2006). Perinatal exposure to PCB 153, but not PCB 126, alters bone tissue composition in female goat offspring. Toxicology. 228(1):33–40. Luthe G, Jacobus JA, Robertson LW (2008). Receptor interactions by polybrominated diphenyl ethers versus polychlorinated biphenyls: a theoretical structure–activity assessment. Environ Toxicol Pharmacol. 25:202–10. WHO Food Additives Series No. 71-S1, 2016 Lyche JL, Oskam IC, Skaare JU, Reksen O, Sweeney T, Dahl E et al. (2004a). Effects of gestational and lactational exposure to low doses of PCBs 126 and 153 on anterior pituitary and gonadal hormones and on puberty in female goats. Reprod Toxicol. 19:87–95. Lyche JL, Larsen HJS, Skaare JU, Tverdal A, Dahl E, Johansen GM et al. (2004b). Effects of perinatal exposure to low doses of PCB 153 and PCB 126 on lymphocyte proliferation and hematology in goat kids. J Toxicol Environ Health A. 67:889–904. Lyche JL, Larsen HJS, Skaare JU, Tverdal A, Johansen GM, Ropstad E (2006). Perinatal exposure to low doses of PCB 153 and PCB 126 affects maternal and neonatal immunity in goat kids. J Toxicol Environ Health A. 69:139–58. Lynch CD, Jackson LW, Kostyniak PJ, McGuinness BM, Buck Louis GM (2012). The effect of prenatal and postnatal exposure to polychlorinated biphenyls and child neurodevelopment at age twenty four months. Reprod Toxicol. 34:451–6.

310

Supplement 1: Non-dioxin-like polychlorinated biphenyls

Machala M, Bláha LK, Vondráček J, Trosko JE, Scott J, Upham BL (2003). Inhibition of gap junctional intercellular communication by noncoplanar polychlorinated biphenyls: inhibitory potencies and screening for potential mode(s) of action. Toxicol Sci. 76:102–11. Machala M, Bláha L, Lehmler HJ, Plísková M, Májková Z, Kapplová P et al. (2004). Toxicity of hydroxylated and quinoid PCB metabolites: inhibition of gap junctional intercellular communication and activation of aryl hydrocarbon and estrogen receptors in hepatic and mammary cells. Chem Res Toxicol. 17:340–7. Maervoet J, Vermeir G, Covaci A, Van Larebeke N, Koppen G, Schoeters G et al. (2007). Association of thyroid hormone concentrations with levels of organochlorine compounds in cord blood of neonates. Environ Health Perspect. 115:1780–6. Maifredi G, Donato F, Magoni M, Orizio G, Gelatti U, Maiolino P et al. (2011). Polychlorinated biphenyls and non-Hodgkin’s lymphoma: a case–control study in northern Italy. Environ Res. 111:254–9. Malarvannan G, Isobe T, Covaci A, Prudente M, Tanabe S (2013). Accumulation of brominated flame retardants and polychlorinated biphenyls in human breast milk and scalp hair from the Philippines: levels, distribution and profiles. Sci Total Environ. 442:366–79. Malisch R, Dilara P (2007). PCDD/Fs and PCBs in butter samples from new European Union member states and a candidate country: analytical quality control, results and certain PCB-specific aspects. Chemosphere. 67:S79–89. Malisch R, Kotz A (2014). Dioxins and PCBs in feed and food – Review from European perspective. Sci Total Environ. 491:2–10. Malisch R, Kotz A, Wahl K (2009). Analytical requirements for determination of non dioxin-like PCBs in food (PCB # 28, 52, 101, 138, 153, 180). Organohalogen Compounds. 71:2123–6. Mandalakis M, Holmstrand H, Andersson P, Gustafsson O (2008). Compound-specific chlorine isotope analysis of polychlorinated biphenyls isolated from Aroclor and Clophen technical mixtures. Chemosphere. 71:299–305. Mansour SA (2011). Chemical pollutants threatening food safety and security: an overview. In: Hefnawy M, editor. Advances in food protection. Focus on food safety and defense. NATO Science for Peace and Security Series A: Chemistry and Biology. Springer Science+Business Media B.V.; 73–117. Marabini L, Calò R, Fucile S (2011). Genotoxic effects of polychlorinated biphenyls (PCB 153, 138, 101, 118) in a fish cell line (RTG-2). Toxicol In Vitro. 25:1045–52. Maranghi F, Tassinari R, Moracci G, Altieri I, Rasinger JD, Carroll TS et al. (2013). Dietary exposure of juvenile female mice to polyhalogenated seafood contaminants (HBCD, BDE-47, PCB-153, TCDD): comparative assessment of effects in potential target tissues. Food Chem Toxicol. 56:443–9. Mariussen E, Morch Andersen J, Fonnum F (1999). The effect of polychlorinated biphenyls on the uptake of dopamine and other neurotransmitters into rat brain synaptic vesicles. Toxicol Appl Pharmacol. 161:274–82. Marmon SK, Liljelind P, Undeland I (2009). Removal of lipids, dioxins, and polychlorinated biphenyls during production of protein isolates from Baltic herring (Clupea harengus) using pH-shift processes. J Agric Food Chem. 57:7819–25. Martinez JM, Stephens LC, Jones LA (2005). Long-term effects of neonatal exposure to hydroxylated polychlorinated biphenyls in the BALB/cCrgl mouse. Environ Health Perspect. 113:1022–6. 311

Safety evaluation of certain food additives and contaminants Eightieth JECFA

Masci M, Orban E, Nevigato T (2015). Fish contamination by polychlorobiphenyls: the mass spectrometric ortho effect in a new and easy gas chromatography–mass spectrometry method for the analysis of the seven indicators. The case of bluefin tuna. J Chromatogr A. 1375:110–22. Masuda Y (2001). Fate of PCDF/PCB congeners and change of clinical symptoms in patients with Yusho PCB poisoning for 30 years. Chemosphere. 43:925–30. Masuda Y, Haraguchi K, Kono S, Tsuji H, Papke O (2005). Concentrations of dioxins and related compounds in the blood of Fukuoka residents. Chemosphere. 58:329–44. Matthews HB, Dedrick RL (1984). Pharmacokinetics of PCBs. Annu Rev Pharmacol Toxicol. 24:85–103. Mayes BA, McConnell EE, Neal BH, Brunner MJ, Hamilton SB, Sullivan TM et al. (1998). Comparative carcinogenicity in Sprague-Dawley rats of the polychlorinated biphenyl mixtures Aroclors 1016, 1242, 1254, and 1260. Toxicol Sci. 41:62–76. McAuliffe ME, Williams PL, Korrick SA, Altshul LM, Perry MJ (2012). Environmental exposure to polychlorinated biphenyls and p,p′-DDE and sperm sex-chromosome disomy. Environ Health Perspect. 120:535–40. McGlynn KA, Quraishi SM, Graubard BI, Weber JP, Rubertone MV, Erickson RL (2009a). Polychlorinated biphenyls and risk of testicular germ cell tumors. Cancer Res. 69:1901–9. McGlynn KA, Guo X, Graubard BI, Brock JW, Klebanoff MA, Longnecker MP (2009b). Maternal pregnancy levels of polychlorinated biphenyls and risk of hypospadias and cryptorchidism in male offspring. Environ Health Perspect. 117:1472–6. McGraw JE Sr, Waller DP (2006). Specific human CYP 450 isoform metabolism of a pentachlorobiphenyl (PCB-IUPAC# 101). Biochem Biophys Res Commun. 344:129–33. McLachlan MS (1993). Digestive tract absorption of polychlorinated dibenzo-p-dioxins, dibenzofurans, and biphenyls in a nursing infant. Toxicol Appl Pharmacol. 123:68–72. Meeker JD, Altshul L, Hauser R (2007). Serum PCBs, p,p′-DDE and HCB predict thyroid hormone levels in men. Environ Res. 104:296–304. Meerts IA, Assink Y, Cenijn PH, Van Den Berg JH, Weijers BM, Bergman A et al. (2002). Placental transfer of a hydroxylated polychlorinated biphenyl and effects on fetal and maternal thyroid hormone homeostasis in the rat. Toxicol Sci. 68:361–71. WHO Food Additives Series No. 71-S1, 2016 Meerts IA, Hoving S, van den Berg JH, Weijers BM, Swarts HJ, van der Beek EM et al. (2004). Effects of in utero exposure to 4-hydroxy-2,3,3′,4′,5-pentachlorobiphenyl (4-OH-CB107) on developmental landmarks, steroid hormone levels, and female estrous cyclicity in rats. Toxicol Sci. 82:259–67. Meisner LF, Roloff B, Sargent L, Pitot H (1992). Interactive cytogenetic effects on rat bone marrow due to chronic ingestion of 2,5,2′,5′ and 3,4,3′,4′ PCBs. Mutat Res. 283:179–83. Melis M, Zuccato E (2012). Chapter 2. Analysis of PCBs in food. In: Nollet LML, editor. Analysis of endocrine disrupting compounds in food. Ames (IA): Blackwell Publishing Ltd; 19–48. Mendez MA, Garcia-Esteban R, Guxens M, Vrijheid M, Kogevinas M, Goñi F et al. (2011). Prenatal organochlorine compound exposure, rapid weight gain, and overweight in infancy. Environ Health Perspect. 119:272–8. Mes J, Davies DJ (1979). Presence of polychlorinated biphenyl and organochlorine pesticide residues and the absence of polychlorinated terphenyls in Canadian human milk samples. Bull Environ Contam Toxicol. 21:381–7.

312

Supplement 1: Non-dioxin-like polychlorinated biphenyls

Meyer HW, Frederiksen M, Göen T, Ebbehøj NE, Gunnarsen L, Brauer C et al. (2013). Plasma polychlorinated biphenyls in residents of 91 PCB-contaminated and 108 non-contaminated dwellings – An exposure study. Int J Hyg Environ Health. 216(6):755–62. Mihats D, Moche W, Prean M, Rauscher-Gabernig E (2015). Dietary exposure to non-dioxin-like PCBs of different population groups in Austria. Chemosphere. 126:53–9. Mikoczy Z, Rylander L (2009). Mortality and cancer incidence in cohorts of Swedish fishermen and fishermen’s wives: updated findings. Chemosphere. 74:938–43. Mills SA III, Thal DI, Barney J (2007). A summary of the 209 PCB congener nomenclature. Chemosphere. 68:1603–12. Min JY, Kim R, Min KB (2014). Serum polychlorinated biphenyls concentrations and hearing impairment in adults. Chemosphere. 102:6–11. Ministry of Health of the People’s Republic of China (2012). National Food Safety Standard. Maximum levels of contaminants in food. National Standard of the People’s Republic of China. GB 2762-2012 (http://www.seafish.org/media/publications/China_Max_levels_of_contaminants_in_food.pdf, accessed 3 January 2016). Ministry of Health of the People’s Republic of China (2014). National Food Safety Standard. Determination of indicator polychlorinated biphenyls in foods. National Standard of the People’s Republic of China. GB 5009.190-2014. Morck TA, Erdmann SE, Long M, Mathiesen L, Nielsen F, Siersma VD et al. (2014). PCB concentrations and dioxin-like activity in blood samples from Danish school children and their mothers living in urban and rural areas. Basic Clin Pharmacol Toxicol. 115:134–44. Mori C, Morsey B, Levin M, Nambiar PR, DeGuise S (2006). Immunomodulatory effects of in vitro exposure to organochlorines on T-cell proliferation in marine mammals and mice. J Toxicol Environ Health A. 69:282–302. Mori C, Morsey B, Levin M, Gorton TS, DeGuise S (2008). Effects of organochlorines, individually and in mixtures, on B cell proliferation in marine mammals and mice. J Toxicol Environ Health A. 71:266–75. Moser GA, McLachlan MS (2001). The influence of dietary concentration on the absorption and excretion of persistent lipophilic organic pollutants in the human intestinal tract. Chemosphere. 45:201–11. Mrema E, Rubino F, Mandic-Rajcevic S, Sturchio E, Turci R, Osculati A et al. (2014). Exposure to priority organochlorine contaminants in the Italian general population. Part 2: Fifteen priority polychlorinated biphenyl congeners in blood serum. Human Exp Toxicol. 33:170–84. Muckle G, Ayotte P, Dewailly E, Jacobson SW, Jacobson JL (2001). Prenatal exposure of the northern Québec Inuit infants to environmental contaminants. Environ Health Perspect. 109:1291–9. Muir D, Sverko E (2006). Analytical methods for PCBs and organochlorine pesticides in environmental monitoring and surveillance: a critical appraisal. Anal Bioanal Chem. 386:769–89. Mullin MD, Pochini CM, McCrindle S, Romkes M, Safe SH, Safe LM (1984). High-resolution PCB analysis: synthesis and chromatographic properties of all 209 PCB congeners. Environ Sci Technol. 18:468–76. Murk AJ, Rijntjes E, Blaauboer BJ, Clewell R, Crofton KM, Dingemans MM et al. (2013). Mechanismbased testing strategy using in vitro approaches for identification of thyroid hormone disrupting chemicals. Toxicol In Vitro. 27:1320–46.

313

Safety evaluation of certain food additives and contaminants Eightieth JECFA

Murphy LE, Gollenberg AL, Buck Louis GM, Kostyniak PJ, Sundaram R (2010). Maternal serum preconception polychlorinated biphenyl concentrations and infant birth weight. Environ Health Perspect. 118:297–302. Naert C, Piette M, Bruneel N, Van Peteghem C (2006). Occurrence of polychlorinated biphenyls and polybrominated diphenyl ethers in Belgian human adipose tissue samples. Arch Environ Contam Toxicol. 50:290–6. Nagayama J, Kohno H, Kunisue T, Kataoka K, Shimomura H, Tanabe S et al. (2007). Concentrations of organochlorine pollutants in mothers who gave birth to neonates with congenital hypothyroidism. Chemosphere. 68:972–6. Nakajima S, Saijo Y, Kato S, Sasaki S, Uno A, Kanagami N et al. (2006). Effects of prenatal exposure to polychlorinated biphenyls and dioxins on mental and motor development in Japanese children at 6 months of age. Environ Health Perspect. 114:773–8. Ndountse LT, Chan HM (2009). Role of N-methyl-D-aspartate receptors in polychlorinated biphenyl mediated neurotoxicity. Toxicol Lett. 184:50–5. Needham LL, Barr DB, Caudill SP, Pirkle JL, Turner WE, Osterloh J et al. (2005). Concentrations of environmental chemicals associated with neurodevelopmental effects in U.S. population. Neurotoxicology. 26:531–45. Ness DK, Schantz SL, Moshtaghian J, Hansen LG (1993). Effects of perinatal exposure to specific PCB congeners on thyroid hormone concentrations and thyroid histology in the rat. Toxicol Lett. 68:311–23. Newman J, Aucompaugh AG, Schell LM, Denham M, DeCaprio AP, Gallo MV et al. (2006). PCBs and cognitive functioning of Mohawk adolescents. Neurotoxicol Teratol. 28:439–45. Newman J, Gallo MV, Schell LM, DeCaprio AP, Denham M, Deane GD; Akwesasne Task Force on the Environment (2009). Analysis of PCB congeners related to cognitive functioning in adolescents. Neurotoxicology. 30:686–96. Newsome WH, Davies DJ, Sun WF (1998). Residues of polychlorinated biphenyls (PCB) in fatty foods of the Canadian diet. Food Addit Contam. 15:19–29. Ng E, Salihovic S, Lind PM, Mahajan A, Syvänen AC, Axelsson T et al. (2015). Genome-wide association study of plasma levels of polychlorinated biphenyls disclose an association with the CYP2B6 gene in a population-based sample. Environ Res. 140:95–101. Nie Z, Liu G, Liu W, Zhang B, Zheng M (2012). Characterization and quantification of unintentional POP emissions from primary and secondary copper metallurgical processes in China. Atmos Environ. 57:109–15. Niskar AS, Needham LL, Rubin C, Turner WE, Martin CA, Patterson DG Jr et al. (2009). Serum dioxins, polychlorinated biphenyls, and endometriosis: a case–control study in Atlanta. Chemosphere. 74:944–9. Noma Y, Yamamoto T, Falandysz J, Mostrag A, Sakai S (2006). By-side chlorobenzenes and chlorophenols in technical chlorobiphenyl formulations of Aroclor 1268, Chlorofen, Clophen T 64, Kanechlor 600, and Kanechlor 1000. J Environ Sci Health A Tox Hazard Subst Environ Eng. 41:35–46. Nøst TH, Breivik K, Fuskevåg O-M, Nieboer E, Øyvind Odland J, Manning Sandanger T (2013). Persistent organic pollutants in Norwegian men from 1979 to 2007: intraindividual changes, age–period–cohort effects, and model predictions. Environ Health Perspect. 121:1292–8.

314

WHO Food Additives Series No. 71-S1, 2016

Supplement 1: Non-dioxin-like polychlorinated biphenyls

NTP (2006a). Toxicology and carcinogenesis studies of 2,2′,4,4′,5,5′-hexachlorobiphenyl (PCB 153) (CAS No. 35065-27-1) in female Harlan Sprague-Dawley rats (gavage studies). Research Triangle Park (NC): United States Department of Health and Human Services, Public Health Service, National Institutes of Health, National Toxicology Program (Technical Report Series No. 529; NIH Publication No. 06-4465). NTP (2006b).Toxicology and carcinogenesis studies of a binary mixture of 3,3′,4,4′,5-pentachlorobiphenyl (PCB 126) (CAS No. 57465-28-8) and 2,2′,4,4′,5,5′-hexachlorobiphenyl (PCB 153) (CAS No. 3506527-1) in female Harlan Sprague-Dawley rats (gavage studies). Research Triangle Park (NC): United States Department of Health and Human Services, Public Health Service, National Institutes of Health, National Toxicology Program (Technical Report Series No. 530; NIH Publication No. 06-4466). NTP (2006c). Toxicology and carcinogenesis studies of 3,3′,4,4′,5 pentachlorobiphenyl (PCB 126) (CAS No. 57465-28-8) in female Harlan Sprague-Dawley rats (gavage studies). Research Triangle Park (NC): United States Department of Health and Human Services, Public Health Service, National Institutes of Health, National Toxicology Program (Technical Report Series No. 520; NIH Publication No. 06-4454). Oliveira Ribeiro CA, Vollaire Y, Coulet E, Roche H (2008). Bioaccumulation of polychlorinated biphenyls in the eel (Anguilla anguilla) at the Camargue Nature Reserve – France. Environ Pollut. 153:424–31. Olivero-Verbel J, Ganey PE (1998). Effect of three polychlorinated biphenyls on f-met-leu-phe-induced degranulation in rat neutrophils. Toxicol Lett. 98:195–202. Onozuka D, Hirata T, Furue M (2011). Relative survival after exposure to polychlorinated biphenyls and dioxins: a follow-up of Japanese patients affected in the Yusho incident. Sci Total Environ. 409:2361–5. Onozuka D, Hirata T, Furue M (2014). Net survival after exposure to polychlorinated biphenyls and dioxins: the Yusho study. Environ Int. 73:28–32. Onozuka D, Yoshimura T, Kaneko S, Furue M (2009). Mortality after exposure to polychlorinated biphenyls and polychlorinated dibenzofurans: a 40-year follow-up study of Yusho patients. Am J Epidemiol. 169:86–95. Oterhals Å, Kvamme B (2013). Optimization of an oil leaching process to reduce the level of dioxins and dioxin-like PCBs in fishmeal. J Sci Food Agric. 93:1649–59. Oterhals Å, Berntssen MHG (2010). Effects of refining and removal of persistent organic pollutants by short-path distillation on nutritional quality and oxidative stability of fish oil. J Agric Food Chem. 58:12250–9. Oterhals Å, Solvang M, Nortvedt R, Berntssen MHG (2007). Optimization of activated carbon-based decontamination of fish oil by response surface methodology. Eur J Lipid Sci Technol. 109:691–705. Ounnas F, Feidt C, Toussaint H, Marchand P, Le Bizec B, Rychen G et al. (2010). Polychlorinated biphenyl and low polybrominated diphenyl ether transfer to milk in lactating goats chronically exposed to contaminated soil. Environ Sci Technol. 44:2682–8. Overmann SR, Kostas J, Wilson LR, Shain W, Bush B (1987). Neurobehavioral and somatic effects of perinatal PCB exposure in rats. Environ Res. 44:56–70. Pandelova M, Schramm K-W (2012). Human and environmental biomonitoring of polychlorinated biphenyls and hexachlorobenzene in Saxony, Germany based on the German Environmental Specimen Bank. Int J Hyg Environ Health. 215:220–3.

315

Safety evaluation of certain food additives and contaminants Eightieth JECFA

Papadopoulou E, Caspersen IH, Kvalem HE, Knutsen HK, Duarte-Salles T, Alexander J et al. (2013). Maternal dietary intake of dioxins and polychlorinated biphenyls and birth size in the Norwegian Mother and Child Cohort Study (MoBa). Environ Int. 60:209–16. Park H, Lee SJ, Kang JH, Chang YS (2007). Congener-specific approach to human PCB concentrations by serum analysis. Chemosphere. 68:1699–1706. Park HY, Hertz-Picciotto I, Petrik J, Palkovicova L, Kocan A, Trnovec T (2008). Prenatal PCB exposure and thymus size at birth in neonates in eastern Slovakia. Environ Health Perspect. 116:104–9. Park HY, Park JS, Sovcikova E, Kocan A, Linderholm L, Bergman A et al. (2009). Exposure to hydroxylated polychlorinated biphenyls (OH-PCBs) in the prenatal period and subsequent neurodevelopment in eastern Slovakia. Environ Health Perspect. 117:1600–6. Park HY, Hertz-Picciotto I, Sovcikova E, Kocan A, Drobna B, Trnovec T (2010). Neurodevelopmental toxicity of prenatal polychlorinated biphenyls (PCBs) by chemical structure and activity: a birth cohort study. Environ Health. 9:51. Patterson DG Jr, Todd GD, Turner WE, Maggio V, Alexander LR, Needham LL (1994). Levels of nonortho-substituted (coplanar), mono- and di-ortho-substituted polychlorinated biphenyls, dibenzo-pdioxins, and dibenzofurans in human serum and adipose tissue. Environ Health Perspect. 102(Suppl. 1):195–204. Patterson DG Jr, Barr JR, DiPietro ES, Granger J, Green VE, Lapeza CR Jr et al. (1996). Comprehensive two-dimensional and fast single-dimensional GC-HRMS analysis of human serum for PCDDs, PCDFs, cPCBs, congener PCBs, persistent and nonpersistent pesticides, and PAHs. Organohalogen Compounds. 27:309–14. Patterson DG Jr, Wong LY, Turner WE, Caudill SP, Dipietro ES, McClure PC et al. (2009). Levels in the U.S. population of those persistent organic pollutants (2003–2004) included in the Stockholm Convention or in other long range transboundary air pollution agreements. Environ Sci Technol. 43(4):1211–8. Pavuk M, Olson JR, Sjodin A, Wolff P, Turner WE, Shelton C et al. (2014). Serum concentrations of polychlorinated biphenyls (PCBs) in participants of the Anniston Community Health Survey. Sci Total Environ. 473–474:286–97. Pereg D, Tampal N, Espandiari P, Robertson LW (2001). Distribution and macromolecular binding of benzo[a]pyrene and two polychlorinated biphenyl congeners in female mice. Chem Biol Interact. 137:243–58. Persky V, Piorkowski J, Turyk M, Freels S, Chatterton R Jr, Dimos J et al. (2011). Associations of polychlorinated biphenyl exposure and endogenous hormones with diabetes in post-menopausal women previously employed at a capacitor manufacturing plant. Environ Res. 111:817–24. Pesatori AC, Grillo P, Consonni D, Caironi M, Sampietro G, Olivari L et al. (2013). Update of the mortality study of workers exposed to polychlorinated biphenyls (PCBs) in two Italian capacitor manufacturing plants. Med Lav. 104:107–14. Pessah IN, Cherednichenko G, Lein PJ (2010). Minding the calcium store: ryanodine receptor activation as a convergent mechanism of PCB toxicity. Pharmacol Ther. 125:260–85. Pessah IN, Hansen LG, Albertson TE, Garner CE, Ta TA, Do Z et al. (2006). Structure–activity relationship for noncoplanar polychlorinated biphenyl congeners toward the ryanodine receptor–Ca2+ channel complex type 1 (RyR1). Chem Res Toxicol. 19:92–101.

316

WHO Food Additives Series No. 71-S1, 2016

Supplement 1: Non-dioxin-like polychlorinated biphenyls

Petersen MS, Halling J, Damkier P, Nielsen F, Grandjean P, Weihe P et al. (2007). Polychlorinated biphenyl (PCB) induction of CYP3A4 enzyme activity in healthy Faroese adults. Toxicol Appl Pharmacol. 224:202–6. Petrik J, Drobna B, Pavuk M, Jursa S, Wimmerova S, Chovancova J (2006). Serum PCBs and organochlorine pesticides in Slovakia: age, gender, and residence as determinants of organochlorine concentrations. Chemosphere. 65:410–8. Philibert A, Schwartz H, Mergler D (2009). An exploratory study of diabetes in a First Nation community with respect to serum concentrations of p,p′-DDE and PCBs and fish consumption. Int J Environ Res Public Health. 6:3179–89. Piedrafita B, Erceg S, Cauli O, Felipo V (2008a). Developmental exposure to polychlorinated biphenyls or methylmercury, but not to its combination, impairs the glutamate–nitric oxide–cyclic GMP pathway and learning in 3-month-old rats. Neuroscience. 154:1408–16. Piedrafita B, Erceg S, Cauli O, Monfort P, Felipo V (2008b). Developmental exposure to polychlorinated biphenyls PCB153 or PCB126 impairs learning ability in young but not in adult rats. Eur J Neurosci. 27:177–82. Pietrogrande MC, Michi M, Nunez Plasencia M, Dondi F (2002). Analysis of PCB by on-line coupled HPLC-HRGC. Chromatographia. 55:189–96. Pirard C, Focant JF, De PE (2002). An improved clean-up strategy for simultaneous analysis of polychlorinated dibenzo-p-dioxins (PCDD), polychlorinated dibenzofurans (PCDF), and polychlorinated biphenyls (PCB) in fatty food samples. Anal Bioanal Chem. 372:373–81. Polder A, Becher G, Savinova TN, Skaare JU (1998). Dioxins, PCBs and some chlorinated pesticides in human milk from the Kola Peninsula, Russia. Chemosphere. 37:1795–806. Polder A, Thomsen C, Lindstrom G, Loken KB, Skaare JU (2008a). Levels and temporal trends of chlorinated pesticides, polychlorinated biphenyls and brominated flame retardants in individual human breast milk samples from northern and southern Norway. Chemosphere. 73:14–23. Polder A, Gabrielsen GW, Odland JO, Savinova TN, Tkachev A, Loken KB et al. (2008b). Spatial and temporal changes of chlorinated pesticides, PCBs, dioxins (PCDDs/PCDFs) and brominated flame retardants in human breast milk from northern Russia. Sci Total Environ. 391:41–54. Polsky JY, Aronson KJ, Heaton JP, Adams MA (2007). Pesticides and polychlorinated biphenyls as potential risk factors for erectile dysfunction. J Androl. 28:28–37. Poon E, Bandara SB, Allen JB, Sadowski RN, Schantz SL (2015). Developmental PCB exposure increases susceptibility to audiogenic seizures in adulthood. Neurotoxicology. 46:117–24. Porpora MG, Ingelido AM, di Domenico A, Ferro A, Crobu M, Pallante D et al. (2006). Increased levels of polychlorobiphenyls in Italian women with endometriosis. Chemosphere. 63:1361–7. Porpora MG, Medda E, Abballe A, Bolli S, De Angelis I, di Domenico A et al. (2009). Endometriosis and organochlorinated environmental pollutants: a case–control study on Italian women of reproductive age. Environ Health Perspect. 117:1070–5. Porta M, López T, Gasull M, Rodríguez-Sanz M, Garí M, Pumarega J et al. (2012). Distribution of blood concentrations of persistent organic pollutants in a representative sample of the population of Barcelona in 2006, and comparison with levels in 2002. Sci Total Environ. 423:51–61.

317

Safety evaluation of certain food additives and contaminants Eightieth JECFA

Powers BE, Widholm JJ, Lasky RE, Schantz SL (2006). Auditory deficits in rats exposed to an environmental PCB mixture during development. Toxicol Sci. 89:415–22. Preau L, Fini JB, Morvan-Dubois G, Demeneix B (2015). Thyroid hormone signaling during early neurogenesis and its significance as a vulnerable window for endocrine disruption. Biochim Biophys Acta. 1849:112–21. Prince MM, Hein MJ, Ruder AM, Waters MA, Laber PA, Whelan EA (2006a). Update: cohort mortality study of workers highly exposed to polychlorinated biphenyls (PCBs) during the manufacture of electrical capacitors, 1940–1998. Environ Health. 5:13. Prince MM, Ruder AM, Hein MJ, Waters MA, Whelan EA, Nilsen N et al. (2006b). Mortality and exposure response among 14,458 electrical capacitor manufacturing workers exposed to polychlorinated biphenyls (PCBs). Environ Health Perspect. 114:1508–14. Pulkrabová J, Hrádková P, Hajslová J, Poustka J, Nápravníková M, Polácek V (2009). Brominated flame retardants and other organochlorine pollutants in human adipose tissue samples from the Czech Republic. Environ Int. 35:63–8. Purdue MP, Engel LS, Langseth H, Needham LL, Andersen A, Barr DB et al. (2009). Prediagnostic serum concentrations of organochlorine compounds and risk of testicular germ cell tumors. Environ Health Perspect. 117:1514–9. Quinete N, Schettgen T, Bertram J, Kraus T (2014). Occurrence and distribution of PCB metabolites in blood and their potential health effects in humans: a review. Environ Sci Pollut Res. 21:11951–72. Quinn CL, Wania F (2012). Understanding differences in the body burden–age relationships of bioaccumulating contaminants based on population cross sections versus individuals. Environ Health Perspect. 120:554–9. Raaschou-Nielsen O, Pavuk M, LeBlanc A, Dumas P, Weber JP, Olsen A et al. (2005). Adipose organochlorine concentrations and risk of breast cancer among postmenopausal Danish women. Cancer Epidemiol Biomarkers Prev. 14:67–74. Ramos L, Hernández LM, González MJ (1999). Simultaneous separation of coplanar and chiral polychlorinated biphenyls by off-line pyrenyl-silica liquid chromatography and gas chromatography. Enantiomeric ratios of chiral congeners. Anal Chem. 71:70–7. WHO Food Additives Series No. 71-S1, 2016 Rawn DFK, Sadler AR, Quade SC, Sun W-F, Kosarac I, Hayward S et al. (2012). The impact of production type and region on polychlorinated biphenyl (PCB), polychlorinated dibenzo-p-dioxin and dibenzofuran (PCDD/F) concentrations in Canadian chicken egg yolks. Chemosphere. 89:929–35. Rawn DFK, Breakell K, Verigin V, Tittlemier SA, Del Gobbo L, Diamond M et al. (2013). Impacts of cooking technique on polychlorinated biphenyl and polychlorinated dioxins/furan concentrations in fish and fish products with intake estimates. J Agric Food Chem. 61:989–97. Recio-Vega R, Velazco-Rodriguez V, Ocampo-Gómez G, Hernandez-Gonzalez S, Ruiz-Flores P, LopezMarquez F (2011). Serum levels of polychlorinated biphenyls in Mexican women and breast cancer risk. J Appl Toxicol. 31:270–8. Recio-Vega R, Mendez-Henandez A, Gabriel AP, Jacobo-Avila A, Portales-Castanedo A, HernandezGonzalez S et al. (2013). Potentially estrogenic polychlorinated biphenyls congeners serum levels and its relation with lung cancer. J Appl Toxicol. 33:906–14.

318

Supplement 1: Non-dioxin-like polychlorinated biphenyls

Redding LE (2010). Population physiologically-based pharmacokinetic modeling for the human lactational transfer of PCB 153 with consideration of worldwide human biomonitoring results. Berkeley (CA): Lawrence Berkeley National Laboratory. Redding LE, Sohn MD, McKone TE, Chen J-W, Wang S-L, Hsieh DPH et al. (2008). Population physiologically based pharmacokinetic modeling for the human lactational transfer of PCB-153 with consideration of worldwide human biomonitoring results. Environ Health Perspect. 116:1629–34. Reddy BS, Rozati R, Reddy S, Kodampur S, Reddy P, Reddy R (2006). High plasma concentrations of polychlorinated biphenyls and phthalate esters in women with endometriosis: a prospective case control study. Fertil Steril. 85:775–9. Richardson VM, Staskal DF, Ross DG, Diliberto JJ, DeVito MJ, Birnbaum LS (2008). Possible mechanisms of thyroid hormone disruption in mice by BDE 47, a major polybrominated diphenyl ether congener. Toxicol Appl Pharmacol. 226:244–50. Ridgway K, Lalljie SPD, Smith RM (2007). Sample preparation techniques for the determination of trace residues and contaminants in foods. J Chromatogr A. 1153:36–53. Rignall B, Grote K, Gavrilov A, Weimer M, Kopp-Schneider A, Krause E et al. (2013). Biological and tumor-promoting effects of dioxin-like and non-dioxin-like polychlorinated biphenyls in mouse liver after single or combined treatment. Toxicol Sci. 133:29–41. Rignell-Hydbom A, Rylander L, Hagmar L (2007). Exposure to persistent organochlorine pollutants and type 2 diabetes mellitus. Hum Exp Toxicol. 26:447–52. Rignell-Hydbom A, Lidfeldt J, Kiviranta H, Rantakokko P, Samsioe G, Agardh CD et al. (2009). Exposure to p,p′-DDE: a risk factor for type 2 diabetes. PLoS One. 4:e7503. Rignell-Hydbom A, Elfving M, Ivarsson SA, Lindh C, Jönsson BA, Olofsson P et al. (2010). A nested case–control study of intrauterine exposure to persistent organochlorine pollutants in relation to risk of type 1 diabetes. PLoS One. 5:e11281. Ritchie JM, Vial SL, Fuortes LJ, Robertson LW, Guo H, Reedy VE et al. (2005). Comparison of proposed frameworks for grouping polychlorinated biphenyl congener data applied to a case–control pilot study of prostate cancer. Environ Res. 98:104–13. Ritter R, Scheringer M, MacLeod M, Moeckel C, Jones KC, Hungerbühler K (2011). Intrinsic human elimination half-lives of polychlorinated biphenyls derived from the temporal evolution of crosssectional biomonitoring data from the United Kingdom. Environ Health Perspect. 119:225–31. Robertson LW, Gupta RC (2000). Metabolism of polychlorinated biphenyls (PCBs) generates electrophiles and reactive oxygen species that damage DNA. In: Williams GM, Aruoma OI, editors. Molecular drug metabolism and toxicology. OICA International (UK); 16–32. Robertson LW, Ludewig G (2011). Polychlorinated biphenyl (PCB) carcinogenicity with special emphasis on airborne PCBs. Gefahrst Reinhalt Luft. 71:25–32. Robertson LW, Berberian I, Borges T, Chen LC, Chow CK, Glauert HP et al. (2007). Suppression of peroxisomal enzyme activities and cytochrome P450 4A isozyme expression by congeneric polybrominated and polychlorinated biphenyls. PPAR Res. 2007:15481. Rocheleau CM, Bertke SJ, Deddens JA, Ruder AM, Lawson CC, Waters MA et al. (2011). Maternal exposure to polychlorinated biphenyls and the secondary sex ratio: an occupational cohort study. Environ Health. 10:20. 319

Safety evaluation of certain food additives and contaminants Eightieth JECFA

Röllin HB, Sandanger TM, Hansen L, Channa K, Odland JØ (2009). Concentration of selected persistent organic pollutants in blood from delivering women in South Africa. Sci Total Environ. 408:146–52. Roos R, Andersson PL, Halldin K, Håkansson H, Westerholm E, Hamers T et al. (2011). Hepatic effects of a highly purified 2,2′,3,4,4′,5,5′-heptachlorbiphenyl (PCB 180) in male and female rats. Toxicology. 284:42–53. Rubin CH, Lanier A, Kieszak S, Brock JW, Koller KR, Strosnider H et al. (2006). Breast cancer among Alaska Native women potentially exposed to environmental organochlorine chemicals. Int J Circumpolar Health. 65:18–27. Ruder AM, Hein MJ, Nilsen N, Waters MA, Laber P, Davis-King K et al. (2006). Mortality among workers exposed to polychlorinated biphenyls (PCBs) in an electrical capacitor manufacturing plant in Indiana: an update. Environ Health Perspect. 114:18–23. Ruder AM, Hein MJ, Hopf NB, Waters MA (2014). Mortality among 24,865 workers exposed to polychlorinated biphenyls (PCBs) in three electrical capacitor manufacturing plants: a ten-year update. Int J Hyg Environ Health. 217:176–87. Ruiz P, Faroon O, Moudgal CJ, Hansen H, De Rosa CT, Mumtaz M (2008). Prediction of the health effects of polychlorinated biphenyls (PCBs) and their metabolites using quantitative structure–activity relationship (QSAR). Toxicol Lett. 181:53–65. Runge-Morris M, Kocarek TA, Falany CN (2013). Regulation of the cytosolic sulfotransferases by nuclear receptors. Drug Metab Rev. 45:15–33. Rusiecki JA, Matthews A, Sturgeon S, Sinha R, Pellizzari E, Zheng T et al. (2005). A correlation study of organochlorine levels in serum, breast adipose tissue, and gluteal adipose tissue among breast cancer cases in India. Cancer Epidemiol Biomark Prev. 14:1113–24. Ryan JJ, Rawn DF (2014). Polychlorinated dioxins, furans (PCDD/Fs), and polychlorinated biphenyls (PCBs) and their trends in Canadian human milk from 1992 to 2005. Chemosphere. 102:76–86. Ryan JJ, Levesque D, Panopio LG, Sun WF, Masuda Y, Kuroki H (1993). Elimination of polychlorinated dibenzofurans (PCDFs) and polychlorinated biphenyls (PCBs) from human blood in the Yusho and YuCheng rice oil poisonings. Arch Environ Contam Toxicol. 24:504–12. Rylander C, Sandanger TM, Petrenya N, Konoplev A, Bojko E, Odland JO (2011). Indications of decreasing human PTS concentrations in north west Russia. Global Health Action. 4:10.3402/gha.v4i0.8427. Rylander C, Lund E, Frøyland L, Sandanger TM (2012). Predictors of PCP, OH-PCBs, PCBs and chlorinated pesticides in a general female Norwegian population. Environ Int. 43:3–20. Rylander L, Rignell-Hydbom A, Hagmar L (2005). A cross-sectional study of the association between persistent organochlorine pollutants and diabetes. Environ Health. 4:28–33. Rylander L, Wallin E, Jönsson BA, Stridsberg M, Erfurth EM, Hagmar L (2006). Associations between CB-153 and p,p′-DDE and hormone levels in serum in middle-aged and elderly men. Chemosphere. 65:375–81. Safe S (1984). Polychlorinated biphenyls (PCBs) and polybrominated biphenyls (PBBs): biochemistry, toxicology, and mechanism of action. Crit Rev Toxicol. 13:319–95. Safe S (1989). Polychlorinated biphenyls (PCBs): mutagenicity and carcinogenicity. Mutat Res. 220:31–47.

320

WHO Food Additives Series No. 71-S1, 2016

Supplement 1: Non-dioxin-like polychlorinated biphenyls

Safe S (1993). Toxicology, structure–function relationship, and human and environmental health impacts of polychlorinated biphenyls: progress and problems. Environ Health Perspect. 100:259–68. Safe SH (1994). Polychlorinated biphenyls (PCBs): environmental impact, biochemical and toxic responses, and implications for risk assessment. Crit Rev Toxicol. 24:87–149. Sagiv SK, Tolbert PE, Altshul LM, Korrick SA (2007). Organochlorine exposures during pregnancy and infant size at birth. Epidemiology. 18:120–9. Sagiv SK, Nugent JK, Brazelton TB, Choi AL, Tolbert PE, Altshul LM et al. (2008). Prenatal organochlorine exposure and measures of behavior in infancy using the Neonatal Behavioral Assessment Scale (NBAS). Environ Health Perspect. 116:666–73. Sagiv SK, Thurston SW, Bellinger DC, Tolbert PE, Altshul LM, Korrick SA (2010). Prenatal organochlorine exposure and behaviors associated with attention deficit hyperactivity disorder in school-aged children. Am J Epidemiol. 171:593–601. Saint-Amour D, Roy MS, Bastien C, Ayotte P, Dewailly E, Després C et al. (2006). Alterations of visual evoked potentials in preschool Inuit children exposed to methylmercury and polychlorinated biphenyls from a marine diet. Neurotoxicology. 27:567–78. Salama AA, Mohamed MAM, Duval B, Potter TL, Levin RE (1998). Polychlorinated biphenyl concentration in raw and cooked North Atlantic bluefish (Pomatomus saltatrix) fillets. J Agric Food Chem. 46:1359–62. Salgovicová D, Pavlovicová D (2007). Exposure of the population of the Slovak Republic to dietary polychlorinated biphenyls. Food Chem Toxicol. 45:1641–9. Sandal S, Yilmaz B, Carpenter DO (2008). Genotoxic effects of PCB 52 and PCB 77 on cultured human peripheral lymphocytes. Mutat Res. 654:88–92. Sandal S, Yilmaz B, Godekmerdan A, Kelestimur H, Carpenter DO (2005). Effects of PCBs 52 and 77 on Th1/Th2 balance in mouse thymocyte cell cultures. Immunopharm Immunotoxicol. 27:601–13. Sandanger TM, Sinotte M, Dumas P, Marchand M, Sandau CD, Pereg D et al. (2007). Plasma concentrations of selected organobromine compounds and polychlorinated biphenyls in postmenopausal women of Québec, Canada. Environ Health Perspect. 115:1429–34. Sargent L, Rollof B, Meisner L (1989). In vitro chromosome damage due to PCB interactions. Mutat Res. 224:77–88. Sargent L, Sattler GL, Rollof B, Xu YH, Sattler CA, Meisner L et al. (1992). Ploidy and specific karyotypic changes during promotion with phenobarbital, 2,5,2′,5′-tetrachlorobiphenyl and/or 3,4,3′,4′-tetrachlorobiphenyl in rat liver. Cancer Res. 52:955–62. Sawada N, Iwasaki M, Inoue M, Itoh H, Sasazuki S, Yamaji T et al. (2010). Plasma organochlorines and subsequent risk of prostate cancer in Japanese men: a nested case–control study. Environ Health Perspect. 118:659–65. Schantz SL, Moshtaghian J, Ness DK (1995). Spatial learning deficits in adult rats exposed to orthosubstituted PCB congeners during gestation and lactation. Fundam Appl Toxicol. 26:117–26. Schantz SL, Seo BW, Wong PW, Pessah IN (1997). Long-term effects of developmental exposure to 2,2′,3,5′,6-pentachlorobiphenyl (PCB 95) on locomotor activity, spatial learning and memory and brain ryanodine binding. Neurotoxicology. 18:457–67.

321

Safety evaluation of certain food additives and contaminants Eightieth JECFA

Schecter A, Dellarco M, Päpke O, Olson J (1998). A comparison of dioxins, dibenzofurans and coplanar PCBs in uncooked and broiled ground beef, catfish and bacon. Chemosphere. 37:1723–30. Schell LM, Gallo MV, Denham M, Ravenscroft J, DeCaprio AP, Carpenter DO (2008). Relationship of thyroid hormone levels to levels of polychlorinated biphenyls, lead, p,p′-DDE, and other toxicants in Akwesasne Mohawk youth. Environ Health Perspect. 116:806–13. Schettgen T, Gube M, Alt A, Fromme H, Kraus T (2011). Pilot study on the exposure of the German general population to non-dioxin-like and dioxin-like PCBs. Int J Hyg Environ Health. 214:319–25. Schroeder AC, Privalsky ML (2014). Thyroid hormones, T3 and T4, in the brain. Front Endocrinol. (Lausanne). 5:40. Schulz DE, Petrick G, Duinker JC (1989). Complete characterization of polychlorinated biphenyl congeners in commercial Aroclor and Clophen mixtures by multidimensional gas chromatography– electron capture detection. Environ Sci Technol. 23:852–9. Schüpfer PY, Gülaçar FO (2000). Relative stabilities of cholestadienes calculated by molecular mechanics and semi-empirical methods: application to the acid-catalyzed rearrangement reactions of cholesta-3,5-diene. Org Geochem. 31:1589–96. Schwarz MA, Schneider K, Lindtner O, Blume K, Fiddicke U, Heinemeyer G (2010). [LExUKon – Dietary intake of environmental contaminants. Data processing support and standardization of exposure estimates on the basis of the National Consumption Study II; report on exposure estimates for NDL-PCBs.] Forschungs- und Beratungsinstitut Gefahrstoffe GmbH (FoBiG) and Bundesinstitut für Risikobewertung (BfR) (Forschungsvorhaben FKZ UM 08 61 712, July; in German). Schwarz MA, Lindtner O, Blume K, Heinemeyer G, Schneider K (2014). Cadmium exposure from food: the German LExUKon Project. Food Addit Contam A. 31:1038–51. Seegal RF, Brosch KO, Bush B (1986). Regional alterations in serotonin metabolism induced by oral exposure of rats to polychlorinated biphenyls. Neurotoxicology. 7:155–65. Seegal RF, Brosch KO, Okoniewski RJ (1997). Effects of in utero and lactational exposure of the laboratory rat to 2,4,2′,4′- and 3,4,3′,4′-tetrachlorobiphenyl on dopamine function. Toxicol Appl Pharmacol. 146:95–103. Seegal RF, Bush B, Brosch K (1985). Polychlorinated biphenyls induce regional changes in brain norepinephrine concentrations in adult rats. Neurotoxicology. 6:13–23. Seegal RF, Bush B, Brosch KO (1991a). Sub-chronic exposure of the adult rat to Aroclor 1254 yields regionally-specific changes in central dopaminergic function. Neurotoxicology. 12:55–65. Seegal RF, Bush B, Brosch KO (1991b). Comparison of effects of Aroclors 1016 and 1260 on non-human primate catecholamine function. Toxicology. 66:145–63. Seegal RF, Bush B, Shain W (1990). Lightly chlorinated ortho-substituted PCB congeners decrease dopamine in nonhuman primate brain and in tissue culture. Toxicol Appl Pharmacol. 106:136–44. Seegal RF, Fitzgerald EF, McCaffrey RJ, Shrestha S, Hills EA, Wolff MS et al. (2013). Tibial bone lead, but not serum polychlorinated biphenyl, concentrations are associated with neurocognitive deficits in former capacitor workers. J Occup Environ Med. 55:552–62. Seelbach M, Chen L, Powell A, Choi YJ, Zhang B, Hennig B et al. (2010). Polychlorinated biphenyls disrupt blood–brain barrier integrity and promote brain metastasis formation. Environ Health Perspect. 118:479–84.

322

WHO Food Additives Series No. 71-S1, 2016

Supplement 1: Non-dioxin-like polychlorinated biphenyls

Senthilkumar PK, Robertson LW, Ludewig G (2012). PCB153 reduces telomerase activity and telomere length in immortalized human skin keratinocytes (HaCaT) but not in human foreskin keratinocytes (NFK). Toxicol Appl Pharmacol. 259:115–23. Senthilkumar PK, Klingelhutz AJ, Jacobus JA, Lehmler H, Robertson LW, Ludewig G (2011). Airborne polychlorinated biphenyls (PCBs) reduce telomerase activity and shorten telomere length in immortal human skin keratinocytes (HaCat). Toxicol Lett. 204:64–70. Serra-Majem L, Ribas L, Salvador G, Castells C, Serra J, Jover L et al. (2003). [Evaluation of the nutritional state of the Catalan population from 2002 to 2003. Evolution of dietary habits and consumption of food and nutrients in Catalonia (1992–2003).] Direcció General de Salut Pública, Departament de Sanitat i Seguretat Social, Generalitat de Catalunya, Barcelona, Spain (in Catalan). Shafer TJ, Mundy WR, Tilson HA, Kodavanti PR (1996). Disruption of inositol phosphate accumulation in cerebellar granule cells by polychlorinated biphenyls: a consequence of altered Ca2+ homeostasis. Toxicol Appl Pharmacol. 141:448–55. Shain W, Bush B, Seegal R (1991). Neurotoxicity of polychlorinated biphenyls: structure–activity relationship of individual congeners. Toxicol Appl Pharmacol. 111:33–42. Shao Y, Wang J, Chen X, Wu Y (2014). The consolidation of food contaminants standards in China. Food Control. 43:213–6. Sharma R, Kodavanti PR (2002). In vitro effects of polychlorinated biphenyls and hydroxy metabolites on nitric oxide synthases in rat brain. Toxicol Appl Pharmacol. 178:127–36. Shen H, Han J, Tie X, Xu W, Ren Y, Ye C (2008). Polychlorinated dibenzo-p-dioxins/furans and polychlorinated biphenyls in human adipose tissue from Zhejiang Province, China. Chemosphere. 74:384–8. Shen H, Ding G, Wu Y, Pan G, Zhou X, Han J et al. (2012). Polychlorinated dibenzo-p-dioxins/furans (PCDD/Fs), polychlorinated biphenyls (PCBs), and polybrominated diphenyl ethers (PBDEs) in breast milk from Zhejiang, China. Environ Int. 42:84–90. Sherer RA, Price PS (1993). The effect of cooking processes on PCB levels in edible fish tissue. Qual Assur. 2:396–407. Shirai JH, Kissel JC (1996). Uncertainty in estimated half-lives of PCBs in humans: impact on exposure assessment. Sci Total Environ. 187:199–210. Shiu W, Ma K (2000). Temperature dependence of physical-chemical properties of selected chemicals of environmental interest. II. Chlorobenzenes, polychlorinated biphenyls, polychlorinated dibenzo-pdioxins, and dibenzofurans. J Phys Chem Ref Data. 29:387–462. Silver SR, Whelan EA, Deddens JA, Steenland NK, Hopf NB, Waters MA et al. (2009). Occupational exposure to polychlorinated biphenyls and risk of breast cancer. Environ Health Perspect. 117:276–82. Silverstone AE, Rosenbaum PF, Weinstock RS, Bartell SM, Foushee HR, Shelton C et al. (2012). Polychlorinated biphenyl (PCB) exposure and diabetes: results from the Anniston Community Health Survey. Environ Health Perspect. 120:727–32. Šimečková P, Vondrácek J, Andrysík Z, Zatloukalová J, Krcmár P, Kozubík A et al. (2009a). The 2,2′,4,4′,5,5′-hexachlorobiphenyl-enhanced degradation of connexin 43 involves both proteasomal and lysosomal activities. Toxicol Sci. 107:9–18.

323

Safety evaluation of certain food additives and contaminants Eightieth JECFA

Šimečková P, Vondrácek J, Procházková J, Kozubík A, Krcmár P, Machala M (2009b). 2,2′,4,4′,5,5′Hexachlorobiphenyl (PCB 153) induces degradation of adherens junction proteins and inhibits betacatenin-dependent transcription in liver epithelial cells. Toxicology. 260:104–11. Simon T, Britt JK, James RC (2007). Development of a neurotoxic equivalence scheme of relative potency for assessing the risk of PCB mixtures. Regul Toxicol Pharmacol. 48:148–70. Sioen I, Den Hond E, Nelen V, Van de Mieroop E, Croes K, Van Larebeke N et al. (2013). Prenatal exposure to environmental contaminants and behavioural problems at age 7–8 years. Environ Int. 59:225–31. Sipka S, Eum SY, Son KW, Xu S, Gavalas VG, Hennig B et al. (2008). Oral administration of PCBs induces proinflammatory and prometastatic responses. Environ Toxicol Pharmacol. 25:251–9. Sirot V, Tard A, Venisseau A, Brosseaud A, Marchand P, Le Bizec B et al. (2012). Dietary exposure to polychlorinated dibenzo-p-dioxins, polychlorinated dibenzofurans and polychlorinated biphenyls of the French population: results of the second French Total Diet Study. Chemosphere. 88:492–500. Sissons D, Welti D (1971). Structural identification of polychlorinated biphenyls in commercial mixtures by gas–liquid chromatography, nuclear magnetic resonance and mass spectrometry. J Chromatogr A. 60:15–32. Skrbic B, Durisic-Mladenovic N (2007). Non-dioxin-like PCBs in crops and related products: levels and intakes in Serbia. Food Addit Contam. 24:652–62. Smeds A, Saukko P (2001). Identification and quantification of polychlorinated biphenyls and some endocrine disrupting pesticides in human adipose tissue from Finland. Chemosphere. 44:1463–71. Smialowicz RJ, DeVito MJ, Riddle MM, Williams WC, Birnbaum LS (1997). Opposite effects of 2,2′,4,4′,5,5′-hexachlorobiphenyl and 2,3,7,8-tetrachlorodibenzo-p-dioxin on the antibody response to sheep erythrocytes in mice. Fundam Appl Toxicol. 37:141–9. Smithwick LA, Smith A, Quensen JF III, Stack A, London L, Morris PJ (2003). Inhibition of LPS-induced splenocyte proliferation by ortho-substituted polychlorinated biphenyl congeners. Toxicology. 188:319–33. Sonneborn D, Park HY, Petrik J, Kocan A, Palkovicova L, Trnovec T et al. (2008). Prenatal polychlorinated biphenyl exposures in eastern Slovakia modify effects of social factors on birthweight. Paediatr Perinat Epidemiol. 22:202–13. WHO Food Additives Series No. 71-S1, 2016 Spanò M, Toft G, Hagmar L, Eleuteri P, Rescia M, Rignell-Hydbom A et al. (2005). Exposure to PCB and p,p′-DDE in European and Inuit populations: impact on human sperm chromatin integrity. Hum Reprod. 20:3488–99. Sparr Eskilsson C, Björklund E (2000). Analytical-scale microwave-assisted extraction. J Chromatogr A. 902:227–50. Spector JT, De Roos AJ, Ulrich CM, Sheppard L, Sjödin A, Wener MH et al. (2014). Plasma polychlorinated biphenyl concentrations and immune function in postmenopausal women. Environ Res. 131:174–80. Spencer WA, Lehmler HJ, Robertson LW, Gupta RC (2009). Oxidative DNA adducts after Cu2+-mediated activation of dihydroxy PCBs: role of reactive oxygen species. Free Radic Biol Med. 46:1346–52. Spinelli JJ, Ng CH, Weber JP, Connors JM, Gascoyne RD, Lai AS et al. (2007). Organochlorines and risk of non-Hodgkin lymphoma. Int J Cancer. 121:2767–75. Srinivasan A, Lehmler HJ, Robertson LW, Ludewig G (2001). Production of DNA strand breaks in vitro and reactive oxygen species in vitro and in HL-60 cells by PCB metabolites. Toxicol Sci. 60:92–102.

324

Supplement 1: Non-dioxin-like polychlorinated biphenyls

Stack AS, Altman-Hamamdzic S, Morris PJ, London SD, London L (1999). Polychlorinated biphenyl mixtures (Aroclors) inhibit LPS-induced murine splenocyte proliferation in vitro. Toxicology. 139:137– 54. Stadnicki SS, Allen JR (1979). Toxicity of 2,2′,5,5′-tetrachlorobiphenyl and its metabolites 2,2′,5,5′tetrachlorobiphenyl-3,4-oxide and 2,2′,5,5′-tetrachlorobiphenyl-4-ol to cultured cells in vitro. Bull Environ Contam Toxicol. 23:788–96. Steenland K, Hein MJ, Cassinelli RT II, Prince MM, Nilsen NB, Whelan EA et al. (2006). Polychlorinated biphenyls and neurodegenerative disease mortality in an occupational cohort. Epidemiology. 17:8–13. Stellman SD, Djordjevic MV, Muscat JE, Gong L, Bernstein D, Citron ML et al. (1998). Relative abundance of organochlorine pesticides and polychlorinated biphenyls in adipose tissue and serum of women in Long Island, New York. Cancer Epidemiol Biomarkers Prev. 7:489–96. Stewart P, Reihman J, Gump B, Lonky E, Darvill T, Pagano J (2005). Response inhibition at 8 and 9 1/2 years of age in children prenatally exposed to PCBs. Neurotoxicol Teratol. 27:771–80. Stewart PW, Sargent DM, Reihman J, Gump BB, Lonky E, Darvill T et al. (2006). Response inhibition during differential reinforcement of low rates (DRL) schedules may be sensitive to low-level polychlorinated biphenyl, methylmercury, and lead exposure in children. Environ Health Perspect. 114:1923–9. Stewart PW, Lonky E, Reihman J, Pagano J, Gump BB, Darvill T (2008). The relationship between prenatal PCB exposure and intelligence (IQ) in 9-year-old children. Environ Health Perspect. 116:1416–22. Stockholm Convention (2009). Stockholm Convention on Persistent Organic Pollutants, as amended in 2009. Châtelaine (Geneva): Secretariat of the Stockholm Convention (http://chm.pops.int/ TheConvention/Overview/TextoftheConvention/tabid/2232/Default.aspx, accessed 3 January 2016). Stølevik SB, Nygaard UC, Namork E, Haugen M, Kvalem HE, Meltzer HM (2011). Prenatal exposure to polychlorinated biphenyls and dioxins is associated with increased risk of wheeze and infections in infants. Food Chem Toxicol. 49:1843–8. Stølevik SB, Nygaard UC, Namork E, Haugen M, Meltzer HM, Alexander J (2013). Prenatal exposure to polychlorinated biphenyls and dioxins from the maternal diet may be associated with immunosuppressive effects that persist into early childhood. Food Chem Toxicol. 51:165–72. Storelli MM, Barone G, D’Addabbo R, Marcotrigiano GO (2006). Concentrations and composition of organochlorine contaminants in different species of cephalopod molluscs from the Italian waters (Adriatic Sea). Chemosphere. 64:129–34. Storelli MM, Casalino E, Barone G, Marcotrigiano GO (2008). Persistent organic pollutants (PCBs and DDTs) in small size specimens of bluefin tuna (Thunnus thynnus) from the Mediterranean Sea (Ionian Sea). Environ Int. 34:509–13. Storelli MM, Perrone VG, Busco VP, Spedicato D, Barone G (2012a). Persistent organic pollutants (PCBs and DDTs) in European conger eel, Conger conger L., from the Ionian Sea (Mediterranean Sea). Bull Environ Contam Toxicol. 88:928–32. Storelli MM, Scarano C, Spanu C, De Santis EPL, Busco VP, Storelli A et al. (2012b). Levels and congener profiles of polychlorinated dibenzo-p-dioxins (PCDDs), polychlorinated dibenzofurans (PCDFs) and polychlorinated biphenyls (PCBs) in sheep milk from an industrialised area of Sardinia, Italy. Food Chem Toxicol. 50:1413–7.

325

Safety evaluation of certain food additives and contaminants Eightieth JECFA

Storr-Hansen E, Cederberg T (1992). Determination of coplanar polychlorinated biphenyl (CB) congeners in seal tissues by chromatography on active carbon, dual-column high resolution GC/ECD and high resolution GC/high resolution MS. Chemosphere. 24:1181–96. Strathmann J, Schwarz M, Tharappel JC, Glauert HP, Spear BT, Robertson LW et al. (2006). PCB 153, a non-dioxin-like tumor promoter, selects for beta-catenin (Catnb)-mutated mouse liver tumors. Toxicol Sci. 93:34–40. Strauss HS, Heiger-Bernays W (2012). Methodological limitations may prevent the observation of nonHodgkin’s lymphoma in bioassays of polychlorinated biphenyls. Toxicol Pathol. 40:995–1003. Su PH, Huang PC, Lin CY, Ying TH, Chen JY, Wang SL (2012). The effect of in utero exposure to dioxins and polychlorinated biphenyls on reproductive development in eight year-old children. Environ Int. 39:181–7. Suarez L, Gilani Z, Felkner M, Brender J, Henry J, Hendricks K (2005). Exposure to polychlorinated biphenyls and risk of neural-tube defects in a Mexican American population. Int J Occup Environ Health. 11:233–7. Sunyer J, Torrent M, Muñoz-Ortiz L, Ribas-Fitó N, Carrizo D, Grimalt J et al. (2005). Prenatal dichlorodiphenyldichloroethylene (DDE) and asthma in children. Environ Health Perspect. 113:1787– 90. Sunyer J, Garcia-Esteban R, Alvarez M, Guxens M, Goñi F, Basterrechea M et al. (2010). DDE in mothers’ blood during pregnancy and lower respiratory tract infections in their infants. Epidemiology. 21:729– 35. Suzuki K, Nakai K, Sugawara T, Nakamura T, Ohba T, Shimada M et al. (2010). Neurobehavioral effects of prenatal exposure to methylmercury and PCBs, and seafood intake: neonatal behavioral assessment scale results of Tohoku study of child development. Environ Res. 110:699–704. Takser L, Mergler D, Baldwin M, de Grosbois S, Smargiassi A, Lafond J (2005). Thyroid hormones in pregnancy in relation to environmental exposure to organochlorine compounds and mercury. Environ Health Perspect. 113:1039–45. Tampal N, Lehmler HJ, Espandiari P, Malmberg T, Robertson LW (2002). Glucuronidation of hydroxylated polychlorinated biphenyls (PCBs). Chem Res Toxicol. 15:1259–66. WHO Food Additives Series No. 71-S1, 2016 Tan J, Li QQ, Loganath A, Chong YS, Xiao M, Obbard JP (2008). Multivariate data analyses of persistent organic pollutants in maternal adipose tissue in Singapore. Environ Sci Technol. 42:2681–7. Tan J, Loganath A, Chong YS, Obbard JP (2009). Exposure to persistent organic pollutants in utero and related maternal characteristics on birth outcomes: a multivariate data analysis approach. Chemosphere. 74:428–33. Tanabe S, Madhusree B, Ozturk AA, Tatsukawa R, Miyazaki N, Ozdamar E et al. (1997). Isomer-specific analysis of polychlorinated biphenyls in harbour porpoise (Phocoena phocoena) from the Black Sea. Mar Pollut Bull. 34:712–20. Tanaka T, Morita A, Kato M, Hirai T, Mizoue T, Terauchi Y et al. (2011). Congener-specific polychlorinated biphenyls and the prevalence of diabetes in the Saku Control Obesity Program (SCOP). Endocr J. 58:589–96. Tang-Péronard JL, Heitmann BL, Andersen HR, Steuerwald U, Grandjean P, Weihe P et al. (2014). Association between prenatal polychlorinated biphenyl exposure and obesity development at ages 5 and 7 y: a prospective cohort study of 656 children from the Faroe Islands. Am J Clin Nutr. 99:5–13.

326

Supplement 1: Non-dioxin-like polychlorinated biphenyls

Tatsuta N, Nakai K, Murata K, Suzuki K, Iwai-Shimada M, Kurokawa N et al. (2014). Impacts of prenatal exposures to polychlorinated biphenyls, methylmercury, and lead on intellectual ability of 42-monthold children in Japan. Environ Res. 133:321–6. Terrell ML, Berzen AK, Small CM, Cameron LL, Wirth JJ, Marcus M (2009). A cohort study of the association between secondary sex ratio and parental exposure to polybrominated biphenyl (PBB) and polychlorinated biphenyl (PCB). Environ Health. 8:35. Thomas GO, Sweetman AJ, Jones KC (1999a). Input–output balance of polychlorinated biphenyls in a long-term study of lactating dairy cows. Environ Sci Technol. 33:104–12. Thomas GO, Sweetman AJ, Jones KC (1999b). Metabolism and body-burden of PCBs in lactating dairy cows. Chemosphere. 39:1533–44. Thomas GO, Wilkinson M, Hodson S, Jones KC (2006). Organohalogen chemicals in human blood from the United Kingdom. Environ Pollut. 141:30–41. Tiido T, Rignell-Hydbom A, Jönsson B, Giwercman YL, Rylander L, Hagmar L et al. (2005). Exposure to persistent organochlorine pollutants associates with human sperm Y:X chromosome ratio. Hum Reprod. 20:1903–9. Tiido T, Rignell-Hydbom A, Jönsson BA, Giwercman YL, Pedersen HS, Wojtyniak B et al. (2006). Impact of PCB and p,p′-DDE contaminants on human sperm Y:X chromosome ratio: studies in three European populations and the Inuit population in Greenland. Environ Health Perspect. 114:718–24. Tilson HA, Kodavanti PRS (1998). The neurotoxicity of polychlorinated biphenyls. Neurotoxicology. 19:517–26. Tilson HA, Kodavanti PRS, Mundy WR, Bushnell PJ (1998). Neurotoxicity of environmental chemicals and the mechanism of action. Toxicol Lett. 102–103:631–5. Todaka T, Hori T, Hirakawa H, Kajiwara J, Yasutake D, Onozuka D et al. (2008). Congener-specific analysis of non-dioxin-like polychlorinated biphenyls in blood collected from 127 elderly residents in Nakagawa Town, Fukuoka Prefecture, Japan. Chemosphere. 73:865–72. Todaka T, Hirakawa H, Kajiwara J, Hori T, Tobiishi K, Yasutake D et al. (2010). Relationship between the concentrations of polychlorinated dibenzo-p-dioxins, polychlorinated dibenzofurans, and polychlorinated biphenyls in maternal blood and those in breast milk. Chemosphere. 78:185–92. Todaka T, Hirakawa H, Kajiwara J, Onozuka D, Sasaki S, Miyashita C et al. (2011). Concentrations of polychlorinated dibenzo-p-dioxins, polychlorinated dibenzofurans, and polychlorinated biphenyls in blood and breast milk collected from pregnant women in Sapporo City, Japan. Chemosphere. 85:1694–700. Toft G, Rignell-Hydbom A, Tyrkiel E, Shvets M, Giwercman A, Lindh CH et al. (2006). Semen quality and exposure to persistent organochlorine pollutants. Epidemiology. 17:450–8. Tomasallo C, Anderson H, Haughwout M, Imm P, Knobeloch L (2010). Mortality among frequent consumers of Great Lakes sport fish. Environ Res. 110:62–9. Törnkvist A, Glynn A, Aune M, Darnerud PO, Ankarberg EH (2011). PCDD/F, PCB, PBDE, HBCD and chlorinated pesticides in a Swedish market basket from 2005 – levels and dietary intake estimations. Chemosphere. 83:193–9.

327

Safety evaluation of certain food additives and contaminants Eightieth JECFA

Traag WA, Kan CA, van der Weg G, Onstenk C, Hoogenboom LAP (2006). Residues of dioxins (PCDD/ Fs) and PCBs in eggs, fat and livers of laying hens following consumption of contaminated feed. Chemosphere. 65:1518–25. Trabert B, De Roos AJ, Schwartz SM, Peters U, Scholes D, Barr DB et al. (2010). Non-dioxin-like polychlorinated biphenyls and risk of endometriosis. Environ Health Perspect. 118:1280–5. Trnovec T, Sovcíková E, Pavlovcinová G, Jakubíková J, Jusko TA, Husták M et al. (2010). Serum PCB concentrations and cochlear function in 12-year-old children. Environ Sci Technol. 44:2884–9. Tsai PC, Ko YC, Huang W, Liu HS, Guo YL (2007). Increased liver and lupus mortalities in 24-year followup of the Taiwanese people highly exposed to polychlorinated biphenyls and dibenzofurans. Sci Total Environ. 374:216–22. Tsuji M, Vogel CF, Koriyama C, Akiba S, Katoh T, Kawamoto T et al. (2012). Association of serum levels of polychlorinated biphenyls with IL-8 mRNA expression in blood samples from asthmatic and nonasthmatic Japanese children. Chemosphere. 87:1228–34. Tsuji M, Aiko Y, Kawamoto T, Hachisuga T, Kooriyama C, Myoga M et al. (2013). Polychlorinated biphenyls (PCBs) decrease the placental syncytiotrophoblast volume and increase placental growth factor (PlGF) in the placenta of normal pregnancy. Placenta. 34:619–23. Tsukimori K, Tokunaga S, Shibata S, Uchi H, Nakayama D, Ishimaru T et al. (2008). Long-term effects of polychlorinated biphenyls and dioxins on pregnancy outcomes in women affected by the Yusho incident. Environ Health Perspect. 116:626–30. Tsukimori K, Yasukawa F, Uchi H, Furue M, Morokuma S (2012). Sex ratio in two generations of the Yusho cohort. Epidemiology. 23:349–50. Tsukino H, Hanaoka T, Sasaki H, Motoyama H, Hiroshima M, Tanaka T et al. (2005). Associations between serum levels of selected organochlorine compounds and endometriosis in infertile Japanese women. Environ Res. 99:118–25. Tsutsumi T, Iida T, Hori T, Nakagawa R, Tobiishi K, Yanagi T et al. (2002). Recent survey and effects of cooking processes on levels of PCDDs, PCDFs and Co-PCBs in leafy vegetables in Japan. Chemosphere. 46:1443–9. Tuinstra LG, Traag WA, Keukens HJ (1980). Quantitative determination of individual chlorinated biphenyls in milkfat by splitless glass capillary gas chromatography. J Assoc Off Anal Chem. 63:952–8. Turci R, Mariani G, Marinaccio A, Balducci C, Bettinelli M, Fanelli R et al. (2004). Critical evaluation of a high-throughput analytical method for polychlorinated biphenyls in human serum: which detector for the establishment of the reference values? Rapid Commun Mass Spectrom. 18:421–34. Turci R, Finozzi E, Catenacci G, Marinaccio A, Balducci C, Minoia C (2006). Reference values of coplanar and non-coplanar PCBs in serum samples from two Italian population groups. Toxicol Lett. 162:250–5. Turci R, Balducci C, Brambilla G, Colosio C, Imbriani M, Mantovani A et al. (2010). A simple and fast method for the determination of selected organohalogenated compounds in serum samples from the general population. Toxicol Lett. 192:66–71. Turrio-Baldassarri L, Abate V, Battistelli CL, Carasi S, Casella M, Iacovella N et al. (2008). PCDD/F and PCB in human serum of differently exposed population groups of an Italian city. Chemosphere. 73:S228–34.

328

WHO Food Additives Series No. 71-S1, 2016

Supplement 1: Non-dioxin-like polychlorinated biphenyls

Turrio-Baldassarri L, Alivernini S, Carasi S, Casella M, Fuselli S, Iacovella N et al. (2009). PCB, PCDD and PCDF contamination of food of animal origin as the effect of soil pollution and the cause of human exposure in Brescia. Chemosphere. 76:278–85. Turunen AW, Verkasalo PK, Kiviranta H, Pukkala E, Jula A, Männistö S et al. (2008). Mortality in a cohort with high fish consumption. Int J Epidemiol. 37:1008–17. Turyk ME, Anderson HA, Persky VW (2007). Relationships of thyroid hormones with polychlorinated biphenyls, dioxins, furans, and DDE in adults. Environ Health Perspect. 115:1197–203. Turyk ME, Anderson HA, Freels S, Chatterton R Jr, Needham LL, Patterson DG Jr et al. (2006). Associations of organochlorines with endogenous hormones in male Great Lakes fish consumers and nonconsumers. Environ Res. 102:299–307. Turyk M, Anderson HA, Knobeloch L, Imm P, Persky VW (2009a). Prevalence of diabetes and body burdens of polychlorinated biphenyls, polybrominated diphenyl ethers, and p,p′-diphenyldichloroethene in Great Lakes sport fish consumers. Chemosphere. 75:674–9. Turyk M, Anderson H, Knobeloch L, Imm P, Persky V (2009b). Organochlorine exposure and incidence of diabetes in a cohort of Great Lakes sport fish consumers. Environ Health Perspect. 117:1076–82. Uçar Y, Traag W, Immerzeel J, Kraats C, van der Lee M, Hoogenboom R et al. (2011). Levels of PCDD/Fs, PCBs and PBDEs in butter from Turkey and estimated dietary intake from dairy products. Food Addit Contam B Surveill. 4:141–51. Ukropec J, Radikova Z, Huckova M, Koska J, Kocan A, Sebokova E et al. (2010). High prevalence of prediabetes and diabetes in a population exposed to high levels of an organochlorine cocktail. Diabetologia. 53:899–906. Ulbrich B, Stahlmann R (2004). Developmental toxicity of polychlorinated biphenyls (PCBs): a systematic review of experimental data. Arch Toxicol. 78:252–68. Ulutaş OK, Çok I, Darendeliler F, Aydin B, Çoban A, Henkelmann B et al. (2015). Blood levels of polychlorinated biphenyls and organochlorinated pesticides in women from Istanbul, Turkey. Environ Monit Assess. 187:132. UNEP (1999). Guidelines for the identification of PCBs and materials containing PCBs. First issue. August 1999. Châtelaine (Geneva): United Nations Environment Programme, UNEP Chemicals, Switzerland (http://chm.pops.int/Implementation/PCBs/DocumentsPublications/tabid/665/Default. aspx, accessed 7 July 2016). UNEP (2013). Guidance on the global monitoring plan for persistent organic pollutants. Stockholm Convention on Persistent Organic Pollutants (UNEP/POPS/COP.6/INF/31; http://chm.pops.int/ Implementation/GlobalMonitoringPlan/Overview/tabid/83/Default.aspx, accessed 3 January 2016). UNEP/WHO (2013). Human exposure to POPs across the globe: POPs levels and human health implications. Results of the WHO/UNEP Human Milk Survey. Geneva: United Nations Environment Programme and World Health Organization (UNEP/POPS/COP.6/INF/33; http://chm.pops.int/ Implementation/GlobalMonitoringPlan/MonitoringActivities/Humanmilksurvey/tabid/270/Default. aspx, accessed 3 January 2016). USEPA (1996). Method 8082. Polychlorinated biphenyls (PCBs) by gas chromatography. Revision 0, December 1996. Washington (DC): United States Environmental Protection Agency (https://www. o2si.com/docs/epa-method-8082.pdf, accessed 28 February 2016).

329

Safety evaluation of certain food additives and contaminants Eightieth JECFA

USEPA (2000). Method 8082A. Polychlorinated biphenyls (PCBs) by gas chromatography. Revision 1, November 2000. Washington (DC): United States Environmental Protection Agency (http://www. caslab.com/EPA-Methods/PDF/EPA-Method-8082A.pdf, accessed 28 February 2016). USEPA (2008). Method 1668B. Chlorinated biphenyl congeners in water, soil, sediment, biosolids, and tissue by HRGC/HRMS. November 2008. Washington (DC): United States Environmental Protection Agency (http://nepis.epa.gov/Exe/ZyPDF.cgi/P1005EUE.PDF?Dockey=P1005EUE.PDF, accessed 28 February 2016). USEPA (2010). EPA’s reanalysis of key issues related to dioxin toxicity and response to NAS comments (external review draft). Washington (DC): United States Environmental Protection Agency (EPA/600/ R-10/038A; http://cfpub.epa.gov/ncea/risk/recordisplay.cfm?deid=222203, accessed 28 February 2016). USFDA (2014). Part 109. Unavoidable contaminants in food for human consumption and foodpackaging material. Subpart B. Tolerances for unavoidable poisonous or deleterious substances. Sec. 109.30 Tolerances for polychlorinated biphenyls (PCB’s). Silver Spring (MD): United States Department of Health and Human Services, Food and Drug Administration, Code of Federal Regulations Title 21 (http://www.accessdata.fda.gov/SCRIPTS/cdrh/cfdocs/cfcfr/CFRSearch.cfm?fr=109.30, accessed 7 April 2015). Vaclavik E, Tjonneland A, Stripp C, Overvad K, Weber JP, Raaschou-Nielsen O (2006). Organochlorines in Danish women: predictors of adipose tissue concentrations. Environ Res. 100:363–70. Valera B, Jørgensen ME, Jeppesen C, Bjerregaard P (2013a). Exposure to persistent organic pollutants and risk of hypertension among Inuit from Greenland. Environ Res. 122:65–73. Valera B, Ayotte P, Poirier P, Dewailly E (2013b). Associations between plasma persistent organic pollutant levels and blood pressure in Inuit adults from Nunavik. Environ Int. 59:282–9. Valvi D, Mendez MA, Martinez D, Grimalt JO, Torrent M, Sunyer J et al. (2012). Prenatal concentrations of polychlorinated biphenyls, DDE, and DDT and overweight in children: a prospective birth cohort study. Environ Health Perspect. 120:451–7. van den Berg M, Birnbaum L, Bosveld AT, Brunstrom B, Cook P, Feeley M et al. (1998). Toxic equivalency factors (TEFs) for PCBs, PCDDs, PCDFs for humans and wildlife. Environ Health Perspect. 106:775–92. WHO Food Additives Series No. 71-S1, 2016 van den Berg M, Birnbaum LS, Denison M, De Vito M, Farland W, Feeley M et al. (2006). The 2005 World Health Organization reevaluation of human and mammalian toxic equivalency factors for dioxins and dioxin-like compounds. Toxicol Sci. 93:223–41. van den Hurk P, Kubiczak GA, Lehmler HJ, James MO (2002). Hydroxylated polychlorinated biphenyls as inhibitors of the sulfation and glucuronidation of 3-hydroxy-benzo[a]pyrene. Environ Health Perspect. 110:343–8. van der Kolk J, van Birgelen APJM, Poiger H, Schlatter C (1992). Interactions of 2,2′,4,4′,5,5′hexachlorobiphenyl and 2,3,7,8-tetrachlorodibenzo-p-dioxin in a subchronic feeding study in the rat. Chemosphere. 25:2023–7. van der Velde EG, Hijman WC, Linders SHMA, Liem AKD (1996). SFE as clean up technique for ppt-levels of PCBs in fatty samples. Organohalogen Compounds. 27:247–52. Van Leeuwen FXR, Younes MM, editors (2000). Consultation on assessment of the health risk of dioxins: re-evaluation of the tolerable daily intake (TDI) – executive summary. Food Addit Contam. 17:223–40.

330

Supplement 1: Non-dioxin-like polychlorinated biphenyls

Van Leeuwen SPJ, Leonards PEG, Traag WA, Hoogenboom LAP, De Boer J (2007). Polychlorinated dibenzo-p-dioxins, dibenzofurans and biphenyls in fish from the Netherlands: concentrations, profiles and comparison with DR CALUX bioassay results. Anal Bioanal Chem. 389:321–33. Vansell NR, Klaassen CD (2002). Effect of microsomal enzyme inducers on the biliary excretion of triiodothyronine (T3) and its metabolites. Toxicol Sci. 65:184–91. Vasiliu O, Cameron L, Gardiner J, Deguire P, Karmaus W (2006). Polybrominated biphenyls, polychlorinated biphenyls, body weight, and incidence of adult-onset diabetes mellitus. Epidemiology. 17:352–9. Venkatesha VA, Venkataraman S, Sarsour EH, Kalen AL, Buettner GR, Robertson LW et al. (2008). Catalase ameliorates polychlorinated biphenyl–induced cytotoxicity in non-malignant human breast epithelial cells. Free Radic Biol Med. 45:1094–102. Verger P, Tressou J, Clémençon S (2007). Integration of time as a description parameter in risk characterisation: application to methyl mercury. Regul Toxicol Pharmacol. 49:25–30. Verhulst SL, Nelen V, Hond ED, Koppen G, Beunckens C, Vael C et al. (2009). Intrauterine exposure to environmental pollutants and body mass index during the first 3 years of life. Environ Health Perspect. 117:122–6. Verner MA, McDougall R, Glynn A, Andersen ME, Clewell HJ III, Longnecker MP (2013). Is the relationship between prenatal exposure to PCB-153 and decreased birth weight attributable to pharmacokinetics? Environ Health Perspect. 121:1219–24. Vichi S, Medda E, Ingelido AM, Ferro A, Resta S, Porpora MG et al. (2012). Glutathione transferase polymorphisms and risk of endometriosis associated with polychlorinated biphenyls exposure in Italian women: a gene–environment interaction. Fertil Steril. 97:1143–51.e3. Viel JF, Floret N, Deconinck E, Focant JF, De Pauw E, Cahn JY (2011). Increased risk of non-Hodgkin lymphoma and serum organochlorine concentrations among neighbors of a municipal solid waste incinerator. Environ Int. 37:449–53. Vigh É, Colombo A, Benfenati E, Håkansson H, Berglund M, Bódis J et al. (2013). Individual breast milk consumption and exposure to PCBs and PCDD/Fs in Hungarian infants: a time-course analysis of the first three months of lactation. Sci Total Environ. 449:336–44. Villeneuve S, Cyr D, Lynge E, Orsi L, Sabroe S, Merletti F et al. (2010). Occupation and occupational exposure to endocrine disrupting chemicals in male breast cancer: a case–control study in Europe. Occup Environ Med. 67:837–44. Viluksela M, Heikkinen P, van der Ven LT, Rendel F, Roos R, Esteban J et al. (2014). Toxicological profile of ultrapure 2,2′,3,4,4′,5,5′-heptachlorbiphenyl (PCB 180) in adult rats. PLoS One. 9(8):e104639. Voie ØA, Wiik P, Fonnum F (1998). Ortho substituted polychlorinated biphenyls activate respiratory burst measured as luminol-amplified chemoluminescence in human granulocytes. Toxicol Appl Pharmacol. 150:369–75. Voie ØA, Tysklind M, Andersson PL, Fonnum F (2000). Activation of respiratory burst in human granulocytes by polychlorinated biphenyls: a structure–activity study. Toxicol Appl Pharmacol. 167:118–24. Wallin E, Rylander L, Jönsson BA, Lundh T, Isaksson A, Hagmar L (2005). Exposure to CB-153 and p,p′DDE and bone mineral density and bone metabolism markers in middle-aged and elderly men and women. Osteoporos Int. 16:2085–94. 331

Safety evaluation of certain food additives and contaminants Eightieth JECFA

Wang SL, Su PH, Jong SB, Guo YL, Chou WL, Päpke O (2005). In utero exposure to dioxins and polychlorinated biphenyls and its relations to thyroid function and growth hormone in newborns. Environ Health Perspect. 113:1645–50. Wang SS, Lu Y, Rothman N, Abdou AM, Cerhan JR, De Roos A et al. (2011). Variation in effects of nonHodgkin lymphoma risk factors according to the human leukocyte antigen (HLA)-DRB1*01:01 allele and ancestral haplotype 8.1. PLoS One. 6:e26949. Wangpradit O, Teesch LM, Mariappan SV, Duffel MW, Norstrom K, Robertson LW et al. (2009). Oxidation of 4-chlorobiphenyl metabolites to electrophilic species by prostaglandin H synthase. Chem Res Toxicol. 22:64–71. Ward MH, Colt JS, Metayer C, Gunier RB, Lubin J, Crouse V et al. (2009). Residential exposure to polychlorinated biphenyls and organochlorine pesticides and risk of childhood leukemia. Environ Health Perspect. 117:1007–13. Warner NA, Martin JW, Wong CS (2009). Chiral polychlorinated biphenyls are biotransformed enantioselectively by mammalian cytochrome P-450 isozymes to form hydroxylated metabolites. Environ Sci Technol. 43:114–21. Waxman EA, Lynch DR (2005). N-Methyl-D-aspartate receptor subtypes: multiple roles in excitotoxicity and neurological disease. Neuroscientist. 11:37–49. Webb RG, McCall AC (1973). Quantitative PCB standards for electron capture gas chromatography. J Chromatogr Sci. 11:366–73. Weber R, Albrecht M, Ballschmiter K, Berger J, Bruns-Weller E, Kamphues J et al. (2014). Safe food production from free range beef – minimizing TEQ-levels in meat by tracking PCB-sources. Organohalogen Compounds. 76:815–8. Wei W, Zhang C, Liu AL, Xie SH, Chem XM, Lu WQ (2009). Effect of PCB153 on BaP-induced genotoxicity in Hep2G cells via modulation of metabolic enzymes. Mutat Res. 675:71–6. Weiss J, Papke O, Bergman A (2005). A worldwide survey of polychlorinated dibenzo-p-dioxins, dibenzofurans, and related contaminants in butter. Ambio. 34:589–97. Weiss J, Wallin E, Axmon A, Jönsson BA, Akesson H, Janák K et al. (2006). Hydroxy-PCBs, PBDEs, and HBCDDs in serum from an elderly population of Swedish fishermen’s wives and associations with bone density. Environ Sci Technol. 40:6282–9. Weisskopf MG, Anderson HA, Hanrahan LP, Kanarek MS, Falk CM, Steenport DM et al. (2005). Maternal exposure to Great Lakes sport-caught fish and dichlorodiphenyl dichloroethylene, but not polychlorinated biphenyls, is associated with reduced birth weight. Environ Res. 97:149–62. Weistrand C, Noren K (1993). Liquid–gel partitioning using Lipidex in the determination of polychlorinated biphenyls in cod liver oil. J Chromatogr. 630:179–86. Wells DE, Hess P (2000). Chapter 2. Separation, clean-up and recoveries of persistent trace organic contaminants from soils, sediment and biological matrices. Tech Instrum Anal Chem. 21:73–113. Whitehead TP, Crispo Smith S, Park JS, Petreas MX, Rappaport SM, Metayer C (2015). Concentrations of persistent organic pollutants in California women’s serum and residential dust. Environ Res. 136:57–66. WHO (1985). Guidelines for the study of dietary intakes of chemical contaminants. Geneva: World Health Organization (http://apps.who.int/iris/bitstream/10665/39255/1/WHO_OFFSET_87.pdf, accessed 28 February 2016).

332

WHO Food Additives Series No. 71-S1, 2016

Supplement 1: Non-dioxin-like polychlorinated biphenyls

WHO (1998). Assessment of the health risks of dioxins: re-evaluation of the tolerable daily intake (TDI). Executive summary of the WHO consultation, 25–29 May 1998, Geneva, Switzerland. Geneva: World Health Organization (http://www.who.int/ipcs/publications/en/exe-sum-final.pdf, accessed 28 February 2016). WHO (1999). Definition, diagnosis and classification of diabetes mellitus and its complications: report of a WHO consultation. Part 1: Diagnosis and classification of diabetes mellitus. Geneva: World Health Organization (http://apps.who.int/iris/bitstream/10665/66040/1/WHO_NCD_NCS_99.2.pdf, accessed 21 March 2016). WHO (2015). Infant and young child feeding. Geneva: World Health Organization (Fact Sheet No. 342; http://www.who.int/mediacentre/factsheets/fs342/en/, accessed 3 January 2016). Wigestrand MB, Stenberg M, Walaas SI, Fonnum F, Andersson PL (2013). Non-dioxin-like PCBs inhibit [3H]WIN-35,428 binding to the dopamine transporter: a structure–activity relationship study. Neurotoxicology. 39:18–24. Wilhelm M, Ranft U, Krämer U, Wittsiepe J, Lemm F, Fürst P et al. (2008). Lack of neurodevelopmental adversity by prenatal exposure of infants to current lowered PCB levels: comparison of two German birth cohort studies. J Toxicol Environ Health A. 71:700–2. Wilson ND, Shear NM, Paustenbach DJ, Price PS (1998). The effect of cooking practices on the concentration of DDT and PCB compounds in the edible tissue of fish. J Expos Anal Environ Epidemiol. 8:423–40. Windham GC, Pinney SM, Sjodin A, Lum R, Jones RS, Needham LL et al. (2010). Body burdens of brominated flame retardants and other persistent organohalogenated compounds and their descriptors in U.S. girls. Environ Res. 110:251–7. Wingfors H, Hansson M, Papke O, Bergek S, Nilsson C, Haglund P (2005). Sorbent-assisted liquid–liquid extraction (Chem-Elut) of polychlorinated biphenyls, dibenzo-p-dioxins and dibenzofurans in the lipid fraction of human blood plasma. Chemosphere. 58:311–20. Winneke G, Krämer U, Sucker K, Walkowiak J, Fastabend A, Heinzow B et al. (2005). PCB-related neurodevelopmental deficit may be transient: follow-up of a cohort at 6 years of age. Environ Toxicol Pharmacol. 19:701–6. Winneke G, Ranft U, Wittsiepe J, Kasper-Sonnenberg M, Fürst P, Krämer U et al. (2014). Behavioral sexual dimorphism in school-age children and early developmental exposure to dioxins and PCBs: a follow-up study of the Duisburg cohort. Environ Health Perspect. 122:292–8. Witczak A, Ciereszko W (2006a). Effect of smoking process on changes in the content of selected nonortho- and mono-ortho-PCB congeners in mackerel slices. J Agric Food Chem. 54:5664–71. Witczak A, Ciereszko W (2006b). Changes in total PCB content in selected fish products during hot- and cold smoking. Acta Ichthyol Piscatoria. 36:11–6. Wojtyniak BJ, Rabczenko D, Jönsson BA, Zvezday V, Pedersen HS, Rylander L et al. (2010). Association of maternal serum concentrations of 2,2′,4,4′,5,5′-hexachlorobiphenyl (CB-153) and 1,1-dichloro-2,2bis(p-chlorophenyl)-ethylene (p,p′-DDE) levels with birth weight, gestational age and preterm births in Inuit and European populations. Environ Health. 9:56. Wolff MS, Fischbein A, Selikoff IJ (1992). Changes in PCB serum concentrations among capacitor manufacturing workers. Environ Res. 59:202–16.

333

Safety evaluation of certain food additives and contaminants Eightieth JECFA

Wolff MS, Camann D, Gammon M, Stellman SD (1997). Proposed PCB congener groupings for epidemiological studies. Environ Health Perspect. 105:13–14. Wolff MS, Engel S, Berkowitz G, Teitelbaum S, Siskind J, Barr DB et al. (2007). Prenatal pesticide and PCB exposures and birth outcomes. Pediatr Res. 61:243–50. Wolff MS, Britton JA, Boguski L, Hochman S, Maloney N, Serra N et al. (2008). Environmental exposures and puberty in inner-city girls. Environ Res. 107:393–400. Wu H, Bertrand KA, Choi AL, Hu FB, Laden F, Grandjean P et al. (2013). Persistent organic pollutants and type 2 diabetes: a prospective analysis in the Nurses’ Health Study and meta-analysis. Environ Health Perspect. 121:153–61. Wu K, Xu X, Liu J, Guo Y, Huo X (2011). In utero exposure to polychlorinated biphenyls and reduced neonatal physiological development from Guiyu, China. Ecotoxicol Environ Saf. 74:2141–7. Wyndham C, Devenish J, Safe S (1976). The in vitro metabolism, macromolecular binding and bacterial mutagenicity of 4-chorobiphenyl, a model PCB substrate. Res Commun Chem Pathol Pharmacol. 15:563–70. Xie W, Wang K, Robertson LW, Ludewig G (2010). Investigation of mechanism(s) of DNA damage induced by 4-monochlorobiphenyl (PCB3) metabolites. Environ Int. 36:950–61. Xing GH, Wu SC, Wong MH (2010). Dietary exposure to PCBs based on food consumption survey and food basket analysis at Taizhou, China – the world’s major site for recycling transformers. Chemosphere. 81:1239–44. Xiong Q, Wilson WK, Pang J (2007). The Liebermann-Burchard reaction: sulfonation, desaturation, and rearrangment of cholesterol in acid. Lipids. 42:87–96. Xu W, Cai Z (2015). Risk assessment of dietary exposure to PCDD/Fs, DL-PCBs and NDL-PCBs of Hong Kong resident. Sci China Chem. 58(6):1082–8. Xu Y, Yu RM, Zhang X, Murphy MB, Giesy JP, Lam MH et al. (2006). Effects of PCBs and MeSO2-PCBs on adrenocortical steroidogenesis in H295R human adrenocortical carcinoma cells. Chemosphere. 63:772–84. Yakushiji T, Watanabe I, Kuwabara K, Tanaka R, Kashimoto T, Kunita N (1984). Rate of decrease and half-life of polychlorinated biphenyls (PCBs) in the blood of mothers and their children occupationally exposed to PCBs. Arch Environ Contam Toxicol. 13:341–5. Yamazaki K, Suzuki M, Itoh T, Yamamoto K, Kanemitsu M, Matsumura C et al. (2011). Structural basis of species differences between human and experimental animal CYP1A1s in metabolism of 3,3′,4,4′,5-pentachlorobiphenyl. J Biochem. 149:487–94. Yang CY, Yu ML, Guo HR, Lai TJ, Hsu CC, Lambert G et al. (2005). The endocrine and reproductive function of the female Yucheng adolescents prenatally exposed to PCBs/PCDFs. Chemosphere. 61:355–60. Yang CY, Wang YJ, Chen PC, Tsai SJ, Guo YL (2008). Exposure to a mixture of polychlorinated biphenyls and polychlorinated dibenzofurans resulted in a prolonged time to pregnancy in women. Environ Health Perspect. 116:599–604. Yang CY, Huang TS, Lin KC, Kuo P, Tsai PC, Guo YL (2011). Menstrual effects among women exposed to polychlorinated biphenyls and dibenzofurans. Environ Res. 111:288–94.

334

WHO Food Additives Series No. 71-S1, 2016

Supplement 1: Non-dioxin-like polychlorinated biphenyls

Yard EE, Terrell ML, Hunt DR, Cameron LL, Small CM, McGeehin MA et al. (2011). Incidence of thyroid disease following exposure to polybrominated biphenyls and polychlorinated biphenyls, Michigan, 1974–2006. Chemosphere. 84:863–8. Yilmaz B, Sandal S, Chen C-H, Carpenter DO (2006). Effects of PCB 52 and PCB 77 on cell viability, [Ca2+]i levels and membrane fluidity in mouse thymocytes. Toxicology. 217:184–93. Yorita Christensen KL, White P (2011). A methodological approach to assessing the health impact of environmental chemical mixtures: PCBs and hypertension in the National Health and Nutrition Examination Survey. Int J Environ Res Public Health. 8:4220–37. Yoshimura T (2012). Yusho: 43 years later. Kaohsiung J Med Sci. 28:S49–52. Yurdakok B, Tekin K, Daskin A, Filazi A (2015). Effects of polychlorinated biphenyls 28, 30 and 118 on bovine spermatozoa in vitro. Reprod Domest Anim. 50:41–7. Zabik ME, Booren A, Zabik MJ, Welch R, Humphrey H (1996). Pesticide residues, PCBs and PAHs in baked, charbroiled, salt boiled and smoked Great Lakes lake trout. Food Chem. 55:231–9. Zani C, Donato F, Magoni M, Feretti D, Covolo L, Vassallo F et al. (2013). Polychlorinated biphenyls, glycaemia and diabetes in a population living in a highly polychlorinated biphenyls–polluted area in northern Italy: a cross-sectional and cohort study. J Public Health Res. 2:2–8. Zettner MA, Flor S, Ludewig G, Wagner J, Robertson LW, Lehmann L (2007). Quinoid metabolites of 4-monochlorobiphenyl induce gene mutations in cultured Chinese hamster V79 cells. Toxicol Sci. 100:88–98. Zhang LI, Bao S, Merzenich MM (2002). Disruption of primary auditory cortex by synchronous auditory inputs during a critical period. Proc Natl Acad Sci USA. 99:2309–14. Zhao G, Xu Y, Li W, Han G, Ling B (2007). Prenatal exposures to persistent organic pollutants as measured in cord blood and meconium from three localities of Zhejiang, China. Sci Total Environ. 377:179–91. Zhao S, Narang A, Ding X, Eadon G (2004). Characterization and quantitative analysis of DNA adducts formed from lower chlorinated PCB-derived quinones. Chem Res Toxicol. 17:502–11. Zhou J, Zhang J, Xie W (2005). Xenobiotic nuclear receptor–mediated regulation of UDP-glucuronosyltransferases. Curr Drug Metab. 6:289–98. Zimmer KE, Gutleb AC, Lyche JL, Dahl E, Oskam IC, Krogenaes A et al. (2009). Altered stress-induced cortisol levels in goats exposed to polychlorinated biphenyls (PCB 126 and PCB 153) during fetal and postnatal development. J Toxicol Environ Health A. 72:164–72. Zubero MB, Aurrekoetxea JJ, Murcia M, Ibarluzea JM, Goñi F, Jiménez C et al. (2015). Time trends in serum organochlorine pesticides and polychlorinated biphenyls in the general population of Biscay, Spain. Arch Environ Contam Toxicol. 68:476–88. Zuccato E, Calvarese S, Mariani G, Mangiapan S, Grasso P, Guzzi A et al. (1999). Level, sources and toxicity of polychlorinated biphenyls in the Italian diet. Chemosphere. 38:2753–65. Zuccato E, Grassi P, Davoli E, Valdicelli L, Wood D, Reitano G et al. (2008). PCB concentrations in some foods from four European countries. Food Chem Toxicol. 46:1062–7.

335

336 Concentration (ng/kg ww) PCB 52 Min ND 0.007 0.003 0.012 0.006 47.0 0.005 0.002 0.027 ND ND ND ND ND ND ND ND ND 3 450 12  000 10  600 ND ND 300 4 600 310 4 500 ND ND ND ND ND ND ND ND ND 5 590 5 000 23  500 ND ND 477 1 850 ND 16 100 ND ND ND ND ND ND ND ND ND 34  820 25  000 72  800 ND ND 300 3 310 ND 4 110 ND ND ND ND ND ND ND ND ND 0.020 190 57.0 0.160 60.0 670 180 0.29 0.007 0.009 0.009 0.009 41.0 0.005 0.003 0.080 0.066 530 200 0.110 100 890 680 1.00 0.008 0.022 0.012 0.010 30.0 0.004 0.006 0.140 0.541 960 370 0.092 330 4  000 2 400 2.10 0.007 0.045 0.020 0.019 36.0 0.008 0.010 0.370 0.976 1  300 550 0.170 490 6 600 400 4.40 31  000 32  000 126  000 ND ND 1  050 4 800 ND 5 060 ND ND ND ND ND ND ND Max Min Max Min Max Min Max Min ND 0.003 0.014 0.009 0.008 8.90 0.002 0.004 0.210 ND ND ND ND ND ND ND ND ND PCB 101 PCB 138 PCB 153 PCB 180 Max ND 190 760 220 0.051 190 2  000 1  500 1.70 34  680 21  000 25  700 ND ND 600 1 900 ND 3 820

Appendix 1. Indicator PCB concentrations for each food category tested by country

Note: In some cases, items have been moved to a different category to more accurately reflect food type (e.g. olive oil moved from fruit and fruit products to fats and oils).

PCB 28

Min

Max

ND

ND

Country / food category Australia Fish and products Austria Cereals and products Eggs and egg products Fish and products Food for the very young Herbs, spices, etc. Meat and meat products Milk and dairy products Special nutritional use Belgium Eggs and egg products Fats and oils Fish and products Food for the very young Legumes and pulses Meat and meat products Milk and dairy products Other foods Special nutritional use

0.009 0.010 0.023 0.006 65.0 0.006 0.003 0.052

0.032 730 3.20 0.320 80.0 3  000 280 0.170

ND ND ND ND ND ND ND ND ND

3 400 3 000 5 440 ND ND 300 300 ND 686

Concentration (ng/kg ww) PCB 52 Min 0.142 2.09 2.21 1.75 0.485 1.74 1.53 0.115 4.92 17.7 9.19 6.54 ND ND ND ND ND ND ND ND ND ND ND ND 13.1 ND 30.5 ND 61.9 ND 70.0 39.8 12.7 ND ND ND ND ND 14.7 22.2 2 960 43.2 24.9 83.8 ND ND ND ND ND ND 55.5 66.7 2 850 16.6 101 45.2 ND ND ND ND ND ND 39.5 1 250 3 850 153 39.5 410 223 12.2 11.4 ND ND ND ND ND ND ND ND ND ND ND 99.7 84.4 4 750 67.1 168 305 102 14.1 16.2 ND ND ND ND ND ND ND ND ND ND ND 7.51 2.68 7.45 344 14.7 3.24 51.8 12.3 8.41 84.8 16.1 10.2 NR NR NR NR NR NR NR NR NR NR NR NR NR NR NR NR NR NR NR NR NR NR NR NR 0.142 14.5 2.36 20.0 0.415 0.142 5.16 0.126 4.30 2.20 0.793 41.1 44.2 34.2 7.78 1 240 12.0 42.9 72.6 181 7.66 8.49 11.5 44.2 0.140 15.4 2.21 21.4 0.414 4.68 5.25 0.122 2.10 4.74 0.693 45.6 48.8 33.0 5.99 1 180 14.3 8.18 69.7 205 8.63 8.16 13.3 51.8 0.056 8.09 0.779 9.06 0.236 1.53 2.87 0.041 1.10 2.16 0.325 17.3 32.2 16.8 4.34 478 4.87 4.55 30.9 58.4 6.14 4.17 7.70 18.6 39.6 43.8 4 250 24.3 49.6 116 89.8 5.59 11.8 ND Max Min Max Min Max Min Max Min Max PCB 101 PCB 138 PCB 153 PCB 180

PCB 28

Min

Max

Country / food category Canada Composite food Eggs and egg products Fats and oils Fish and products Food for the very young Herbs, spices, etc. Meat and meat products Milk and dairy products Nuts and oilseeds Snacks and desserts Starchy roots and tubers Sugar and confectionary China, mainland Cereals and products Eggs and egg products Fish and products Food for the very young Legumes and pulses Meat and meat products

0.220 14.6 1.91 2.65 0.466 2.90 1.88 0.103 6.44 28.8 4.31 8.19

9.10 45.9 15.1 126 8.76 3.77 21.0 11.1 14.9 32.0 8.86 11.3

ND ND ND ND ND ND

33.9 498 5 420 74.3 74.6 191

Milk and dairy products

ND

67.2

Starchy roots and tubers

ND

70.5

Vegetables and products

ND

73.6

Supplement 1: Non-dioxin-like polychlorinated biphenyls

China, Hong Kong SAR

Cereals and products

ND

ND

337

338 Concentration (ng/kg ww) PCB 52 Min ND ND ND ND ND ND ND ND ND ND ND ND ND ND 0.006 ND ND 0.011 ND ND ND ND ND 10.0 ND ND 0.595 ND 0.005 12.8 ND ND ND 0.005 38.8 1.78 0.009 10.0 ND ND ND ND 0.040 82.1 4 650 0.012 725 104 0.020 ND ND 0.004 ND ND 0.007 94.1 5 220 0.010 675 120 0.015 ND ND 0.008 ND ND 0.009 249 15 010 0.015 1 050 567 0.021 0.007 14.7 2.92 0.016 0.136 ND ND 0.008 ND ND 0.016 ND ND ND ND ND 406 34 900 0.016 2 290 702 0.025 0.010 44.8 3.99 ND 258 ND ND 0.003 ND ND 0.006 ND ND ND ND ND 1 000 7 910 0.006 928 165 0.011 ND 183 1.35 0.004 74.8 Max ND ND ND 850 ND ND ND ND ND ND ND ND Min ND ND ND ND ND ND ND ND ND ND ND ND Max ND ND ND 1 500 ND ND ND ND ND ND ND ND Min ND ND ND ND ND ND ND ND ND ND ND ND Max ND ND ND 1 500 ND 60.0 200 ND ND ND ND 40.0 Min ND ND ND ND ND ND ND ND ND ND ND ND Max ND ND ND 2 400 ND 130 280 ND ND ND ND 60.0 Min ND ND ND ND ND ND ND ND ND ND ND ND Max ND ND ND 840 ND ND ND ND ND ND ND ND PCB 101 PCB 138 PCB 153 PCB 180 Safety evaluation of certain food additives and contaminants Eightieth JECFA

WHO Food Additives Series No. 71-S1, 2016

PCB 28

Min ND ND ND ND ND ND ND ND ND ND ND ND

Max ND ND ND 870 ND ND ND ND ND ND ND ND

Country / food category Composite food Eggs and egg products Fats and oils Fish and products Herbs, spices, etc. Meat and meat products Milk and dairy products Non-alcoholic beverages Snacks and desserts Starchy roots and tubers Stimulant beverages Sugar and confectionary Cyprus Eggs and egg products Fish and products Food for the very young Meat and meat products Milk and dairy products Other foods Czech Republic

ND ND 0.013 ND ND 0.050

377 19 900 0.026 609 252 0.084

Alcoholic beverages

ND

ND

Cereals and products

ND

85.1

Composite food

ND

0.043

Drinking-water

ND

ND

Eggs and egg products

ND

96.1

Concentration (ng/kg ww) PCB 52 Min ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND 974 974 3 630 3 630 1 260 3.13 2 320 0.023 0.038 1 100 ND ND ND ND ND ND 860 5.55 1 250 0.002 0.120 1 400 ND 0.006 ND ND ND ND 6 740 2 640 9.04 2 810 0.156 0.565 1 000 6 740 ND ND ND ND ND ND 10 300 3 130 10.6 2 300 0.231 0.465 1 000 10 300 ND ND ND ND ND ND 3 850 1 310 2.61 549 0.072 0.074 1 000 3 850 Max 20.0 20.0 ND ND ND 0.053 ND 2 830 15.0 0.011 ND 0.344 ND ND 0.290 20 Min ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND Max 20.0 101 ND 0.050 ND 0.131 ND 1 960 128 ND 0.092 0.056 0.027 ND 0.492 20.0 Min ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND Max 29.2 147 0.139 0.081 ND ND ND 449 282 0.007 0.184 0.061 0.056 ND 0.486 20.0 Min ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND Max 184 531 0.139 0.086 ND ND ND 815 529 0.004 0.188 0.128 0.064 ND 0.407 20.0 Min ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND Max 27.3 201 0.102 ND ND 0.010 ND 206 247 ND 0.178 0.059 0.007 ND 0.404 20.0 PCB 101 PCB 138 PCB 153 PCB 180

PCB 28

Min ND ND ND ND ND ND ND ND ND ND 0.236 ND ND ND ND ND

Max 214 20.0 0.101 0.045 ND 0.178 0.062 1 370 180 0.004 0.763 0.072 0.079 ND 0.192 111

Country / food category Fats and oils Fish and products Food for the very young Fruit and fruit products Fruit and vegetable juices Herbs, spices, etc. Legumes and pulses Meat and meat products Milk and dairy products Non-alcoholic beverages Nuts and oilseeds Snacks and desserts Starchy roots and tubers Stimulant beverages Sugar and confectionary Vegetables and products Estonia Eggs and egg products Fats and oils Fish and products Meat and meat products Milk and dairy products Sugar and confectionary Finland Composite food

ND ND ND ND ND ND

1 630 1.60 1 010 0.024 0.003 1 000

Supplement 1: Non-dioxin-like polychlorinated biphenyls

825

825

339

340 Concentration (ng/kg ww) PCB 52 Min 3.68 ND 308 1.28 ND ND 0.005 ND 0.001 ND ND ND ND ND ND ND ND ND ND ND 2.24 ND ND ND 8.43 600 912 10 800 189 000 150 33.1 ND 9.40 2 710 308 ND ND ND ND ND ND ND ND ND 5.82 ND ND ND 13 150 287 22 600 243 000 90.0 31.0 ND 19.9 240 000 855 ND ND ND ND ND ND 20.0 ND ND 7.24 ND ND ND 18.8 1 300 5 000 37 800 255 000 50.0 30.8 ND 40.4 12 900 5 740 ND ND ND ND ND ND 20.0 ND ND 6.15 ND ND ND 18.6 1 800 4 000 44 800 411 000 70.0 31.0 ND 32.9 17 300 16 900 81.9 0.269 25 870 0.028 14.2 0.720 ND 0.008 ND 0.001 ND ND 66.0 0.291 31 100 0.061 14.6 1.54 ND 0.006 ND 0.002 ND ND 1 000 0.141 70 400 0.088 3 150 64.2 ND 0.010 ND 0.008 ND ND 1 120 0.275 134 000 0.216 3 270 68.7 ND 0.004 ND 0.002 ND ND ND ND ND ND ND ND 10.0 ND ND 1.78 ND ND 1 170 0.068 29 300 0.028 90.9 34.2 ND 9.93 900 5 990 21 800 165 000 40.0 10.4 ND 14.7 198 000 4 210 Max 6.82 5 790 1 510 71.6 77.5 Min 4.90 ND 423 0.836 ND Max 15.0 25 500 3 140 8.44 13.8 Min 30.8 ND 385 2.31 ND Max 209 223 000 4 143 287 442 Min 42.7 ND 577 3.74 ND Max 261 359 000 5 920 796 529 Min 15.5 ND 115 0.900 ND Max 82.9 130 000 1 740 840 190 PCB 101 PCB 138 PCB 153 PCB 180 Safety evaluation of certain food additives and contaminants Eightieth JECFA

WHO Food Additives Series No. 71-S1, 2016

PCB 28

Min 19.3 ND 346 3.85 ND

Max 121 3 050 1 110 563 1 020

Country / food category Eggs and egg products Fish and products Food for the very young Meat and meat products Milk and dairy products France Eggs and egg products Fats and oils Fish and products Food for the very young Meat and meat products Milk and dairy products Germany Alcoholic beverages Cereals and products Composite food Eggs and egg products Fats and oils Fish and products Food for the very young Fruit and fruit products Fruit and vegetable juices Legumes and pulses Meat and meat products Milk and dairy products

ND 0.006 ND <0.001 ND ND

52.0 0.17 9 700 0.016 400 0.785

ND ND ND ND ND ND ND 1.16 ND 2.83 ND ND

ND 12.3 400 4 070 15 300 59 100 70 175 ND 15.2 16 100 822

Concentration (ng/kg ww) PCB 52 Min ND 0.750 ND ND ND ND ND ND 0.010 ND ND 75.1 ND 520 ND ND ND 417 4.00 10.0 10.0 ND 10.0 10.0 53.0 50.0 20.0 ND 60.0 20.0 88.7 10.0 10.0 8.78 75.1 6.88 19 700 6 230 28 52.8 6 580 4.00 ND ND 1 660 ND ND ND 15 700 4.00 ND 13.5 35 400 9 130 73.5 52.8 57 100 4.00 ND 243 2 720 ND ND 33.2 33 200 4.00 ND 987 77 200 14 700 222 1 670 373 500 4.00 10.0 10.0 5 490 ND 251 3 550 ND 19.7 38.0 129 000 ND 10.0 10.0 8.78 ND 1 020 66 100 17 000 232 2 110 499 000 ND 10.0 10.0 10 100 6 710 62 300 9 060 0.111 2 710 1 250 ND ND ND 0.009 ND ND 8 410 20 600 3 110 0.109 10 400 1 820 ND ND ND 0.007 ND ND 3 480 58 400 3 730 0.126 10 730 2 240 ND ND ND 0.009 ND ND 1 700 36 900 4160 0.155 14 400 797 ND ND ND 0.004 ND ND ND 85.9 1 190 ND 5.23 12.2 115 000 ND 10.0 10.0 3.25 7 490 16 500 1 300 0.052 9 110 456 ND 634 25 100 5 650 114 807 365 000 ND 10.0 10.0 15 800 Max 1 400 14.1 ND ND 81.8 Min ND 1.65 ND ND ND Max 4 800 20.0 ND ND 246 Min 330 1.84 ND ND ND Max 7 900 34.6 ND ND 494 Min 300 1.74 ND ND ND Max 10 900 30.4 ND ND 323 Min 450 0.530 ND ND ND Max 4 000 10.2 ND ND 79.3 PCB 101 PCB 138 PCB 153 PCB 180

PCB 28

Min ND 1.37 ND ND ND

Max 300 62.9 ND ND 85.5

ND ND ND 0.006 ND ND

6 800 18 900 444 0.053 2 700 5 250

Country / food category Special nutritional use Starchy roots and tubers Stimulant beverages Sugar and confectionary Vegetables and products Greece Eggs and egg products Fats and oils Fish and products Food for the very young Meat and meat products Milk and dairy products Iceland Cereals and products Eggs and egg products Fats and oils Fish and products Meat and meat products Milk and dairy products Special nutritional use Starchy roots and tubers Ireland Cereals and products Composite food Eggs and egg products

353 ND 100 ND ND ND ND 8.00

353 37.0 4 530 1 600 96.3 ND 500 8.00

Supplement 1: Non-dioxin-like polychlorinated biphenyls

10.0 10.0 ND

10.0 10.0 3 230

341

342 Concentration (ng/kg ww) PCB 52 Min ND ND ND ND 0.477 0.316 ND ND ND ND ND ND ND ND ND NR ND 21.2 ND 0.001 16.0 47.5 314 120 000 96.4 21.2 ND 0.001 9.12 5 700 1 230 11 700 95.0 140 230 000 110 NR ND 5 700 ND 3 700 208 ND 0.003 286 5 800 21 500 745 148 482 000 656 57 2 780 267 51.0 0.005 378 490 ND 5 700 20.2 ND ND ND ND ND ND ND ND 490 ND 15 200 75.0 ND ND ND ND ND ND ND ND 490 ND 31 000 28.2 ND ND ND ND ND ND ND ND 490 ND 51 000 45.5 ND ND 10 000 17 200 985 357 650 000 718 ND ND ND ND ND ND 16 1 500 160 ND 0.002 170 1 270 ND 6 500 41.4 ND ND 2 400 14 300 768 123 280 000 392 Max 145 1 840 ND ND 21.9 112 8 260 ND ND Min ND ND ND ND 0.477 0.316 ND ND ND Max 307 3 730 ND ND 9.50 183 22 700 ND ND Min ND ND ND ND 4.56 1.88 ND ND ND Max 1 410 5 700 ND ND 119 199 33 800 ND ND Min ND ND ND ND 3.05 2.45 ND ND ND Max 1 480 9 460 ND ND 188 239 51 100 ND ND Min ND ND ND ND 2.03 0.840 ND ND ND Max 355 2 360 ND ND 73.6 87.7 14 900 ND ND PCB 101 PCB 138 PCB 153 PCB 180 Safety evaluation of certain food additives and contaminants Eightieth JECFA

WHO Food Additives Series No. 71-S1, 2016

PCB 28

Min ND ND ND ND 0.955 0.525 ND ND ND

Max 396 742 ND ND 39.0 47.5 2 330 ND ND

ND ND ND ND ND ND

490 ND 2 500 5.63 ND ND

4.7

1 000

Country / food category Fats and oils Fish and products Fruit and fruit products Herbs, spices, etc. Meat and meat products Milk and dairy products Special nutritional use Starchy roots and tubers Vegetables and products Italy Eggs and egg products Fats and oils Fish and products Meat and meat products Milk and dairy products Sugar and confectionary Japan Fish and products Luxembourg Fish and products Netherlands Eggs and egg products Fats and oils Fish and products Meat and meat products

262

3 630

66.1 111 <0.001 30.8

100 991 19 400 158

Concentration (ng/kg ww) PCB 52 Min ND 80 NR 12.8 ND ND ND 5.34 ND ND ND 26.1 61.2 186 ND ND 0.009 <0.001 ND ND ND 1 280 ND 3 230 ND 19.2 2 700 106 9.28 ND ND ND ND ND ND ND ND 11.1 9 570 87.9 4.88 0.009 ND ND ND ND ND ND ND ND 68.6 19 200 312 45.0 ND 4 540 ND ND ND 0.005 0.009 ND ND ND 103 22 500 766 46.4 0.015 6 480 228 63.3 1 770 48 000 5.34 ND 378 44.0 26.1 61.2 186 9.98 ND ND ND 3.32 ND ND ND 34.8 48.1 301 384 67.4 2 910 125 000 3.32 ND 681 157 34.8 48.1 301 8.66 9.20 ND ND 2.54 ND ND ND 167 229 210 252 645 7 060 237 000 2.54 ND 4 190 1 090 167 229 210 12.8 18.7 ND ND 3.62 ND ND ND 221 321 331 390 514 7 950 300 000 3.62 ND 4 750 944 221 321 331 8.45 ND ND ND 1.23 ND ND ND 104 141 60.8 ND ND ND ND 0.002 ND ND NR ND ND NR NR NR NR NR NR 73.7 249 2 100 65 000 1.23 ND 1 750 566 104 141 60.8 ND 58.6 6 230 502 32.5 ND 2 770 Max 281 100 Min ND 100 Max 319 160 Min 32.9 680 Max 1 370 860 Min 40.3 1 000 Max 1 800 1 200 Min 13.7 400 Max 732 520 PCB 101 PCB 138 PCB 153 PCB 180

PCB 28

Min ND 200

Max 431 220

ND

ND

11.1 ND ND ND 3.80 ND ND ND 13.4 56.6 41.6

45.4 89.9 524 15 000 3.80 ND 273 251 13.4 56.6 41.6

Country / food category Milk and dairy products Snacks and desserts New Zealand Milk and dairy products Norway Cereals and products Eggs and egg products Fats and oils Fish and products Fruit and fruit products Legumes and pulses Meat and meat products Milk and dairy products Other foods Snacks and desserts Vegetables and products Poland Composite food Eggs and egg products Fish and products Meat and meat products Milk and dairy products Sugar and confectionary Republic of Korea Fish and products

ND ND ND 0.002 0.003 ND

ND 150 2 230 296 55.8 ND

Supplement 1: Non-dioxin-like polychlorinated biphenyls

ND

15 300

343

344 Concentration (ng/kg ww) PCB 52 Min ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND 2 540 0.017 610 000 0.001 70 ND ND ND ND ND ND ND 2 500 60.0 1  200  000 0.001 1 020 ND ND ND ND ND 60.0 101 0.042 0.017 6.36 0.056 0.004 ND ND ND ND ND 0.044 0.016 7.03 0.046 0.006 ND ND ND ND ND 0.657 0.063 16.2 0.299 0.626 1 580 1  800 250  000 50.0 980 358 ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND 60.0 133 ND 1.01 0.092 40.8 0.571 0.908 1 370 2  000 812  000 70.0 1 200 449 ND ND ND ND ND ND ND ND ND ND ND 45.0 ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND 0.587 0.036 12.6 0.195 0.312 244 562 68  000 0.001 500 157 Max Min Max Min Max Min Max Min Max PCB 101 PCB 138 PCB 153 PCB 180 Safety evaluation of certain food additives and contaminants Eightieth JECFA

WHO Food Additives Series No. 71-S1, 2016

PCB 28

Min

Max

ND ND ND ND ND ND ND ND ND

ND ND ND ND ND ND ND ND ND

ND

ND

Country / food category Romania Cereals and products Composite food Fats and oils Fish and products Meat and meat products Milk and dairy products Other foods Stimulant beverages Sugar and confectionary Singapore Food for the very young Slovakia Eggs and egg products Fats and oils Fish and products Meat and meat products Milk and dairy products Slovenia Eggs and egg products Fats and oils Fish and products Food for the very young Meat and meat products Milk and dairy products

ND ND ND ND ND

0.288 0.081 7.70 0.150 0.064

35 ND ND ND ND ND

679 272 103 000 0.001 70 317

Concentration (ng/kg ww) PCB 52 Min ND 0.014 ND ND NR NR <0.001 NR ND 1 500 ND ND ND ND 0.002 ND ND ND 0.001 0.001 ND ND 22.5 1 100 ND ND 21.3 2 040 ND ND 18.6 5 360 ND ND 13.0 2 100 20.3 109 4 610 69.0 3.00 99.6 97.8 630 2.60 1.5 ND 6 200 ND ND ND ND 0.003 ND ND ND 0.002 0.005 24.0 8 800 210 1 430 18 200 160 3.60 99.6 97.8 3 100 2.10 2.20 ND 15 000 ND ND ND ND 0.002 ND ND ND 0.001 0.002 22 22 000 4 200 5 630 36 000 200 5.60 1 200 364 7 900 1.20 0.880 ND 16 000 ND ND ND ND 0.002 ND ND ND 0.002 0.003 36.0 23 000 4 900 7 900 56 000 290 6.50 2 760 485 10 400 2.00 1.60 25.6 6 730 NR NR <0.001 NR ND ND 0.001 ND ND ND 0.001 ND ND ND 0.003 ND ND ND 0.003 ND ND ND 0.001 ND ND ND 0.001 ND ND ND 0.001 ND ND 5 400 ND ND ND ND 0.001 ND ND ND <0.001 0.001 ND ND ND ND 0.001 ND 14.0 8 500 2 400 4 790 21 400 75.0 2.80 1 210 240 7 850 0.680 0.650 0.370 2 730 Max ND 120 ND ND Min ND 0.143 ND ND Max ND 750 44.0 ND Min ND 0.728 ND ND Max 703 2 000 195 ND Min ND 0.638 ND ND Max 624 2 300 230 ND Min ND 0.360 ND ND Max 464 1 500 147 ND PCB 101 PCB 138 PCB 153 PCB 180

PCB 28

Min ND 0.004 ND ND

Max 37 30 193 ND

ND ND <0.001 ND

ND ND <0.001 ND

Country / food category Nuts and oilseeds Special nutritional use Sugar and confectionary Vegetables and products Spain Eggs and egg products Fats and oils Fish and products Milk and dairy products Sweden Cereals and products Composite food Eggs and egg products Fats and oils Fish and products Food for the very young Fruit and fruit products Meat and meat products Milk and dairy products Special nutritional use Starchy roots and tubers Vegetables and products United Kingdom Cereals and products Composite food

ND 510 ND ND ND ND <0.001 ND ND ND <0.001 <0.001

11 730 220 174 1 500 33 5.6 297 97.8 190 6.1 2.3

Supplement 1: Non-dioxin-like polychlorinated biphenyls

0.367 ND

33.1 439

345

346 Concentration (ng/kg ww) PCB 52 Min ND ND ND ND ND ND 11.6 ND 0.264 110 ND Max 29.4 1 740 1 630 8.5 18.7 92.7 11.6 14 510 0.264 110 21.1 Min ND ND ND ND ND ND ND ND 0.116 80 ND Max 233 4 980 3 540 11.9 39.3 218 ND 15 100 0.116 80 32.1 Min 3.88 ND ND ND ND 0.960 ND ND 0.095 220 ND Max 15 800 13 690 7 050 8.70 1 320 543 ND 57 000 0.095 220 27.9 Min 3.88 ND ND ND ND ND 9.97 ND 0.106 240 ND Max 16 200 12 100 8 310 11.8 1 720 821 9.97 71 000 0.106 240 32.5 Min ND ND ND ND ND ND 8.31 ND ND 140 ND Max 8 790 4 360 8 150 0.093 403 190 8.31 15 100 ND 140 10.3 PCB 101 PCB 138 PCB 153 PCB 180 Safety evaluation of certain food additives and contaminants Eightieth JECFA

WHO Food Additives Series No. 71-S1, 2016

PCB 28

Country / food category Eggs and egg products Fats and oils Fish and products Fruit and fruit products Meat and meat products Milk and dairy products Nuts and oilseeds Special nutritional use Starchy roots and tubers Sugar and confectionary Vegetables and products

Min 2.72 ND ND ND ND ND 26.6 ND 0.423 160 ND

Max 374 1 150 1 110 12 664 60.3 26.6 1 280 0.423 160 27.7

Max: maximum; Min: minimum; ND: not detected; NR: not reported; SAR: Special Administrative Region; ww: wet weight

Supplement 1: Non-dioxin-like polychlorinated biphenyls

Appendix 2. Mean indicator PCB concentrations in foods reported by countries responding to JECFA call for data Note: In some cases, items have been moved to a different category to more accurately reflect food type (e.g. olive oil moved from fruit and fruit products to fats and oils). Country / food category Australia Fish and products Austria Cereals and products Eggs and egg products Fish and products Food for the very young Herbs, spices, etc. Meat and products Milk and dairy products Specialty products Belgium Eggs and egg products Fats and oils Fish and products Food for the very young Legumes and pulses Meat and products Milk and dairy products Other foods Specialty products Canada Composite food Eggs and egg products Fats and oils Fish and products Food for the very young Herbs, spices, etc. Meat and products Milk and dairy products Nuts and oilseeds Snacks and desserts Starchy roots, tubers Sugar and confectionary China, mainland Cereals and products Eggs and egg products PCB 28 ND 0.020 95.0 0.278 0.058 72.5 113 19.7 0.093 287 430 158 ND ND 14.5 19.0 ND 97.9 3.70 30.3 6.03 32.3 2.36 3.33 6.07 1.97 10.7 30.4 6.58 9.74 4.43 53.1 PCB 52 ND 0.014 28.6 ND 0.029 53.5 47.7 13.2 0.12 328 588 2.81 ND ND 19.8 281 62.0 321 3.19 2.38 4.12 75.6 3.36 2.49 6.89 2.52 6.66 51.3 12.6 8.37 1.50 2.70 Mean concentration (ng/kg ww) PCB 101 PCB 138 ND 0.036 81.4 9.55 0.034 70.5 67.0 40.6 0.504 396 467 1 740 ND ND 39.9 107 ND 1 290 NR NR NR NR NR NR NR NR NR NR NR NR 5.25 5.30 ND 0.274 224 17.6 0.040 180 278 160 0.820 1 660 936 960 ND ND 14.5 305 ND 349 9.72 24.4 5.28 266 2.97 5.98 23.3 26.1 5.35 6.34 6.16 42.7 2.40 37.4 PCB 153 ND 0.491 294 26.4 0.069 263 445 250 1.65 2 320 1 530 1 940 ND ND 54.8 547 ND 718 10.9 24.2 4.39 283 3.49 6.43 26.1 29.8 5.37 6.45 6.98 48.7 5.82 16.1 PCB 180 ND 0.192 149 10.4 0.028 99.5 166 99.4 0.792 1 420 793 427 ND ND 25.6 131 ND 469 6.02 12.4 3.05 93.4 1.46 3.04 12.8 9.57 3.62 3.17 4.01 18.0 1.87 6.11

347

Country / food category Fish and products Food for the very young Legumes and pulses Meat and products Milk and dairy products Starchy roots, tubers Vegetables and products China, Hong Kong SAR Cereals and products Composite food Eggs and egg products Fats and oils Fish and products Herbs, spices, etc. Meat and products Milk and dairy products Non-alcoholic beverages Snacks and desserts Starchy roots, tubers Stimulant beverages Sugar and confectionary Cyprus Eggs and egg products Fish and products Food for the very young Meat and products Milk and dairy products Other foods Czech Republic Alcoholic beverages Cereals and products Composite food Drinking-water Eggs and egg products Fats and oils Fish and products Food for the very young Fruit and fruit products Herbs, spices, etc. Legumes and pulses Meat and products Milk and dairy products Non-alcoholic drinks Nuts and oilseeds

PCB 28 97.8 9.86 13.6 11.3 6.55 7.54 9.82 ND ND ND ND 36.3 ND ND ND ND ND ND ND ND 21.1 1 550 0.018 77.5 17.9 0.067 ND 2.72 0.145 ND 16.1 17.8 0.498 0.011 0.002 0.059 0.010 31.6 6.24 0.001 0.500

PCB 52 75.9 7.47 3.77 5.38 3.83 2.57 2.51 ND ND ND ND 71.2 ND ND ND ND ND ND ND ND 6.06 376 0.008 80.8 6.57 0.016 ND 0.853 0.122 ND 1.67 1.45 0.503 ND ND 0.018 ND 62.7 0.686 0.003 ND

Mean concentration (ng/kg ww) PCB 101 PCB 138 76.5 131 2.28 11.4 5.98 2.90 2.32 12.2 2.08 7.35 3.14 1.62 2.08 1.27 ND ND ND ND 172 ND ND ND ND ND ND ND ND 7.23 484 0.007 69.3 7.41 0.011 0.001 3.28 0.305 0.007 1.68 1.47 2.49 ND 0.002 0.044 ND 41.6 3.10 ND 0.046 ND ND ND ND 170 ND 2.92 26.3 ND ND ND ND 10 15.3 1 350 0.012 97.2 23.7 0.015 0.002 1.25 0.406 0.008 0.087 3.83 4.32 0.027 0.007 ND ND 10.1 10.8 0.003 0.092

PCB 153 129 8.67 7.25 9.82 6.67 1.86 1.69 ND ND ND ND 359 ND 6.88 35.4 ND ND ND ND 15 24.2 3 170 0.013 218 40.5 0.021 0.002 3.28 0.546 ND 43.1 16.5 13.0 0.036 0.013 ND ND 18.6 18.5 0.002 0.094

PCB 180 52.3 2.99 2.03 3.51 2.90 0.768 0.670 ND ND ND ND 86.6 ND ND ND ND ND ND ND ND 47.0 629 0.005 71.0 11.2 0.008 ND 5.22 0.183 0.002 12.5 3.72 5.12 0.011 ND 0.003 ND 5.67 8.29 ND 0.089

348

Supplement 1: Non-dioxin-like polychlorinated biphenyls

Country / food category Snacks and desserts Starchy roots, tubers Stimulant beverages Sugar and confectionary Vegetables and products Estonia Eggs and egg products Fats and oils Fish and products Meat and products Milk and dairy products Sugar and confectionary Finland Composite food Eggs and egg products Fish and products Food for the very young Meat and products Milk and dairy products France Eggs and egg products Fats and oils Fish and products Food for the very young Meat and products Milk and dairy products Germany Alcoholic beverages Cereals and products Composite food Eggs and egg products Fats and oils Fish and products Food for the very young Fruit and fruit products Fruit, vegetable juices Legumes and pulses Meat and products Milk and dairy products Specialty products Starchy roots, tubers Stimulant beverages Sugar and confectionary Vegetables and products

PCB 28 0.012 0.014 ND 0.009 4.06 452 0.738 59.3 0.006 <0.001 83.3 825 36.6 350 797 62.8 177 8.94 0.045 60.4 0.006 2.61 0.108 ND 6.26 80.0 60.6 1 100 363 37.0 25.7 ND 6.42 273 51.7 64.3 10.7 ND ND 9.21

PCB 52 0.057 ND ND 0.008 0.874 467 1.52 78.0 0.004 0.003 91.7 974 5.19 628 999 10.3 16.6 3.85 0.052 134 0.009 0.142 0.112 ND 3.98 120 43.2 1 060 2 820 77.0 8.01 ND 4.71 127 0.347 254 3.37 ND ND 6.48

Mean concentration (ng/kg ww) PCB 101 PCB 138 0.009 0.020 0.002 0.003 ND ND 0.018 0.017 0.873 0.881 381 2.48 90.2 <0.001 0.011 117 3 630 8.22 2 700 1 840 3.88 0.922 5.52 0.061 356 0.016 0.147 0.146 ND 7.43 30.0 14.2 3 310 5 190 31.9 9.77 ND 10.6 526 30.5 936 6.01 ND ND 17.9 444 4.07 181 0.028 0.197 83.3 6 740 119 6 640 2 590 28.8 46.3 42.5 0.042 817 0.028 13.2 5.58 ND 9.38 260 339 13 600 6 090 28.9 9.51 ND 15.5 280 105 1 930 7.52 ND ND 33.6

PCB 153 0.039 0.010 ND 0.016 0.889 518 5.27 207 0.056 0.162 83.3 10 300 152 8 870 3 640 74.3 53.0 50.3 0.067 1540 0.056 13.9 6.55 ND 9.68 360 351 22 300 6 300 38.9 10.2 ND 13.7 358 136 2 650 7.27 ND ND 25.6

PCB 180 0.010 <0.001 ND 0.018 0.875 376 1.38 57.5 0.015 0.029 83.3 3 850 53.7 3 350 1 070 51.6 17.8 26.0 0.019 473 0.011 0.382 2.88 ND 3.35 180 269 7 060 5 390 21.5 2.67 ND 4.62 460 59.2 1 280 2.23 ND ND 7.48

349

Safety evaluation of certain food additives and contaminants Eightieth JECFA

Country / food category Greece Eggs and egg products Fats and oils Fish and products Food for the very young Meat and products Milk and dairy products Iceland Cereals and products Eggs and egg products Fats and oils Fish and products Meat and products Milk and dairy products Specialty products Starchy roots, tubers Ireland Cereals and products Composite food Eggs and egg products Fats and oils Fish and products Fruit and fruit products Herbs, spices, etc. Meat and products Milk and dairy products Specialty products Starchy roots, tubers Vegetables and products Italy Eggs and egg products Fats and oils Fish and products Meat and products Milk and dairy products Sugar and confectionary Japan Fish and products Luxembourg Fish and products Netherlands Eggs and egg products Fats and oils Fish and products

PCB 28 292 285 38.4 0.031 24.8 113 210 13.7 1 100 106 23.7 ND 225 8.00 10.0 10.0 75.2 67.8 241 ND ND 8.88 5.07 360 ND ND 18.1 ND 224 0.158 ND ND 89.4 740 83.0 362 2 450

PCB 52 787 1 340 196 0.068 37.1 52.0 75.1 2.53 5 890 287 7.35 6.21 3 520 4.00 10.0 10.0 4.46 22.8 654 ND ND 5.16 6.90 1 230 ND ND 18.1 ND 392 0.453 ND ND NR 1 490 34.4 96.3 12 590

Mean concentration (ng/kg ww) PCB 101 PCB 138 1 170 553 179 0.072 124 83.8 ND 4.76 10 800 464 14.3 11.7 35 650 4.00 53.3 20.0 7.83 65.1 1 430 ND ND 2.99 11.2 2 650 ND ND 18.1 ND 740 1.45 ND ND 458 5 000 58.1 75.2 23 500 387 602 311 0.071 188 102 ND 448 20 300 803 72.7 403 185 000 4.00 10.0 10.0 227 266 2 080 ND ND 29.6 13.7 6 790 ND ND 18.1 ND 2 390 0.553 ND ND 517 11 400 476 63.8 58 300

PCB 153 105 495 361 0.100 161 24.6 ND 466 22 500 872 86.5 495 329 000 ND 10.0 10.0 380 367 3 180 ND ND 54.0 16.3 9 680 ND ND 18.1 ND 2 210 0.901 ND ND 878 8 620 626 154 77 000

PCB 180 317 160 105 0.031 74.0 23.6 ND 233 8 420 292 34.9 171 265 000 ND 10.0 10.0 479 131 863 ND ND 13.9 5.10 3 110 ND ND 47.2 ND 535 0.785 ND ND 273 6 050 464 45.4 28 800

350

WHO Food Additives Series No. 71-S1, 2016

Supplement 1: Non-dioxin-like polychlorinated biphenyls

Country / food category Meat and products Milk and dairy products Snacks and desserts New Zealand Milk and dairy products Norway Cereals and products Eggs and egg products Fats and oils Fish and products Fruit and fruit products Legumes and pulses Meat and products Milk and dairy products Other foods Snacks and desserts Vegetables and products Poland Composite food Eggs and egg products Fish and products Meat and products Milk and dairy products Sugar and confectionary Republic of Korea Fish and products Romania Cereals and products Composite food Fats and oils Fish and products Meat and products Milk and dairy products Other foods Stimulant beverages Sugar and confectionary Singapore Food for the very young Slovakia Eggs and egg products Fats and oils Fish and products Meat and products Milk and dairy products

PCB 28 71.1 147 210 ND 22.7 39.6 54.4 2 420 3.80 ND 43.5 64.6 13.4 56.6 41.6 ND 20.9 158 7.95 4.84 ND 234 ND ND ND ND ND ND ND ND ND ND 0.067 0.020 0.310 0.011 0.006

PCB 52 40.3 82.4 90.0 NR 68.0 13.6 80.1 4 860 5.34 ND 37.7 11.0 26.1 61.2 186 ND 2.72 272 3.05 1.09 ND 163 ND ND ND ND ND ND ND ND ND ND 0.008 0.004 0.201 0.002 0.001

Mean concentration (ng/kg ww) PCB 101 PCB 138 56.1 424 72.8 562 130 770 ND 107 12.9 122 14 700 3.32 ND 53.9 33.4 34.8 48.1 301 ND 1.49 1 080 2.86 0.590 0.002 170 ND ND ND ND ND ND ND ND ND ND 0.012 0.003 0.231 0.003 0.002 NR 74.2 337 613 30 200 2.54 ND 402 342 167 229 210 ND 8.50 2 050 23.3 8.90 ND 194 ND ND ND ND ND ND ND ND ND ND 0.079 0.016 0.519 0.017 0.024

PCB 153 562 776 100 NR 111 280 412 37 200 3.62 ND 516 344 221 321 331 ND 18.3 2 560 51.7 12.5 0.009 216 ND ND ND ND ND ND ND ND ND ND 0.132 0.023 1.16 0.033 0.064

PCB 180 279 290 ND NR 28.6 97.7 211 7 730 1.23 ND 196 178 104 141 60.8 ND 11.6 571 25.5 4.75 ND 136 ND ND ND ND ND ND ND ND ND ND 0.061 0.009 0.373 0.017 0.026

351

Safety evaluation of certain food additives and contaminants Eightieth JECFA

Country / food category Slovenia Eggs and egg products Fats and oils Fish and products Food for the very young Meat and products Milk and dairy products Nuts and oilseeds Specialty products Sugar and confectionary Vegetables and products Spain Eggs and egg products Fats and oils Fish and products Milk and dairy products Sweden Cereals and products Composite food Eggs and egg products Fats and oils Fish and products Food for the very young Fruit and fruit products Meat and products Milk and dairy products Specialty products Starchy roots, tubers Vegetables and products United Kingdom Cereals and products Composite food Eggs and egg products Fats and oils Fish and products Fruit and fruit products Meat and products Milk and dairy products Nuts and oilseeds Specialty products Starchy roots, tubers Sugar and confectionary Vegetables and products

PCB 28 189 52.6 2 150 <0.001 24.2 45.3 4.63 12.0 38.6 ND ND ND <0.001 ND 0.657 655 35.8 18.3 344 4.34 2.44 15.9 4.55 17.3 2.73 0.714 11.9 67.2 52.5 110 112 3.48 7.93 5.76 26.6 332 0.423 160 2.00

PCB 52 508 1.90 10 800 <0.001 26.6 ND ND 46.1 ND ND NR NR 0.001 NR 0.663 1 830 3.84 8.15 795 7.39 1.24 7.76 4.49 57.3 1.20 0.463 7.64 170 2.91 155 184 1.70 1.57 2.83 11.6 3 810 0.264 110 1.52

Mean concentration (ng/kg ww) PCB 101 PCB 138 500 4.48 19 600 <0.001 246 ND ND 262 8.80 ND ND ND 0.001 ND 1.20 7 880 23.4 31.2 3 300 13.6 1.73 7.90 3.91 282 0.901 0.705 6.55 344 10.9 372 295 2.38 2.28 6.02 ND 4 960 0.116 80.0 2.30 388 238 8 850 1.79 353 199 87.9 701 39 ND ND ND 0.003 ND 1.10 19 800 394 189 6 050 16.5 1.58 79.6 87.5 719 0.533 0.318 3.84 805 729 1 180 722 1.75 77.2 85.2 ND 21 600 0.095 220 2.02

PCB 153 336 429 14 100 2.50 452 285 78.0 781 46.0 ND ND ND 0.001 ND 1.80 20 800 499 361 9 560 24.8 2.26 144 116 947 0.926 0.543 6.39 898 844 1 090 875 2.37 99.4 99.0 9.97 26 600 0.106 240 2.35

PCB 180 80.2 148 1 430 <0.001 174 99.3 58.0 501 29.4 ND ND ND 0.001 ND 0.701 7 230 168 155 3 120 6.68 0.875 56.9 42.1 714 0.280 0.188 0.086 243 370 575 427 0.020 33.8 36.4 8.31 6 010 ND 140 0.778

WHO Food Additives Series No. 71-S1, 2016 352

ND: not detected; NR: not reported; SAR: Special Administrative Region; ww: wet weight

Appendix 3. Summary of the concentration data for food groups and major contributors to dietary exposure for the national dietary exposure estimates for the sum of the six indicator PCB congeners estimated by the Committee using the CIFOCOss consumption data and concentration data from the GEMS/Food database Note: A blank per cent contribution in the tables below either represents a zero lowerbound concentration or indicates that the food group was not consumed by the population group. Table A3.1 Concentration and dietary exposure data for Belgium Mean concentration (ng/kg) Lower Upper bound bound 416 766 0 4 000 2 214 5 238 17 772 19 865 1 161 0 338 4 000 3 249 0 62 10 458 1 000 3 671 4 000 9 082 2 875 3 929 2 1 <1 <1 2 2 Contribution to dietary exposure (%) Other Adoleschildren cents Adults Elderly <1 <1 <1 <1 30 <1 <1 <1 3 10 9 <1 <1 1 18 10 1 <1 3 20 10 3 <1 5

Food name Dairy products, nes Fermented milks (plain) Cheese Animal or vegetable fats, nes 02.2.1 Butter 06.8.3 Soya bean curd (tofu) 08.0 Unprocessed meat and offal, nes 09.0 Fishes and aquatic animals, nes 13.6 Dietary supplements, food supplements 13.7 Food for infants and small children, nes 17 Other foods (foods which cannot be included in any other group) FA 0142 Marine animal fats, nes IM 0150 Molluscs, including cephalopods, nes IM 1003 Mussels IM 1005 Scallops IM 1008 Squids MF 0100 Mammalian fats (except milk fats), and skin, nes ML 0106 Other and nes milks

Food code 01.0 01.2.1 01.6 02.1

Toddlers <1 36 <1 <1 1 <1

Very elderly <1 19 9 4 <1 4

0 157 2 877 499 0 756 485

1 980 3 934 4 811 3 832 4 000 10 200 848 58 <1 <1

<1 <1 <1

<1 <1 <1

<1 1 <1 <1 <1

61

<1

<1

<1

<1

353

Safety evaluation of certain food additives and contaminants Eightieth JECFA

Table A3.1 (continued) Mean concentration (ng/kg) Lower Upper bound bound 485 848 2 935 6 760 378 378 910 910 139 286 317 0 236 0 0 0 0 0 3 546 233 3 373 8 095 53 401 0 2 374 6 800 7 316 0 624 7 595 10 272 182 379 4 123 850 1 740 5 208 1 981 1 715 1 030 1 605 1 477 10 213 3 855 3 373 14 031 2 712 3 401 4 000 4 435 8 400 9 418 4 000 3 291 7 995 11 072 3 682 4 045 5 723 Contribution to dietary exposure (%) Other Adoleschildren cents Adults Elderly 61 <1 <1 <1 64 45 33

WHO Food Additives Series No. 71-S1, 2016

Food code ML 0106 ML 0812 ML 0814 ML 0822 MM 0816 MM 0822 OR 0172 OR 0305 OR 0495 OR 0541 OR 0645 OR 0665 OR 0696 OR 0697 OR 0702 PE 0112 PE 0840 PM 0840 WC 0143 WC 0978 WD 0123 WD 0890 WF 0115 WF 0864 WS 0927 WS 0937 WS 0941 WS 0945 WS 0948 WS 0951

Food name Other and nes milks Cow milk Goat milk Sheep milk Horse and other equines Sheep and other ovines Vegetable oils, nes Olive oil Rape seed oil (incl. canola) Soya bean oil Maize oil Coconut oil Palm oil Peanut oil and butter Sunflower seed oil Eggs, nes Chicken eggs Chicken meat Crustaceans, nes Lobsters Trout Eels Freshwater fish, nes Perch Cod Herring Mackerel Plaice Rays Sole

Toddlers 58

Very elderly <1 40 <1 <1 <1

<1

<1 <1

<1 <1

<1 <1

<1 <1

<1

<1

<1

<1

<1

<1

<1 <1

<1 <1 <1 <1

<1

<1 <1 <1 11 <1 <1

<1 <1 <1 15 <1 <1

<1 <1 15 <1

<1 <1 14 <1

<1

<1

<1

<1 <1

<1 <1 1 <1

<1 <1 <1 <1

<1 1 1 <1 <1

<1 4 2 <1 3

<1 2 <1 <1 1

incl.: including; nes: not elsewhere specified

354

Supplement 1: Non-dioxin-like polychlorinated biphenyls

Table A3.2 Concentration and dietary exposure data for China Mean concentration (ng/kg) Food code 01.0 13.1.1.1 CF 1211 GC 0649 IM 1000 IM 1003 ML 0812 MM 0812 MM 0818 MM 0822 MO 1284 MO 1285 PE 0840 PM 0840 PE 0841 PM 0841 VB 0041 VC 0046 VC 0424 VC 0427 VD 0541 VL 0053 VL 0466 VL 0467 VL 0483 VL 0485 VL 0502 VO 0448 VP 0064 VR 0075 VR 0506 VR 0577 VR 0589 VS 0622 Food name Dairy products, nes Ready-to-eat meal for infants and young children Wheat flour Rice (excl. wild) Clams Mussels Cow milk Beef and other bovine meat Pork and other porcines Sheep and other ovines Pig kidney Pig liver Chicken eggs Chicken meat Duck eggs Duck meat Cabbages, head Melons, except watermelon Cucumber Loofah, angled Soya bean Leafy vegetables, unprocessed, nes Chinese cabbage, type pak-choi Chinese cabbage, type pe-tsai Lettuce, leaf Mustard greens Spinach Tomato Peas, shelled (succulent seeds) Root vegetables, unprocessed, nes Turnip, garden Carrot Potato Bamboo shoots Lower bound 57 43 36 14 769 798 20 29 61 0 5 4 100 3 344 7 37 41 22 63 37 17 0 8 0 0 10 11 28 0 0 0 21 12 Upper bound 57 43 36 14 769 798 20 29 61 1 5 4 101 5 345 9 37 41 22 63 37 17 1 8 1 1 10 12 28 0.4 0.4 0.3 21 12 2 2 Contribution to dietary exposure (%) General Children population

10 10 <1 <1 3 <1 13 <1 <1 12 <1 2 <1 <1 <1 <1 <1 <1 <1

10 10 <1 <1 2 <1 12 <1 <1 8 <1 1 <1 <1 <1 <1 <1 <1 <1

<1 <1 <1

<1 <1 <1

3 <1

3 <1

355

Safety evaluation of certain food additives and contaminants Eightieth JECFA

Table A3.2 (continued) Mean concentration (ng/kg) Food code WC 0979 WC 0978 WC 0146 WD 0890 WF 0115 WS 0125 WF 0859 WC 0143 Food name Shrimps or prawns Lobsters Crabs Eels Freshwater fish, nes Marine fish, nes Carps Crustaceans, nes Lower bound 147 608 488 0 580 663 0 555 Upper bound 147 608 488 0.9 581 663 0.4 555 Contribution to dietary exposure (%) General Children population 2 1 <1 2 3 26 13 27 17

excl.: excluding; nes: not elsewhere specified

Table A3.3 Concentration and dietary exposure data for Cyprus Food code 01.0 01.6 09.0 13.2b 13.7 17 IM 0150 ML 0812 ML 0814 ML 0822 MM 0095 MM 0812 MM 0818 MM 0819 MM 0822 PE 0112 PM 0840 WF 0115 WS 0130 WS 0952 Food name Dairy products, nes Cheese Fishes and aquatic animals, nes Cereal-based food for infants and young children Food for infants and small children, nes Other foods (foods which cannot be included in any other group) Molluscs, including cephalopods, nes Cow milk Goat milk Sheep milk Meat from mammals other than marine mammals, nes Beef and other bovine meat Pork and other porcines Rabbit meat Sheep and other ovines Eggs, nes Chicken meat Freshwater fish, nes Sardine and sardine-like fishes Tuna Mean concentration (ng/kg) Contribution to dietary exposure (%) Lower bound Upper bound Adolescents 0.1 0.1 <1 1 1 <1 3 3 <1 0.1 0.1 <1 0.1 0.1 4 536 820 0 0.6 0 882 1 931 768 756 1 901 27 460 3 12 0.1 0.1 4 3 870 7 820 18 600 0.6 10 000 2 882 5 265 768 756 1 901 27 460 3 12 <1 <1 56

WHO Food Additives Series No. 71-S1, 2016

14 1 <1 1 28

<1

nes: not elsewhere specified

356

Supplement 1: Non-dioxin-like polychlorinated biphenyls

Table A3.4 Concentration and dietary exposure data for Czech Republic Mean concentration (ng/kg) Food code 01.0 01.4.1 01.6 02.1 01.2.1 02.2.1 02.2.2 03.0 04.1.2.4 04.1.2.5 04.2.2 05.1 07.2.1 08.0 09.0 11.5 12.6.1a 12.6.1b 13.1.1.1 13.7 14.1 14.1.1.1 14.1.1.6 14.2.1 14.2.3.1 14.2.6 15.1 15.4 16.1 16.2 16.8 16.10 CF 1211 Food name Lower bound Upper bound Dairy products, nes 196 836 Cream 3 4 Cheese 1 215 Animal or vegetable fats, nes 16 61 Fermented milks (plain) 0.13 161 Butter 402 402 Margarine 1 1 Edible ices, including sherbet and 0.18 400 sorbet Canned or bottled (pasteurized) fruits 0 0.8 Jams, jellies, marmalades 0 0.8 Other vegetables, nes, other 70 71 processing Other cocoa products (incl. choco0.26 0.5 late), nes Cakes, cookies and pies (e.g. fruit0.32 0.4 filled or custard types) Unprocessed meat and offal, nes 385 477 Fishes and aquatic animals, nes 0 0.8 Honey 0 17 Mustard sauce 0.18 0.8 Mayonnaise 0 0.6 Ready-to-eat meal for infants and 0.11 172 young children Food for infants and small children, 0 2 nes Non-alcoholic (“soft”) beverages, nes 0.01 20 Natural mineral waters and source 0.03 0 waters Water, nes 0.01 0 Beer and malt beverages, nes 0 0 Still grape wine 0.01 0 Distilled spirituous beverages con0.01 20 taining more than 15% alcohol Potato crisps 0.17 0.6 Snacks, nes 0 800 Cereal-based composite food 0.09 0.2 Vegetable-based composite food 0.36 201 (incl. mushroom-based) Fish-based composite food 4 338 Composite food, nes 0.39 0.8 Wheat flour 0.02 0.3 Contribution to dietary exposure (%) Other children Adolescents Adults 5 4 5 <1 <1 35 <1 <1 <1 32 <1 <1 <1 27 <1

3 <1 <1 8

5 <1 <1 12

7 <1 <1 9

<1

<1

<1

<1 <1

<1 <1

<1 <1

<1 <1 <1 <1

<1 <1 <1 <1 <1

<1 <1 <1 <1 <1

<1 <1

<1 <1

<1 <1

357

Safety evaluation of certain food additives and contaminants Eightieth JECFA

Table A3.4 (continued) Mean concentration (ng/kg) Food code CP 0179 CP 0179a DF 5259 DM 0659 DM 0715 DT 1114 FB 0269 FB 0275 FC 0001 FC 0004 FI 0327 FI 0341 FP 0226 FP 0230 FS 0014 FS 0240 FS 0247 GC 0080 GC 0649 HS 0093 JF 0175 ML 0106 ML 0812 MM 0095 MM 0812 MM 0814 MM 0818 MM 0819 MO 0105 MO 1285 OR 0172 PE 0112 PE 0840 PF 0840 PM 0110 PM 0840 PM 0847 PM 0848 SO 0697 Food name Hominy/mugunzá Wheat white bread Vine fruits (currants, raisins and sultanas), dried Sugarcane molasses Cocoa powder Tea, dried leaves Grapes Strawberry Citrus fruits, nes Orange, sweet, sour + orange-like hybrid Banana Kiwi fruit Apple Pear Plums (incl. prunes) Apricot Peach Cereals grains, nes Rice (excl. wild) Spices and condiments, nes Fruit juice, nes Other and nes milks Cow milk Meat from mammals other than marine mammals, nes Beef and other bovine meat Goat and other caprines Pork and other porcines Rabbit meat Offal from mammals, nes Pig liver Vegetable oils, nes Eggs, nes Chicken eggs Chicken fat Poultry meat, nes Chicken meat Quail meat Turkey meat Peanut Lower bound Upper bound 30 143 40 40 0 400 0 1 0 0.04 0.04 0 0.01 0.02 0 0.04 0 0 0.06 0.07 0.02 0 0.19 0 0.31 3 0.04 0.78 0 211 0.09 1 50 0 0.48 230 478 0 22 0 30 0.33 220 201 20 0.4 400 400 267 267 0.8 267 400 0.8 400 400 150 267 0.6 0 134 39 5 134 0.9 244 51 1 51 0.4 0.6 430 478 3 122 3 31 0.9 Contribution to dietary exposure (%) Other children Adolescents Adults <1 <1 <1

<1 <1 <1 <1 <1 <1

<1 <1 <1 <1 <1 <1

<1 <1 <1 <1 <1 <1

<1 <1 <1 <1 <1 4

<1 <1 <1 <1 <1 2

<1 <1 <1 <1 <1 1

<1 19 <1 <1 <1

<1 22 <1 <1

<1 28 <1 <1 <1

WHO Food Additives Series No. 71-S1, 2016

21

18

18

4 <1 <1

3 <1 <1

3 <1 <1

358

Supplement 1: Non-dioxin-like polychlorinated biphenyls

Mean concentration (ng/kg) Food code TN 0664 VA 0035 VB 0041 VB 0400 VB 0404 VB 0405 VC 0424 VC 0432 VD 0070 VD 0533 VD 0541 VL 0466 VL 0470 VL 0480 VL 0483 VL 0502 VO 0448 VO 0450 VP 0061 VR 0494 VR 0577 VR 0578 VR 0587 VR 0589 VR 0596 VS 0624 WD 0123 WF 0115 WF 0858 WF 0859 WS 0941 Food name Chestnuts Bulb vegetables, unprocessed, nes Cabbages, head Broccoli Cauliflower Kohlrabi Cucumber Watermelon Pulses, nes Lentil Soya bean Chinese cabbage, type pak-choi Corn salad (lamb’s lettuce) Kale Lettuce, leaf Spinach Tomato Mushrooms and fungi Beans, except broad bean and soya bean Radish Carrot Celeriac Parsley, turnip-rooted Potato Sugar beet Celery Trout Freshwater fish, nes Bream Carps Mackerel Lower bound Upper bound 1 2 2 336 28 162 0.03 267 0.02 267 0.04 267 0.03 267 0.06 0.5 0 0.8 0 0.8 0.03 400 0.02 267 0.11 0.6 0.26 267 0 800 0.02 200 0.01 400 0.07 400 0 400 0.03 0.05 0 0.04 0.03 0.02 0.05 0.8 3 2 51 4 267 267 800 0.6 267 0.1 0.5 3 97 2 52 337

Contribution to dietary exposure (%) Other children Adolescents Adults

<1 <1 <1 <1 <1 <1

<1 <1 <1 <1 <1 <1

<1 <1 <1 <1 <1 <1

<1

<1 <1

<1 <1

<1 <1

<1 <1 <1 <1 <1

<1 <1 <1 <1 <1

<1 <1 <1 <1 <1

<1 <1

<1 <1

<1 <1

excl.: excluding; incl.: including; nes: not elsewhere specified

359

Safety evaluation of certain food additives and contaminants Eightieth JECFA

Table A3.5 Concentration and dietary exposure data for Finland Mean concentration (ng/kg) Lower bound 988 10 927 26 319 23 329 96 129 18 91 29 45 191 428 47 314 43 815 24 839 15 436 6 921 7 471 14 233 23 799 5 980 Upper bound 990 10 927 26 319 118 329 96 129 18 91 29 45 191 428 47 317 43 817 24 839 15 436 6 921 7 471 14 233 23 799 5 980 Contribution to dietary exposure (%) Other Other children children Toddlers (DIPP) (STRIP) Adults 5 54 6 59 92 15 9 81 <1

Food code 01.6 13.1.1.1 16.8 ML 0106 MM 0095 MM 0095a MM 0819 MO 1280 MO 1281 MO 1284 MO 1285 MO 1289 PE 0112 PM 0841 WC 0143 WD 0123 WF 0115 WF 0858 WF 0864 WS 0130 WS 0932 WS 0937 WR 0140

WHO Food Additives Series No. 71-S1, 2016

Food name Cheese Ready-to-eat meal for infants and young children Fish-based composite food Other and nes milks Meat from mammals other than marine mammals, nes Game (mammalian) meat Rabbit meat Cattle, kidney Cattle, liver Pig kidney Pig liver Sheep liver Eggs, nes Duck meat Crustaceans, nes Trout Freshwater fish, nes Bream Perch Sardine and sardine-like fishes Flounders Herring Fish roe, nes

Elderly 18

<1

<1

<1

<1

<1

<1

<1

<1 <1 <1

<1

<1 <1

<1 <1

<1

<1

<1

<1

1 6

<1 1

3

3 9

<1 2

1 22

2 <1

2 33 1

70 <1

DIPP: Finnish Type I Diabetes Prediction and Prevention Nutrition Study; nes: not elsewhere specified; STRIP: Turku Coronary Risk Factor Intervention Project for Children

360

Supplement 1: Non-dioxin-like polychlorinated biphenyls

Table A3.6 Concentration and dietary exposure data for France Mean concentration (ng/kg) Lower Upper Food code Food name bound bound 02.2.1 Butter 1 716 1 716 13.1.1.1 Ready-to-eat meal for infants 0.1 0.1 and young children IM 0150 Molluscs, including cephalo28 28 pods, nes IM 1003 Mussels 28 28 IM 1004 Oysters (incl. cupped oysters) 439 439 IM 1005 Scallops 0.7 0.7 IM 1008 Squids 3 3 IM 5173 Octopuses 547 547 ML 0106 Other and nes milks 4 4 ML 0812 Cow milk 0.1 8 ML 0814 Goat milk 0 162 MM 0095 Meat from mammals other 0 660 than marine mammals, nes MM 0095a Game (mammalian) meat 1 1 MM 0812 Beef and other bovine meat 0.8 0.8 MM 0818 Pork and other porcines 1 1 MM 0819 Rabbit meat 0.2 141 MM 0822 Sheep and other ovines 0.7 0.7 MO 0105 Offal from mammals, nes 583 583 MO 1284 Pig kidney 0 660 MO 1289 Sheep liver 0.5 0.5 OR 0172 Vegetable oils, nes 0.2 0.2 OR 0305 Olive oil 0.6 0.6 OR 0495 Rape seed oil (incl. canola) 0.3 0.3 OR 0541 Soya bean oil 0.2 0.2 OR 0696 Palm oil 0.3 0.3 OR 0702 Sunflower seed oil 0.1 0.1 PE 0112 Eggs, nes 327 327 PE 0840 Chicken eggs 0.6 379 PE 0847 Quail eggs 0 1 578 PM 0110 Poultry meat, nes 0.2 128 PM 0840 Chicken meat 10 79 PM 0841 Duck meat 0.1 402 PM 0847 Quail meat 0 2 PM 0848 Turkey meat 18 105 WC 0979 Shrimps or prawns 6 6 WC 0143 Crustaceans, nes 38 38 WC 0146 Crabs 2 2 Contribution to exposure (%) Adolescents Adults Elderly 52 40 34

Other children 54 <1

Very elderly 28

<1 <1 <1 <1 <1 <1 <1 <1 <1 <1 <1 <1 <1

<1 <1 <1 <1 <1 <1 <1

<1 <1 <1 <1 <1 <1 <1

<1 <1 <1 <1

<1 <1

<1 <1 <1 <1 <1 <1 <1 <1 <1 <1 <1 <1 <1 <1 <1 <1 <1 <1 <1 <1

<1 <1 <1 <1 <1 1 <1 <1 <1 <1 <1 <1 <1 <1 <1 <1 <1 <1 <1 <1

<1 <1 <1 <1 <1 2 <1 <1 <1 <1 <1 <1 <1 <1 <1 <1 <1 <1 <1 <1 <1

<1 <1 <1 <1 <1 2 <1 <1 <1 <1

<1 <1 <1 <1 <1 2

<1 <1 <1 <1 <1 <1 <1 <1 <1 <1 <1 <1

<1 <1 <1 <1 <1 <1 <1 <1 <1

361

Safety evaluation of certain food additives and contaminants Eightieth JECFA

Table A3.6 (continued) Mean concentration (ng/kg) Lower Upper bound bound 1 948 1 948 196 1 261 66 66 6 673 7 416 137 137 0 15 333 0.4 0.4 6 6 3 235 3 235 15 334 5 346 8 17 840 2 506 5 800 358 1 068 15 334 5 346 8 17 840 2 506 5 800 358 1 068 Contribution to exposure (%) Adolescents Adults Elderly <1 <1 <1 <1 <1 <1 <1 <1 <1 <1 <1 <1 <1 <1 <1 <1 <1 <1 <1 <1

Food code WC 0978 WD 0123 WD 0890 WF 0115 WF 0858 WF 0859 WF 0864 WS 0125 WS 0130 WS 0920 WS 0927 WS 0937 WS 0941 WS 0948 WS 0945 WS 0952 WS 0951

Food name Lobsters Trout Eels Freshwater fish, nes Bream Carps Perch Marine fish, nes Sardine and sardine-like fishes Anchovies Cod Herring Mackerel Rays Plaice Tuna Sole

Other children <1

Very elderly <1

24 <1 15 <1 2 2

27 <1 10 2 2 2 <1

28 <1 24 1 <1 2 1

36 <1 22 2 1 2

30 <1 35 1 <1 2

incl.: including; nes: not elsewhere specified

Table A3.7 Concentration and dietary exposure data for Germany Mean concentration (ng/kg) Toddlers Upper (DONALD bound 2008) 255 <1 317 8 1 770 12 <1 61 515 5 3 719 120 650 7 726 9

WHO Food Additives Series No. 71-S1, 2016

Food code 01.0 01.2.1 01.4.1 01.6 02.1 02.2.1 02.2.2 13.6

Food name Dairy products, nes Fermented milks (plain) Cream Cheese Animal or vegetable fats, nes Butter Margarine Dietary supplements, food supplements

Lower bound 228 317 1 770 6 61 515 3 608 120 650 7 107

Contribution to exposure (%) Other children (DONALD Adoles2008) cents Adults Elderly <1 <1 <1 <1 12 6 6 6 <1 3 14 <1 21 20 <1 19 15 <1 12 16

Very elderly <1 5 <1 13 19

362

Supplement 1: Non-dioxin-like polychlorinated biphenyls

Mean concentration (ng/kg) Toddlers Upper (DONALD bound 2008) 13 118 18 3 4 309 665 10 <1

Food code 08.0

Food name Unprocessed meat and offal, nes 09.0 Fishes and aquatic animals, nes 11.5 Honey 13.1.1.1 Ready-to-eat meal for infants and young children 13.7 Food for infants and small children, nes 14.1.2 Fruit juice and herbal tea for infants and young children 14.2.3.1 Still grape wine 16.10 Composite food, nes 16.6 Meat-based composite food (incl. seafood) FB 0275 Strawberry FP 0226 Apple GC 0650 Rye GC 0654 Wheat JF 0175 Fruit juice, nes JF 0226 Apple juice JF 0269 Grape juice IM 0150 Molluscs, including cephalopods, nes IM 1003 Mussels IM 1005 Scallops ML 0106 Other and nes milks ML 0812 Cow milk ML 0814 Goat milk MM 0095 Meat from mammals other than marine mammals, nes MM 0095a Game (mammalian) meat MM 0812 Beef and other bovine meat MM 0814 Goat and other caprines MM 0816 Horse and other equines MM 0818 Pork and other porcines MM 0819 Rabbit meat MM 0822 Sheep and other ovines MO 1281 Cattle, liver MO 1285 Pig liver

Lower bound 13 112 2 0 253 221 0

Contribution to exposure (%) Other children (DONALD Adoles2008) cents Adults Elderly 29 12 12 11 <1 <1 <1 <1

Very elderly 6 <1

23 4

1 <1

0 0 5 150 25 105 30 46 0 0 0 0.05 3 0.04 337 23 142 49 0.2 216 0 0 794 0 174 916 0

10 2 5 150 25 105 30 46 10 10 10 3 3 0.1 337 26 151 112 186 288 0.7 1 795 0.9 328 1 015 160

10 <1 2 <1

11 <1 4 <1 <1

16 <1 4 <1

9 <1 3 <1 <1

5 <1 4 <1

5 <1 4 <1

<1 <1 <1 1 2

<1 <1 <1 1 <1 <1

<1 <1 <1 1 <1 <1

9 4 <1

5 3 <1 <1

1 <1 <1

<1

<1

<1 1

<1 <1

<1 <1

<1 <1

<1

3 <1

9 <1

7 <1 <1

5 <1 <1

4 <1 <1

363

Safety evaluation of certain food additives and contaminants Eightieth JECFA

Table A3.7 (continued) Mean concentration (ng/kg)

Food code MO 0105 MO 1289 PE 0112 PE 0840 PE 0847 PM 0110 PM 0840 PM 0841 PM 0848 PO 0111 VB 0041 VB 0404 VB 0405 VC 0431 VL 0053

Food name Offal from mammals, nes Sheep liver Eggs, nes Chicken eggs Quail eggs Poultry meat, nes Chicken meat Duck meat Turkey meat Offal from poultry, nes Cabbages, head Cauliflower Kohlrabi Squash Leafy vegetables, unprocessed, nes VP 0061 Beans, except broad bean and soya bean VR 0589 Potato VR 0577 Carrot WC 0143 Crustaceans, nes WC 0979 Shrimps or prawns WD 0123 Trout WD 0890 Eels WF 0115 Freshwater fish, nes WF 0858 Bream WF 0859 Carps WF 0864 Perch WS 0130 Sardine and sardine-like fishes WS 0952 Tuna WS 0927 Cod WS 0932 Flounders WS 0941 Mackerel WS 0945 Plaice WS 0937 Herring WS 0951 Sole WR 0140a Offal of fish, nes

Lower bound 0.7 4 0.2 1 226 130 1 5 38 13 0 28 45 24 219 125 55 19 56 0 0.003 1 435 22 585 777 61 669 8 31 000 2 420 282 800 84 907 0 12 548 7 056 258 416

Contribution to exposure (%) Other Toddlers children Upper (DONALD (DONALD Adolesbound 2008) 2008) cents Adults Elderly 2 <1 <1 5 <1 2 1 415 7 11 3 4 3 287 <1 3 <1 <1 <1 27 <1 <1 <1 <1 <1 115 <1 <1 <1 35 <1 <1 <1 <1 <1 4 28 <1 <1 <1 <1 <1 45 <1 <1 <1 <1 <1 24 <1 <1 <1 <1 <1 219 125 <1 <1 1 1 <1 55 19 56 1 541 8 2 878 22 965 1 380 65 2 966 93 37 667 <1 <1 <1 <1 <1

Very elderly <1

3 <1 <1 <1 <1 <1 <1 <1 <1 <1

<1

<1 <1

<1 <1

<1 <1 <1 <1 <1 <1

<1 <1 1

<1 <1 1

WHO Food Additives Series No. 71-S1, 2016

<1

<1

<1

<1 <1

<1 <1

3 176 1 018 977 93 646 1 008 15 275 9 989 258 416

<1

<1 <1

1 <1 2

1 <1 8 12 <1 <1

<1 <1 14 18 <1

<1 <1 8 23 1 3

1

3

incl.: including; nes: not elsewhere specified

364

Supplement 1: Non-dioxin-like polychlorinated biphenyls

Table A3.8 Concentration and dietary exposure data for Grecce Mean concentration (ng/kg) Lower bound Upper bound 0.04 0.038 36 38 8 861 8 861 88 88 3 967 22 839 602 2 532 0.2 0.2 0.5 2 552 1 230 1 210 27 38 0.2 136 31 259 0.2 723 0.4 649 259 274 3 221 0.3 89 2 923 20 365 4 445 2 568 1 225 0.5 5 803 1 230 1 611 28 38 0.2 136 31 260 0.2 723 0.4 649 259 296 4 073 0.3 89 2 923 20 558 4 445 2 568 1 225 Contribution to exposure (%) Other children <1

Food code 01.0 01.1.2 01.4.1 01.6 02.1 08.0 13.1.1.1 13.2b IM 0150 IM 1003 ML 0106 ML 0812 ML 0814 ML 0822 MM 0812 MM 0818 MM 0822 MO 0105 MO 1281 OR 0172 OR 0305 OR 0645 OR 0702 PE 0112 PE 0840 PM 0840 WD 0123 WD 0890 WF 0115 WF 0858 WS 0927 WS 0941

Food name Dairy products, nes Flavoured milk Cream Cheese Animal or vegetable fats, nes Unprocessed meat and offal, nes Ready-to-eat meal for infants and young children Cereal-based food for infants and young children Molluscs, including cephalopods, nes Mussels Other and nes milks Cow milk Goat milk Sheep milk Beef and other bovine meat Pork and other porcines Sheep and other ovines Offal from mammals, nes Cattle, liver Vegetable oils, nes Olive oil Maize oil Sunflower seed oil Eggs, nes Chicken eggs Chicken meat Trout Eels Freshwater fish, nes Bream Cod Mackerel

<1 <1 <1 <1

79 1 <1 <1 <1 <1 <1 <1 <1 2 <1 10 <1 <1

6 <1

nes: not elsewhere specified

365

Safety evaluation of certain food additives and contaminants Eightieth JECFA

Table A3.9 Concentration and dietary exposure data for Ireland Mean concentration (ng/kg) Lower bound Upper bound 29 29 29 29 284 284 880 880 1 131 1 131 0 1 000 23 820 24 436 70 70 0 200 0 200 0 200 103 103 1 008 1 008 759 1 176 19 19 33 33 121 121 248 248 22 22 158 158 257 1 090 1 909 1 909 713 752 1 387 1 387 917 917 46 46 154 154 0 200 0 200 0 200 0 200 0 200 0 200 0 200 0 200 9 669 9 669 6 012 6 012 2 626 2 626 6 479 6 479 3 970 3 988 Contribution to exposure (%) Other children <1 <1 8 13 <1 32 <1

Food code 01.0 01.2.1 01.6 02.2.1 02.1 12.1.1 13.6 16.10 FB 0272 FB 0275 FP 0226 GC 0080 IM 1004 MF 0100 ML 0812 ML 0814 MO 0105 MO 1281 MO 1285 MO 1289 OR 0172 PE 0112 PE 0840 PF 0111 PF 0840 PO 0840a PO 0848 VA 0387 VB 0041 VL 0053 VL 0502 VO 0448 VO 0450 VR 0577 VR 0589 WD 0123 WS 0937 WS 0130 WS 0941 WS 0952

Food name Dairy products, nes Fermented milks (plain) Cheese Butter Animal or vegetable fats, nes Salt Dietary supplements, food supplements Composite food, nes Raspberries, red, black Strawberry Apple Cereals grains, nes Oysters (incl. cupped oysters) Mammalian fats (except milk fats), and skin, nes Cow milk Goat milk Offal from mammals, nes Cattle, liver Pig liver Sheep liver Vegetable oils, nes Eggs, nes Chicken eggs Poultry fat and skin, nes Chicken fat Chicken liver only Turkey, offal of Onion, Welsh Cabbages, head Leafy vegetables, unprocessed, nes Spinach Tomato Mushrooms and fungi Carrot Potato Trout Herring Sardine and sardine-like fishes Mackerel Tuna

<1 <1 8 <1 <1 <1 <1 <1 2 14

WHO Food Additives Series No. 71-S1, 2016

<1

<1 32

incl.: including; nes: not elsewhere specified

366

Supplement 1: Non-dioxin-like polychlorinated biphenyls

Table A3.10 Concentration and dietary exposure data for Italy Mean concentration (ng/kg) Lower Upper bound bound 0 12 793 0 0 0 0 12 500 104 58 118

Food code 02.1 11.5 ML 0106 ML 0812 MM 0095

MM 0812 MM 0818 MM 0819 MM 0822 MO 0105 MO 1280 MO 1285 PE 0112 PM 0840 WD 0123 WD 0890 WF 0115 WF 0858

Food name Animal or vegetable fats, nes Honey Other and nes milks Cow milk Meat from mammals other than marine mammals, nes Beef and other bovine meat Pork and other porcines Rabbit meat Sheep and other ovines Offal from mammals, nes Cattle, kidney Pig liver Eggs, nes Chicken meat Trout Eels Freshwater fish, nes Bream

Contribution to exposure (%) Other AdolesInfants Toddlers children cents Adults Elderly

Very elderly

15 0 0 0 0 0 0 28 0 1 905 1 845 9 637 6 553

225 290 106 375 10 33 74 335 120 2 375 5 345 9 978 10 124

100

100

4

5

3

4

8

<1 96

<1 95

1 96

1 95 92

nes: not elsewhere specified

Table A3.11 Concentration and dietary exposure data for the Netherlands Mean concentration (ng/kg) Food code 01.6 02.2.1 02.2.2 15.4 IM 0150 Food name Cheese Butter Margarine Snacks, nes Molluscs, including cephalopods, nes Lower bound Upper bound 1 149 1 169 3 945 3 945 1 860 2 047 2 760 2 760 11 900 11 900 Contribution to exposure (%) Other Toddlers children Adults 49 39 23 2 3 5 9 9 2

367

Safety evaluation of certain food additives and contaminants Eightieth JECFA

Table A3.11 (continued) Mean concentration (ng/kg) Food code ML 0106 MM 0812 MM 0818 OR 0172 PE 0112 PM 0840 WC 0143 WC 0979 WD 0123 WD 0890 WF 0115 WF 0864 WR 0140a WS 0125 WS 0130 WS 0920 WS 0927 WS 0937 WS 0941 WS 0952 Food name Other and nes milks Beef and other bovine meat Pork and other porcines Vegetable oils, nes Eggs, nes Chicken meat Crustaceans, nes Shrimps or prawns Trout Eels Freshwater fish, nes Perch Offal of fish, nes Marine fish, nes Sardine and sardine-like fishes Anchovies Cod Herring Mackerel Tuna Lower bound Upper bound 112 124 929 929 570 570 1 030 1 330 1 397 1 397 1 398 1 398 7 502 7 502 1 475 1 475 13 133 13 133 247 901 247 901 28 864 28 864 50 460 50 460 267 173 267 173 33 100 33 100 32 350 32 350 63 900 63 900 25 25 15 500 15 500 10 000 10 000 23 867 23 867 Contribution to exposure (%) Other Toddlers children Adults 7 10 2 9 8 8 3 3 11 <1 <1 <1 11 12 17 2 <1 <1 <1 5 14

2

2

<1 5 3

<1 2 2 6

<1 2 1 10

nes: not elsewhere specified

Table A3.12 Concentration and dietary exposure data for Republic of Korea Food code 09.0 IM 0150 IM 1000 IM 1003 IM 1004 IM 1005 IM 1008 IM 5173 WC 0146 WC 0978 WC 0979 WD 0123 WD 0890 Food name Fishes and aquatic animals, nes Molluscs, including cephalopods, nes Clams Mussels Oysters (incl. cupped oysters) Scallops Squids Octopuses Crabs Lobsters Shrimps or prawns Trout Eels Mean concentration (ng/kg) Lower bound Upper bound 1 259 1 480 1 211 1 438 1 026 1 266 1 214 1 441 1 584 1 792 831 1 087 1 436 1 641 1 982 2 104 1 474 1 640 0 364 978 1 235 927 1 000 806 948

WHO Food Additives Series No. 71-S1, 2016

Contribution to exposure (%) Children General population

5

40 7 7

368

Supplement 1: Non-dioxin-like polychlorinated biphenyls

Food code WD 4917 WF 0115 WF 0858 WM 0141 WS 0125 WS 0927 WS 0932 WS 0937 WS 0941 WS 0945 WS 0946 WS 0951 WS 0952 WS 5015

Food name Salmon, Pacific Freshwater fish, nes Bream Marine mammals, nes Marine fish, nes Cod Flounders Herring Mackerel Plaice Pollack Sole Tuna Shark

Mean concentration (ng/kg) Lower bound Upper bound 1 644 1 720 1 084 1 202 386 586 3 408 3 450 878 999 513 670 558 711 2 184 2 197 1 311 1 372 936 1 033 367 605 648 721 747 906 642 779

Contribution to exposure (%) Children General population

100

24 7 9

incl.: including; nes: not elsewhere specified

Table A3.13 Concentration and dietary exposure data for Sweden Mean concentration (ng/kg) Food code 01.0 01.6 02.1 02.2.1 06.3b 08.0 13.1.1.1 13.6 16.8 CP 0179 FA 0142 FB 0275 FP 0226 GC 0650 GC 0654 ML 0106 ML 0812 MM 0095 Food name Dairy products, nes Cheese Animal or vegetable fats, nes Butter Porridge Unprocessed meat and offal, nes Ready-to-eat meal for infants and young children Dietary supplements, food supplements Fish-based composite food Hominy/mugunzá Marine animal fats, nes Strawberry Apple Rye Wheat Other and nes milks Cow milk Meat from mammals other than marine mammals, nes Lower bound Upper bound 3 3 0.3 0.3 789 928 520 645 20 22 35 35 73 73 2 736 58 080 0 9 904 5 12 0.03 0.02 39 172 67 2 737 58 080 800 9 988 5 12 0.03 0.02 39 694 67 Contribution to exposure (%)a Other children Adolescents Adults <1 <1 <1 <1 <1 <1 <1 <1 <1 <1 <1 <1 <1

1 43

<1 43 72

<1 <1 <1 1 34

<1 <1

<1 <1

1 39

<1 11

369

Safety evaluation of certain food additives and contaminants Eightieth JECFA

Table A3.13 (continued) Mean concentration (ng/kg) Food code MM 0818 MO 0105 MO 1281 MO 1289 OR 0172 OR 0665 OR 0696 PE 0112 PE 0840 PF 0111 PF 0840 PO 0840a VB 0041 VB 0402 VB 0404 VL 0482 VL 0496 VL 0502 VR 0577 VR 0589 WD 0123 WD 0890 WF 0115 WF 0864 WR 0140 WS 0130 WS 0927 WS 0932 WS 0937 WS 0941 Food name Pork and other porcines Offal from mammals, nes Cattle, liver Sheep liver Vegetable oils, nes Coconut oil Palm oil Eggs, nes Chicken eggs Poultry fat and skin, nes Chicken fat Chicken liver only Cabbages, head Brussels sprouts Cauliflower Lettuce, head Rucola Spinach Carrot Potato Trout Eels Freshwater fish, nes Perch Fish roe, nes Sardine and sardine-like fishes Cod Flounders Herring Mackerel Lower bound Upper bound 0 405 28 28 79 79 2 2 0.3 0.3 0 2 160 0 2 203 1 402 2 076 776 789 1 2 981 981 69 69 8 8 0.01 0.01 8 8 0.02 0.02 0.1 0.1 0.04 0.04 9 9 0.01 0.01 44 457 44 468 15 561 15 587 10 719 10 730 11 169 11 169 19 286 19 286 13 564 13 564 2 589 2 622 7 7 21 337 21 353 13 732 13 732 Contribution to exposure (%)a Other children Adolescents Adults <1

<1

<1

<1

3 <1

3 <1

3

<1 <1 <1

<1 <1 <1

<1 <1 <1

<1 <1 <1

<1 <1 <1

<1 <1 <1 <1

<1 9 4 2 2

<1 5 4 1 3

<1 8 <1 <1 5 1

WHO Food Additives Series No. 71-S1, 2016 370

nes: not elsewhere specified a Other children –National Food Agency survey, n = 1473; Adolescents – National Food Agency survey, n = 1018; Adults – survey Riksmaten (National Dietary Survey) 1997–1998, n = 1210.

Supplement 1: Non-dioxin-like polychlorinated biphenyls

Table A3.14 Concentration and dietary exposure data for the United Kingdom Mean concentration (ng/kg) Lower bound Upper bound 59 68 418 510 4 122 4 361 12 51 11 24 203 240 2 429 2 443 950 950 63 258 63 705 2 528 2 560 1 1 45 95 1 135 2 766 2 2 2 2 750 800 180 232 237 792 2 041 2 057 52 68 56 112 1 1 6 6 66 66 4 224 4 557 25 25 54 54 12 12 0.2 0.2 9 9 8 8 4 4 Contribution to exposure (%)a Adults <1 8 <1 <1 <1 7 2 50

Food code 01.0 01.6 02.1 04.1.1 04.2.2 08.0 09.0 11.1.3 13.6 16.8 CP 0179 GC 0080 IM 0150 IM 1003 IM 1004 MM 0095a MO 0105 OR 0172 PE 0840 PM 0110 TN 0085 VR 0589 WD 0123 WD 0890 WF 0115 WF 0858 WF 0859 WF 0864 WS 0927 WS 0932 WS 0937 WS 0941

Food name Dairy products, nes Cheese Animal or vegetable fats, nes Fruits Other vegetables, nes, other processing Unprocessed meat and offal, nes Fishes and aquatic animals, nes Sugar products and confectionaries, nes Dietary supplements, food supplements Fish-based composite food Hominy/mugunzá Cereals grains, nes Molluscs, including cephalopods, nes Mussels Oysters (incl. cupped oysters) Game (mammalian) meat Offal from mammals, nes Vegetable oils, nes Chicken eggs Poultry meat, nes Tree nuts, nes Potato Trout Eels Freshwater fish, nes Bream Carps Perch Cod Flounders Herring Mackerel

<1 <1 <1 <1 <1 <1 <1 32 <1 <1 <1 <1

<1 <1 <1

incl.: including; nes: not elsewhere specified a Number of subjects = 1724. National Diet and Nutrition Survey.

371

Appendix 4. Details of the national dietary exposure for the six indicator PCB congeners individually as estimated by the Committee using the CIFOCOss consumption data and concentration data from the GEMS/Food database Table A4.1 National dietary exposures to PCB 28 Dietary exposure (ng/kg bw per day) Mean High Lower bound Upper bound Lower bound Upper bound 0.2 5.7 0.4 11.3 0.2 4.5 0.4 9.0 0.2 4.3 0.4 8.6 0.2 14.0 0.4 28.0 0.3 17.5 0.7 35.1 0.2 4.7 0.4 9.5 0.3 0.3 0.5 0.5 0.1 0.1 0.3 0.3 0.5 2.8 1.1 5.7 0.1 0.6 0.2 1.1 0.1 0.3 0.1 0.6 0.1 0.8 0.2 1.7 0.3 0.3 0.6 0.6 0.2 0.2 0.4 0.4 0.5 0.5 1.0 1.0 0.7 0.7 1.3 1.3 1.0 1.0 2.1 2.1 0.02 0.1 0.04 0.12 0.02 0.1 0.05 0.11 0.03 0.1 0.05 0.11 0.03 0.1 0.07 0.21 0.02 0.05 0.05 0.1 0.2 0.2 0.4 0.5 0.3 0.3 0.5 0.6 0.3 0.4 0.5 0.7 0.8 0.9 1.7 1.9 0.8 0.9 1.5 1.7 0.8 0.8 1.6 1.7 1.1 1.4 2.2 2.9 1.1 1.4 2.2 2.8 1.2 1.7 2.5 3.4 0.2 0.3 0.5 0.7 0.4 1.0 0.8 2.1

Country Belgium

  China   Cyprus Czech Republic   Finland

  France

  Germany

  Greece

Population group (survey) Adolescents Adults Elderly Other children Toddlers Very elderly Children General population Adolescents Adolescents Adults Other children Adults Elderly Other children (DIPP) Other children (STRIP) Toddlers Adolescents Adults Elderly Other children Very elderly Adolescents Adults Elderly Other children (DONALD 2006) Other children (DONALD 2007) Other children (DONALD 2008) Toddlers (DONALD 2006) Toddlers (DONALD 2007) Toddlers (DONALD 2008) Very elderly Other children

372

Supplement 1: Non-dioxin-like polychlorinated biphenyls

Country Ireland Italy

  Japan   Netherlands   Republic of Korea  Sweden

United Kingdom

Population group (survey) Adults Adolescents Adults Elderly Infants Other children Toddlers Very elderly Children General population Adults Other children Toddlers Children General population Adolescents Adults Other children Adults

Dietary exposure (ng/kg bw per day) Mean High Lower bound Upper bound Lower bound Upper bound 0.04 0.1 0.1 0.2 0.02 0.1 0.03 0.3 0.01 0.1 0.03 0.3 0.01 0.1 0.02 0.2 0.000  2 1.2 0.000  4 2.5 0.03 0.4 0.1 0.7 0.001 0.6 0.002 1.2 0.005 0.1 0.01 0.2 0.03 0.03 0.1 0.1 0.02 0.02 0.04 0.04 0.1 0.1 0.2 0.2 0.1 0.1 0.3 0.3 0.1 0.2 0.3 0.3 0.03 0.03 0.1 0.1 0.1 0.1 0.2 0.2 0.1 1.2 0.1 2.5 0.1 0.6 0.2 1.2 0.1 2.1 0.2 4.3 0.01 0.02 0.02 0.05

DIPP: Finnish Type I Diabetes Prediction and Prevention Nutrition Study; STRIP: Turku Coronary Risk Factor Intervention Project for Children

Table A4.2 National dietary exposures to PCB 52 Dietary exposure (ng/kg bw per day) Mean High Lower bound Upper bound Lower bound Upper bound 1.5 5.6 3.1 11.3 1.2 4.5 2.3 8.9 1.0 4.2 2.0 8.4 1.6 13.7 3.3 27.4 2.4 17.2 4.8 34.3 1.1 4.6 2.1 9.2 0.1 0.1 0.2 0.2 0.1 0.1 0.1 0.1 0.3 2.6 0.7 5.3 0.04 0.5 0.1 1.1 0.02 0.3 0.05 0.5 0.04 0.8 0.1 1.6 0.04 0.04 0.1 0.1 0.1 0.1 0.1 0.1

Country Belgium

  China   Cyprus Czech Republic   Finland

Population group (survey) Adolescents Adults Elderly Other children Toddlers Very elderly Children General population Adolescents Adolescents Adults Other children Adults Elderly

373

Safety evaluation of certain food additives and contaminants Eightieth JECFA

Table A4.2 (continued) Dietary exposure (ng/kg bw per day) Mean High Lower bound Upper bound Lower bound Upper bound 0.1 0.1 0.2 0.2 0.4 0.4 0.8 0.8 0.9 0.9 1.9 1.9 0.02 0.1 0.04 0.1 0.02 0.1 0.05 0.1 0.03 0.1 0.07 0.1 0.03 0.1 0.07 0.2 0.03 0.1 0.07 0.1 0.1 0.3 0.3 0.5 0.2 0.3 0.4 0.6 0.2 0.4 0.5 0.8 0.5 0.9 1.0 1.9 0.7 1.2 1.4 2.4 0.4 0.8 0.8 1.7 1.3 2.2 2.5 4.3 1.0 2.0 2.0 4.1 1.2 2.3 2.5 4.5 0.2 0.4 0.4 0.7 1.5 2.0 3.1 4.1 0.04 0.1 0.1 0.2 0.04 0.2 0.1 0.3 0.03 0.1 0.1 0.3 0.02 0.1 0.05 0.3 0.001 1.2 0.001 2.5 0.1 0.4 0.1 0.8 0.004 0.6 0.01 1.2 0.01 0.1 0.02 0.2 0.1 0.1 0.2 0.2 0.1 0.1 0.2 0.3 0.1 0.1 0.2 0.2 0.02 0.03 0.05 0.1 0.1 0.1 0.1 0.2 0.1 1.0 0.2 1.9 0.2 0.5 0.3 1.1 0.2 1.7 0.3 3.3 0.05 0.1 0.1 0.1

Country

  France

  Germany

  Greece Ireland Italy

WHO Food Additives Series No. 71-S1, 2016

  Netherlands   Republic of Korea  Sweden   United Kingdom

Population group (survey) Other children (DIPP) Other children (STRIP) Toddlers Adolescents Adults Elderly Other children Very elderly Adolescents Adults Elderly Other children (DONALD 2006) Other children (DONALD 2007) Other children (DONALD 2008) Toddlers (DONALD 2006) Toddlers (DONALD 2007) Toddlers (DONALD 2008) Very elderly Other children Adults Adolescents Adults Elderly Infants Other children Toddlers Very elderly Adults Other children Toddlers Children General population Adolescents Adults Other children Adults

DIPP: Finnish Type I Diabetes Prediction and Prevention Nutrition Study; STRIP: Turku Coronary Risk Factor Intervention Project for Children

374

Supplement 1: Non-dioxin-like polychlorinated biphenyls

Table A4.3 National dietary exposures to PCB 101 Dietary exposure (ng/kg bw per day) Mean High Lower bound Upper bound Lower bound Upper bound 0.8 5.4 1.5 10.8 0.7 4.4 1.3 8.9 0.7 4.4 1.3 8.7 1.2 13.8 2.4 27.6 1.6 17.2 3.1 34.4 0.6 4.8 1.3 9.5 0.1 0.1 0.3 0.3 0.1 0.1 0.1 0.1 0.3 2.6 0.6 5.2 0.1 0.6 0.2 1.1 0.1 0.3 0.1 0.6 0.1 0.8 0.2 1.6 0.05 0.05 0.1 0.1 0.2 0.2 0.4 0.4 0.1 0.1 0.2 0.2 1.3 1.3 2.5 2.5 1.7 1.7 3.4 3.4 0.04 0.1 0.08 0.2 0.1 0.1 0.1 0.2 0.1 0.1 0.2 0.2 0.1 0.2 0.2 0.3 0.1 0.1 0.2 0.3 0.2 0.2 0.4 0.5 0.3 0.4 0.7 0.8 0.4 0.5 0.9 1.0 0.6 0.7 1.2 1.5 1.0 1.1 2.0 2.2 0.6 0.7 1.2 1.3 0.8 1.6 1.7 3.2 0.9 1.5 1.7 2.9 1.0 1.7 2.0 3.5 0.4 0.4 0.8 0.9 3.0 3.4 6.0 6.7 0.1 0.2 0.2 0.3 0.05 0.2 0.1 0.4 0.04 0.2 0.1 0.3 0.03 0.1 0.1 0.3 0.000  3 1.2 0.001 2.5 0.1 0.4 0.2 0.9 0.002 0.6 0.004 1.2

Country Belgium

  China   Cyprus Czech Republic   Finland

  France

  Germany

  Greece Ireland Italy

Population group (survey) Adolescents Adults Elderly Other children Toddlers Very elderly Children General population Adolescents Adolescents Adults Other children Adults Elderly Other children (DIPP) Other children (STRIP) Toddlers Adolescents Adults Elderly Other children Very elderly Adolescents Adults Elderly Other children (DONALD 2006) Other children (DONALD 2007) Other children (DONALD 2008) Toddlers (DONALD 2006) Toddlers (DONALD 2007) Toddlers (DONALD 2008) Very elderly Other children Adults Adolescents Adults Elderly Infants Other children Toddlers

375

Safety evaluation of certain food additives and contaminants Eightieth JECFA

Table A4.3 (continued) Dietary exposure (ng/kg bw per day) Mean High Lower bound Upper bound Lower bound Upper bound 0.01 0.1 0.03 0.2 0.2 0.2 0.3 0.3 0.1 0.1 0.2 0.2 0.1 0.2 0.3 0.3 0.1 0.2 0.2 0.3 0.1 0.1 0.2 0.3 0.03 0.03 0.1 0.1 0.1 0.1 0.2 0.2 0.3 1.7 0.6 3.4 0.6 1.2 1.3 2.5 0.6 3.0 1.2 6.0 0.1 0.1 0.1 0.2

Country   Japan   Netherlands   Republic of Korea  Sweden   United Kingdom

Population group (survey) Very elderly Children General population Adults Other children Toddlers Children General population Adolescents Adults Other children Adults

DIPP: Finnish Type I Diabetes Prediction and Prevention Nutrition Study; STRIP: Turku Coronary Risk Factor Intervention Project for Children

Table A4.4 National dietary exposures to PCB 138 Dietary exposure (ng/kg bw per day) Mean High Lower bound Upper bound Lower bound Upper bound 2.6 5.4 5.2 10.8 1.7 4.4 3.4 8.8 1.3 4.2 2.6 8.5 3.2 14.2 6.3 28.3 4.1 17.5 8.1 35.0 1.5 4.7 2.9 9.3 0.2 0.2 0.4 0.4 0.1 0.1 0.2 0.2 0.8 3.1 1.6 6.2 0.1 0.6 0.2 1.1 0.05 0.3 0.1 0.6 0.1 0.9 0.2 1.7 0.2 0.2 0.3 0.3 0.5 0.5 0.9 0.9 0.3 0.3 0.6 0.6 2.4 2.4 4.7 4.7 2.5 2.5 4.9 5.0 0.2 0.2 0.4 0.5 0.2 0.2 0.4 0.5

Country Belgium

WHO Food Additives Series No. 71-S1, 2016

  China   Cyprus Czech Republic   Finland

  France

Population group (survey) Adolescents Adults Elderly Other children Toddlers Very elderly Children General population Adolescents Adolescents Adults Other children Adults Elderly Other children (DIPP) Other children (STRIP) Toddlers Adolescents Adults

376

Supplement 1: Non-dioxin-like polychlorinated biphenyls

Country

  Germany

  Greece Ireland Italy

  Japan   Netherlands   Republic of Korea  Sweden   United Kingdom

Population group (survey) Elderly Other children Very elderly Adolescents Adults Elderly Other children (DONALD 2006) Other children (DONALD 2007) Other children (DONALD 2008) Toddlers (DONALD 2006) Toddlers (DONALD 2007) Toddlers (DONALD 2008) Very elderly Other children Adults Adolescents Adults Elderly Infants Other children Toddlers Very elderly Children General population Adults Other children Toddlers Children General population Adolescents Adults Other children Adults

Dietary exposure (ng/kg bw per day) Mean High Lower bound Upper bound Lower bound Upper bound 0.3 0.3 0.5 0.6 0.4 0.5 0.8 0.9 0.3 0.3 0.5 0.6 0.7 0.7 1.4 1.4 0.9 0.9 1.7 1.8 1.0 1.0 2.0 2.0 1.7 1.8 3.5 3.5 1.9 2.0 3.8 3.9 1.5 1.6 3.1 3.1 2.1 2.1 4.2 4.2 2.0 2.0 3.9 4.0 2.1 2.2 4.3 4.3 1.1 1.1 2.1 2.2 4.3 4.7 8.6 9.4 0.2 0.3 0.4 0.5 0.1 0.2 0.2 0.5 0.1 0.2 0.2 0.4 0.1 0.2 0.1 0.4 0.001 1.2 0.001 2.5 0.2 0.5 0.4 1.0 0.003 0.6 0.01 1.2 0.03 0.1 0.1 0.3 0.1 0.1 0.3 0.3 0.1 0.1 0.2 0.2 0.8 0.8 1.5 1.5 1.1 1.1 2.1 2.1 1.1 1.1 2.3 2.3 0.03 0.04 0.1 0.1 0.1 0.1 0.2 0.2 1.1 1.3 2.2 2.7 1.7 1.8 3.5 3.7 2.2 2.5 4.3 5.1 0.3 0.3 0.7 0.7

DIPP: Finnish Type I Diabetes Prediction and Prevention Nutrition Study; STRIP: Turku Coronary Risk Factor Intervention Project for Children

377

Safety evaluation of certain food additives and contaminants Eightieth JECFA

Table A4.5 National dietary exposures to PCB 153 Dietary exposure (ng/kg bw per day) Mean High Lower bound Upper bound Lower bound Upper bound 4.6 6.2 9.2 12.4 3.0 4.8 6.0 9.6 2.3 4.4 4.7 8.9 4.9 12.1 9.9 24.2 6.5 15.7 13.1 31.5 2.6 5.0 5.2 9.9 0.2 0.2 0.4 0.4 0.1 0.1 0.2 0.2 1.2 3.5 2.4 7.0 0.2 0.7 0.4 1.4 0.1 0.4 0.2 0.7 0.3 1.0 0.5 2.1 0.2 0.2 0.4 0.4 0.6 0.6 1.2 1.2 0.4 0.4 0.7 0.7 3.5 3.5 7.0 7.1 3.4 3.5 6.9 6.9 0.3 0.3 0.6 0.7 0.3 0.4 0.7 0.7 0.4 0.5 0.9 0.9 0.6 0.6 1.2 1.3 0.5 0.5 0.9 1.0 0.9 1.0 1.8 1.9 1.1 1.1 2.2 2.3 1.2 1.2 2.4 2.4 2.2 2.4 4.4 4.7 2.4 2.6 4.7 5.1 1.9 2.1 3.8 4.2 2.8 3.1 5.6 6.1 2.6 3.0 5.3 6.0 2.8 3.1 5.6 6.3 1.3 1.3 2.6 2.7 0.5 1.2 1.1 2.3 0.3 0.3 0.5 0.7 0.1 0.2 0.2 0.5 0.1 0.2 0.2 0.4 0.1 0.2 0.1 0.4 0.001 1.2 0.002 2.5 0.2 0.5 0.4 1.0 0.01 0.6 0.01 1.2

Country Belgium

  China   Cyprus Czech Republic   Finland

  France

  Germany

  Greece Ireland Italy

Population group (survey) Adolescents Adults Elderly Other children Toddlers Very elderly Children General population Adolescents Adolescents Adults Other children Adults Elderly Other children (DIPP) Other children (STRIP) Toddlers Adolescents Adults Elderly Other children Very elderly Adolescents Adults Elderly Other children (DONALD 2006) Other children (DONALD 2007) Other children (DONALD 2008) Toddlers (DONALD 2006) Toddlers (DONALD 2007) Toddlers (DONALD 2008) Very elderly Other children Adults Adolescents Adults Elderly Infants Other children Toddlers

378

WHO Food Additives Series No. 71-S1, 2016

Supplement 1: Non-dioxin-like polychlorinated biphenyls

Country   Japan   Netherlands   Republic of Korea  Sweden   United Kingdom

Population group (survey) Very elderly Children General population Adults Other children Toddlers Children General population Adolescents Adults Other children Adults

Dietary exposure (ng/kg bw per day) Mean High Lower bound Upper bound Lower bound Upper bound 0.03 0.1 0.1 0.3 0.3 0.3 0.5 0.5 0.2 0.2 0.3 0.3 1.0 1.0 2.1 2.1 1.4 1.4 2.8 2.8 1.5 1.5 3.1 3.1 0.04 0.04 0.1 0.1 0.1 0.1 0.2 0.2 1.4 1.7 2.8 3.5 2.0 2.2 4.0 4.3 2.7 3.3 5.4 6.6 0.4 0.4 0.8 0.8

DIPP: Finnish Type I Diabetes Prediction and Prevention Nutrition Study; STRIP: Turku Coronary Risk Factor Intervention Project for Children

Table A4.6 National dietary exposures to PCB 180 Dietary exposure (ng/kg bw per day) Mean High Lower bound Upper bound Lower bound Upper bound 1.0 5.5 2.1 11.1 0.7 4.5 1.5 9.0 0.6 4.3 1.2 8.6 1.6 13.8 3.2 27.6 2.0 17.3 4.1 34.5 0.6 4.7 1.3 9.4 0.1 0.1 0.2 0.2 0.04 0.04 0.1 0.1 0.4 2.7 0.8 5.4 0.1 0.6 0.2 1.1 0.05 0.3 0.1 0.6 0.1 0.9 0.2 1.7 0.1 0.1 0.1 0.1 0.2 0.2 0.5 0.5 0.1 0.1 0.2 0.2 1.3 1.3 2.6 2.6 1.0 1.0 2.0 2.1 0.1 0.1 0.2 0.3 0.1 0.1 0.2 0.3 0.1 0.2 0.3 0.3 0.2 0.3 0.4 0.5

Country Belgium

  China   Cyprus Czech Republic   Finland

  France

Population group (survey) Adolescents Adults Elderly Other children Toddlers Very elderly Children General population Adolescents Adolescents Adults Other children Adults Elderly Other children (DIPP) Other children (STRIP) Toddlers Adolescents Adults Elderly Other children

379

Safety evaluation of certain food additives and contaminants Eightieth JECFA

Table A4.6 (continued) Dietary exposure (ng/kg bw per day) Mean High Lower bound Upper bound Lower bound Upper bound 0.1 0.2 0.3 0.3 0.5 0.6 0.9 1.1 0.6 0.7 1.1 1.3 0.6 0.7 1.3 1.5 1.1 1.5 2.2 3.0 1.5 2.0 3.1 3.9 1.0 1.4 2.0 2.7 1.3 2.0 2.7 4.0 1.2 2.0 2.4 4.0 1.3 2.1 2.7 4.1 0.6 0.7 1.2 1.4 0.7 1.2 1.5 2.5 0.1 0.2 0.2 0.4 0.04 0.2 0.1 0.3 0.04 0.1 0.1 0.3 0.03 0.1 0.1 0.3 0.001 1.2 0.002 2.5 0.1 0.4 0.1 0.8 0.005 0.6 0.01 1.2 0.01 0.1 0.03 0.2 0.1 0.1 0.1 0.1 0.05 0.05 0.1 0.1 0.4 0.4 0.9 0.9 0.6 0.6 1.2 1.2 0.7 0.7 1.3 1.3 0.02 0.03 0.04 0.1 0.1 0.1 0.1 0.2 0.4 0.5 0.9 1.1 0.7 0.7 1.4 1.4 0.9 1.1 1.8 2.1 0.1 0.1 0.2 0.2

Country   Germany

  Greece Ireland Italy

  Japan   Netherlands   Republic of Korea 

WHO Food Additives Series No. 71-S1, 2016

Sweden   United Kingdom

Population group (survey) Very elderly Adolescents Adults Elderly Other children (DONALD 2006) Other children (DONALD 2007) Other children (DONALD 2008) Toddlers (DONALD 2006) Toddlers (DONALD 2007) Toddlers (DONALD 2008) Very elderly Other children Adults Adolescents Adults Elderly Infants Other children Toddlers Very elderly Children General population Adults Other children Toddlers Children General population Adolescents Adults Other children Adults

DIPP: Finnish Type I Diabetes Prediction and Prevention Nutrition Study; STRIP: Turku Coronary Risk Factor Intervention Project for Children

380

Appendix 5. Dietary exposure calculations for each cluster, including consumption, concentration (lower bound and upper bound) and dietary exposure (lower bound and upper bound) Clusters with specific concentration data: Clusters 6, 7, 8, 9, 10, 11, 15 (only showing information for foods where concentration data were available) Table A5.1 Dietary exposure calculations for cluster 6 Lev2 CODE 84 85 90 91 100 101 102 104 110 120 121 123 124 170 Consumption Number of (g/day) samples 33.5 22 0.1 133 146.9 20 15.6 25 35.6 50 35.5 5.0 0.3 17.6 8.7 0.8 0.7 8.4 0.4 1 994 9 12 62 41 55 4 7 5 5 LB concentration (ng/kg) 249 3 967 390 437 934 89 121 602 2 986 885 1 897 2 363 2 342 0.4 UB concenLB UB tration exposure exposure (ng/kg) (ng/day) (ng/day) 250 8.3 8.4 22 839 0.4 2.4 531 57.3 77.9 437 6.8 6.8 94 3.3 3.3 89 121 2 532 3 775 1 053 1 897 5 150 2 342 0.4 3.2 0.6 0.2 52.4 7.7 1.6 1.7 19.6 0.000  2 163 3 3.2 0.6 0.7 66.3 9.2 1.6 3.6 19.6 0.000  2 203 3

Lev2NAME Plant origin fat Animal or vegetable fat, nes Milks (no other ingredients) Dairy products (incl. whey, excl. milk fats) Mammalian (not marine) meat, unprocessed (incl. home-cooked) Poultry (incl. pigeon) meat, unprocessed (incl. home-cooked) Mammalian offal, unprocessed (incl. home-cooked) Meat and offal, nes (incl. reptiles and amphibians), unprocessed (incl. home-cooked) Eggs Freshwater fish, unprocessed (incl. homecooked) Marine fish, unprocessed (incl. homecooked) Molluscs and cephalopods, unprocessed (incl. home-cooked) Other fishes and aquatic animals, unprocessed (incl. home-cooked) Food for infants and small children Total Mean dietary exposure (ng/kg bw per day)

excl.: excluding; incl.: including; LB: lower bound; nes: not elsewhere specified; UB: upper bound

381

Safety evaluation of certain food additives and contaminants Eightieth JECFA

Table A5.2 Dietary exposure calculations for cluster 7 LB concentration (ng/kg) 13 567 11 1 130 11 330 69 950 1 987 22 242 1 481 3 427 22 219 286 889 159 203 414 811 7 071 13 448 891 273 92 734 2 528 3 497 13 443 UB concenUB LB tration exposure exposure (ng/kg) (ng/day) (ng/day) 46 0.2 0.6 112 0.3 0.6 182 0.001 0.001 1 130 23.4 23.4 24 430 69 950 2 342 22 265 2 021 3 945 63 248 334 168 241 240 414 947 7 457 13 482 1 018 439 92 837 2 560 3 497 13 443 0.4 71.6 2.1 93.9 41.7 75.5 13.3 6.1 10.5 36.1 6.0 2.4 0.8 6.6 23.3 37.3 8.6 1.1 1.6 287.8 73.8 2.2 222.3 1 049 17 0.8 93.2 2.1 93.9 49.2 103.0 15.3 17.4 11.9 42.3 11.4 3.7 0.9 6.6 27.3 39.3 8.7 1.2 2.6 288.2 74.7 2.2 222.3 1 143 19

Lev2 CODE 15 22 31 40 48 60 61 70 80 83 84 85 90 91 100 101 102 104 105 110 120

WHO Food Additives Series No. 71-S1, 2016

121 122 123 124 127 170 190

Lev2NAME Fresh fruit, nes Tree nuts (excl. groundnut) Roots and tubers, nes Brassica (cole or cabbage) vegetables, head cabbages, flowerhead cabbages Other and mixed vegetables Cereal grains & flours Further processed cereals and by-products Sugars, honey, candies (excl. chocolate), sweeteners and molasses Milk fats Marine animal fat Plant origin fat Animal or vegetable fat, nes Milks (no other ingredients) Dairy products (incl. whey, excl. milk fats) Mammalian (not marine) meat, unprocessed (incl. home-cooked) Poultry (incl. pigeon) meat, unprocessed (incl. home-cooked) Mammalian offal, unprocessed (incl. home-cooked) Meat and offal, nes (incl. reptiles and amphibians), unprocessed (incl. home-cooked) Meat and offal, processed (excl. marine) Eggs Freshwater fish, unprocessed (incl. homecooked) Marine fish, unprocessed (incl. homecooked) Crustaceans, unprocessed (incl. homecooked) Molluscs ans cephalopods, unprocessed (incl. home-cooked) Other fishes and aquatic animals, unprocessed (incl. home-cooked) Processed aquatic animals Food for infants and small children Out of classifying Total Mean dietary exposure (ng/kg bw per day)

Consumption Number (g/day) of samples 14.0 6 5.2 1 0.1 2 20.7 1 33.4 216.7 30.1 98.9 21.0 0.0 51.0 3.9 276.9 48.1 126.6 67.9 15.2 3.9 15.9 28.8 5.3 0.6 1.2 5.9 3.1 29.2 0.6 16.5 1 994 14 6 4 1 18 44 32 27 45 80 131 102 194 80 5 201 771 560 55 296 1 014 163 25 2

excl.: excluding; incl.: including; LB: lower bound; nes: not elsewhere specified; UB: upper bound

382

Supplement 1: Non-dioxin-like polychlorinated biphenyls

Table A5.3 Dietary exposure calculations for cluster 8 LB concentration (ng/kg) 25 105 37 34 219 0.04 125 55 256 1 222 44 2 0 3 545 120 650 61 515 331 741 322 72 459 0 33 580 1 865 560 27 476 0.002 2 UB LB UB concentration exposure exposure (ng/kg) (ng/day) (ng/day) 25 0.6 0.6 105 6.1 6.1 37 4.94E-05 4.94E-05 34 1.3 1.3 219 0.04 125 55 256 1 222 44 2 3 3 643 120 650 61 515 331 743 374 93 520 4 33 580 9 998 1 401 30 663 519 3 7.5 0.002 1.9 0.5 0 1.6 10.8 0.03 0.0 90.5 6 484.8 64.6 77.3 39.5 47.3 3.7 2.4 0 47.4 0.004 29.4 1.5 5.4 2E-07 0.001 7.5 0.002 1.9 0.5 0 1.6 10.8 0.03 0.3 93.0 6 484.8 64.6 77.3 39.6 55.0 4.7 2.7 0.003 47.4 0.04 33.9 3.8 6.1 0.05 0.002

Lev2 CODE 10 12 31 40 42 43 44 45 50 53 60 61 70 80 84 85 90 91 100 101 102 103 104 105 110 120 121 122 123

Lev2NAME Berries and other small fruits, fresh Pome fruits, fresh Roots and tubers, nes Brassica (cole or cabbage) vegetables, head cabbages, flowerhead cabbages Fruiting vegetables, cucurbits Fruiting vegetables (other than cucurbits) and mushrooms Leafy vegetables (including brassica leafy vegetables and seaweed) Legume vegetables (fresh/green) Herbs Spices & condiments Cereal grains & flours Further processed cereals and by-products Sugars, honey, candies (excl. chocolate), sweeteners and molasses Milk fats Plant origin fat Animal or vegetable fat, nes Milks (no other ingredients) Dairy products (incl. whey, excl. milk fats) Mammalian (not marine) meat, unprocessed (incl. home-cooked) Poultry (incl. pigeon) meat, unprocessed (incl. home-cooked) Mammalian offal, unprocessed (incl. home-cooked) Poultry offal, unprocessed (incl. homecooked) Meat and offal, nes (incl. reptiles and amphibians), unprocessed (incl. home-cooked) Meat and offal, processed (excl. marine) Eggs Freshwater fish, unprocessed (incl. homecooked) Marine fish, unprocessed (incl. homecooked) Crustaceans, unprocessed (incl. homecooked) Molluscs and cephalopods, unprocessed (incl. home-cooked)

Number Consumption (g/day) of samples 23.9 9 57.8 11 0.0 20 39.8 20 34.2 61.0 15.4 8.9 0.0 1.3 245.0 15.5 110.4 25.5 53.7 1.0 233.3 53.3 146.9 50.9 5.2 0.7 1.4 4.6 34.0 2.7 0.2 0.1 0.9 10 1 10 10 1 1 10 1 18 28 1 9 4 808 431 422 159 44 1 29 48 372 354 831 39 18

383

Safety evaluation of certain food additives and contaminants Eightieth JECFA

Table A5.3 (continued) LB concentration (ng/kg) 35 643 0 0 163 858 UB LB UB concentration exposure exposure (ng/kg) (ng/day) (ng/day) 36 241 116.9 118.9 10 10 351 1 083 0 0 0.001 1.2 7 042 117 0.1 0.7 0.002 1.5 7 065 118

Lev2 CODE Lev2NAME 124 Other fishes and aquatic animals, unprocessed (incl. home-cooked) 130 Fruit & vegetable juices 151 Other alcoholic beverages 170 Food for infants and small children 190 Out of classifying Total Mean dietary exposure (ng/kg bw per day)

Consumption Number (g/day) of samples 3.3 389 9.9 67.7 0.0 1.4 1 994 21 11 39 6

excl.: excluding; incl.: including; LB: lower bound; nes: not elsewhere specified; UB: upper bound

Table A5.4 Dietary exposure calculations for cluster 9 LB concentration (ng/kg) 37 16 37 36 11 13 28 12 0 17 0 25 343 0 19 51 46 3 UB LB UB concentration exposure exposure (ng/kg) (ng/day) (ng/day) 37 0.7 0.7 42 0.01 0.02 37 1.9 1.9 36 11 14 28 12 300 83 300 300 543 300 28 84 64 110 2.0 0.6 0.4 0.2 0.1 0 6.5 0 0.0 0.1 0 0.8 0.0 3.6 0.1 2.0 0.7 0.5 0.2 0.1 0.9 32.6 1.0 0.1 0.2 5.8 1.1 0.0 5.1 2.6

Lev2 CODE 20 31 40 42 43 44

45 47 54 60 61 71 80 84 90 91 100 101

Lev2NAME Pulses (dry, prepared and composites) Roots and tubers, nes Brassica (cole or cabbage) vegetables, head cabbages, flowerhead cabbages Fruiting vegetables, cucurbits Fruiting vegetables (other than cucurbits) and mushrooms Leafy vegetables (including brassica leafy vegetables and seaweed) Legume vegetables (fresh/green) Stalk and stem vegetables Sauces & vinegars Cereal grains & flours Further processed cereals and by-products Cocoa, cola and their non-liquid derivates Milk fats Plant origin fat Milks (no other ingredients) Dairy products (incl. whey, excl. milk fats) Mammalian (not marine) meat, unprocessed (incl. home-cooked) Poultry (incl. pigeon) meat, unprocessed (incl. home-cooked)

Consumption Number (g/day) of samples 20.1 216 0.4 276 50.6 12 56.7 58.3 33.4 7.2 7.2 3.1 391.9 3.4 0.3 0.4 19.4 40.3 0.6 79.5 23.8 42 24 162 36 6 24 324 240 24 24 24 1 698 648 2 256 204

384

WHO Food Additives Series No. 71-S1, 2016

Supplement 1: Non-dioxin-like polychlorinated biphenyls

Lev2 CODE Lev2NAME 102 Mammalian offal, unprocessed (incl. home-cooked) 105 Meat and offal, processed (excl. marine) 110 Eggs 120 Freshwater fish, unprocessed (incl. homecooked) 121 Marine fish, unprocessed (incl. homecooked) 122 Crustaceans, unprocessed (incl. homecooked) 123 Molluscs and cephalopods, unprocessed (incl. home-cooked) 124 Other fishes and aquatic animals, unprocessed (incl. home-cooked) 131 Non-alcoholic beverages (excl. milk-based beverages, stimulants and water) 140 Coffee (or substitute) based beverages 142 Tea and mate beverages 170 Food for infants and small children 190 Out of classifying Total Mean dietary exposure (ng/kg bw per day)

Consumption Number (g/day) of samples 6.3 54 1.7 31.6 27.0 1.3 5.3 12.2 7.2 11.2 0.7 1.1 0.1 0.5 1 994 72 360 1 590 474 276 180 12 24 24 24 822 240

LB concentration (ng/kg) 2 0 97 648 597 280 922 483 0 0 0 43 0

UB LB UB concentration exposure exposure (ng/kg) (ng/day) (ng/day) 136 0.0 0.9 300 157 670 625 318 1 015 484 300 300 300 43 300 0 3.0 17.5 0.8 1.5 11.3 3.5 0 0 0 0.004 0 55 1 0.5 5.0 18.1 0.8 1.7 12.4 3.5 3.4 0.2 0.3 0.004 0.2 103 2

excl.: excluding; incl.: including; LB: lower bound; nes: not elsewhere specified; UB: upper bound

Table A5.5 Dietary exposure calculations for cluster 10 LB concentration (ng/kg) 32 36 21 271 128 80 24 18 0 UB LB UB concentration exposure exposure (ng/kg) (ng/day) (ng/day) 32 0.1 0.1 36 0.1 0.1 21 0.0 0.0 15 688 35.1 2 031.5 128 81 25 18 12 793 0.5 0.7 0.0 1.1 0 0.5 0.7 0.0 1.1 13.6

Lev2 CODE 22 32 53 70 71 80 83 84 85

Lev2NAME Tree nuts (excl. groundnut) Roots and tubers processed Spices & condiments Sugars, honey, candies (excl. chocolate), sweeteners and molasses Cocoa, cola and their non-liquid derivates Milk fats Marine animal fat Plant origin fat Animal or vegetable fat, nes

Consumption Number (g/day) of samples 4.2 10 3.2 10 1.9 10 129.5 144 4.1 8.7 0.0 59.7 1.1 10 58 18 26 66

385

Safety evaluation of certain food additives and contaminants Eightieth JECFA

Table A5.5 (continued) Lev2 CODE 90 91 100 101 102 103 105 110 120 121 122 123 124 170 190 Consumption Number (g/day) of samples 246.2 304 32.8 156 108.7 532 84.4 6.8 0.3 3.1 39.1 3.8 0.5 1.3 5.5 6.6 0.3 2.9 1 994 58 28 10 30 364 2 738 1 510 485 1 422 1 416 68 112 LB concentration (ng/kg) 112 37 163 788 50 88 86 2 796 3 149 1 084 957 1 170 1 394 11 36 UB LB UB concentration exposure exposure (ng/kg) (ng/day) (ng/day) 1 706 27.7 420.0 37 1.2 1.2 652 17.7 70.9 825 76 88 86 2 933 3 395 1 255 1 193 1 387 1 748 11 36 66.5 0.3 0.03 0.3 109.5 12.1 0.5 1.3 6.5 9.2 0.003 0.1 290 5 69.7 0.5 0.03 0.3 114.8 13.0 0.6 1.6 7.7 11.5 0.003 0.1 2 760 46

Lev2NAME Milks (no other ingredients) Dairy products (incl. whey, excl. milk fats) Mammalian (not marine) meat, unprocessed (incl. home-cooked) Poultry (incl. pigeon) meat, unprocessed (incl. home-cooked) Mammalian offal, unprocessed (incl. home-cooked) Poultry offal, unprocessed (incl. homecooked) Meat and offal, processed (excl. marine) Eggs Freshwater fish, unprocessed (incl. homecooked) Marine fish, unprocessed (incl. homecooked) Crustaceans, unprocessed (incl. homecooked) Molluscs and cephalopods, unprocessed (incl. home-cooked) Other fishes and aquatic animals, unprocessed (incl. home-cooked) Food for infants and small children Out of classifying Total Mean dietary exposure (ng/kg bw per day)

excl.: excluding; incl.: including; LB: lower bound; nes: not elsewhere specified; UB: upper bound

WHO Food Additives Series No. 71-S1, 2016

Table A5.6 Dietary exposure calculations for cluster 11 LB concentration (ng/kg) 0 1 956 756 4 355 0 UB LB UB concentration exposure exposure (ng/kg) (ng/day) (ng/day) 1 000 0 6.4 8 597 41.5 182.6 10 200 9.5 127.8 11 688 2.5 6.8 1 980 0 0

Lev2 CODE 20 80 81 82 83

Lev2NAME Pulses (dry, prepared and composites) Milk fats Mammalian fats (no milk fat) Poultry fats Marine animal fat

Consumption Number (g/day) of samples 6.4 1 21.2 7 12.5 3 0.6 5 0.0 2

386

Supplement 1: Non-dioxin-like polychlorinated biphenyls

Lev2 CODE 84 85 90 91 100 101 104 110 120 121 122 123 124 170 190

Lev2NAME Plant origin fat Animal or vegetable fat, nes Milks (no other ingredients) Dairy products (incl. whey, excl. milk fats) Mammalian (not marine) meat, unprocessed (incl. home-cooked) Poultry (incl. pigeon) meat, unprocessed (incl. home-cooked) Meat and offal, nes (incl. reptiles and amphibians), unprocessed (incl. home-cooked) Eggs Freshwater fish, unprocessed (incl. homecooked) Marine fish, unprocessed (incl. homecooked) Crustaceans, unprocessed (incl. homecooked) Molluscs and cephalopods, unprocessed (incl. home-cooked) Other fishes and aquatic animals, unprocessed (incl. home-cooked) Food for infants and small children Out of classifying Total Mean dietary exposure (ng/kg bw per day)

Consumption Number (g/day) of samples 55.7 100 8.7 43 266.9 91 69.0 25 117.1 18 51.2 0.3 42.1 1.2 0.2 0.4 1.1 11.4 0.0 0.0 1 994 11 1 60 129 32 7 22 178 8 7

LB concentration (ng/kg) 161 17 772 1 609 1 477 364 298 338 7 163 12 186 8 891 2 679 2 006 214 614 0 833

UB LB UB concentration exposure exposure (ng/kg) (ng/day) (ng/day) 2 529 9.0 140.9 19 865 155.4 173.7 3 552 429.4 947.9 3 440 101.9 237.4 671 42.6 78.6 2 473 3 671 12 012 13 943 10 162 4 108 4 673 214 831 2 875 3 595 15.2 0.1 301.8 14.6 1.9 1.1 2.3 2 444.0 0 0 3 573 60 126.7 1.0 506.1 16.7 2.1 1.6 5.3 2 446.5 0 0 5 008 83

excl.: excluding; incl.: including; LB: lower bound; nes: not elsewhere specified; UB: upper bound

Table A5.7 Dietary exposure calculations for cluster 15 LB concentration (ng/kg) 1 0 4 0.1 0 0 0 0.02 UB LB UB concentration exposure exposure (ng/kg) (ng/day) (ng/day) 173 0.0 3.8 321 0.0 9.3 190 0.3 12.3 320 0.0 10.1 201 0.0 6.1 400 0 0.9 0.8 0 0.01 201 0.000  2 2.1

Lev2 CODE 10 11 12 13 14 16 17 20

Lev2NAME Berries and other small fruits, fresh Citrus fruits, fresh Pome fruits, fresh Stone fruits, fresh Tropical and subtropical fruits, fresh Dried fruits Prepared fruits (no dried & juice) Pulses (dry, prepared and composites)

Consumption Number (g/day) of samples 22.1 7 29.0 5 64.6 14 31.4 5 30.6 4 2.2 2 8.4 1 10.5 4

387

Safety evaluation of certain food additives and contaminants Eightieth JECFA

Table A5.7 (continued) LB concentration (ng/kg) 457 1 1 0.1 11 2 0.04 0.02 0.1 0 0.05 30 0.1 15 18 17 0.6 521 759 924 9 904 55 888 72 38 52 20 77 61 809 1 1 044 19 488 UB LB UB concentration exposure exposure (ng/kg) (ng/day) (ng/day) 807 0.3 0.5 301 0.003 1.2 249 2.15E-06 0.000  4 1 0.000  9 0.005 200 0.6 11.7 302 201 350 200 400 1 31 172 210 147 815 81 624 1 176 959 9 988 481 1 031 217 288 211 101 121 84 1 043 80 1 134 19 520 0.1 0.001 0.002 0.001 0 1.28E-05 1.2 0.000  2 3.9 0.4 1.5 0.004 7.5 10.7 0.3 0 2.4 1.4 25.3 1.4 5.9 1.1 0.8 0.1 0.8 0.01 31.3 40.0 10.9 6.9 25.9 1.5 4.2 0.000  1 1.3 0.6 55.1 2.9 71.2 0.5 9.0 16.6 0.3 0 20.8 1.6 76.5 10.3 24.0 5.6 1.2 0.1 1.0 0.6 34.0 40.1

Lev2 CODE 21 22 31 32 40 41 42 43 44 45 47 48 53 60 61 70 71 80 81 82 83 84 85 90 91 100 101 102 103 104 105 110 120

Lev2NAME Oilseed (incl. flour) Tree nuts (excl. groundnut) Roots and tubers, nes Roots and tubers processeded Brassica (cole or cabbage) vegetables, head cabbages, flowerhead cabbages Bulb vegetables Fruiting vegetables, cucurbits Fruiting vegetables (other than cucurbits) and mushrooms Leafy vegetables (including brassica leafy vegetables and seaweed) Legume vegetables (fresh/green) Stalk and stem vegetables Other and mixed vegetables Spices & condiments Cereal grains & flours Further processed cereals and by-products Sugars, honey, candies (excl. chocolate), sweeteners and molasses Cocoa, cola and their non-liquid derivates Milk fats Mammalian fats (no milk fat) Poultry fats Marine animal fat Plant origin fat Animal or vegetable fat, nes Milks (no other ingredients) Dairy products (incl. whey, excl. milk fats) Mammalian (not marine) meat, unprocessed (incl. home-cooked) Poultry (incl. pigeon) meat, unprocessed (incl. home-cooked) Mammalian offal, unprocessed (incl. home-cooked) Poultry offal, unprocessed (incl. homecooked) Meat and offal, nes (incl. reptiles and amphibians), unprocessed (incl. home-cooked) Meat and offal, processed (excl. marine) Eggs Freshwater fish, unprocessed (incl. homecooked)

Consumption Number (g/day) of samples 0.6 4 4.0 2 0.0 21 9.5 2 58.7 18 36.1 34.4 73.9 7.5 10.4 0.3 40.8 3.3 262.3 19.4 87.4 6.3 14.4 14.1 0.3 0.0 43.2 1.6 352.3 35.8 114.1 55.6 10.0 0.9 1.0 7.5 30.0 2.1 4 4 8 17 2 2 7 7 21 42 47 5 23 6 40 2 18 149 149 51 160 38 66 7 22 37 207 111

388

WHO Food Additives Series No. 71-S1, 2016

Supplement 1: Non-dioxin-like polychlorinated biphenyls

Lev2 CODE Lev2NAME 121 Marine fish, unprocessed (incl. homecooked) 123 Molluscs and cephalopods, unprocessed (incl. home-cooked) 124 Other fishes and aquatic animals, unprocessed (incl. home-cooked) 127 Processed aquatic animals 130 Fruit & vegetable juices 131 Non-alcoholic beverages (excl. milk based beverages, stimulants and water) 142 Tea and mate beverages 150 Beers 151 Other alcoholic beverages 160 Water 170 Food for infants and small children 190 Out of classifying Total Mean dietary exposure (ng/kg bw per day)

Consumption Number (g/day) of samples 0.1 194 1.6 3.5 21.7 13.5 55.7 0.9 225.2 61.4 4.6 0.6 6.1 1 994 5 310 2 1 4 3 1 3 2 61 21

LB concentration (ng/kg) 17 240 1 008 32 603 7 0 0.01 0 0 0.01 0.02 31 11 066

UB LB UB concentration exposure exposure (ng/kg) (ng/day) (ng/day) 17 257 2.0 2.0 1 008 32 705 7 0.04 20 1 047 0.04 13 0.04 133 11 986 1.6 112.7 0.1 0 0.000  5 0 0 0.000  4 7.79E-05 0.02 67.0 321 5 1.6 113.1 0.1 0.0 1.1 0.9 0.01 0.8 0.000  2 0.1 72.6 673 11

excl.: excluding; incl.: including; LB: lower bound; nes: not elsewhere specified; UB: upper bound

Clusters where concentration data from all countries were used for the dietary exposure calculations: Clusters 1, 2, 3, 4, 5, 12, 13, 14, 16, 17 Table A5.8 Concentration data used for the dietary exposure estimates for clusters with no concentration data Lev2 CODE 10 11 12 13 14 15 16 17 20 21 22 Lev2NAME Berries and other small fruits, fresh Citrus fruits, fresh Pome fruits, fresh Stone fruits, fresh Tropical and subtropical fruits, fresh Fresh fruit, nes Dried fruits Prepared fruits (no dried & juice) Pulses (dry, prepared and composites) Oilseed (incl. flour) Tree nuts (excl. groundnut) Number of samples 96 30 150 30 24 36 12 6 246 24 28 LB concentration (ng/kg) 15 0.008 49 0.05 0.02 13 0 0 32 457 24 UB concentration (ng/kg) 90 320 153 320 201 46 400 0.8 76 807 172

389

Safety evaluation of certain food additives and contaminants Eightieth JECFA

Table A5.8 (continued) Lev2 CODE 30 31 32 40 41 42 43 44 45 46 47 48 50 52 53 54 60 61 70 71 80 81 82 83 84 85 90 91 100 101 102 103 104 105 110 111 120 Lev2NAME Roots and tubers raw or boiled Roots and tubers, nes Roots and tubers processeded Brassica (cole or cabbage) vegetables, head cabbages, flowerhead cabbages Bulb vegetables Fruiting vegetables, cucurbits Fruiting vegetables (other than cucurbits) and mushrooms Leafy vegetables (including brassica leafy vegetables and seaweed) Legume vegetables (fresh/green) Root vegetables Stalk and stem vegetables Other and mixed vegetables Herbs Hops Spices & condiments Sauces & vinegars Cereal grains & flours Further processed cereals and by-products Sugars, honey, candies (excl. chocolate), sweeteners and molasses Cocoa, cola and their non-liquid derivates Milk fats Mammalian fats (no milk fat) Poultry fats Marine animal fat Plant origin fat Animal or vegetable fat, nes Milks (no other ingredients) Dairy products (incl. whey, excl. milk fats) Mammalian (not marine) meat, unprocessed (incl. home-cooked) Poultry (incl. pigeon) meat, unprocessed (incl. homecooked) Mammalian offal, unprocessed (incl. home-cooked) Poultry offal, unprocessed (incl. home-cooked) Meat and offal, nes (incl. reptiles and amphibians), unprocessed (incl. home-cooked) Meat and offal, processed (excl. marine) Eggs Egg products and processed eggs Freshwater fish, unprocessed (incl. home-cooked) Number of samples 0 534 22 246 24 126 78 324 108 0 18 126 6 0 58 24 546 522 540 64 566 64 275 306 1 115 2 261 33 046 4 521 7 542 2 192 2 024 58 1 169 642 6 072 0 12 848 LB concentration (ng/kg) 17 18 50 2 116 3 30 40 4 17 256 129 0 40 12 92 33 1 815 650 1 280 19 579 1 026 5 894 327 546 189 96 179 61 5 366 25 1 431 5 154 UB concentration (ng/kg) 89 19 133 302 154 219 93 85 4 26 256 254 300 131 212 4 620 174 2 884 5 085 2 291 19 680 2 585 13 323 412 719 280 231 250 78 6 053 89 1 927 5 599

390

WHO Food Additives Series No. 71-S1, 2016

Supplement 1: Non-dioxin-like polychlorinated biphenyls

Lev2 CODE 121 122 123 124 127 130 131 140 142 150 151 160 170 190

Lev2NAME Marine fish, unprocessed (incl. home-cooked) Crustaceans, unprocessed (incl. home-cooked) Molluscs and cephalopods, unprocessed (incl. homecooked) Other fishes and aquatic animals, unprocessed (incl. home-cooked) Processed aquatic animals Fruit & vegetable juices Non-alcoholic beverages (excl. milk-based beverages, stimulants and water) Coffee (or substitute) based beverages Tea and mate beverages Beers Other alcoholic beverages Water Food for infants and small children Out of classifying

Number of samples 11 710 1 367 3 690 12 806 990 132 48 24 42 6 84 12 1 728 568

LB concentration (ng/kg) 17 556 693 734 73 356 2 497 0 0.004 0 0 0 0.002 0.02 353 2 859

UB concentration (ng/kg) 18 971 948 1 029 73 589 2 529 10 160 300 620 0.04 11 0.04 495 3 412

excl.: excluding; incl.: including; LB: lower bound; nes: not elsewhere specified; UB: upper bound

391

392 UB exposure (ng/day) 1 10 2 3 7 2 0.5 0.001 1 1 0.6 0 0.3 0.004 1 0.0 1.0 46.7 6E-05 0.02 2 0.000 3 0.02 6 0.4 3.3 314.4 0.2 0.08 0 0.3 0.6 0 1.8 5.1 469.5 1.4 56.3 0.9 7.1 0.2 0.5 23.7 0.8 0.8 0.1 0 0.02 1 0.05 35.9 0.9 1.4 0.5 0 0 2 0.4 0.001 9.1 0.1 0.6 0.1 0 0 0.4 0.03 0.000 5 2 1 0.9 0 0.1 1 0.1 64.6 2.9 13.5 19.2 2.5 8.6 56.9 1.8 4.9 11.6 11.4 33.2 22.7 9.0 0.000 1 2 0.001 0.000 2 4 5 7 2 16.7 0.6 0.2 123.1 0.000 1 0.03 1E-05 0.002 5 0.09 0.08 25 72.7 20.2 10.5 117.0 0.000 6 1 0.000 5 0.002 0.8 0 0 0.6 1 0.2 0 0.03 0.09 0.6 Consumption (g/day) 13.4 Cluster 2 LB exposure (ng/day) 0.2 UB exposure (ng/day) 1 Consumption (g/day) 0.1 UB exposure (ng/day) 0.007 Consumption (g/day) 18.4 UB exposure (ng/day) 2 23 3 3 23 3 1 0.01 1 2 1 0 0.2 0.09 2 Cluster 3 LB exposure (ng/day) 0.001 Cluster 4 LB exposure (ng/day) 0.3 Safety evaluation of certain food additives and contaminants Eightieth JECFA 10 3 46.4 61.8 0.07 7 14 10 4.7 7.9 0.007 0.9 1 1 41.4 43.2 0.06 5 12 7

WHO Food Additives Series No. 71-S1, 2016

Table A5.9 Estimated dietary exposures for clusters 1, 2, 3 and 4 using concentration data from all countries

Consumption (g/day) 14.4

Cluster 1 LB exposure (ng/day) 0.2

31.7 16.3 10.8 32.8

0.000 3 0.8 0.000 6 0.000 6

42.7 1.3 1.6

0.6 0 0

15.6

0.5

1.3 3.6 59.3

0.6 0.09 0

Lev2 CODE Lev2NAME 10 Berries and other small fruits, fresh 11 Citrus fruits, fresh 12 Pome fruits, fresh 13 Stone fruits, fresh 14 Tropical and subtropical fruits, fresh 15 Fresh fruit, nes 16 Dried fruits 17 Prepared fruits (no dried & juice) 20 Pulses (dry, prepared and composites) 21 Oilseed (incl. flour) 22 Tree nuts (excl. groundnut) 30 Roots and tubers raw or boiled 31 Roots and tubers, nes 32 Roots and tubers processed 40 Brassica (cole or cabbage) vegetables, head cabbages, flowerhead cabbages 41 Bulb vegetables 42 Fruiting vegetables, cucurbits

3.1 0.2 7.5

0.05 0.004 0.4

34.3 20.7

0.05 2

Consumption (g/day) 49.6

Cluster 1 LB exposure (ng/day) 0.2 UB exposure (ng/day) 11 Consumption (g/day) 80.8 UB exposure (ng/day) 18 Consumption (g/day) 20.9 UB exposure (ng/day) 5 Consumption (g/day) 98.6 UB exposure (ng/day) 22

Cluster 2 LB exposure (ng/day) 0.3

Cluster 3 LB exposure (ng/day) 0.07

Cluster 4 LB exposure (ng/day) 0.3

1.3

0.04

0.1

0.6

0.02

0.06

0.8

0.03

0.08

7.9

0.2

0.7

7.7 0 0.003 0.6 0.1 0 0.5 0.7 49 2 469 91.5 8 423 32.9 3 152 30.8 0.0 46.3 0.0 0.0 1.3 1.3 337.8 13.0 0 5E-05 0.8 0.000 2 0 0.2 0 13 0.2 0 5E-05 1 0.000  2 0 0.3 0.4 44 3 0.4 0.0 29.4 0.0 0.0 2.0 1.6 205.9 4.0 0 2E-05 0.5 0 0 0.3 0 8 0.05 0 2E-05 0.8 0 0 0.5 0.5 27 0.8 8.7 0.3 52.2 0.0 0.0 6.5 4.6 355.3 21.9 85.3

0.3

0.7

1.5

0.06

0.1

0.5

0.02

0.04

3.0

0.1 0 0.001 0.9 0 0 0.8 0 14 0.3 8

0.2 0 0.002 1 0 0 2 1 47 5 394

9.5 0.7 23.4 0.5 0.0 2.1 2.3 372.4 7.2

0 0.003 0.4 0.1 0 0.3 0 15 0.09

101.5

9

Supplement 1: Non-dioxin-like polychlorinated biphenyls

Lev2 CODE Lev2NAME 43 Fruiting vegetables (other than cucurbits) and mushrooms 44 Leafy vegetables (including brassica leafy vegetables and seaweed) 45 Legume vegetables (fresh/ green) 46 Root vegetables 47 Stalk and stem vegetables 48 Other and mixed vegetables 50 Herbs 52 Hops 53 Spices & condiments 54 Sauces & vinegars 60 Cereal grains & flours 61 Further processed cereals and by-products 70 Sugars, honey, candies (excl. chocolate), sweeteners and molasses 71 Cocoa, cola and their nonliquid derivates 80 Milk fats 81 Mammalian fats (no milk fat) 82 Poultry fats 83 Marine animal fat 84 Plant origin fat 0.1 16 11 0.003 0.002 81 5.1 4.4 0.1 0.0 29.9 9 3 0.1 0 31 4.2 0.1 0.7 15 22 0.2 0 77 0.6 0.1 0.7 0.0 0.0 20.7 0.02 0.2 0.4 0 0 21 0.1 0.3 3 0 0 54 4.2 4.8 1.0 0.1 0.0 44.8 0.1 9 0.6 0.1 0.6 46 0.7 14 5 0.2 0.6 116

0.7

0.02

5.4 2.1 0.0 0.0 31.5

10 1 0.002 0.002 32

393

394 UB exposure (ng/day) 1 106 2 9 69.4 13 19 14.8 3 4 45.6 9 13 Consumption (g/day) 1.4 456.5 10.7 Cluster 2 LB exposure (ng/day) 8 149 6 UB exposure (ng/day) 18 188 8 Consumption (g/day) 0.1 20.4 0.3 UB exposure (ng/day) 2 8 0.2 Consumption (g/day) 0.7 126.8 17.6 UB exposure (ng/day) 10 52 13 Cluster 3 LB exposure (ng/day) 0.9 7 0.2 Cluster 4 LB exposure (ng/day) 4 41 10 3 29.2 3 7 8.0 0.8 2 127.6 12 29 1 0.009 3 0.0 0.09 0.1 6.5 0.1 0.005 0.007 0.1 0.007 35 9.7 2 2 3.0 0.5 0.7 0.009 39 5.5 5.4 0.6 1 0.3 3 1 0.4 3 Safety evaluation of certain food additives and contaminants Eightieth JECFA 0.02 18 27.1 0.1 1.9 1.4 10 24 39 52 3.3 0.08 0.3 1.1 3.4 0.0 11 26 6.3 2.8 0.03 5 0.1 7 3.3 17.1 0.2 32 49 35 53 3.5 2.5 0.08 24 0.3 33 11 12 18 44 19 48

WHO Food Additives Series No. 71-S1, 2016

Table A5.9 (continued)

Lev2 CODE 85 90 91

Consumption (g/day) 0.1 258.4 2.7

Cluster 1 LB exposure (ng/day) 0.7 84 1

100

30.9

6

101

14.6

1

102

4.8

0.9

103

0.1

0.007

104

0.5

3

105

0.2

0.004

110 111

9.1 0.1

13

120

2.0

10

121

Lev2NAME Animal or vegetable fat, nes Milks (no other ingredients) Dairy products (incl. whey, excl. milk fats) Mammalian (not marine) meat, unprocessed (incl. home-cooked) Poultry (incl. pigeon) meat, unprocessed (incl. homecooked) Mammalian offal, unprocessed (incl. home-cooked) Poultry offal, unprocessed (incl. home-cooked) Meat and offal, nes (incl. reptiles and amphibians), unprocessed (incl. homecooked) Meat and offal, processed (excl. marine) Eggs Egg products and processed eggs Freshwater fish, unprocessed (incl. home-cooked) Marine fish, unprocessed (incl. home-cooked)

0.6

11

Consumption (g/day) 0.1 0.1 0.4 0.3 0.4 0.0 0.04 0.05 0.5 0.4 0.5

Cluster 1 LB exposure (ng/day) 0.07 UB exposure (ng/day) 0.1 Consumption (g/day) 0.5 UB exposure (ng/day) 0.5 Consumption (g/day) 0.3 UB exposure (ng/day) 0.3 Consumption (g/day) 1.4 UB exposure (ng/day) 1

Cluster 2 LB exposure (ng/day) 0.4

Cluster 3 LB exposure (ng/day) 0.2

Cluster 4 LB exposure (ng/day) 1

0.1

0.08

4.4

324

325

4.2

305

306

4.4

325

326

9.4

688

690

1.5 3.3 7.7

4 0 3E-05

4 0.03 1

14.1 4.5 74.6

35 0 0.000 3

36 0.04 12

9.8 0.6 3.0

24 0 1E-05

25 0.006 0.5

14.2 12.4 17.2

35 0 7E-05

36 0.1 3

1.2 1 0.000 2 0.008 1E-05 0.09 4 1 188 20 2 121 684 11 3.7 11 13 1 358 23 1.2 1 195 2.0 93.8 12.7 39.7 0.3 0 0 2E-05 0.000 7 0.09 1 0.004 0.1 0.002 0.1 1.5 50.3 12.7 0.5 0.1 0 0 2E-05 8E-06 0.03 4 524 9

0

0.4

2.0

0

0.6

1.3

0

0.4 0.9 0.002 0.1 2E-05 0.05 4 791 13

3.6 2.5 12.8 1.4 7.4 1.9 13.4 1 664

0 0 0 2E-06 0.000 1 0.7 38 1 022 17

1 2 0.000 5 0.01 0.000 3 0.9 46 1 696 28

Lev2 CODE Lev2NAME 122 Crustaceans, unprocessed (incl. home-cooked) 123 Molluscs and cephalopods, unprocessed (incl. homecooked) 124 Other fishes and aquatic animals, unprocessed (incl. home-cooked) 127 Processed aquatic animals 130 Fruit & vegetable juices 131 Non-alcoholic beverages (excl. milk-based beverages, stimulants and water) 140 Coffee (or substitute) based beverages 142 Tea and mate beverages 150 Beers 151 Other alcoholic beverages 160 Water 170 Food for infants and small children 190 Out of classifying

2.4 4.9 0.7 0.2 0.2

0 0 1E-06 4E-06 0.06

1.0

3

Total mean exposure (ng/day) Total mean exposure (ng/kg bw per day)

1 267

538 9

Supplement 1: Non-dioxin-like polychlorinated biphenyls

excl.: excluding; incl.: including; LB: lower bound; nes: not elsewhere specified; UB: upper bound

395

396 UB exposure (ng/day) 0.4 11 1 0.7 19 0.8 0.1 0.001 3 1 3 0 0.07 0.09 2 0.6 6.6 19.4 0.01 0.1 1 0.05 0.1 3 0.4 28.3 86.3 0.2 0.7 0 0.3 5 0 2.9 6.7 165.9 28.0 23.6 6.3 36.7 1 3 41.1 1 1 0.2 0 0.5 0.4 0.3 43.5 1.8 2.6 0.6 0 0 2 0.7 0.002 18.8 0.1 1.1 0.2 0 0 0.9 0.05 0.000  8 3 2 1 0 2 0.4 0.8 73.9 0.6 1.3 17.1 0.9 157.2 168.1 4.9 8.7 5.0 465.0 2.0 0.1 135.8 0.004 0.1 5E-06 0.002 149 0.3 0.03 27 20.8 0.3 0.1 74.5 0.000  2 0.01 4E-06 0.001 7 0.04 0.03 15 2.1 2.2 0.0 132.4 2E-05 0.1 1E-06 0.002 1 0 0 0.6 0.4 4 0 0.08 0.2 0.3 Consumption (g/day) 0.8 Cluster 12 LB exposure (ng/day) 0.01 UB exposure (ng/day) 0.08 Consumption (g/day) 0.2 UB exposure (ng/day) 0.01 Consumption (g/day) 11.5 UB exposure (ng/day) 1 0.7 0.3 0.003 27 3 0.3 0.001 1 0.7 27 0 0.4 0.2 0.7 Cluster 13 LB exposure (ng/day) 0.002 Cluster 14 LB exposure (ng/day) 0.2 Safety evaluation of certain food additives and contaminants Eightieth JECFA 6 2 20.2 22.4 0.03 3 6 3 11.3 6.2 0.02 0.7 3 1 23.8 14.2 0.04 2 7 2

WHO Food Additives Series No. 71-S1, 2016

Table A5.10 Estimated dietary exposures for clusters 5, 12, 13 and 14 using concentration data from all countries

Consumption (g/day) 4.1

Cluster 5 LB exposure (ng/day) 0.06

33.7 7.6 2.2 93.1

0.000  3 0.4 0.000  1 0.002

17.5 0.3 1.4

0.2 0 0

33.1

1

1.6 15.9 67.2

0.7 0.4 0

Lev2 CODE Lev2NAME 10 Berries and other small fruits, fresh 11 Citrus fruits, fresh 12 Pome fruits, fresh 13 Stone fruits, fresh 14 Tropical and subtropical fruits, fresh 15 Fresh fruit, nes 16 Dried fruits 17 Prepared fruits (no dried & juice) 20 Pulses (dry, prepared and composites) 21 Oilseed (incl. flour) 22 Tree nuts (excl. groundnut) 30 Roots and tubers raw or boiled 31 Roots and tubers, nes 32 Roots and tubers processed 40 Brassica (cole or cabbage) vegetables, head cabbages, flowerhead cabbages 41 Bulb vegetables 42 Fruiting vegetables, cucurbits

0.8 5.1 17.8

0.01 0.09 0.9

21.1 10.5

0.03 1

Consumption (g/day) 46.1

Cluster 5 LB exposure (ng/day) 0.2 UB exposure (ng/day) 10 Consumption (g/day) 10.4 UB exposure (ng/day) 2 Consumption (g/day) 24.2 UB exposure (ng/day) 5 Consumption (g/day) 33.7 UB exposure (ng/day) 7

Cluster 12 LB exposure (ng/day) 0.03

Cluster 13 LB exposure (ng/day) 0.08

Cluster 14 LB exposure (ng/day) 0.1

2.0

0.06

0.2

2.5

0.08

0.2

0.5

0.01

0.05

0.0

0.001

0.003

5.1 0 0.001 1 0 0 1 0.1 47 1 398 107.2 10 495 35.8 3 166 8.7 0.0 41.5 0.0 0.0 2.5 1.2 247.0 26.0 0 0.000  2 0.7 0 0 0.3 0 10 0.3 0 0.000  2 1 0 0 0.6 0.4 32 6 2.1 0.0 46.4 0.0 0.0 2.2 0.6 330.5 2.8 0 2E-05 0.8 0 0 0.3 0 13 0.03 0 2E-05 1 0 0 0.6 0.2 43 0.6 3.0 0.0 32.7 0.0 0.0 7.0 0.3 321.3 3.6 107.7

0.2

0.4

0.2

0.007

0.01

0.6

0.02

0.05

3.2

0.1 0 2E-06 0.6 0 0 0.9 0 13 0.04 10

0.3 0 2E-06 0.9 0 0 2 0.1 42 0.8 498

2.8 0.3 43.9 0.0 0.0 4.4 0.4 358.4 6.6

0 0.001 0.8 0 0 0.6 0 14 0.08

86.2

8

Supplement 1: Non-dioxin-like polychlorinated biphenyls

Lev2 CODE Lev2NAME 43 Fruiting vegetables (other than cucurbits) and mushrooms 44 Leafy vegetables (including brassica leafy vegetables and seaweed) 45 Legume vegetables (fresh/ green) 46 Root vegetables 47 Stalk and stem vegetables 48 Other and mixed vegetables 50 Herbs 52 Hops 53 Spices & condiments 54 Sauces & vinegars 60 Cereal grains & flours 61 Further processed cereals and by-products 70 Sugars, honey, candies (excl. chocolate), sweeteners and molasses 71 Cocoa, cola and their nonliquid derivates 80 Milk fats 81 Mammalian fats (no milk fat) 82 Poultry fats 83 Marine animal fat 84 Plant origin fat 0.07 14 8 4E-06 0 70 1.7 5.5 0.0 0.0 14.4 3 4 0 0 15 5.3 0.2 0.9 5 28 0 0 37 0.1 0.6 0.9 0.0 0.0 22.4 0.003 1 0.6 0 0 23 0.02 2 5 0 0 58 0.9 0.7 0.9 0.0 0.0 10.7 0.03 1 0.6 0 0 11 0.2 2 5 0 0 28

0.4

0.01

5.0 1.6 0.0 0.0 26.9

9 1 2E-06 0 28

397

398 UB exposure (ng/day) 1 72 1 9 30.8 6 9 27.0 5 8 16.8 3 5 Consumption (g/day) 9.7 94.1 12.0 Cluster 12 LB exposure (ng/day) 57 31 7 UB exposure (ng/day) 129 39 9 Consumption (g/day) 0.2 98.3 1.5 UB exposure (ng/day) 2 40 1 Consumption (g/day) 0.1 22.8 0.4 UB exposure (ng/day) 1 9 0.3 Cluster 13 LB exposure (ng/day) 1 32 0.8 Cluster 14 LB exposure (ng/day) 0.6 7 0.2 6 82.6 8 19 3.9 0.4 0.9 11.8 1 3 1 0.02 2 0.7 4 5 2.4 0.0 0 0 0.0 0.001 13 3.2 0.6 0.8 4.6 0.8 1 0.002 15 2.0 0.7 34.1 0.4 0.04 183 0.5 0.05 206 Safety evaluation of certain food additives and contaminants Eightieth JECFA 0.03 22 10.6 0.0 2.7 3.2 14 56 15 20 18.8 0.5 2 0.3 4.5 0.0 15 61 3.4 0.4 0.009 6 0.03 9 0.6 4.0 1.0 18 8 19 8 4.1 3.8 0.01 6 0.05 8 37 27 21 67 23 72

WHO Food Additives Series No. 71-S1, 2016

Table A5.10 (continued)

Lev2 CODE 85 90 91

Consumption (g/day) 0.1 174.9 1.8

Cluster 5 LB exposure (ng/day) 0.6 57 1

100

33.0

6

101

24.9

2

102

3.8

0.7

103

0.2

0.01

104

0.3

2

105

0.4

0.009

110 111

11.4 0.2

16

120

6.6

34

121

Lev2NAME Animal or vegetable fat, nes Milks (no other ingredients) Dairy products (incl. whey, excl. milk fats) Mammalian (not marine) meat, unprocessed (incl. home-cooked) Poultry (incl. pigeon) meat, unprocessed (incl. homecooked) Mammalian offal, unprocessed (incl. home-cooked) Poultry offal, unprocessed (incl. home-cooked) Meat and offal, nes (incl. reptiles and amphibians), unprocessed (incl. homecooked) Meat and offal, processed (excl. marine) Eggs Egg products and processed eggs Freshwater fish, unprocessed (incl. home-cooked) Marine fish, unprocessed (incl. home-cooked)

1.4

25

Consumption (g/day) 1.0 0.7 4.2 3 4 0.0 0.03 0.04 0.2 0.2 0.2

Cluster 5 LB exposure (ng/day) 0.7 UB exposure (ng/day) 1 Consumption (g/day) 4.2 UB exposure (ng/day) 4 Consumption (g/day) 0.2 UB exposure (ng/day) 0.2 Consumption (g/day) 0.8 UB exposure (ng/day) 0.8

Cluster 12 LB exposure (ng/day) 3

Cluster 13 LB exposure (ng/day) 0.1

Cluster 14 LB exposure (ng/day) 0.6

0.7

0.5

3.9

285

286

8.0

586

588

2.8

203

204

26.5

1 946

1 952

3.0 1.5 5.9

8 0 2E-05

8 0.01 0.9

12.0 27.9 43.1

30 0 0.000  2

30 0.3 7

5.5 1.1 2.1

14 0 9E-06

14 0.01 0.3

12.7 1.1 8.0

32 0 3E-05

32 0.01 1

1.8 1 0.002 0.04 3E-05 0.06 2 1  081 18 1  835 977 16 37.1 106 127 1  881 31 1.0 1  137 0.8 65.3 6.8 1.4 2.8 0 0 1E-05 2E-05 1 0.5 0.003 0.07 6E-05 1 1.6 48.0 49.1 0.4 0.1 0 0 8E-05 6E-06 0.03 3 352 6

0

0.6

5.9

0

2

0.9

0

0.3 1 0.002 0.5 2E-05 0.04 3 646 11

1.2 5.3 11.4 0.2 0.2 0.1 1.6 1  321

0 0 0 0 4E-06 0.04 5 2  318 39

0.4 3 0.000  5 0.003 1E-05 0.06 5 2  982 50

Lev2 CODE Lev2NAME 122 Crustaceans, unprocessed (incl. home-cooked) 123 Molluscs and cephalopods, unprocessed (incl. homecooked) 124 Other fishes and aquatic animals, unprocessed (incl. home-cooked) 127 Processed aquatic animals 130 Fruit & vegetable juices 131 Non-alcoholic beverages (excl. milk-based beverages, stimulants and water) 140 Coffee (or substitute) based beverages 142 Tea and mate beverages 150 Beers 151 Other alcoholic beverages 160 Water 170 Food for infants and small children 190 Out of classifying

2.3 39.3 3.6 0.6 0.1

0 0 6E-06 1E-05 0.04

0.6

2

Total mean exposure (ng/day) Total mean exposure (ng/kg bw per day)

1 247

508 8

Supplement 1: Non-dioxin-like polychlorinated biphenyls

excl.: excluding; incl.: including; LB: lower bound; nes: not elsewhere specified; UB: upper bound

399

Safety evaluation of certain food additives and contaminants Eightieth JECFA

Table A5.11 Estimated dietary exposures for clusters 16 and 17 using concentration data from all countries Consumption (g/day) 0.0 0.1 0.1 0.0 455.7 8.2 0.0 0.0 70.1 8.4 0.0 539.7 0.0 19.7 0.0 9.7 13.6 1.0 0.0 0.0 0.1 0.0 31.2 0.0 0.0 0.5 0.2 136.8 1.4 29.7 0.2 0.1 0.9 0.0 0.0 Cluster 16 LB exposure (ng/day) 0.000  2 1E-06 0.006 0 0.008 0.1 0 0 2 4 0.000  9 0 0.000  3 0.4 0.001 0.01 2 0.003 0.000  1 0.002 0 2E-06 0.5 0 0 0.06 0 5 0.02 3 0.006 0.2 0.6 0 0 UB exposure (ng/day) 0.001 0.04 0.02 0.000  8 91 0.4 0.002 3E-05 5 7 0.007 0 0.002 0.4 0.003 3 2 0.2 0.000  3 0.003 0 2E-06 0.8 0 0 0.1 0.07 18 0.3 137 0.03 0.3 5 0 0 Consumption (g/day) 0.3 4.4 1.5 0.0 411.2 23.6 0.3 18.9 9.0 0.9 347.3 348.7 39.7 0.2 0.0 8.7 0.1 0.3 0.0 0.0 0.0 0.0 61.6 0.0 0.0 1.4 3.7 182.9 18.3 59.8 2.3 3.1 2.5 0.0 0.0 Cluster 17 LB exposure (ng/day) 0.004 4E-05 0.07 0 0.007 0.3 0 0 0.3 0.4 8 0 0.7 0.004 0 0.01 0.006 0.000  9 0 0 0 0 1 0 0 0.2 0 7 0.2 5 0.08 6 2 0 0 UB exposure (ng/day) 0.03 1 0.2 0 82 1 0.1 0.02 0.7 0.7 60 0 4 0.004 0 3 0.008 0.06 0 0 0 0 2 0 0 0.3 1 24 4 276 0.4 9 13 0 0

Lev2 CODE 10 11 12 13 14 15 16 17 20 21 22 30 31 32 40 41 42 43 44 45 46 47 48 50 52 53 54 60 61 70 71 80 81 82 83

Lev2NAME Berries and other small fruits, fresh Citrus fruits, fresh Pome fruits, fresh Stone fruits, fresh Tropical and subtropical fruits, fresh Fresh fruit, nes Dried fruits Prepared fruits (no dried & juice) Pulses (dry, prepared and composites) Oilseed (incl. flour) Tree nuts (excl. groundnut) Roots and tubers raw or boiled Roots and tubers, nes Roots and tubers processed Brassica (cole or cabbage) vegetables, head cabbages, flowerhead cabbages Bulb vegetables Fruiting vegetables, cucurbits Fruiting vegetables (other than cucurbits) and mushrooms Leafy vegetables (including brassica leafy vegetables and seaweed) Legume vegetables (fresh/green) Root vegetables Stalk and stem vegetables Other and mixed vegetables Herbs Hops Spices & condiments Sauces & vinegars Cereal grains & flours Further processed cereals and by-products Sugars, honey, candies (excl. chocolate), sweeteners and molasses Cocoa, cola and their non-liquid derivates Milk fats Mammalian fats (no milk fat) Poultry fats Marine animal fat

400

WHO Food Additives Series No. 71-S1, 2016

Supplement 1: Non-dioxin-like polychlorinated biphenyls

Lev2 CODE 84 85 90 91 100 101 102 103 104

105 110 111 120 121 122 123 124 127 130 131 140 142 150 151 160 170 190

Lev2NAME Plant origin fat Animal or vegetable fat, nes Milks (no other ingredients) Dairy products (incl. whey, excl. milk fats) Mammalian (not marine) meat, unprocessed (incl. home-cooked) Poultry (incl. pigeon) meat, unprocessed (incl. home-cooked) Mammalian offal, unprocessed (incl. home-cooked) Poultry offal, unprocessed (incl. homecooked) Meat and offal, nes (incl. reptiles and amphibians), unprocessed (incl. homecooked) Meat and offal, processed (excl. marine) Eggs Egg products and processed eggs Freshwater fish, unprocessed (incl. home-cooked) Marine fish, unprocessed (incl. homecooked) Crustaceans, unprocessed (incl. homecooked) Molluscs and cephalopods, unprocessed (incl. home-cooked) Other fishes and aquatic animals, unprocessed (incl. home-cooked) Processed aquatic animals Fruit & vegetable juices Non-alcoholic beverages (excl. milk-based beverages, stimulants and water) Coffee (or substitute) based beverages Tea and mate beverages Beers Other alcoholic beverages Water Food for infants and small children Out of classifying Total mean exposure (ng/day) Total mean exposure (ng/kg bw per day)

Consumption (g/day) 14.8 0.1 59.0 0.1 20.4 5.0 3.3 0.0 3.8

Cluster 16 LB exposure (ng/day) 15 0.7 19 0.08 4 0.5 0.6 0.000  6 20

UB exposure (ng/day) 38 2 24 0.1 6 1 0.8 0.000  7 23

Consumption (g/day) 22.6 3.7 27.7 1.9 66.3 55.2 4.0 0.0 0.0

Cluster 17 LB exposure (ng/day) 23 22 9 1 13 5 0.7 0 0

UB exposure (ng/day) 58 49 11 1 19 13 1 0 0

0.2 1.3 0.0 18.4 0.1 0.0 0.0 0.9 2.4 0.2 7.0 2.9 0.6 153.7 138.6 0.1 0.1 0.9 1 762

0.006 2 0 95 2 0.03 0.03 68 6 0 3E-05 0 0 0 0.000 2 2E-06 0.03 3 253 4

0.02 3 0 103 2 0.04 0.04 68 6 0.002 1 0.9 0.3 0.006 1 5E-06 0.05 3 554 9

6.2 8.8 0.0 0.1 3.7 1.4 4.9 38.0 30.7 5.4 29.7 6.4 0.9 52.2 31.9 6.1 1.0 21.9 1 981

0.2 13 0 0.4 65 1 4 2 787 77 0 0.000  1 0 0 0 5E-05 0.000  1 0.4 62 3 115 52

0.6 17 0 0.4 70 1 5 2 796 78 0.05 5 2 0.5 0.002 0.3 0.000  3 0.5 75 3 686 61

excl.: excluding; incl.: including; LB: lower bound; nes: not elsewhere specified; UB: upper bound

401

Appendix 6. GEMS/Food cluster diets 2012 – Countries by cluster Cluster Country 1 Afghanistan, Algeria, Azerbaijan, Iraq, Jordan, Libya, Mauritania, Mongolia, Morocco, Occupied Paestinian Territory, Pakistan, Syrian Arab Republic, Tunisia, Turkmenistan, Uzbekistan, Yemen 2 Albania, Bosnia and Herzegovina, Georgia, Kazakhstan, Kyrgyzstan, Montenegro, Republic of Moldova, Ukraine 3 Angola, Benin, Burundi, Cameroon, Congo, Côte d’lvoire, Democratic Republic of the Congo, Ghana, Guinea, Liberia, Madagascar, Mozambique, Paraguay, Togo, Zambia 4 Antigua and Barbuda, Bahamas, Brunei Darussalam, French Polynesia, Grenada, Israel, Jamaica, Kuwait, Netherlands Antilles, Saint Kitts and Nevis, Saint Lucia, Saint Vincent and the Grenadines, Saudi Arabia, United Arab Emirates 5 Argentina, Bolivia (Plurinational State of), Brazil, Cape Verde, Chile, Colombia, Costa Rica, Djibouti, Dominican Republic, Ecuador, El Salvador, Guatemala, Guyana, Honduras, India, Malaysia, Maldives, Mauritius, Mexico, New Caledonia, Nicaragua, Panama, Peru, Seychelles, South Africa, Suriname, Tajikistan, The former Yugoslav Republic of Macedonia, Trinidad and Tobago, Venezuela (Bolivarian Republic of) 6 Armenia, Cuba, Egypt, Greece, Iran (Islamic Republic of), Lebanon, Turkey 7 Australia, Bermuda, Finland, France, Iceland, Luxembourg, Norway, Switzerland, United Kingdom, Uruguay 8 Austria, Germany, Poland, Spain 9 Bangladesh, Cambodia, China, Democratic People’s Republic of Korea, Guinea Bissau, Indonesia, Lao People’s Democratic Republic, Myanmar, Nepal, Philippines, Sierra Leone, Thailand, Timor Leste, Viet Nam 10 Belarus, Bulgaria, Canada, Croatia, Cyprus, Estonia, Italy, Japan, Latvia, Malta, New Zealand, Republic of Korea, Russian Federation, United States of America 11 Belgium, Netherlands 12 Belize, Dominica 13 Botswana, Burkina Faso, Central African Republic, Chad, Ethiopia, Gambia, Haiti, Kenya, Malawi, Mali, Namibia, Niger, Nigeria, Senegal, Somalia, Sudan, Swaziland, United Republic of Tanzania, Zimbabwe 14 Comoros, Fiji Islands, Kiribati, Papua New Guinea, Solomon Islands, Sri Lanka, Vanuatu 15 Czech Republic, Denmark, Hungary, Ireland, Lithuania, Montenegro, Portugal, Romania, Serbia, Slovakia, Slovenia, Sweden 16 Gabon, Rwanda, Uganda 17 Samoa, Sao Tome and Principe

402

ANNEX 1 Reports and other documents resulting from previous meetings of the Joint FAO/WHO Expert Committee on Food Additives 1. General principles governing the use of food additives (First report of the Joint FAO/WHO Expert Committee on Food Additives). FAO Nutrition Meetings Report Series, No. 15, 1957; WHO Technical Report Series, No. 129, 1957 (out of print). Procedures for the testing of intentional food additives to establish their safety for use (Second report of the Joint FAO/WHO Expert Committee on Food Additives). FAO Nutrition Meetings Report Series, No. 17, 1958; WHO Technical Report Series, No. 144, 1958 (out of print). Specifications for identity and purity of food additives (antimicrobial preservatives and antioxidants) (Third report of the Joint FAO/WHO Expert Committee on Food Additives). These specifications were subsequently revised and published as Specifications for identity and purity of food additives, Vol. I. Antimicrobial preservatives and antioxidants, Rome, Food and Agriculture Organization of the United Nations, 1962 (out of print). Specifications for identity and purity of food additives (food colours) (Fourth report of the Joint FAO/ WHO Expert Committee on Food Additives). These specifications were subsequently revised and published as Specifications for identity and purity of food additives, Vol. II. Food colours, Rome, Food and Agriculture Organization of the United Nations, 1963 (out of print). Evaluation of the carcinogenic hazards of food additives (Fifth report of the Joint FAO/WHO Expert Committee on Food Additives). FAO Nutrition Meetings Report Series, No. 29, 1961; WHO Technical Report Series, No. 220, 1961 (out of print). Evaluation of the toxicity of a number of antimicrobials and antioxidants (Sixth report of the Joint FAO/ WHO Expert Committee on Food Additives). FAO Nutrition Meetings Report Series, No. 31, 1962; WHO Technical Report Series, No. 228, 1962 (out of print). Specifications for the identity and purity of food additives and their toxicological evaluation: emulsifiers, stabilizers, bleaching and maturing agents (Seventh report of the Joint FAO/WHO Expert Committee on Food Additives). FAO Nutrition Meetings Series, No. 35, 1964; WHO Technical Report Series, No. 281, 1964 (out of print). Specifications for the identity and purity of food additives and their toxicological evaluation: food colours and some antimicrobials and antioxidants (Eighth report of the Joint FAO/WHO Expert Committee on Food Additives). FAO Nutrition Meetings Series, No. 38, 1965; WHO Technical Report Series, No. 309, 1965 (out of print). Specifications for identity and purity and toxicological evaluation of some antimicrobials and antioxidants. FAO Nutrition Meetings Report Series, No. 38A, 1965; WHO/Food Add/24.65 (out of print).

2.

3.

4.

5.

6.

7.

8.

9.

10. Specifications for identity and purity and toxicological evaluation of food colours. FAO Nutrition Meetings Report Series, No. 38B, 1966; WHO/Food Add/66.25. 403

Safety evaluation of certain food additives and contaminants Eightieth JECFA

11. Specifications for the identity and purity of food additives and their toxicological evaluation: some antimicrobials, antioxidants, emulsifiers, stabilizers, flour treatment agents, acids, and bases (Ninth report of the Joint FAO/WHO Expert Committee on Food Additives). FAO Nutrition Meetings Series, No. 40, 1966; WHO Technical Report Series, No. 339, 1966 (out of print). 12. Toxicological evaluation of some antimicrobials, antioxidants, emulsifiers, stabilizers, flour treatment agents, acids, and bases. FAO Nutrition Meetings Report Series, No. 40A, B, C; WHO/Food Add/67.29. 13. Specifications for the identity and purity of food additives and their toxicological evaluation: some emulsifiers and stabilizers and certain other substances (Tenth report of the Joint FAO/WHO Expert Committee on Food Additives). FAO Nutrition Meetings Series, No. 43, 1967; WHO Technical Report Series, No. 373, 1967. 14. Specifications for the identity and purity of food additives and their toxicological evaluation: some flavouring substances and non-nutritive sweetening agents (Eleventh report of the Joint FAO/WHO Expert Committee on Food Additives). FAO Nutrition Meetings Series, No. 44, 1968; WHO Technical Report Series, No. 383, 1968. 15. Toxicological evaluation of some flavouring substances and non-nutritive sweetening agents. FAO Nutrition Meetings Report Series, No. 44A, 1968; WHO/Food Add/68.33. 16. Specifications and criteria for identity and purity of some flavouring substances and non-nutritive sweetening agents. FAO Nutrition Meetings Report Series, No. 44B, 1969; WHO/Food Add/69.31. 17. Specifications for the identity and purity of food additives and their toxicological evaluation: some antibiotics (Twelfth report of the Joint FAO/WHO Expert Committee on Food Additives). FAO Nutrition Meetings Series, No. 45, 1969; WHO Technical Report Series, No. 430, 1969. 18. Specifications for the identity and purity of some antibiotics. FAO Nutrition Meetings Series, No. 45A, 1969; WHO/Food Add/69.34. 19. Specifications for the identity and purity of food additives and their toxicological evaluation: some food colours, emulsifiers, stabilizers, anticaking agents, and certain other substances (Thirteenth report of the Joint FAO/WHO Expert Committee on Food Additives). FAO Nutrition Meetings Series, No. 46, 1970; WHO Technical Report Series, No. 445, 1970. 20. Toxicological evaluation of some food colours, emulsifiers, stabilizers, anticaking agents, and certain other substances. FAO Nutrition Meetings Report Series, No. 46A, 1970; WHO/Food Add/70.36. WHO Food Additives Series No. 71-S1, 2016 404 21. Specifications for the identity and purity of some food colours, emulsifiers, stabilizers, anticaking agents, and certain other food additives. FAO Nutrition Meetings Report Series, No. 46B, 1970; WHO/ Food Add/70.37. 22. Evaluation of food additives: specifications for the identity and purity of food additives and their toxicological evaluation: some extraction solvents and certain other substances; and a review of the technological efficacy of some antimicrobial agents (Fourteenth report of the Joint FAO/WHO Expert Committee on Food Additives). FAO Nutrition Meetings Series, No. 48, 1971; WHO Technical Report Series, No. 462, 1971. 23. Toxicological evaluation of some extraction solvents and certain other substances. FAO Nutrition Meetings Report Series, No. 48A, 1971; WHO/Food Add/70.39. 24. Specifications for the identity and purity of some extraction solvents and certain other substances. FAO Nutrition Meetings Report Series, No. 48B, 1971; WHO/Food Add/70.40.

Supplement 1: Non-dioxin-like polychlorinated biphenyls

25. A review of the technological efficacy of some antimicrobial agents. FAO Nutrition Meetings Report Series, No. 48C, 1971; WHO/Food Add/70.41. 26. Evaluation of food additives: some enzymes, modified starches, and certain other substances: Toxicological evaluations and specifications and a review of the technological efficacy of some antioxidants (Fifteenth report of the Joint FAO/WHO Expert Committee on Food Additives). FAO Nutrition Meetings Series, No. 50, 1972; WHO Technical Report Series, No. 488, 1972. 27. Toxicological evaluation of some enzymes, modified starches, and certain other substances. FAO Nutrition Meetings Report Series, No. 50A, 1972; WHO Food Additives Series, No. 1, 1972. 28. Specifications for the identity and purity of some enzymes and certain other substances. FAO Nutrition Meetings Report Series, No. 50B, 1972; WHO Food Additives Series, No. 2, 1972. 29. A review of the technological efficacy of some antioxidants and synergists. FAO Nutrition Meetings Report Series, No. 50C, 1972; WHO Food Additives Series, No. 3, 1972. 30. Evaluation of certain food additives and the contaminants mercury, lead, and cadmium (Sixteenth report of the Joint FAO/WHO Expert Committee on Food Additives). FAO Nutrition Meetings Series, No. 51, 1972; WHO Technical Report Series, No. 505, 1972, and corrigendum. 31. Evaluation of mercury, lead, cadmium and the food additives amaranth, diethylpyrocarbamate, and octyl gallate. FAO Nutrition Meetings Report Series, No. 51A, 1972; WHO Food Additives Series, No. 4, 1972. 32. Toxicological evaluation of certain food additives with a review of general principles and of specifications (Seventeenth report of the Joint FAO/WHO Expert Committee on Food Additives). FAO Nutrition Meetings Series, No. 53, 1974; WHO Technical Report Series, No. 539, 1974, and corrigendum (out of print). 33. Toxicological evaluation of some food additives including anticaking agents, antimicrobials, antioxidants, emulsifiers, and thickening agents. FAO Nutrition Meetings Report Series, No. 53A, 1974; WHO Food Additives Series, No. 5, 1974. 34. Specifications for identity and purity of thickening agents, anticaking agents, antimicrobials, antioxidants and emulsifiers. FAO Food and Nutrition Paper, No. 4, 1978. 35. Evaluation of certain food additives (Eighteenth report of the Joint FAO/WHO Expert Committee on Food Additives). FAO Nutrition Meetings Series, No. 54, 1974; WHO Technical Report Series, No. 557, 1974, and corrigendum. 36. Toxicological evaluation of some food colours, enzymes, flavour enhancers, thickening agents, and certain other food additives. FAO Nutrition Meetings Report Series, No. 54A, 1975; WHO Food Additives Series, No. 6, 1975. 37. Specifications for the identity and purity of some food colours, enhancers, thickening agents, and certain food additives. FAO Nutrition Meetings Report Series, No. 54B, 1975; WHO Food Additives Series, No. 7, 1975. 38. Evaluation of certain food additives: some food colours, thickening agents, smoke condensates, and certain other substances (Nineteenth report of the Joint FAO/WHO Expert Committee on Food Additives). FAO Nutrition Meetings Series, No. 55, 1975; WHO Technical Report Series, No. 576, 1975. 39. Toxicological evaluation of some food colours, thickening agents, and certain other substances. FAO Nutrition Meetings Report Series, No. 55A, 1975; WHO Food Additives Series, No. 8, 1975. 405

Safety evaluation of certain food additives and contaminants Eightieth JECFA

40. Specifications for the identity and purity of certain food additives. FAO Nutrition Meetings Report Series, No. 55B, 1976; WHO Food Additives Series, No. 9, 1976. 41. Evaluation of certain food additives (Twentieth report of the Joint FAO/WHO Expert Committee on Food Additives). FAO Food and Nutrition Meetings Series, No. 1, 1976; WHO Technical Report Series, No. 599, 1976. 42. Toxicological evaluation of certain food additives. WHO Food Additives Series, No. 10, 1976. 43. Specifications for the identity and purity of some food additives. FAO Food and Nutrition Series, No. 1B, 1977; WHO Food Additives Series, No. 11, 1977. 44. Evaluation of certain food additives (Twenty-first report of the Joint FAO/WHO Expert Committee on Food Additives). WHO Technical Report Series, No. 617, 1978. 45. Summary of toxicological data of certain food additives. WHO Food Additives Series, No. 12, 1977. 46. Specifications for identity and purity of some food additives, including antioxidants, food colours, thickeners, and others. FAO Nutrition Meetings Report Series, No. 57, 1977. 47. Evaluation of certain food additives and contaminants (Twenty-second report of the Joint FAO/WHO Expert Committee on Food Additives). WHO Technical Report Series, No. 631, 1978. 48. Summary of toxicological data of certain food additives and contaminants. WHO Food Additives Series, No. 13, 1978. 49. Specifications for the identity and purity of certain food additives. FAO Food and Nutrition Paper, No. 7, 1978. 50. Evaluation of certain food additives (Twenty-third report of the Joint FAO/WHO Expert Committee on Food Additives). WHO Technical Report Series, No. 648, 1980, and corrigenda. 51. Toxicological evaluation of certain food additives. WHO Food Additives Series, No. 14, 1980. 52. Specifications for identity and purity of food colours, flavouring agents, and other food additives. FAO Food and Nutrition Paper, No. 12, 1979. 53. Evaluation of certain food additives (Twenty-fourth report of the Joint FAO/WHO Expert Committee on Food Additives). WHO Technical Report Series, No. 653, 1980. WHO Food Additives Series No. 71-S1, 2016 406 54. Toxicological evaluation of certain food additives. WHO Food Additives Series, No. 15, 1980. 55. Specifications for identity and purity of food additives (sweetening agents, emulsifying agents, and other food additives). FAO Food and Nutrition Paper, No. 17, 1980. 56. Evaluation of certain food additives (Twenty-fifth report of the Joint FAO/WHO Expert Committee on Food Additives). WHO Technical Report Series, No. 669, 1981. 57. Toxicological evaluation of certain food additives. WHO Food Additives Series, No. 16, 1981. 58. Specifications for identity and purity of food additives (carrier solvents, emulsifiers and stabilizers, enzyme preparations, flavouring agents, food colours, sweetening agents, and other food additives). FAO Food and Nutrition Paper, No. 19, 1981. 59. Evaluation of certain food additives and contaminants (Twenty-sixth report of the Joint FAO/WHO Expert Committee on Food Additives). WHO Technical Report Series, No. 683, 1982. 60. Toxicological evaluation of certain food additives. WHO Food Additives Series, No. 17, 1982.

Supplement 1: Non-dioxin-like polychlorinated biphenyls

61. Specifications for the identity and purity of certain food additives. FAO Food and Nutrition Paper, No. 25, 1982. 62. Evaluation of certain food additives and contaminants (Twenty-seventh report of the Joint FAO/WHO Expert Committee on Food Additives). WHO Technical Report Series, No. 696, 1983, and corrigenda. 63. Toxicological evaluation of certain food additives and contaminants. WHO Food Additives Series, No. 18, 1983. 64. Specifications for the identity and purity of certain food additives. FAO Food and Nutrition Paper, No. 28, 1983. 65. Guide to specifications – General notices, general methods, identification tests, test solutions, and other reference materials. FAO Food and Nutrition Paper, No. 5, Rev. 1, 1983. 66. Evaluation of certain food additives and contaminants (Twenty-eighth report of the Joint FAO/WHO Expert Committee on Food Additives). WHO Technical Report Series, No. 710, 1984, and corrigendum. 67. Toxicological evaluation of certain food additives and contaminants. WHO Food Additives Series, No. 19, 1984. 68. Specifications for the identity and purity of food colours. FAO Food and Nutrition Paper, No. 31/1, 1984. 69. Specifications for the identity and purity of food additives. FAO Food and Nutrition Paper, No. 31/2, 1984. 70. Evaluation of certain food additives and contaminants (Twenty-ninth report of the Joint FAO/WHO Expert Committee on Food Additives). WHO Technical Report Series, No. 733, 1986, and corrigendum. 71. Specifications for the identity and purity of certain food additives. FAO Food and Nutrition Paper, No. 34, 1986. 72. Toxicological evaluation of certain food additives and contaminants. WHO Food Additives Series, No. 20. Cambridge University Press, 1987. 73. Evaluation of certain food additives and contaminants (Thirtieth report of the Joint FAO/WHO Expert Committee on Food Additives). WHO Technical Report Series, No. 751, 1987. 74. Toxicological evaluation of certain food additives and contaminants. WHO Food Additives Series, No. 21. Cambridge University Press, 1987. 75. Specifications for the identity and purity of certain food additives. FAO Food and Nutrition Paper, No. 37, 1986. 76. Principles for the safety assessment of food additives and contaminants in food. WHO Environmental Health Criteria, No. 70. Geneva, World Health Organization, 1987 (out of print). The full text is available electronically at www.who.int/pcs. 77. Evaluation of certain food additives and contaminants (Thirty-first report of the Joint FAO/WHO Expert Committee on Food Additives). WHO Technical Report Series, No. 759, 1987, and corrigendum. 78. Toxicological evaluation of certain food additives. WHO Food Additives Series, No. 22. Cambridge University Press, 1988. 79. Specifications for the identity and purity of certain food additives. FAO Food and Nutrition Paper, No. 38, 1988.

407

Safety evaluation of certain food additives and contaminants Eightieth JECFA

80. Evaluation of certain veterinary drug residues in food (Thirty-second report of the Joint FAO/WHO Expert Committee on Food Additives). WHO Technical Report Series, No. 763, 1988. 81. Toxicological evaluation of certain veterinary drug residues in food. WHO Food Additives Series, No. 23. Cambridge University Press, 1988. 82. Residues of some veterinary drugs in animals and foods. FAO Food and Nutrition Paper, No. 41, 1988. 83. Evaluation of certain food additives and contaminants (Thirty-third report of the Joint FAO/WHO Expert Committee on Food Additives). WHO Technical Report Series, No. 776, 1989. 84. Toxicological evaluation of certain food additives and contaminants. WHO Food Additives Series, No. 24. Cambridge University Press, 1989. 85. Evaluation of certain veterinary drug residues in food (Thirty-fourth report of the Joint FAO/WHO Expert Committee on Food Additives). WHO Technical Report Series, No. 788, 1989. 86. Toxicological evaluation of certain veterinary drug residues in food. WHO Food Additives Series, No. 25, 1990. 87. Residues of some veterinary drugs in animals and foods. FAO Food and Nutrition Paper, No. 41/2, 1990. 88. Evaluation of certain food additives and contaminants (Thirty-fifth report of the Joint FAO/WHO Expert Committee on Food Additives). WHO Technical Report Series, No. 789, 1990, and corrigenda. 89. Toxicological evaluation of certain food additives and contaminants. WHO Food Additives Series, No. 26, 1990. 90. Specifications for identity and purity of certain food additives. FAO Food and Nutrition Paper, No. 49, 1990. 91. Evaluation of certain veterinary drug residues in food (Thirty-sixth report of the Joint FAO/WHO Expert Committee on Food Additives). WHO Technical Report Series, No. 799, 1990. 92. Toxicological evaluation of certain veterinary drug residues in food. WHO Food Additives Series, No. 27, 1991. 93. Residues of some veterinary drugs in animals and foods. FAO Food and Nutrition Paper, No. 41/3, 1991. 94. Evaluation of certain food additives and contaminants (Thirty-seventh report of the Joint FAO/WHO Expert Committee on Food Additives). WHO Technical Report Series, No. 806, 1991, and corrigenda. 95. Toxicological evaluation of certain food additives and contaminants. WHO Food Additives Series, No. 28, 1991. 96. Compendium of food additive specifications (Joint FAO/WHO Expert Committee on Food Additives (JECFA)). Combined specifications from 1st through the 37th meetings, 1956–1990. Rome, Food and Agriculture Organization of the United Nations, 1992 (2 volumes). 97. Evaluation of certain veterinary drug residues in food (Thirty-eighth report of the Joint FAO/WHO Expert Committee on Food Additives). WHO Technical Report Series, No. 815, 1991. 98. Toxicological evaluation of certain veterinary drug residues in food. WHO Food Additives Series, No. 29, 1991. 99. Residues of some veterinary drugs in animals and foods. FAO Food and Nutrition Paper, No. 41/4, 1991. 100. Guide to specifications – General notices, general analytical techniques, identification tests, test solutions, and other reference materials. FAO Food and Nutrition Paper, No. 5, Ref. 2, 1991.

408

WHO Food Additives Series No. 71-S1, 2016

Supplement 1: Non-dioxin-like polychlorinated biphenyls

101. Evaluation of certain food additives and naturally occurring toxicants (Thirty-ninth report of the Joint FAO/WHO Expert Committee on Food Additives). WHO Technical Report Series No. 828, 1992. 102. Toxicological evaluation of certain food additives and naturally occurring toxicants. WHO Food Additives Series, No. 30, 1993. 103. Compendium of food additive specifications: addendum 1. FAO Food and Nutrition Paper, No. 52, 1992. 104. Evaluation of certain veterinary drug residues in food (Fortieth report of the Joint FAO/WHO Expert Committee on Food Additives). WHO Technical Report Series, No. 832, 1993. 105. Toxicological evaluation of certain veterinary drug residues in food. WHO Food Additives Series, No. 31, 1993. 106. Residues of some veterinary drugs in animals and food. FAO Food and Nutrition Paper, No. 41/5, 1993. 107. Evaluation of certain food additives and contaminants (Forty-first report of the Joint FAO/WHO Expert Committee on Food Additives). WHO Technical Report Series, No. 837, 1993. 108. Toxicological evaluation of certain food additives and contaminants. WHO Food Additives Series, No. 32, 1993. 109. Compendium of food additive specifications: addendum 2. FAO Food and Nutrition Paper, No. 52, Add. 2, 1993. 110. Evaluation of certain veterinary drug residues in food (Forty-second report of the Joint FAO/WHO Expert Committee on Food Additives). WHO Technical Report Series, No. 851, 1995. 111. Toxicological evaluation of certain veterinary drug residues in food. WHO Food Additives Series, No. 33, 1994. 112. Residues of some veterinary drugs in animals and foods. FAO Food and Nutrition Paper, No. 41/6, 1994. 113. Evaluation of certain veterinary drug residues in food (Forty-third report of the Joint FAO/WHO Expert Committee on Food Additives). WHO Technical Report Series, No. 855, 1995, and corrigendum. 114. Toxicological evaluation of certain veterinary drug residues in food. WHO Food Additives Series, No. 34, 1995. 115. Residues of some veterinary drugs in animals and foods. FAO Food and Nutrition Paper, No. 41/7, 1995. 116. Evaluation of certain food additives and contaminants (Forty-fourth report of the Joint FAO/WHO Expert Committee on Food Additives). WHO Technical Report Series, No. 859, 1995. 117. Toxicological evaluation of certain food additives and contaminants. WHO Food Additives Series, No. 35, 1996. 118. Compendium of food additive specifications: addendum 3. FAO Food and Nutrition Paper, No. 52, Add. 3, 1995. 119. Evaluation of certain veterinary drug residues in food (Forty-fifth report of the Joint FAO/WHO Expert Committee on Food Additives). WHO Technical Report Series, No. 864, 1996. 120. Toxicological evaluation of certain veterinary drug residues in food. WHO Food Additives Series, No. 36, 1996. 121. Residues of some veterinary drugs in animals and foods. FAO Food and Nutrition Paper, No. 41/8, 1996.

409

Safety evaluation of certain food additives and contaminants Eightieth JECFA

122. Evaluation of certain food additives and contaminants (Forty-sixth report of the Joint FAO/WHO Expert Committee on Food Additives). WHO Technical Report Series, No. 868, 1997. 123. Toxicological evaluation of certain food additives. WHO Food Additives Series, No. 37, 1996. 124. Compendium of food additive specifications, addendum 4. FAO Food and Nutrition Paper, No. 52, Add. 4, 1996. 125. Evaluation of certain veterinary drug residues in food (Forty-seventh report of the Joint FAO/WHO Expert Committee on Food Additives). WHO Technical Report Series, No. 876, 1998. 126. Toxicological evaluation of certain veterinary drug residues in food. WHO Food Additives Series, No. 38, 1996. 127. Residues of some veterinary drugs in animals and foods. FAO Food and Nutrition Paper, No. 41/9, 1997. 128. Evaluation of certain veterinary drug residues in food (Forty-eighth report of the Joint FAO/WHO Expert Committee on Food Additives). WHO Technical Report Series, No. 879, 1998. 129. Toxicological evaluation of certain veterinary drug residues in food. WHO Food Additives Series, No. 39, 1997. 130. Residues of some veterinary drugs in animals and foods. FAO Food and Nutrition Paper, No. 41/10, 1998. 131. Evaluation of certain food additives and contaminants (Forty-ninth report of the Joint FAO/WHO Expert Committee on Food Additives). WHO Technical Report Series, No. 884, 1999. 132. Safety evaluation of certain food additives and contaminants. WHO Food Additives Series, No. 40, 1998. 133. Compendium of food additive specifications: addendum 5. FAO Food and Nutrition Paper, No. 52, Add. 5, 1997. 134. Evaluation of certain veterinary drug residues in food (Fiftieth report of the Joint FAO/WHO Expert Committee on Food Additives). WHO Technical Report Series, No. 888, 1999. 135. Toxicological evaluation of certain veterinary drug residues in food. WHO Food Additives Series, No. 41, 1998. 136. Residues of some veterinary drugs in animals and foods. FAO Food and Nutrition Paper, No. 41/11, 1999. 137. Evaluation of certain food additives (Fifty-first report of the Joint FAO/WHO Expert Committee on Food Additives). WHO Technical Report Series, No. 891, 2000. 138. Safety evaluation of certain food additives. WHO Food Additives Series, No. 42, 1999. 139. Compendium of food additive specifications, addendum 6. FAO Food and Nutrition Paper, No. 52, Add. 6, 1998. 140. Evaluation of certain veterinary drug residues in food (Fifty-second report of the Joint FAO/WHO Expert Committee on Food Additives). WHO Technical Report Series, No. 893, 2000. 141. Toxicological evaluation of certain veterinary drug residues in food. WHO Food Additives Series, No. 43, 2000. 142. Residues of some veterinary drugs in animals and foods. FAO Food and Nutrition Paper, No. 41/12, 2000.

410

WHO Food Additives Series No. 71-S1, 2016

Supplement 1: Non-dioxin-like polychlorinated biphenyls

143. Evaluation of certain food additives and contaminants (Fifty-third report of the Joint FAO/WHO Expert Committee on Food Additives). WHO Technical Report Series, No. 896, 2000. 144. Safety evaluation of certain food additives and contaminants. WHO Food Additives Series, No. 44, 2000. 145. Compendium of food additive specifications, addendum 7. FAO Food and Nutrition Paper, No. 52, Add. 7, 1999. 146. Evaluation of certain veterinary drug residues in food (Fifty-fourth report of the Joint FAO/WHO Expert Committee on Food Additives). WHO Technical Report Series, No. 900, 2001. 147. Toxicological evaluation of certain veterinary drug residues in food. WHO Food Additives Series, No. 45, 2000. 148. Residues of some veterinary drugs in animals and foods. FAO Food and Nutrition Paper, No. 41/13, 2000. 149. Evaluation of certain food additives and contaminants (Fifty-fifth report of the Joint FAO/WHO Expert Committee on Food Additives). WHO Technical Report Series, No. 901, 2001. 150. Safety evaluation of certain food additives and contaminants. WHO Food Additives Series, No. 46, 2001. 151. Compendium of food additive specifications: addendum 8. FAO Food and Nutrition Paper, No. 52, Add. 8, 2000. 152. Evaluation of certain mycotoxins in food (Fifty-sixth report of the Joint FAO/WHO Expert Committee on Food Additives). WHO Technical Report Series, No. 906, 2002. 153. Safety evaluation of certain mycotoxins in food. WHO Food Additives Series, No. 47/FAO Food and Nutrition Paper 74, 2001. 154. Evaluation of certain food additives and contaminants (Fifty-seventh report of the Joint FAO/WHO Expert Committee on Food Additives). WHO Technical Report Series, No. 909, 2002. 155. Safety evaluation of certain food additives and contaminants. WHO Food Additives Series, No. 48, 2002. 156. Compendium of food additive specifications: addendum 9. FAO Food and Nutrition Paper, No. 52, Add. 9, 2001. 157. Evaluation of certain veterinary drug residues in food (Fifty-eighth report of the Joint FAO/WHO Expert Committee on Food Additives). WHO Technical Report Series, No. 911, 2002. 158. Toxicological evaluation of certain veterinary drug residues in food. WHO Food Additives Series, No. 49, 2002. 159. Residues of some veterinary drugs in animals and foods. FAO Food and Nutrition Paper, No. 41/14, 2002. 160. Evaluation of certain food additives and contaminants (Fifty-ninth report of the Joint FAO/WHO Expert Committee on Food Additives). WHO Technical Report Series, No. 913, 2002. 161. Safety evaluation of certain food additives and contaminants. WHO Food Additives Series, No. 50, 2003. 162. Compendium of food additive specifications: addendum 10. FAO Food and Nutrition Paper, No. 52, Add. 10, 2002. 163. Evaluation of certain veterinary drug residues in food (Sixtieth report of the Joint FAO/WHO Expert Committee on Food Additives). WHO Technical Report Series, No. 918, 2003. 411

Safety evaluation of certain food additives and contaminants Eightieth JECFA

164. Toxicological evaluation of certain veterinary drug residues in food. WHO Food Additives Series, No. 51, 2003. 165. Residues of some veterinary drugs in animals and foods. FAO Food and Nutrition Paper, No. 41/15, 2003. 166. Evaluation of certain food additives and contaminants (Sixty-first report of the Joint FAO/WHO Expert Committee on Food Additives). WHO Technical Report Series, No. 922, 2004. 167. Safety evaluation of certain food additives and contaminants. WHO Food Additives Series, No. 52, 2004. 168. Compendium of food additive specifications: addendum 11. FAO Food and Nutrition Paper, No. 52, Add. 11, 2003. 169. Evaluation of certain veterinary drug residues in food (Sixty-second report of the Joint FAO/WHO Expert Committee on Food Additives). WHO Technical Report Series, No. 925, 2004. 170. Residues of some veterinary drugs in animals and foods. FAO Food and Nutrition Paper, No. 41/16, 2004. 171. Toxicological evaluation of certain veterinary drug residues in food. WHO Food Additives Series, No. 53, 2005. 172. Compendium of food additive specifications: addendum 12. FAO Food and Nutrition Paper, No. 52, Add. 12, 2004. 173. Evaluation of certain food additives (Sixty-third report of the Joint FAO/WHO Expert Committee on Food Additives). WHO Technical Report Series, No. 928, 2005. 174. Safety evaluation of certain food additives. WHO Food Additives Series, No. 54, 2005. 175. Compendium of food additive specifications: addendum 13. FAO Food and Nutrition Paper, No. 52, Add. 13 (with Errata), 2005. 176. Evaluation of certain food contaminants (Sixty-fourth report of the Joint FAO/WHO Expert Committee on Food Additives). WHO Technical Report Series, No. 930, 2005. 177. Safety evaluation of certain contaminants in food. WHO Food Additives Series, No. 55/FAO Food and Nutrition Paper, No. 82, 2006. WHO Food Additives Series No. 71-S1, 2016 412 178. Evaluation of certain food additives (Sixty-fifth report of the Joint FAO/WHO Expert Committee on Food Additives). WHO Technical Report Series, No. 934, 2006. 179. Safety evaluation of certain food additives. WHO Food Additives Series, No. 56, 2006. 180. Combined compendium of food additive specifications. FAO JECFA Monographs 1, Volumes 1–4, 2005, 2006. 181. Evaluation of certain veterinary drug residues in food (Sixty-sixth report of the Joint FAO/WHO Expert Committee on Food Additives). WHO Technical Report Series, No. 939, 2006. 182. Residue evaluation of certain veterinary drugs. FAO JECFA Monographs 2, 2006. 183. Toxicological evaluation of certain veterinary drug residues in food. WHO Food Additives Series, No. 57, 2006. 184. Evaluation of certain food additives and contaminants (Sixty-seventh report of the Joint FAO/WHO Expert Committee on Food Additives). WHO Technical Report Series, No. 940, 2007.

Supplement 1: Non-dioxin-like polychlorinated biphenyls

185. Compendium of food additive specifications. FAO JECFA Monographs 3, 2006. 186. Safety evaluation of certain food additives and contaminants. WHO Food Additives Series, No. 58, 2007. 187. Evaluation of certain food additives and contaminants (Sixty-eighth report of the Joint FAO/WHO Expert Committee on Food Additives). WHO Technical Report Series, No. 947, 2007. 188. Safety evaluation of certain food additives and contaminants. WHO Food Additives Series, No. 59, 2008. 189. Compendium of food additive specifications. FAO JECFA Monographs 4, 2007. 190. Evaluation of certain food additives (Sixty-ninth report of the Joint FAO/WHO Expert Committee on Food Additives). WHO Technical Report Series, No. 952, 2009. 191. Safety evaluation of certain food additives. WHO Food Additives Series, No. 60, 2009. 192. Compendium of food additive specifications. FAO JECFA Monographs 5, 2009. 193. Evaluation of certain veterinary drug residues in food (Seventieth report of the Joint FAO/WHO Expert Committee on Food Additives). WHO Technical Report Series, No. 954, 2009. 194. Toxicological evaluation of certain veterinary drug residues in food. WHO Food Additives Series, No. 61, 2009. 195. Residue evaluation of certain veterinary drugs. FAO JECFA Monographs 6, 2009. 196. Evaluation of certain food additives (Seventy-first report of the Joint FAO/WHO Expert Committee on Food Additives). WHO Technical Report Series, No. 956, 2010. 197. Safety evaluation of certain food additives. WHO Food Additives Series, No. 62, 2010. 198. Compendium of food additive specifications. FAO JECFA Monographs 7, 2009. 199. Evaluation of certain contaminants in food (Seventy-second report of the Joint FAO/WHO Expert Committee on Food Additives). WHO Technical Report Series, No. 959, 2011. 200. Safety evaluation of certain contaminants in food. WHO Food Additives Series, No. 63/FAO JECFA Monographs 8, 2011. 201. Residue evaluation of certain veterinary drugs. FAO JECFA Monographs 9, 2010. 202. Evaluation of certain food additives and contaminants (Seventy-third report of the Joint FAO/WHO Expert Committee on Food Additives). WHO Technical Report Series, No. 960, 2011. 203. Safety evaluation of certain food additives and contaminants. WHO Food Additives Series, No. 64, 2011. 204. Compendium of food additive specifications. FAO JECFA Monographs 10, 2010. 205. Evaluation of certain food additives and contaminants (Seventy-fourth report of the Joint FAO/WHO Expert Committee on Food Additives). WHO Technical Report Series, No. 966, 2011. 206. Safety evaluation of certain food additives and contaminants. WHO Food Additives Series, No. 65, 2011. 207. Compendium of food additive specifications. FAO JECFA Monographs 11, 2011. 208. Evaluation of certain veterinary drug residues in food (Seventy-fifth report of the Joint FAO/WHO Expert Committee on Food Additives). WHO Technical Report Series, No. 969, 2012. 209. Toxicological evaluation of certain veterinary drug residues in food. WHO Food Additives Series, No. 66, 2012. 413

Safety evaluation of certain food additives and contaminants Eightieth JECFA

210. Residue evaluation of certain veterinary drugs. FAO JECFA Monographs 12, 2012. 211. Evaluation of certain food additives (Seventy-sixth report of the Joint FAO/WHO Expert Committee on Food Additives). WHO Technical Report Series, No. 974, 2012. 212. Safety evaluation of certain food additives. WHO Food Additives Series, No. 67, 2012. 213. Compendium of food additive specifications. FAO JECFA Monographs 13, 2012. 214. Evaluation of certain food additives and contaminants (Seventy-seventh report of the Joint FAO/WHO Expert Committee on Food Additives). WHO Technical Report Series, No. 983, 2013. 215. Safety evaluation of certain food additives and contaminants. WHO Food Additives Series, No. 68, 2013. 216. Compendium of food additive specifications. FAO JECFA Monographs 14, 2013. 217. Evaluation of certain veterinary drug residues in food (Seventy-eighth report of the Joint FAO/WHO Expert Committee on Food Additives). WHO Technical Report Series, No. 988, 2014. 218. Toxicological evaluation of certain veterinary drug residues in food. WHO Food Additives Series, No. 69, 2014. 219. Residue evaluation of certain veterinary drugs. FAO JECFA Monographs 15, 2014. 220. Evaluation of certain food additives (Seventy-ninth report of the Joint FAO/WHO Expert Committee on Food Additives). WHO Technical Report Series, No. 990, 2015. 221. Safety evaluation of certain food additives. WHO Food Additives Series, No. 70, 2015. 222. Compendium of food additive specifications. FAO JECFA Monographs 16, 2014. 223. Evaluation of certain food additives and contaminants (Eightieth report of the Joint FAO/WHO Expert Committee on Food Additives). WHO Technical Report Series, No. 995, 2016. 224. Safety evaluation of certain food additives and contaminants. WHO Food Additives Series, No. 71, 2015.

414

WHO Food Additives Series No. 71-S1, 2016

Annex 2 Abbreviations used in the monographs ADHD attention deficit hyperactivity disorder AH aniline hydroxylase; ancestral haplotype AHH acetanilide-4-hydroxylase AhR aryl hydrocarbon receptor AMICS Asthma Multicentre Infants Cohort Study ANOVA analysis of variance APDM aminopyrine N-demethylase ATHON Assessing the Toxicity and Hazard of Non-dioxin-like PCBs Present in Food ATSDR Agency for Toxic Substances and Disease Registry (USA) BAEP brainstem auditory evoked potential BDE brominated diphenyl ether BDE-47 2,2′,4,4′-tetrabromodiphenyl ether BF breastfed as infants BMD benchmark dose BMDL lower 95% confidence limit on the benchmark dose BMI body mass index BSID Bayley Scales of Infant Development bw body weight CAR constitutive androstane receptor CARDIA Coronary Artery Risk Development in Young Adults CDC Centers for Disease Control and Prevention (USA) cGMP cyclic guanosine monophosphate CED critical effect dose CEDL lower bound of the confidence interval on the critical effect dose CES critical effect size CI confidence interval CIFOCOss FAO/WHO Chronic Individual Food Consumption database – summary statistics CPT Continuous Performance Test CYP cytochrome P450 DAT dopamine transporter DBP diastolic blood pressure DDE 1,1-dichloro-2,2-bis(p-chlorophenyl) ethylene DDT 2,2-bis(p-chlorophenyl)-1,1,1-trichloroethane 415

Safety evaluation of certain food additives and contaminants Eightieth JECFA

DEN diethylnitrosamine DIPP Finnish Type I Diabetes Prediction and Prevention Nutrition Study DL-PCB dioxin-like polychlorinated biphenyl DMSO dimethylsulfoxide DNA deoxyribonucleic acid ECD electron capture detection/detector E-CPT Extended Continuous Performance Test EFSA European Food Safety Authority ELISA enzyme-linked immunosorbent assay ENRIECO Environmental Health Risks in European Birth Cohorts EPA Environmental Protection Agency (USA) EROD ethoxyresorufin-O-deethylase F female F0 parental generation F1 first filial generation FAO Food and Agriculture Organization of the United Nations FISH fluorescence in situ hybridization FSH follicle stimulating hormone GC gas chromatography GD gestation day GEMS/Food Global Environment Monitoring System – Food Contamination Monitoring and Assessment Programme GJIC gap junctional intercellular communication GR glucocorticoid receptor GST glutathione S-transferase HBCD hexabromocyclododecane HCB hexachlorobenzene HDL high-density lipoprotein HR hazard ratio HRGC-MS high-resolution gas chromatography with mass spectrometry HSD hydroxysteroid dehydrogenase IARC International Agency for Research on Cancer IC50 median inhibitory concentration ICAM-1 intercellular adhesion molecule-1 ICD-8 International Classification of Diseases, Eighth Revision ICD-9-CM International Classification of Diseases, Ninth Revision, Clinical Modification IGF-1 insulin-like growth factor-1 IL interleukin IM-GSM grey-scale median of the intima-media complex IMT intima-media thickness

416

WHO Food Additives Series No. 71-S1, 2016

Supplement 1: Non-dioxin-like polychlorinated biphenyls

INCA-2 INMA IQ IU IUPAC JECFA K-ABC LB LD50 LDL LH LOAEL LOD LOQ M MB MCP-1 MDI MMP-7 MNC MOE MR mRNA MSCA NADPH NBF ND NDL-PCB nes NES2 NHANES NHL NMDA NOAEL NOS NTD NTP OECD OR P5 P90

Etude individuelle nationale des consommations alimentaires 2 INfancia y Medio Ambiente (Child and Environment) intelligence quotient International Units International Union of Pure and Applied Chemistry Joint FAO/WHO Expert Committee on Food Additives Kaufman Assessment Battery for Children lower bound median lethal dose low-density lipoprotein luteinizing hormone lowest-observed-adverse-effect level limit of detection limit of quantification male middle bound monocyte chemoattractant protein-1 mental development index matrix metalloproteinase-7 magnocellular neuroendocrine cell margin of exposure muscarinic receptor messenger ribonucleic acid McCarthy Scales of Children’s Abilities nicotinamide adenine dinucleotide phosphate (reduced) non-breastfed as infants non-detects; not detected non-dioxin-like polychlorinated biphenyl not elsewhere specified Neurobehavioural Evaluation System 2 National Health and Nutrition Examination Survey (USA) non-Hodgkin lymphoma N-methyl-D-aspartate no-observed-adverse-effect level nitric oxide synthase neural tube defect National Toxicology Program (USA) Organisation for Economic Co-operation and Development odds ratio 5th percentile 90th percentile 417

Safety evaluation of certain food additives and contaminants Eightieth JECFA

P95 P99 PBB PBDE PBK PBPK PCB PCDD PCDF PDI PFC PIVUS PKC PND POPUP PROD PXR POP PPAR QSAR ROCK RR RyR SAR SBP SD SHBG SIR SMR SON SRBC SRS STEM STRIP SULT T3 T4 TCDD TDI TEF TEQ

95th percentile 99th percentile polybrominated biphenyl polybrominated diphenyl ether physiologically based kinetic physiologically based pharmacokinetic polychlorinated biphenyl polychlorinated dibenzo-p-dioxin polychlorinated dibenzofuran psychomotor development index plaque-forming cell Prospective Investigation of the Vasculature in Uppsala Seniors protein kinase C postnatal day Persistent Organic Pollutants in Uppsala Primiparas pentoxyresorufin-O-deethylase pregnane X receptor persistent organic pollutant peroxisome proliferator–activated receptor quantitative structure–activity relationship Rho-associated kinase relative risk ryanodine receptor structure–activity relationship; Special Administrative Region systolic blood pressure standard deviation sex hormone binding globulin standardized incidence rate standardized mortality rate supraoptic nucleus sheep red blood cell Social Responsiveness Scale STatistical Exposure Model Turku Coronary Risk Factor Intervention Project for Children sulfotransferase triiodothyronine thyroxine 2,3,7,8-tetrachlorodibenzo-p-dioxin tolerable daily intake toxic equivalency factor toxic equivalents

418

WHO Food Additives Series No. 71-S1, 2016

Supplement 1: Non-dioxin-like polychlorinated biphenyls

TSH UB UDS UGT UNEP USA USEPA VCAM-1 VEP VMAT WAM WGM WHO w/w ww

thyroid stimulating hormone upper bound unscheduled DNA synthesis uridine diphosphate-glucuronosyltransferase United Nations Environment Programme United States of America United States Environmental Protection Agency vascular cell adhesion molecule-1 visual evoked potential vesicular monoamine transporter weighted arithmetic mean weighted geometric mean World Health Organization weight per weight wet weight

419

Annex 3 Joint FAO/WHO Expert Committee on Food Additives1 Rome, 16–25 June 2015 Members Dr D. Benford, Risk Assessment Unit, Science Evidence and Research Division, Food Standards Agency, London, England, United Kingdom Dr M. DiNovi, Office of Food Additive Safety, Center for Food Safety and Applied Nutrition, United States Food and Drug Administration, College Park, MD, USA Dr M. Feeley, Bureau of Chemical Safety, Food Directorate, Health Canada, Ottawa, Ontario, Canada Dr D. Folmer, Office of Food Additive Safety, Center for Food Safety and Applied Nutrition, United States Food and Drug Administration, College Park, MD, USA Dr Y. Kawamura, Division of Food Additives, National Institute of Health Sciences, Tokyo, Japan Dr Madduri Veerabhadra Rao, Quality Control Department, Department of the President’s Affairs, Al Ain, United Arab Emirates Dr A. Mattia, Office of Food Additive Safety, Center for Food Safety and Applied Nutrition, United States Food and Drug Administration, College Park, MD, USA (Vice-Chairperson) Mrs I. Meyland, Birkerød, Denmark (Chairperson) Dr U. Mueller, Food Standards Australia New Zealand, Barton, ACT, Australia (Joint Rapporteur) Dr J. Schlatter, Zurich, Switzerland Mrs H. Wallin, Helsinki, Finland (Joint Rapporteur) Professor G.M. Williams, Department of Pathology, New York Medical College, Valhalla, NY, USA

Secretariat Dr A. Agudo, Unit of Nutrition and Cancer, Cancer Epidemiology Research Program, Institut Català d’Oncologia, L’Hospitalet de Llobregat, Spain (WHO Expert) Dr B. Amzal*, LASER, United Kingdom Headquarters, London, England, United Kingdom (WHO Expert) 1

Participants marked with an asterisk (*) did not attend the entire meeting.

420

Supplement 1: Non-dioxin-like polychlorinated biphenyls

Dr J.H. Andersen, National Food Institute, Technical University of Denmark, Søborg, Denmark (WHO Expert) Dr S. Barlow, Brighton, East Sussex, England, United Kingdom (WHO Expert) Ms A. Bruno, Joint FAO/WHO Food Standards Programme, Food and Agriculture Organization of the United Nations, Rome, Italy (Codex Secretariat) Ms A.S. Bulder, Centre for Nutrition, Prevention and Health Services, National Institute for Public Health and the Environment (RIVM), Bilthoven, the Netherlands (WHO Expert) Dr S. Cahill, Food Safety and Quality Unit, Agriculture and Consumer Protection Department, Food and Agriculture Organization of the United Nations, Rome, Italy (FAO Secretariat) Dr R. Cantrill, AOCS, Urbana, IL, USA (FAO Expert) Mr P. Cressey, Risk and Response Group, ESR (Institute of Environmental Science and Research Ltd), Christchurch, New Zealand (FAO Expert) Dr M. De Nijs, RIKILT Wageningen UR, Wageningen, the Netherlands (FAO Expert) Dr E. Dessipri, General Chemical State Laboratory, Athens, Greece (FAO Expert) Dr J.A. Edgar, CSIRO Food and Nutritional Sciences, North Ryde, NSW, Australia (FAO Expert) Dr V. Fattori, Agriculture and Consumer Protection Department, Food and Agriculture Organization of the United Nations, Rome, Italy (FAO Joint Secretary) Professor H. Håkansson, Institute of Environmental Medicine, Karolinska Institutet, Stockholm, Sweden (WHO Expert) Ms T. Hambridge, Food Data Analysis Section, Food Standards Australia New Zealand, Barton, ACT, Australia (WHO Expert) Dr S.M.F. Jeurissen, Centre for Nutrition, Prevention and Health Services, National Institute for Public Health and the Environment (RIVM), Bilthoven, the Netherlands (WHO Expert) Professor F. Kayama, Department of Environmental & Preventive Medicine, School of Medicine, Jichi Medical University, Yakushiji, Shimotsuke-shi, Tochigi-ken, Japan (WHO Expert) Mr J. Kim, Department of Food Safety and Zoonoses, World Health Organization, Geneva, Switzerland (WHO Secretariat) Dr J.C. Leblanc, Food Safety and Quality Unit, Agriculture and Consumer Protection Department, Food and Agriculture Organization of the United Nations, Rome, Italy (FAO Secretariat) Dr T. Rawn, Food Research Division, Health Canada, Ottawa, Ontario, Canada (FAO Expert) Dr K. Schneider, Forschungs- und Beratungsinstitut Gefahrstoffe GmbH (FoBiG), Freiburg, Germany (WHO Expert) 421

Safety evaluation of certain food additives and contaminants Eightieth JECFA

Ms M. Sheffer, Orleans, Ontario, Canada (WHO Technical Editor) Dr J.R. Srinivasan, Office of Food Additive Safety, Center for Food Safety and Applied Nutrition, United States Food and Drug Administration, College Park, MD, USA (FAO Expert) Professor I. Stankovic, Department of Bromatology, Faculty of Pharmacy, University of Belgrade, Belgrade, Serbia (FAO Expert) Dr A. Tritscher, Department of Food Safety and Zoonoses, World Health Organization, Geneva, Switzerland (WHO Joint Secretary) Dr T. Umemura, Division of Pathology, Biological Safety Research Center, National Institute of Health Sciences, Tokyo, Japan (WHO Expert) Professor Dr M. Van den Berg*, Toxicology and Veterinary Pharmacology Division, Institute for Risk Assessment Sciences, Utrecht University, Utrecht, the Netherlands (WHO Expert) Dr Y. Wu, China National Center for Food Safety Risk Assessment, Beijing, China (FAO Expert) Dr X. Yang, Guangdong Provincial Center for Disease Control and Prevention, Guangzhou, Guangdong Province, China (WHO Expert) Dr H.J. Yoon, Food Contaminants Division, Food Safety Evaluation Department, Ministry of Food and Drug Safety, Cheongwon-gun, Chungcheongbuk-do, Republic of Korea (FAO Expert) Dr Y. Zang, Office of Food Additive Safety, Center for Food Safety and Applied Nutrition, United States Food and Drug Administration, College Park, MD, USA (WHO Expert)

422

WHO Food Additives Series No. 71-S1, 2016

This volume contains a monograph prepared at the eightieth meeting of the Joint FAO/WHO Expert Committee on Food Additives (JECFA), which met in Rome, Italy, from 16 to 25 June 2015. The toxicological and dietary exposure monograph in this volume summarizes the safety and dietary exposure data on a contaminant group (non-dioxin-like polychlorinated biphenyls) discussed at the eightieth meeting. Monographs on seven food additives discussed at that meeting have been previously published in the WHO Food Additives series (FAS 71), and a monograph on a second contaminant group (pyrrolizidine alkaloids) will be published as a separate supplement in the WHO Food Additives series. This volume and others in the WHO Food Additives series contain information that is useful to those who produce and use food additives and veterinary drugs and those involved with controlling contaminants in food, government and food regulatory officers, industrial testing laboratories, toxicological laboratories and universities.

ISBN 978 92 4 166171 3