World Health Organization (WHO) · Publications

Effect of increased potassium intake on blood pressure, renal function, blood lipids and other potential adverse effects

December 31, 2012 World Health Organization

The full text is hosted by the publishing organisation. lawenc.com indexes the metadata and links to the official source.

Full text

WHO Library Cataloguing-in-Publication Data Effect of increased potassium intake on blood pressure, renal function, blood lipids and other potential adverse effects. 1.Potassium, Dietary. 2.Blood pressure. 3.Lipids. 4.Kidney. 5.Triglycerides. 6.Cholesterol. 7.Adult. 8.Review literature. 9.Meta-analysis 10.Chronic disease – prevention and control. I. World Health Organization. ISBN 978 92 4 150488 1 (NLM classification: QV 277) © World Health Organization 2012 All rights reserved. Publications of the World Health Organization are available on the WHO web site (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; e-mail: bookorders@who.int). Requests for permission to reproduce or translate WHO publications – whether for sale or for noncommercial distribution – should be addressed to WHO Press through the WHO web site (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 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.

Printed by the WHO Document Production Services, Geneva, Switzerland

C o n t e n ts Acknowledgements.......................................................................................................... v Abbreviations and acronyms ........................................................................................... vi 1 Introduction ........................................................................................................ 1 1.1 1.2 1.3 2 Background ...................................................................................................... 1 Need for this review ........................................................................................ 1 Objectives ........................................................................................................ 2

Methods .............................................................................................................. 3 2.1 2.2 2.3 Criteria for considering studies for this review ............................................... 3 Identification of studies ................................................................................... 4 Data collection and analysis ............................................................................ 5 2.3.1 Selection of studies ............................................................................. 5 2.3.2 Data extraction and management ...................................................... 5 2.3.3 Assessment of risk of bias in included studies.................................... 6 2.3.4 Measures of treatment effect ............................................................ 8 2.3.5 Missing data ........................................................................................ 8 2.3.6 Data synthesis ..................................................................................... 8 2.3.7 Subgroup analysis ............................................................................... 8 2.3.8 Sensitivity analysis .............................................................................. 9

Results............................................................................................................... 10 3.1 3.2 3.3 Results of the search...................................................................................... 10 Retrieval of missing data................................................................................ 10 Included studies ............................................................................................. 10 3.3.1 3.3.2 3.3.3 3.3.4 3.3.5 3.4 3.5 Settings ............................................................................................. 10 Types of studies ................................................................................ 11 Participants ....................................................................................... 11 Interventions..................................................................................... 11 Outcome measures ........................................................................... 13

Excluded studies ............................................................................................ 14 Effects of interventions.................................................................................. 14 3.5.1 Resting blood pressure (systolic and diastolic) – primary findings ............................................................................................. 14 3.5.2 Resting blood pressure (systolic and diastolic) – secondary findings ............................................................................................. 16

iii

3.5.3 3.5.4 3.5.5 3.5.6 3.5.7 3.6

Ambulatory blood pressure – primary findings ................................ 18 Ambulatory blood pressure – secondary findings ............................ 19 Blood lipids ....................................................................................... 19 Catecholamine levels ........................................................................ 20 Renal function ................................................................................... 20

Sensitivity analysis and risk of bias ................................................................ 20 3.6.1 Quality of the body of evidence ....................................................... 20

3.7

Characteristics of studies ............................................................................... 21 3.7.1 Characteristics of included studies ................................................... 21 3.7.2 Characteristics of studies awaiting classification ............................. 45 3.7.3 Characteristics of ongoing studies .................................................... 47 3.7.4 Excluded studies and reasons for exclusion ..................................... 49 Effect estimate tables .................................................................................... 50 Figures .......................................................................................................... 54

3.8 3.9 4

References to studies ......................................................................................... 75 4.1 4.2 4.3 4.4 4.5 Included studies ............................................................................................. 75 Excluded studies ............................................................................................ 77 Studies awaiting classification ....................................................................... 80 Ongoing studies ............................................................................................. 80 Other references ............................................................................................ 81

Annex 1: Electronic search strategy ................................................................................ 83 Annex 2: Example data extraction template.................................................................... 91 Annex 3: Funnel plots..................................................................................................... 94 Annex 4: Risk of bias summary ....................................................................................... 96 Annex 5: Risk of bias graph............................................................................................. 97 Annex 6: GRADE evidence profiles .................................................................................. 98 Annex 7: Lists of tables and figures ............................................................................... 103 Full list of references .................................................................................................... 107

Ac k n o w l e d g e m e n t s This systematic review was prepared by Dr Nancy J Aburto, who was working as a scientist at the Nutrition Policy and Scientific Advice Unit (NPU) of the World Health Organization (WHO) Department of Nutrition for Health and Development (NHD), Ms Sarah Hanso, who was working as an intern in NPU at time of the preparation of this review, and Ms Hialy Gutierrez (Independent consultant). This review was one of three systematic reviews prepared to inform the development of the WHO guideline on potassium. All systematic reviews were presented to the WHO Nutrition Guidance Expert Advisory Group Subgroup on Diet and Health, which assisted WHO in the interpretation of the results and in the generation of the guideline informed by those results. WHO expresses special appreciation to the Ministry of Health, Labour and Welfare of the Government of Japan and the International Kidney Evaluation Association Japan for providing financial support for undertaking of the systematic reviews. Technical editing of the document was undertaken by Dr Hilary Cadman from Cadman Editing Services in Australia, and cover design was undertaken by Ms Sue Hobbs from Minimum Graphics in New Zealand.

Ab b r e vi a t i o n s a n d a c r o n y m s CCNFSDU CI CVD HDL HIV ICTRP ITT LDL LILACS MD NCD NUGAG WHO Symbols > < ≥ ≤ greater than less than equal to or greater than equal to or less than Codex Committee on Nutrition and Food for Special Dietary Uses confidence interval cardiovascular disease high-density lipoprotein human immunodeficiency virus International Clinical Trials Registry Platform intention-to-treat low-density lipoprotein Latin American and Caribbean Health Science Literature Database mean difference noncommunicable diseases Nutrition Guidance Expert Advisory Group World Health Organization

1 1.1

Introduction Background Noncommunicable diseases (NCDs) are the main contributor to mortality and morbidity globally (WHO, 2005). The major NCDs currently account for approximately 60% of all deaths and 43% of disease burden globally, and these levels are expected to continue to rise (Strong et al., 2005; WHO, 2005). In 2005, cardiovascular disease (CVD) itself accounted for 30% of all deaths – the equivalent of infectious disease, nutritional deficiency, and maternal and perinatal conditions combined (WHO, 2005). Hypertension is considered a major risk factor for heart attack, and the most important risk factor for stroke. Suboptimal systolic blood pressure (>115 mmHg) is estimated to contribute to 49% of all coronary heart disease and 62% of all strokes (Mackay & Mensah, 2004). Potassium consumption is of interest in public health nutrition, mainly due to its association with hypertension and CVDs. Increased consumption of potassium is thought to counteract the negative effects of sodium consumption on blood pressure. The biological plausibility of this theory is based on potassium's role in the physiological mechanisms through which the kidney reduces blood volume. Several large cohort studies have found an inverse association between potassium intake and risk of stroke (Ascherio et al., 1998; Khaw & Barrett-Connor, 1987). Additionally, three meta-analyses of trials comparing increased potassium with lower or usual potassium intake found that, on average, increased potassium intake lowers blood pressure (Cappuccio & MacGregor, 1991; Geleijnse et al., 2003; Whelton et al., 1997). However, another meta-analysis undertaken exclusively in individuals with hypertension did not detect a significant effect of potassium on blood pressure (Dickinson et al., 2006).

1.2

Need for this review The 32nd Session of the Codex Committee on Nutrition and Food for Special Dietary Uses (CCNFSDU) – held in Santiago, Chile on 1–5 November 2010 – made a special request to the World Health Organization (WHO) to consider establishing daily potassium intake values for the general population on the basis of dietary adequacy or reduction of chronic NCD risk (or both). The CCNFSDU requested that this work be included as part of the update of recommendations on sodium intake by the WHO Nutrition Guidance Expert Advisory Group (NUGAG). A 2002 joint WHO/Food and Agriculture Organization of the United Nations (FAO) Expert Consultation (WHO, 2003) concluded that the general population should consume a sufficient amount of potassium to maintain the molar ratio of sodium to potassium of 1:1. This ratio can generally be achieved with an intake of 70–80 mmol potassium/day if sodium is consumed at the recommended level. This recommendation is based on an improvement in blood pressure, and a blunting of the detrimental effects of higher sodium consumption on blood pressure. The expert consultation recommended that this intake of potassium be achieved through fruit and vegetable consumption. A report from the Institute of Medicine recommends a much higher level of potassium as an adequate intake – 120.5 mmol/day – based on literature citing benefits in blood pressure, risk of developing kidney stones and possibly decreased bone loss (Institute of Medicine, 2005). The latter two outcomes were not considered by the 2002 joint WHO/FAO Expert Consultation when recommending the level of potassium intake (WHO, 2003). The review presented here was undertaken in light of more recent evidence regarding potassium and blood pressure, and the association with

renal function and potential bone loss, plus the contradictory findings of recent systematic reviews. NUGAG will make use of this document when generating, reviewing and updating WHO guidelines on potassium intake.

1.3

Objectives The overall objective was to assess the effect, in adults, of increased potassium intake compared with normal, lower or usual potassium intake on blood pressure and renal function, and on adverse effects, such as changes in blood lipids and catecholamine levels,. Specific objectives were to assess whether: • consuming more potassium differentially affects blood pressure, renal function and adverse effects such as increased lipids, cholesterol and triglycerides relative to consuming less potassium; potassium intake resulting in a urinary potassium excretion at or above 70 mmol/day has a greater effect on blood pressure, renal function and adverse effects such as increased lipids, cholesterol and triglycerides than less potassium intake; potassium intake resulting in a urinary potassium excretion at or above 90 mmol/day has a greater effect on blood pressure, renal function and adverse effects such as increased lipids, cholesterol and triglycerides than less potassium intake; potassium intake resulting in a urinary potassium excretion at or above 120 mmol/day has a greater effect on blood pressure, renal function and adverse effects such as increased lipids, cholesterol and triglycerides than less potassium intake.

•

Estimating potassium intake Urinary potassium excretion is a common and valid form of estimating potassium intake. Data from 4680 men and women from 17 cities in four countries showed that average urinary potassium excretion was approximately 77% of intake (Stamler et al., 2003).Therefore, a factor of 1.30 is used to convert urinary potassium excretion to potassium intake. Because original studies reported urinary potassium excretion, data were analysed based on that value. Using the factor of 1.30: • • • 70 mmol urinary potassium/day equals approximately 91 mmol potassium intake/day; 90 mmol urinary potassium/day equals approximately 117 mmol intake/day; 120 mmol urinary potassium/day equals approximately 156 mmol intake/day.

2 2.1

Methods Criteria for considering studies for this review Study types

We included in the review randomized controlled trials (RCTs) – both individual and cluster randomized – that allocated at least one group of participants to increased potassium intake, and at least one control group to lower potassium intake, and that were of at least 4 weeks’ duration. We excluded studies that had concomitant interventions (i.e. nonpharmacological interventions, antihypertensive or other medications) in the intervention group unless those interventions were also applied to the control group (i.e. the only difference between the groups was the level of potassium intake). Participants

Studies considered for inclusion were those involving adults (≥16 years of age) of either gender, from the general population (free living) or specific groups (e.g. refugee populations). We considered studies in apparently healthy populations who may have been at risk of or have had hypertension, were known to have hypertension or were known to have normal blood pressure. We excluded studies targeting those who were acutely ill, infected with human immunodeficiency virus (HIV) or hospitalized. Interventions

We were interested in comparisons between increased potassium intake achieved through any means (e.g. supplements, food or dietary advice) and lower or usual potassium intake. If a manuscript presented multiple follow-up time points, data from the last follow-up were included in the overall analysis. Data from each time point were used in the subgroup analysis by duration of follow-up, without calculating an overall effect estimate across all subgroups. Outcome measures

The primary outcome measures were: • • • blood pressure (systolic or diastolic, or both); renal function; adverse effects including: –

increased total cholesterol, low-density lipoprotein (LDL) cholesterol and triglycerides; decreased high-density lipoprotein (HDL) cholesterol; increased adrenaline or noradrenaline (urinary or plasma); any other adverse effects reported by study authors.

– – –

As secondary outcome measures, any other outcomes reported in the studies were noted.

2.2

Identification of studies We searched for studies in two phases. In the first phase, we searched for high-quality systematic reviews on increased potassium consumption that included the outcomes of interest. If the inclusion criteria for an identified review were similar or equivalent to those of the current review, we used the references from that review as a list of potential studies, and completed the list by searching the literature subsequent to the search date used in that review. In some cases, we contacted the original authors of the systematic review and sought to collaborate, to explore the data in such a way as to meet our objectives. In the second phase, we undertook a complete search for data published since the date of of the search performed in the identified systematic review (see Electronic databases and Other resources, below). 2.2.1 Search for systematic reviews and meta-analyses

We first identified in the literature any high-quality systematic reviews of RCTs on the effect of increased potassium intake on blood pressure, renal function and adverse effects. If the inclusion criteria were in agreement with, or were broader than, the inclusion criteria defined for the specific objectives of the current literature review, the reference list of included studies was examined. Each of the original references was reviewed and compared against the inclusion criteria for the current review. Those references that met the inclusion criteria were included in the current review. Electronic databases

We searched the following electronic databases: • • • • • the Cochrane Central Register of Controlled Trials (searched 6 September 2011); MEDLINE (PubMed searched 28 August 2011); EMBASE (searched 25 August 2011); WHO International Clinical Trials Registry Platform (ICTRP) for ongoing trials (searched 1 September 2011); the Latin American and Caribbean Health Science Literature Database (LILACS) (searched 1 September 2011).

The detailed search strategy used for the electronic search is given in Annex 1. Other resources

We also searched the web site of WHO,1 and scanned the reference lists of identified papers for further trials. For assistance in identifying continuing or unpublished studies, we contacted the WHO Department of Nutrition for Health and Development, authors of recent systematic reviews and meta-analyses, and other academics and international partners with a known interest in this field.

www.who.int/nutrition

2.3

Data collection and analysis 2.3.1 Selection of studies

Identified references were independently assessed for potential relevance by two reviewers (HG, SH). These reviewers also independently scanned the title, abstract and keywords of every record retrieved, to determine which studies required further assessment. In the case of RCTs, the full reference was retrieved when the information given in the title, abstract and keywords suggested that the study: • included an intervention that targeted or achieved an increased potassium intake (however, to remain in the review, an increase in potassium had to have been documented); had a prospective design and a control group; included the random allocation of participants (individually or within clusters) to the intervention or control group; did not specifically target people identified as being infected with HIV or acquired immunodeficiency syndrome (AIDS), or acutely ill or hospitalized individuals; reported results of at least one of the outcomes of interest; involved an intervention of at least 4 weeks’ duration; had a measure of potassium excretion through 24-hour urinary potassium excretion.

• • • • • •

We also retrieved the full reference when it was unclear from scanning the title and abstract whether a study met the above criteria. If studies were published only as abstracts, or study reports contained little information on methods, we attempted to contact the authors to obtain further details of study design and results. Differences in opinion between the two review authors were resolved by consultation with a third reviewer (NA) and by consensus. If resolving disagreement through consensus was not possible, the reference was added to a list of those “awaiting assessment”, and authors were contacted for clarification. An adapted preferred reporting items for systematic reviews and meta-analyses (PRISMA) flowchart of study selection was generated (see Figure 3.1) (Liberati et al., 2009). 2.3.2 Data extraction and management

For studies that fulfilled inclusion criteria, two authors independently abstracted relevant population and intervention characteristics using a standard data extraction form (see Annex 2), with any disagreements resolved through discussion. Any relevant missing information about a trial was sought from the authors of the original reference. Data concerning details of study population, intervention and outcomes were extracted independently by two reviewers (HG, SH). Another author (NA) checked extracted data. The data extraction form included the following items: • General information – published or unpublished, title, authors, reference or source, contact address, country, language of publication, year of publication, duplicate publications, sponsor and setting.

•

Trial characteristics – design, duration of intervention or follow-up, method of randomization, allocation concealment and blinding (patients, people administering treatment and outcome assessors). Interventions or exposure – placebo or comparison included, interventions (dose, route and timing), comparison interventions (dose, route and timing), co-medications, potassium intake achieved at baseline and follow-up, and method of evaluation of potassium intake. Participants – sampling and randomization methods, inclusion and exclusion criteria of original study, total number and number in comparison groups, sex, age, baseline characteristics, diagnostic criteria, similarity of groups at baseline (including any comorbidities), assessment of compliance, withdrawals or losses to follow-up (reasons or description), subgroups analysed in original study, status of blood pressure and status of consumption of medication for controlling blood pressure. Outcomes – outcomes specified above and any other outcomes assessed, length of follow-up, and quality and completeness of reporting of outcomes. Results – outcomes, times of assessment, and a measure of variation (if necessary, converted to measures of effect specified below) and intention-to-treat (ITT) analysis. Stated objective of the study.

•

• • •

Duplicate publications

In the case of duplicate publications and companion papers of a primary study, we attempted to maximize the yield of information by simultaneously evaluating all available data. In cases of doubt, the original publication (usually the oldest version) was given priority. 2.3.3 Assessment of risk of bias in included studies

Data were entered into Review Manager software (RevMan 2008) and checked for accuracy by a second author. In cases of disagreement, a third party was consulted, and a judgement made based on consensus. We assessed risk of bias of RCTs by the quality criteria specified in the Cochrane handbook for systematic reviews of interventions 5.0.2 (Higgins et al., 2009), outlined below. Randomization (checking for possible selection bias) For each included study, we described the method used for randomization in sufficient detail to allow an assessment of whether it should produce comparable groups. Methods of randomization were categorized as one of the following: • • • • adequate – any truly random process (e.g. random number table or computer random number generator); inadequate – any non-random process ( e.g. odd or even date of birth, or hospital or clinic record number); no randomization; unclear.

Allocation concealment (checking for possible selection bias) We gave a detailed description of the method used to conceal the allocation, and determined whether intervention allocation could have been foreseen in advance of, or

during, recruitment, and whether it could have been changed after assignment. Methods were categorized as one of the following: • • • • adequate – for example, telephone or central randomization, or consecutively numbered sealed opaque envelopes; inadequate – open random allocation, unsealed or non-opaque envelopes, alternation or date of birth; concealment not used or not applicable; unclear.

Blinding (checking for possible performance bias) We described how studies blinded study participants and personnel from knowledge of which intervention a participant received. We judged studies to be at low risk of bias if they were blinded, or if we judged that the lack of blinding was unlikely to have affected the results. We assessed blinding separately for different outcomes or classes of outcomes. Methods were categorized as adequate, inadequate or unclear – for participants, personnel and outcome assessors. Incomplete outcome data (checking for possible attrition bias through withdrawals, dropouts and protocol deviations) We stated whether attrition and exclusions were reported, the numbers included in the analysis at each stage (compared with the total number of randomized participants), reasons for attrition or exclusion when reported, and whether missing data were balanced across groups or were related to outcomes. Methods regarding loss to follow-up or attrition were categorized as one of the following: • • • adequate – trials where few drop-outs or losses to follow-up were noted, and an ITT analysis was possible; inadequate – the rate of exclusion was at least 20%, or there were wide differences in exclusions between groups, whether or not ITT was used; unclear.

Selective reporting bias Methods regarding attempts to reduce selective reporting bias were categorized as one of the following: • • adequate – when it was clear that all of the study's prespecified outcomes and all expected outcomes of interest to the review were reported; inadequate – when not all of the study’s prespecified outcomes were reported, one or more reported primary outcomes were not prespecified, outcomes of interest were reported incompletely and could not be used, or the study failed to include results of a key outcome that would have been expected to have been reported; unclear.

•

Other sources of bias We noted any important concerns we had about other possible sources of bias, such as similarity of the groups at baseline. We assessed whether each study was free of other problems that could put it at risk of bias, and answered yes, no or unclear.

2.3.4

Measures of treatment effect

Dichotomous data were expressed as risk ratio with 95% confidence intervals (CI). Continuous variables were expressed as mean difference (MD) with 95% CI. 2.3.5 Missing data

Where feasible, we obtained relevant missing data from the authors of the original study, and evaluated important numerical data such as screened and randomized patients, as well as ITT. 2.3.6 Data synthesis

Data were summarized statistically if they were available, sufficiently similar and of sufficient quality. Statistical analyses were performed according to the statistical guidelines referenced in the newest version of the Cochrane handbook for systematic reviews of interventions (Higgins et al., 2009), using the random-effects model. Assessment of heterogeneity

In the event of substantial clinical, methodological or statistical heterogeneity, study results were not reported as meta-analytically pooled effect estimates were instead summarized in a narrative format. We identified heterogeneity by visual inspection of the forest plots, and by using a standard Chi-squared (I2) test and a significance level of α = 0.1, in view of the low power of this test. We specifically examined heterogeneity with the I2 statistic quantifying inconsistency across studies, to assess the impact of heterogeneity on the meta-analysis (Higgins et al., 2002; Higgins & Thompson, 2002) – where an I2 statistic of 75% or greater indicates a considerable level of inconsistency (Higgins et al., 2009). Where heterogeneity was found, we attempted to determine the potential causes by examining individual study and subgroup characteristics. Statistical comparisons

For each of the objectives discussed above, we analysed the data through both direct and indirect comparisons: • • direct comparisons looked within each study for multiple groups randomized into the different specific potassium levels; indirect comparisons involved a subgroup analysis of all studies testing the effect of potassium on the specified outcomes by potassium intake achieved in the intervention group. Subgroup analysis

2.3.7

We conducted both overall analysis and subgroup analyses, to explore effect-size differences between groups, as follows: • • • • by gender (male vs female vs combined); by hypertensive status of participants at baseline (hypertensive vs not hypertensive vs undetermined status or heterogeneous); by level of potassium intake achieved in the intervention groups (<70 mmol/day vs 70– 90 mmol/day vs 90–120 mmol/day vs ≥120 mmol/day urinary potassium excretion); by achieved difference in intake between intervention and control (<30 mmol/day vs 30–60 mmol/day vs ≥60 mmol/day urinary potassium excretion);

• • • • • • •

by population average level of potassium intake at baseline (<40 mmol/day vs 40– 60 mmol/day vs ≥60 mmol/day urinary potassium excretion); by population average level of sodium intake at baseline (≤2 g/day vs 2–4 g/day vs ≥4 g/day); by duration of intervention (<2 months vs 2–4 months vs >4 months); by type of intervention (diet or feeding vs supplementation vs advice); by type of blood pressure device used (automatic, manual); by method of measurement of blood pressure (supine office, seated office, standing office, supine home, seated home, standing home); by study design (parallel, cross-over). Sensitivity analysis

2.3.8

Sensitivity analysis was used to examine the effects of removing studies at high risk of bias from the analysis. We considered a study to be of high risk of bias if it was graded as “inadequate” in both the randomization and allocation concealment, and in either blinding or loss to follow-up. Quality of the body of evidence

We used funnel plots to assess the potential existence of small-study bias, and carefully interpreted the results (Lau et al., 2006; Sterne & Egger, 2001). We generated a “risk of bias summary” (Annex 4) and a “risk of bias graph” (Annex 5), and assessed the impact of individual bias domains on study results at end point and study levels. GRADEProfiler software (version 3.6) was used to assess the quality of the body of evidence according to the methodology of Grading of recommendations assessment, development and evaluation (Guyatt et al., 2008).

3 3.1

R e s u l ts Results of the search The search for RCTs of potassium intake and outcomes of interest in PubMed, the Cochrane Central Register of Controlled Trials, EMBASE, the WHO ICTRP and LILACS resulted in the identification of 4882 publications and study reports. The scan of the reference lists of previous high-quality systematic reviews and the reference lists of included studies resulted in the identification of another 44 studies of possible interest. Thus, 4926 publications or study reports for possible inclusion in the systematic review were identified, of which: • • • 4646 were excluded after review of the titles, due to obvious lack of relevance to this review; 50 of the remaining 280 studies were removed for being duplicates or companion papers to primary references to studies, leaving a total of 237 potential studies; 172 of the remaining 237 studies were excluded for not meeting basic criteria for inclusion.

The remaining 58 studies were fully reviewed. Of these, 27 were excluded on the basis of the methods and one was identified as a duplicate only after full review, leaving 30 studies for inclusion in the review. Of these, seven are awaiting author communications, and one does not report quantitative data and will thus not contribute to the meta-analysis. Hence, a total of 22 studies, contributed to the meta-analyses. The process of selection of studies is shown in Figure 3.1.

3.2

Retrieval of missing data The following authors were contacted and generously provided additional information about their studies: Anne Barden (Barden et al., 1986), Norman Kaplan (Kaplan et al., 1985) and Thomas Sanders (Berry et al., 2010).

3.3

Included studies Details of the characteristics of the included studies are shown in Section 3.7.1 (Tables 3.1– 3.51). 3.3.1 Settings

All included studies were published in English. Of the 23 included studies, two were undertaken in Australia (Barden et al., 1986; Chalmers et al., 1986), one in Chile (Valdés et al., 1991), one in China (Gu et al., 2001), one in Germany (Overlack et al., 1991), one in India (Patki et al., 1990), two in Italy (Siani et al., 1991; Siani et al., 1987), one in Jamaica (Forrester & Grell, 1988), one in Japan (Kawano et al., 1998), one in Kenya (Obel, 1989), one in the Netherlands (Grobbee et al., 1987), one in New Zealand (Richards et al., 1984), one in South Africa (Matlou et al., 1986), five in the United Kingdom (Berry et al., 2010; Bulpitt et al., 1985; Fotherby & Potter, 1992; He et al., 2010; MacGregor et al., 1982) and four in the United States of America (USA) (Kaplan et al., 1985; Smith et al., 1985; Trial Hyp Prv Col, 1992; Whelton et al., 1995).

3.3.2

Types of studies

All studies reporting blood pressure, renal function, blood lipids or catecholamine levels were RCTs. Eight trials had a parallel design (Bulpitt et al., 1985; Chalmers et al., 1986; Gu et al., 2001; Obel, 1989; Siani et al., 1991; Siani et al., 1987; Trial Hyp Prv Col, 1992; Whelton et al., 1995). The remaining 15 trials had a cross-over design (Barden et al., 1986; Berry et al., 2010; Forrester & Grell, 1988; Fotherby & Potter, 1992; Grobbee et al., 1987; He et al., 2010; Kaplan et al., 1985; Kawano et al., 1998; MacGregor et al., 1982; Matlou et al., 1986; Overlack et al., 1991; Patki et al., 1990; Richards et al., 1984; Smith et al., 1985; Valdés et al., 1991). The duration of the RCTs ranged from 4 weeks (Barden et al., 1986; Forrester & Grell, 1988; Fotherby & Potter, 1992; He et al., 2010; Kawano et al., 1998; MacGregor et al., 1982; Richards et al., 1984; Smith et al., 1985; Valdés et al., 1991) to 12 months (Siani et al., 1991). Most of the studies (15) had a duration of less than 3 months (Barden et al., 1986; Berry et al., 2010; Forrester & Grell, 1988; Fotherby & Potter, 1992; Grobbee et al., 1987; He et al., 2010; Kaplan et al., 1985; Kawano et al., 1998; MacGregor et al., 1982; Matlou et al., 1986; Overlack et al., 1991; Patki et al., 1990; Richards et al., 1984; Smith et al., 1985; Valdés et al., 1991). The Fotherby (1992) study included a subgroup in which 8 of the 18 original patients were followed up for 3 months of intervention, after the 1-month original study. The Gu (2001) study reported data after 6 weeks of intervention, in addition to the 3-month data. The Matlou (1986) study reported data after 4 and 6 weeks of intervention. The Obel (1989) study reported data after 1, 2, 3 and 4 months of intervention. The Trial Hyp Prv Col (1992) and Whelton (1995) studies reported data after 3 and 6 months of intervention. Three studies were undertaken in individuals with normal blood pressure (Barden et al., 1986; Trial Hyp Prv Col, 1992; Whelton et al., 1995), 18 in individuals with hypertension (Bulpitt et al., 1985; Chalmers et al., 1986; Forrester & Grell, 1988; Fotherby & Potter, 1992; Grobbee et al., 1987; He et al., 2010; Kaplan et al., 1985; Kawano et al., 1998; MacGregor et al., 1982; Matlou et al., 1986; Obel, 1989; Overlack et al., 1991; Patki et al., 1990; Richards et al., 1984; Siani et al., 1991; Siani et al., 1987; Smith et al., 1985; Valdés et al., 1991), and two in a heterogeneous group of individuals with hypertension or normal blood pressure (Berry et al., 2010; Gu et al., 2001). Two studies reported on an all-female population (Barden et al., 1986; Matlou et al., 1986). The remaining 21 studies were conducted in a heterogeneous population of men and women (Berry et al., 2010; Bulpitt et al., 1985; Chalmers et al., 1986; Forrester & Grell, 1988; Fotherby & Potter, 1992; Grobbee et al., 1987; Gu et al., 2001; He et al., 2010; Kaplan et al., 1985; Kawano et al., 1998; MacGregor et al., 1982; Obel, 1989; Overlack et al., 1991; Patki et al., 1990; Richards et al., 1984; Siani et al., 1991; Siani et al., 1987; Smith et al., 1985; Trial Hyp Prv Col, 1992; Valdés et al., 1991; Whelton et al., 1995). 3.3.3 Participants

The number of participants in the trials ranged from 12 (Overlack et al., 1991) to 353 (Whelton et al., 1995). There were a total of 1606 participants: 445 in cross-over trials and 1161 in parallel trials, of whom 719 had hypertension, 689 had normal blood pressure and 198 had an undisclosed hypertensive status at baseline. 3.3.4 Interventions

All of the included studies intended to compare health outcomes between a group of participants consuming a normal or usual potassium intake to a group consuming increased potassium. In one study, the intervention was dietary advice or education, plus a tablet

supplement (Berry et al., 2010). In two studies, the intervention was dietary advice or education (Chalmers et al., 1986; Siani et al., 1991). The remaining 20 studies used a supplement intervention (Barden et al., 1986; Bulpitt et al., 1985; Forrester & Grell, 1988; Fotherby & Potter, 1992; Grobbee et al., 1987; Gu et al., 2001; He et al., 2010; Kaplan et al., 1985; Kawano et al., 1998; MacGregor et al., 1982; Matlou et al., 1986; Obel, 1989; Overlack et al., 1991; Patki et al., 1990; Richards et al., 1984; Siani et al., 1987; Smith et al., 1985; Trial Hyp Prv Col, 1992; Valdés et al., 1991; Whelton et al., 1995). Compliance to intervention and control was monitored in all studies using 24-hour urinary potassium excretion. Twenty-two studies contributed only one comparison between an increased potassium intake group and a corresponding control group (Barden et al., 1986; Bulpitt et al., 1985; Chalmers et al., 1986; Forrester & Grell, 1988; Fotherby & Potter, 1992; Grobbee et al., 1987; Gu et al., 2001; He et al., 2010; Kaplan et al., 1985; Kawano et al., 1998; MacGregor et al., 1982; Matlou et al., 1986; Obel, 1989; Overlack et al., 1991; Patki et al., 1990; Richards et al., 1984; Siani et al., 1991; Siani et al., 1987; Smith et al., 1985; Trial Hyp Prv Col, 1992; Valdés et al., 1991; Whelton et al., 1995). However, one of these studies did not contribute to the meta-analysis, because additional information on outcomes of interest was requested from study authors and had not been received at the time of writing (Overlack et al., 1991). One study contributed two comparisons between two potassium groups and one corresponding control group, with the different interventions administered in a cross-over design (Berry et al., 2010). Thus, the 23 studies contributed 22 comparisons between an increased potassium intake group and a corresponding control group for the generation of overall estimates of effect on health outcomes, with an additional comparison from the Berry (2010) study used only in the subgroup analysis. In two studies, the achieved potassium intake in the intervention group was less than that necessary for a urinary potassium excretion of 70 mmol/day (Forrester & Grell, 1988; Gu et al., 2001). In five studies, the potassium intake at follow-up in the intervention group was at least equal to that necessary for a urinary potassium excretion of 70 mmol/day, but less than 90 mmol/day (Berry et al., 2010; Kaplan et al., 1985; Patki et al., 1990; Siani et al., 1991; Siani et al., 1987). In 11 studies, the potassium intake at follow-up was at least equal to that necessary for a urinary potassium excretion of 90 mmol/day, but less than 120 mmol/day for the intervention group (Barden et al., 1986; Bulpitt et al., 1985; Chalmers et al., 1986; Fotherby & Potter, 1992; Kawano et al., 1998; MacGregor et al., 1982; Matlou et al., 1986; Obel, 1989; Smith et al., 1985; Trial Hyp Prv Col, 1992; Whelton et al., 1995). In four studies, the potassium intake at follow-up was greater than that necessary for a urinary potassium excretion of 120 mmol/day for the intervention group (Grobbee et al., 1987; He et al., 2010; Richards et al., 1984; Valdés et al., 1991). One study reported a baseline sodium intake of less than 2 g/day (Smith et al., 1985). At baseline, 18 studies reported a sodium intake of 2–4 g/day (Barden et al., 1986; Berry et al., 2010; Bulpitt et al., 1985; Chalmers et al., 1986; Forrester & Grell, 1988; Fotherby & Potter, 1992; Grobbee et al., 1987; He et al., 2010; Kaplan et al., 1985; Kawano et al., 1998; MacGregor et al., 1982; Matlou et al., 1986; Obel, 1989; Overlack et al., 1991; Siani et al., 1987; Trial Hyp Prv Col, 1992; Valdés et al., 1991; Whelton et al., 1995). Four studies reported a baseline sodium intake of more than 4 g/day (Gu et al., 2001; Kawano et al., 1998; Patki et al., 1990; Siani et al., 1987). In 15 studies, participants were not taking any medical therapy to control blood pressure (Berry et al., 2010; Chalmers et al., 1986; Fotherby & Potter, 1992; Gu et al., 2001; He et al., 2010; MacGregor et al., 1982; Matlou et al., 1986; Obel, 1989; Overlack et al., 1991; Patki et al., 1990; Richards et al., 1984; Smith et al., 1985; Trial Hyp Prv Col, 1992; Valdés et al., 1991;

Whelton et al., 1995). In five studies, participants were taking medical therapy to control blood pressure (Bulpitt et al., 1985; Forrester & Grell, 1988; Kaplan et al., 1985; Siani et al., 1991; Siani et al., 1987). In one study, participants included both treated and untreated individuals (Kawano et al., 1998). In two studies, the status of consumption of medication to control blood pressure was unspecified or unknown (Barden et al., 1986; Grobbee et al., 1987). 3.3.5 Outcome measures

Resting blood pressure (systolic and diastolic)

Twenty-one studies contributed a total of 21 comparisons to the combined analysis of resting blood pressure (Barden et al., 1986; Berry et al., 2010; Bulpitt et al., 1985; Chalmers et al., 1986; Forrester & Grell, 1988; Fotherby & Potter, 1992; Grobbee et al., 1987; Gu et al., 2001; He et al., 2010; Kaplan et al., 1985; Kawano et al., 1998; MacGregor et al., 1982; Matlou et al., 1986; Patki et al., 1990; Richards et al., 1984; Siani et al., 1991; Siani et al., 1987; Smith et al., 1985; Trial Hyp Prv Col, 1992; Valdés et al., 1991; Whelton et al., 1995). One study was excluded from the resting blood pressure meta-analysis because it introduced excessive heterogeneity (Tau2 = 88.59, 2 = 555.78, P <0.00001, I2 = 96%) (Obel, 1989). The removal of this study from the meta-analysis decreased the heterogeneity (Tau2 = 6.52, 2 =56.52, P <0.0001, I2 = 65%). Another study was removed from most meta-analyses of resting diastolic blood pressure because it introduced excessive heterogeneity into the statistical analyses (Patki et al., 1990). Ten studies measured resting blood pressure with an automatic device (Barden et al., 1986; Berry et al., 2010; Chalmers et al., 1986; Fotherby & Potter, 1992; He et al., 2010; MacGregor et al., 1982; Richards et al., 1984; Siani et al., 1991; Smith et al., 1985; Valdés et al., 1991). Eleven studies measured blood pressure using a manual device (Bulpitt et al., 1985; Forrester & Grell, 1988; Grobbee et al., 1987; Gu et al., 2001; Kaplan et al., 1985; Kawano et al., 1998; Matlou et al., 1986; Patki et al., 1990; Siani et al., 1987; Trial Hyp Prv Col, 1992; Whelton et al., 1995). Thirteen studies measured supine office blood pressure (Barden et al., 1986; Berry et al., 2010; Forrester & Grell, 1988; Fotherby & Potter, 1992; Grobbee et al., 1987; Kaplan et al., 1985; MacGregor et al., 1982; Patki et al., 1990; Richards et al., 1984; Siani et al., 1991; Siani et al., 1987; Smith et al., 1985; Valdés et al., 1991). Seven studies measured seated office blood pressure (Barden et al., 1986; Chalmers et al., 1986; Gu et al., 2001; Kawano et al., 1998; Matlou et al., 1986; Trial Hyp Prv Col, 1992; Whelton et al., 1995). Eight studies measured standing office blood pressure (Forrester & Grell, 1988; Fotherby & Potter, 1992; MacGregor et al., 1982; Patki et al., 1990; Richards et al., 1984; Siani et al., 1987; Smith et al., 1985; Valdés et al., 1991). Two studies reported resting blood pressure data, but did not specify further the method by which the measurements were taken (Bulpitt et al., 1985; He et al., 2010). Ambulatory blood pressure (systolic and diastolic)

Four studies measured ambulatory blood pressure (Berry et al., 2010; Fotherby & Potter, 1992; He et al., 2010; Kawano et al., 1998). All four measured 24-hour ambulatory blood pressure, as well as day and night ambulatory blood pressure, and measured ambulatory blood pressure with an automatic device. Blood lipids

Four studies reported total cholesterol (Berry et al., 2010; Grobbee et al., 1987; Kaplan et al., 1985; Patki et al., 1990). One study reported quantitative data on LDL cholesterol (Berry et al., 2010). Two studies reported quantitative data on total triglycerides (Berry et al., 2010; Kaplan et al., 1985). Two studies reported quantitative data on HDL cholesterol (Berry et al., 2010; Kaplan et al., 1985).

Catecholamine levels

Three studies reported results on plasma adrenaline and noradrenaline (Grobbee et al., 1987; Richards et al., 1984; Valdés et al., 1991). No studies measured urinary catecholamine levels. Renal function Renal function was measured by serum creatinine in three studies (Bulpitt et al., 1985; Patki et al., 1990; Smith et al., 1985).

3.4

Excluded studies Reasons for exclusion of the 28 ineligible studies are given in Section 3.7.4. In 12 studies, the duration of the intervention was less than 4 weeks (Agnoli et al., 1992a; Agnoli et al., 1992b; Ceglia et al., 2009; Fujita & Ando, 1984; Iimura et al., 1981; Khaw & Thom, 1982; Krishnan, 2010; Lennon & Lemann J, 1968; Parfrey et al., 1981; Poulter & Sever, 1986; Sanclemente, 1987; Smith et al., 1992). In nine studies, 24-hour urinary potassium excretion was not reported (Barcelo et al., 1993; CWP, 1987; Gamarra et al., 1994; Grimm, 1999.; Grimm et al., 1988; Grimm et al., 1990; Jardim et al., 1988; Rahimi et al., 2007; Svetkey et al., 1987). In four studies, other nutrients were manipulated as part of the intervention (Agnoli et al., 1994; Langford et al., 1991; Overlack et al., 1995; WHOICTRP, 2008). In one study, there was no control group (Heller et al., 1998). One study used a child population (Gomez-Marin et al., 1991). One study was a duplicate (MacGregor et al., 1984).

3.5

Effects of interventions The effects of increased potassium versus control in adults are summarized in Tables 3.53– 3.58, and in Figures 3.2–3.27. 3.5.1 Resting blood pressure (systolic and diastolic) – primary findings

Indirect comparisons

The meta-analysis of change in systolic blood pressure is shown in Figure 3.2 and Table 3.53. Systolic blood pressure was reduced by increased potassium intake relative to normal potassium intake by 3.49 mmHg (95%CI: 1.82, 5.15). The reduction in systolic blood pressure in studies specifically targeting individuals with hypertension was 5.53 mmHg (95%CI: 3.51, 7.56), which was statistically significantly greater than the reduction from the trials targeting individuals with normal blood pressure (0.09 mmHg, 95%CI: –0.77, 0.95). The two studies reporting on heterogeneous populations including some individuals with hypertension and some with normal blood pressure showed a reduction in systolic blood pressure of 2.95 mmHg (95%CI: 0.26, 5.65); this reduction was significant, but was not significantly different from that found in individuals with hypertension (Table 3.54). After removing the studies in individuals with normal blood pressure, the overall effect of increased potassium on systolic blood pressure was a reduction of 4.77 mmHg (95%CI: 3.14, 6.41). In studies where the achieved potassium intake of the intervention was that necessary for a urinary potassium excretion of less than 70 mmol/day, the reduction of systolic blood pressure was 3.65 mmHg (95%CI: 0.62, 6.69) (Figure 3.4 and Table 3.53). In studies where the achieved potassium intake of the intervention was equal to that necessary for a urinary potassium excretion of 70–90 mmol/day, the reduction of systolic blood pressure was 7.16 mmHg (95%CI: 1.91, 12.41). In studies where the achieved potassium intake of the intervention was equal to that necessary for a urinary potassium excretion of 90–

120 mmol/day, the reduction in systolic blood pressure was a borderline significant 1.71 mmHg (95%CI: 0.00, 3.42). In studies where achieved potassium intake was greater than that necessary for a urinary potassium excretion of 120 mmol/day, the change in resting systolic blood pressure was non-significant, at 3.00 mmHg (95%CI: –0.27, 6.28). When the studies in participants with normal blood pressure were removed from the analysis, the reduction in those studies that achieved an intake of potassium equal to that necessary for a urinary potassium excretion of 90–120 mmol/day was 4.69 mmHg (95%CI: 2.56, 6.82); this reduction was significant, but was not significantly different from studies achieving a potassium intake equal to that necessary for a urinary potassium excretion of less than 70 mmol/day or 70–90 mmol/day. No other groups were affected. The difference in achieved potassium intake between the intervention and control groups had no significant effect on the reduction in systolic blood pressure (Figure 3.5 and Table 3.53). In studies where the difference in achieved potassium intake between the intervention and the control was less than 30 mmol/day, systolic blood pressure was reduced by 4.89 mmHg (95%CI: 2.20, 7.59). In studies where the difference in achieved potassium intake between the intervention and the control was 30–60 mmol/day, systolic blood pressure was reduced by 1.97 mmHg (95%CI: 0.09, 3.85). In studies where the difference in achieved potassium intake between the intervention and the control was greater than 60 mmol/day, systolic blood pressure was reduced by 3.01 mmHg (95%CI: – 1.02,7.03), which was not statistically significant. When the studies in participants with normal blood pressure were excluded from the meta-analysis, the decrease in systolic blood pressure was 4.38 mmHg (95%CI: 1.57, 7.18) in the studies that achieved a difference of 30– 60 mmol/day in potassium intake, and 6.02 mmHg (95%CI: 0.99, 11.05) in the studies that achieved a difference of more than 60 mmol/day; however, these reductions were not statistically different from the group with the achieved difference of less than 30 mmol/day. The meta-analysis of change in diastolic blood pressure is presented in Figure 3.6 and Table 3.54. With increased potassium intake relative to normal potassium intake, diastolic blood pressure was reduced by 3.02 mmHg (95%CI: 1.17, 4.86). The reduction in diastolic blood pressure was not statistically significant in studies specifically targeting individuals with normal blood pressure (0.56 mmHg, 95%CI:-0.42, 1.55) or in studies targeting heterogeneous groups of individuals with hypertension or normal blood pressure (0.17 mmHg, 95%CI: -1.48,1.82), but had statistical significance in studies targeting those with hypertension with a reduction of 3.91 mmHg (95%CI: 1.28, 6.54,) (Figure 3.7 and Table 3.54). Achieved potassium intake had little effect on the reduction in diastolic blood pressure (Figure 3.8 and Table 3.54). The reduction in diastolic blood pressure was not significant in any of the subgroups of achieved intake (<70 mmol urinary excretion/day 1.35 mmHg [95%CI: –2.60, 5.31]; 70–90 mmol urinary excretion/day 4.01 mmHg [95%CI: –0.42, 8.44]; 90–120 mmol urinary excretion/day 0.83 mmHg [95%CI: –0.17, 1.82]; ≥120 mmol urinary excretion/day 1.75 mmHg [95%CI: –0.74, 4.23]). When only the studies including individuals with hypertension were included in the analysis, the reduction in the group that achieved an intake resulting in 70–90 mmol urinary excretion/day became a statistically significant 7.78 mmHg (95%CI: 1.80, 13.77). No other groups reached statistical significance. The reduction in diastolic blood pressure was not significantly affected by the difference in achieved potassium intake between the intervention and control groups (Figure 3.9 and Table 3.54). In studies where the difference in achieved potassium intake was less than 30 mmol/day, the reduction in diastolic blood pressure was not statistically significant (1.87 mmHg, 95%CI: –4.11, 0.37). In studies where the difference in achieved potassium

intake was 30–60 mmol/day, diastolic blood pressure was reduced by 1.63 mmHg (95%CI: 0.21, 3.04). In studies where the difference in achieved potassium intake was greater than 60 mmol/day, diastolic blood pressure was reduced by 3.57 mmHg (95%CI: 0.82, 6.32). When only studies of hypertensive participants were included in the meta-analysis, there was a significant reduction in diastolic blood pressure when the difference in potassium intake was less than 30 mmol/day (3.89 mmHg, 95%CI: 2.34, 5.43) and when the difference in intake was 30–60 mmol/day (2.82 mmHg, 95%CI: 0.46, 5.18). Direct comparisons

In one study (Berry et al., 2010), participants were randomized into two levels of potassium intake. The dietary advice interventions targeted an increase of 20 mmol/day versus 40 mmol/day of potassium via fruits and vegetables in the two intervention groups. The group with the 40 mmol/day increase in potassium did not achieve a statistically significant reduction in systolic blood pressure (1.90 mmHg, 95%CI: –6.82,3.02) or diastolic blood pressure (0.70 mmHg, 95%CI: –3.78,2.38) relative to the group with the 20 mmol/day increase in potassium intake. The absolute intake each group achieved was 75 mmol/day versus 84 mmol/day. 3.5.2 Resting blood pressure (systolic and diastolic) – secondary findings

Baseline potassium intake had no statistically significant impact on reduction in systolic blood pressure with increased potassium intake (Figure 3.10 and Table 3.53). In studies where the baseline potassium urinary excretion was less than 40 mmol, systolic blood pressure decreased by 3.89 mmHg (95%CI: 0.74, 7.03). In studies where the baseline potassium urinary excretion was 40–60 mmol/day, systolic blood pressure decreased by 3.39 mmHg (95%CI: 1.28, 5.51). In studies where the baseline potassium urinary excretion was greater than 60 mmol/day, systolic blood pressure decreased by 4.11 mmHg (95%CI: 1.97, 6.26). When only the studies including individuals with hypertension, or heterogeneous populations, were included in the analysis, the group of studies with a baseline potassium urinary excretion of 40–60 mmol/day had a further decrease of 5.78 mmHg (95%CI: 2.86, 8.71), which was not statistically different from the other subgroups. Baseline sodium intake also had no statistically significant impact on the reduction in systolic blood pressure with increased potassium intake (Figure 3.11 and Table 3.53). In the one study with a baseline sodium value less than 2 g/day, the reduction in systolic blood pressure was not statistically significant (2.00 mmHg, 95%CI: –7.70, 11.70). In studies with baseline sodium intake of 2–4 g/day, systolic blood pressure was reduced by 1.97 mmHg (95%CI: 0.52, 3.41). Studies where the sodium intake at baseline was more than 4 g/day showed a greater (but not statistically significant) reduction in systolic blood pressure (6.91 mmHg, 95%CI: 2.29, 11.53). When studies in individuals with normal blood pressure were excluded from the meta-analysis, the reduction detected in the group of studies with baseline sodium intake of 2–4 g/day was 4.07 mmHg (95%CI: 2.37, 5.76), and the other groups remained unchanged. There was a decrease in systolic blood pressure of 3.36 mmHg (95%CI: 1.76, 4.94) in the studies with a duration of less than 2 months (Figure 3.12 and Table 3.53). In studies with a duration of 2–4 months, the decrease was 3.83 mmHg (95%CI: 0.95, 6.72). In studies with a duration of more than 4 months, the reduction in systolic blood pressure was not

statistically significant (0.02 mmHg, 95%CI: –0.90, 0.85). Removing the studies in individuals with normal blood pressure had little effect on the results. The reduction in systolic blood pressure was not affected by measurement device (automatic device 3.45 mmHg [95%CI: 1.77, 5.13] vs manual device 3.63 mmHg [95%CI: 1.28, 5.98]). The measurement method also did not affect the reduction in systolic blood pressure (supine office 4.88 mmHg [95%CI: 2.13, 7.64] vs seated office 1.83 mmHg [95%CI: 0.12, 3.54] vs standing office 6.94 mmHg [95%CI: 3.25, 10.63]). When the measurement method was unspecified, the reduction in systolic blood pressure was not significant (2.51 mmHg, 95%CI: –2.85, 7.87) (Figures 3.13–3.14 and Table 3.53). The exclusion of studies in only individuals with normal blood pressure had little effect on the results. The reduction in systolic blood pressure with increased potassium intake was detected in both individuals not taking medication to control blood pressure (3.63 mmHg, 95%CI: 1.57, 5.69) and those taking medication to control blood pressure (5.85 mmHg, 95%CI: 1.08, 10.61). In studies where the medication status was undetermined, the reduction in blood pressure was not statistically significant (1.16 mmHg, 95%CI: –1.43, 3.74) (Figure 3.15 and Table 3.53). When studies in only individuals with normal blood pressure were removed from the analysis, systolic blood pressure in individuals not taking medication to control blood pressure was further reduced by 5.13 mmHg (95%CI: 3.19, 7.06); in those of unspecified medication status, it was further reduced by 2.07 (95%CI: –1.41, 5.56). The reduction in systolic blood pressure in individuals with increased potassium intake was not affected by type of intervention (supplement 3.31 mmHg [95%CI: 1.55, 5.07] vs dietary advice or education 4.19 mmHg [95%CI: 1.92, 6.46]) (Figure 3.16 and Table 3.53). The exclusion of studies in only individuals with normal blood pressure had little effect on the results. The reduction in systolic blood pressure in individuals with increased potassium intake was also not affected by study design (parallel design 2.74 mmHg [95%CI: 0.39, 5.08] vs crossover 3.99 mmHg [95%CI: 1.86, 6.13]) (Figure 3.17 and Table 3.53). The exclusion of studies in only individuals with normal blood pressure increased the reduction detected with both study designs (parallel 5.19 [95%CI: 2.23, 8.16] vs cross-over 4.57 [95%CI: 2.44, 6.70]). In studies where the baseline potassium urinary excretion was less than 40 mmol/day, the reduction in diastolic blood pressure was not significant (2.41 mmHg, 95%CI: –3.07, 7.90) (Table 3.54). However, this reduction became significant with the inclusion of only individuals with hypertension (5.80 mmHg, 95%CI: 0.53, 11.07). In studies where the baseline potassium urinary excretion was 40–60 mmol/day, diastolic blood pressure decreased by 1.53 mmHg (95%CI: 0.25, 2.80). In studies where the baseline potassium urinary excretion was greater than 60 mmol/day, diastolic blood pressure decreased by 3.38 mmHg (95%CI: 2.02, 4.74). Sodium intake at baseline did not have a significant effect on the reduction of diastolic blood pressure detected with increased potassium intake (Table 3.54). In the one study where the sodium intake at baseline was less than 2 g/day, a reduction in diastolic blood pressure was not statistically significant (0.00 mmHg, 95%CI: –6.12, 6.12). In studies where the sodium intake at baseline was 2–4 g/day, the reduction in diastolic blood pressure was 1.96 mmHg (95%CI: 0.76, 3.16). Studies where the sodium intake at baseline was greater than 4 g/day did not have a statistically significant reduction in diastolic blood pressure (2.87 mmHg; 95%CI: –1.22, 6.96).

When only studies including individuals with hypertension were included in the metaanalysis, the diastolic blood pressure was further reduced in those studies where the baseline sodium intake was 2–4 g/day (3.42 mmHg; 95%CI: 2.21, 4.62). There was little change in the other groups. The reduction in diastolic blood pressure with increased potassium intake was only significant in studies of a duration of less than 2 months (1.99 mmHg, 95%CI: 0.87, 3.11) (Table 3.54). In studies where the duration of intervention was 2–4 months, the reduction in diastolic blood pressure was not statistically significant (1.86 mmHg, 95%CI: –0.02, 3.75). In studies with a duration of more than 4 months, the reduction was 0.35 mmHg (95%CI: –0.35, 1.06). There was no meaningful change in results when only studies with individuals with hypertension were included in the meta-analysis. Type of device used to measure blood pressure did not affect the reduction in diastolic blood pressure (automatic device 2.84 mmHg [95%CI: 1.71, 3.96] vs manual 1.54 mmHg [95%CI: 0.11, 2.96]) (Table 3.54). The measurement method also did not affect the reduction in diastolic blood pressure (supine office 4.34 [95%CI: 1.10, 7.57] vs seated office 1.30 mmHg [95%CI: –0.04, 2.64] vs standing office 4.78 [95%CI: 0.18, 9.38]). When only studies with individuals with hypertension were included in the meta-analysis, the reduction in diastolic blood pressure when measured by the seated office method became statistically significant (3.46 mmHg, 95%CI: 1.22, 5.70). Status of use of medication to control blood pressure had little effect on the reduction in diastolic blood pressure. The reduction in diastolic blood pressure with increased potassium intake was 1.37 mmHg (95%CI: 0.23, 2.50) in studies of individuals not taking medical therapy to control blood pressure, 3.80 mmHg (95%CI: –0.66, 8.25) in studies of individuals taking medication, and 2.32 mmHg (95%CI: 0.17, 4.46) in studies in which the participants were of undetermined status in relation to blood pressure. When only studies of individuals with hypertension were included in the meta-analysis, the diastolic blood pressure in those not taking medical therapy to control blood pressure was further reduced (4.54 mmHg, 95%CI: 0.69, 8.39). Type of intervention had little effect on the reduction in diastolic blood pressure (supplement 3.04 mmHg [95%CI: 0.99, 5.09] versus dietary advice or education 2.44 mmHg [95%CI: –0.17, 5.04]). When only individuals with hypertension were included in the metaanalysis, the decrease in diastolic blood pressure in studies administering intervention through dietary advice or education reached statistical significance (3.45 mmHg, 95%CI: 0.85, 6.05). In studies with a parallel design, the reduction in diastolic blood pressure in those with increased potassium intake was not statistically significant (1.59 mmHg, 95%CI: –0.18, 3.35). In studies with a cross-over design, the reduction in diastolic blood pressure with increased potassium intake was 4.21 mmHg (95%CI: 1.29, 7.12); this reduction was statistically significant, but was not statistically different from the parallel design group. These results were not altered in a meaningful way when only individuals with hypertension were included in the meta-analysis. 3.5.3 Ambulatory blood pressure – primary findings

In the four studies that measured ambulatory blood pressure, an increase in potassium intake resulted in a significant reduction of 3.04 mmHg (95%CI: 0.66, 5.42) in systolic blood pressure, and a non-significant reduction of 1.24 mmHg (95%CI: –0.66, 3.13) in diastolic blood pressure (Figures 3.18–3.19 and Tables 3.55–3.56). The reduction in ambulatory

systolic blood pressure was significant in studies targeting individuals with hypertension (3.37 mmHg, 95%CI: 0.69, 6.05). There were no studies targeting individuals with normal blood pressure, and the reduction was not significant in the one study of heterogeneous populations of individuals with normal blood pressure or hypertension (1.80 mmHg, 95%CI: –3.42, 7.02). The diastolic blood pressure change was non-significant, regardless of blood pressure status. There were no studies in which the potassium intake in the intervention group resulted in less than 70 mmol urinary potassium excretion/day. The one study with a potassium intake in the intervention group resulting in 70–90 mmol urinary potassium excretion/day reported a non-significant decrease in ambulatory systolic blood pressure (1.80 mmHg, 95%CI: –3.42, 7.02). In the two studies in which the potassium intake of the intervention resulted in 90– 120 mmol urinary potassium excretion/day, the reduction of systolic blood pressures was 3.65 mmHg (95%CI: 0.09, 7.21). Only one study had an intake resulting in more than 120 mmol urinary potassium excretion/day; it reported a non-significant decrease in ambulatory systolic blood pressure (3.00 mmHg, 95%CI: –1.07, 7.07) (Figure 3.20 and Table 3.56). 3.5.4 Ambulatory blood pressure – secondary findings

Ambulatory systolic blood pressure was significantly reduced (3.34 mmHg, 95%CI: 0.66, 5.52) by increased potassium intake when 24-hour ambulatory pressure was measured. Daytime ambulatory systolic blood pressure was reduced by 2.74 mmHg (95%CI: 0.28, 5.20), but night-time ambulatory systolic pressure was not reduced significantly with increased potassium (2.37 mmHg, 95%CI: –0.34, 5.09). Studies with a duration of intervention of less than 2 months had a significant reduction in blood pressure of 3.04 mmHg (95%CI: 0.66, 5.42). The one study with a duration of 2–4 months did not have a significant reduction in ambulatory systolic blood pressure (15.00 mmHg, 95%CI:–0.27, 29.73). All studies had a cross-over design. 3.5.5 Blood lipids

Total cholesterol

Total cholesterol was quantified in four studies with 104 participants, but one of these studies (Patki et al., 1990) was not included in the meta-analysis because it introduced excessive heterogeneity (Figure 3.21 and Table 3.57). Increased potassium intake relative to normal potassium intake had no significant effect on total cholesterol (MD: –0.12 mmol/L, 95%CI:-0.33,0.09). The Patki (1990) study reported a significant reduction in total cholesterol with increased potassium intake (1.40 mmol/L, 95%CI: 1.23, 1.56). This result was not affected by blood pressure status, achieved potassium or sodium intake, duration, medication status or study design. HDL cholesterol

HDL cholesterol was quantified in two studies with 64 participants. Increased potassium intake relative to normal potassium intake had no significant effect on HDL cholesterol (MD: –0.01 mmol/L, 95%CI: -0.13, 0.11) (Figure 3.22 and Table 3.57). LDL cholesterol

LDL cholesterol was quantified in one study with 48 participants. Increased potassium intake relative to normal potassium intake had no significant effect on LDL cholesterol (MD: –0.10 mmol/L, 95%CI: -0.38, 0.18) (Figure 3.23 and Table 3.57).

Total triglyceride concentration

Total triglyceride concentration was quantified in two studies with 64 participants. Increased potassium intake relative to normal potassium intake had no significant effect on total triglyceride concentration (-0.11mmol/L, 95%CI: -0.48, 0.26) (Figure 3.24 and Table 3.57). 3.5.6 Catecholamine levels

Plasma adrenaline concentrations were reported in three studies with 76 participants; these studies detected no effect of increased potassium intake on plasma adrenaline (MD:– 3.94 pg/mL, 95%CI: –9.22, 1.34) (Figure 3.25 and Table 3.58). These studies also reported on plasma noradrenaline concentration, and again detected no effect of increased potassium intake on plasma noradrenaline (MD:–4.32 pg/mL, 95%CI: –23.78, 15.13) (Figure 3.26 and Table 3.58). 3.5.7 Renal function

No included studies reported quantitative measurements for urinary protein excretion, protein:creatinine ratio, or urinary albumin excretion. Three studies quantified serum creatinine concentrations and found a non-significant decrease of 4.86 µmol/L (95%CI: – 0.39, 13.59) with increased potassium intake (Figure 3.27 and Table 3.59).

3.6

Sensitivity analysis and risk of bias A sensitivity analysis was performed to remove the studies considered to have a high risk of bias. Only one study was determined to be at high risk (Forrester & Grell, 1988). Removal of that study did not greatly alter the effect estimate for systolic (MD: –3.50 mmHg; 95%CI: – 5.15, –1.82) or diastolic (MD: –2.95 mmHg; 95%CI: –4.84, –1.06) blood pressure. 3.6.1 Quality of the body of evidence

Funnel plots for each of the main outcomes to assess publication bias indicated little risk of publication bias (Annex 3). The risk of bias summary (Annex 4) and risk of bias graph (Annex 5) suggested that the entire body of evidence is not at risk of serious problems due to bias. Several studies reported that personnel were not blinded, but most reported that participants were blinded. Blinding of outcome assessors was reported in half of the studies and was not mentioned in most of the remaining studies. There was no bias due to selective reporting or incomplete outcomes. One study was assessed as being at high risk of bias in randomization, allocation concealment and blinding. Most studies did not report on how randomization and allocation concealment were achieved. We generated GRADE evidence profiles for each of the specific objectives of the review. Each profile contained the assessment of the quality of evidence for all indicators of blood pressure, blood lipids, renal function and catecholamine levels (Annex 6). The evidence for increased potassium leading to a reduction in blood pressure was of high or moderate quality. The evidence of no effect of increased potassium on renal function, blood lipids and catecholamine levels was all of high quality. There was high- and moderate-quality evidence that increasing potassium intake to a level resulting in less than 70 mmol urinary excretion/day reduced blood pressure; however, that evidence came from only two studies, and should thus be treated with caution. There was moderate-quality evidence that increased potassium intake resulting in 70–90 mmol urinary excretion/day reduced blood pressure, and high-quality evidence that it had no effect on blood lipids, renal function and catecholamine levels; however, the evidence for these outcomes was all derived from one or two studies and should be treated with caution. There

was moderate-quality evidence that increased potassium intake resulting in 90–120 mmol urinary excretion/day reduced blood pressure, and high-quality evidence that it had no effect on renal function; however, these data came from only two studies. There was moderate-quality evidence that increased potassium intake resulting in greater than 120 mmol/day reduced blood pressure and high-quality evidence of no effect on cholesterol; however, that data came from only one study. There was also high-quality evidence of no effect on catecholamine levels.

3.7

Characteristics of studies In this section, tables are labelled by the study identifier (e.g. Barden BPARA 1986). The specific references for a particular study are listed below the table. The reference list in Chapter 4 also indicates which references are included in a particular study, and which is the primary reference. 3.7.1 Characteristics of included studies

Table 3.1 Barden BPARA1986 Methods Cross-over design study of increased K via supplements. Participants randomized to receive KCl supplement tablets or placebo. Conducted in Australia. 43 adult women, normotensive, not specified if taking BP medication. Group1 – placebo Group2 – K-supplemented diet (80 mmol K/day) via tablets Tablet type: KCl. Supine BP Treatment-period interaction 1) K intake in intervention ≥90 mmol/day 2) Na intake at baseline 2–4 g/d 3) Age – adult (15 years or greater) 4) Group – normotensive 5) Duration of follow-up – 1 month (4 weeks) 6) Sex – women only 7) BP device – automatic 8) BP method – supine office SBP/DBP, seated office SBP/DBP BP, blood pressure; Cl: chloride; DBP: diastolic blood pressure; K, potassium; Na, sodium References: (Barden et al., 1987; Barden et al., 1986)

Participants Interventions

Outcomes Notes

Table 3.2 Risk of bias table Barden BPARA1986 Bias Random sequence generation (selection bias) Allocation concealment (selection bias) Blinding of participants and personnel (performance bias) Blinding of outcome assessment (detection bias) Authors' judgement Unclear risk Unclear risk Unclear risk Unclear risk Support for judgement No description of method of sequence generation No description of method of concealment of allocation Providers were blinded but the blinding of participants was unclear and unlikely Unclear if outcome assessor was blinded Low loss to follow-up (2%) All outcomes reported

Incomplete outcome data (attrition Low risk bias) Selective reporting (reporting bias) Low risk

Table 3.3 Methods

Berry BPA2010 Cross-over design study of increased K education or advice to increase fruits and vegetables and via supplements. Participants randomized to receive K-citrate supplemented diet (fruit and vegetable diet), supplement tablets, or placebo. Conducted in the United Kingdom. 57 adult men and women, hypertensive status not specified, not taking BP medication. Group1 – K-supplemented diet (20 mmol K/day) of fruits and vegetables Group2 – K-supplemented diet (40 mmol K/day) of fruits and vegetables Group3 – K-supplemented diet (20 mmol K/day) via tablets Group4 – K-supplemented diet (40 mmol K/day) via tablets Group5 – unchanged diet (control) tablet type: K-citrate Resting BP Carotid to femoral pulse wave velocity Radial pulse wave analysis Flow-mediated dilation of the brachial artery Endothelial dilation in response to 25 µg glycerol trinitrate Serum total cholesterol, high-density lipoprotein-cholesterol, triacylglycerol, glucose Plasma adrenaline, plasma noradrenaline 1) K intake in intervention ≥70 mmol/day 2) Na intake at baseline 2–4 g/d 3) Age – adult (15 years or greater) 4) Group – both 5) Duration of follow-up – 1.5 months (6 weeks) 6) Sex – both (heterogeneous) 7) BP device – automatic 8) BP method – ambulatory SBP/DBP (24-hour/day/night), supine DBP/SBP

Participants Interventions

Outcomes

Notes

BP, blood pressure; DBP, diastolic blood pressure; K, potassium; Na, sodium; SBP, systolic blood pressure Reference: (Berry et al., 2010)

Table 3.4 Risk of bias table Berry BPA2010 Bias Random sequence generation (selection bias) Allocation concealment (selection bias) Blinding of participants and personnel (performance bias) Blinding of outcome assessment (detection bias) Incomplete outcome data (attrition bias) Selective reporting (reporting bias) Authors' judgement Support for judgement Low risk Low risk Low risk Low risk Low risk Low risk Randomized via computer algorithm Computer allocation Participants blinded Outcome assessor was blinded 16% loss to follow-up All outcomes reported

Table 3.5 Bulpitt BPA1985 Methods Parallel design study of increased K via supplements. Participants randomized to receive placebo or KCl supplement. Conducted in the United Kingdom. 33 adult men and women, hypertensive, taking BP medication Group1 – placebo Group2 – K-supplemented diet (64 mmol K/day) via tablets Type: KCl Changes in resting BP Change in medication dosage 1) K intake in intervention ≥90 mmol/day 2) Na intake at baseline 2–4 g/d 3) Age – adult (15 years or greater) 4) Group – hypertensive 5) Duration of follow-up – 3 months (12 weeks) 6) Sex – both (heterogeneous) 7) BP device – manual 8) BP method – SBP (unspecified resting), DBP (unspecified resting) BP, blood pressure; Cl, chloride; DBP, diastolic blood pressure; K, potassium; Na, sodium; SBP, systolic blood pressure Reference: (Bulpitt et al., 1985)

Participants Interventions

Outcomes Notes

Table 3.6 Risk of bias table Bulpitt BPA1985 Bias Authors' judgement Support for judgement No description of method of sequence generation No description of method of concealment of allocation Not clear how or if participants and providers blinded No description of blinding of outcome assessor 2% loss to follow-up All outcomes reported

Random sequence generation (selection Unclear risk bias) Allocation concealment (selection bias) Blinding of participants and personnel (performance bias) Blinding of outcome assessment (detection bias) Incomplete outcome data (attrition bias) Selective reporting (reporting bias) Unclear risk Unclear risk Unclear risk Low risk Low risk

Table 3.7 Chalmers BPA1986 Methods Participants Interventions Outcomes Parallel design study of increased K diet. Participants randomized to receive normal diet or high-K diet. Conducted in Australia. 107 adult men and women, hypertensive, not taking BP medication Group1 – normal diet Group2 – high K diet through diet advice or education Resting BP Urinary Na, K, creatinine excretion Serum K, creatinine, cholesterol, gamma-glutamyl transferase 1) K intake in intervention ≥90 mmol/day 2) Na intake at baseline 2–4 g/d 3) Age – adult (15 years or greater) 4) Group – hypertensive 5) Duration of follow-up – 3 months (12 weeks) 6) Sex – both (heterogeneous) 7) BP device – automatic 8) BP method – seated office BP, blood pressure; K, potassium; Na, sodium Reference: (Chalmers et al., 1986)

Notes

Table 3.8 Risk of bias table Chalmers BPA1986 Bias Authors' judgement Support for judgement No description of method of sequence generation No description of method of concealment of allocation Participants and providers not blinded Outcome assessor not blinded Loss to follow-up not reported All outcomes reported

Table 3.9 Forrester BPA1988 Methods Cross-over design study of increased K via supplements. Participants randomized to receive bendroflumethiazide or bendroflumethiazide + KCl. Conducted in Jamaica. 23 adult men and women, hypertensive status, taking BP medication Group1 – bendroflumethiazide Group2 – bendroflumethiazide + K-supplement (600 mg K) via tablets (type: KCl) Resting BP Mean BP Serum K Blood glucose Serum urate Urine Na Urine K Red cell Na Red cell K 1) K intake in intervention ≥60 mmol/day 2) Na intake at baseline 2–4 g/d 3) Age – adult (15 years or greater) 4) Group – hypertensive 5) Duration of follow-up – 1 month (4 weeks) 6) Sex – both (heterogeneous) 7) BP device – manual 8) BP method – Supine office SBP, supine office DBP, standing office SBP, standing office DBP BP, blood pressure; Cl, chloride; DBP, diastolic blood pressure; K, potassium; Na, sodium; SBP, systolic blood pressure Reference: (Forrester & Grell, 1988)

Participants Interventions

Outcomes

Notes

Table 3.10 Bias

Risk of bias table Forrester BPA1988 Authors' judgement High risk High risk High risk Unclear risk Low risk Low risk Support for judgement Patients were chosen from clinic by volunteering Allocation not concealed Participants and providers not blinded No description of blinding of outcome assessor 4% loss to follow-up reported All outcomes reported

Random sequence generation (selection bias) Allocation concealment (selection bias) Blinding of participants and personnel (performance bias) Blinding of outcome assessment (detection bias) Incomplete outcome data (attrition bias) Selective reporting (reporting bias)

Table 3.11 Methods

Fotherby BPA1992 Cross-over design study of increased K via elixir (supplement). Participants randomized to receive KCl supplement tablets or placebo. Conducted in the United Kingdom. 18 adult men and women, hypertensive, not taking BP medication Group1 – placebo Group2 – K-supplemented diet (60 mmol K/day) via elixir Elixir type: KCl Resting BP Pulse rate Ambulatory BP (24-hour/day/night) Serum electrolytes Creatinine Plasma renin activity Urinary electrolytes Body weight changes 1) Potassium intake in intervention ≥90 mmol/day 2) Na intake at baseline 2–4 g/d 3) Age – adult (15 years or greater) 4) Group – hypertensive 5) Duration of follow-up – 1 month (4 weeks) 6) Sex – both (heterogeneous) 7) BP device – automatic 8) BP method – ambulatory SBP (24-hour/day/night), ambulatory DBP (24hour/day/night), supine office systolic BP, supine office DBP, standing office SBP, standing office DBP

Participants Interventions

Outcomes

Notes

BP, blood pressure; Cl, chloride; DBP, diastolic blood pressure; K, potassium; Na, sodium; SBP, systolic blood pressure References: (Fotherby & Potter, 1992; Fotherby & Potter, 1997)

Table 3.12 Bias

Risk of bias table Fotherby BPA1992 Authors' judgement Unclear risk Unclear risk Low risk Support for judgement Random sequence generation not described Allocation concealment not described Both participants and personnel blinded Outcome assessor blinded Low loss to follow-up (0%) All outcomes reported

Random sequence generation (selection bias) Allocation concealment (selection bias) Blinding of participants and personnel (performance bias)

Blinding of outcome assessment (detection Low risk bias) Incomplete outcome data (attrition bias) Selective reporting (reporting bias) Low risk Low risk

Table 3.13 Methods

Grobbee BPA1987 Cross-over design study of increased K via supplements. Participants randomized to receive KCl supplement tablets or placebo. Conducted in the Netherlands. 40 adult men and women, hypertensive, not specified if taking BP medication Group1 – placebo Group2 – K-supplemented diet (8 mmol K) via slow-release tablets Tablet type: slow-KCl Urinary excretion SBP, DBP Urinary electrolyte excretion Body weight Pulse rate Plasma catecholamine levels Plasma renin Cardiac output Cardiac index 1) Potassium intake in intervention ≥120 mmol/day 2) Na intake at baseline 2–4 g/d 3) Age – adult (15 years or greater) 4) Group – hypertensive 5) Duration of follow-up – 1.5 months (6 weeks) 6) Sex – both (heterogeneous) 7) BP device – manual 8) BP method – supine office

Participants Interventions

Outcomes

Notes

BP, blood pressure; Cl: chloride; K, potassium; Na, sodium Reference: (Grobbee et al., 1987)

Table 3.14 Bias

Risk of bias table Grobbee BPA1987 Authors' judgement Support for judgement No description of method of sequence generation No description of method of concealment of allocation Participants and providers blinded No description of blinding of outcome assessor No loss to follow-up reported All outcomes reported

Table 3.15 Methods Participants Interventions

Gu BPA2001 Parallel study of increased K via supplements. Participants randomized to receive KCl supplement tablets or placebo. Conducted in China. 43 adult women, heterogeneous hypertensive status, not taking BP medication Group1 – placebo Group2 – K-supplemented diet (60 mmol K/day) via tablets Tablet type: KCl Resting BP Body weight 1) K intake intervention <70 mmol/day (~57 mmol/day) 2) Na intake at baseline >4 g/d 3) Age – adult (15 years or greater) 4) Group – both 5) Duration of follow-up – 3 months (12 weeks) 6) Sex – women only 7) BP device – manual 8) BP method – seated SBP, seated DBP 9) Subgroup analysis – 6-week timepoint

Outcomes Notes

BP, blood pressure; Cl, chloride; DBP, diastolic blood pressure; Na, sodium; SBP, systolic blood pressure Reference: (Gu et al., 2001)

Table 3.16 Bias

Risk of bias table Gu BPA2001 Authors' judgement Low risk Low risk Low risk Low risk Low risk Low risk Support for judgement Generated by computer program Concealed in an ordered set of sealed envelopes Participants and providers blinded Outcome assessor blinded Low loss to follow-up reported (<10%) All outcomes reported

Table 3.17 Methods Participants Interventions

He BPA2010

Cross-over design study of increased K via supplements. Participants randomized to receive supplement tablets or placebo. Conducted in the United Kingdom. 46 adult men and women, hypertensive, not taking BP medication Group1 – placebo Group2 – K-supplemented diet (64 mmol K/day) via tablets (type KCl) Group3 – K-supplemented diet (64 mmol K/day) via tablets (type KHCO3) Tablet type: KCl and KHCO3 Resting BP Ambulatory BP 24-hour albumin excretion Pulse wave velocity Vascular function: changes in left ventricular geometry and function Change in bone metabolism markers: urinary calcium, pH 1) K intake in intervention ≥120 mmol/day 2) Na intake at baseline 2–4 g/day 3) Age – adult (15 years or greater) 4) Group – hypertensive 5) Duration of follow-up – 1 month (4 weeks) 6) Sex – both (heterogeneous) 7) BP device – automatic 8) BP method – SBP/DBP (unspecified resting) , ambulatory SBP/DBP (24hour/day/ night)

Outcomes

Notes

BP, blood pressure; Cl, chloride; DBP, diastolic blood pressure; HCO3, bicarbonate; K, potassium; Na, sodium; SBP, systolic blood pressure Reference: (He et al., 2010)

Table 3.18 Bias

Risk of bias table He BPA2010 Authors' Support for judgement judgement Computer generated program by independent company Allocated in random order to take intervention capsules or placebo capsules; all were blinded to treatment allocation Participants and providers blinded Outcome assessor blinded Low loss to follow-up reported (<10%) All outcomes reported

Random sequence generation Low risk (selection bias) Allocation concealment (selection bias) Low risk

Blinding of participants and Low risk personnel (performance bias) Blinding of outcome assessment (detection bias) Incomplete outcome data (attrition bias) Low risk Low risk

Selective reporting (reporting Low risk bias)

Table 3.19 Methods

Kaplan BPA1985 Cross-over design study of increased K via supplements. Participants randomized to receive KCl supplement tablets or placebo. Conducted in the United States of America. 16 adult men and women, hypertensive, taking BP medication Group1 – placebo Group2 – K-supplemented diet (60 mmol K/day) via tablets Tablet type: KCl Resting BP Serum K and Na levels Plasma renin activity Plasma aldosterone Body weight 1) Potassium intake in intervention ≥70 mmol/day 2) Na intake at baseline 2–4 g/d 3) Age – adult (15 years or greater) 4) Group – hypertensive 5) Duration of follow-up – 1.5 months (6 weeks) 6) Sex – both (heterogeneous) 7) BP device – manual 8) BP method – supine office SBP, supine office DBP

Participants Interventions

Outcomes

Notes

BP, blood pressure; Cl, chloride; DBP, diastolic blood pressure; K, potassium; Na, sodium; SBP, systolic blood pressure Reference: (Kaplan et al., 1985)

Table 3.20 Bias

Risk of bias table Kaplan BPA1985 Authors' judgement Support for judgement No description of method of sequence generation No description of method of concealment of allocation Participants and providers blinded Outcome assessor blinded 20% loss to follow-up reported All outcomes reported

Table 3.21 Methods

Kawano BPA1998 Cross-over design study of increased K via supplements. Participants randomized to receive KCl supplement tablets or placebo. Conducted in Japan. 55 adult men and women, hypertensive, heterogeneous medication status Group1 – placebo Group2 – K-supplemented diet (64 mmol K/day) via tablets Tablet type: KCl Resting BP Ambulatory BP Serum and urinary electrolytes 1) K intake in intervention ≥90 mmol/day 2) Na intake at baseline >4 g/day 3) Age – adult (15 years or greater) 4) Group – hypertensive 5) Duration of follow-up – 1 month (4 weeks) 6) Sex – both (heterogeneous) 7) BP device – automatic (ambulatory), manual (resting) 8) BP method – ambulatory SBP/DBP (24-hour/day/night), seated office SBP, seated office DBP

Participants Interventions

Outcomes

Notes

BP, blood pressure; Cl, chloride; DBP, diastolic blood pressure; K, potassium; Na, sodium; SBP, systolic blood pressure Reference: (Kawano et al., 1998)

Table 3.22 Bias

Risk of bias table Kawano BPA1998 Authors' judgement Support for judgement No description of method of sequence generation No description of method of concealment of allocation Unclear if participants and providers blinded No description of blinding of outcome assessor Low loss to follow-up reported (<10%) All outcomes reported

Table 3.23 Methods

MacGregor AEBPA1982 Cross-over design study of increased K via supplements. Participants randomized to receive KCl tablets or placebo. Conducted in the United Kingdom. 23 hypertensive men and women, not taking BP medication Group1 – control (placebo) Group2 – supplement (64 mmol K/day) Tablet type: KCl Resting BP Pulse rate Weight 24-hour urinary sodium, potassium, creatinine, urea, creatinine, electrolytes, plasma renin activity, aldosterone 1) K intake in intervention ≥90 mmol/day 2) Na intake at baseline 2–4 g/day 3) Age – adult (15 years or greater) 4) Group – hypertensive 5) Duration of follow-up – 1 month (4 weeks) 6) Sex – both (heterogeneous) 7) BP device – automatic 8) BP method – supine office DBP, supine office SBP, standing office SBP, standing office DBP

Participants Interventions

Outcomes

Notes

BP, blood pressure; Cl, chloride; DBP, diastolic blood pressure; K, potassium; Na, sodium; SBP, systolic blood pressure References: (MacGregor et al., 1982; MacGregor et al., 1984; Smith et al., 1985)

Table 3.24 Bias

Risk of bias table MacGregor AEBPA1982 Authors' judgement Support for judgement No description of method of sequence generation No description of method of concealment of allocation Participants and providers blinded No description of blinding of outcome assessor 0% loss to follow-up reported All outcomes reported

Table 3.25 Methods Participants Interventions

Matlou BPA1986 Cross-over design study of increased K via supplements. Participants randomized to receive KCl tablets or placebo. Conducted in South Africa. 32 hypertensive women, not taking BP medication Group1 – control (placebo) Group2 – supplement (65 mmol K/day) Tablet type: KCl Resting BP Serum K, Na Urinary K, Na 1) K intake in intervention ≥90 mmol/day 2) Na intake at baseline 2–4 g/d 3) Age – adult (15 years or greater) 4) Group – hypertensive 5) Duration of follow-up – 1.5 months (6 weeks) 6) Sex – women only 7) BP device – manual 8) BP method – seated office SBP, seated office DBP 9) Subgroup analyses – 4 week time point

Outcomes

Notes

BP, blood pressure; Cl, chloride; DBP, diastolic blood pressure; K, potassium; Na, sodium; SBP, systolic blood pressure Reference: (Matlou et al., 1986)

Table 3.26 Bias

Risk of bias table Matlou BPA1986 Authors' judgement Support for judgement No description of method of sequence generation No description of method of concealment of allocation Participants and providers were not blinded Reported that observer bias was eliminated Low loss to follow-up reported (10%) All outcomes reported

Table 3.27 Methods

Obel BPA1989 Parallel design study of increased K via supplements. Participants randomized to receive KCl tablets or placebo. Conducted in the United Kingdom. 48 hypertensive men and women, not taking BP medication Group1 – control (placebo) Group2 – supplement (64 mmol K/day) Tablet type: KCl Standing BP Supine BP 1) K intake in intervention ≥90 mmol/day 2) Na intake at baseline 2–4 g/d 3) Age – adult (15 years or greater) 4) Group – hypertensive 5) Duration of follow-up – 4 months (16 weeks) 6) Sex – both (heterogeneous) 7) BP device – manual 8) BP method – supine office DBP, supine office SBP, standing DBP, standing office SBP 9) Subgroup analysis – 4-, 8- and 12-week time points

Participants Interventions

Outcomes Notes

BP, blood pressure; Cl, chloride; DBP, diastolic blood pressure; K, potassium; Na, sodium; SBP, systolic blood pressure Reference: (Obel, 1989)

Table 3.28 Bias

Risk of bias table Obel BPA1989 Authors' judgement Unclear risk Unclear risk Low risk Unclear risk Low risk Low risk Support for judgement No description of method of sequence generation No description of method of concealment of allocation Participants and providers blinded No description of blinding of outcome assessor 0% loss to follow-up reported Most outcomes reported; urinary K and Na not reported for subanalyses (4, 8, 12 week time points)

Table 3.29 Methods Participants Interventions

Overlack BPARAAEA1991 Cross-over design study of increased K via supplements. Participants randomized to receive K tablets or placebo. Conducted in Germany. 12 hypertensive men and women, not taking BP medication Group1 – control (placebo) Group2 – supplement (120 mmol K/day) Tablet type: K + citrate + bicarbonate Mean arterial BP Serum Na, K concentrations Intracellular Na, K concentrations Plasma renin activity Plasma aldosterone 1) K intake in intervention ≥120 mmol/day 2) Na intake at baseline 2–4 g/d 3) Age – adult (15 years or greater) 4) Group – hypertensive 5) Duration of follow-up – 2 months (8 weeks) 6) Sex – both (heterogeneous) 7) BP device – not specified 8) BP method – SBP, DBP 9) Other – Does not contribute to meta-analyses. Waiting on author reply

Outcomes

Notes

BP, blood pressure; Cl, chloride; DBP, diastolic blood pressure; K, potassium; Na, sodium; SBP, systolic blood pressure Reference: (Overlack et al., 1991)

Table 3.30 Bias

Risk of bias table Overlack BPARAAEA1991 Authors' judgement Unclear risk Support for judgement No description of method of sequence generation; stated that there was a "randomization plan" No description of method of concealment of allocation Participants and providers not blinded Outcome assessor blinded 0% loss to follow-up reported All outcomes reported

Unclear risk High risk Low risk Low risk Low risk

Table 3.31 Methods Participants Interventions

Patki BPARA1990 Cross-over design study of increased K via supplements. Participants randomized to receive KCl tablets or placebo. Conducted in India. 37 hypertensive men and women, not taking BP medication Group1 – control (placebo) Group2 – supplement (30 mmol K/day) Tablet type: KCl Resting BP Serum cholesterol Serum creatinine Serum urea 1) K intake in intervention ≥70 mmol/day 2) Na intake at baseline >4 g/day 3) Age – adult (15 years or greater) 4) Group – hypertensive 5) Duration of follow-up – 2 months (8 weeks) 6) Sex – both (heterogeneous) 7) BP device – manual 8) BP method – supine SBP, supine DBP, standing office SBP, standing office DBP

Outcomes

Notes

BP, blood pressure; Cl, chloride; DBP, diastolic blood pressure; K, potassium; Na, sodium; SBP, systolic blood pressure Reference: (Patki et al., 1990)

Table 3.32 Bias

Risk of bias table Patki BPARA1990 Authors' judgement Support for judgement No description of method of sequence generation No description of method of concealment of allocation Participants and providers blinded No description of blinding of outcome assessor Loss to follow-up not reported All outcomes reported

Table 3.33 Methods Participants Interventions

Richards BPAAEA1984 Cross-over design study of increased K via supplements. Participants randomized to receive KCl tablets or placebo. Conducted in New Zealand. 12 hypertensive men and women, not taking BP medication Group1 – control (placebo) Group2 – supplement (200 mmol K/day) Tablet type: KCl Resting BP Plasma renin activity Anguitensin II Aldosterone Noradrenaline Adrenaline Urine Na, K, creatinine excretions 1) K intake in intervention ≥120 mmol/day 2) Na intake at baseline >4 g/d (estimate based on figure) 3) Age – adult (15 years or greater) 4) Group – hypertensive 5) Duration of follow-up – 1 month (4 weeks) 6) Sex – both (heterogeneous) 7) BP device – automatic 8) BP method – supine SBP, supine DBP, standing office SBP, standing office DBP

Outcomes

Notes

BP, blood pressure; Cl, chloride; DBP, diastolic blood pressure; K, potassium; Na, sodium; SBP, systolic blood pressure Reference: (Richards et al., 1984)

Table 3.34 Bias

Risk of bias table Richards BPAAEA1984 Authors' judgement Support for judgement No description of method of sequence generation No description of method of concealment of allocation Providers and participants were not blinded Unclear if outcome assessors were blinded 25% loss to follow-up All outcomes reported

Table 3.35 Methods Participants Interventions

Siani BPA1987 Parallel study of increased K via supplements. Participants randomized to receive KCl tablets or placebo. Conducted in Italy. 37 hypertensive men and women, not taking BP medication Group1 – control (placebo) Group2 – supplement (24 mmol K/day) Tablet type: KCl Resting BP Urinary creatinine excretion 1) K intake in intervention ≥70 mmol/day 2) Na intake at baseline >4 g/d 3) Age – adult (15 years or greater) 4) Group – hypertensive 5) Duration of follow-up – 3.75 months (7 weeks) 6) Sex – both (heterogeneous) 7) BP device – manual 8) BP method – supine office SBP, supine office DBP, standing office SBP, standing office DBP

Outcomes Notes

BP, blood pressure; Cl, chloride; DBP, diastolic blood pressure; K, potassium; Na, sodium; SBP, systolic blood pressure Reference: (Siani et al., 1987)

Table 3.36 Bias

Risk of bias table Siani BPA1987 Authors' judgement Unclear risk Low risk Low risk Low risk Unclear risk Low risk Support for judgement No description of method of sequence generation Concealment of allocation through prepackaged identical containers Providers and participants were blinded Outcome assessor was blinded Loss to follow-up not reported All outcomes reported

Table 3.37 Methods Participants Interventions

Siani BPARA1991 Parallel design study of increased K diet. Participants randomized to receive regular (unchanged) diet or high K diet. Conducted in Italy. 47 hypertensive men and women, taking BP medication Group1 – control (unchanged diet) Group2 – high K diet (≥30 mmol K/day) Tablet type: N/A Resting BP Mean energy and nutrient intake Urinary K and Na Rate of treatment discontinuation Pills per day (drug consumption) 1) K intake in intervention ≥70 mmol/day 2) Na intake at baseline 2–4 g/day 3) Age – adult (15 years or greater) 4) Group – hypertensive 5) Duration of follow-up – 12 months (52 weeks) 6) Sex – both (heterogeneous) 7) BP device – automatic 8) BP method – Supine office SBP, supine office DBP

Outcomes

Notes

BP, blood pressure; Cl, chloride; DBP, diastolic blood pressure; K, potassium; Na, sodium; SBP, systolic blood pressure Reference: (Siani et al., 1991)

Table 3.38 Bias

Risk of bias table Siani BPARA1991 Authors' judgement Support for judgement No description of method of sequence generation No description of method of concealment of allocation Providers and participants were not blinded Unclear if outcome assessors were blinded 13% loss to follow-up All outcomes reported

Table 3.39 Methods

Smith BPARA1985 Cross-over design study of increased K via supplements. Participants randomized to receive KCl tablets or placebo. Conducted in the United States of America. 20 hypertensive men and women, not taking BP medication Group1 – control (placebo) Group2 – supplement (64 mmol K/day) Tablet type: KCl Resting BP Supine heart rate Weight Urinary Na, K, creatinine Plasma renin activity Plasma aldosterone, K Blood creatinine 1) K intake in intervention ≥90 mmol/day 2) Na intake at baseline <2 g/d 3) Age – adult (15 years or greater) 4) Group – hypertensive 5) Duration of follow-up – 1 month (4 weeks) 6) Sex – both (heterogeneous) 7) BP device – automatic 8) BP method – Supine SBP, supine DBP, standing SBP, standing DBP

Participants Interventions

Outcomes

Notes

BP, blood pressure; Cl, chloride; DBP, diastolic blood pressure; K, potassium; Na, sodium; SBP, systolic blood pressure Reference: (Smith et al., 1985)

Table 3.40 Bias

Risk of bias table Smith BPARA1985 Authors' judgement Support for judgement No description of method of sequence generation No description of method of concealment of allocation Providers and participants were blinded Outcome assessors was blinded 10% loss to follow-up All outcomes reported

Table 3.41 Methods

Trial Hyp Prv Col BPA1992 Parallel design study of increased K via supplements. Participants randomized to receive KCl tablets or placebo. Conducted in the United States of America. 286 normotensive men and women, not taking BP medication Group1 – control (placebo) Group2 – supplement (60 mmol K/day) Tablet type: KCl Resting BP Na reduction Weight reduction Stress management Incidence of hypertension 1) K intake in intervention ≥90 mmol/day 2) Na intake at baseline 2–4 g/d 3) Age – adult (15 years or greater) 4) Group – normotensive 5) Duration of follow-up – 6 months (24 weeks) 6) Sex – both (heterogeneous) 7) BP device – manual 8) BP method – Seated office SBP, seated office DBP 9) Subgroup analysis – 3-month time point

Participants Interventions

Outcomes

Notes

BP, blood pressure; Cl, chloride; DBP, diastolic blood pressure; K, potassium; Na, sodium; SBP, systolic blood pressure Reference: (Trial Hyp Prv Col, 1992)

Table 3.42 Bias

Risk of bias table Trial Hyp Prv Col BPA1992 Authors' judgement Low risk Low risk Low risk Low risk Low risk Low risk Support for judgement Sequence generation by telephone Sealed opaque envelopes Providers and participants were blinded Outcome assessors were blinded 18–20% loss to follow-up among groups All outcomes reported

Table 3.43 Methods Participants Interventions

Valdes BPA1991 Cross-over design study of increased K via supplements. Participants randomized to receive KCl tablets or placebo. Conducted in Chile. 24 hypertensive men and women, not taking BP medication Group1 – control (placebo) Group2 – supplement (64 mmol K/day) Tablet type: KCl Mean arterial BP Serum Na, K concentrations Intracellular Na, K concentrations Plasma renin activity Plasma aldosterone activity 1) K intake in intervention ≥120 mmol/day 2) Na intake at baseline 2–4 g/d 3) Age – adult (15 years or greater) 4) Group – hypertensive 5) Duration of follow-up – 1 month (4 weeks) 6) Sex – both (heterogeneous) 7) BP device – automatic 8) BP method – supine office SBP, supine office DBP, standing office SBP, standing office DBP

Outcomes

Notes

BP, blood pressure; Cl, chloride; DBP, diastolic blood pressure; K, potassium; Na, sodium; SBP, systolic blood pressure Reference: (Valdés et al., 1991)

Table 3.44 Bias

Risk of bias table Valdes BPA1991 Authors' judgement Support for judgement No description of method of sequence generation No description of method of concealment of allocation Providers and participants were blinded Unclear if outcome assessors were blinded No loss to follow-up reported All outcomes reported

Table 3.45 Methods

Whelton BPA1995 Parallel design study of increased K via supplements. Participants randomized to receive KCl tablets or placebo. Conducted in the United States of America. 353 normotensive men and women, not taking BP medication Group1 – control (placebo) Group2 – supplement (60 mmol K/day) Tablet type: KCl BP Urinary excretion Dietary assessment 1) K intake in intervention ≥90 mmol/day 2) Na intake at baseline 2–4 g/d 3) Age – adult (15 years or greater) 4) Group – normotensive 5) Duration of follow-up – 6 months (24 weeks) 6) Sex – both (heterogeneous) 7) BP device – manual 8) BP method – SBP, DBP 9) Subgroup analysis – 3-month time point

Participants Interventions

Outcomes

Notes

BP, blood pressure; Cl, chloride; DBP, diastolic blood pressure; K, potassium; Na, sodium; SBP, systolic blood pressure References: (Whelton et al., 1997; Whelton et al., 1995)

Table 3.46 Bias

Risk of bias table Whelton BPA1995 Authors' judgement Support for judgement No description of method of sequence generation No description of method of concealment of allocation Providers and participants were blinded Outcome assessors were blinded 4–10% loss to follow-up among groups of interest All outcomes reported

3.7.2

Characteristics of studies awaiting classification Braschi BPA2008

Table 3.47 Methods Participants

Parallel design study of increased K diet. Participants randomized to receive placebo, K-citrate supplement, or KCl supplement. Conducted in the United Kingdom. 90 adult men and women, heterogeneous hypertensive status, heterogeneous BP medication status

Interventions Group1 – placebo Group2 – K-supplemented diet (30 mmol K/day) via tablets (type: KCl) Group3 – K-supplemented diet (30 mmol K/day) via tablets (type: K-citrate) Outcomes Notes BP Urinary electrolyte and creatinine, haematocrit, erythrocyte water and K content 1) K intake in intervention ≥70 mmol/day 2) Na intake at baseline 2–4 g/d 3) Age – adult (15 years or greater) 4) Group – both 5) Duration of follow-up – 1.5 months (6 weeks) 6) Sex – both (heterogeneous) 7) BP device – automatic 8) BP method – seated office SBP, seated office DBP BP, blood pressure; Cl, chloride; DBP, diastolic blood pressure; K, potassium; Na, sodium; SBP, systolic blood pressure Reference: (Braschi & Naismith, 2008)

Table 3.48 Methods Participants

Cushman BPA1988

Parallel design study of increased K diet. Participants randomized to receive placebo or KCl supplement. Conducted in the United States of America. 58 adult men, hypertensive status, not taking BP medication

Interventions Group1 – placebo Group2 – K-supplemented diet (80 mmol K/day) via tablets Tablet type: KCl Outcomes Notes Urinary K excretion BP (type not specified and data in current form not usable for meta-analysis) 1) Potassium intake in intervention ≥90 mmol/day 2) Na intake at baseline not reported 3) Age – adult (15 years or greater) 4) Group – hypertensive 5) Duration of follow-up – 2.5 months 6) Sex – men 7) BP device – not specified 8) BP method – SBP, DBP BP, blood pressure; Cl, chloride; DBP, diastolic blood pressure; K, potassium; Na, sodium; SBP, systolic blood pressure Reference: (Cushman & Langford, 1988)

Table 3.49 Methods

Hilary Green BPA2000 Cross-over design study of increased K diet. Participants randomized to receive high-calcium milk or high-calcium milk enriched with K. Conducted in New Zealand. 38 adult men and women, heterogeneous hypertensive status, not taking BP medication Group1 – high-calcium milk Group2 – K-supplemented high-calcium milk (1585 mg K/50 g milk) Ambulatory BP Resting BP Excretion of calcium, K, Na, magnesium 1) K intake (as measured by urinary excretion) – Author contacted 2) Na intake at baseline not reported 3) Age – adult (15 years or greater) 4) Group – both 5) Duration of follow-up – 1 month (4 weeks) 6) Sex – both (heterogeneous) 7) BP device – automatic 8) BP method – ambulatory SBP (day), ambulatory DBP (day), seated office DBP, seated office SBP, standing office SBP, standing office DBP

Participants Interventions Outcomes

Notes

BP, blood pressure; DBP, diastolic blood pressure; K, potassium; Na, sodium; SBP, systolic blood pressure Reference: (Hilary Green et al., 2000)

The full text was unavailable for the following studies: • • • • Barros BPA1984 – reference: Barros and Brito (1984); Iimura BPA1979 – reference: Iimura et al. (1981); Kawano BPA1997 – reference: Kawano et al. (1997); Morris BPA 1995 – reference: Morris et al. (1995).

3.7.3

Characteristics of ongoing studies

Table 3.50 Study name

Mullan BPA2010 The renin-angiotensin-aldosterone axis, endothelial function and hypertension: diagnostic strategies, and therapeutic role of K supplementation – a randomized cross-over trial and an observational study. WHO International Clinical Trials Registry Platform. [Other: ISRCTN55798944] Study 1: Prospective randomized cross-over investigator-blinded trial Study 2: Observational study Study 1: 1.1. Patients (both male and female) aged 40–70 with moderate (>10%) cardiovascular disease risk (Joint British Societies' guidelines) Study 2: Three groups of participants (both male and female, all between age 18– 70): 2.1. Patients with essential hypertension 2.2. Patients with hyperaldosteronism 2.3. Healthy volunteers

Methods Participants

Interventions

Study 1: Dietary intake of Na and K will be assessed and a 24-hour urine collection for estimation of Na and K excretion will be taken. The participants will be randomized to either placebo or K supplementation (4.8 g/day; oral) for 6 weeks. There will be a 6-week washout period. Participants will be studied at baseline, at the end of washout and after each intervention (four in total). Study 2: The target numbers of participants for the three groups are as follows: i. Patients with essential hypertension: n = 20 ii. Patients with hyperaldosteronism: n = 8 iii. Healthy volunteers: n = 15 43 participants in total Patients with essential hypertension and patients with hyperaldosteronism will undergo a 250-mcg synacthen test with blood and saliva sampled at 0, 30 and 60 minutes for cortisol and aldosterone. This test will be performed after 30 minutes of recumbency. In addition the healthy volunteers and patients with hyperaldosteronism will also undergo a GnRH test to assess aldosterone response.

Outcomes

Study 1 – Global endothelial function assessed by determining the change in augmentation index in response to the administration of nitroglycerin and salbutamol. All primary and secondary outcomes will be assessed at baseline, at the end of washout and after each intervention. Study 2 – Aldosterone response to synacthen/GnRH

Starting date Contact information Notes

20/11/2008 karen.mullan@hscni.net http://www.controlled-trials.com/ISRCTN55798944/ Expected to have ended in 2010, but no publications listed

GnRH, gonadotropin-releasing hormone; K, potassium; Na, sodium

Table 3.51 Study name Methods

Turban BPA RA2009 Potassium intake in patients with chronic kidney disease. Clinicaltrials.gov. [Other: NCT00949585] Allocation: randomized Endpoint classification: safety/efficacy study Intervention model: cross-over assignment Masking: double-blind (patient, caregiver, investigator, outcomes assessor) Primary purpose: treatment Inclusion criteria: Stage 3 chronic kidney disease (estimated glomerular filtration rate 30– 2 59 mL/min/1.73 m by the 4-variable Modification of Diet in Renal Disease Study Equation Systolic BP 120–159 mmHg and diastolic BP <100 mmHg Willingness to follow strict dietary rules for 9 weeks and to come to the clinical research unit at least 3 weekdays per week for one meal during the two study periods Exclusion criteria: Baseline serum K of at least 5 mEq/L (or history of hyperkalemia) Baseline serum K of less than 3.5 mEq/L (or history of hypokalemia) Insulin-requiring or uncontrolled (HbA1C >9 g/dL) diabetes mellitus Use of K-sparing diuretics, K supplements, non-steroidal anti-inflammatory agents, steroids, digoxin, or calcineurin inhibitors History of any organ transplant 2 Body mass index >40 kg/m Chronic disease(s) that may interfere with trial participation Pregnancy or lactation >14 alcoholic drinks/week Major food allergies or intolerances Participants will be given one of two diets: one contains 100 mmol K/day, the other contains 40 mmol K/day: Primary outcome measures: 24-hour ambulatory systolic BP [Time frame: at screening, and at the end of each intervention period] [Designated as safety issue: Yes] Secondary outcome measures: other measures of peripheral BP (other types of ambulatory BP measurements as well as clinic BP) [Time frame: ambulatory BP: same as primary outcome. Clinic BP: at screening, run-in, and weekly during intervention] [Designated as safety issue: Yes] Measures of central BP (pulse wave velocity and augmentation index) [Time frame: at screening and at the end of each intervention period] [Designated as safety issue: No] Serum K [Time frame: at screening, run-in, and at least three times during each intervention period] [Designated as safety issue: Yes] Serum inflammatory markers [Time frame: before and at the end of each intervention period] [Designated as safety issue: No]

Participants

Interventions Outcomes

Starting date Contact information Notes

July 2009 CKD@jhmi.edu http://clinicaltrials.gov/ct2/show/NCT00949585

BP, blood pressure; K, potassium

3.7.4

Excluded studies and reasons for exclusion

The excluded studies, and the reasons for their exclusions, are as shown in Table 3.52, below. Table 3.52 Excluded studies and reasons for exclusion Reason for exclusion Intervention <4 weeks Intervention <4 weeks 24-hour urinary K excretion values not measured Intervention <4 weeks Na levels varied; intervention <4 weeks 24-hour urinary K excretion values not measured 24-hour urinary K excretion values not measured No control 24-hour urinary K excretion values not measured Intervention <4 weeks Intervention <4 weeks K intake not only difference between control and intervention (Na levels varied) Intervention <4 weeks 24-hour urinary K excretion values not measured K intake not only difference between control and intervention Intervention <4 weeks Intervention <4 weeks 24-hour urinary K excretion values not measured Intervention <4 weeks Intervention <4 weeks 24-hour urinary K excretion values not measured K intake not only difference between control and intervention (calcium levels varied)

Study ID Agnoli RA 1992 Agnoli RA 1992 Barcelo RA 1993 Ceglia RA 2009 Fujita AEA 1984 Gamarra BPA 1994 Grimm BPA 1990 Heller RA 1998 Jardim BPA 1988 Khaw BPA 1982 Krishnan RA 2010 Langford BPA 1991 Lennon RA 1968 Med Res CWP BPA 1987 Overlack BPA RA AEA 1995 Parfrey AEA 1981 Poulter BPA 1986 Rahimi BPA 2007 Sanclemente BPA 1987 Smith BPA 1992 Svetkey BPA 1987 WHO ICTRP RA 2008 K, potassium; Na, sodium

3.8

Effect estimate tables Table 3.53 Resting systolic blood pressure Studies 22 22 3 16 2 2 5 11 4 6 12 4 2 15 5 1 16 5 15 8 3 10 12 14 7 9 2 14 5 3 20 3 8 14 757 818 233 183 286 1187 236 501 1169 222 169 1372 351 40 1470 382 933 1074 718 608 1284 692 1169 351 117 1421 195 276 1744 244 1026 866 Participants 1892 Effect Estimate -3.49 [-5.15, -1.82] Subtotals only 0.09 [-0.77, 0.95] -5.53 [-7.56, -3.51] -2.95 [-5.65, -0.26] -3.65 [-6.69, -0.62] -7.16 [-12.41, -1.91] -1.71 [-3.42, -0.00] -3.00 [-6.28, 0.27] -4.89 [-7.59, -2.20] -1.97 [-3.85, -0.09] -3.01 [-7.03, 1.02] -3.89 [-7.03, -0.74] -3.39 [-5.51, -1.28] -4.11 [-6.26, -1.97] -2.00 [-11.70, 7.70] -1.97 [-3.41, -0.52] -6.91 [-11.53, -2.29] -3.36 [-4.94, -1.78] -3.83 [-6.72, -0.95] 0.02 [-0.85, 0.90] -3.45 [-5.13, -1.77] -3.63 [-5.98, -1.28] -4.88 [-7.64, -2.13] -1.83 [-3.54, -0.12] -6.94 [-10.63, -3.25] -2.51 [-7.87, 2.85] -3.63 [-5.69, -1.57] -5.85 [-10.61, -1.08] -1.16 [-3.74, 1.43] -3.31 [-5.07, -1.55] -4.19 [-6.46, -1.92] -2.74 [-5.08, -0.39] -3.99 [-6.13, -1.86]

Outcome or Subgroup* 1.1 Resting systolic blood pressure (ALL) 1.2 Resting systolic blood pressure (SUBGROUPS)) 1.2.1 BP Status (NORMOTENSIVE) 1.2.2 BP Status (HYPERTENSIVE) 1.2.3 BP Status (HETEROGENEOUS) 1.2.4 1.2.5 1.2.6 1.2.7 Achieved Achieved Achieved Achieved K K K K intake intervention ( <70mmol) in intervention (>=70mmol v <90mmol) in intervention (>=90mmol v <120mmol) in intervention (>120mmol)

1.2.8 Difference in achieved K intake inter. v control: <30 1.2.9 Difference in achieved K intake inter. v control: 30-60 1.2.10 Difference in achieved K intake inter. v control: >60 1.2.11 K intake at baseline (lower: <40) 1.2.12 K intake at baseline (intermediate: 40-60) 1.2.13 K intake at baseline (higher: >60) 1.2.14 Na intake at baseline (<2g/d) 1.2.15 Na intake at baseline (2-4g/d) 1.2.16 Na intake at baseline (>4g/d) 1.2.17 Duration (<2 months) 1.2.18 Duration (2 - 4 months) 1.2.19 Duration (> 4 months) 1.2.20 Type of BP device (automatic) 1.2.21 Type of BP device (manual) 1.2.22 1.2.23 1.2.24 1.2.25 Type Type Type Type of of of of BP BP BP BP measure measure measure measure (supine office SBP) (seated office SBP) (standing office SBP) (unspecified SBP)

1.2.26 Hypertension medication status (not taking medication) 1.2.27 Hypertension medication status (population taking medication) 1.2.28 Hypertension medication status (not specified/heterogeneous) 1.2.29 Type of intervention (supplement) 1.2.30 Type of intervention (advice) 1.2.31 Study design (parallel) 1.2.32 Study design (cross-over) *Statistical method used: Mean Difference (IV, Random, 95%CI)

Table 3.54

Resting diastolic blood pressure Outcome or Subgroup* Studies 22 22 3 17 2 2 5 10 4 6 12 4 2 15 5

Participants 1857

Effect Estimate -3.02 [-4.86, -1.17] Subtotals only

1.3 Resting diastolic blood pressure (ALL) 1.4 Resting diastolic blood pressure (SUBGROUPS)) 1.4.1 BP Status (NORMOTENSIVE) 1.4.2 BP Status (HYPERTENSIVE) 1.4.3 BP Status (HETEROGENEOUS) 1.4.4 Achieved K intake in intervention (<70mmol) 1.4.5 Achieved K in intervention (>=70mmol v <90mmol) 1.4.6 Achieved K in intervention (>=90mmol v <120mmol) 1.4.7 Achieved K in intervention (>120mmol) 1.4.8 Difference in achieved K intake inter. v control: <30 1.4.9 Difference in achieved K intake inter. v control: 30-60 1.4.10 Difference in achieved K intake inter. v control: >60 1.4.11 K intake at baseline (lower: <40) 1.4.12 K intake at baseline (intermediate: 40-60) 1.4.13 K intake at baseline (higher: >60) 1.4.14 Na intake at baseline (<2g/d) 1.4.15 Na intake at baseline (2-4g/d) 1.4.16 Na intake at baseline (>4g/d) 1.4.17 Duration (<2 months) 1.4.18 Duration (2 - 4 months) 1.4.19 Duration (> 4 months) 1.4.20 Type of BP device (automatic) 1.4.21 Type of BP device (manual) 1.4.22 Type of BP measure (supine office DBP) 1.4.23 Type of BP measure (seated office DBP) 1.4.24 Type of BP measure (standing office DBP) 1.4.25 Type of BP measure (unspecified DBP) 1.4.26 Hypertension medication status (not taking medication) 1.4.27 Hypertension medication status (population taking medication) 1.4.28 Hypertension medication status (not specified/heterogeneous) 1.4.29 Type of intervention (supplement) 1.4.30 Type of intervention (advice) 1.4.31 Study design (parallel) 1.4.32 Study design (cross-over) *Statistical method used: Mean Difference (IV, Random, 95%CI)

722 902 233 183 212 1051 236 427 1134 222 169 1263 351

-0.56 [-1.55, 0.42] -3.91 [-6.54, -1.28] -0.17 [-1.82, 1.48] -1.35 [-5.31, 2.60] -4.01 [-8.44, 0.42] -0.83 [-1.82, 0.17] -1.75 [-4.23, 0.74] -1.87 [-4.11, 0.37] -1.63 [-3.04, -0.21] -3.57 [-6.32, -0.82] -2.41 [-7.90, 3.07] -1.53 [-2.80, -0.25] -3.38 [-4.74, -2.02] 0.00 [-6.12, 6.12] -1.96 [-3.16, -0.76] -2.87 [-6.96, 1.22] -1.99 [-3.11, -0.87] -1.86 [-3.75, 0.02] -0.35 [-1.06, 0.35] -2.84 [-3.96, -1.71] -1.54 [-2.96, -0.11] -4.34 [-7.57, -1.10] -1.30 [-2.64, 0.04] -4.78 [-9.38, -0.18] 0.83 [-4.45, 6.11] -1.37 [-2.50, -0.23] -3.80 [-8.25, 0.66] -2.32 [-4.46, -0.17] -3.04 [-5.09, -0.99] -2.44 [-5.04, 0.17] -1.59 [-3.35, 0.18] -4.21 [-7.12, -1.29]

1 16 5 15 8 3 10 12 14 7 9 2 14 5 3 20 3 8 15

40 1435 308 933 965 683 608 1175 692 1134

351 117 1312 195 276 1709 244 991 903

Table 3.55

Ambulatory systolic blood pressure Outcome or Subgroup* Studies 4 4 0 1 2 1 0 3 1 0 96 142 84 0 226 96 Participants 322 Effect Estimate -3.04 [-5.42, -0.66] Subtotals only Not estimable -1.80 [-7.02, 3.42] -3.65 [-7.21, -0.09] -3.00 [-7.07, 1.07] Not estimable -3.37 [-6.05, -0.69] -1.80 [-7.02, 3.42]

1.5 Ambulatory systolic blood pressure (ALL) 1.6 Ambulatory systolic blood pressure (SUBGROUPS)) 1.6.1 Achieved K intake inter. v control (both <70mmol) 1.6.2 Achieved K in intervention (>=70mmol v <90mmol) 1.6.3 Achieved K in intervention (>=90mmol v <120mmol) 1.6.4 Achieved K in intervention (>120mmol) 1.6.5 BP Status (NORMOTENSIVE) 1.6.6 BP Status (HYPERTENSIVE) 1.6.7 BP Status (HETEROGENEOUS) *Statistical method used: Mean Difference (IV, Random, 95%CI)

Table 3.56

Ambulatory diastolic blood pressure Outcome or Subgroup* 1.7 Ambulatory diastolic blood pressure (ALL)

Studies 4

Participants 322

Effect Estimate -1.24 [-3.13, 0.66]

1.8 Ambulatory diastolic blood pressure (SUBGROUPS)) 1.8.1 BP Status (NORMOTENSIVE) 1.8.2 BP Status (HYPERTENSIVE) 1.8.3 BP Status (HETEROGENEOUS) *Statistical method used: Mean Difference (IV, Random, 95%CI)

4 0 3 1 0 226 96

Subtotals only Not estimable -1.18 [-3.38, 1.02] -1.40 [-5.14, 2.34]

Table 3.57 1.9 Total cholesterol (ALL)

Blood lipids Studies 4 2 1 2 Participants 208 128 96 128 Effect Estimate -0.12 [-0.33, 0.09] -0.0.1 [-0.13, 0.11] -0.10 [-0.38, 0.18] -0.11 [-0.48, 0.26]

Outcome or Subgroup*

1.10 HDL cholesterol (ALL) 1.11 LDL cholesterol (ALL) 1.12 Triglycerides (ALL) *Statistical method used: Mean Difference (IV, Random, 95%CI)

Table 3.58

Catecholamine levels Outcome or Subgroup* Studies 3 3 Participants 152 152 Effect Estimate -3.94 [-9.22, 1.34] -4.32 [-23.78, 15.13]

1.14 Adrenaline (plasma) (ALL) 1.15 Noradrenaline (plasma) (ALL) *Statistical method used: Mean Difference (IV, Random, 95%CI)

Table 3.59 1.16 Serum creatinine (ALL)

Renal function Outcome or Subgroup* Studies 3 Participants 147 Effect Estimate -4.86 [-13.59, 3.87]

*Statistical method used: Mean Difference (IV, Random, 95%CI)

3.9

Figures Figure 3.1 Flow through screening, inclusion, exclusion

Figure 3.2 Study or Subgroup

Resting systolic blood pressure Increased potassium Mean SD Total Control Mean SD Total Weight Mean Difference IV, Random, 95% CI Mean Difference IV, Random, 95% CI

Barden BPARA1986 Berry BPA2010 Bulpitt BPA1985 Chalmers BPA1986 Forrester BPA1988 Fotherby BPA1992 Grobbee BPA1987 Gu BPA2001 He BPA2010 Kaplan BPA1985 Kawano BPA1998 MacGregor AEBPA1982

113.67 120.7 -5.2 -8.9 129.8 176 135 -13.1 142 127.6 147.9 148

8.87 9.9 27.5 7 13.5 20 13.28 10.7 11 11.6 11.7 13.91

43 113.69 48 19 49 23 18 40 68 42 16 55 23

9.44 13 23.9 7.4 12.7 24

43 48 14 52 23 18 40 69 42 16 55 23

6.9% 6.0% 0.8% 8.4% 3.4% 1.2% 5.1% 7.7% 4.9% 2.2% 6.1% 2.9%

-0.02 [-3.89, 3.85] -1.50 [-6.12, 3.12] 2.30 [-15.30, 19.90] -5.10 [-7.91, -2.29] -3.40 [-10.97, 4.17] -10.00 [-24.43, 4.43] -0.90 [-6.32, 4.52] -3.70 [-7.01, -0.39] -3.00 [-8.63, 2.63] -5.60 [-15.60, 4.40] -2.90 [-7.45, 1.65] -7.00 [-15.32, 1.32]

122.2 -7.5 -3.8 133.2 186

135.9 11.38 -9.4 145 133.2

9 15 16.8

150.8 12.61 155 14.87

Matlou BPA1986 Obel BPA1989 Patki BPARA1990 Richards BPAAEA1984 Siani BPA1987 Siani BPARA1991 Smith BPARA1985 Trial Hyp Prv Col BPA1992 Valdes BPA1991 Whelton BPA1995 Total (95% CI)

144 133 143.6 148 131.8 142.1 160 -0.78 138 -0.87

18.1 4 10.8 14.2 12.7 11.5 16.1 5.88 14.7 5.94

32 24 37 12 18 21 20 178 24 161 947

151 172 155.7 149.9 145.8 145.5 162 -0.84 145 -1

19.2 7 11.4 14.5 11.3 9.7 15.2 5.82 9.8 5.69

32 24 37 12 19 26 20 175 24 157

2.6%

-7.00 [-16.14, 2.14]

0.0% -39.00 [-42.23, -35.77] 5.5% 1.8% 3.3% 4.4% 2.3% 10.5% 3.7% 10.4% -12.10 [-17.16, -7.04] -1.90 [-13.38, 9.58] -14.00 [-21.76, -6.24] -3.40 [-9.57, 2.77] -2.00 [-11.70, 7.70] 0.06 [-1.16, 1.28] -7.00 [-14.07, 0.07] 0.13 [-1.15, 1.41] -3.49 [-5.15, -1.82] -20 -10 Favours increased K 0 10 20 Favours control

945 100.0%

Heterogeneity: Tau² = 6.52; Chi² = 56.52, df = 20 (P < 0.0001); I² = 65% Test for overall effect: Z = 4.10 (P < 0.0001)

Figure 3.3 Study or Subgroup

Resting systolic blood pressure – blood pressure status subgroups Increased potassium Control Mean Difference Mean Difference

Mean

Total

Mean

SD Total Weight

IV, Random, 95% CI

5.6.1 BP Status (NORMOTENSIVE)

Barden BPARA1986 Trial Hyp Prv Col BPA1992

113.67 -0.78

8.87 5.88

43 113.69 178 -0.84

9.44 5.82

43 175

4.9% 49.7%

-0.02 [-3.89, 3.85] 0.06 [-1.16, 1.28]

Whelton BPA1995 Subtotal (95% CI)

-0.87

5.94

161 382

-1

5.69

157

45.3%

0.13 [-1.15, 1.41] 0.09 [-0.77, 0.95]

375 100.0%

Heterogeneity: Tau² = 0.00; Chi² = 0.01, df = 2 (P = 1.00); I² = 0%

Test for overall effect: Z = 0.20 (P = 0.84) 5.6.2 BP Status (HYPERTENSIVE) Bulpitt BPA1985 Chalmers BPA1986 Forrester BPA1988

-5.2 -8.9 129.8

27.5 7 13.5

19 49 23

-7.5 -3.8 133.2

23.9 7.4 12.7

14 52 23

1.3% 19.4% 5.8%

2.30 [-15.30, 19.90] -5.10 [-7.91, -2.29] -3.40 [-10.97, 4.17]

Fotherby BPA1992 Grobbee BPA1987 Kaplan BPA1985

176 135 127.6

20 13.28 11.6

18 40 16

186

18 40 16

1.8% 9.6% 3.6%

-10.00 [-24.43, 4.43] -0.90 [-6.32, 4.52] -5.60 [-15.60, 4.40]

135.9 11.38 133.2 16.8

Kawano BPA1998 MacGregor AEBPA1982 Matlou BPA1986

147.9 148 144

11.7 13.91 18.1

55 23 32

150.8 12.61 155 14.87 151 19.2

55 23 32

12.1% 5.0% 4.2%

-2.90 [-7.45, 1.65] -7.00 [-15.32, 1.32] -7.00 [-16.14, 2.14]

Obel BPA1989 Patki BPARA1990 Richards BPAAEA1984

133 143.6 148

10 10.8 14.2

24 37 12

172 155.7 149.9

7 11.4 14.5

24 37 12

0.0% -39.00 [-43.88, -34.12] 10.5% 2.8% -12.10 [-17.16, -7.04] -1.90 [-13.38, 9.58]

Siani BPA1987 Siani BPARA1991 Smith BPARA1985

131.8 142.1 160

12.7 11.5 16.1

18 21 20

145.8 145.5 162

11.3 9.7 15.2

19 26 20

5.6% 8.0% 3.8%

-14.00 [-21.76, -6.24] -3.40 [-9.57, 2.77] -2.00 [-11.70, 7.70]

Valdes BPA1991 Subtotal (95% CI)

138

14.7

24 407

145

9.8

6.5%

-7.00 [-14.07, 0.07] -5.53 [-7.56, -3.51]

411 100.0%

Heterogeneity: Tau² = 3.45; Chi² = 18.39, df = 14 (P = 0.19); I² = 24%

Test for overall effect: Z = 5.36 (P < 0.00001) 5.6.3 BP Status (HETEROGENEOUS)

Berry BPA2010 Gu BPA2001 Subtotal (95% CI)

120.7 -13.1

9.9 10.7

48 68 116

122.2 -9.4

13 9

48 69

33.9% 66.1%

-1.50 [-6.12, 3.12] -3.70 [-7.01, -0.39] -2.95 [-5.65, -0.26]

117 100.0%

Heterogeneity: Tau² = 0.00; Chi² = 0.57, df = 1 (P = 0.45); I² = 0% Test for overall effect: Z = 2.15 (P = 0.03)

-20 Test for subgroup differences: Chi² = 27.43, df = 2 (P < 0.00001), I² = 92.7%

-10

Favours increased K

Favours control

Figure 3.4

Resting systolic blood pressure – achieved intake subgroups (based on urinary potassium excretion*) Increased potassium Control Mean SD Total Weight

Mean Difference IV, Random, 95% CI

Study or Subgroup

Mean

Total

2.1.1 Achieved K intake in intervention (<70mmol) Forrester BPA1988 129.8 13.5 23 133.2 12.7 23 16.1% -3.40 [-10.97, 4.17]

Gu BPA2001 Subtotal (95% CI)

-13.1

10.7

68 91

-9.4

83.9%

-3.70 [-7.01, -0.39] -3.65 [-6.69, -0.62]

92 100.0%

Heterogeneity: Tau² = 0.00; Chi² = 0.01, df = 1 (P = 0.94); I² = 0%

Test for overall effect: Z = 2.36 (P = 0.02) 2.1.2 Achieved K in intervention (>=70mmol v <90mmol) Berry BPA2010 120.7 9.9 48 122.2 13 48 23.9% -1.50 [-6.12, 3.12]

Kaplan BPA1985 Patki BPARA1990 Siani BPA1987

127.6 143.6 131.8

11.6 10.8 12.7

16 37 18

133.2 155.7 145.8

16.8 11.4 11.3

16 37 19

14.2% 23.1% 17.9%

-5.60 [-15.60, 4.40] -12.10 [-17.16, -7.04] -14.00 [-21.76, -6.24]

Siani BPARA1991 Subtotal (95% CI)

142.1

11.5

21 140

145.5

9.7

20.9%

-3.40 [-9.57, 2.77] -7.16 [-12.41, -1.91]

146 100.0%

Heterogeneity: Tau² = 24.42; Chi² = 13.78, df = 4 (P = 0.008); I² = 71% Test for overall effect: Z = 2.67 (P = 0.007)

2.1.3 Achieved K in intervention (>=90mmol v <120mmol) Barden BPARA1986 113.67 8.87 43 113.69 9.44 43 11.7% -0.02 [-3.89, 3.85]

Bulpitt BPA1985 Chalmers BPA1986

-5.2 -8.9

27.5 7

19 49

-7.5 -3.8

23.9 7.4

14 52

0.9% 16.3%

2.30 [-15.30, 19.90] -5.10 [-7.91, -2.29]

Fotherby BPA1992 Kawano BPA1998

176 147.9

20 11.7

18 55

186

18 55

1.3% 9.5%

-10.00 [-24.43, 4.43] -2.90 [-7.45, 1.65]

150.8 12.61

MacGregor AEBPA1982 Matlou BPA1986 Obel BPA1989

148 144 133

13.91 18.1 10

23 32 24

155 14.87 151 172 19.2 7

23 32 24

3.7% 3.1%

-7.00 [-15.32, 1.32] -7.00 [-16.14, 2.14]

0.0% -39.00 [-43.88, -34.12]

Smith BPARA1985 Trial Hyp Prv Col BPA1992

160 -0.78

16.1 5.88

20 178

162 -0.84

15.2 5.82

20 175

2.8% 25.5%

-2.00 [-11.70, 7.70] 0.06 [-1.16, 1.28]

Whelton BPA1995 Subtotal (95% CI)

-0.87

5.94

161 598

-1

5.69

157

25.1%

0.13 [-1.15, 1.41] -1.71 [-3.42, -0.00]

589 100.0%

Heterogeneity: Tau² = 2.60; Chi² = 19.28, df = 9 (P = 0.02); I² = 53% Test for overall effect: Z = 1.96 (P = 0.05) 2.1.4 Achieved K in intervention (>120mmol) Grobbee BPA1987 135 13.28 40 135.9 11.38 40 36.5% -0.90 [-6.32, 4.52]

He BPA2010 Richards BPAAEA1984 Valdes BPA1991

142 148 138

11 14.2 14.7

42 12 24

145 149.9 145

15 14.5 9.8

42 12 24

33.9% 8.1% 21.5%

-3.00 [-8.63, 2.63] -1.90 [-13.38, 9.58] -7.00 [-14.07, 0.07]

Subtotal (95% CI)

118

118 100.0%

-3.00 [-6.28, 0.27]

Heterogeneity: Tau² = 0.00; Chi² = 1.84, df = 3 (P = 0.61); I² = 0%

Test for overall effect: Z = 1.80 (P = 0.07)

-20 Test for subgroup differences: Chi² = 4.46, df = 3 (P = 0.22), I² = 32.8%

-10

Favours increased K

Favours control

 Urinary potassium excretion is a common, valid form of estimating potassium intake. A factor of 1.30 is used to convert urinary potassium excretion to potassium intake (Stamler et al., 2003).

Figure 3.5

Resting systolic blood pressure – achieved difference subgroups (based on urinary potassium excretion*) Increased potassium Control Mean Difference Mean Difference

Study or Subgroup

Mean

Total

Mean

SD Total Weight

IV, Random, 95% CI

5.5.1 Difference in achieved K intake inter. v control: <30

Berry BPA2010 Chalmers BPA1986

120.7 -8.9

9.9 7

48 49

122.2 -3.8

13 7.4

48 52

16.8% 24.4%

-1.50 [-6.12, 3.12] -5.10 [-7.91, -2.29]

Forrester BPA1988 Gu BPA2001 Patki BPARA1990

129.8 -13.1 143.6

13.5 10.7 10.8

23 68 37

133.2 -9.4 155.7

12.7 9 11.4

23 69 37

9.2% 22.1% 15.3%

-3.40 [-10.97, 4.17] -3.70 [-7.01, -0.39] -12.10 [-17.16, -7.04]

Siani BPARA1991 Subtotal (95% CI)

142.1

11.5

21 246

145.5

9.7

12.1%

-3.40 [-9.57, 2.77] -4.89 [-7.59, -2.20]

255 100.0%

Heterogeneity: Tau² = 5.67; Chi² = 10.75, df = 5 (P = 0.06); I² = 53%

Test for overall effect: Z = 3.56 (P = 0.0004) 5.5.2 Difference in achieved K intake inter. v control: 30-60

Bulpitt BPA1985 Fotherby BPA1992

-5.2 176

27.5 20

19 18

-7.5 186

23.9 24

14 18

1.1% 1.6%

2.30 [-15.30, 19.90] -10.00 [-24.43, 4.43]

Grobbee BPA1987 He BPA2010 Kaplan BPA1985

135 142 127.6

13.28 11 11.6

40 42 16

135.9 11.38 145 133.2

40 42 16

8.6% 8.2% 3.2%

-0.90 [-6.32, 4.52] -3.00 [-8.63, 2.63] -5.60 [-15.60, 4.40]

15 16.8

Kawano BPA1998 MacGregor AEBPA1982 Obel BPA1989

147.9 148 133

11.7 13.91 10

55 23 24

150.8 12.61 155 14.87 172 7

55 23 24

11.0% 4.4%

-2.90 [-7.45, 1.65] -7.00 [-15.32, 1.32]

0.0% -39.00 [-43.88, -34.12]

Siani BPA1987 Smith BPARA1985 Trial Hyp Prv Col BPA1992

131.8 160 -0.78

12.7 16.1 5.88

18 20 178

145.8 162 -0.84

11.3 15.2 5.82

19 20 175

4.9% 3.4% 27.0%

-14.00 [-21.76, -6.24] -2.00 [-11.70, 7.70] 0.06 [-1.16, 1.28]

Whelton BPA1995 Subtotal (95% CI)

-0.87

5.94

161 590

-1

5.69

157

26.7%

0.13 [-1.15, 1.41] -1.97 [-3.85, -0.09]

579 100.0%

Heterogeneity: Tau² = 3.04; Chi² = 20.36, df = 10 (P = 0.03); I² = 51%

Test for overall effect: Z = 2.05 (P = 0.04) 5.5.3 Difference in achieved K intake inter. v control: >60

Barden BPARA1986 Matlou BPA1986

113.67 144

8.87 18.1

43 113.69 32 151

9.44 19.2

43 32

49.3% 16.0%

-0.02 [-3.89, 3.85] -7.00 [-16.14, 2.14]

Richards BPAAEA1984 Valdes BPA1991 Subtotal (95% CI)

148 138

14.2 14.7

12 24 111

149.9 145

14.5 9.8

12 24

10.8% 23.9%

-1.90 [-13.38, 9.58] -7.00 [-14.07, 0.07] -3.01 [-7.03, 1.02]

111 100.0%

Heterogeneity: Tau² = 4.65; Chi² = 4.05, df = 3 (P = 0.26); I² = 26% Test for overall effect: Z = 1.46 (P = 0.14)

-20 Test for subgroup differences: Chi² = 3.04, df = 2 (P = 0.22), I² = 34.2% 

-10

Favours increased K

Favours control

Urinary potassium excretion is a common, valid form of estimating potassium intake. A factor of 1.30 is used to convert urinary potassium excretion to potassium intake (Stamler et al., 2003).

Figure 3.6 Study or Subgroup

Resting diastolic blood pressure – all adults Increased potassium Mean SD Control SD Total Weight Mean Difference IV, Random, 95% CI Mean Difference IV, Random, 95% CI Total Mean

Barden BPARA1986 Berry BPA2010 Bulpitt BPA1985 Chalmers BPA1986 Forrester BPA1988 Fotherby BPA1992 Grobbee BPA1987 Gu BPA2001 He BPA2010 Kaplan BPA1985 Kawano BPA1998 MacGregor AEBPA1982

68.02 76.6 0.5 -5.8 87.7 94 71.6 -3.8 90 91.9 86.8 95

7.24 6.54 11.8 4.2 12.1 20 10.75 6.2 9 7.6 8.16 7.67

43 71.67 48 19 49 23 18 40 68 42 16 55 23

9.68 7.5 11.2 4.5 9.2 14 9.49 6 9 7.6 9.64 8.15

43 48 14 52 23 18 40 69 42 16 55 23

5.4% 5.9% 3.0% 6.5% 3.9% 1.9% 4.9% 6.3% 5.3% 4.4% 5.6% 4.8%

-3.65 [-7.26, -0.04] -0.30 [-3.12, 2.52] 4.80 [-3.11, 12.71] -4.20 [-5.90, -2.50] -4.60 [-10.81, 1.61] -6.00 [-17.28, 5.28] -2.10 [-6.54, 2.34] -0.10 [-2.14, 1.94] -1.00 [-4.85, 2.85] -5.80 [-11.07, -0.53] -1.30 [-4.64, 2.04] -4.00 [-8.57, 0.57]

76.9 -4.3 -1.6 92.3 100 73.7 -3.7 91 97.7 88.1 99

Matlou BPA1986 Obel BPA1989 Patki BPARA1990 Richards BPAAEA1984 Siani BPA1987 Siani BPARA1991 Smith BPARA1985 Trial Hyp Prv Col BPA1992 Valdes BPA1991 Whelton BPA1995 Total (95% CI)

144 83 84.5 91.4 82 83.1 103 -0.27 89 -0.29

18.1 4 3.5 11.4 8.9 9.2 8.94 4.56 9.8 4.68

32 24 37 12 18 21 20 161 24 161 930

151 100 97.6 92.4 92.5 84.2

19.2 4 5.4 12.1 8.7 6.6

32 24 37 12 19 26 20 157 24 157

2.6%

-7.00 [-16.14, 2.14]

0.0% -17.00 [-19.26, -14.74] 6.3% -13.10 [-15.17, -11.03] 2.5% 4.2% 4.8% 3.9% 6.7% 4.2% 6.7% -1.00 [-10.41, 8.41] -10.50 [-16.18, -4.82] -1.10 [-5.78, 3.58] 0.00 [-6.12, 6.12] -0.41 [-1.42, 0.60] -3.00 [-8.54, 2.54] -0.26 [-1.27, 0.75] -3.02 [-4.86, -1.17] -10 -5 0 5 10

103 10.73 0.14 92 -0.03 4.64 9.8 4.5

927 100.0%

Heterogeneity: Tau² = 12.84; Chi² = 160.62, df = 20 (P < 0.00001); I² = 88% Test for overall effect: Z = 3.21 (P = 0.001)

Favours increased K

Favours control

Figure 3.7 Study or Subgroup

Resting diastolic blood pressure – blood pressure status subgroups High potassium Control Mean Difference Mean Difference

Mean

SD Total Mean

SD Total Weight

IV, Random, 95% CI

3.2.16 BP Status (NORMOTENSIVE)

Barden BPARA1986 Trial Hyp Prv Col BPA1992

68.02 -0.27

7.24 4.56

43 71.67 161 0.14

9.68 4.64

43 157

6.9% 46.5%

-3.65 [-7.26, -0.04] -0.41 [-1.42, 0.60]

Whelton BPA1995 Subtotal (95% CI)

-0.29

4.68

161 365

-0.03

4.5

157

46.6%

-0.26 [-1.27, 0.75] -0.56 [-1.55, 0.42]

357 100.0%

Heterogeneity: Tau² = 0.28; Chi² = 3.15, df = 2 (P = 0.21); I² = 37%

Test for overall effect: Z = 1.12 (P = 0.26) 3.2.17 BP Status (HYPERTENSIVE)

Bulpitt BPA1985 Chalmers BPA1986 Forrester BPA1988

0.5 -5.8 87.7

11.8 4.2 12.1

19 49 23

-4.3 -1.6 92.3

11.2 4.5 9.2

14 52 23

4.9% 8.4% 5.9%

4.80 [-3.11, 12.71] -4.20 [-5.90, -2.50] -4.60 [-10.81, 1.61]

Fotherby BPA1992 Grobbee BPA1987

18 40

100 73.7

14 9.49

18 40

3.4% 7.0%

-6.00 [-17.28, 5.28] -2.10 [-6.54, 2.34]

71.6 10.75

He BPA2010 Kaplan BPA1985 Kawano BPA1998

90 91.9 86.8

9 7.6 8.16

42 16 55

91 97.7 88.1

9 7.6 9.64

42 16 55

7.4% 6.5% 7.7%

-1.00 [-4.85, 2.85] -5.80 [-11.07, -0.53] -1.30 [-4.64, 2.04]

MacGregor AEBPA1982 Matlou BPA1986 Obel BPA1989

95 144 83

7.67 18.1 4

23 32 24

99 151 100

8.15 19.2 4

23 32 24

6.9% 4.3%

-4.00 [-8.57, 0.57] -7.00 [-16.14, 2.14]

0.0% -17.00 [-19.26, -14.74]

Patki BPARA1990 Richards BPAAEA1984 Siani BPA1987

84.5 91.4 82

3.5 11.4 8.9

37 12 18

97.6 92.4 92.5

5.4 12.1 8.7

37 12 19

8.3% -13.10 [-15.17, -11.03] 4.1% 6.2% -1.00 [-10.41, 8.41] -10.50 [-16.18, -4.82]

Siani BPARA1991 Smith BPARA1985 Valdes BPA1991 Subtotal (95% CI)

83.1 103 89

9.2 8.94 9.8

21 20 24 449

84.2

6.6

26 20 24

6.9% 5.9% 6.3%

-1.10 [-5.78, 3.58] 0.00 [-6.12, 6.12] -3.00 [-8.54, 2.54] -3.91 [-6.54, -1.28]

103 10.73 92 9.8

453 100.0%

Heterogeneity: Tau² = 20.54; Chi² = 85.68, df = 15 (P < 0.00001); I² = 82%

Test for overall effect: Z = 2.91 (P = 0.004) 3.2.18 BP Status (HETEROGENEOUS) Berry BPA2010 Gu BPA2001 Subtotal (95% CI)

76.6 -3.8

6.54 6.2

48 68 116

76.9 -3.7

7.5 6

48 69

34.5% 65.5%

-0.30 [-3.12, 2.52] -0.10 [-2.14, 1.94] -0.17 [-1.82, 1.48]

117 100.0%

Heterogeneity: Tau² = 0.00; Chi² = 0.01, df = 1 (P = 0.91); I² = 0% Test for overall effect: Z = 0.20 (P = 0.84)

-20 Test for subgroup differences: Chi² = 6.16, df = 2 (P = 0.05), I² = 67.5%

-10

10 Favours control

Favours high potassium

Figure 3.8

Resting diastolic blood pressure – achieved intake subgroups (based on urinary potassium excretion*) Increased potassium Control Mean Difference Mean Difference

Study or Subgroup

Mean

Total Mean

SD Total Weight

IV, Random, 95% CI

4.1.1 Achieved K intake in intervention (<70mmol) Forrester BPA1988 87.7 12.1 23 92.3 9.2 23 27.9% -4.60 [-10.81, 1.61]

Gu BPA2001 Subtotal (95% CI)

-3.8

6.2

68 91

-3.7

72.1%

-0.10 [-2.14, 1.94] -1.35 [-5.31, 2.60]

92 100.0%

Heterogeneity: Tau² = 4.56; Chi² = 1.82, df = 1 (P = 0.18); I² = 45%

Test for overall effect: Z = 0.67 (P = 0.50) 4.1.2 Achieved K in intervention (>=70mmol v <90mmol) Berry BPA2010 Kaplan BPA1985 76.6 91.9 6.54 7.6 48 16 76.9 97.7 7.5 7.6 48 16 30.1% 23.1% -0.30 [-3.12, 2.52] -5.80 [-11.07, -0.53]

Patki BPARA1990 Siani BPA1987

84.5 82

3.5 8.9

37 18

97.6 92.5

5.4 8.7

37 19

0.0% -13.10 [-15.17, -11.03] 22.0% -10.50 [-16.18, -4.82]

Siani BPARA1991 Subtotal (95% CI)

83.1

9.2

21 103

84.2

6.6

24.8%

-1.10 [-5.78, 3.58] -4.01 [-8.44, 0.42]

109 100.0%

Heterogeneity: Tau² = 14.90; Chi² = 11.82, df = 3 (P = 0.008); I² = 75%

Test for overall effect: Z = 1.77 (P = 0.08) 4.1.3 Achieved K in intervention (>=90mmol v <120mmol) Barden BPARA1986 68.02 7.24 43 71.67 9.68 43 6.7% -3.65 [-7.26, -0.04]

Bulpitt BPA1985 Fotherby BPA1992 Kawano BPA1998

0.5 94 86.8

11.8 20 8.16

19 18 55

-4.3 100 88.1

11.2 14 9.64

14 18 55

1.5% 0.8% 7.7%

4.80 [-3.11, 12.71] -6.00 [-17.28, 5.28] -1.30 [-4.64, 2.04]

MacGregor AEBPA1982 Matlou BPA1986

95 144

7.67 18.1

23 32

99 151

8.15 19.2

23 32

4.4% 1.2%

-4.00 [-8.57, 0.57] -7.00 [-16.14, 2.14]

Obel BPA1989 Smith BPARA1985 Trial Hyp Prv Col BPA1992

83 103 -0.27

4 8.94 4.56

24 20 161

100

24 20 157

0.0% -17.00 [-19.26, -14.74] 2.5% 37.6% 0.00 [-6.12, 6.12] -0.41 [-1.42, 0.60]

103 10.73 0.14 4.64

Whelton BPA1995 Subtotal (95% CI)

-0.29

4.68

161 532

-0.03

4.5

157

37.6%

-0.26 [-1.27, 0.75] -0.83 [-1.82, 0.17]

519 100.0%

Heterogeneity: Tau² = 0.42; Chi² = 10.19, df = 8 (P = 0.25); I² = 21% Test for overall effect: Z = 1.63 (P = 0.10)

4.1.4 Achieved K in intervention (>120mmol) Grobbee BPA1987 71.6 10.75 40 73.7 9.49 40 31.3% -2.10 [-6.54, 2.34]

He BPA2010 Richards BPAAEA1984

90 91.4

9 11.4

42 12

91 92.4

9 12.1

42 12

41.7% 7.0%

-1.00 [-4.85, 2.85] -1.00 [-10.41, 8.41]

Valdes BPA1991 Subtotal (95% CI)

9.8

24 118

9.8

20.1%

-3.00 [-8.54, 2.54] -1.75 [-4.23, 0.74]

118 100.0%

Heterogeneity: Tau² = 0.00; Chi² = 0.39, df = 3 (P = 0.94); I² = 0%

Test for overall effect: Z = 1.38 (P = 0.17)

-20 Test for subgroup differences: Chi² = 2.22, df = 3 (P = 0.53), I² = 0%

-10

Favours increased K

Favours control

 Urinary potassium excretion is a common, valid form of estimating potassium intake. A factor of 1.30 is used to convert urinary potassium excretion to potassium intake (Stamler et al., 2003).

Figure 3.9

Resting diastolic blood pressure – achieved difference subgroups (based on urinary potassium excretion*) High potassium Control Mean Difference Mean Difference

Study or Subgroup

Mean

SD Total Mean

SD Total Weight

IV, Random, 95% CI

3.3.13 Difference in achieved K intake inter. v control: <30

Berry BPA2010 Chalmers BPA1986

76.6 -5.8

6.54 4.2

48 49

76.9 -1.6

7.5 4.5

48 52

22.1% 28.4%

-0.30 [-3.12, 2.52] -4.20 [-5.90, -2.50]

Forrester BPA1988 Gu BPA2001 Patki BPARA1990

87.7 -3.8 84.5

12.1 6.2 3.5

23 68 37

92.3 -3.7 97.6

9.2 6 5.4

23 69 37

9.4% 26.4%

-4.60 [-10.81, 1.61] -0.10 [-2.14, 1.94]

0.0% -13.10 [-15.17, -11.03]

Siani BPARA1991 Subtotal (95% CI)

83.1

9.2

21 209

84.2

6.6

13.7%

-1.10 [-5.78, 3.58] -1.87 [-4.11, 0.37]

218 100.0%

Heterogeneity: Tau² = 3.86; Chi² = 11.92, df = 4 (P = 0.02); I² = 66%

Test for overall effect: Z = 1.63 (P = 0.10) 3.3.14 Difference in achieved K intake inter. v control: 30-60

Bulpitt BPA1985 Fotherby BPA1992

0.5 94

11.8 20

19 18

-4.3 100

11.2 14

14 18

2.9% 1.5%

4.80 [-3.11, 12.71] -6.00 [-17.28, 5.28]

Grobbee BPA1987 He BPA2010 Kaplan BPA1985

71.6 10.75 90 91.9

40 42 16

73.7 91 97.7

9.49 9 7.6

40 42 16

7.3% 8.9% 5.7%

-2.10 [-6.54, 2.34] -1.00 [-4.85, 2.85] -5.80 [-11.07, -0.53]

9 7.6

Kawano BPA1998 MacGregor AEBPA1982 Obel BPA1989

86.8 95 83

8.16 7.67 4

55 23 24

88.1 99 100

9.64 8.15 4

55 23 24

10.7% 7.0%

-1.30 [-4.64, 2.04] -4.00 [-8.57, 0.57]

0.0% -17.00 [-19.26, -14.74]

Siani BPA1987 Smith BPARA1985 Trial Hyp Prv Col BPA1992

82 103 -0.27

8.9 8.94 4.56

18 20 161

92.5

8.7

19 20 157

5.0% 4.4% 23.3%

-10.50 [-16.18, -4.82] 0.00 [-6.12, 6.12] -0.41 [-1.42, 0.60]

103 10.73 0.14 4.64

Whelton BPA1995 Subtotal (95% CI)

-0.29

4.68

161 573

-0.03

4.5

157

23.3%

-0.26 [-1.27, 0.75] -1.63 [-3.04, -0.21]

561 100.0%

Heterogeneity: Tau² = 1.97; Chi² = 21.39, df = 10 (P = 0.02); I² = 53%

Test for overall effect: Z = 2.26 (P = 0.02) 3.3.15 Difference in achieved K intake inter. v control: >60

Barden BPARA1986 Matlou BPA1986

68.02 144

7.24 18.1

43 71.67 32 151

9.68 19.2

43 32

57.9% 9.0%

-3.65 [-7.26, -0.04] -7.00 [-16.14, 2.14]

Richards BPAAEA1984 Valdes BPA1991 Subtotal (95% CI)

91.4 89

11.4 9.8

12 24 111

92.4 92

12.1 9.8

12 24

8.5% 24.6%

-1.00 [-10.41, 8.41] -3.00 [-8.54, 2.54] -3.57 [-6.32, -0.82]

111 100.0%

Heterogeneity: Tau² = 0.00; Chi² = 0.87, df = 3 (P = 0.83); I² = 0% Test for overall effect: Z = 2.54 (P = 0.01)

-20 Test for subgroup differences: Chi² = 1.53, df = 2 (P = 0.47), I² = 0% 

-10

10 Favours control

Favours high potassium

Urinary potassium excretion is a common, valid form of estimating potassium intake. A factor of 1.30 is used to convert urinary potassium excretion to potassium intake (Stamler et al., 2003).

Figure 3.10 Resting systolic blood pressure – baseline potassium subgroups (based on urinary potassium excretion*) Increased potassium Control Mean Difference Mean Difference

Study or Subgroup

Mean

Total

Mean

SD Total Weight

IV, Random, 95% CI

5.13.1 K intake at baseline (lower: <40)

Gu BPA2001 Kaplan BPA1985 Subtotal (95% CI)

-13.1 127.6

10.7 11.6

68 16 84

-9.4 133.2

9 16.8

69 16

90.1% 9.9%

-3.70 [-7.01, -0.39] -5.60 [-15.60, 4.40] -3.89 [-7.03, -0.74]

85 100.0%

Heterogeneity: Tau² = 0.00; Chi² = 0.12, df = 1 (P = 0.72); I² = 0% Test for overall effect: Z = 2.42 (P = 0.02)

5.13.2 K intake at baseline (intermediate: 40-60) Barden BPARA1986 113.67 8.87 43 113.69 9.44 43 10.4% -0.02 [-3.89, 3.85]

Berry BPA2010 Bulpitt BPA1985 Forrester BPA1988

120.7 -5.2 129.8

9.9 27.5 13.5

48 19 23

122.2 -7.5 133.2

13 23.9 12.7

48 14 23

9.1% 1.3% 5.2%

-1.50 [-6.12, 3.12] 2.30 [-15.30, 19.90] -3.40 [-10.97, 4.17]

Fotherby BPA1992 Kawano BPA1998

176 147.9

20 11.7

18 55

186

18 55

1.9% 9.2%

-10.00 [-24.43, 4.43] -2.90 [-7.45, 1.65]

150.8 12.61

Matlou BPA1986 Obel BPA1989 Patki BPARA1990

144 133 143.6

18.1 10 10.8

32 24 37

151 172 155.7

19.2 7 11.4

32 24 37

4.0%

-7.00 [-16.14, 2.14]

0.0% -39.00 [-43.88, -34.12] 8.3% -12.10 [-17.16, -7.04]

Richards BPAAEA1984 Siani BPA1987 Siani BPARA1991

148 131.8 142.1

14.2 12.7 11.5

12 18 21

149.9 145.8 145.5

14.5 11.3 9.7

12 19 26

2.8% 5.1% 6.8%

-1.90 [-13.38, 9.58] -14.00 [-21.76, -6.24] -3.40 [-9.57, 2.77]

Trial Hyp Prv Col BPA1992 Valdes BPA1991 Whelton BPA1995 Subtotal (95% CI)

-0.78 138 -0.87

5.88 14.7 5.94

178 24 161 689

-0.84 145 -1

5.82 9.8 5.69

175 24 157

15.2% 5.7% 15.1%

0.06 [-1.16, 1.28] -7.00 [-14.07, 0.07] 0.13 [-1.15, 1.41] -3.39 [-5.51, -1.28]

683 100.0%

Heterogeneity: Tau² = 7.29; Chi² = 42.84, df = 13 (P < 0.0001); I² = 70% Test for overall effect: Z = 3.15 (P = 0.002)

5.13.3 K intake at baseline (higher: >60) Chalmers BPA1986 Grobbee BPA1987 He BPA2010

-8.9 135 142

7 13.28 11

49 40 42

-3.8

7.4

52 40 42

58.3% 15.7% 14.5%

-5.10 [-7.91, -2.29] -0.90 [-6.32, 4.52] -3.00 [-8.63, 2.63]

135.9 11.38 145 15

MacGregor AEBPA1982 Smith BPARA1985 Subtotal (95% CI)

148 160

13.91 16.1

23 20 174

155 14.87 162 15.2

23 20

6.6% 4.9%

-7.00 [-15.32, 1.32] -2.00 [-11.70, 7.70] -4.11 [-6.26, -1.97]

177 100.0%

Heterogeneity: Tau² = 0.00; Chi² = 2.62, df = 4 (P = 0.62); I² = 0% Test for overall effect: Z = 3.76 (P = 0.0002)

-20 Test for subgroup differences: Chi² = 0.22, df = 2 (P = 0.89), I² = 0% 

-10

Favours increased K

Favours control

Urinary potassium excretion is a common, valid form of estimating potassium intake. A factor of 1.30 is used to convert urinary potassium excretion to potassium intake (Stamler et al., 2003).

Figure 3.11 Resting systolic blood pressure – baseline sodium subgroups Increased potassium Control Mean Difference Mean Difference

Study or Subgroup

Mean

Total

Mean

SD Total Weight

IV, Random, 95% CI

5.12.1 Na intake at baseline (<2g/d)

Smith BPARA1985 Subtotal (95% CI) Heterogeneity: Not applicable

160

16.1

20 20

162

15.2

20 100.0% 20 100.0%

-2.00 [-11.70, 7.70] -2.00 [-11.70, 7.70]

Test for overall effect: Z = 0.40 (P = 0.69) 5.12.2 Na intake at baseline (2-4g/d) Barden BPARA1986 Berry BPA2010 113.67 120.7 8.87 9.9 43 113.69 48 122.2 9.44 13 43 48 8.9% 7.0% -0.02 [-3.89, 3.85] -1.50 [-6.12, 3.12]

Bulpitt BPA1985 Chalmers BPA1986 Forrester BPA1988

-5.2 -8.9 129.8

27.5 7 13.5

19 49 23

-7.5 -3.8 133.2

23.9 7.4 12.7

14 52 23

0.7% 12.7% 3.2%

2.30 [-15.30, 19.90] -5.10 [-7.91, -2.29] -3.40 [-10.97, 4.17]

Fotherby BPA1992 Grobbee BPA1987

176 135

20 13.28

18 40

186

18 40

1.0% 5.5%

-10.00 [-24.43, 4.43] -0.90 [-6.32, 4.52]

135.9 11.38

He BPA2010 Kaplan BPA1985 MacGregor AEBPA1982

142 127.6 148

11 11.6 13.91

42 16 23

145 133.2

15 16.8

42 16 23

5.2% 1.9% 2.7%

-3.00 [-8.63, 2.63] -5.60 [-15.60, 4.40] -7.00 [-15.32, 1.32]

155 14.87

Matlou BPA1986 Obel BPA1989 Siani BPARA1991

144 133 142.1

18.1 10 11.5

32 24 21

151 172 145.5

19.2 7 9.7

32 24 26

2.3%

-7.00 [-16.14, 2.14]

0.0% -39.00 [-43.88, -34.12] 4.5% -3.40 [-9.57, 2.77]

Trial Hyp Prv Col BPA1992 Valdes BPA1991 Whelton BPA1995 Subtotal (95% CI)

-0.78 138 -0.87

5.88 14.7 5.94

178 24 161 737

-0.84 145 -1

5.82 9.8 5.69

175 24 157

20.7% 3.6% 20.4%

0.06 [-1.16, 1.28] -7.00 [-14.07, 0.07] 0.13 [-1.15, 1.41] -1.97 [-3.41, -0.52]

733 100.0%

Heterogeneity: Tau² = 2.25; Chi² = 24.32, df = 14 (P = 0.04); I² = 42% Test for overall effect: Z = 2.66 (P = 0.008)

5.12.3 Na intake at baseline (>4g/d) Gu BPA2001 Kawano BPA1998 Patki BPARA1990

-13.1 147.9 143.6

10.7 11.7 10.8

68 55 37

-9.4

69 55 37

26.6% 23.7% 22.5%

-3.70 [-7.01, -0.39] -2.90 [-7.45, 1.65] -12.10 [-17.16, -7.04]

150.8 12.61 155.7 11.4

Richards BPAAEA1984 Siani BPA1987 Subtotal (95% CI)

148 131.8

14.2 12.7

12 18 190

149.9 145.8

14.5 11.3

12 19

10.6% 16.5%

-1.90 [-13.38, 9.58] -14.00 [-21.76, -6.24] -6.91 [-11.53, -2.29]

192 100.0%

Heterogeneity: Tau² = 18.05; Chi² = 13.78, df = 4 (P = 0.008); I² = 71% Test for overall effect: Z = 2.93 (P = 0.003)

-20 Test for subgroup differences: Chi² = 4.01, df = 2 (P = 0.13), I² = 50.1%

-10

Favours increased K

Favours control

Figure 3.12 Resting systolic blood pressure – duration subgroups Increased potassium Control Mean Difference Mean Difference

Study or Subgroup 5.7.1 Duration (<2 months) Barden BPARA1986

Mean

Total

Mean

SD Total Weight

IV, Random, 95% CI

113.67

8.87

43 113.69

9.44

16.7%

-0.02 [-3.89, 3.85]

Berry BPA2010 Forrester BPA1988 Fotherby BPA1992

120.7 129.8 176

9.9 13.5 20

48 23 18

122.2 133.2 186

13 12.7 24

48 23 18

11.7% 4.4% 1.2%

-1.50 [-6.12, 3.12] -3.40 [-10.97, 4.17] -10.00 [-24.43, 4.43]

Grobbee BPA1987 Gu BPA2001

135 -12.1

13.28 11.4

40 70

135.9 11.38 -5.7 10.6

40 71

8.5% 18.9%

-0.90 [-6.32, 4.52] -6.40 [-10.03, -2.77]

He BPA2010 Kaplan BPA1985 Kawano BPA1998

142 127.6 147.9

11 11.6 11.7

42 16 55

145 133.2

15 16.8

42 16 55

7.9% 2.5% 12.1%

-3.00 [-8.63, 2.63] -5.60 [-15.60, 4.40] -2.90 [-7.45, 1.65]

150.8 12.61

MacGregor AEBPA1982 Matlou BPA1986

148 144

13.91 18.1

23 32

155 14.87 151 19.2

23 32

3.6% 3.0%

-7.00 [-15.32, 1.32] -7.00 [-16.14, 2.14]

Obel BPA1989 Richards BPAAEA1984 Smith BPARA1985

133 148 160

4 14.2 16.1

24 12 20

172 149.9 162

7 14.5 15.2

24 12 20

0.0% -39.00 [-42.23, -35.77] 1.9% 2.7% -1.90 [-13.38, 9.58] -2.00 [-11.70, 7.70]

Valdes BPA1991 Subtotal (95% CI)

138

14.7

24 466

145

9.8

5.0%

-7.00 [-14.07, 0.07] -3.36 [-4.94, -1.78]

467 100.0%

Heterogeneity: Tau² = 0.00; Chi² = 10.52, df = 13 (P = 0.65); I² = 0% Test for overall effect: Z = 4.17 (P < 0.0001)

5.7.2 Duration (2 - 4 months) Bulpitt BPA1985 -5.2 27.5 19 -7.5 23.9 14 2.4% 2.30 [-15.30, 19.90]

Chalmers BPA1986 Gu BPA2001

-8.9 123.8

7 10.7

49 68

-3.8 124.6

7.4 9

52 69

18.0% 16.9%

-5.10 [-7.91, -2.29] -0.80 [-4.11, 2.51]

Obel BPA1989 Patki BPARA1990 Siani BPA1987

133 143.6 131.8

10 10.8 12.7

24 37 18

172 155.7 145.8

7 11.4 11.3

24 37 19

0.0% -39.00 [-43.88, -34.12] 13.0% 8.4% -12.10 [-17.16, -7.04] -14.00 [-21.76, -6.24]

Trial Hyp Prv Col BPA1992 Whelton BPA1995 Subtotal (95% CI)

-0.78 -0.6

5.88 6.82

178 171 540

-0.84 -0.14

5.82 7.27

175 168

20.9% 20.5%

0.06 [-1.16, 1.28] -0.46 [-1.96, 1.04] -3.83 [-6.72, -0.95]

534 100.0%

Heterogeneity: Tau² = 10.01; Chi² = 40.90, df = 6 (P < 0.00001); I² = 85%

Test for overall effect: Z = 2.60 (P = 0.009) 5.7.3 Duration (> 4 months)

Siani BPARA1991 Trial Hyp Prv Col BPA1992

142.1 -0.78

11.5 5.88

21 178

145.5 -0.84

9.7 5.82

26 175

2.0% 51.3%

-3.40 [-9.57, 2.77] 0.06 [-1.16, 1.28]

Whelton BPA1995 Subtotal (95% CI)

-0.87

5.94

161 360

-1

5.69

157

46.7%

0.13 [-1.15, 1.41] 0.02 [-0.85, 0.90]

358 100.0%

Heterogeneity: Tau² = 0.00; Chi² = 1.21, df = 2 (P = 0.55); I² = 0%

Test for overall effect: Z = 0.05 (P = 0.96)

-20 Test for subgroup differences: Chi² = 17.55, df = 2 (P = 0.0002), I² = 88.6%

-10

Favours increased K

Favours control

Figure 3.13 Resting systolic blood pressure – device subgroups Increased potassium Control Mean Difference Mean Difference

Study or Subgroup

Mean

Total

Mean

SD Total Weight

IV, Random, 95% CI

5.9.1 Type of BP device (automatic) Barden BPARA1986 Berry BPA2010 Chalmers BPA1986 Fotherby BPA1992

113.67 120.7 -8.9 176

8.87 9.9 7 20

43 113.69 48 49 18

9.44 13 7.4 24

43 48 52 18

18.8% 13.2% 35.7% 1.4%

-0.02 [-3.89, 3.85] -1.50 [-6.12, 3.12] -5.10 [-7.91, -2.29] -10.00 [-24.43, 4.43]

122.2 -3.8 186

He BPA2010 MacGregor AEBPA1982 Richards BPAAEA1984 Siani BPARA1991

142 148 148 142.1

11 13.91 14.2 11.5

42 23 12 21

145

42 23 12 26

8.9% 4.1% 2.1% 7.4%

-3.00 [-8.63, 2.63] -7.00 [-15.32, 1.32] -1.90 [-13.38, 9.58] -3.40 [-9.57, 2.77]

155 14.87 149.9 145.5 14.5 9.7

Smith BPARA1985 Valdes BPA1991 Subtotal (95% CI)

160 138

16.1 14.7

20 24 300

162 145

15.2 9.8

20 24

3.0% 5.6%

-2.00 [-11.70, 7.70] -7.00 [-14.07, 0.07] -3.45 [-5.13, -1.77]

308 100.0%

Heterogeneity: Tau² = 0.00; Chi² = 7.66, df = 9 (P = 0.57); I² = 0% Test for overall effect: Z = 4.03 (P < 0.0001) 5.9.2 Type of BP device (manual) Bulpitt BPA1985 Forrester BPA1988 Grobbee BPA1987 Gu BPA2001

-5.2 129.8 135 -13.1

27.5 13.5 13.28 10.7

19 23 40 68

-7.5 133.2

23.9 12.7

14 23 40 69

1.6% 6.3% 9.2% 13.2%

2.30 [-15.30, 19.90] -3.40 [-10.97, 4.17] -0.90 [-6.32, 4.52] -3.70 [-7.01, -0.39]

135.9 11.38 -9.4 9

Kaplan BPA1985 Kawano BPA1998 Matlou BPA1986

127.6 147.9 144

11.6 11.7 18.1

16 55 32

133.2

16.8

16 55 32

4.2% 10.7% 4.8%

-5.60 [-15.60, 4.40] -2.90 [-7.45, 1.65] -7.00 [-16.14, 2.14]

150.8 12.61 151 19.2

Obel BPA1989 Patki BPARA1990 Siani BPA1987 Trial Hyp Prv Col BPA1992

133 143.6 131.8 -0.78

10 10.8 12.7 5.88

24 37 18 178

172 155.7 145.8 -0.84

7 11.4 11.3 5.82

24 37 19 175

0.0% -39.00 [-43.88, -34.12] 9.8% 6.1% 17.1%

-12.10 [-17.16, -7.04] -14.00 [-21.76, -6.24] 0.06 [-1.16, 1.28]

Whelton BPA1995 Subtotal (95% CI)

-0.87

5.94

161 647

-1

5.69

157

17.0%

0.13 [-1.15, 1.41] -3.63 [-5.98, -1.28]

637 100.0%

Heterogeneity: Tau² = 8.00; Chi² = 41.19, df = 10 (P < 0.0001); I² = 76% Test for overall effect: Z = 3.03 (P = 0.002)

-20 Test for subgroup differences: Chi² = 0.02, df = 1 (P = 0.90), I² = 0%

-10

Favours increased K

Favours control

Figure 3.14 Resting systolic blood pressure – method subgroups Increased potassium Study or Subgroup Mean SD Total

Control Mean SD Total Weight

Mean Difference IV, Random, 95% CI

5.10.1 Type of BP measure (supine office SBP) Barden BPARA1986 Berry BPA2010

113.67 120.7

8.87 9.9

43 113.69 48 122.2

9.44 13

43 48

12.4% 11.2%

-0.02 [-3.89, 3.85] -1.50 [-6.12, 3.12]

Forrester BPA1988 Fotherby BPA1992

129.8 176

13.5 20

23 18

133.2 186

12.7 24

23 18

7.3% 3.0%

-3.40 [-10.97, 4.17] -10.00 [-24.43, 4.43]

Grobbee BPA1987 Kaplan BPA1985

135 127.6

13.28 11.6

40 16

135.9 11.38 133.2 16.8

40 16

10.0% 5.2%

-0.90 [-6.32, 4.52] -5.60 [-15.60, 4.40]

MacGregor AEBPA1982 Obel BPA1989

148 133

13.91 10

23 24

155 14.87 172 7

23 24

6.6%

-7.00 [-15.32, 1.32]

0.0% -39.00 [-43.88, -34.12]

Patki BPARA1990 Richards BPAAEA1984 Siani BPA1987 Siani BPARA1991 Smith BPARA1985 Valdes BPA1991

143.6 148 131.8 142.1 160 138

10.8 14.2 12.7 11.5 16.1 14.7

37 12 18 21 20 24

155.7 149.9 145.8 145.5 162 145

11.4 14.5 11.3 9.7 15.2 9.8

37 12 19 26 20 24

10.6% 4.3% 7.1% 9.0% 5.4% 7.9%

-12.10 [-17.16, -7.04] -1.90 [-13.38, 9.58] -14.00 [-21.76, -6.24] -3.40 [-9.57, 2.77] -2.00 [-11.70, 7.70] -7.00 [-14.07, 0.07]

Subtotal (95% CI)

343

349 100.0%

-4.88 [-7.64, -2.13]

Heterogeneity: Tau² = 12.08; Chi² = 24.89, df = 12 (P = 0.02); I² = 52% Test for overall effect: Z = 3.47 (P = 0.0005)

5.10.2 Type of BP measure (seated office SBP) Barden BPARA1986 Chalmers BPA1986

112.66 -8.9

9.44 7

43 113.52 49 -3.8

6.99 7.4

43 52

12.5% 15.5%

-0.86 [-4.37, 2.65] -5.10 [-7.91, -2.29]

Gu BPA2001 Kawano BPA1998

-13.1 147.9

10.7 11.7

68 55

-9.4

69 55

13.3% 9.2%

-3.70 [-7.01, -0.39] -2.90 [-7.45, 1.65]

150.8 12.61

Matlou BPA1986 Trial Hyp Prv Col BPA1992

144 -0.78

18.1 5.88

32 178

151 -0.84

19.2 5.82

32 175

3.1% 23.3%

-7.00 [-16.14, 2.14] 0.06 [-1.16, 1.28]

Whelton BPA1995 Subtotal (95% CI)

-0.87

5.94

161 586

-1

5.69

157

23.1%

0.13 [-1.15, 1.41] -1.83 [-3.54, -0.12]

583 100.0%

Heterogeneity: Tau² = 2.89; Chi² = 18.42, df = 6 (P = 0.005); I² = 67% Test for overall effect: Z = 2.09 (P = 0.04) 5.10.3 Type of BP measure (standing office SBP) Forrester BPA1988 Fotherby BPA1992 MacGregor AEBPA1982 Obel BPA1989 Patki BPARA1990 126.2 170 147 130 143.2 13.6 15 15.83 9 10.8 23 18 23 24 37 129 178 14.3 19 23 18 23 24 37 13.0% 8.3% 11.0% -2.80 [-10.87, 5.27] -8.00 [-19.18, 3.18] -8.00 [-17.15, 1.15]

155 15.83 168 156.4 7 10.9

0.0% -38.00 [-42.56, -33.44] 21.1% -13.20 [-18.14, -8.26]

Richards BPAAEA1984 Siani BPA1987

148.7 134.8

15.2 13.6

12 18

151.7 145.9

15.6 14

12 19

7.1% 11.4%

-3.00 [-15.32, 9.32] -11.10 [-19.99, -2.21]

Smith BPARA1985 Valdes BPA1991 Subtotal (95% CI)

160 138

9.84 14.7

20 24 175

160 15.65 143 9.8

20 24

12.9% 15.2%

0.00 [-8.10, 8.10] -5.00 [-12.07, 2.07] -6.94 [-10.63, -3.25]

176 100.0%

Heterogeneity: Tau² = 10.42; Chi² = 11.33, df = 7 (P = 0.12); I² = 38%

Test for overall effect: Z = 3.68 (P = 0.0002) 5.10.4 Type of BP measure (unspecified SBP) Bulpitt BPA1985 -5.2 27.5 19 -7.5 23.9 14 9.3% 2.30 [-15.30, 19.90]

He BPA2010 Subtotal (95% CI)

142

42 61

145

90.7%

-3.00 [-8.63, 2.63] -2.51 [-7.87, 2.85]

56 100.0%

Heterogeneity: Tau² = 0.00; Chi² = 0.32, df = 1 (P = 0.57); I² = 0% Test for overall effect: Z = 0.92 (P = 0.36)

-20 Test for subgroup differences: Chi² = 7.89, df = 3 (P = 0.05), I² = 62.0%

-10

Favours increased K

Favours control

Figure 3.15 Resting systolic blood pressure – medication status subgroups Increased potassium Control Mean Difference Mean Difference

Study or Subgroup

Mean

Total

Mean

SD Total Weight

IV, Random, 95% CI

5.11.1 Hypertension medication status (not taking medication)

Berry BPA2010 Chalmers BPA1986

120.7 -8.9

9.9 7

48 49

122.2 -3.8

13 7.4

48 52

8.9% 12.3%

-1.50 [-6.12, 3.12] -5.10 [-7.91, -2.29]

Fotherby BPA1992 Gu BPA2001 He BPA2010

176 -13.1 142

20 10.7 11

18 68 42

186 -9.4 145

24 9 15

18 69 42

1.8% 11.3% 7.3%

-10.00 [-24.43, 4.43] -3.70 [-7.01, -0.39] -3.00 [-8.63, 2.63]

MacGregor AEBPA1982 Matlou BPA1986 Obel BPA1989

148 144 133

13.91 18.1 10

23 32 24

155 14.87 151 172 19.2 7

23 32 24

4.4% 3.9%

-7.00 [-15.32, 1.32] -7.00 [-16.14, 2.14]

0.0% -39.00 [-43.88, -34.12]

Patki BPARA1990 Richards BPAAEA1984 Smith BPARA1985

143.6 148 160

10.8 14.2 16.1

37 12 20

155.7 149.9 162

11.4 14.5 15.2

37 12 20

8.1% 2.7% 3.5%

-12.10 [-17.16, -7.04] -1.90 [-13.38, 9.58] -2.00 [-11.70, 7.70]

Trial Hyp Prv Col BPA1992 Valdes BPA1991

-0.78 138

5.88 14.7

178 24

-0.84 145

5.82 9.8

175 24

15.1% 5.6%

0.06 [-1.16, 1.28] -7.00 [-14.07, 0.07]

Whelton BPA1995 Subtotal (95% CI)

-0.87

5.94

161 712

-1

5.69

157

15.0%

0.13 [-1.15, 1.41] -3.63 [-5.69, -1.57]

709 100.0%

Heterogeneity: Tau² = 6.95; Chi² = 43.52, df = 12 (P < 0.0001); I² = 72%

Test for overall effect: Z = 3.45 (P = 0.0006) 5.11.2 Hypertension medication status (population taking medication)

Bulpitt BPA1985 Forrester BPA1988 Kaplan BPA1985

-5.2 129.8 127.6

27.5 13.5 11.6

19 23 16

-7.5 133.2 133.2

23.9 12.7 16.8

14 23 16

6.5% 23.9% 16.5%

2.30 [-15.30, 19.90] -3.40 [-10.97, 4.17] -5.60 [-15.60, 4.40]

Siani BPA1987 Siani BPARA1991 Subtotal (95% CI)

131.8 142.1

12.7 11.5

18 21 97

145.8 145.5

11.3 9.7

19 26

23.2% 29.9%

-14.00 [-21.76, -6.24] -3.40 [-9.57, 2.77] -5.85 [-10.61, -1.08]

98 100.0%

Heterogeneity: Tau² = 9.84; Chi² = 6.07, df = 4 (P = 0.19); I² = 34% Test for overall effect: Z = 2.40 (P = 0.02)

5.11.3 Hypertension medication status (not specified/heterogeneous) Barden BPARA1986 113.67 8.87 43 113.69 9.44 43 44.7% -0.02 [-3.89, 3.85]

Grobbee BPA1987 Kawano BPA1998 Subtotal (95% CI)

135 147.9

13.28 11.7

40 55 138

135.9 11.38 150.8 12.61

40 55

22.8% 32.4%

-0.90 [-6.32, 4.52] -2.90 [-7.45, 1.65] -1.16 [-3.74, 1.43]

138 100.0%

Heterogeneity: Tau² = 0.00; Chi² = 0.90, df = 2 (P = 0.64); I² = 0% Test for overall effect: Z = 0.87 (P = 0.38)

-20 Test for subgroup differences: Chi² = 3.68, df = 2 (P = 0.16), I² = 45.6%

-10

Favours increased K

Favours control

Figure 3.16 Resting systolic blood pressure – type of intervention subgroups Increased potassium Control Mean Difference Mean Difference

Study or Subgroup

Mean

Total

Mean

SD Total Weight

IV, Random, 95% CI

5.14.1 Type of intervention (supplement) Barden BPARA1986 Berry BPA2010 Bulpitt BPA1985

113.67 120.7 -5.2

8.87 9.9 27.5

43 113.69 48 19

9.44 13 23.9

43 48 14

8.0% 6.9% 0.9%

-0.02 [-3.89, 3.85] -1.50 [-6.12, 3.12] 2.30 [-15.30, 19.90]

122.2 -7.5

Forrester BPA1988 Fotherby BPA1992 Grobbee BPA1987 Gu BPA2001

129.8 176 135 -13.1

13.5 20 13.28 10.7

23 18 40 68

133.2 186

12.7 24

23 18 40 69

3.8% 1.3% 5.8% 8.9%

-3.40 [-10.97, 4.17] -10.00 [-24.43, 4.43] -0.90 [-6.32, 4.52] -3.70 [-7.01, -0.39]

135.9 11.38 -9.4 9

He BPA2010 Kaplan BPA1985 Kawano BPA1998

142 127.6 147.9

11 11.6 11.7

42 16 55

145 133.2

15 16.8

42 16 55

5.6% 2.5% 7.0%

-3.00 [-8.63, 2.63] -5.60 [-15.60, 4.40] -2.90 [-7.45, 1.65]

150.8 12.61

MacGregor AEBPA1982 Matlou BPA1986 Obel BPA1989 Patki BPARA1990

148 144 133 143.6

13.91 18.1 10 10.8

23 32 24 37

155 14.87 151 172 155.7

23 32 24 37

3.3% 2.9%

-7.00 [-15.32, 1.32] -7.00 [-16.14, 2.14]

19.2 7 11.4

0.0% -39.00 [-43.88, -34.12] 6.3% -12.10 [-17.16, -7.04]

Richards BPAAEA1984 Siani BPA1987 Smith BPARA1985

148 131.8 160

14.2 12.7 16.1

12 18 20

149.9 145.8 162

14.5 11.3 15.2

12 19 20

2.0% 3.7% 2.6%

-1.90 [-13.38, 9.58] -14.00 [-21.76, -6.24] -2.00 [-11.70, 7.70]

Trial Hyp Prv Col BPA1992 Valdes BPA1991 Whelton BPA1995 Subtotal (95% CI)

-0.78 138 -0.87

5.88 14.7 5.94

178 24 161 877

-0.84 145 -1

5.82 9.8 5.69

175 24 157

12.2% 4.2% 12.1%

0.06 [-1.16, 1.28] -7.00 [-14.07, 0.07] 0.13 [-1.15, 1.41] -3.31 [-5.07, -1.55]

867 100.0%

Heterogeneity: Tau² = 6.22; Chi² = 48.47, df = 18 (P = 0.0001); I² = 63% Test for overall effect: Z = 3.68 (P = 0.0002) 5.14.2 Type of intervention (advice)

Berry BPA2010 Chalmers BPA1986 Siani BPARA1991 Subtotal (95% CI)

122.9 -8.9 142.1

12.37 7 11.5

48 49 21 118

124.8 -3.8 145.5

12.2 7.4 9.7

48 52 26

21.3% 65.2% 13.5%

-1.90 [-6.82, 3.02] -5.10 [-7.91, -2.29] -3.40 [-9.57, 2.77] -4.19 [-6.46, -1.92]

126 100.0%

Heterogeneity: Tau² = 0.00; Chi² = 1.30, df = 2 (P = 0.52); I² = 0% Test for overall effect: Z = 3.62 (P = 0.0003)

-20 Test for subgroup differences: Chi² = 0.36, df = 1 (P = 0.55), I² = 0%

-10

Favours increased K

Favours control

Figure 3.17 Resting systolic blood pressure – trial design subgroups Increased potassium Control Mean Difference Mean Difference

Study or Subgroup 5.15.1 Study design (parallel) Bulpitt BPA1985 Chalmers BPA1986 Gu BPA2001 Obel BPA1989

Mean

Total

Mean

SD Total Weight

IV, Random, 95% CI

-5.2 -8.9 -13.1 133

27.5 7 10.7 10

19 49 68 24

-7.5 -3.8 -9.4 172

23.9 7.4 9 7

14 52 69 24

1.7% 18.5% 16.8%

2.30 [-15.30, 19.90] -5.10 [-7.91, -2.29] -3.70 [-7.01, -0.39]

0.0% -39.00 [-43.88, -34.12]

Siani BPA1987 Siani BPARA1991 Trial Hyp Prv Col BPA1992 Whelton BPA1995 Subtotal (95% CI)

131.8 142.1 -0.78 -0.87

12.7 11.5 5.88 5.94

18 21 178 161 514

145.8 145.5 -0.84 -1

11.3 9.7 5.82 5.69

19 26 175 157

6.7% 9.2% 23.6% 23.5%

-14.00 [-21.76, -6.24] -3.40 [-9.57, 2.77] 0.06 [-1.16, 1.28] 0.13 [-1.15, 1.41] -2.74 [-5.08, -0.39]

512 100.0%

Heterogeneity: Tau² = 5.69; Chi² = 27.79, df = 6 (P = 0.0001); I² = 78% Test for overall effect: Z = 2.28 (P = 0.02) 5.15.2 Study design (cross-over) Barden BPARA1986 Berry BPA2010 113.67 120.7 8.87 9.9 43 113.69 48 122.2 9.44 13 43 48 13.4% 11.3% -0.02 [-3.89, 3.85] -1.50 [-6.12, 3.12]

Forrester BPA1988 Fotherby BPA1992 Grobbee BPA1987 He BPA2010

129.8 176 135 142

13.5 20 13.28 11

23 18 40 42

133.2 186

12.7 24

23 18 40 42

6.0% 2.0% 9.4% 9.0%

-3.40 [-10.97, 4.17] -10.00 [-24.43, 4.43] -0.90 [-6.32, 4.52] -3.00 [-8.63, 2.63]

135.9 11.38 145 15

Kaplan BPA1985 Kawano BPA1998 MacGregor AEBPA1982

127.6 147.9 148

11.6 11.7 13.91

16 55 23

133.2

16.8

16 55 23

3.8% 11.5% 5.2%

-5.60 [-15.60, 4.40] -2.90 [-7.45, 1.65] -7.00 [-15.32, 1.32]

150.8 12.61 155 14.87

Matlou BPA1986 Patki BPARA1990 Richards BPAAEA1984 Smith BPARA1985

144 143.6 148 160

18.1 10.8 14.2 16.1

32 37 12 20

151 155.7 149.9 162

19.2 11.4 14.5 15.2

32 37 12 20

4.4% 10.2% 3.0% 4.0%

-7.00 [-16.14, 2.14] -12.10 [-17.16, -7.04] -1.90 [-13.38, 9.58] -2.00 [-11.70, 7.70]

Valdes BPA1991 Subtotal (95% CI)

138

14.7

24 433

145

9.8

6.6%

-7.00 [-14.07, 0.07] -3.99 [-6.13, -1.86]

433 100.0%

Heterogeneity: Tau² = 4.94; Chi² = 19.16, df = 13 (P = 0.12); I² = 32% Test for overall effect: Z = 3.66 (P = 0.0002)

-20 Test for subgroup differences: Chi² = 0.60, df = 1 (P = 0.44), I² = 0%

-10

Favours increased K

Favours control

Figure 3.18 Ambulatory systolic blood pressure – all adults Increased potassium Study or Subgroup Berry BPA2010 Fotherby BPA1992 He BPA2010 Kawano BPA1998 Total (95% CI) Mean 132.3 154 139 133.9 SD 12 15 9 10.4 Control SD Total Weight 14 18 10 48 16 42 55 20.9% 4.3% 34.3% 40.5% Mean Difference IV, Random, 95% CI -1.80 [-7.02, 3.42] -6.00 [-17.48, 5.48] -3.00 [-7.07, 1.07] -3.40 [-7.15, 0.35] -3.04 [-5.42, -0.66] -20 -10 0 10 20 Mean Difference IV, Random, 95% CI Total Mean 48 134.1 16 42 160 142

55 137.3 9.64 161

161 100.0%

Heterogeneity: Tau² = 0.00; Chi² = 0.51, df = 3 (P = 0.92); I² = 0%

Test for overall effect: Z = 2.50 (P = 0.01)

Favours increased K

Favours control

Figure 3.19 Ambulatory diastolic blood pressure – all adults Increased potassium Study or Subgroup Berry BPA2010 Fotherby BPA1992 He BPA2010 Kawano BPA1998 Total (95% CI) Mean 81.9 87 87 80.1 SD 10.6 13 8 8.16 Control SD Total Weight 7.9 12 9 48 16 42 55 25.7% 4.8% 27.1% 42.4% Mean Difference IV, Random, 95% CI -1.40 [-5.14, 2.34] -2.00 [-10.67, 6.67] -1.00 [-4.64, 2.64] -1.20 [-4.11, 1.71] -1.24 [-3.13, 0.66] -10 -5 0 5 10 Mean Difference IV, Random, 95% CI Total Mean 48 16 42 55 161 83.3 89 88

81.3 7.42

161 100.0%

Heterogeneity: Tau² = 0.00; Chi² = 0.05, df = 3 (P = 1.00); I² = 0%

Test for overall effect: Z = 1.28 (P = 0.20)

Favours increased K

Favours control

Figure 3.20 Ambulatory systolic blood pressure – achieved intake subgroups (based on urinary potassium excretion*) Increased potassium Study or Subgroup Mean SD Control SD Total Weight Mean Difference IV, Random, 95% CI Mean Difference IV, Random, 95% CI Total Mean

4.1.1 Achieved K intake inter. v control (both <70mmol) Subtotal (95% CI) 0 0 Not estimable

Heterogeneity: Not applicable Test for overall effect: Not applicable 4.1.2 Achieved K in intervention (>=70mmol v <90mmol) Berry BPA2010 Subtotal (95% CI) Heterogeneity: Not applicable Test for overall effect: Z = 0.68 (P = 0.50)

132.3

48 134.1 48

48 100.0% 48 100.0%

-1.80 [-7.02, 3.42] -1.80 [-7.02, 3.42]

4.1.3 Achieved K in intervention (>=90mmol v <120mmol) Fotherby BPA1992 Kawano BPA1998 Subtotal (95% CI)

154 133.9

15 10.4

16 71

160

16 55

9.6% 90.4%

-6.00 [-17.48, 5.48] -3.40 [-7.15, 0.35] -3.65 [-7.21, -0.09]

55 137.3 9.64

71 100.0%

Heterogeneity: Tau² = 0.00; Chi² = 0.18, df = 1 (P = 0.67); I² = 0% Test for overall effect: Z = 2.01 (P = 0.04) 4.1.4 Achieved K in intervention (>120mmol) He BPA2010 Subtotal (95% CI) Heterogeneity: Not applicable Test for overall effect: Z = 1.45 (P = 0.15)

139

42 42

142

42 100.0% 42 100.0%

-3.00 [-7.07, 1.07] -3.00 [-7.07, 1.07]

-20 Test for subgroup differences: Chi² = 0.33, df = 2 (P = 0.85), I² = 0% 

-10

Favours increased K

Favours control

Urinary potassium excretion is a common, valid form of estimating potassium intake. A factor of 1.30 is used to convert urinary potassium excretion to potassium intake (Stamler et al., 2003).

Figure 3.21 Total cholesterol – all adults

Figure 3.22 HDL cholesterol – all adults

Figure 3.23 LDL cholesterol – all adults

Figure 3.24 Total triglycerides – all adults

Figure 3.25 Plasma adrenaline – all adults Increased potassium Study or Subgroup Grobbee BPA1987 Richards BPAAEA1984 Valdes BPA1991 Total (95% CI) Mean 67 73 42 SD 44.27 7 39.19 Control SD Total Weight 40 12 24 Mean Difference IV, Random, 95% CI Mean Difference IV, Random, 95% CI Total Mean 40 12 24 76

77 56.92 77 7

5.6% -10.00 [-32.35, 12.35] 88.8% 5.7% -4.00 [-9.60, 1.60] 3.00 [-19.17, 25.17] -3.94 [-9.22, 1.34] -20 -10 0 10 20

39 39.19

76 100.0%

Heterogeneity: Tau² = 0.00; Chi² = 0.66, df = 2 (P = 0.72); I² = 0% Test for overall effect: Z = 1.46 (P = 0.14)

Favours increased K

Favours control

Figure 3.26 Plasma noradrenaline – all adults Increased potassium Study or Subgroup Grobbee BPA1987 Richards BPAAEA1984 Valdes BPA1991 Total (95% CI) Mean 225 404 203 SD 132.82 28 97.98 Total Mean 40 12 24 76 Control SD Total Weight 40 12 24 Mean Difference IV, Random, 95% CI Mean Difference IV, Random, 95% CI

239 120.17 405 218 28 97.98

12.3% -14.00 [-69.51, 41.51] 75.4% -1.00 [-23.40, 21.40]

12.3% -15.00 [-70.44, 40.44] -4.32 [-23.78, 15.13] -100 -50 0 50 Favours control 100

76 100.0%

Heterogeneity: Tau² = 0.00; Chi² = 0.34, df = 2 (P = 0.84); I² = 0% Test for overall effect: Z = 0.44 (P = 0.66)

Favours increased K

Figure 3.27 Serum creatinine – all adults Increased potassium Study or Subgroup Bulpitt BPA1985 Patki BPARA1990 Smith BPARA1985 Total (95% CI) Mean 84 73.38 89 SD 18.7 6.96 16.99 Control SD Total Weight 39.2 14 37 20 Mean Difference IV, Random, 95% CI Mean Difference IV, Random, 95% CI Total Mean 19 110

12.4% -26.00 [-48.19, -3.81] 52.1% 35.4% -1.76 [-6.84, 3.32] -2.00 [-11.74, 7.74] -4.86 [-13.59, 3.87] -100 -50 0 50 Favours control 100

37 75.14 14.15 20 76 91 14.31

71 100.0%

Heterogeneity: Tau² = 31.33; Chi² = 4.39, df = 2 (P = 0.11); I² = 54% Test for overall effect: Z = 1.09 (P = 0.28)

Favours increased K

References to studies

An asterisk indicates that a reference is the primary reference for a study.

4.1

Included studies Barden BPARA1986

* Barden A, Vandongen R, Beilin LJ. Increases in urinary kallikrein activity and prostanoid synthesis after dietary potassium supplementation. Clinical and Experimental Pharmacology and Physiology, 1987, 14:565–572. Barden AE, Vandongen R, Beilin LJ et al. Potassium supplementation does not lower blood pressure in normotensive women. Journal of Hypertension, 1986, 4:339–343. Berry BPA2010

* Berry SE, Mulla UZ, Chowienczyk PJ et al. Increased potassium intake from fruit and vegetables or supplements does not lower blood pressure or improve vascular function in UK men and women with early hypertension: a randomised controlled trial. British Journal of Nutrition, 2010, 104:1839–1847. Bulpitt BPA1985

Bulpitt CJ, Ferrier G, Lewis PJ et al. Potassium supplementation fails to lower blood pressure in hypertensive patients receiving a potassium losing diuretic. Annals of Clinical Research, 1985, 17:126–130. Chalmers BPA1986

* Chalmers J, Morgan T, Doyle A et al. Australian National Health and Medical Research Council dietary salt study in mild hypertension. Journal of Hypertension – Supplement, 1986, 4:S629–S637. Forrester BPA1988

Forrester TE, Grell GA. Changes in red cell sodium content and blood pressure levels with potassium supplementation in black hypertensive patients. West Indian Medical Journal, 1988, 37:92–96. Fotherby BPA1992

Fotherby MD, Potter JF. Long-term potassium supplementation lowers blood pressure in elderly hypertensive subjects. International Journal of Clinical Practice, 1997, 51:219–222. * Fotherby MD, Potter JF. Potassium supplementation reduces clinic and ambulatory blood pressure in elderly hypertensive patients. Journal of Hypertension, 1992, 10:1403–1408. Grobbee BPA1987

Grobbee DE, Hofman A, Roelandt JT et al. Sodium restriction and potassium supplementation in young people with mildly elevated blood pressure. Journal of Hypertension, 1987, 5:115–119.

Gu BPA2001

* Gu D, He J, Wu X et al. Effect of potassium supplementation on blood pressure in Chinese: a randomized, placebo-controlled trial. Journal of Hypertension, 2001, 19:1325–1331. He BPA2010

Effect of potassium bicarbonate and potassium chloride on blood pressure and markers of target organ damage in hypertensives. Clinicaltrials.gov. [Other: NCT00160368] * He FJ, Marciniak M, Carney C et al. Effects of potassium chloride and potassium bicarbonate on endothelial function, cardiovascular risk factors, and bone turnover in mild hypertensives. Hypertension, 2010, 55:681–688. Kaplan BPA1985

Kaplan NM, Carnegie A, Raskin P et al. Potassium supplementation in hypertensive patients with diuretic-induced hypokalemia. New England Journal of Medicine, 1985, 312:746–749. Kawano BPA1998

Kawano Y, Minami J, Takishita S et al. Effects of potassium supplementation on ofﬁce, home and 24-hour blood pressure in patients with essential hypertension. American Journal of Hypertension, 1998, 11:1141–1146. MacGregor AEBPA1982

* MacGregor GA, Smith SJ, Markandu ND et al. Moderate potassium supplementation in essential hypertension. Archives des Maladies du Coeur et des Vaisseaux, 1984, 77:Spec No. 93–Spec No.100. MacGregor GA, Smith SJ, Markandu ND et al. Moderate potassium supplementation in essential hypertension. Lancet, 1982, 11:567–570. Smith SJ, Markandu ND, Banks RA et al. Does moderate potassium supplementation lower blood-pressure in essential-hypertension – a double-blind randomized crossover trial using slow potassium and placebo. Clinical Science, 1982, 63(3):123. Matlou BPA1986

* Matlou SM, Isles CG, Higgs A et al. Potassium supplementation in blacks with mild to moderate essential hypertension. Journal of Hypertension, 1986, 4:61–64. Obel BPA1989

Obel AO. Placebo-controlled trial of potassium supplements in black patients with mild essential hypertension. Journal of Cardiovascular Pharmacology, 1989, 42:294–296. Overlack BPARAAEA1991

Overlack A, Conrad H, Stumpe KO. The inﬂuence of oral potassium citrate/bicarbonate on blood pressure in essential hypertension during unrestricted salt intake. Klinische Wochenschrift, 1991, 69:S79–S83. Patki BPARA1990

Patki PS, Singh J, Gokhale SV et al. Efficacy of potassium and magnesium in essential hypertension: a double blind, placebo controlled, crossover study. BMJ, 1990, 301:521–523.

Richards BPAAEA1984

Richards AM, Nicholls MG, Espiner EA et al. Blood pressure response to moderate sodium restriction and to potassium supplementation in mild essential hypertension. Lancet, 1984, 1:757–761. Siani BPA1987

Siani A, Strazzullo P, Russo L et al. Controlled trial of long term oral potassium supplements in patients with mild hypertension. BMJ (Clinical Research Edition), 1987, 294:1453–1456. Siani BPARA1991

Siani A, Strazzullo P, Giacco A et al. Increasing the dietary potassium intake reduces the need for antihypertensive medication. Annals of Internal Medicine, 1991, 115:753–759. Smith BPARA1985

Smith SJ, Markandu ND, Sagnella GA et al. Moderate potassium chloride supplementation in essential hypertension: is it additive to moderate sodium restriction? BMJ (Clinical Research Edition), 1985, 290:110–113. Trial Hyp Prv Col BPA1992

The effects of nonpharmacologic interventions on blood pressure of persons with high normal levels. Results of the Trials of Hypertension Prevention, Phase I. Journal of the American Medical Association, 1992, 267:1213–1220. Valdes BPA1991

Valdés G, Vio CP, Montero J et al. Potassium supplementation lowers blood pressure and increases urinary kallikrein in essential hypertensives. Journal of Human Hypertension, 1991, 5:91–96. Whelton BPA1995

Whelton PK, Buring J, Borhani NO et al. Efficacy of nonpharmacologic interventions in adults with high-normal blood pressure: results from phase 1 of the Trials of Hypertension Prevention. Trials of Hypertension Prevention Collaborative Research Group. Annals of Epidemiology, 1997, 5:85–95. * Whelton PK, Kumanyika SK, Cook NR et al. The effect of potassium supplementation in persons with a high-normal blood pressure: results from phase I of the Trials of Hypertension Prevention. The American Journal of Clinical Nutrition, 1995, 65:S652–S660.

4.2

Excluded studies Agnoli RA1992

* Agnoli GC, Borgatti R, Cacciari M et al. Urinary prostanoid excretion in healthy women with different degrees of induced potassium depletion. Prostaglandins Leukot Essent Fatty Acids, 1992, 46(1):21–26. Agnoli GC, Borgatti R, Cacciari M et al. Effects of experimental potassium depletion on renal function and urinary prostanoid excretion in normal women during moderate anti-diuresis. Clinical Physiology, 1992, 12:79–93.

Agnoli RA1994

* Agnoli GC, Borgatti R, Cacciari M et al. Interactions between the renin-angiotensin system and prostanoids in modulating renal function in potassium-depleted healthy women. Prostaglandins, Leukotrienes and Essential Fatty Acids, 1994, 50:347–352. Barcelo RA1993

* Barcelo P, Wuhl O, Servitge E et al. Randomized double-blind study of potassium citrate in idiopathic hypocitraturic calcium nephrolithiasis. Journal of Urology, 1993, 150:1761–1764. Ceglia RA2009

* Ceglia L, Harris SS, Abrams SA et al. Potassium bicarbonate attenuates the urinary nitrogen excretion that accompanies an increase in dietary protein and may promote calcium absorption. Journal of Clinical Endocrinology & Metabolism, 2009, 94:645–653. Fujita AEA1984

* Fujita T, Ando K. Hemodynamic and endocrine changes associated with potassium supplementation in sodium-loaded hypertensives. Hypertension, 1984, 6:184–192. Gamarra BPA1994

* Gamarra G, Balaguera H, Corzo D et al. Tratamiento de la hipertensión arterial esencial con dieta rica en potasio. Acta Medica Colombiana, 1994, 19(1):15–23. Grimm BPA1990

* Grimm RH, Neaton JD, Elmer PJ et al. The inﬂuence of oral potassium chloride on blood pressure in hypertensive men on a low-sodium diet. New England Journal of Medicine, 1990, 322:569–574. Grimm RH, Kofron PM, Neaton JD et al. Effect of potassium supplementation combined with dietary sodium reduction on blood pressure in men taking antihypertensive medication. Journal of Hypertension, 1988, 6:S591–S593. Potassium and sodium to control blood pressure in hypertensives. ClinicalTrials.gov. [Other: NCT00000509] Heller RA1998

* Heller HJ, Reza-Albarrán AA, Breslau NA et al. Sustained reduction in urinary calcium during long-term treatment with slow release neutral potassium phosphate in absorptive hypercalciuria. Journal of Urology, 1998, 159:1451–1455. Jardim BPA1988

* Jardim PC, Branco RF, Silva EG. Evaluation of potassium supplementation in patients under furosemide therapy. Arquivos Brasileiros de Cardiologia, 1988, 51:153–156. Khaw BPA1982

* Khaw KT, Thom S. Randomized double-blind cross-over trial of potassium on blood pressure in normal subjects. Lancet, 1982, 2:1127–1129. Krishnan RA2010

* Krishnan AV. AUSSPRINT: Australian study of the effects of strict potassium restriction on neuropathy in chronic kidney disease (ACTRN12610000538044). 2010

Langford BPA1991

* Langford HG, Davis BR, Blaufox D et al. Effect of drug and diet treatment of mild hypertension on diastolic blood pressure. Hypertension, 1991, 17:210–217. Lennon RA1968

* Lennon EJ, Lemann J Jr. The effect of a potassium-deficient diet on the pattern of recovery from experimental metabolic acidosis. Clinical Science, 1968, 34:365–378. Med Res CWP BPA1987

* Comparison of the antihypertensive efficacy and adverse reactions to two doses of bendrofluazide and hydro-chlorothiazide and the effect of potassium supplementation on the hypotensive action of bendrofluazide: sub-studies of the Medical Research Council's Trials of Treatment of Mild Hypertension. Journal of Clinical Pharmacology, 1987, 27:271– 277. Overlack BPA RA AEA1995

* Overlack A, Maus B, Ruppert M et al. Potassium citrate vs potassium chloride in essential hypertension: effect on haemodynamic, hormonal and metabolic parameters. Deutsche Medizinische Wochenschrift, 1995, 120:631–635. Parfrey AEA1981

* Parfrey PS, Condon K, Wright P et al. Blood pressure and hormonal changes following alterations in dietary sodium and potassium in young men with and without a familial predisposition to hypertension. Lancet, 1981, 1:113–117. Poulter BPA1986

* Poulter NR, Sever PS. Moderate potassium supplementation: ineffective in black normotensives. East African Medical Journal, 1986, 63:798–802. Rahimi BPA2007

* Rahimi AR, Mahmoodpoor A, Sanaie S. The effect of high-calcium and high-potassium diet on grade-I hypertension and high normal blood pressure. Pakistan Journal of Medical Sciences, 2007, 23:589–592. Sanclemente BPA1987

* Sanclemente P. Essential arterial hypertension and dietary potassium supplement. CES Medicina, 1987, 1:65–70. Smith BPA1992

* Smith SR, Klotman PE, Svetkey LP. Potassium chloride lowers blood pressure and causes natriuresis in older patients with hypertension. Journal of the American Society of Nephrology, 1992, 2:1302–1309. Svetkey BPA1987

* Svetkey LP, Yarger WE, Feussner JR et al. Double-blind placebo-controlled trial of potassium chloride in the treatment of mild hypertension. Hypertension, 1987, 9:444–450.

WHOICTRP RA2008

WHOICTRP. The effects of potassium on physiological and cardiovascular variables in healthy persons (NCT00801034). Clinicaltrials.gov, World Health Organization International Clinical Trials Registry Platform (WHOICTRP), 2008.

4.3

Studies awaiting classification Barros BPA1984

* Barros MA, Brito IM. Potassium in essential arterial hypertension. CCS, 1984, 6:48–55. Braschi BPA2008

* Braschi A, Naismith DJ. The effect of a dietary supplement of potassium chloride or potassium citrate on blood pressure in predominantly normotensive volunteers. British Journal of Nutrition, 2008, 99:1284–1292. Cushman BPA1988

Cushman WC, Langford HG. Randomized controlled trial of potassium chloride versus placebo in mildly hypertensive blacks and whites. Circulation, 1988, 17:S370–S370. Hilary Green BPA2000

* Hilary Green J, Richards JK, Bunning RL. Blood pressure responses to high-calcium skim milk and potassium-enriched high-calcium skim milk. Journal of Hypertension, 2000, 18:1331–1339. Iimura BPA1979

* Iimura O, Kijima T, Kikuchi K et al. Studies on the hypotensive effect of high potassium intake in patients with essential hypertension. Clinical Science, 1981, 61:77s–80s. Kawano BPA1997

* Kawano Y, Minami J, Takishita S. Effects of potassium, calcium and magnesium supplementation in patients with essential hypertension – assessment by blood pressure monitoring. Therapeutic Research, 1997, 18(7):44–47. Morris BPA1995

* Morris R, O'Connor M, Forman A et al. Supplemental dietary potassium with KHCO3 but not KCl attenuates essential hypertension [abstract]. Journal of the American Society of Nephrology, 1995, 6:645–645.

4.4

Ongoing studies Mullan BPA2010

Mullan K. The renin-angiotensin-aldosterone (RAAS) axis, endothelial function and hypertension: diagnostic strategies, and therapeutic role of potassium supplementation - a randomised cross-over trial and an observational study (ISRCTN55798944). 2010 Turban BPA RA2009

Turban S. Potassium intake in patients with chronic kidney disease (NCT00949585). Clinicaltrials.gov, 2009

4.5

Other references Ascherio A, Rimm EB, Hernán MA et al. Intake of potassium, magnesium, calcium, and fiber and risk of stroke among US men. Circulation, 1998, 98:1198–1204. Cappuccio FP, MacGregor GA. Does potassium supplementation lower blood pressure? A meta-analysis of published trials. Journal of Hypertension, 1991, 9(5):465–473. Deeks JJ, Dinnes J, D'Amico R et al. Evaluating non-randomised intervention studies. Health Technol Assess, 2003, 7(27):iii-x, 1-173. Dickinson HO, Nicolson DJ, Campbell F et al. Potassium supplementation for the management of primary hypertension in adults. Cochrane Database of Systematic Reviews, 2006, 3:CD004641. Geleijnse JM, Kok FJ, Grobbee DE. Blood pressure response to changes in sodium and potassium intake: a metaregression analysis of randomised trials. Journal of Human Hypertension, 2003, 17(7):471–480. Guyatt GH, Oxman AD, Vist GE et al. GRADE: an emerging consensus on rating quality of evidence and strength of recommendations. BMJ, 2008, 336(7650):924-926. Higgins J, Thompson S, Deeks J et al. Statistical heterogeneity in systematic reviews of clinical trials: a critical appraisal of guidelines and practice. Journal of Health Services Research & Policy, 2002, 7(1):51–61. Higgins JP, Thompson SG. Quantifying heterogeneity in a meta-analysis. Statistics in Medicine, 2002, 21(11):1539–1558. Higgins JP, Thompson SG, Spiegelhalter DJ. A re-evaluation of random-effects meta-analysis. Journal of The Royal Statistical Society Series A-Statistics in Society, 2009, 172(1):137–159. Institute of Medicine. Dietary reference intakes: water, potassium, sodium, chloride, and sulfate. Washington, DC, National Academies Press, 2005 Khaw KT, Barrett-Connor E. Dietary potassium and stroke-associated mortality. A 12-year prospective population study. New England Journal of Medicine, 1987, 316:235–240. Lau J, Ioannidis JP, Terrin N et al. The case of the misleading funnel plot. BMJ, 2006, 333(7568):597–600. Liberati A, Altman DG, Tetzlaff J et al. The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate healthcare interventions: explanation and elaboration. BMJ, 2009, 339:b2700. Mackay J, Mensah G. Atlas of heart disease and stroke. Geneva, World Health Organization, 2004. Stamler J, Elliott P, Appel L et al. Higher blood pressure in middle-aged American adults with less education-role of multiple dietary factors: the INTERMAP study. Journal of Human Hypertension, 2003, 17(9):655–775. Sterne JA, Egger M. Funnel plots for detecting bias in meta-analysis: guidelines on choice of axis. Journal of Clinical Epidemiology, 2001, 54(10):1046–1055. 81

Strong K, Mathers C, Leeder S et al. Preventing chronic diseases: how many lives can we save? Lancet, 2005, 366(9496):1578–1582. Whelton PK, Appel LJ, Espeland MA et al. Sodium reduction and weight loss in the treatment of hypertension in older persons: a randomized controlled trial of nonpharmacologic interventions in the elderly (TONE). TONE Collaborative Research Group. JAMA, 1998, 279(11):839-846. Whelton PK, He J, Cutler JA et al. Effects of oral potassium on blood pressure. Meta-analysis of randomized controlled clinical trials. Journal of the American Medical Association, 1997, 277(20):1624–1632. WHO. Diet, nutrition and the prevention of chronic disease. Report of a Joint WHO/FAO Expert Consultation. Geneva, World Health Organization (WHO), 2003. WHO. Preventing chronic disease: A vital investment. Geneva, World Health Organization (WHO), 2005.

An n e x 1 : E l e c t r o n i c s e a r c h s t r a t e g y A1.1 Overview For each outcome, two EMBASE searches were conducted: one broad search according to the original protocol and another, more restrictive, search using more specific terms for each concept and an indexer limit for controlled trials. This strategy was supplied by the WHO librarian to facilitate data retrieval. MEDLINE was searched through the PubMed database for the previous 6 months only, because all references in MEDLINE are also found in EMBASE. EMBASE requires more time to update its database and, therefore, it is possible that some very recent studies could be captured in a PubMed search that would not be captured in EMBASE. All other databases were searched without any date limits. All electronic searches were first run to search for RCTs.

A1.2 Search for randomized controlled trials A1.2.1 EMBASE searches

Searches conducted on 25 August 2011 in EMBASE version available at http://www.embase.com . Note EMBASE.COM contains over 24 million indexed records and more than 7500 current, mostly peer-reviewed journals with over 2000 biomedical titles not currently offered by MEDLINE. MEDLINE citations are included in EMBASE. 1) Blood pressure

No language limits; dates needed: Jan 1 2004 to present. Restricted search Step Search terms Step 1 'potassium'/exp OR 'potassium chloride'/exp Step 2 'hypertension'/exp OR 'blood pressure'/exp AND ([cochrane review]/lim OR [controlled clinical trial]/lim OR [meta analysis]/lim OR [randomized controlled trial]/lim OR [systematic review]/lim) AND [2004–2012]/py Step 3 'dietary intake'/exp OR 'diet'/exp OR restrict*:ab,ti OR reduce*:ab,ti OR 'reduction'/exp OR intake:ab,ti OR diet:ab,ti OR dietary:ab,ti AND [2004– 2012]/py Step 4 Step 1 AND Step 2 AND Step 3

# Citations 111,188 339

324

169

Broader search Step Search terms Step 1 'potassium'/exp OR 'potassium' OR 'potassium chloride'/exp OR 'potassium chloride' OR potassium:ab,ti Step 2 'hypertension'/exp OR 'blood pressure'/exp OR 'hypertension'/exp OR 'blood pressure':ab,ti OR hypertensive:ab,ti OR 'blood pressure'/exp AND 'intravascular pressure':ab,ti OR normotension:ab,ti OR 'vascular pressure':ab,ti OR 'blood pressure monitoring'/exp Step 3 'dietary intake'/exp OR 'diet'/exp OR restrict*:ab,ti OR reduce*:ab,ti OR 'reduction'/exp OR intake:ab,ti OR diet:ab,ti OR dietary:ab,ti AND [2004– 2012]/py Step 4 Step 1 AND Step 2 AND Step 3 Step 5 'randomized controlled trial'/exp OR 'controlled trial':ab,ti OR 'randomized':ab,ti OR 'randomised':ab,ti OR placebo:ab,ti OR randomly:ab,ti OR trial:ab,ti Step 6 Step 4 AND Step 5 Step 5 Step 4 AND ([cochrane review]/lim OR [controlled clinical trial]/lim OR [meta analysis]/lim OR [randomized controlled trial]/lim OR [systematic review]/lim) Step 6 Step 4 AND Step 5 Step 7 Step 6 NOT [animals]/lim Step 8 Step 6 AND [animals]/lim AND [humans]/lim Step 9 Step 7 OR Step 8 Step 10 Step 9 NOT (Citations found in Restricted Search Step 3)

# Citations 301,383 684,004

925,155

62,873 833,589

1,256 930

468 721 20 741 575

2) Adverse effects

No language limits; no date limits. Restricted search Step Search terms Step 1 'potassium'/exp OR 'potassium chloride'/exp Step 2 'noradrenalin'/exp OR 'adrenor':ab,ti OR 'alginodia':ab,ti OR 'arterenal':ab,ti OR 'arterenol':ab,ti OR 'baycain green':ab,ti OR 'd noradrenalin':ab,ti OR 'dextro noradrenalin':ab,ti OR 'dextro noradrenaline':ab,ti OR 'dl arterenol':ab,ti OR 'dl noradrenalin':ab,ti OR 'dl noradrenalin hydrochloride':ab,ti OR 'l alpha aminomethyl 3, 4 dihydroxybenzyl alcohol':ab,ti OR 'l noradrenalin':ab,ti OR 'l noradrenalin hydrochloride':ab,ti OR 'l noradrenaline':ab,ti OR 'l norepinephrine':ab,ti OR 'levarterenol':ab,ti OR 'levo noradrenalin':ab,ti OR 'levo noradrenaline':ab,ti OR 'levo norepinephrine':ab,ti OR 'levonor':ab,ti OR 'levophed':ab,ti OR 'neomelubrin':ab,ti OR 'neurogenic noradrenalin':ab,ti OR 'noradrec':ab,ti OR 'noradrenalin hydrochloride':ab,ti OR 'noradrenalin reduction':ab,ti OR 'noradrenaline':ab,ti OR 'noradrine':ab,ti OR 'norepinephrin':ab,ti OR 'norepinephrine':ab,ti OR 'norepinephrine hydrochloride':ab,ti OR 'norexadrin':ab,ti OR 'revarterenol':ab,ti OR 'sympathin':ab,ti OR 'sympathin e':ab,ti OR 'catecholamine'/exp OR 'catechol amine; catecholamin':ab,ti OR 'catecholamines':ab,ti OR 'cathecholamine':ab,ti OR 'dextro pyrocatecholamine':ab,ti OR 'endogenous catecholamine':ab,ti OR 'pyrocatechinamine':ab,ti OR 'pyrocatecholamine':ab,ti OR 'hydroxy 5 cholestene':ab,ti OR '3beta hydroxy 5 cholestene':ab,ti OR '3beta hydroxycholest 5 ene':ab,ti OR '5 cholesten 3beta ol':ab,ti OR 'beta cholesterol':ab,ti OR 'cholest 5 en 3beta ol':ab,ti OR 'cholest 5 ene 3 ol':ab,ti OR 'cholesterin':ab,ti OR 'cholesterine':ab,ti OR 'cholesterol release':ab,ti OR 'dythol':ab,ti OR 'nsc 8798':ab,ti OR 'cholesterol'/exp OR 'riacylglycerol' OR 'acylglycerol, tri':ab,ti OR 'fatty acid triglyceride':ab,ti OR 'triacyl glyceride':ab,ti OR 'triglyceride':ab,ti OR 'triglycerides':ab,ti OR 'tryglyceride':ab,ti OR 'beta lipoprotein':ab,ti OR 'ldl':ab,ti OR 'lipoprotein, beta':ab,ti OR 'lipoprotein, low density':ab,ti OR 'lipoproteins, ldl'/exp OR 'low density lipoprotein'/exp OR 'lpha 7 lipoprotein':ab,ti OR 'alpha lipoprotein':ab,ti OR 'hdl':ab,ti OR 'high density lipoprotein phospholipid':ab,ti OR 'lipoprotein, alpha':ab,ti OR 'lipoprotein, high density':ab,ti OR 'lipoproteins, hdl':ab,ti OR 'pre alpha lipoprotein':ab,ti OR 'very high density lipoprotein'/exp OR 'high density lipoprotein'/exp AND ([cochrane review]/lim OR [controlled clinical trial]/lim OR [meta analysis]/lim OR [randomized controlled trial]/lim OR [systematic review]/lim) Step 3 Step 1 AND Step 2

# Citations 111,188 25,043

276

Broader search Step Search terms Step 1 sium'/exp OR 'potassium chloride'/exp OR potassium:ab,ti Step 2 'noradrenalin'/exp OR 'adrenor':ab,ti OR 'alginodia':ab,ti OR 'arterenal':ab,ti OR 'arterenol':ab,ti OR 'baycain green':ab,ti OR 'd noradrenalin':ab,ti OR 'dextro noradrenalin':ab,ti OR 'dextro noradrenaline':ab,ti OR 'dl arterenol':ab,ti OR 'dl noradrenalin':ab,ti OR 'dl noradrenalin hydrochloride':ab,ti OR 'l alpha aminomethyl 3, 4 dihydroxybenzyl alcohol':ab,ti OR 'l noradrenalin':ab,ti OR 'l noradrenalin hydrochloride':ab,ti OR 'l noradrenaline':ab,ti OR 'l norepinephrine':ab,ti OR 'levarterenol':ab,ti OR 'levo noradrenalin':ab,ti OR 'levo noradrenaline':ab,ti OR 'levo norepinephrine':ab,ti OR 'levonor':ab,ti OR 'levophed':ab,ti OR 'neomelubrin':ab,ti OR 'neurogenic noradrenalin':ab,ti OR 'noradrec':ab,ti OR 'noradrenalin hydrochloride':ab,ti OR 'noradrenalin reduction':ab,ti OR 'noradrenaline':ab,ti OR 'noradrine':ab,ti OR 'norepinephrin':ab,ti OR 'norepinephrine':ab,ti OR 'norepinephrine hydrochloride':ab,ti OR 'norexadrin':ab,ti OR 'revarterenol':ab,ti OR 'sympathin':ab,ti OR 'sympathin e':ab,ti OR 'catecholamine'/exp OR 'catechol amine; catecholamin':ab,ti OR 'catecholamines':ab,ti OR 'cathecholamine':ab,ti OR 'dextro pyrocatecholamine':ab,ti OR 'endogenous catecholamine':ab,ti OR 'pyrocatechinamine':ab,ti OR 'pyrocatecholamine':ab,ti OR 'hydroxy 5 cholestene':ab,ti OR '3beta hydroxy 5 cholestene':ab,ti OR '3beta hydroxycholest 5 ene':ab,ti OR '5 cholesten 3beta ol':ab,ti OR 'beta cholesterol':ab,ti OR 'cholest 5 en 3beta ol':ab,ti OR 'cholest 5 ene 3 ol':ab,ti OR 'cholesterin':ab,ti OR 'cholesterine':ab,ti OR 'cholesterol release':ab,ti OR 'dythol':ab,ti OR 'nsc 8798':ab,ti OR 'cholesterol'/exp OR 'riacylglycerol' OR 'acylglycerol, tri':ab,ti OR 'fatty acid triglyceride':ab,ti OR 'triacyl glyceride':ab,ti OR 'triglyceride':ab,ti OR 'triglycerides':ab,ti OR 'tryglyceride':ab,ti OR 'beta lipoprotein':ab,ti OR 'ldl':ab,ti OR 'lipoprotein, beta':ab,ti OR 'lipoprotein, low density':ab,ti OR 'lipoproteins, ldl'/exp OR 'low density lipoprotein'/exp OR 'lpha 7 lipoprotein':ab,ti OR 'alpha lipoprotein':ab,ti OR 'hdl':ab,ti OR 'high density lipoprotein phospholipid':ab,ti OR 'lipoprotein, alpha':ab,ti OR 'lipoprotein, high density':ab,ti OR 'lipoproteins, hdl':ab,ti OR 'pre alpha lipoprotein':ab,ti OR 'very high density lipoprotein'/exp OR 'high density lipoprotein'/exp 'randomized controlled trial'/exp OR 'controlled trial':ab,ti OR 'randomized':ab,ti OR 'randomised':ab,ti OR placebo:ab,ti OR randomly:ab,ti OR trial:ab,ti Step 1 AND Step 2 AND Step 3 Step 1 AND Step 2 AND ([cochrane review]/lim OR [controlled clinical trial]/lim OR [meta analysis]/lim OR [randomized controlled trial]/lim OR [systematic review]/lim) Step 4 OR Step 5 Step 6 NOT [animals]/lim Step 6 AND [animals]/lim AND [humans]/lim Step 6 OR Step 7 Step 9 NOT (Citations found in Restricted Search Step 3)

# Citations 185,013 588,203

Step 3

833,589

Step 4 Step 5

731 488

Step 6 Step 7 Step 8 Step 9 Step 10

842 720 22 742 468

3) Renal function

No language limits; no date limits Restricted search Step Search terms Step 1 'potassium'/exp OR 'potassium chloride'/exp Step 2 'kidney diseases':ab,ti OR 'kidney disorder':ab,ti OR 'kidney pathology':ab,ti OR 'nephropathy':ab,ti OR 'perinephritis':ab,ti OR 'perirenal infection':ab,ti OR 'renal disease':ab,ti OR 'renal disorder':ab,ti OR 'unilateral kidney disease':ab,ti OR 'kidney disease'/exp OR renal:ab,ti OR 'analgesic'/exp AND nephropathy:ab,ti OR 'chronic kidney disease':ab,ti OR 'cystinuria':ab,ti OR 'diabetic nephropathy':ab,ti OR 'fabry disease':ab,ti OR 'gitelman syndrome':ab,ti OR 'glomerulopathy':ab,ti OR 'gordon syndrome':ab,ti OR 'hepatorenal syndrome':ab,ti OR 'hiv associated nephropathy':ab,ti OR 'immunoglobulin a nephropathy':ab,ti OR 'kidney amyloidosis':ab,ti OR 'kidney calcification':ab,ti OR 'kidney colic':ab,ti OR 'kidney cyst':ab,ti OR 'kidney dysfunction':ab,ti OR 'kidney failure':ab,ti OR 'kidney fibrosis':ab,ti OR 'kidney hemorrhage':ab,ti OR 'kidney hypertrophy':ab,ti OR 'kidney infarction':ab,ti OR 'kidney infection':ab,ti OR 'kidney injury':ab,ti OR 'kidney ischemia':ab,ti OR 'kidney malformation':ab,ti OR 'kidney necrosis':ab,ti OR 'kidney pain':ab,ti OR 'kidney papilla necrosis':ab,ti OR 'kidney polycystic disease':ab,ti OR 'kidney rupture':ab,ti OR 'kidney scar':ab,ti OR 'kidney tubule acidosis':ab,ti OR 'kidney tubule damage':ab,ti OR 'kidney tubule disorder':ab,ti OR 'kidney tumor':ab,ti OR 'liddle syndrome':ab,ti OR 'lowe syndrome':ab,ti OR 'meckel syndrome':ab,ti OR 'medullary sponge kidney':ab,ti OR 'nephritis':ab,ti OR 'nephrogenic diabetes insipidus':ab,ti OR 'nephrolithiasis':ab,ti OR 'nephronophthisis':ab,ti OR 'nephrosis':ab,ti OR 'nephrotoxicity':ab,ti OR 'perirenal abscess':ab,ti OR 'prune belly syndrome':ab,ti OR 'pyelectasis':ab,ti OR 'reflux nephropathy':ab,ti OR 'renal diabetes':ab,ti OR 'renal graft dysfunction':ab,ti OR 'renovascular disease':ab,ti OR 'silent kidney':ab,ti OR 'uric acid nephropathy':ab,ti OR 'kidney disease'/exp AND ([cochrane review]/lim OR [controlled clinical trial]/lim OR [meta analysis]/lim OR [randomized controlled trial]/lim OR [systematic review]/lim) Step 3 Step 1 AND Step 2

# Citations 111,188 22,310

324

Broader search Step Search terms Step 1 'potassium'/exp OR 'potassium' OR 'potassium chloride'/exp OR 'potassium chloride' OR potassium:ab,ti Step 2 'kidney diseases':ab,ti OR 'kidney disorder':ab,ti OR 'kidney pathology':ab,ti OR 'nephropathy':ab,ti OR 'perinephritis':ab,ti OR 'perirenal infection':ab,ti OR 'renal disease':ab,ti OR 'renal disorder':ab,ti OR 'unilateral kidney disease':ab,ti OR 'kidney disease'/exp OR renal:ab,ti OR 'analgesic'/exp AND nephropathy:ab,ti OR 'chronic kidney disease':ab,ti OR 'cystinuria':ab,ti OR 'diabetic nephropathy':ab,ti OR 'fabry disease':ab,ti OR 'gitelman syndrome':ab,ti OR 'glomerulopathy':ab,ti OR 'gordon syndrome':ab,ti OR 'hepatorenal syndrome':ab,ti OR 'hiv associated nephropathy':ab,ti OR 'immunoglobulin a nephropathy':ab,ti OR 'kidney amyloidosis':ab,ti OR 'kidney calcification':ab,ti OR 'kidney colic':ab,ti OR 'kidney cyst':ab,ti OR 'kidney dysfunction':ab,ti OR 'kidney failure':ab,ti OR 'kidney fibrosis':ab,ti OR 'kidney hemorrhage':ab,ti OR 'kidney hypertrophy':ab,ti OR 'kidney infarction':ab,ti OR 'kidney infection':ab,ti OR 'kidney injury':ab,ti OR 'kidney ischemia':ab,ti OR 'kidney malformation':ab,ti OR 'kidney necrosis':ab,ti OR 'kidney pain':ab,ti OR 'kidney papilla necrosis':ab,ti OR 'kidney polycystic disease':ab,ti OR 'kidney rupture':ab,ti OR 'kidney scar':ab,ti OR 'kidney tubule acidosis':ab,ti OR 'kidney tubule damage':ab,ti OR 'kidney tubule disorder':ab,ti OR 'kidney tumor':ab,ti OR 'liddle syndrome':ab,ti OR 'lowe syndrome':ab,ti OR 'meckel syndrome':ab,ti OR 'medullary sponge kidney':ab,ti OR 'nephritis':ab,ti OR 'nephrogenic diabetes insipidus':ab,ti OR 'nephrolithiasis':ab,ti OR 'nephronophthisis':ab,ti OR 'nephrosis':ab,ti OR 'nephrotoxicity':ab,ti OR 'perirenal abscess':ab,ti OR 'prune belly syndrome':ab,ti OR 'pyelectasis':ab,ti OR 'reflux nephropathy':ab,ti OR 'renal diabetes':ab,ti OR 'renal graft dysfunction':ab,ti OR 'renovascular disease':ab,ti OR 'silent kidney':ab,ti OR 'uric acid nephropathy':ab,ti OR 'kidney disease'/exp Step 3 'randomized controlled trial'/exp OR 'controlled trial':ab,ti OR 'randomized':ab,ti OR 'randomised':ab,ti OR placebo:ab,ti OR randomly:ab,ti OR trial:ab,ti Step 4 Step 1 AND Step 2 AND Step 3 Step 5 Step 1 AND Step 2 AND ([cochrane review]/lim OR [controlled clinical trial]/lim OR [meta analysis]/lim OR [randomized controlled trial]/lim OR [systematic review]/lim) Step 6 Step 4 OR Step 5 Step 7 Step 6 NOT [animals]/lim Step 8 AND [animals]/lim AND [humans]/lim Step 9 Step 7 OR Step 8 Step 10 'dietary intake'/exp OR 'diet'/exp OR restrict*:ab,ti OR reduce*:ab,ti OR reduction Step 11 Step 9 AND Step 10 Step 12 Step 11 NOT (Citations found in Restricted Search Step 3)

# Citations 301,383 573,788

833,589

1256 930

1609 1154 40 1194 2,617,056 806 626

A1.2.2 PubMed searches No language limits; date conducted: 28 Aug 2011; date limit: previous 180 days.

1) Blood pressure

(blood pressure[MeSH] OR hypertension[MeSH] OR blood pressure[tiab] OR hypertension[tiab]) AND (potassium[MeSH] OR potassium chloride[MeSH] OR potassium[tiab] OR potassium chloride[tiab]) AND (diet[MeSH] OR dietary[MeSH] OR intake[MeSH] OR restriction[MeSH] or reduction[MeSH] OR diet[tiab] OR dietary[tiab] OR intake[tiab] OR restriction[tiab] or reduction[tiab]) AND (randomized controlled trial [pt] OR controlled clinical trial [pt] OR randomized [tiab] OR placebo [tiab] OR drug therapy [sh] OR randomly [tiab] OR trial [tiab] OR groups [tiab]) NOT (animals [mh] NOT humans [mh])

2) Adverse effects

(potassium[MeSH] OR potassium[tiab]) AND (noradrenaline[MeSH] OR norepinephrine[MeSH] OR noradrenaline[tiab] OR norepinephrine[tiab] OR catecholamine[MeSH] OR catecholamine[tiab] OR cholesterol[MeSH] OR triglycerides[MeSH] OR low density lipoprotein[MeSH] OR high density lipoprotein[MeSH] OR LDL[tiab] OR HDL[tiab] OR cholesterol[tiab] OR triglyceride[tiab]) AND (randomized controlled trial [pt] OR controlled clinical trial [pt] OR randomized [tiab] OR placebo [tiab] OR drug therapy [sh] OR randomly [tiab] OR trial [tiab] OR groups [tiab]) NOT (animals [mh] NOT humans [mh])

3) Renal function

(potassium[MeSH] OR potassium[tiab]) AND (kidney disease[MeSH] OR renal[tiab]) AND (dietary[MeSH] OR diet[MeSH] OR diets[MeSH] OR restriction[MeSH] OR reduction[MeSH] OR reduce[MeSH] OR restrict[MeSH]) AND (randomized controlled trial [pt] OR controlled clinical trial [pt] OR randomized [tiab] OR placebo [tiab] OR drug therapy [sh] OR randomly [tiab] OR trial [tiab] OR groups [tiab]) NOT (animals [mh] NOT humans [mh]) A1.2.3 LILACS searches No language limits; date conducted: 01 Sept 2011; date limit: none

Query Blood pressure Adverse effects

Search potassium AND blood pressure potassium AND hypertension potassium AND noradrenaline potassium AND norepinephrine potassium AND catecholamine potassium AND lipoprotein potassium AND hdl potassium AND ldl potassium AND cholesterol potassium AND triglyceride potassium AND renal

Renal disease

A1.2.4 WHO International Clinical Trials Registry Platform searches No language limits; date conducted: 01 Sept 2011; date limit: none

Query Blood pressure Adverse effects

Search (potassium AND blood pressure) OR (potassium AND hypertension) (potassium and noradrenaline) OR (potassium and norepinephrine) OR (potassium and catecholamine) OR (potassium and lipoprotein) OR (potassium and hdl) OR (potassium and ldl) OR (potassium and cholesterol) OR (potassium and triglyceride) potassium AND renal

Renal disease

A1.2.5 Cochrane Central Register of Controlled Trials searches No language limits; date conducted: 06 Sept 2011; date limit: none

Query Blood pressure

Search (( #1 OR #2 OR ( blood AND pressure ) OR hypertension ) AND ( #3 OR #4 OR potassium OR ( potassium AND chloride ) ) AND ( #5 OR diet OR dietary OR intake OR restriction OR reduction ) AND ( ( randomized AND controlled AND trial ) OR ( controlled AND clinical AND trial ) OR randomized OR placebo OR ( drug AND therapy ) OR randomly OR trial OR groups ))* (( #3 OR #4 OR potassium OR ( potassium AND chloride ) ) AND ( #6 OR #7 OR #8 OR noradrenaline OR norepinephrine OR noradrenaline OR catecholamine OR cholesterol OR triglycerides OR ( low AND density AND lipoprotein ) OR ( high AND density AND lipoprotein ) OR LDL OR HDL ) AND ( ( randomized AND controlled AND trial ) OR ( controlled AND clinical AND trial ) OR randomized OR placebo OR ( drug AND therapy ) OR randomly OR trial OR groups )) (( ( renal AND disease ) OR renal ) AND ( #3 OR #4 OR potassium OR ( potassium AND chloride ) ) AND ( #5 OR diet OR dietary OR intake OR restriction OR reduction ) AND ( ( randomized AND controlled AND trial ) OR ( controlled AND clinical AND trial ) OR randomized OR placebo OR ( drug AND therapy ) OR randomly OR trial OR groups ))

Adverse effects

Renal disease

#1 = MeSH descriptor Blood Pressure explode all trees #2 = MeSH descriptor Hypertension explode all trees #3 = MeSH descriptor Potassium explode all trees #4 = MeSH descriptor Potassium Chloride explode all trees #5 = MeSH descriptor Diet explode all trees #6 = MeSH descriptor Norepinephrine explode all trees #7 = MeSH descriptor Cholesterol explode all trees #8 = MeSH descriptor Triglycerides explode all trees * “dietary”, “intake”, “reduction”, “restriction” did not retrieve MeSH terms **”noradrenaline”, “catecholamine”, “low density lipoprotein”, “high density lipoprotein” did not retrieve MeSH terms

An n e x 2 : E x a m p l e d a ta e x t r a c t i o n t e m p l a t e The data extraction form was an Excel workbook divided into worksheets per topic area. The following sheets are examples demonstrating the data collected in the Excel workbook. Sheet 1: Reference data Author communication needed? Comparisons made in study Final Exclusion (Y/N)?

Trial ID

Ref ID

Duplicate

Extract ID

Extract date

Author

Author contact

Year

Journal

Volume

Pages

Rep Type

Language

Country

Sponsor

Inclusion

Exclusion

Sheet 2: Risk of bias information Random Allocation sequence concealment generation risk of bias risk of bias Blinding Subject Provider Incomplete Selective outcome reporting Outcome Risk of bias data risk of risk of assessment bias bias Risk of other bias Risk of bias due to systematic differences in care

Trial ID

Sheet 3: Study design Parallel design or crossover study Method of measurement of potassium intake Dietary 24-hour urinary excretion Outcomes measured Type of intervention

Trial ID

Description of assessment of compliance

Adults All

Adults Normotensive

Adults Hypertensive

Subgroup analyses performed Children Feeding

Supplement

Diet advice / Education

Other

Sheet 4: Participant characteristics Part 1 BP Group Intervention and Group control groups descript Hyper Normo comparable at Both Unspecified ion tensive tensive baseline (Y/N) Age group Age (Mean and SD) Sex (%M) N originally randomiz ed Final samples (n) % Loss to follow up (%)

Ref ID

Comparison ID

Group

Adult

Children

Both

Part 2 Potassium intake at follow-up Potassium intake at baseline (mmol K/day) Potassium intake achieved at follow-up (mmol K/day) Sodium intake at baseline

> 70 > 90 Other (if >120 mmol/day mmol/day 'Other', see mmol/day intervention intervention column T: >4 intervention <2g/d 2-4 g/d vs < 70 vs < 90 Potassium g/day vs < 120 mmol mmol intake achieved mmol control control control at follow-up)

Other

Starting Ending time of time of interventi interventi on on

Duration of follow up (months)

Sheet 5: Outcomes (continuous) Part 1

Ref ID

Comparison ID

Group

Systolic blood n Mean (SD)

Diastolic blood n Mean (SD)

Adrenaline n Mean (SD)

Nonadrenaline n Mean (SD)

Cholesterol n Mean (SD)

Part 2

Ref ID

Comparison ID

Group

Triglyceride n Mean (SD)

HDL n

Mean (SD)

LDL n

Mean (SD)

NPS n

Mean (SD)

An n e x 3 : F u n n e l p l o ts Funnel plot 1: Resting systolic blood pressure

Funnel plot 2: Resting diastolic blood pressure

An n e x 4 : R i s k o f b i a s s u m m a r y

Blinding of participants and personnel (performance bias)

Blinding of outcome assessment (detection bias)

Random sequence generation (selection bias)

Incomplete outcome data (attrition bias) + + +

Allocation concealment (selection bias)

Barden BPARA1986

? + ?

Berry BPA2010 Bulpitt BPA1985 Chalmers BPA1986 Forrester BPA1988 Fotherby BPA1992 Grobbee BPA1987 Gu BPA2001

? – ? ? + + ?

– – + + + + +

– ? + ? + + +

? + + ? + + –

He BPA2010 Kaplan BPA1985 Kawano BPA1998 MacGregor AEBPA1982 Matlou BPA1986 Obel BPA1989

? ? ? ? ? ?

? + – + – +

? ? + ? + ?

+ + + + + ?

Overlack BPARAAEA1991 Patki BPARA1990 Richards BPAAEA1984 Siani BPA1987 Siani BPARA1991 Smith BPARA1985 Trial Hyp Prv Col BPA1992

? ? ? ? + ? ?

? + ? ? + ? ?

– + – + + + +

? + ? + + ? +

– ? + + + ? +

Valdes BPA1991 Whelton BPA1995

Selective reporting (reporting bias) + + +

+ + + + + + +

+ + + + + +

+ + + + + + +

An n e x 5 : R i s k o f b i a s g r a p h

Random sequence generation (selection bias) Allocation concealment (selection bias) Blinding of participants and personnel (performance bias) Blinding of outcome assessment (detection bias)

Incomplete outcome data (attrition bias) Selective reporting (reporting bias) 0% 25% 50% 75% 100%

Low risk of bias

Unclear risk of bias

High risk of bias

An n e x 6 : G R AD E e vi d e n c e p r o f i l e s Research question: What is the effect of increased potassium relative to usual intake in adults? Quality assessment No of studies/ comparisons Design Risk of bias Inconsistency Indirectness Participants Other consider Increased Imprecision ations Potassium Control Effect Relative (95% CI) Quality  MODERATE Importance

Resting systolic blood pressure (follow-up 1 - 12 months; units mmHg; better indicated by lower values) 21/21 randomised trials no serious serious1 risk of bias no serious indirectness no serious imprecision none 947 945 MD 3.49 lower (5.15 to 1.82 lower) CRITICAL

Resting diastolic blood pressure (follow-up 1 - 12 months; units mmHg; better indicated by lower values) 21/21 randomised trials no serious serious1 risk of bias no serious indirectness no serious imprecision none 930 927 MD 3.02 lower (4.9 to 1.2 lower)  MODERATE CRITICAL

Ambulatory systolic blood pressure (follow-up 1 - 1.5 months; units mmHg; better indicated by lower values) 4/4 randomised trials no serious no serious risk of bias inconsistency no serious indirectness serious2 none 161 161 MD 3.04 lower (5.4 to 0.7 lower)  MODERATE CRITICAL

Ambulatory diastolic blood pressure (follow-up 1 - 1.5 months; units mmHg; better indicated by lower values) 4/4 randomised trials no serious no serious risk of bias inconsistency no serious indirectness serious3 none 161 161 MD 1.24 lower (3.1 lower to 0.7 higher)  MODERATE CRITICAL

Serum creatinine (follow-up 1 - 3 months; units µmol/L; better indicated by lower values) 3/3 randomised trials no serious no serious risk of bias inconsistency no serious indirectness no serious imprecision4 none 76 71 MD 4.86 lower (13.6 lower to 3.9 higher)  HIGH IMPORTANT

Total cholesterol (follow-up 1.5 months; units mmol/L ; better indicated by lower values) 3/3 randomised trials no serious no serious risk of bias inconsistency no serious indirectness no serious imprecision4 none 104 104 MD 0.12 lower (0.33 lower to 0.09 higher)  HIGH IMPORTANT

HDL cholesterol (follow-up mean 1.5 months; units mmol/L; better indicated by higher values) 2/2 randomised trials no serious no serious risk of bias inconsistency no serious indirectness no serious imprecision4 none 64 64 MD 0.01 lower (0.13 lower to 0.11 higher)  HIGH5  HIGH6  HIGH5  HIGH IMPORTANT

LDL cholesterol (follow-up 1.5 months; units mmol/L; better indicated by lower values) 1/1 randomised trials no serious no serious risk of bias inconsistency no serious indirectness no serious imprecision4 none 48 48 MD 0.10 lower (0.38 lower to 0.18 higher) IMPORTANT

Triglycerides (follow-up mean 1.5 months; units mmol/L ; better indicated by lower values) 2/2 randomised trials no serious no serious risk of bias inconsistency no serious indirectness no serious imprecision4 none 64 64 MD 0.11 lower (0.48 lower to 0.26 higher) IMPORTANT

Adrenaline (plasma) (follow-up 1 - 1.5 months; units pg/mL; better indicated by lower values) 3/3 randomised trials no serious no serious risk of bias inconsistency no serious indirectness no serious imprecision4 none 76 76 MD 3.94 lower (9.2 lower to 1.3 higher) IMPORTANT

Noradrenaline (plasma) (follow-up 1 - 1.5 months; units pg/mL; better indicated by lower values) 3/3 1 2

randomised trials

no serious no serious risk of bias inconsistency

no serious indirectness

no serious imprecision4

none

MD 4.32 lower (23.8 lower to 15.1 higher)

 HIGH

IMPORTANT

95%CI do not always overlap 95%CI very near zero 3 95%CI crosses zero 4 Outcome is a measure of adverse effect and thus a null value is the perferred value and crossing zero is not considered a matter of imprecision 5 Only two studies included in generation of estimate 6 Only one study included in generation of estimate

Research question: What is the effect of increased potassium intake to a level that results in <70 mmol urinary potassium excretion/day relative to usual intake in adults? Participants Other consider Increased Imprecision ations Potassium

Quality assessment No of studies/ comparisons Design Risk of bias Inconsistency Indirectness

Effect Relative

Control

(95% CI)

Quality

Importance

Resting systolic blood pressure (follow-up 1.5 - 12 months; units mmHg; better indicated by lower values) 2/2 randomised trials no serious no serious risk of bias inconsistency no serious indirectness no serious imprecision none 91 92 MD 3.7 lower (6.7 to 0.6 lower)  HIGH2  MODERATE2 CRITICAL

Resting diastolic blood pressure (follow-up 1.5 - 12 months; units mmHg; better indicated by lower values) 2/2 randomised trials no serious no serious risk of bias inconsistency no serious indirectness serious1 none 91 92 MD 1.35 lower (5.3 lower to 2.6 higher) CRITICAL

Ambulatory systolic blood pressure (better indicated by lower values) 0/0 CRITICAL

Ambulatory diastolic blood pressure (better indicated by lower values) 0/0 CRITICAL

Serum creatinine (better indicated by lower values) 0/0 IMPORTANT

Total cholesterol (better indicated by lower values) 0/0 IMPORTANT

HDL cholesterol (better indicated by higher values) 0/0 IMPORTANT

LDL cholesterol (better indicated by lower values) 0/0 IMPORTANT

Triglycerides (better indicated by lower values) 0/0 IMPORTANT

Adrenaline (plasma) (better indicated by lower values) 0/0 IMPORTANT

Noradrenaline (plasma) (better indicated by lower values) 0/0 IMPORTANT

1 2

95%CI crosses zero Only two studies contributed to the effect estimate

Research question: What is the effect of increased potassium intake to a level that results in 70–90 mmol urinary potassium excretion/day relative to usual intake in adults? Quality assessment No of studies/ comparisons Design Risk of bias Inconsistency Indirectness Participants Other consider Increased Imprecision ations Potassium Control Effect Relative (95% CI) Quality Importance

Resting systolic blood pressure (follow-up 1.5 - 12 months; units mmHg; better indicated by lower values) 5/5 randomised trials no serious serious1 risk of bias no serious indirectness no serious imprecision none 140 146 MD 7.16 lower (12.41 to 1.91 lower)  MODERATE CRITICAL

Resting diastolic blood pressure (follow-up 1.5 - 12 months; units mmHg; better indicated by lower values) 4/4 randomised trials no serious no serious risk of bias inconsistency no serious indirectness serious 2

none

103

109

MD 4.01 lower (8.44 lower to 0.42 higher)

 MODERATE

CRITICAL

Ambulatory systolic blood pressure (follow-up 1.5 months; units mmHg; better indicated by lower values) 1/1 randomised trials no serious no serious risk of bias inconsistency no serious indirectness serious2 none 48 48 MD 1.8 lower (7.02 lower to 3.42 higher)  MODERATE  MODERATE3  HIGH3  HIGH5  HIGH5  HIGH3  MODERATE5 3

CRITICAL

Ambulatory diastolic blood pressure (follow-up 1.5 months; units mmHg; better indicated by lower values) 1/1 randomised trials no serious no serious risk of bias inconsistency no serious indirectness serious 2

none

MD 1.4 lower (5.14 lower to 2.34 higher)

CRITICAL

Serum creatinine (follow-up 1 - 3 months; units µmol/L; better indicated by lower values) 1/1 randomised trials no serious no serious risk of bias inconsistency no serious indirectness no serious imprecision4 none 37 37 MD 1.8 lower (6.8 lower to 3.3 higher) IMPORTANT

Total cholesterol (follow-up 1.5 months; units mmol/L; better indicated by lower values) 2/2 randomised trials no serious no serious risk of bias inconsistency no serious indirectness no serious imprecision4 none 64 64 MD 0.17 lower (0.43 lower to 0.09 higher) IMPORTANT

Triglycerides (follow-up mean 1.5 months; units mmol/L ; better indicated by lower values) 2/2 randomised trials no serious serious risk of bias inconsistency6 no serious indirectness no serious imprecision4 none 64 64 MD 0.11 lower (0.48 lower to 0.26 higher) IMPORTANT

Adrenaline (plasma) (better indicated by lower values) 0/0 IMPORTANT

Noradrenaline (plasma) (better indicated by lower values) 0/0 IMPORTANT

1 2 3 4 5 6

95%CI not always overlap 95%CI crosses zero Only one study contributed to effect estimate Outcome is a measure of adverse effect and thus a null value is the perferred value and crossing zero is not considered a matter of imprecision Only two studies contributed to effect estimate Effect point estimates fall on both sides of zero

100

Research question: What is the effect of increased potassium intake to a level that results in 90–120 mmol urinary potassium excretion/day relative to usual intake in adults?

Quality assessment No of studies/ comparisons Design Risk of bias Inconsistency Indirectness

Participants Other consider Increased Imprecision ations Potassium

Effect Relative

Control

(95% CI)

Quality

Importance

Resting systolic blood pressure (follow-up 1.5 - 12 months; units mmHg; better indicated by lower values) 10 / 10 randomised trials no serious no serious risk of bias inconsistency no serious indirectness serious1 none 598 589 MD 1.71 lower (3.42 lower to 0 higher)  MODERATE CRITICAL

Resting diastolic blood pressure (follow-up 1.5 - 12 months; units mmHg; better indicated by lower values) 9/9 randomised trials no serious no serious risk of bias inconsistency no serious indirectness serious2 none 532 519 MD 0.83 lower (1.82 lower to 0.17 higher)  MODERATE CRITICAL

Ambulatory systolic blood pressure (follow-up 1.5 months; units mmHg; better indicated by lower values) 2/2 randomised trials no serious no serious risk of bias inconsistency no serious indirectness serious1 none 71 71 MD 3.65 lower (7.21 to 0.09 lower)  MODERATE CRITICAL

Ambulatory diastolic blood pressure (follow-up 1.5 months; units mmHg; better indicated by lower values) 2/2 randomised trials no serious no serious risk of bias inconsistency no serious indirectness serious2 none 71 71 MD 1.28 lower  CRITICAL (4.04 lower to 1.48 higher) MODERATE3  HIGH3

Serum creatinine (follow-up 1 - 3 months; units µmol/L; better indicated by lower values) 2/2 randomised trials no serious no serious risk of bias inconsistency no serious indirectness no serious imprecision4 none 39 34 MD 11.8 lower (35.0 lower to 11.3 higher) IMPORTANT

Total cholesterol (better indicated by lower values) 0/0 IMPORTANT

HDL cholesterol (better indicated by higher values) 0/0 IMPORTANT

LDL cholesterol (better indicated by lower values) 0/0 IMPORTANT

Triglycerides (better indicated by lower values) 0/0 IMPORTANT

Adrenaline (plasma) (better indicated by lower values) 0/0 IMPORTANT

Noradrenaline (plasma) (better indicated by lower values) 0/0 IMPORTANT

1 2 3 4

95%CI reaches/is very close to zero 95%CI crosses zero Only two studies contributed to effect estimate Outcome is a measure of adverse effect and thus a null value is the perferred value and crossing zero is not considered a matter of imprecision

101

Research question: What is the effect of increased potassium intake to a level that results in >120 mmol urinary potassium excretion/day relative to usual intake in adults? Participants Other consider Increased Imprecision ations Potassium Control

Quality assessment No of studies/ comparisons Design Risk of bias Inconsistency Indirectness

Effect Relative (95% CI)

Quality

Importance

Resting systolic blood pressure (follow-up 1.5 - 12 months; units mmHg; better indicated by lower values) 4/4 randomised trials no serious no serious risk of bias inconsistency no serious indirectness serious1 none 118 118 MD 3 lower (6.28 lower to 0.27 higher)  MODERATE CRITICAL

Ambulatory systolic blood pressure (follow-up 1.5 months; units mmHg; better indicated by lower values) 1/1 randomised trials no serious no serious risk of bias inconsistency no serious indirectness serious1 none 42 42 MD 3 lower (7.07 lower to 1.07 higher)  MODERATE2  CRITICAL

Serum creatinine (better indicated by lower values) 0/0 IMPORTANT

Total cholesterol (follow-up 1.5 months; units mmol/L; better indicated by lower values) 1/1 randomised trials no serious no serious risk of bias inconsistency no serious indirectness no serious imprecision3 none 40 40 MD 0.03 lower (0.39 lower to 0.33 higher)  HIGH2 IMPORTANT

HDL cholesterol (better indicated by higher values) 0/0 IMPORTANT

LDL cholesterol (better indicated by lower values) 0/0 IMPORTANT

Triglycerides (better indicated by lower values) 0/0 IMPORTANT

Adrenaline (plasma) (follow-up 1 - 1.5 months; units pg/mL; better indicated by lower values) 3/3 randomised trials no serious no serious risk of bias inconsistency no serious indirectness no serious imprecision3 none 76 76 MD 3.9 lower (9.2 lower to 1.3 higher)  HIGH IMPORTANT

Noradrenaline (plasma) (follow-up 1 - 1.5 months; units pg/mL; better indicated by lower values) 3/3 randomised trials no serious no serious risk of bias inconsistency no serious indirectness no serious imprecision3 none 76 76 MD 4.3 lower (23.8 lower to 15.1 higher)  HIGH IMPORTANT

1 2 3

95%CI crosses zero Only one study contributed to effect estimate Outcome is a measure of adverse effect and thus a null value is the perferred value and crossing zero is not considered a matter of imprecision

102

An n e x 7 : L i s ts o f ta b l e s a n d f i g u r e s Tables Table 3.1 Table 3.2 Table 3.3 Table 3.4 Table 3.5 Table 3.6 Table 3.7 Table 3.8 Table 3.9 Table 3.10 Table 3.11 Table 3.12 Table 3.13 Table 3.14 Table 3.15 Table 3.16 Table 3.17 Table 3.18 Table 3.19 Table 3.20 Table 3.21 Table 3.22 Table 3.23 Table 3.24 Table 3.25 Table 3.26 Barden BPARA1986........................................................................... 21 Risk of bias table Barden BPARA1986............................................... 22 Berry BPA2010 .................................................................................. 22 Risk of bias table Berry BPA2010 ...................................................... 23 Bulpitt BPA1985 ................................................................................ 23 Risk of bias table Bulpitt BPA1985 .................................................... 24 Chalmers BPA1986............................................................................ 24 Risk of bias table Chalmers BPA1986................................................ 25 Forrester BPA1988 ............................................................................ 25 Risk of bias table Forrester BPA1988 ................................................ 26 Fotherby BPA1992 ............................................................................ 26 Risk of bias table Fotherby BPA1992 ................................................ 27 Grobbee BPA1987............................................................................. 27 Risk of bias table Grobbee BPA1987................................................. 28 Gu BPA2001 ...................................................................................... 28 Risk of bias table Gu BPA2001 .......................................................... 29 He BPA2010 ...................................................................................... 29 Risk of bias table He BPA2010 .......................................................... 30 Kaplan BPA1985 ................................................................................ 30 Risk of bias table Kaplan BPA1985 .................................................... 31 Kawano BPA1998 .............................................................................. 31 Risk of bias table Kawano BPA1998 .................................................. 32 MacGregor AEBPA1982 .................................................................... 32 Risk of bias table MacGregor AEBPA1982 ........................................ 33 Matlou BPA1986 ............................................................................... 33 Risk of bias table Matlou BPA1986 ................................................... 34

103

Table 3.27 Table 3.28 Table 3.29 Table 3.30 Table 3.31 Table 3.32 Table 3.33 Table 3.34 Table 3.35 Table 3.36 Table 3.37 Table 3.38 Table 3.39 Table 3.40 Table 3.41 Table 3.42 Table 3.43 Table 3.44 Table 3.45 Table 3.46 Table 3.47 Table 3.48 Table 3.49 Table 3.50 Table 3.51 Table 3.52 Table 3.53 Table 3.54 Table 3.55 Table 3.56

Obel BPA1989 ................................................................................... 34 Risk of bias table Obel BPA1989 ....................................................... 35 Overlack BPARAAEA1991.................................................................. 35 Risk of bias table Overlack BPARAAEA1991 ..................................... 36 Patki BPARA1990 .............................................................................. 36 Risk of bias table Patki BPARA1990 .................................................. 37 Richards BPAAEA1984 ...................................................................... 37 Risk of bias table Richards BPAAEA1984 .......................................... 38 Siani BPA1987 ................................................................................... 38 Risk of bias table Siani BPA1987 ....................................................... 39 Siani BPARA1991............................................................................... 39 Risk of bias table Siani BPARA1991................................................... 40 Smith BPARA1985 ............................................................................. 40 Risk of bias table Smith BPARA1985 ................................................. 41 Trial Hyp Prv Col BPA1992 ................................................................ 41 Risk of bias table Trial Hyp Prv Col BPA1992 .................................... 42 Valdes BPA1991 ................................................................................ 42 Risk of bias table Valdes BPA1991 .................................................... 43 Whelton BPA1995 ............................................................................. 43 Risk of bias table Whelton BPA1995................................................. 44 Braschi BPA2008 ............................................................................... 45 Cushman BPA1988 ............................................................................ 45 Hilary Green BPA2000 ...................................................................... 46 Mullan BPA2010 ............................................................................... 47 Turban BPA RA2009 .......................................................................... 48 Excluded studies and reasons for exclusion ..................................... 49 Resting systolic blood pressure ........................................................ 50 Resting diastolic blood pressure ....................................................... 51 Ambulatory systolic blood pressure ................................................. 52 Ambulatory diastolic blood pressure................................................ 52

104

Table 3.57 Table 3.58 Table 3.59

Blood lipids ....................................................................................... 53 Catecholamine levels ........................................................................ 53 Renal function ................................................................................... 53

Figures Figure 3.1 Figure 3.2 Figure 3.3 Figure 3.4 Figure 3.5 Figure 3.6 Figure 3.7 Figure 3.10 Figure 3.11 Figure 3.12 Figure 3.13 Figure 3.14 Figure 3.16 Figure 3.17 Figure 3.18 Figure 3.19 Figure 3.20 Figure 3.21 Figure 3.22 Figure 3.23 Figure 3.24 Figure 3.25 Flow through screening, inclusion, exclusion .................................... 54 Resting systolic blood pressure ......................................................... 55 Resting systolic blood pressure – blood pressure status subgroups .......................................................................................... 56 Resting systolic blood pressure – achieved intake subgroups (based on urinary potassium excretion*) .......................................... 57 Resting systolic blood pressure – achieved difference subgroups (based on urinary potassium excretion*) ........................ 58 Resting diastolic blood pressure – all adults ..................................... 59 Resting diastolic blood pressure – blood pressure status subgroups .......................................................................................... 60 Resting systolic blood pressure – baseline potassium subgroups (based on urinary potassium excretion*) ........................ 63 Resting systolic blood pressure – baseline sodium subgroups ......... 64 Resting systolic blood pressure – duration subgroups ...................... 65 Resting systolic blood pressure – device subgroups ......................... 66 Resting systolic blood pressure – method subgroups ....................... 67 Resting systolic blood pressure – type of intervention subgroups .......................................................................................... 69 Resting systolic blood pressure – trial design subgroups .................. 70 Ambulatory systolic blood pressure – all adults ................................ 71 Ambulatory diastolic blood pressure – all adults .............................. 71 Ambulatory systolic blood pressure – achieved intake subgroups (based on urinary potassium excretion*) ........................ 72 Total cholesterol – all adults.............................................................. 73 HDL cholesterol – all adults ............................................................... 73 LDL cholesterol – all adults ................................................................ 73 Total triglycerides – all adults ............................................................ 73 Plasma adrenaline – all adults ........................................................... 73

105

Figure 3.26 Figure 3.27

Plasma noradrenaline – all adults..................................................... 74 Serum creatinine – all adults ............................................................. 74

106

Full list of references Agnoli GC, Borgatti R, Cacciari M et al. Interactions between the renin-angiotensin system and prostanoids in modulating renal function in potassium-depleted healthy women. Prostaglandins, Leukotrienes and Essential Fatty Acids, 1994, 50:347–352. Agnoli GC, Borgatti R, Cacciari M et al. Effects of experimental potassium depletion on renal function and urinary prostanoid excretion in normal women during moderate antidiuresis. Clinical Physiology, 1992a, 12:79–93. Agnoli GC, Borgatti R, Cacciari M et al. Urinary prostanoid excretion in healthy women with different degrees of induced potassium depletion. Prostaglandins Leukot Essent Fatty Acids, 1992b, 46(1):21–26. Ascherio A, Rimm EB, Hernán MA et al. Intake of potassium, magnesium, calcium, and fiber and risk of stroke among US men. Circulation, 1998, 98:1198–1204. Barcelo P, Wuhl O, Servitge E et al. Randomized double-blind study of potassium citrate in idiopathic hypocitraturic calcium nephrolithiasis. Journal of Urology, 1993, 150:1761–1764. Barden A, Vandongen R, Beilin LJ. Increases in urinary kallikrein activity and prostanoid synthesis after dietary potassium supplementation. Clinical and Experimental Pharmacology and Physiology, 1987, 14:565–572. Barden AE, Vandongen R, Beilin LJ et al. Potassium supplementation does not lower blood pressure in normotensive women. Journal of Hypertension, 1986, 4:339–343. Barros MA, Brito IM. Potassium in essential arterial hypertension CCS, 1984, 6:48–55. Berry SE, Mulla UZ, Chowienczyk PJ et al. Increased potassium intake from fruit and vegetables or supplements does not lower blood pressure or improve vascular function in UK men and women with early hypertension: a randomised controlled trial. British Journal of Nutrition, 2010, 104:1839–1847. Braschi A, Naismith DJ. The effect of a dietary supplement of potassium chloride or potassium citrate on blood pressure in predominantly normotensive volunteers. British Journal of Nutrition, 2008, 99:1284–1292. Bulpitt CJ, Ferrier G, Lewis PJ et al. Potassium supplementation fails to lower blood pressure in hypertensive patients receiving a potassium losing diuretic. Annals of Clinical Research, 1985, 17:126–130. Cappuccio FP, MacGregor GA. Does potassium supplementation lower blood pressure? A meta-analysis of published trials. Journal of Hypertension, 1991, 9(5):465–473. Ceglia L, Harris SS, Abrams SA et al. Potassium bicarbonate attenuates the urinary nitrogen excretion that accompanies an increase in dietary protein and may promote calcium absorption. Journal of Clinical Endocrinology & Metabolism, 2009, 94:645–653.

107

Chalmers J, Morgan T, Doyle A et al. Australian National Health and Medical Research Council dietary salt study in mild hypertension. Journal of Hypertension – Supplement, 1986, 4:S629–S637. Cushman WC, Langford HG. Randomized controlled trial of potassium chloride versus placebo in mildly hypertensive blacks and whites. Circulation, 1988, 17:S370–S370. CWP MR. Comparison of the antihypertensive efficacy and adverse reactions to two doses of bendrofluazide and hydro-chlorothiazide and the effect of potassium supplementation on the hypotensive action of bendrofluazide: substudies of the Medical Research Council's Trials of Treatment of Mild Hypertension. Journal of Clinical Pharmacology, 1987, 27:271–277. Dickinson HO, Nicolson DJ, Campbell F et al. Potassium supplementation for the management of primary hypertension in adults. Cochrane Database of Systematic Reviews, 2006, 3:CD004641. Forrester TE, Grell GA. Changes in red cell sodium content and blood pressure levels with potassium supplementation in black hypertensive patients. West Indian Medical Journal, 1988, 37:92–96. Fotherby MD, Potter JF. Potassium supplementation reduces clinic and ambulatory blood pressure in elderly hypertensive patients. Journal of Hypertension, 1992, 10:1403– 1408. Fotherby MD, Potter JF. Long-term potassium supplementation lowers blood pressure in elderly hypertensive subjects. International journal of clinical practice, 1997, 51:219–222. Fujita T, Ando K. Hemodynamic and endocrine changes associated with potassium supplementation in sodium-loaded hypertensives. Hypertension, 1984, 6:184–192. Gamarra G, Balaguera H, Corzo D et al. Tratamiento de la hipertensión arterial esencial con dieta rica en potasio. Acta Medica Colombiana, 1994, 19(1):15–23. Geleijnse JM, Kok FJ, Grobbee DE. Blood pressure response to changes in sodium and potassium intake: a metaregression analysis of randomised trials. Journal of Human Hypertension, 2003, 17(7):471–480. Gomez-Marin O, Prineas RJ, Sinaiko AR. The sodium-potassium blood pressure trial in children. Design, recruitment, and randomization: the children and adolescent blood pressure program. Control Clin Trials, 1991, 12(3):408–423. Grimm RH. Potassium and sodium to control blood pressure in hypertensives ClinicalTrials.gov, 1999. Grimm RH, Kofron PM, Neaton JD et al. Effect of potassium supplementation combined with dietary sodium reduction on blood pressure in men taking antihypertensive medication. Journal of Hypertension, 1988, 6:S591–S593. Grimm RH, Neaton JD, Elmer PJ et al. The inﬂuence of oral potassium chloride on blood pressure in hypertensive men on a low-sodium diet. New England Journal of Medicine, 1990, 322:569–574.

108

Grobbee DE, Hofman A, Roelandt JT et al. Sodium restriction and potassium supplementation in young people with mildly elevated blood pressure. Journal of Hypertension, 1987, 5:115–119. Gu D, He J, Wu X et al. Effect of potassium supplementation on blood pressure in Chinese: a randomized, placebo-controlled trial. Journal of Hypertension, 2001, 19:1325–1331. Guyatt GH, Oxman AD, Vist GE et al. GRADE: an emerging consensus on rating quality of evidence and strength of recommendations. BMJ, 2008, 336(7650):924-926. He FJ, Marciniak M, Carney C et al. Effects of potassium chloride and potassium bicarbonate on endothelial function, cardiovascular risk factors, and bone turnover in mild hypertensives. Hypertension, 2010, 55:681–688. Heller HJ, Reza-Albarrán AA, Breslau NA et al. Sustained reduction in urinary calcium during long-term treatment with slow release neutral potassium phosphate in absorptive hypercalciuria. Journal of Urology, 1998, 159:1451–1455. Higgins J, Thompson S, Deeks J et al. Statistical heterogeneity in systematic reviews of clinical trials: a critical appraisal of guidelines and practice. Journal of Health Services Research & Policy, 2002, 7(1):51–61. Higgins JP, Thompson SG. Quantifying heterogeneity in a meta-analysis. Statistics in Medicine, 2002, 21(11):1539–1558. Higgins JP, Thompson SG, Spiegelhalter DJ. A re-evaluation of random-effects meta-analysis. Journal of The Royal Statistical Society Series A-Statistics in Society, 2009, 172(1):137–159. Hilary Green J, Richards JK, Bunning RL. Blood pressure responses to high-calcium skim milk and potassium-enriched high-calcium skim milk. Journal of Hypertension, 2000, 18:1331–1339. Iimura O, Kijima T, Kikuchi K et al. Studies on the hypotensive effect of high potassium intake in patients with essential hypertension. Clinical Science, 1981, 61:77s–80s. Institute of Medicine. Dietary reference intakes: water, potassium, sodium, chloride, and sulfate. Washington, DC, National Academies Press, 2005 Jardim PC, Branco RF, Silva EG. Evaluation of potassium supplementation in patients under furosemide therapy. Arquivos Brasileiros de Cardiologia, 1988, 51:153–156. Kaplan NM, Carnegie A, Raskin P et al. Potassium supplementation in hypertensive patients with diuretic-induced hypokalemia. New England Journal of Medicine, 1985, 312:746–749. Kawano Y, Minami J, Takishita S. Effects of potassium, calcium and magnesium supplementation in patients with essential hypertension – assessment by blood pressure monitoring. Therapeutic Research, 1997, 18(7):44–47. Kawano Y, Minami J, Takishita S et al. Effects of potassium supplementation on ofﬁce, home and 24-hour blood pressure in patients with essential hypertension. American Journal of Hypertension, 1998, 11:1141–1146.

109

Khaw KT, Barrett-Connor E. Dietary potassium and stroke-associated mortality. A 12-year prospective population study. New England Journal of Medicine, 1987, 316:235–240. Khaw KT, Thom S. Randomized double-blind cross-over trial of potassium on blood pressure in normal subjects. Lancet, 1982, 2:1127–1129. Krishnan AV. AUSSPRINT: Australian study of the effects of strict potassium restriction on neuropathy in chronic kidney disease (ACTRN12610000538044). 2010 Langford HG, Davis BR, Blaufox D et al. Effect of drug and diet treatment of mild hypertension on diastolic blood pressure. Hypertension, 1991, 17:210–217. Lau J, Ioannidis JP, Terrin N et al. The case of the misleading funnel plot. BMJ, 2006, 333(7568):597–600. Lennon EJ, Lemann J Jr. The effect of a potassium-deficient diet on the pattern of recovery from experimental metabolic acidosis. Clinical Science, 1968, 34:365–378. Liberati A, Altman DG, Tetzlaff J et al. The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate healthcare interventions: explanation and elaboration. BMJ, 2009, 339:b2700. MacGregor GA, Smith SJ, Markandu ND et al. Moderate potassium supplementation in essential hypertension. Lancet, 1982, 11:567–570. MacGregor GA, Smith SJ, Markandu ND et al. Moderate potassium supplementation in essential hypertension. Archives des maladies du coeur et des vaisseaux, 1984, 77:Spec No. 93–Spec No.100. Mackay J, Mensah G. Atlas of heart disease and stroke. Geneva, World Health Organization, 2004 Matlou SM, Isles CG, Higgs A et al. Potassium supplementation in blacks with mild to moderate essential hypertension. Journal of Hypertension, 1986, 4:61–64. Morris R, O'Connor M, Forman A et al. Supplemental dietary potassium with KHCO3 but not KCl attenuates essential hypertension [abstract]. Journal of the American Society of Nephrology, 1995, 6:645–645. Obel AO. Placebo-controlled trial of potassium supplements in black patients with mild essential hypertension. Journal of Cardiovascular Pharmacology, 1989, 42:294–296. Overlack A, Conrad H, Stumpe KO. The inﬂuence of oral potassium citrate/ bicarbonate on blood pressure in essential hypertension during unrestricted salt intake. Klinische Wochenschrift, 1991, 69:S79–S83. Overlack A, Maus B, Ruppert M et al. Potassium citrate vs potassium chloride in essential hypertension: Effect on haemodynamic, hormonal and metabolic parameters. Deutsche Medizinische Wochenschrift, 1995, 120:631–635. Parfrey PS, Condon K, Wright P et al. Blood pressure and hormonal changes following alterations in dietary sodium and potassium in young men with and without a familial predisposition to hypertension. Lancet, 1981, 1:113–117.

110

Patki PS, Singh J, Gokhale SV et al. Efficacy of potassium and magnesium in essential hypertension: a double blind, placebo controlled, crossover study. BMJ, 1990, 301:521–523. Poulter NR, Sever PS. Moderate potassium supplementation: ineffective in black normotensives. East African Medical Journal, 1986, 63:798–802. Rahimi AR, Mahmoodpoor A, Sanaie S. The effect of high-calcium and high-potassium diet on grade-I hypertension and high normal blood pressure. Pakistan Journal of Medical Sciences, 2007, 23:589–592. Richards AM, Nicholls MG, Espiner EA et al. Blood pressure response to moderate sodium restriction and to potassium supplementation in mild essential hypertension. Lancet, 1984, 1:757–761. Sanclemente P. Essential arterial hypertension and dietary potassium supplement CES Medicina, 1987, 1:65–70. Siani A, Strazzullo P, Giacco A et al. Increasing the dietary potassium intake reduces the need for antihypertensive medication. Annals of Internal Medicine, 1991, 115:753–759. Siani A, Strazzullo P, Russo L et al. Controlled trial of long term oral potassium supplements in patients with mild hypertension. BMJ (Clinical Research Edition), 1987, 294:1453– 1456. Smith SJ, Markandu ND, Sagnella GA et al. Moderate potassium chloride supplementation in essential hypertension: is it additive to moderate sodium restriction? BMJ (Clinical Research Edition), 1985, 290:110–113. Smith SR, Klotman PE, Svetkey LP. Potassium chloride lowers blood pressure and causes natriuresis in older patients with hypertension. Journal of the American Society of Nephrology, 1992, 2:1302–1309. Stamler J, Elliott P, Appel L et al. Higher blood pressure in middle-aged American adults with less education-role of multiple dietary factors: the INTERMAP study. Journal of Human Hypertension, 2003, 17(9):655–775. Sterne JA, Egger M. Funnel plots for detecting bias in meta-analysis: guidelines on choice of axis. Journal of Clinical Epidemiology, 2001, 54(10):1046–1055. Strong K, Mathers C, Leeder S et al. Preventing chronic diseases: how many lives can we save? Lancet, 2005, 366(9496):1578–1582. Svetkey LP, Yarger WE, Feussner JR et al. Double-blind placebo-controlled trial of potassium chloride in the treatment of mild hypertension. Hypertension, 1987, 9:444–450. Trial Hyp Prv Col. The effects of nonpharmacologic interventions on blood pressure of persons with high normal levels. Results of the Trials of Hypertension Prevention, Phase I. Journal of the American Medical Association, 1992, 267:1213–1220. Valdés G, Vio CP, Montero J et al. Potassium supplementation lowers blood pressure and increases urinary kallikrein in essential hypertensives. Journal of Human Hypertension, 1991, 5:91–96.

111

Whelton PK, He J, Cutler JA et al. Effects of oral potassium on blood pressure. Meta-analysis of randomized controlled clinical trials. Journal of the American Medical Association, 1997, 277(20):1624–1632. Whelton PK, Kumanyika SK, Cook NR et al. The effect of potassium supplementation in persons with a high-normal blood pressure: results from phase I of the Trials of Hypertension Prevention. The American Journal of Clinical Nutrition, 1995, 65:S652– S660. WHO. Diet, nutrition and the prevention of chronic disease. Report of a Joint WHO/FAO Expert Consultation. Geneva, World Health Organization (WHO), 2003 WHO. Preventing chronic disease: A vital investment. Geneva, World Health Organization (WHO), 2005 WHOICTRP. The effects of potassium on physiological and cardiovascular variables in healthy persons (NCT00801034). Clinicaltrials.gov, World Health Organization International Clinical Trials Registry Platform (WHOICTRP), 2008

112

For further information please contact: Nutrition Policy and Scientific Advice Unit (NPU) Department of Nutrition for Health and Development (NHD) World Health Organization (WHO) 20, Avenue Appia CH–1211 Geneva 27 Switzerland Fax: +41.22.791.4156 E-mail: NPUinfo@who.int NHD website: http://www.who.int//nutrition

ISBN 978 92 4 150488 1

Key facts

Organisation World Health Organization (WHO)

Document type Publications

Adoption date December 31, 2012

Source World Health Organization