Transfer of DDT used in malaria control to infants via breast milk.

Transfer of DDT used in malaria control to infants via breast milk H. Bouwman,' P.J. Becker,2 R.M. Cooppan,3 & A.J. Reinecke4 The transfer of p,p'-DDT (1,1, 1-trichloro-2,2-bis(4-chlorophenyl)ethane) and its metabolites to infants via breast-feeding was studied in an area of KwaZulu, South Africa, where DDT is used to interrupt malaria transmission. Samples of whole blood were collected from 23 infants, together with samples of breast milk from their respective mothers. The mean XDDT (total DDT) in the whole blood was 127.03 jig. 1 and that in the breast milk, 15.06 mg.kg -1 (milk fat). The % DDT (% DDT of XDDT) was significantly higher in the infant blood than in the breast milk (P <0.05). A multiplicative regression analysis indicated that XDDT increased significantly (P <0.01) in infant whole blood with infant age. Multiple regression showed that 70.0% of the variation in X:DDT was due to the variation in parity of the mother, age of the infant, and the I0DDT in breast milk. These variables accounted also for 76.3% of the variation in p,p'-DDE but only for 38.2% of that in p,p'-DDT. Organochlorines were therefore largely transferred to the infant from the mother, with DDT in the envir- onment playing a secondary role. Introduction p,p '-DDT (1,1,1 -trichloro-2,2-bis(4-chlorophen- yl)ethane) is used to interrupt malaria transmis- sion in the malaria-endemic areas of KwaZulu, South Africa (1). For this purpose, it is applied annually on the inner walls of all dwellings at a coverage of 2g /m2. The levels of DDT and its metabolites in the breast milk of lactating mothers significantly ex- ceed (2) the estimated acceptable daily intake (ADI) for total DDT (XDDT) of 0.02 mg.kg-'. day-' (3). The mean XDDT (p,p'-DDT + p,p'- DDE + p,p '-DDD)a b in the breast milk of 129 mothers was 15.83 mg.kg-' (milk fat: DDT = 6.77 mg.kg-'; DDD = 0.3 mg.kg-'; and DDE = 6.2 mg.kg-'). In a survey on human exposure to selected orga- nochlorine compounds, Slorach & Vaz reported the I Senior Lecturer, Department of Zoology, Potchefstroom University for Christian Higher Education, Private Bag X6001, 2520 Potchefstroom, South Africa. Requests for reprints should be sent to this author. 2 Specialist Scientist, Institute for Biostatistics, Medical Re- search Council, 0001 Pretoria, South Africa. 3Chief Senior Medical Research Officer, Research Institute for Diseases in a Tropical Environment, Medical Research Council, Congella, South Africa. 4Professor, Department of Zoology, Potchefstroom University for Christian Higher Education Potchefstroom, South Africa. a DDE = 1,1-dichloro-2,2-bis(4-chlorophenyl)ethylene. b DDD = 1,1-dichloro-2,2-bis(4-chlorophenyl)ethane. Reprint No. 5272 mean national levels of DDT and DDE in milk from various countries (4). The highest median levels in milk fat were from China (DDE = 4.4mg.kg-1 and DDT = 1.8 mg.kg-') and India (DDE = 4.8 mg.kg-' and DDT = 4.4 mg.kg-'), where DDT is still used in agriculture and for vector control. The samples were taken from mothers who lived in urban areas and who were nursing their first or second child. First-boms in KwaZulu were exposed to signifi- cantly higher levels of XDDT (24.82 mg.kg-' in milk fat) than were subsequent infants (12.21 mg.kg-' in milk fat) (5). Because of the practice of breast- feeding for as long as possible (2 years or longer), large amounts of DDT can be transferred to the in- fants. The proportion of DDT (% DDT of XDDT) in breast milk, on the other hand, increased significant- ly (P <0.05) with parity, from a mean of 34.5% for primiparous mothers to means of 46.4%, 50.9%, 50.6%, and 47.3% for mothers who were breast-feed- ing their second, third, fourth, and fifth to tenth child combined, respectively. It was hypothesized that this increase was due to the uptake and faster elimination of DDT via milk rather than to the uptake and endo- genous formation of DDE in the mother (5). Infants cannot be considered to be recipients of xenobiotics in the same way as older children or adults (6, 7), because the rate at which infants accu- mulate DDT together with associated factors (such as breast-feeding, metabolism and immunological status) change with age. Rogan et al., in the only study that has examined the effects of DDE in breast milk on infants, found a significant increase in the Bulletin of the World Health Organization, 70(2): 241-250 (1992) © World Health Organization 1992 241 H. Bouwman et al. number of infants showing hyporeflexia that was associated with an increase in the DDE concentration in breast milk (8). These effects became apparent at DDE levels of >4 mg.kg-' in milk fat. There is there- fore still a lack of understanding about the levels and effects of DDT on infants. Since in developing coun- tries mothers and infants are exposed to DDT via malaria control and agricultural activities, the levels and associated factors need to be determined to assess risk. The aims of the present study were there- fore as follows: to determine the levels of DDT and its metabolites in blood from infants who had been exposed to high levels of DDT in breast milk; and to design statistical models from the results. Materials and methods The study design and the methods used were appro- ved by the Ethical Committee of the Research Institute for Diseases in a Tropical Environment. For the purpose of the study infants were defined as chil- dren aged 0-2 years. Corresponding milk and blood samples were collected at the Mseleni Hospital, northem KwaZulu, during 1987. The mothers who attend routine well-baby clinics at Mseleni Hospital live in an area where DDT is used for malaria control. Mseleni is situated on the shore of Lake Sibaya and is removed from the major routes of population migration. During a pilot study it was found that migrants from Mozambique, where hardly any malaria control is practised, had very low levels of DDT (2, 5). Dwellings in KwaZulu are usually constructed of mud, branches, and thatch. Home- steads consist of three to seven such structures and house 4-22 people. The major source of income is from migrant labour in mines or on farms. The diet consists of staples such as maize and rice together with fish and meat from goats, chickens or cattle. This is supplemented by collecting wild fruits, nuts, and roots. Northem KwaZulu is a malaria-endemic area where DDT has been applied only for malaria control. No agricultural use of this pesticide has been recorded, and DDT has been banned for this purpose in South Africa since 1976 (2). Sample collection and analysis All mothers who attended the hospital on a routine baby-clinic day were approached. Questionnaires were filled out for all mothers who participated, and information on maternal and infant age, parity, dura- tion of residence in DDT-sprayed dwellings, and maternal occupation was obtained. Mothers were supplied with a 100-ml glass beaker and asked to express manually about 10ml of milk, which was transferred immediately to 10-ml blood-collecting tubes and capped. The tubes were stored on ice until they were frozen on the same day. Samples of blood were only taken from infants whose mothers produced enough milk. The big toe of the infant was cleaned with Hibitane and allowed to dry; a toe prick with a sterile lancet was then made and a capillary tubec (44.7 gl) was tilted at a slightly upward angle and filled by capillary action (a downward angle resulted in air bubbles being drawn in). The capillary tube was then placed in a 3-ml test-tube containing 2 ml of doubly distilled water. The test-tube was capped and shaken vigor- ously to dislodge and lyse the blood from the capilla- ry tube. The test-tubes were also stored on ice until they were frozen on the same day. The milk and blood samples were then analysed as described pre- viously (9, 10). The results are reported on a milk-fat basis to allow comparison with those from other studies. Results The socioeconomic circumstances of the study group were such that no activity (e.g., employment in a pesticide factory) or location (e.g., near an intensive farm that used DDT) could have caused additional exposure to DDT, other than from background levels or from malaria control activities. There were also no obvious factors that would have resulted in a major variation in exposure to DDT for either the mother or infant. Only three of the mothers who were approached refused to allow a blood sample to be taken from their infants. A summary of the statistics on the cha- racteristics of the 23 infants and mothers is presented in Tables 1 and 2. Matemal age was difficult to esta- blish, as in many cases it was not known; an estima- ted age was therefore used. The same was true for the age of the infants, but here more accurate in- ference was possible. The analytical technique employed was not sen- sitive enough to detect DDD in the infant blood. The proportion of DDD in the breast milk, however, was less than 3% and its contribution could therefore be ignored. DDE and DDT were detected in all blood samples (Table 1); marked variations in the levels of XDDT, DDT and DDE resulted in large standard deviations. The maximum XDDT concentration was 316.5 jg.1-'. Infant age and parity did not differ c Coulter Pipets, Coulter Diagnostics, Hialeah, FL, USA. 242 WHO Bulletin OMS. Vol 70 1992 Transfer of DDT to infants via breast milk Table 1: Summary of statistics for all variables relating to the infants tested In the study (in each column the data are for 23 Infants) Concentration in whole blood (gg.l-1): Infant age DDE DDT XDDT %DDTa (days) Parity Mean 67.12 (52.92)b 59.91 (30.17) 127.03 (70.84) 50.88 (16.16) 252.4 (144.99) 3.17 (1.85) Median 49.7 58.10 114.2 49.37 240 3 Minimum 5.6 16.9 29.4 22.15 5 1 Maximum 218.4 135.3 316.5 80.95 540 7 a % DDT = % DDT of EDDT. b Figures in parentheses are the standard deviations. Table 2: Summary of statistics for all variables relating to the mothers of the infants tested in the study (in each column the data are for 23 mothers) Concentration in Concentration whole milk (gg.1-1): in milk fat (mg.kg-1) Maternal Milk fat DDE DDD DDT XDDT DDE DDD DDT XDDT %DDT age (years) (%) Mean 297.0 13.60 222.9 535.1 8.87 0.44 6.74 15.06 45.40 25.83 3.50 (345/1)a (10.34) (184.7) (519.9) (7.81) (0.36) (4.31) (11.08) (5.97) (5.53) (1.58) Median 201.5 9.89 200.9 427.1 7 0.34 5.9 13.1 45.06 25 3.15 Minimum 67.1 1.05 35.6 144.1 1.9 0.02 1.22 3.1 35.24 17 1.43 Maxi- 1758.8 43.7 967.9 2743.4 36.9 1.6 20.3 57.6 57.28 35 6.29 mum a Figures in parentheses are percentages. significantly (P >0.05) from those found in three pre- vious surveys that examined serial changes in the levels of DDT in breast milk (3). DDE, DDD, and DDT were detected in all milk samples (Table 2); marked variations were again in evidence. The maximum value of 57.6 mg.kg-1 for XDDT in milk fat should be noted. The concentrations (but not the % DDT) were skewed, and logarithmic transformations were used in subsequent analyses. A significant multiplicative relationship (ln (% DDT in infant blood) = -4.1 + 2.lx(In %DDT in breast milk)), coefficient of determination = 60.6%, was found between the % DDT in infant blood and the % DDT in the corresponding sample of breast milk (P <0.001; Fig. 1). The % DDT in milk (correc- ted for DDD by subtraction) and the % DDT in infant blood were compared using a paired Student's t-test. The proportion of DDT was 5.35% higher in infant blood than in breast milk, which was sig- nificant (P <0.05). The coefficient of determination from the linear regression of XDDT on infant age (days) was 30.3% (XDDT = 59.18 + 0.268x(infant age); P <0.01). The relevant statistics for the multiplicative regres- sion, which exhibited an improved coefficient of determination of 44.95%, are shown in Fig. 2. Multiple regression with stepwise variable selec- tion was used to model the levels of XDDT, DDT and DDE in infant blood. Variables such as parity, infant age, matemal age and percentage milk fat were introduced. The best model for describing XDDT is illustrated in Fig. 3. Matemal age in combi- nation with any of the other variables reduced the accuracy of the model. The best models for DDE and DDT are shown in Fig. 4 and 5, respectively. Discussion The levels of DDT and its metabolites in whole blood differ from those in plasma or serum, in that 75% of the XDDT occurs in the plasma fraction (11). Morgan et al. reported that less than 18% of the DDT and DDE was associated with red blood cells (12). If we assume the same relationship for infant blood and a haematocrit of 50%, the levels found should at least be doubled for comparison with WHO Bulletin OMS. Vol 70 1992 243 H. Bouwman et al. Fig. 1. Multiplicative regression of percentage DDT in samples of infant's whole blood against percentage DDT in the corresponding samples of mother's breast milk. CD = coefficient of determination. . P < 0.001 CD = 60.6% InY = 4.1 + 2.1 (InX) 0 * 0 0 0 . 0 35 40 45 %DDT (milkfat) 50 55 010iR alazi 60 Fig. 2. Multiplicative regression of XDDT in samples of infant's whole blood against infant age. CD = coefficient of determination. 100 200 300 Infant age (days) 400 500 0 600 WHO Bulletin OMS. Vol 70 1992 * 0 0 0 80 70 _ g0 60 -0 0 -r 50 O 40 0 30 _ 20 30 I P < 0.001 - CD = 44.95% - InY = 2.73 + 0.37(InX) 0~~~~~~ _ * . _~~~~~~~ 400 _300 z-,CD V 0 0 B c 200 0H- 100 0 0 - . . . 244 Transfer of DDT to infants via breast milk Fig. 3. Multiple regression model of XDDT in samples of infant's whole blood. PAR = parity; IA = infant age; CD = coefficient of determination. 0 ~0 a) a) 0 c to 3.3 3.8 4.3 4.8 5.3 5.8 In (predicted l: DDT) Fig. 4. Multiple regression model of DDE in samples of infant's whole blood. PAR = parity; IA = infant age; CD = coef- ficient of determination. 5.7 InY = 2.24- 0.69(In PAR) + 0.33 (In IA) + 0.34 (In DDE) 'O _CD (adjusted) =76.3%0 04.7 D| X ~~~~* *-00 .o2.7 - _1 -.7 1.7 L 1.7 2.7 3.7 4.7 5.7 In (predicted DDE) WHO Bulletin OMS. Vol 70 1992 245 H. Bouwman et al. Fig. 5. Multiple regression model of DDT in samples of infant's whole blood. PAR = parity; IA = infant age; CD = coef- ficient of determination. 5.2 I 4.8 InY = 2.91 0.31 (In PAR) + 0.17 (In IA) + 0.25 (In DDT) CD (adjusted) = 36.2% 0~~~~~~~~~~ 0) 4.0 c:3.6 / 3.2 2.8l , 2.8 3.2 3.6 4.0 4.4 4.8 5.2 In (predicted DDT) serum data. This gives a mean serum level for the infants of 254 jig. 1` for YDDT, which exceeds that found for the general population in the study area (140.9 jig.l1-; n = 71) (13). The mean serum level for 3-9-year-olds (n = 24) in this population was 168.6 ,ug.l- for XDDT (13). Kreiss et al. found a mean XDDT level in serum of 159.4 jig.l- for a population with exceptional exposure to DDT (through consumption of contaminated fish and water) (14); the serum level for 1-9-month-olds in the same population was ±30 gg.-l for XDDT (deter- mined graphically), and increased with age. Breast- fed children had higher levels of XDDT than those who were not breast-fed (14), but the determination was carried out some years after breast-feeding had stopped. After controlling for consumption of fish and age, the relation to duration of breast-feeding was no longer significant. Breast milk and the environment (including other food and hand-mouth activity) were conside- red for the purpose of the analysis as the two major routes of uptake of DDT by the infants (as compared to malaria control as a source). Bouwman et al. have reported the levels of DDT and its metabolites in fish from the Pongolo Flood Plain, which is about 30 km from Mseleni, and where the same conditions pre- vail; the maximum intake from eating tigerfish (the most contaminated fish), assuming a mean body weight of 60 kg and a daily fish consumption of 200 g, was 0.98 jig.kg- .day-1 (15). For the data reported by Kreiss et al. (14), assuming the same level of consumption and body weight, we calculated the maximum intake to be 75.3 jig.kg- .day- . Intake of DDT from fish is therefore not responsible for the high levels found in the present and other investiga- tions in the study area (2, 5, 13). If the environment were a major source of DDT in the study area, XDDT would increase with infant age. The multiplicative model showed that 44.95% of the variation in XDDT was accounted for by infant age; however, the difference in the levels of XDDT in milk from primiparous mothers relative to that from mothers with more children permitted identifi- cation of two extra variables: parity and concentra- tion of XDDT in milk. Parity and infant age are not linearly related, and the regression of the levels of EDDT in infant blood with either of these variables provided no information about the duration of expo- sure. Multiple regression with stepwise selection of variables (Fig. 3), which allowed the introduction of WHO Bulletin OMS. Vol 70 1992246 Transfer of DDT to infants via breast milk time as a variable, indicated that 60.5% of the varia- tion was accounted for by parity, 8% by infant age, and only 1.5% by XDDT in the milk. Maternal age and percentage milk fat did not contribute significantly to the model and were exclu- ded. The non-significance of maternal age, which varied linearly with WDDT in breast milk, should be noted. The percentage milk fat did not contribute probably because no relationship existed between the levels of XDDT, DDT or DDE and this parameter (which, in turn, was also not significantly related to any other parameter). Also, maternal age, although related to parity, was a worse indicator of the level of contamination of breast milk for this group (5). Unexpectedly, the level of XDDT in breast milk had very little influence, and its contribution to the model containing the other two variables was not significant (P = 0.1747). It was nevertheless included because it improved the model (increasing the coef- ficient of determination from 68.5% to 70%) and re- duced the mean square error by 30%. Also, for the model that accounted for DDE (Fig. 4), the con- centration variable was significant. Two considerations could explain the lack of influence of XDDT in breast milk - parity (as a major determinant of concentration) and time. The amount of DDT taken up is determined by its concentration in milk and the length of the lactation period (concentration x time = exposure). If there is little variation in the amount of milk consumed and a constant concentration of XDDT in the milk, the quantity of DDT compounds taken in with the milk will only be determined by time. The uptake from other sources (environment) will also be governed by time. Infant age is therefore a direct measure of exposure to DDT; and both parity and infant age would then be the best predictors of such exposure. This holds only if the exposure of the mothers was relatively uniform. In this case, the annual malaria control operations with DDT could be considered as constant, with no other environmental source of this substance. The time of application of DDT was about the same for all cases studied (1). Furthermore, the level of DDT in breast milk changes after appli- cation of the pesticide (2). In cases of different expo- sure of the mother, the concentration of DDT in breast milk would conceivably have more influence on a predictive model. In the model infant age alone accounted for 42.3% of the variation of WDDT, while parity alone accounted for 60.5%. Parity together with infant age accounted for 68.5% of the variation. It seems that these two variables explain at least part (and not necessarily mutually exclusive parts) of the same variation; this would imply a relationship between them. Infant age, however, reveals nothing about parity, and introducing mathematical interaction be- tween these two variables did not improve the mod- el. Although there might be confounding influences (e.g., the rate of weight gain of the infant, the length of lactation period per infant, and the nonlactating interval between infants), the relationship between parity and XDDT in the infant (i.e., the relative contributions of parity and infant age in the model) implies that breast milk rather than the environment was the major source of XDDT in the study infants. Further support for this conclusion is provided by the fact that infant blood had 5% more DDT than the breast milk. Although metabolism and subse- quent excretion of DDE by the infant or preferential uptake of DDT from the intestinal to the blood com- partment are possible, there is a more logical expla- nation. Malaria control involves the spraying of DDT (which contains <4% of DDE); uptake of DDT from this source (possibly from indoor air or by contact) by the infant could account for the difference. The magnitude of the difference in the % DDT in milk or infant blood could therefore indicate the relative importance of uptake from milk and the environ- ment. This serves as evidence for a secondary route of uptake, other than from breast milk. Even more support for the above argument is provided by the differences in the models for DDE (Fig. 4) and DDT (Fig. 5). The model for DDE accounted for 76.3% of its variation (P <0.001); parity for 62.7%, and infant age for 7.6%. The concentration of DDE in milk contributed 6%. All the variables were significant. In contrast, only 38.2% of the variation in DDT was accounted for by its model (P = 0.0067). Parity accounted for 26.5%; infant age for 6%, and the DDT concentration in milk for 5.7%. All three variables were not signifi- cant. Also, in the study that determined the levels of DDT in serum of the general population, children aged 3-9 years had significantly higher (P <0.05) levels of XDDT (168.6 Jg.l-) than adults aged 20-29 years (60.5 ,ug.l-1) (13). The initially high levels of serum XDDT were ascribed to exposure to contaminated breast milk, and the subsequent reduc- tion to dilution by growth and/or elimination from the body. A regression analysis suggested pharmaco- dynamic differences between DDE and DDT for the younger (3-29 months), and older (30 to .60 months) groups, which are consistent with elimina- tion as a probable, but not the only mechanism of reduction. This in turn indicates that the initial up- take which produced the elevated levels of DDT in the infant must have come from breast milk, since uptake from the environment will remain constant after breast-feeding has stopped and could even in- crease upon consumption of other food such as fish. WHO Bulletin OMS. Vol 70 1992 247 H. Bouwman et al. If it is assumed that the infant's only significant source of DDE is breast milk, the predictive value of a model that includes variables which influence exposure will be higher. The DDT in the infant can be taken to be derived from two routes: breast milk and the environment. A model that includes infant age as the only variable associated with the environ- ment will therefore have less predictive value. Variable contamination of food and air (both in- and outdoors) with DDT, as well as the duration of expo- sure, are some of the uncertainties associated with the model. Few studies have examined the changes in the concentration of organochlorine compounds in infants as a function of breast-feeding. The levels of polychlorinated biphenyls (PCBs) in samples of blood from a group of infants who were breast-fed increased from 1.1 tg.l1- at birth to 3.6 gg.[-' three months later, while for a group who were not breast- fed the corresponding increase was from 1.1 jg.l-1 to 1.6,g.1- (16). Engst et al. reported a slight increase in the level of DDT in body fat from 15 mg.kg-1 at birth to 16.4 mg.kg-1 thirteen days later, and attribu- ted this to a modest weight loss in the infants (17). Eckenhausen et al. found an increase in the level of DDE in blood from 3 jg.l-1 at 2 weeks of age to 3.9 gg.l' at 3 months postpartum. A slight increase in the levels of PCBs in infant blood due to breast- feeding was also found (18). The group of infants who were bottle-fed in the same study exhibited hardly any change in this respect (18). A reverse trend, i.e., reduction in the level of XDDT in fat samples by almost 50%, has also been reported by Niessen et al. for infants up to 6 months after birth (19). There have been two reports of the change in the levels of DDT in animals. Ando found that the concentrations of DDT in whole suckling rats that had been exposed to DDT via milk increased rapidly from birth and followed a sigmoid curve (20). Tomatis et al. followed a different experimental desi- gn and analysed mouse fetuses bom to dams that had been exposed to different concentrations of DDT (21). The XDDT concentration in the fetuses was directly related to matemal exposure (dosage) (21). Only one study was found that predicted age- related DDT levels in infants from the levels in milk. In this study Mes et al. plotted lactation time against theoretically estimated accumulated levels of DDT and DDE, derived from the values in milk and in infant body fat (22). A sequential analysis of breast milk from 16 women over a 3-month period was used. At 1, 2, 4, and 14 weeks postpartum the res- pective DDE levels in infant body were estimated graphically to be 0.15, 0.40, 1.0 and 1.85 mg.kg-1; for DDT, the respective values were 0.02, 0.07, 0.12, and 0.25 mg.kg-'. The model assumed a linear increase in the percentage body fat, with no provi- sion for adsorption by other organs and no correction for excretion (22). In this instance, the levels in milk were not correlated with parity. Mes et al. predicted that the infant body burden would reach adult levels within 3 months of breast-feeding. In contrast, the results of the present study show that the infant body burden continued to increase at higher levels of DDT in breast milk. Wickizer et al. estimated the body burden over a 12-week period for infants exposed to 1.5 mg.kg-1 of PCB in milk fat (23). The burdens at 1, 2, 4, 6, 8, 10, and 12 weeks, respectively, were 0.25, 0.45, 0.62, 0.75, 0.90, 1.2, and 1.4 mg.kg-1 total body weight. Whether the levels of DDT found in infants constitute a health hazard has not yet been deter- mined. Rogan et al., in a study that examined the effects of DDE in breast milk on infants (8), found that the proportion of infants (about 2 weeks post- partum) who exhibited more than four (out of 20) delayed or not-elicited responses increased with the concentrations of DDE in milk. This trend became apparent at 4 mg.kg-' milk fat. The mean level of XDDT (containing DDE and the more toxic DDT) was higher. It is therefore possible that infants thus exposed are at risk. The level of DDT in blood (after conversion to serum) was higher than for the general population. Many workers have expressed concern about the possible higher susceptibility of various body systems, e.g., the nervous, immunological and renal systems of infants, which reach maturity after birth (8, 17, 24-27). Our study shows that the levels of DDT and DDE in infants increase with age. This increase could be modelled using parity, infant age, and concentration of the contaminant in the breast milk. For the case of malaria control as a single source of DDT, more than one route of uptake seemed to play a role. However, the major contribution was con- sidered to be uptake from breast-feeding. Acknowledgements This work was supported by the Medical Research Council and is published with their permission. We thank Professor C.H.J. Schutte and Dr P.W. le R. Murray for their assistance and encouragement; Dr V. Friendlund, Dr G.M. Short and Mr S. Ngxongo, KwaZulu Department of Health and Welfare, and Mr M.J. Botha, South African Department of Health and Population Development, for their invaluable assistance; Mrs J. Nkomokazi and Mr G. Ngcobo for their excellent technical assistance; and Mrs B. Pflugler, Mr R. Byng and Mrs B. Bouwman for help in preparing the manuscript. WHO Bulletin OMS. Vol 70 1992248 Transfer of DDT to infants via breast milk Resume Transfert aux nourrissons, via le lait maternel, du DDT utilis6 en lutte antipaludique Dans une partie du KwaZulu, en Afrique du Sud, ou on utilise le DDT pour stopper la transmission du paludisme, on a etudie le transfert aux nour- rissons du p,p'-DDT (1,1,1-trichloro-2,2-bis(4- chlorophenyl)ethane) et de ses metabolites par le biais de I'allaitement au sein. Dans cette region, le DDT est applique une fois par an sur les murs interieurs de toutes les habitations au taux de 2 gIm2. Des echantillons de sang complet ont ete recueillis au moyen d'un tube capillaire (44,7 iii) chez 23 nourrissons par piqure a l'orteil. Des 6chantillons correspondants de lait maternel ont ete recueillis chez les meres. Le lait et le sang ont 6te soumis a une chromatographie en phase gazeuse, technique qui n'est toutefois pas assez sensible pour deceler le DDD (1,1-dichloro-2,2- bis(4-chIorophenyl)6thane), un metabolite du DDT, dans le sang des nourrissons. Le XDDT moyen (DDT total) dans les echan- tillons de sang complet 6tait de 127,03 ,ug/l, tan- dis que le XDDT moyen dans les echantillons de lait maternel 6tait de 15,06 mg/kg (graisses du lait). Une analyse par regression lineaire a montre que le XDDT (P <0,01) dans le sang complet des nourrissons augmentait significativement avec l'age. Une analyse par regression multiple a mon- tre que 70,0% de la variation du XDDT etaient dus a des diff6rences au niveau du degr6 de parit6 de la mere, de l'age de l'enfant et du EDDT dans le lait maternel. Ces memes variables expliquaient 76,3% de la variation du p,p'-DDE (1,1-dichloro- 2,2-bis(4-chloroph6nyl)ethylene), mais seulement 38,2% de la variation du p,p'-DDT. Le sang des nourrissons contenait 5% de DDT de plus que le lait maternel (P <0,05; test-t bilateral de Student). Bien que l'on puisse avancer comme hypotheses le m6tabolisme et l'excr6tion du DDE chez le nourrisson, ou I'absorption pr6f6- rentielle du DDT du compartiment intestinal dans le compartiment sanguin, il existe une explication plus logique. Les op6rations de lutte antipaludique comportent la pulv6risation de DDT, qui contient au maximum 4% de DDE. L'absorption de DDT a partir de cette source par le nourrisson pourrait expliquer la diff6rence de teneur en DDT entre le lait maternel et le sang du nourrisson; I'ampleur de cette difference t6moignerait de l'importance relative de I'absorption de DDT a partir du lait maternel et a partir de 1'environnement, et sem- blerait d6montrer qu'il existe une deuxieme voie d'absorption du DDT, s'ajoutant a I'absorption via le lait maternel. Toutefois, la contribution de cette derniere est consid6ree comme plus importante. De nombreux auteurs considerent comme preoccupante la possibilit6 d'une sensibilite accrue au DDT du systeme nerveux, du systeme immunitaire et du systeme renal chez le nourris- son, systemes qui n'atteignent leur maturite qu'apres la naissance. La seule etude dans laquelle ont et6 examines les effets neurologiques chez le nourrisson du DDE present dans le lait maternel indique que la proportion de nourrissons (ages d'environ 2 semaines) pr6sentant plus de quatre reponses (sur 20) diff6rees ou nulles aug- mentait avec la concentration de DDE dans le lait. Cette tendance se manifestait a partir de 4mg/kg de DDE dans les graisses du lait. Le XDDT moyen (qui comprend le DDE et les m6tabolites plus toxiques du DDT) trouve dans notre etude 6tait superieur a cette valeur. 11 existe par cons6- quent une possibilite reelle de risque pour le nour- risson. References 1. Sharp, B.L. et al. Analysis of 10 years of retrospec- tive malaria data from the KwaZulu areas of Natal. South African journal of science, 84: 102-106 (1988). 2. Bouwman, H. et al. Levels of DDT and metabolites in breast milk from KwaZulu mothers after DDT application for malaria control. Bulletin of the World Health Organization, 68: 761-768 (1990). 3. Pesticide residues in food - 1984: report of the Joint Meeting on Pesticide Residues. Rome, Food and Agriculture Organization, 1985 (FAO Plant Production and Protection Paper, No. 62). 4. Slorach, S.A. & Vaz, R. Assessment of human exposure to selected organochlorine compounds through biological monitoring. Uppsala, UNEP/WHO/ Swedish National Food Administration, 1983. 5. Bouwman, H. et al. Factors affecting levels of DDT and metabolites in human breast milk from Kwa- Zulu. Journal of toxicology and environmental health, 31: 93-115 (1990). 6. Morselli, P.L. Clinical pharmacokinetics in neo- nates. Clinical pharmacokinetics, 1: 81-98 (1976). 7. Principles for evaluating health risks from chemicals during infancy and early childhood: the need for a special approach. Geneva, World Health Organization, 1986 (Environmental Health Criteria No. 59). 8. Rogan, W.J. et al. Neonatal effects of transplacen- tal exposure to PCBs and DDE. Journal of pedi- atrics, 109: 335-341 (1986). 9. Bouwman, H. et al. Single-step extraction and clean-up procedure for determining serum levels of WHO Bulletin OMS. Vol 701992 249 H. Bouwman et al. DDT, DDE and DDD. Chemosphere, 18: 2085-2091 (1989). 10. Bouwman, H. et al. Extraction and clean-up proce- dure to determine DDT and metabolites in milk. Chemosphere, 19: 1563-1571 (1989). 11. Radomski, J.L. et al. Blood levels of organochlo- rine pesticides in Argentina: occupationally and nonoccupationally exposed adults, children and newborn infants. Toxicology and applied pharma- cology, 20: 186-193 (1971). 12. Morgan, D.P. et al. Transport of DDT, DDE and dieldrin in human blood. Bulletin of environmental contamination and toxicology, 8: 321-326 (1972). 13. Bouwman, H. et al. Malaria control and levels of DDT in serum of two populations in KwaZulu. Journal of toxicology and environmental health, 33: 141-155 (1991). 14. Kreiss, K. et al. Cross-sectional study of a commu- nity with exceptional exposure to DDT. Journal of the American Medical Association, 254: 1926-1930 (1981). 15. Bouwman, H. et al. Environmental and health impli- cations of DDT-contaminated fish from the Pongolo Flood Plain. Journal of African zoology, 104: 275-286 (1990). 16. Kodama, H. & Ota, H. Transfer of polychlorinated biphenyls to infants from their mothers. Archives of environmental health, 35: 95-100 (1981). 17. Engst, R. [The occurrence of DDT and DDE in the adipose tissue and organs of very young children]. Pharmazie, 24: 673-676 (1970) (in German). 18. Eckenhausen, F.W. et al. Organochlorine pesticide concentrations in perinatal samples from mothers and babies. Archives of environmental health, 36: 81-92 (1981). 19. Niessen, K.H. et al. Chlorinated hydrocarbons in adipose tissue of infants and toddlers: inventory and studies on their association with intake of mothers' milk. European journal of pediatrics, 142: 238-243 (1984). 20. Ando, M. Transfer of 2, 4, 5, 2', 4', 5' -hexachloro- biphenyl and 2,2-bis(p-chlorophenyl)-1,1,1-trichloro- ethane (p,p'-DDT) from maternal to newborn and suckling rats. Archives of toxicology, 41: 179-186 (1978). 21. Tomatis, L. et al. Storage levels of DDT metabo- lites in mouse tissue following long-term exposure to technical DDT. Tumori, 57: 377-396 (1971). 22. Mes, J. et al. Polychlorinated biphenyls and organo- chlorine pesticides in milk and blood of Canadian women during lactation. Archives of environmental contamination and toxicology, 13: 217-223 (1984). 23. Wickizer, T.M. et al. Polychlorinated biphenyl contamination of nursing mothers' milk in Michigan. American journal of public health, 71: 132-137 (1981). 24. Egan, H. et al. Organochlorine pesticide residues in human fat and human milk. British medical journal, 2: 66-69 (1965). 25. Quinby, G.E. et al. DDT in human milk. Nature, 207: 726-728 (1965). 26. Knoll, W. & Jayaraman, S. [Contamination of human milk with chlorinated hydrocarbons]. Nahrung, 17: 599-615 (1973) (in German). 27. Klein, D. et al. Cinetique d'elimination des compo- s6s organochlores au cours de la premiere semaine d'allaitement maternel. Food and chemical toxicolo- gy, 24: 869-873 (1986). 250 WHO Bulletin OMS. Vol 70 1992