ABSTRACT
Background
Over the past three decades, the prevalence of hypertension in adults has doubled worldwide, surging from 650 million to 1.3 billion cases between 1990 and 2019. Sodium reduction is a cornerstone of non-pharmacological strategies for managing hypertension. However, recent guidelines increasingly emphasize the importance of boosting potassium intake, supported by robust evidence of its cardiovascular benefits. Despite this, the precise dose-dependent effects of potassium on blood pressure (BP) remain inadequately defined.
Methods
We conducted a systematic review of randomized controlled trials (RCTs) published between 2000 and 2024 to evaluate the impact of potassium supplementation alone—assessed solely via 24-h urinary potassium excretion—on BP. A dose–response meta-analysis was performed using linear, quadratic, and one-stage cubic spline regression models. Subgroup analyses were carried out based on subjects with or without hypertension.
Results
Our meta-analysis included 10 RCTs, comprising 4 studies on subjects without hypertension and 6 studies on subjects with hypertension. The dose–response relationship varied according to BP status. In subjects without hypertension, potassium supplementation had a modest negative linear effect on BP. In contrast, subjects with hypertension exhibited a markedly higher reduction in BP. Specifically, a 50 mmol/day increase in urinary potassium excretion was associated with a 0.5 mmHg reduction in systolic BP (SBP) and a 0.12 mmHg reduction in diastolic BP (DBP) in subjects without hypertension, and a 5.3 mmHg reduction in SBP and a 3.62 mmHg reduction in DBP in subjects with hypertension.
Conclusion
This meta-analysis highlights the dose–response relationship between potassium supplementation and BP reduction, particularly in subjects with hypertension. While the findings offer valuable insights for refining dietary guidelines, caution is warranted due to the limited number of RCTs included in the analysis.
Keywords: blood pressure, dose–response, meta-analysis, potassium
Graphical Abstract
Graphical Abstract.
KEY LEARNING POINTS.
What was known:
Hypertension is the main risk factor for cardiovascular disease, which remains the world's leading cause of death and disability, with around 18.6 million deaths and 393 million years lived with disability each year.
Reducing sodium intake is a fundamental priority for the non-pharmacological management of hypertension; in addition, recent guidelines increasingly emphasize the importance of increasing potassium intake, supported by robust evidence of its cardiovascular benefits.
Despite this, the precise dose-dependent effect of potassium supplementation on blood pressure (BP) remains inadequately defined.
This study adds:
This systematic review of randomized controlled trials analyzed only recent publications (from 2000 to 2024 to reflect current BP measurements) to assess the impact of potassium supplementation alone (assessed solely by 24-h urinary potassium excretion) on BP.
The dose–response relationship between changes in potassium intake and changes in BP was rigorously analyzed using three distinct statistical models, ensuring that the results were free from preconceived assumptions.
The dose–response meta-analysis revealed a modest linear association between increased potassium intake and reduced BP in subjects without hypertension; however, in subjects with hypertension, the reduction in BP was more pronounced, underlining the greater potential benefit of potassium intake in this subpopulation.
Potential impact:
This dose–response meta-analysis reinforces the importance of current nutritional recommendations to increase dietary potassium intake as a key measure for reducing BP, particularly in individuals with hypertension.
Boosting potassium intake can play a critical role in improving BP control and lowering cardiovascular risk, especially in high-risk hypertensive populations.
INTRODUCTION
Cardiovascular diseases (CVD) remain the leading cause of mortality and disability worldwide, with an estimated 18.6 million deaths and 393 million years of disability each year [1]. Hypertension is the primary risk factor for CVD [2, 3]. It is estimated that reducing the population's blood pressure (BP) distribution by 5 mmHg could prevent one-third of strokes and one-fifth of coronary events [4]. Non-pharmacological interventions, such as dietary and lifestyle modifications, are fundamental to hypertension management. Indeed, a close link has been established between nutritional factors and BP, with the identification of various factors likely to reduce BP, including weight reduction, moderation of alcohol consumption, reduction of sodium intake and increase in potassium intake [5–9]. These non-pharmacological measures to lower BP are widely recognized in all national and international dietary recommendations and guidelines [5, 10–15] although they can be challenging to implement in clinical practice. Reducing sodium intake has long been a cornerstone of hypertension management but recent guidelines have emphasized the need to increase potassium intake based on recent trials [16, 17]. The World Health Organization [5], the American Heart Association [18] and the International and European hypertension guidelines [13–15] recommend a dietary potassium intake exceeding 3.5 g/day (∼90 mmol/day). The benefits of a potassium-enriched diet in reducing cardiovascular risk have been the topic of numerous publications and it is now almost established that a higher intake of potassium is associated with a reduction of BP and a lower risk of stroke [19–30]. Non-pharmacological interventions to reduce BP are therefore a top priority for both patients and healthcare providers [30]. They not only improve quality of life by reducing the number of medications needed to achieve BP control [31, 32], they also promote overall health through balanced dietary habits. The beneficial effects of nutrition on cardiovascular risk also enable patients to gain in quantity of life [33]. Additionally, these measures have the potential to reduce healthcare costs associated with antihypertensive medications, hospitalizations, and management of CVD-related complications. Previous dose–response meta-analyses by Filippini et al. (2017 [34] and 2020 [35]) examined the relationship between potassium intake and BP, reporting a significant correlation between the two parameters [35]. Like many other earlier meta-analyses [36–38], their reviews included studies conducted prior the year 2000, which may not fully reflect current BP measurements, and therapeutic practices or dietary recommendations [14, 39–41]. Given the significant evolution of hypertension guidelines over the past two decades—particularly of the increased emphasis on potassium intake—we hypothesized that a meta-analysis based on more recent data and limited to randomized clinical trials (RCTs) devoted to the effects of BP related to changes in potassium intake alone (since modifications in sodium and potassium intake have interrelated effects) would provide more recent and relevant insights, in line with current practice.
The aim of this study was to perform a dose–response meta-analysis of changes in potassium intake alone (assessed by 24-h urinary potassium excretion) with changes in BP, using recent high-quality RCTs, focusing exclusively on post-2000 data. To avoid any preconceived assumptions about the dose–response relationship, three models (linear, quadratic and cubic spline) were examined to analyze this relationship in subjects with or without hypertension. In particular, the non-linear models will enable us to capture the variable effects of a change in potassium intake in different subpopulations, in order to define the dose–response relationship more precisely. Consistent with current clinical and dietary practices, this approach aims to refine and validate potassium intake recommendations for hypertension management.
MATERIALS AND METHODS
Protocol and registration
The protocol was registered a priori on Prospero: CRD42023440909. The methodology complied with the Cochrane Handbook for Systematic Reviews of interventions [42] and the study was reported in accordance with the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) checklist [43] (see Supplementary data, S1).
Search strategy and study selection
A systematic literature search was conducted in PubMed, Cochrane Central and Embase up to September 2024. Our PICOT criteria were as follows: (P) humans, (I) modification of potassium intake, (C) no or other modification of potassium intake, (O) blood pressure, (T) systematic reviews and meta-analysis. The search strategy was a combination of different keywords for each criterion (for details, see Supplementary data, S2). The search was restricted to articles published in English and French. The research focused on systematic reviews and meta-analyses, as numerous meta-analyses had already been conducted on this topic. Additionally, a manual search of references was carried out.
Eligibility was assessed independently by two independent reviewers (M.G. and V.S.). The records were screen by title and abstract, and subsequently in full-text based on PICOT (M.G. and V.S.). Any disagreements were resolved by a third reviewer (A.G.). Only studies estimating potassium intake via 24-h urinary excretion and published after 1 January 2000, were included to ensure methodological consistency with contemporary BP measurement and hypertension management. Studies involving sodium-only interventions, sodium and potassium associated interventions (salt substitute, diet intervention) or pediatric populations were excluded. Authors of studies with incomplete data were contacted, and studies were excluded if no response was received.
To avoid overlap in data extraction, a correspondence table was created to compare the original articles included in the meta-analyses. Clinical studies included in several systematic reviews were considered only once.
Data extraction
Three independent reviewers (M.G., V.S., J.P.F.) extracted the data using a standardized form. When data for a clinical study were missing in the included systematic reviews, we extracted them from the RCT study. Extracted data included: study characteristics (study ID, author, date of publication, country, study design, washout period, number of participants in the intervention and control groups, duration of intervention), population characteristics [age, proportion of female participants, pathology population, chronic kidney disease population, estimated glomerular filtration rate, presence or absence of hypertension, body mass index (BMI)], outcome data [BP measurement method, baseline and follow-up systolic BP (SBP), baseline and follow-up diastolic BP (DBP), and measured 24-h urinary potassium excretion].
For studies not reporting the standard deviation of means for each group, these were converted from standard errors, median interquartile range, or 95% confidence interval (95% CI). For studies that used a crossover design, the effect of the intervention was taken as the difference between endpoint measurements in the two groups. For studies that used a parallel design, we did not adjust for baseline differences in SBP or DBP. It was assumed that randomization would balance baseline characteristics between groups, and any differences would be minimal and unlikely to significantly affect the measured intervention effects. The duration of the intervention was defined as the period from the start of the intervention to the point at which the outcome was measured. Ambulatory BP measurements were prioritized over clinic readings, and if 24-h BP was not available, daytime BP was prioritized over nighttime BP.
Statistical analysis
A random-effects dose–response meta-analysis was performed, with potassium dose expressed as the standardized mean difference (SMD) in urinary potassium excretion between low and high-intake groups, with the low potassium group referenced at 0. The response was defined as the mean difference in BP between the groups at the end of the intervention.
Three regression models (linear, quadratic and cubic spline) were used to assess the dose–response relationship. The cubic spline model included knots at the 15th, 50th and 85th percentiles of potassium intake distribution. Model selection was based on the Akaike Information Criterion (AIC), and the lowest AIC was considered the best fit for the data. A table of AIC values and dose–response curves for the different models by population is available (see Supplementary data, S3 and S4). The selected model was used to predict the dose–response relationship for potassium intake values ranging from 10 to 100 mmol/day.
Given the limited number of studies included in the present meta-analysis, sensitivity analyses were not performed based on study design or intervention duration, as originally planned. Only a subgroup analysis in subjects with (composed of pre-hypertensive and hypertensive subjects) or without hypertension was performed.
Publication bias was assessed using funnel plot and Egger's test [44], independently reviewed by two independent reviewers (M.G. and A.G.).
Analyses were conducted using the R 4.2.2 software [45] and the dosresmeta and metafor [46] packages.
Risk of bias assessment
The Cochrane “Risk of Bias” 2.0 tool [47, 48] was employed to evaluate the risk of bias of the included studies by MG across five domains: randomization process, deviation from planned intervention, missing outcome data, outcome measurement and outcome reporting. Each domain was classified as “low risk of bias,” “concern” or “high risk” based on responses to the guideline questions.
RESULTS
Study selection and characteristics
Out of the 1826 identified references, 3 systematic reviews were included, encompassing a total of 10 RCTs (Fig. 1). These RCTs were published between 2005 and 2021. The majority of them focused on subjects with hypertension (60%).
Figure 1:
Flow diagram PRISMA of included RCTs in the meta-analysis.
Table 1 shows the main characteristics of the 10 RCTs included in the dose–response meta-analysis. The studies involved a total of 684 patients, 342 in the intervention group (potassium intake modification) and 342 in the control group (unchanged potassium intake). RCTs were conducted in adults in the UK (N = 4), USA (N = 1), Italy (N = 1), China (N = 1), Netherlands (N = 1) and Denmark (N = 2). Among the 10 RCTs, one was conducted in parallel, the others in crossover. Intervention durations ranged from 1 to 6 weeks. BP measurements were conducted in clinical settings with office BP in two studies, whereas the other eight employed ambulatory BP monitoring. Participant characteristics varied, with mean ages ranging from 26 to 66 years and mean BMIs from 22 to 31 kg/m². Most study groups were mixed, although two studies focused exclusively on male participants. Four of the study groups included only subjects without hypertension, whereas six groups comprised individuals with hypertension. Subgroup analyses were conducted to assess outcomes based on subjects with or without hypertension.
Table 1:
Characteristics of the 10 RCTs included.
| Reference | Participants (LP) | Participants (HP) | Age (years) | Female (%) | Design | Intervention | Intervention duration (weeks) | HTN status | BP measure method |
|---|---|---|---|---|---|---|---|---|---|
| Berry [49], 2010, UK | 48 | 48 | 45 | 52 | X | Supplement | 6 | W/hypertension | Ambulatory |
| Dreier [50], 2021, Denmark | 25 | 25 | 25.7 | 0 | X | Supplement | 4 | W/O hypertension | Ambulatory |
| Franzoni [51], 2005, Italy | 52 | 52 | 52 | 38 | P | Supplement | 4 | W/hypertension | Ambulatory |
| Gijsbers [52], 2015, Netherlands | 36 | 36 | 65.8 | 33 | X | Supplement | 4 | W/O hypertension | Ambulatory |
| Graham [53], 2014, UK | 40 | 40 | 54.8 | 20 | X | Supplement | 6 | W/hypertension | Clinic |
| He [54], 2010, UK | 42 | 42 | 51 | 29 | X | Supplement | 4 | W/hypertension | Ambulatory |
| Liu [55], 2017, China | 18 | 18 | 50.7 | 0 | X | Supplement | 1 | W/O hypertension | Ambulatory |
| Matthesen [56], 2012, Denmark | 21 | 21 | 26 | 57 | X | Supplement | 4 | W/O hypertension | Ambulatory |
| Stone [57], 2021, UK | 30 | 30 | 46.2 | 50 | X | Supplement | 2 | W/hypertension | Clinic |
| Vongpatanasin [58], 2016, USA | 30 | 30 | 54 | 53 | X | Supplement | 4 | W/O hypertension | Ambulatory |
LP, low potassium intake; HP, high potassium intake; X, Cross-over; P, Parallel; W/, with; W/O, without.
Description of potassium intake modification
The interventions were designed to increase potassium intake through supplementation with potassium chloride, potassium aspartate or potassium bicarbonate or increase consumption of potassium-rich fruits and vegetables. Dietary interventions occasionally included specific dietary instructions. Patients were generally asked to maintain their usual dietary habits, except for the potassium supplementation, and to maintain their usual physical activities. In the high-potassium groups, supplements were administered as potassium tablets or slow-released potassium tablets. For crossover trials, participants were randomly assigned to groups, with some trials including a washout period between intervention phases. A run-in period was frequently used in both groups before administering potassium supplements versus placebo, to assess participants’ ability to adhere to the different dietary interventions.
In all included RCTs, 24-h urinary potassium excretion and BP were evaluated by a third party. BP measurements were performed at the beginning and end of the intervention for parallel studies, and at the end of each interventional phase in crossover studies.
Dose–response meta-analysis
SMDs in potassium excretion ranged from 12.5 to 92 mmol/day, with changes in urinary potassium excretion ranging between 30 to 92 mmol/day in subjects without hypertension and 12.5 to 45 mmol/day in subjects with hypertension.
In subjects without hypertension, a slight negative linear relationship (P = .48) was observed between increased potassium intake and changes in SBP. On the other hand, a more pronounced negative linear relationship (P = .02) was found in subjects with hypertension. For instance, for a 50 mmol/day increase in 24-h urinary potassium excretion the hypotensive effect was estimated at –5.3 mmHg in SBP in subjects with hypertension compared with only –0.5 mmHg in subjects without hypertension (see Table 2 below). Increases in potassium intake had a less pronounced effect on changes in DBP compared with SBP. For a 50 mmol/day increase in 24-h urinary potassium excretion, DBP decreased by –3.6 mmHg in subjects with hypertension, while in subjects without hypertension the decrease was only –0.1 mmHg (Fig. 2).
Table 2:
Dose–response prediction of mean changes in SBP and DBP [mmHg (95% CI)] as a function of change in potassium 24-h urinary excretion (mmol/day).
| Change in 24-h K urinary excretion (mmol/day) | Change in SBP (mmHg) subjects without hypertension (n = 4 studies) | Change in SBP (mmHg) subjects with hypertension (n = 6 studies) | Change in DBP (mmHg) subjects without hypertension (n = 4 studies) | Change in DBP (mmHg) subjects with hypertension (n = 6 studies) |
|---|---|---|---|---|
| 10 | –1.06 (–1.97; –0.15) | –0.72 (–1.84; 0.39) | ||
| 20 | –2.12 (–3.94; –0.31) | –1.45 (–3.68; 0.78) | ||
| 30 | –0.32 (–1.22; 0.58) | –3.18 (–5.91; –0.46) | –0.07 (–0.88; 0.73) | –2.17 (–5.51; 1.17) |
| 40 | –0.43 (–1.63; 0.77) | –4.25 (–7.88; –0.62) | –0.10 (–1.18; 0.98) | –2.90 (–7.35; 1.56) |
| 50 | –0.54 (–2.03; 0.96) | –5.31 (–9.85; –0.77) | –0.12 (–1.47;1.22) | –3.62 (–9.19; 1.95) |
| 60 | –0.64 (–2.44; 1.15) | –0.15 (–1.77; 1.47) | ||
| 70 | –0.75 (–2.84; 1.35) | –0.18 (–2.06; 1.71) | ||
| 80 | –0.85 (–3.25; 1.54) | –0.20 (–2.36; 1.96) | ||
| 90 | –0.96 (–3.66; 1.73) | –0.23 (–2.65; 2.20) | ||
| 100 | –1.07 (–4.06; 1.92) | –0.25 (–2.95; 2.45) |
Linear models for subjects without hypertension and subjects with hypertension dose-effect prediction.
Figure 2:
Dose–response meta-analysis of changes in blood pressure (in mmHg) as a function of differences in 24-h urinary potassium excretion (in mmol/day) between the treatment group (potassium-supplemented group) and the control group with 95% confidence limits. (a) Change in SBP in subjects without hypertension. (b) Change in DBP in subjects without hypertension. (c) Change in subjects with hypertension. (d) Change in DBP in subjects with hypertension.
Dose–response prediction
Predictions of changes in mean SBP and DBP as a function of changes in potassium intake of 10, 20, 30, 40, 50, 60, 70, 80, 90 and 100 mmol/day are summarized in Table 2 based on the linear model for subjects without hypertension and subjects with hypertension. To minimize data extrapolation, SBP and DBP changes are presented within the range of 30–100 mmol/day change in potassium intake in subjects without hypertension and in the range of 10–50 mmol/day in subjects with hypertension.
Risk of bias assessment
The risk of bias analysis was conducted across five domains: randomization process, deviation from the planned intervention, missing outcome data, outcome measurement and outcome reporting. It classified three studies as having a low overall risk of bias, while seven studies were categorized as raising concerns regarding the overall risk of bias. Four studies raised concerns about the blinding of participants or administrators. All studies were considered as low risk in terms of deviations from the planned interventions and missing outcome data. In contrast, one study was deemed to be of concern in terms of “outcome measurement,” as outcome assessors were likely to have been aware of the intervention received by study participants. Despite this, the likelihood of this knowledge influencing the results was deemed minimal. Half of the included studies raised concerns regarding the “selection of the reported outcome,” as it was not specified whether the data that produced this outcome were eligible for multiple analyses (Fig. 3). Detailed information regarding the risk of bias for each included study is provided in the Supplementary data.
Figure 3:
Summary of the risk of bias assessment.
Publication bias assessment
Visual inspection of the funnel for the changes of SBP was not suggestive of a publication bias (Fig. 4). The P-value of Egger's test was P = .44.
Figure 4:
Funnel plot for publication bias for SMD for changes SBP levels (as mmHg) and its standard error (SE).
Visual inspection of the funnel plot showed a slight asymmetry to the right, in the non-significant area with a study heterogenous in the left (Fig. 5). We considered this asymmetry was not suggestive of a publication bias. The P-value of Egger's test was P = .64. The slight asymmetry of funnel plot may be partly due to a study design that was carried out in parallel, increasing inter-individual variability.
Figure 5:
Funnel plot for publication bias for SMD for changes DBP levels (as mmHg) and its standard error (SE).
DISCUSSION
We conducted this meta-analysis to evaluate the dose-dependent effect of changes in potassium intake on BP focusing exclusively on RCTs published post-2000 and reporting 24-h urinary potassium excretion to assess changes in potassium intake. This decision ensures consistency with contemporary dietary and therapeutic management of hypertension. This approach provides updated information to validate or refine current recommendations for potassium intake in the management of hypertension. Our analyses support the negative relationship between increased potassium intake and decreased BP. However, the relationship was weak in subjects without hypertension and more pronounced in subjects with hypertension, aligning with findings from the previously published meta-analysis by Filippini et al. [35]. The weak negative relationship between increased urinary potassium excretion and SBP or DBP was marginal and probably due to a “study effect” in subjects without hypertension. In contrast, the negative relationship between increased urinary potassium excretion and SBP or DBP was more pronounced in subjects with hypertension. The dose–response relationship between changes in potassium intake and changes in BP (SBP and DBP) was best described using linear models in both groups, likely due to the limited number of studies included (subjects without hypertension: n = 4, subjects with hypertension: n = 6), which restricted the ability to capture potential non-linear variations in effect. The subgroup analysis aimed to differentiate the effects of potassium intake based on subjects with or without hypertension, thus addressing the relevance of increasing potassium intake across populations or targeting those at higher cardiovascular risk, such as subjects with hypertension. The results confirm that subjects with hypertension get the greatest benefits of an increase in potassium intake [34, 35]. In addition, subjects with hypertension are likely to benefit more from additional potassium intake than subjects with hypertension because they have a higher cardiovascular risk. Furthermore, dietary modifications to increase potassium intake (increased consumption of fruits and vegetables) may offer additional benefits with fiber and vitamin supplementation for both subjects with or without hypertension [26].
While the impact of increased potassium intake on BP reduction has already been the focus of several published meta-analyses [34, 35, 37, 38, 59], the dose–response relationship has rarely been studied [35]. The strength of our study is that it provides a more contemporary and precise view of the dose–response relationship, by including in the meta-analysis only RCTs published in the last 25 years (after 2000). This approach ensures relevance to current therapeutic and dietary management while minimizing variability associated with outdated BP measurement techniques or obsolete devices. We also restricted our inclusion criteria to studies where the intervention specifically targeted potassium intake and quantified it through 24-h urinary potassium excretion, which is currently considered the “gold standard” for assessing potassium intake, despite its limitations. This avoided underestimation of potassium intake observed with other methodological approaches. Medications prescribed to subjects with hypertension, such as diuretics, could interfere with urinary potassium excretion. However, in chronically treated patients, we have reported that these medications do not influence 24-h kaliuresis [60]. Kaliuresis is generally considered to reflect only about 70% of dietary potassium intake. Since the majority of the studies (9 out of 10 analyzed) were crossover studies, with each subject being is own control, variations in fecal potassium excretion are minimized. Therefore, intra-individual variation in kaliuresis can be considered a satisfactory assessment of dietary changes in potassium in these well-conducted studies. More subjective methods of assessment based on patient recall and self-report, such as the Food Frequency Questionnaire and 24-h recalls [61], or urinary spots are currently identified as unrepresentative [62, 63]. Moreover, the focus on recent studies guarantees current relevance although it may have led to the exclusion of some potentially useful older studies. This choice was deemed necessary to balance the need for high-quality data with practical constraints associated with study selection.
Contrary to our expectations, no other subgroup analyses (by dietary intervention, duration of intervention, study design or BP measurement method—ambulatory or clinical) [64] could be performed, due to the limited number of studies selected for this meta-analysis. Furthermore, we identified a significant gap in the literature: to our knowledge, no meta-analyses meeting our criteria have evaluated the impact of changes in potassium intake on BP in patients with chronic kidney disease, a population at high cardiovascular risk. It is therefore essential to build RCTs measuring the impact of potassium intake (preferably by an increase in fruits and vegetables diet rather than potassium tablet supplementation) on BP in chronic kidney disease patients, in order to better understand its effects. The lack of correlation between kaliemia and kaliuresis [60, 65, 66] and the expected benefit on BP and cardiovascular risk reduction supports the need for such trials in chronic kidney disease patients.
The dose–response analysis makes it possible to consider variations in the effect of modifying potassium intake, helping to refine the understanding and the assessment of the dose–response relationship. However, the limited number of studies (n = 10) is certainly the greatest challenge to interpreting the models. The dose–response meta-analysis has nevertheless provided valuable information on the effect of potassium intake on BP variation, which remains the ignored ion in the sodium–potassium–BP relationship [30]. Recent meta-analyses have highlighted the benefits of salt substitutes—formulated with reduced sodium chloride and enriched potassium chloride—in lowering BP and reducing cardiovascular risk [16, 17]. Unfortunately, these studies could not be included in our meta-analysis due to our strict inclusion criteria, which required a dietary intervention based solely on potassium modification. A simultaneous dietary intervention on sodium and potassium cannot dissociate the beneficial effect of increasing potassium intake and reducing sodium intake on BP. Nevertheless, this may be an easier approach to apply in the future, as tighter regulation of sodium and potassium intakes should have a greater impact on BP control and, consequently, cardiovascular morbidity and mortality. This dual regulation aligns with two ultimate objectives: improving both the quality and quantity of life for patients, and reducing healthcare costs associated with cardiovascular events and/or diseases, and the use of antihypertensive drugs [31].
The meta-analysis conducted herein has several limitations. The somewhat arbitrary selection of studies post-2000 may have overlooked some important research. However, this selection ensures the homogeneity of BP measurement techniques, as 9 out of 10 studies report 24-h ambulatory BP. Additionally, the limitation of including only studies that controlled potassium supplementation through urinary potassium excretion may have resulted in the exclusion of relevant studies. Unfortunately, the present meta-analysis does not include studies conducted on subjects with renal failure. However, international recommendations advise limiting potassium intake, and to our knowledge, there are no studies on potassium supplementation in patients with renal failure. In the studies included in our meta-analysis, variations in potassium intake were limited, especially in subgroups of subjects with hypertension, and a plateau in BP reduction may occur in response to higher potassium intakes. In previous meta-analyses involving mixed populations (subjects with or without hypertension), a U- or J-shaped curve could be artefactual, as the studies with the greatest variations in potassium intake were conducted in subjects with hypertension, in whom the effect on BP is very modest. For this reason, the results and figures of the meta-analysis performed on the “all group” population are not reported.
This study could provide a basis for the formulation of targeted nutritional recommendations and interventions adapted to specific populations, contributing to the prevention of cardiovascular disease and the management of hypertension. These objectives could only be achieved if the difficulties of implementing and enforcing long-term interventions to increase potassium intake are resolved. These difficulties include industrial opacity regarding the composition of processed foods [30] and logistical challenges associated with 24-h urine collections, which remain the standard measurement method. Today, most efforts are focused on salt. Efforts should also be made to make individuals, and particularly those with hypertension, aware of the cardiovascular risks associated with a low potassium intake.
Supplementary Material
ACKNOWLEDGEMENTS
We would like to thank Léonie Belot for her help in revising this article to improve its linguistic quality.
Contributor Information
Maelys Granal, Department of Statistics and Modeling for Health Sciences, Laboratoire de Biométrie et Biologie Evolutive UMR CNRS 5558, Université Lyon 1, Université de Lyon, Villeurbanne, France; Hospices Civils de Lyon, Hôpital Edouard Herriot, Service de Néphrologie, Lyon, France.
Victoria Sourd, Department of Statistics and Modeling for Health Sciences, Laboratoire de Biométrie et Biologie Evolutive UMR CNRS 5558, Université Lyon 1, Université de Lyon, Villeurbanne, France.
Michel Burnier, Faculté de Biologie et Médecine, Université de Lausanne, Lausanne, Switzerland.
Jean Pierre Fauvel, Department of Statistics and Modeling for Health Sciences, Laboratoire de Biométrie et Biologie Evolutive UMR CNRS 5558, Université Lyon 1, Université de Lyon, Villeurbanne, France; Hospices Civils de Lyon, Hôpital Edouard Herriot, Service de Néphrologie, Lyon, France.
Arthur Gougeon, Department of Statistics and Modeling for Health Sciences, Laboratoire de Biométrie et Biologie Evolutive UMR CNRS 5558, Université Lyon 1, Université de Lyon, Villeurbanne, France; RCTs, Lyon, France.
FUNDING
We would like to thank the Fondation Philanthropia, the Droux family and A. Lefranc for their selfless financial support, which made this research possible.
AUTHORS’ CONTRIBUTIONS
M.G., J.P.F. and A.G. formulated the review question and defined the inclusion and exclusion criteria. M.G. wrote the protocol and developed the search strategy in three specific databases. Data extraction was performed by M.G., J.P.F. and A.G. M.G. analyzed data and drafted the paper. All authors read and approved the article. The corresponding author attests that all listed authors meet authorship criteria and that no others meeting the criteria have been omitted.
DATA AVAILABILITY STATEMENT
All databases are protected in a password-protected Excel file and stored on password-protected computers. The passwords are changed every 3 months. The databases will be destroyed in 20 years. All data are available from the authors upon reasonable request.
CONFLICT OF INTEREST STATEMENT
The authors declare no conflict of interest.
REFERENCES
- 1. Vos T, Lim SS, Abbafati C et al. Global burden of 369 diseases and injuries in 204 countries and territories, 1990–2019: a systematic analysis for the Global Burden of Disease Study 2019. Lancet 2020;396:1204–22. 10.1016/S0140-6736(20)30925-9 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Murray CJL, Lauer JA, Hutubessy RCW et al. Effectiveness and costs of interventions to lower systolic blood pressure and cholesterol: a global and regional analysis on reduction of cardiovascular-disease risk. Lancet 2003;361:717–25. 10.1016/S0140-6736(03)12655-4 [DOI] [PubMed] [Google Scholar]
- 3. Vaduganathan M, Mensah GA, Turco JV et al. The global burden of cardiovascular diseases and risk. J Am Coll Cardiol 2022;80:2361–71. 10.1016/j.jacc.2022.11.005 [DOI] [PubMed] [Google Scholar]
- 4. Whelton PK. Epidemiology of hypertension. Lancet 1994;344:101–6. 10.1016/S0140-6736(94)91285-8 [DOI] [PubMed] [Google Scholar]
- 5. World Health Organization . Guideline: Potassium Intake for Adults and Children. Geneva: World Health Organization, 2012. Available from: https://iris.who.int/handle/10665/77986 (14 October 2024, date last accessed). [PubMed] [Google Scholar]
- 6. World Health Organisation . Hypertension. Available from: https://www.who.int/news-room/fact-sheets/detail/hypertension (14 October 2024, date last accessed).
- 7. World Health Organization . Sodium Reduction. September 14, 2023. Available from: https://www.who.int/news-room/fact-sheets/detail/salt-reduction (14 October 2024, date last accessed).
- 8. Kotsis V, Stabouli S, Papakatsika S et al. Mechanisms of obesity-induced hypertension. Hypertens Res 2010;33:386–93. 10.1038/hr.2010.9 [DOI] [PubMed] [Google Scholar]
- 9. Gache P, Leuenberger V, Sutter K et al. Hypertension artérielle et consommation d'alcool. Rev Med Suisse 2006;2:2041–6. 10.53738/REVMED.2006.2.78.2041 [DOI] [PubMed] [Google Scholar]
- 10. Whelton PK, Carey RM, Aronow WS et al. 2017 ACC/AHA/AAPA/ABC/ACPM/AGS/APhA/ASH/ASPC/NMA/PCNA Guideline for the prevention, detection, evaluation, and management of high blood pressure in adults: executive summary: a report of the American College of Cardiology/American Heart Association Task Force on clinical practice guidelines. Hypertension 2018;71:1269–324. [DOI] [PubMed] [Google Scholar]
- 11. Organisation Mondiale de la Santé . Mise en œuvre du Plan d'action pour la prévention et la maîtrise des maladies non transmissibles dans la Région européenne de l'OMS 2016-2025. August 2022, page 6, Available from: https://iris.who.int/bitstream/handle/10665/360924/72wd17f-D-PR-NCDs-220545.pdf?sequence=1&isAllowed=y (14 October 2024, date last accessed). [Google Scholar]
- 12. Leung AA, Nerenberg K, Daskalopoulou SS et al. Hypertension Canada's 2016 Canadian Hypertension Education Program Guidelines for blood pressure measurement, diagnosis, assessment of risk, prevention, and treatment of hypertension. Can J Cardiol 2016;32:569–88. 10.1016/j.cjca.2016.02.066 [DOI] [PubMed] [Google Scholar]
- 13. Charchar FJ, Prestes PR, Mills C et al. Lifestyle management of hypertension: International Society of Hypertension position paper endorsed by the World Hypertension League and European Society of Hypertension. J Hypertens 2024;42:23–49. 10.1097/HJH.0000000000003563 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Mancia G, Kreutz R, Brunström M et al. 2023 ESH Guidelines for the management of arterial hypertension The Task Force for the management of arterial hypertension of the European Society of Hypertension: endorsed by the International Society of Hypertension (ISH) and the European Renal Association (ERA). J Hypertens 2023;41:1874–2071. [DOI] [PubMed] [Google Scholar]
- 15. McEvoy JW, McCarthy CP, Bruno RM et al. 2024 ESC guidelines for the management of elevated blood pressure and hypertension. Eur Heart J 2024;45:3912–4018. 10.1093/eurheartj/ehae178 [DOI] [PubMed] [Google Scholar]
- 16. Neal B, Wu Y, Feng X et al. Effect of salt substitution on cardiovascular events and death. N Engl J Med 2021;385:1067–77. 10.1056/NEJMoa2105675 [DOI] [PubMed] [Google Scholar]
- 17. Yuan Y, Jin A, Neal B et al. Salt substitution and salt-supply restriction for lowering blood pressure in elderly care facilities: a cluster-randomized trial. Nat Med 2023;29:973–81. 10.1038/s41591-023-02286-8 [DOI] [PubMed] [Google Scholar]
- 18. American Heart Association . How Potassium Can Help Control High Blood Pressure. Understanding the Benefits of Potassium. 2024. Available from: https://www.heart.org/en/health-topics/high-blood-pressure/changes-you-can-make-to-manage-high-blood-pressure/how-potassium-can-help-control-high-blood-pressure (23 October 2024, date last accessed). [Google Scholar]
- 19. D'Elia L, Barba G, Cappuccio FP et al. Potassium intake, stroke, and cardiovascular disease a meta-analysis of prospective studies. J Am Coll Cardiol 2011;57:1210–9. 10.1016/j.jacc.2010.09.070 [DOI] [PubMed] [Google Scholar]
- 20. Addison WL. The use of sodium chloride, potassium chloride, sodium bromide, and potassium bromide in cases of arterial hypertension which are amenable to potassium chloride. Can Med Assoc J 1928;18:281–5. [PMC free article] [PubMed] [Google Scholar]
- 21. Dai WS, Kuller LH, Miller G. Arterial blood pressure and urinary electrolytes. J Chronic Dis 1984;37:75–84. 10.1016/0021-9681(84)90128-0 [DOI] [PubMed] [Google Scholar]
- 22. Langford HG. Dietary potassium and hypertension: epidemiologic data. Ann Intern Med 1983;98:770. 10.7326/0003-4819-98-5-770 [DOI] [PubMed] [Google Scholar]
- 23. Lever AF, Beretta-Piccoli C, Brown JJ et al. Sodium and potassium in essential hypertension. BMJ 1981;283:463–8. 10.1136/bmj.283.6289.463 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Walker WG, Whelton PK, Saito H et al. Relation between blood pressure and renin, renin substrate, angiotensin II, aldosterone and urinary sodium and potassium in 574 ambulatory subjects. Hypertension 1979;1:287–91. 10.1161/01.HYP.1.3.287 [DOI] [PubMed] [Google Scholar]
- 25. Tannen RL. The influence of potassium on blood pressure. Kidney Int Suppl 1987;22:S242–8. [PubMed] [Google Scholar]
- 26. St-Jules DE, Fouque D. Etiology-based dietary approach for managing hyperkalemia in people with chronic kidney disease. Nutr Rev 2022;80:2198–205. 10.1093/nutrit/nuac026 [DOI] [PubMed] [Google Scholar]
- 27. Burnier M. Should we eat more potassium to better control blood pressure in hypertension? Nephrol Dial Transplant 2019;34:184–93. 10.1093/ndt/gfx340 [DOI] [PubMed] [Google Scholar]
- 28. Krishna GG. Effect of potassium intake on blood pressure. J Am Soc Nephrol 1990;1:43–52. 10.1681/ASN.V1143 [DOI] [PubMed] [Google Scholar]
- 29. Ma Y, He FJ, Sun Q et al. 24-Hour urinary sodium and potassium excretion and cardiovascular risk. N Engl J Med 2022;386:252–63. 10.1056/NEJMoa2109794 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30. Chan RJ, Parikh N, Ahmed S et al. Blood pressure control should focus on more potassium: controversies in hypertension. Hypertension 2024;81:501–9. 10.1161/HYPERTENSIONAHA.123.20545 [DOI] [PubMed] [Google Scholar]
- 31. Siani A, Strazzullo P, Giacco A et al. Increasing the dietary potassium intake reduces the need for antihypertensive medication. Ann Intern Med 1991;115:753–9. 10.7326/0003-4819-115-10-753 [DOI] [PubMed] [Google Scholar]
- 32. Zisimopoulou S, Pechere-Bertschi A, Burnier M. Potassium et pression artérielle, une vieille histoire revisitée. Rev Med Suisse 2016;12:1502–6. 10.53738/REVMED.2016.12.530.1502 [DOI] [PubMed] [Google Scholar]
- 33. Houston MC. The importance of potassium in managing hypertension. Curr Hypertens Rep 2011;13:309–17. 10.1007/s11906-011-0197-8 [DOI] [PubMed] [Google Scholar]
- 34. Filippini T, Violi F, D'Amico R et al. The effect of potassium supplementation on blood pressure in hypertensive subjects: a systematic review and meta-analysis. Int J Cardiol 2017;230:127–35. 10.1016/j.ijcard.2016.12.048 [DOI] [PubMed] [Google Scholar]
- 35. Filippini T, Naska A, Kasdagli M et al. Potassium intake and blood pressure: a dose-response meta-analysis of randomized controlled trials. J Am Heart Assoc 2020;9:e015719. 10.1161/JAHA.119.015719 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36. Geleijnse JM, Kok FJ, Grobbee DE. Blood pressure response to changes in sodium and potassium intake: a metaregression analysis of randomised trials. J Hum Hypertens 2003;17:471–80. 10.1038/sj.jhh.1001575 [DOI] [PubMed] [Google Scholar]
- 37. Cappuccio FP, MacGregor GA. Does potassium supplementation lower blood pressure? A meta-analysis of published trials. J Hypertens 1991;9:465–73. 10.1097/00004872-199105000-00011 [DOI] [PubMed] [Google Scholar]
- 38. Whelton PK. Effects of oral potassium on blood pressure: meta-analysis of randomized controlled clinical trials. JAMA 1997;277:1624. 10.1001/jama.1997.03540440058033 [DOI] [PubMed] [Google Scholar]
- 39. Haute Autorité de santé—Société française d'hypertension artérielle. Prise en charge de l'hypertension artérielle de l'adulte. Recommandation de bonne pratique. September 7, 2016. Available from: https://www.has-sante.fr/jcms/c_2059286/fr/prise-en-charge-de-l-hypertension-arterielle-de-l-adulte (12 November 2024, date last accessed). [Google Scholar]
- 40. Institut national de la santé et de la recherche médicale (INSERM) . Hypertension artérielle. Une affection cardiovasculaire fréquente aux conséquences sévères. 2018. Available from: https://www.inserm.fr/dossier/hypertension-arterielle-hta/ (12 November 2024, date last accessed). [Google Scholar]
- 41. National Health Service (NHS) . High Blood Pressure. 2023. Available from: https://www.nhs.uk/conditions/high-blood-pressure/ (12 November 2024, date last accessed). [Google Scholar]
- 42. Higgins J, Thomas J, Chandler J et al. Cochrane Handbook for Systematic Reviews of Interventions. Version 6.5, 2024. Available from: https://training.cochrane.org/handbook/current (17 September 2024, date last accessed). [Google Scholar]
- 43. PRISMA-P Group, Moher D, Shamseer L et al. Preferred reporting items for systematic review and meta-analysis protocols (PRISMA-P) 2015 statement. Syst Rev 2015;4. 10.1186/2046-4053-4-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44. Egger M, Smith GD, Schneider M et al. Bias in meta-analysis detected by a simple, graphical test. BMJ 1997;315:629–34. 10.1136/bmj.315.7109.629 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 45. R Core Team 2022. R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna, Austria. Available from: https://www.R-project.org/. n.d . [Google Scholar]
- 46. Balduzzi S, Rücker G, Schwarzer G. How to perform a meta-analysis with R: a practical tutorial. Evid Based Ment Health 2019;22:153–60. 10.1136/ebmental-2019-300117 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 47. Higgins JPT, Altman DG, Gotzsche PC et al. The Cochrane Collaboration's tool for assessing risk of bias in randomised trials. BMJ 2011;343:d5928. 10.1136/bmj.d5928 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 48. Sterne JAC, Savović J, Page MJ et al. RoB 2: a revised tool for assessing risk of bias in randomised trials. BMJ 2019;366:l4898. 10.1136/bmj.l4898 [DOI] [PubMed] [Google Scholar]
- 49. 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. Br J Nutr 2010;104:1839–47. 10.1017/S0007114510002904 [DOI] [PubMed] [Google Scholar]
- 50. Dreier R, Abdolalizadeh B, Asferg CL et al. Effect of increased potassium intake on the renin–angiotensin–aldosterone system and subcutaneous resistance arteries: a randomized crossover study. Nephrol Dial Transplant 2021;36:1282–91. 10.1093/ndt/gfaa114 [DOI] [PubMed] [Google Scholar]
- 51. Franzoni F, Santoro G, Carpi A et al. Antihypertensive effect of oral potassium aspartate supplementation in mild to moderate arterial hypertension. Biomed Pharmacother 2005;59:25–9. 10.1016/j.biopha.2004.11.002 [DOI] [PubMed] [Google Scholar]
- 52. Gijsbers L, Dower JI, Mensink M et al. Effects of sodium and potassium supplementation on blood pressure and arterial stiffness: a fully controlled dietary intervention study. J Hum Hypertens 2015;29:592–8. 10.1038/jhh.2015.3 [DOI] [PubMed] [Google Scholar]
- 53. Graham UM, McCance DR, Young IS et al. A randomised controlled trial evaluating the effect of potassium supplementation on vascular function and the renin–angiotensin–aldosterone system. J Hum Hypertens 2014;28:333–9. 10.1038/jhh.2013.89 [DOI] [PubMed] [Google Scholar]
- 54. 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;55:681–8. 10.1161/HYPERTENSIONAHA.109.147488 [DOI] [PubMed] [Google Scholar]
- 55. Liu FQ, Liu SQ, Zhang Y et al. Effects of salt loading on plasma osteoprotegerin levels and protective role of potassium supplement in normotensive subjects. Circ J 2017;81:77–81. 10.1253/circj.CJ-16-0756 [DOI] [PubMed] [Google Scholar]
- 56. Matthesen SK, Larsen T, Vase H et al. Effect of potassium supplementation on renal tubular function, ambulatory blood pressure and pulse wave velocity in healthy humans. Scand J Clin Lab Invest 2012;72:78–86. 10.3109/00365513.2011.635216 [DOI] [PubMed] [Google Scholar]
- 57. Stone MS, Martin BR, Weaver CM. Short-term RCT of increased dietary potassium from potato or potassium gluconate: effect on blood pressure, microcirculation, and potassium and sodium retention in pre-hypertensive-to-hypertensive adults. Nutrients 2021;13:1610. 10.3390/nu13051610 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 58. Vongpatanasin W, Peri-Okonny P, Velasco A et al. Effects of potassium magnesium citrate supplementation on 24-hour ambulatory blood pressure and oxidative stress marker in prehypertensive and hypertensive subjects. Am J Cardiol 2016;118:849–53. 10.1016/j.amjcard.2016.06.041 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 59. Aburto NJ, Hanson S, Gutierrez H et al. Effect of increased potassium intake on cardiovascular risk factors and disease: systematic review and meta-analyses. BMJ 2013;346:f1378. 10.1136/bmj.f1378 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 60. Granal M, Fouque D, Ducher M et al. Factors associated with kalemia in renal disease. Nephrol Dial Transplant 2023;38:2067–76. 10.1093/ndt/gfad015 [DOI] [PubMed] [Google Scholar]
- 61. Lepiller O. Studying Food and Eaters: a Cocktail of Perspectives and Methods. Éditions Quae. 2023. [PubMed]
- 62. Rhee MY, Kim JH, Shin SJ et al. Estimation of 24-hour urinary sodium excretion using spot urine samples. Nutrients 2014;6:2360–75. 10.3390/nu6062360 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 63. Zhang Y, Peng Y, Li K et al. Assessing whether a spot urine specimen can predict 24-h urinary sodium excretion accurately: a validation study. J Hypertens 2019;37:99–108. 10.1097/HJH.0000000000001879 [DOI] [PubMed] [Google Scholar]
- 64. Drawz PE, Abdalla M, Rahman M. Blood pressure measurement: clinic, home, ambulatory, and beyond. Am J Kidney Dis 2012;60:449–62. 10.1053/j.ajkd.2012.01.026 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 65. Ramos CI, González-Ortiz A, Espinosa-Cuevas A et al. Does dietary potassium intake associate with hyperkalemia in patients with chronic kidney disease? Nephrol Dial Transplant 2021;36:2049–57. 10.1093/ndt/gfaa232 [DOI] [PubMed] [Google Scholar]
- 66. St-Jules DE, Goldfarb DS, Pompeii ML et al. Assessment and misassessment of potassium, phosphorus, and protein in the hemodialysis diet. Semin Dial 2018;31:479–86. 10.1111/sdi.12713 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
All databases are protected in a password-protected Excel file and stored on password-protected computers. The passwords are changed every 3 months. The databases will be destroyed in 20 years. All data are available from the authors upon reasonable request.