Sodium and potassium intake as determinants of cardiovascular and renal health
Kieneker, Lyanne Marriët
IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite from it. Please check the document version below.
Document Version
Publisher's PDF, also known as Version of record
Publication date: 2019
Link to publication in University of Groningen/UMCG research database
Citation for published version (APA):
Kieneker, L. M. (2019). Sodium and potassium intake as determinants of cardiovascular and renal health. Rijksuniversiteit Groningen.
Copyright
Other than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).
Take-down policy
If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim.
Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons the number of authors shown on this cover page is limited to 10 maximum.
Urinary potassium excretion and risk of
developing hypertension: Prevention of
Renal and Vascular End-stage Disease
study
Lyanne M. Kieneker Ron T. Gansevoort Kenneth J. Mukamal Rudolf A. de Boer Gerjan Navis Stephan J.L. Bakker Michel M. Joosten Hypertension 2014; 4: 769-776ABSTRACT
Background: Previous prospective cohort studies on the association between
potassium intake and risk of hypertension have almost exclusively relied on self-reported dietary data, while repeated 24-hour urine excretions, as estimate of dietary uptake, may provide a more objective and quantitative estimate of this association.
Methods: Risk of hypertension (defined as blood pressure ≥140/90 mm Hg
or initiation of blood pressure-lowering drugs) was prospectively studied in 5,511 normotensive subjects aged 28 to 75 years not using blood pressure-lowering drugs at baseline of the Prevention of Renal and Vascular End-Stage Disease (PREVEND) study. Potassium excretion was measured in 2 24-hour urine specimens at baseline (1997-1998) and midway during follow-up (2001-2003).
Results: Baseline median potassium excretion was 70 mmol/24h (interquartile
range, 57-85 mmol/24h), which corresponds to a dietary potassium intake of ~91 mmol/24h. During a median follow-up of 7.6 years (interquartile range, 5.0-9.3 years), 1,172 subjects developed hypertension. The lowest sex-specific tertile of potassium excretion (men: <68 mmol/24h; women: <58 mmol/24h) had an increased risk of hypertension after multivariable adjustment (hazard ratio, 1.20; 95% confidence interval, 1.05-1.37), compared with the upper 2 tertiles (Pnonlinearity=0.008). The proportion of hypertension attributable to low potassium excretion was 6.2% (95% confidence interval, 1.7%-10.9%). No association was found between the sodium to potassium excretion ratio and risk of hypertension after multivariable adjustment.
Conclusions: Low urinary potassium excretion was associated with an increased
risk of developing hypertension. Dietary strategies to increase potassium intake to the recommended level of 90 mmol/day may have the potential to reduce the incidence of hypertension.
INTRODUCTION
Potassium is an essential mineral which is thought to play an important role in blood pressure regulation (1). Potassium supplementation has been shown to significantly reduce blood pressure in some, but not all randomized controlled trials. Although most meta-analyses of these randomized controlled trials (2-4) found an overall blood pressure-lowering effect, a more comprehensive meta-analysis (5), comprising 21 randomized controlled trials that lasted for ≥4 weeks, observed this effect only among hypertensive subjects.
Long-term prospective cohort studies on the association between dietary potassium and risk of hypertension are limited, and the majority observed no independent relationship (6-10), except one, in which an inverse association was found (11). These observational studies predominantly relied on 24-hour dietary recalls (9, 11) or food frequency questionnaires (6, 7, 10) to assess potassium intake. Such self-reported dietary methods, however, are less objective than urinary measures to assess dietary intake (12). Although 24-hour urine collections are considered the most direct method for estimating dietary potassium (12), few large epidemiological cohort studies have collected them for reasons of costs, logistics and burden.
Hence, the aim of this study was to prospectively examine the association between repeated 24-hour urinary potassium excretions and risk of developing hypertension among subjects free of hypertension at baseline in a cohort with long-term follow-up.
MATERIALS AND METHODS
Study design and populationThe Prevention of Renal and Vascular End-stage Disease (PREVEND) study is a prospective investigation of albuminuria, renal, and cardiovascular disease in a large cohort drawn from the general population. Details of this study are described elsewhere (13, 14). In total, 8,592 individuals constitute the PREVEND cohort and completed an extensive examination in 1997 and 1998 (baseline).
We excluded subjects with hypertension at baseline (n=3,040), subjects requiring dialysis (n=12) and subjects with missing values of urinary analytes at baseline (n=29), leaving 5,511 participants for the analyses. Of these, 4,546 participants completed a second examination between 2001 and 2003, 3,928 participants completed a third examination between 2003 and 2006, and 3,528 participants completed a fourth examination between 2006 and 2008.
The PREVEND study has been approved by the medical ethics committee of the University Medical Center Groningen. Written informed consent was obtained from all participants.
Data collection
The procedures at each examination in the PREVEND study have been described in detail previously (15). In brief, each of the examinations included 2 visits to an outpatient clinic separated by 3 weeks. Before the first visit, all participants completed a self-administered questionnaire regarding demographics, cardiovascular and renal disease history, smoking habits, alcohol consumption, and medication use. In the last week before the second visit, subjects had to collect 2 consecutive 24-hour specimens after thorough oral and written instruction. During the urine collection, the participants were asked to avoid heavy exercise as much as possible. Subjects were also instructed to postpone the urine collection in case of urinary tract infection, menstruation or fever. The collected urine was stored cold (4°C) for a maximum of 4 days before the second visit. After handing in the urine collections, the urine specimens were stored at -20°C. Furthermore, fasting blood samples were provided and stored at -80°C.
Assessment of urinary potassium excretion
Determination of urine potassium concentration was performed on the 24-hour urine specimens of the first (baseline) and second examination by indirect potentiometry with a MEGA clinical chemistry analyzer (Merck, Darmstadt, Germany) (16). The potassium concentration in mmol/L was multiplied by the urine volume in L/24h to obtain a value in mmol/24h. For each of the 2 examinations, we calculated the average value of the paired 24-hour collections.
Ascertainment of hypertension
During both visits of each of the 4 examinations, blood pressure was assessed on the right arm in supine position, every minute for 10 and 8 minutes, respectively, with an automatic Dinamap XL Model 9300 series device (Johnson-Johnson Medical, Tampa, Florida) as described previously (17). The mean of the last 2 recordings from each visit was used. Use of antihypertensive medications was ascertained by a questionnaire at each examination and was complemented by information from a pharmacy-dispensing registry, which has complete information on drug use of >90% of subjects in the PREVEND study.
For this study, incident hypertension was defined as hypertension that occurred after baseline, which included systolic blood pressure of ≥140 mm Hg, a diastolic blood pressure of ≥90 mm Hg, or the use of antihypertensive drugs, in concordance with recommendations from the Seventh Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure (18). Antihypertensive medication use, for the definition of hypertension, included 5 second-level Anatomical Therapeutic Chemical codes: C02 (antihypertensives), C03 (diuretics), C07 (β-blockers), C08 (calcium channel blockers), and C09 (inhibitors).
Assessment of covariates
Body mass index (BMI) was calculated as weight (kilograms) divided by height squared (square meter). Smoking status was categorized as never, former, current <6 cigarettes/day, current 6-20 cigarettes/day, and current >20 cigarettes/day. Alcohol intake was categorized as none, 1 to 4 drinks/month, 2 to 7 drinks/week, 1 to 3 drinks/day, and 4 or more drinks/day. Education was categorized into low (primary education up to those completing intermediate vocational education), average (higher secondary education), and high (higher vocational education and university).
Statistical analyses
Baseline characteristics are presented according to sex-specific tertiles of urinary potassium excretion. Continuous data are presented as mean with SD or as median and interquartile range (IQR) in case of skewed distribution. Categorical
data are presented as percentiles. We calculated the Pearson product–moment correlation coefficient for the paired 24-hour urine specimens at the first and second examination, and for the averaged potassium excretions of the first and the second examination, as estimates of the intraclass correlation coefficient of reliability R (19).
To study the association between potassium excretion (as a categorical [tertiles] and a continuous variable) and risk of hypertension, we used time-dependent Cox proportional hazards regression analyses. For events occurring before the second examination (i.e., between 1997-2003), the average of the 2 baseline 24-hour urinary excretions of potassium (and sodium alike) were used. For events occurring after the second examination (i.e., after 2003), the average of the 2 baseline and 2 follow-up 24-hour urinary excretions were used, because using cumulative averages of dietary factors yield stronger associations than either only baseline or most recent dietary factors (20). Nonlinearity was tested by using the likelihood ratio test, comparing nested models with linear or linear and cubic spline terms. Survival time was defined from baseline until the date of last examination round that participants attended, the incidence of hypertension, death, relocation to an unknown destination, or 1 January, 2009 (end of follow-up). Hazard ratios (HRs) are reported with 95% confidence intervals (CIs).
We included covariables in our models as linear variables if appropriate, or as categorical if discrete, or if their association with hypertension was nonlinear. Additional adjustment for income, as marker of socioeconomic status did not provide further information after accounting for education. Adjustment for race/ ethnicity in the multivariable model did not affect the association between urinary potassium excretion and risk of hypertension and was therefore not included as a confounder. We tested for multicollinearity between urinary excretions of electrolyte and creatinine. All variance inflation factors were <5, which indicates that there is no proof for multicollinearity. We evaluated effect modification by age, sex, BMI, smoking behavior, and 24-hour urinary sodium and albumin excretion in the analyses of risk of hypertension by fitting models containing both main effects and their cross-product terms. The population attributable risk was calculated using the formula p(HR-1)/(1 + p[HR-1]), where p is the prevalence of individuals in the high-risk group (low potassium excretion) and HR is the
associated multivariable-adjusted HR. Upper and lower 95% CIs of the population attributable risk were derived using this formula and the upper and lower 95% CI estimates of the multivariable-adjusted HR.
Despite being considered the gold standard, even 24-hour urine collections may be subject to quality control concerns because of collection errors. To account for potential inadequacies in the timed 24-hour urine collections, we examined the difference between expected and actually measured 24-hour urine volume (21). We defined potential inadequate 24-hour urine collections (i.e., over- or undercollection) as the upper and lower 2.5% of the difference between the estimated and measured volume of a subject’s 24-hour urine sample. The estimated 24-hour urine volume was derived from the formula: creatinine clearance=([urinary creatinine]×24-hour urine volume)/[serum creatinine]), where creatinine clearance was estimated using the Cockcroft-Gault formula (22).
In addition to the analyses on urinary potassium excretion and our previous analyses on urinary sodium excretion and risk of hypertension (23), we also examined the association between the urinary sodium to potassium (Na-K) excretion ratio and risk of hypertension with time-dependent Cox proportional hazards regression analyses. We evaluated effect modification by age, sex, BMI, smoking behavior, and 24-hour urinary albumin excretion by fitting models containing both main effects and their cross-product terms.
All P values are 2 tailed. P value <0.05 was considered statistically significant. All analyses were conducted using the statistical package IBM SPSS (version 20.0.1; SPSS, Chicago, IL) and SAS (version 9.2; SAS Institute, Cary, NC) software.
RESULTS
The median 24-hour potassium excretion for the 2 urine specimens at baseline was 70 mmol (IQR, 57-85 mmol), with a higher value in men (77 mmol; IQR, 63-92) than in women (65 mmol; IQR, 53-78). This median urinary excretion corresponds to a daily dietary potassium intake of ≈91 mmol/24h (≈3,550 mg/day), assuming a gastrointestinal absorption of 77% (24, 25). Baseline characteristics are shown according to sex-specific tertiles of urinary potassium excretion (Table 1). At
baseline, subjects who had a higher potassium excretion were more likely to be younger and had a higher BMI. Men and women in the highest tertile of urinary potassium excretion were less likely to smoke and consumed more alcohol and sodium than men and women in the lowest tertile of excretion. Higher urinary potassium levels were univariately associated with higher plasma levels of aldosterone. The within-subject correlations for potassium excretion between the paired 24-hour urine specimens at the first and second examination were
r=0.59 (P<0.0001; n=5,489) and r=0.64 (P<0.0001; n=4,410), respectively. The
within-subject correlation between the averaged potassium excretions of the first and the second examination (separated by a median of 4.3 years [IQR, 4.0-4.8 years]) was r=0.49 (P<0.0001; N=4,429).
Table 1. Baseline characteristics according to sex-specific tertiles of urinary potassium
excretion in 5,511 participants of the PREVEND study.
Tertiles of urinary potassium excretion,
mmol/24h P value for trend* Male <68 68-86 >86 Female <58 58-74 >74 Participants, n 1,836 1,838 1,837 Women, % 54.7 54.7 54.7 Age, y 45.9 ± 11.6 45.7 ± 10.8 44.2 ± 10.1 <0.001 Race, whites, % 90.7 96.6 98.4 <0.001
Parental history of hypertension, % 27.2 29.1 30.8 0.02 Smoking status, never, % 29.0 30.1 32.9 <0.001 Alcohol consumption, none, % 27.6 21.7 17.7 <0.001
Education, high, % 27.9 35.7 43.0 <0.001
BMI, kg/m2 24.7 ± 3.8 25.2 ± 3.8 25.4 ± 3.9 <0.001 Blood pressure
Systolic, mm Hg 118 ± 11 119 ± 11 119 ± 11 0.02
Diastolic, mm Hg 70 ± 7 70 ± 7 70 ± 7 0.76
Total cholesterol, mmol/L 5.5 ± 1.1 5.5 ± 1.1 5.4 ± 1.0 0.001 HDL cholesterol, mmol/L 1.3 ± 0.4 1.4 ± 0.4 1.4 ± 0.4 0.004 Triglycerides, mmol/L 1.1 (0.8-1.5) 1.0 (0.8-1.4) 1.0 (0.8-1.4) 0.002
Lipid-lowering drugs, % 2.9 3.4 2.4 0.37
Table 1. Continued
Tertiles of urinary potassium excretion,
mmol/24h P value for
trend*
Male <68 68-86 >86
Female <58 58-74 >74
Glucose-lowering drugs, % 0.6 0.5 0.4 0.49
eGFR, mL/min per 1.73 m² 87 (77-97) 87 (77-98) 87 (78-97) 0.96
Proton-pump inhibitors, % 2.3 3.0 2.5 0.70
Plasma potassium, mmol/L† 4.4 ± 0.6 4.3 ± 0.3 4.4 ± 0.6 0.13 Plasma sodium, mmol/L† 142 ± 2 142 ± 3 142 ± 2 0.36 Plasma renin, µIU/mL‡ 19 (12-29) 18 (12-28) 19 (12-28) 0.46 Plasma aldosterone, pg/mL§ 117 (93-149) 118 (93-151) 122 (96-160) 0.001 Urinary excretion of:
Potassium, mmol/24-hour 52 (45-57) 70 (65-76) 92 (83-103) <0.001 Sodium, mmol/24-hour 115 (88-145) 138 (110-170) 154 (122-187) <0.001 Sodium to potassium ratio 2.3 (1.8-2.9) 2.0 (1.6-2.4) 1.6 (1.3-2.0) <0.001 Calcium, mmol/24-hour 3.2 (2.1-4.4) 3.8 (2.7-5.0) 4.1 (2.9-5.5) <0.001 Magnesium, mmol/24-hour 3.3 (2.5-4.0) 3.9 (3.1-4.8) 4.4 (3.4-5.4) <0.001 Creatinine, mmol/24-hour 10.4 (8.7-12.8) 11.9 (10.0-14.3) 13.0 (10.9-16.1) <0.001 Albumin, mg/24-hour 7.3 (5.3-11.8) 8.0 (6.0-12.1) 8.9 (6.4-14.2) <0.001 Continuous variables are reported as mean ± SD or median (interquartile range), and categorical variables are reported as percentage. Abbreviations: BMI, body mass index; eGFR, estimated glomerular filtration rate; HDL, high-density lipoprotein; and PREVEND, Prevention of Renal and Vascular End-Stage Disease. * Determined by χ2 test (categorical variables), linear regression (continuous variables). †Available in 4,627 subjects. ‡ Available in 5,336 subjects. § Available in 4,847 subjects.
During a median follow-up of 7.6 years (IQR, 5.0-9.3 years), 1,172 hypertension cases were detected. The association between urinary potassium excretion and risk of hypertension was nonlinear (P=0.008 for nonlinearity; Table 2). The multivariable-adjusted spline curve confirmed the nonlinear inverse association of urinary potassium excretion with risk of hypertension (Figure 1). Because of the nonlinear association between urinary potassium excretion and risk of hypertension, we combined the upper 2 tertiles of the distribution in further analyses because the increased risk of hypertension was observed only for lower levels of potassium excretion. The lowest sex-specific tertile (men: <68 mmol/24h; women: <58 mmol/24h) had an increased risk of developing hypertension after multivariable adjustment (HR, 1.20; 95% CI, 1.05-1.37) compared with the upper 2 tertiles. In further analyses, we included plasma aldosterone or urinary creatinine
excretion in the multivariable model. This did not appreciably alter the association (HR lowest tertile compared to the upper 2 tertiles, 1.20; 95% CI, 1.05-1.37 and 1.22; 1.06-1.39, respectively). The higher risk associated with low potassium excretion was generally similar in analyses stratified by selected characteristics (Supplemental Figure 1), with no evidence for effect modification by age, BMI, sex, smoking status, alcohol consumption, urinary sodium excretion, or urinary albumin excretion (all P>0.10 for interaction). The proportion of hypertension attributable to low potassium excretion was 6.2% (95% CI, 1.7%-10.9%).
Similar results were found when we accounted for potential inadequacies in the timed 24-hour urine collections by excluding subjects with potential over- or under collections based on the difference between a subject’s estimated and measured volume of 24-hour urine sample (HR first tertile compared with the upper 2 tertiles, 1.23; 95% CI, 1.07-1.40; n=5239, n=1115). Consistent with Figure 1, we observed a curvilinear association between lower levels of potassium excretion and increased risk of hypertension when we divided the population into sex-specific deciles instead of tertiles. The HRs for the lowest 4 deciles of the distribution of potassium excretion were 1.30 (95% CI, 1.06-1.61), 1.12 (95% CI, 0.90-1.37), 1.12 (95% CI, 0.93-1.36) and 1.07 (95% CI, 0.88-1.30), respectively, as compared with the upper 6 deciles.
Na-K excretion ratio
The median Na-K excretion ratio at baseline was 2.0 (IQR, 1.5-2.5), and was slightly lower in women (1.9; IQR, 1.5-2.4) than in men (2.0; IQR, 1.6-2.5). The within-subject correlations for the Na-K excretion ratio between the paired 24-hour urine specimens at the first and second examination were r=0.49 (P<0.0001; n=5,492) and r=0.56 (P<0.0001; n=4,415), respectively. The within-subject correlation between the averaged Na-K excretion ratios of first and the second examination (4.3 years later) was r=0.22 (P<0.0001; n=4,431).
There was no evidence for a deviation from linearity in the association between the Na-K excretion ratio and risk of hypertension (P=0.49 for nonlinearity; Supplemental Table 1). After adjustment for age and sex, a higher Na-K excretion ratio was significantly associated with a higher risk of hypertension (P=0.005
for linear trend) with an HR across extreme tertiles of 1.16 (95% CI, 1.01-1.34). However, the age- and sex-adjusted association was no longer significant after additional adjustment for BMI (HR across extreme tertiles, 1.06; 95% CI, 0.92-1.22) or multivariable adjustment (HR across extreme tertiles, 0.99; 95% CI, 0.85-1.15;
P=0.51 for linear trend). There was no evidence for effect modification by age,
BMI, sex, smoking status, alcohol consumption, or urinary albumin excretion (all
P>0.10 for interaction).
Table 2. Hazard ratios (95% confidence intervals) for risk of hypertension according to
sex-specific tertiles of urinary potassium excretion in 5,511 participants of the PREVEND study.
Tertiles of urinary potassium excretion,
mmol/24h P-value for linear association* P-value for non-linear association† Male <68 68-86 >86 Female <58 58-74 >74 Person-years 9,739 10,919 11,152 No. of events 401 400 371
Age and sex adjusted 1.23 (1.07-1.42) 1.00 (ref) 1.02 (0.89-1.18) 0.08 0.002 Multivariable model 1‡ 1.26 (1.09-1.45) 1.00 (ref) 0.97 (0.84-1.13) 0.03 <0.001 Multivariable model 2§ 1.19 (1.03-1.38) 1.00 (ref) 0.99 (0.86-1.15) 0.18 0.008 Multivariable model 3¶ 1.20 (1.04-1.38) 1.00 (ref) 0.99 (0.85-1.14) 0.17 0.007 Hazard ratios were derived from Cox proportional hazards models. Abbreviations: PREVEND, Preven-tion of Renal and Vascular End-Stage Disease. *Derived from a Cox proporPreven-tional hazards model by using urinary potassium excretion as a continuous linear term. † Derived by using the likelihood ratio test, com-paring nested Cox proportional hazards regression models with a linear or linear and cubic spline terms. ‡ Multivariable model 1 was an age- and sex-adjusted model and was additionally adjusted for body mass index, smoking status, alcohol consumption, parental history of hypertension, and urinary sodium excre-tion. § Multivariable model 2 was multivariable model 1 and was additionally adjusted for education and urinary magnesium and calcium excretion. ¶ Multivariable model 3 was multivariable model 2 and was additionally adjusted for plasma aldosterone.
Figure 1. Association between urinary potassium excretion and risk of hypertension. Data were fit by a Cox
proportional hazards regression model with time-varying covariates based on restricted cubic splines with 4 knots placed on the 7th, 37th, 65th, and 93th percentiles and adjusted for age, sex, body mass index, smoking status, alcohol consumption, parental history of hypertension, education, and urinary sodium, calcium, and magnesium excretion. Dashed lines indicate the 95% confidence interval (CI). The spline curve is truncated at the 1st percentile and 99th percentile of the distribution curve. Reference standard was the baseline median potassium excretion of 70 mmol/24h. HR indicates hazard ratio.
DISCUSSION
In this prospective population-based cohort study, 24-hour urinary potassium excretion, as estimate of dietary uptake, was nonlinearly associated with risk of incident hypertension. Subjects in the lowest tertile had a 20% higher risk of developing hypertension compared with the remainder of the cohort. This association remained after adjustment for conventional risk factors and urinary excretions of several cations, including sodium, and was consistent across strata of age, BMI, sex, smoking status, and urinary excretion of sodium or albumin. The proportion of hypertension attributable to low potassium excretion appeared to be 6.2%, suggesting that almost 1 in 16 incident cases of hypertension might have been prevented if all subjects were in the low-risk group of potassium excretion. An initial positive association between the urinary Na-K excretion ratio and risk of hypertension after age and sex adjustment lost significance after further adjustment, mainly attributable to BMI.
Although we found that low urinary potassium was associated with a higher risk of hypertension, results from previous prospective studies have been inconsistent. The majority of studies assessed potassium intake by means of self-reported dietary data which hampers a direct comparison. Our results are in keeping with the China Health and Nutrition Survey (11), which showed a 34% increased prevalence of hypertension associated with lower potassium consumption, when comparing extreme groups of potassium intake. The Nurses’ Health Study (7, 10), the Health Professionals Follow-up Study (6), and the National Health and Nutrition Examination Survey I (9), on the contrary, did not find a significant association between dietary potassium and risk of hypertension after multivariable adjustment. To date, only 1 observational study has investigated the association between urinary potassium excretion and risk of developing hypertension. Among 1520 ethnic Chinese men and women, 24-hour urinary potassium excretion, extrapolated from a single overnight urine collection, was not associated with risk of incident hypertension (HR extreme groups of potassium excretion, 0.98; 95% CI, 0.78-1.23) (8).
A comprehensive meta-analysis comprising 21 randomized controlled trials (or 1606 participants) that required 24-hour urinary potassium measurements
to estimate potassium intake (5) documented an overall decrease of 3.49 (95% CI, 1.82-5.15) mm Hg in systolic blood pressure and of 1.96 (95% CI, 0.86-3.06) mm Hg in diastolic blood pressure. However, blood pressure reductions were only significant in people with hypertension, perhaps attributable to the limited number of studies (3 randomized controlled trials) (26-28) among normotensive subjects in the meta-analysis. Also, although the largest reductions were seen in the subgroup with achieved potassium intakes of 90 to 120 mmol/day, no (linear) dose-response relationship between potassium intake and blood pressure reductions was observed in this meta-analysis (5). This seems to corroborate the nonlinear association between potassium excretion and risk of hypertension in the current study.
Our data indicate that potassium is associated with an increased risk of hypertension when potassium excretion is <65 mmol/24h, which corresponds to a dietary intake of ≈84 mmol/day when taking into account an average fractional intestinal absorption of 77% (24, 25). Such a lower limit seems to support the recommendation of the 2002 Joint WHO/Food and Agriculture Organization Expert Consultation of a minimal potassium intake of ≥3,500 mg/day (90 mmol/d) for adults (24). In this cohort, 48% of the subjects would have been classified as having a potassium intake below this recommendation and even 84% of the subjects would have had a potassium intake below the adequate intake of 4,700 mg/day (120 mmol/d) recommended by the Institute of Medicine (29). A high prevalence of inadequate potassium intake is not only seen in the Netherlands (30), but also in the United States (31), Canada (29), and China (11).
The antihypertensive effect of dietary potassium may have more than one mechanism. One possibility is that dietary potassium enhances natriuresis, thereby lowering blood pressure (32). Experimental studies have also shown that potassium supplementation may stimulate Na+-K+-ATPase in vascular smooth
muscle cells and adrenergic nerve terminals, resulting in vasodilation (33) or may potentiate endothelium-dependent relaxation (34). The association between potassium excretion and risk of developing hypertension was not affected by adjustment for plasma aldosterone at baseline. This suggests that low urinary potassium excretion influences development of hypertension by a mechanism that does not involve aldosterone.
Despite stronger associations with blood pressure in cross-sectional analyses than sodium or potassium excretion alone (35, 36), we and others (8) did not find a prospective association between the urinary Na-K excretion ratio and risk of hypertension, particularly after BMI adjustment. The Na-K excretion ratio has been shown to be independently associated with total body fat (37), raising the possibility that this ratio is a surrogate of (a poor-quality diet associated with) BMI. Using a ratio makes the implicit assumption that the regression coefficients for the 2 variables are equal in magnitude but opposite in direction (38). This was not the case in our sodium (23) and potassium data. Furthermore, although the Na-K excretion ratio may offer a correction for characteristics of the urine collection such as completeness during the 24-hour period and correlated measurement errors (19), it may also introduce bias because of the combined measurement errors of determining sodium and potassium concentration in urine. Both short-term and long-short-term within-subject correlations between the 24-hour urinary excretions were lower for the Na-K ratio than for potassium alone, indicating a higher repeatability for urinary excretions of potassium than for the Na-K excretion ratio. Regardless, the null finding for the multivariable-adjusted Na-K excretion ratio and the absence of effect modification by sodium excretion in the potassium-hypertension association both suggest that potassium excretion per se may affect the risk of hypertension, irrespective of sodium. However, the vast majority of our population (81%) had a sodium excretion of >100 mmol/24h (>2,300 mg/d). We thus had limited power to investigate whether low potassium excretion is also associated with a higher risk of hypertension among those whose sodium intake does not exceed dietary recommendations. Also, it might be possible that a high dietary potassium protected against developing hypertension despite a high dietary sodium.
Some limitations of this study should be noted. First, we had no information on the dietary origin of the excreted potassium because no dietary records were obtained in the PREVEND study. Potassium is mostly present in fruits, vegetables, meat, potatoes, and dairy products (39). Potassium homeostasis is complex; several factors play a role including electrolytes (sodium and magnesium) and several hormones such as not only insulin, norepinephrin, but also aldosterone and renin. Also, other (dietary) components that are closely associated with
potassium intake, such as alkali and dietary fiber, could potentially be involved in the lower risk for hypertension (6). However, we did adjust for urinary sodium, magnesium, and calcium excretion and for BMI as a proxy for energy intake. Second, participants with mildly elevated levels of urinary albumin, an indicator of kidney impairment, were over-represented in the PREVEND study. However, similar results were obtained after stratification for urinary albumin, showing that mild kidney impairment is unlikely to modify the association of potassium excretion with incident hypertension. Third, >95% of the individuals of the PREVEND study are white and our results of a nonlinear association between urinary potassium excretion and risk of hypertension may not be readily generalizable to different demographic populations. Blacks, for instance, typically show a lower urinary excretion rate of potassium than do whites (40, 41). Fourth, we did not have additional information on recreational drug use other than alcohol consumption and tobacco use. Uncontrolled confounding by unmeasured factors, if associated with potassium excretion and risk of hypertension, remains a possibility. Fifth, there is variation in dietary potassium absorption. In carefully performed balance studies, the mean (SEM) potassium absorption was 77.0% (1.7%), consistent with a 95% CI of absorption of 73.7% to 80.3%, with most extreme values in this report ranging from 64% to 95% (25). Also, ageing of the intestine may have influenced potassium absorption. Sixth, because the subjects of the PREVEND study were in a long term observational trial, their behavior might have been affected. However, the participants were unaware of their urinary potassium excretion, which limits the possibility that subjects changed their diets because of a low potassium excretion. Finally, we studied subjects who were normotensive at baseline. Thus, subjects who had already developed hypertension prior to their inclusion in the PREVEND study, possibly representing the population subgroup with this highest risk, were not included in our analysis.
One of the strengths of this study is the use of multiple 24-hour urine collections updated over time to estimate habitual dietary potassium uptake (and indirectly, intake). Twenty-four hour urine collections are regarded as the gold standard to assess dietary potassium and provide an objective and quantitative measure of intake on a population level. Furthermore, dietary potassium based on urine collections tends to have higher repeatability than those based on 24-hour
dietary recall, as characterized by subject correlations (19). Our within-subject correlations for the paired 24-hour urine assessments for potassium were also similar or higher than those observed in the Trials of Hypertension Prevention (TOHP) (42), the Trial of Nonpharmacological Interventions in the Elderly (TONE) trial (19), and the International Cooperative Study on Salt, Other Factors, and Blood Pressure (INTERSALT) study (43, 44). Despite being considered the gold standard, even 24-hour collections are subject to certain concerns. To evaluate the quality and completeness of the urine specimens, we performed a sensitivity analysis to correct for possible over- or under collections. This did not appreciably alter the results. Also, although greatly reducing bias because of measurement error and random error because of within-person variability over time, 2 24-hour urinary collections at baseline and 2 during follow-up may have been suboptimal to represent habitual potassium uptake and to sufficiently reduce the day-to-day variation in urinary potassium excretion. Other strengths of this study were the prospective design, the use of multiple measured blood pressures and participant and pharmacy information on antihypertensive medication to define hypertension, the use of a large sample size, and the availability of detailed and updated (midway through the period of follow-up to reduce potential misclassification) information on the exposure and potential confounders.
Perspectives
In this large population-based cohort of men and women, low urinary excretion of potassium was associated with a higher risk of hypertension. This association persisted after adjustment for both dietary and nondietary factors and was consistent across several subgroups. A higher consumption of potassium, particularly by those with the lowest potassium excretion, while concurrently not exceeding dietary recommendations on sodium intake, may be a promising approach for the primary prevention of hypertension. These data reinforce the importance of dietary potassium in blood pressure control.
REFERENCES
1. Adrogue HJ, Madias NE. Sodium and potassium in the pathogenesis of hypertension. N Engl J Med 2007;356:1966-78.
2. 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.
3. Whelton PK, He J, Cutler JA, Brancati FL, Appel LJ, Follmann D, Klag MJ. Effects of oral potassium on blood pressure. Meta-analysis of randomized controlled clinical trials. JAMA 1997;277:1624-32.
4. Cappuccio FP, MacGregor GA. Does potassium supplementation lower blood pressure? A meta-analysis of published trials. J Hypertens 1991;9:465-73.
5. Aburto NJ, Hanson S, Gutierrez H, Hooper L, Elliott P, Cappuccio FP. Effect of increased potassium intake on cardiovascular risk factors and disease: systematic review and meta-analyses. BMJ 2013;346:f1378.
6. Ascherio A, Rimm EB, Giovannucci EL, Colditz GA, Rosner B, Willett WC, Sacks F, Stampfer MJ. A prospective study of nutritional factors and hypertension among US men. Circulation 1992;86:1475-84.
7. Ascherio A, Hennekens C, Willett WC, Sacks F, Rosner B, Manson J, Witteman J, Stampfer MJ. Prospective study of nutritional factors, blood pressure, and hypertension among US women. Hypertension 1996;27:1065-72.
8. Chien KL, Hsu HC, Chen PC, Su TC, Chang WT, Chen MF, Lee YT. Urinary sodium and potassium excretion and risk of hypertension in Chinese: report from a community-based cohort study in Taiwan. J Hypertens 2008;26:1750-6.
9. Ford ES, Cooper RS. Risk factors for hypertension in a national cohort study. Hypertension 1991;18:598-606.
10. Witteman JC, Willett WC, Stampfer MJ, Colditz GA, Sacks FM, Speizer FE, Rosner B, Hennekens CH. A prospective study of nutritional factors and hypertension among US women. Circulation 1989;80:1320-7.
11. Du S, Neiman A, Batis C, Wang H, Zhang B, Zhang J, Popkin BM. Understanding the patterns and trends of sodium intake, potassium intake, and sodium to potassium ratio and their effect on hypertension in China. Am J Clin Nutr 2013;99:334-43.
12. Day N, McKeown N, Wong M, Welch A, Bingham S. Epidemiological assessment of diet: a comparison of a 7-day diary with a food frequency questionnaire using urinary markers of nitrogen, potassium and sodium. Int J Epidemiol 2001;30:309-17.
13. Hillege HL, Janssen WM, Bak AA, Diercks GF, Grobbee DE, Crijns HJ, Van Gilst WH, De Zeeuw D, De Jong PE, Prevend Study Group. Microalbuminuria is common, also in a nondiabetic, nonhypertensive population, and an independent indicator of cardiovascular risk factors and cardiovascular morbidity. J Intern Med 2001;249:519-26.
14. Joosten MM, Gansevoort RT, Mukamal KJ, van der Harst P, Geleijnse JM, Feskens EJ, Navis G, Bakker SJ, PREVEND Study Group. Urinary and plasma magnesium and risk of ischemic heart disease. Am J Clin Nutr 2013;97:1299-306.
15. Halbesma N, Brantsma AH, Bakker SJ, Jansen DF, Stolk RP, De Zeeuw D, De Jong PE, Gansevoort RT, PREVEND study group. Gender differences in predictors of the decline of renal function in the general population. Kidney Int 2008;74:505-12.
16. Verhave JC, Hillege HL, Burgerhof JG, Janssen WM, Gansevoort RT, Navis GJ, de Zeeuw D, de Jong PE, PREVEND Study Group. Sodium intake affects urinary albumin excretion especially in overweight subjects. J Intern Med 2004;256:324-30.
17. Joosten MM, Gansevoort RT, Mukamal KJ, Kootstra-Ros JE, Feskens EJ, Geleijnse JM, Navis G, Bakker SJ, PREVEND Study Group. Urinary magnesium excretion and risk of hypertension: the prevention of renal and vascular end-stage disease study. Hypertension 2013;61:1161-7.
18. Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults. Executive Summary of The Third Report of The National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, And Treatment of High Blood Cholesterol In Adults (Adult Treatment Panel III). JAMA 2001;285:2486-97.
19. Espeland MA, Kumanyika S, Wilson AC, Reboussin DM, Easter L, Self M, Robertson J, Brown WM, McFarlane M, TONE Cooperative Research Group. Statistical issues in analyzing 24-hour dietary recall and 24-24-hour urine collection data for sodium and potassium intakes. Am J Epidemiol 2001;153:996-1006.
20. Hu FB, Stampfer MJ, Rimm E, Ascherio A, Rosner BA, Spiegelman D, Willett WC. Dietary fat and coronary heart disease: a comparison of approaches for adjusting for total energy intake and modeling repeated dietary measurements. Am J Epidemiol 1999;149:531-40. 21. Joosten MM, Gansevoort RT, Mukamal KJ, Lambers Heerspink HJ, Geleijnse JM, Feskens
EJ, Navis G, Bakker SJ, PREVEND Study Group. Sodium excretion and risk of developing coronary heart disease. Circulation 2014;129:1121-8.
22. Cockcroft DW, Gault MH. Prediction of creatinine clearance from serum creatinine. Nephron 1976;16:31-41.
23. Forman JP, Scheven L, de Jong PE, Bakker SJ, Curhan GC, Gansevoort RT. Association between sodium intake and change in uric acid, urine albumin excretion, and the risk of developing hypertension. Circulation 2012;125:3108-16.
24. World Health Organization. Guideline: potassium intake for adults and children. Geneva: WHO, 2012.
25. Holbrook JT, Patterson KY, Bodner JE, Douglas LW, Veillon C, Kelsay JL, Mertz W, Smith JC,Jr. Sodium and potassium intake and balance in adults consuming self-selected diets. Am J Clin Nutr 1984;40:786-93.
26. Barden A, Vandongen R, Beilin LJ. Increases in urinary kallikrein activity and prostanoid synthesis after dietary potassium supplementation. Clin Exp Pharmacol Physiol 1987;14:565-72.
27. Hypertension Prevention Trial Research Group. The effects of nonpharmacologic interventions on blood pressure of persons with high normal levels. Results of the Trials of Hypertension Prevention, Phase I. JAMA 1992;267:1213-20.
28. Whelton PK, Buring J, Borhani NO, Cohen JD, Cook N, Cutler JA, Kiley JE, Kuller LH, Satterfield S, Sacks FM. The effect of potassium supplementation in persons with a high-normal blood pressure. Results from phase I of the Trials of Hypertension Prevention (TOHP). Trials of Hypertension Prevention (TOHP) Collaborative Research Group. Ann Epidemiol 1995;5:85-95.
29. Panel on Dietary Reference Intakes for Electrolytes and Water, Standing Committee on the Scientific Evaluation of Dietary Reference Intakes. Dietary reference intakes for water, potassium, sodium, chloride, and sulfate. Washington, D.C: National Academies Press, 2005.
30. van Rossum CTM, Fransen HP, Verkaik-Kloosterman J, Buurma-Rethans EJM, Ocké MC. Dutch National Food Consumption Survey 2007–2010: Diet of children and adults aged 7 to 69 years. Bilthoven: RIVM, 2011. no. 350050006/2011.
31. Cogswell ME, Zhang Z, Carriquiry AL, Gunn JP, Kuklina EV, Saydah SH, Yang Q, Moshfegh AJ. Sodium and potassium intakes among US adults: NHANES 2003-2008. Am J Clin Nutr 2012;96:647-57.
32. Smith SR, Klotman PE, Svetkey LP. Potassium chloride lowers blood pressure and causes natriuresis in older patients with hypertension. J Am Soc Nephrol 1992;2:1302-9. 33. Haddy FJ. Ionic control of vascular smooth muscle cells. Kidney Int Suppl 1988;25:S2-8. 34. Raij L, Luscher TF, Vanhoutte PM. High potassium diet augments endothelium-dependent
relaxations in the Dahl rat. Hypertension 1988;12:562-7.
35. Intersalt Cooperative Research Group. Intersalt: an international study of electrolyte excretion and blood pressure. Results for 24hour urinary sodium and potassium excretion. Intersalt Cooperative Research Group. BMJ 1988;297:319-28.
36. Khaw KT, Barrett-Connor E. The association between blood pressure, age, and dietary sodium and potassium: a population study. Circulation 1988;77:53-61.
37. Jain N, Minhajuddin AT, Neeland IJ, Elsayed EF, Vega GL, Hedayati SS. Association of urinary sodium-to-potassium ratio with obesity in a multiethnic cohort. Am J Clin Nutr 2014;99:992-8.
38. Liew G, Sharrett AR, Kronmal R, Klein R, Wong TY, Mitchell P, Kifley A, Wang JJ. Measurement of retinal vascular caliber: issues and alternatives to using the arteriole to venule ratio. Invest Ophthalmol Vis Sci 2007;48:52-7.
39. Cordain L, Eaton SB, Sebastian A, Mann N, Lindeberg S, Watkins BA, O’Keefe JH, Brand-Miller J. Origins and evolution of the Western diet: health implications for the 21st century. Am J Clin Nutr 2005;81:341-54.
40. Pratt JH, Jones JJ, Miller JZ, Wagner MA, Fineberg NS. Racial differences in aldosterone excretion and plasma aldosterone concentrations in children. N Engl J Med 1989;321:1152-7.
41. Turban S, Miller ER,3rd, Ange B, Appel LJ. Racial differences in urinary potassium excretion. J Am Soc Nephrol 2008;19:1396-402.
42. Cook NR, Kumanyika SK, Cutler JA. Effect of change in sodium excretion on change in blood pressure corrected for measurement error. The Trials of Hypertension Prevention, Phase I. Am J Epidemiol 1998;148:431-44.
43. Dyer AR, Elliott P, Shipley M. Urinary electrolyte excretion in 24hours and blood pressure in the INTERSALT Study. II. Estimates of electrolyte-blood pressure associations corrected for regression dilution bias. The INTERSALT Cooperative Research Group. Am J Epidemiol 1994;139:940-51.
44. Dyer AR, Shipley M, Elliott P. Urinary electrolyte excretion in 24hours and blood pressure in the INTERSALT Study. I. Estimates of reliability. The INTERSALT Cooperative Research Group. Am J Epidemiol 1994;139:927-39.
SUPPLEMENTAL DATA: MATERIALS AND METHODS
Study design and populationIn summary, from 1997 to 1998, all inhabitants of the city Groningen, the Netherlands, aged 28 to 75 years (n=85,421), were sent a vial to collect a first morning void urine sample and a short questionnaire on demographics and renal and cardiovascular morbidity. Altogether, 40,856 people (48%) responded and their urinary albumin concentration was assessed. After exclusion of pregnant women and subjects with type I diabetes mellitus, subjects with a urinary albumin concentration of ≥10 mg/L (n=7,768) were invited to participate, of whom 6,000 did so. In addition, a randomly selected group with a urinary albumin concentration of <10 mg/L (n=3,394) was invited to participate in the cohort, of whom 2,592 joined. These 8,592 individuals constitute the PREVEND cohort and completed an extensive examination in 1997 and 1998 (baseline).
Laboratory assays
Sodium, calcium, magnesium, creatinine and albumin in urine and circulating potassium, sodium, total cholesterol, HDL cholesterol, triglycerides and glucose were determined as previously described (1, 2). Estimated glomerular filtration rate (eGFR) was calculated from the Chronic Kidney Disease Epidemiology Collaboration equation (3). Plasma renin was measured using an automated sandwich immunochemiluminescent assay (LIAISON, Diasorin, DiaSorin Ltd, Schiphol Rijk, The Netherlands) as described previously (4). Plasma aldosterone was measured using an enzyme immunoassay (Alpco Diagnostics, Catalog Number: 11-ALDHU-E01, Alpco, Salem, NH).
REFERENCES
1. Verhave JC, Hillege HL, Burgerhof JG, Janssen WM, Gansevoort RT, Navis GJ, de Zeeuw D, de Jong PE, PREVEND Study Group. Sodium intake affects urinary albumin excretion especially in overweight subjects. J Intern Med 2004;256:324-30.
2. Oterdoom LH, Gansevoort RT, Schouten JP, de Jong PE, Gans RO, Bakker SJ. Urinary creatinine excretion, an indirect measure of muscle mass, is an independent predictor of cardiovascular disease and mortality in the general population. Atherosclerosis 2009;207:534-40.
3. Levey AS, Stevens LA, Schmid CH, Zhang YL, Castro AF,3rd, Feldman HI, Kusek JW, Eggers P, Van Lente F, Greene T, et al. A new equation to estimate glomerular filtration rate. Ann Intern Med 2009;150:604-12.
4. de Boer RA, Schroten NF, Bakker SJ, Mahmud H, Szymanski MK, van der Harst P, Gansevoort RT, van Veldhuisen DJ, van Gilst WH, Hillege HL. Plasma renin and outcome in
Su p p le m e n ta l T ab le 1 . H aza rd ra ti o s (9 5% co n fi d ence in te rv al s) fo r ri sk o f h yp er te n sio n ac co rd in g to se x-sp ec ifi c te rti le s o f so d iu m t o p o ta ss iu m e xc re ti o n r ati o i n 5 ,5 11 p ar ti ci p an ts o f t h e P R E V EN D s tud y. Te rti le s o f s o d iu m -p o ta ss iu m e xc re ti o n r a ti o P -v a lu e f o r li n e a r as so ci a ti o n * P -v a lu e f o r no n -l in e a r as so ci a ti o n † Ma le <1 .7 1. 7-2 .3 >2 .3 Fe m al e <1 .6 1. 6 -2 .2 >2 .2 Pe rs o n -y e ar s 10 ,8 47 10 ,7 51 10 ,2 1 3 N u m b e r o f e ve n ts 37 2 41 2 38 8 A ge - a n d s e x-ad ju st e d 0 .9 2 ( 0 .8 0 -1 .0 6) 1 .0 0 ( re f) 1. 0 7 ( 0 .9 3 -1. 2 3) 0. 0 0 5 0. 5 0 M u lti va ri ab le m o d e l 1 ‡ 0 .9 7 ( 0 .8 5 -1 .1 2) 1 .0 0 ( re f) 1 .01 ( 0 .8 8 -1 .1 6) 0. 17 0. 17 M u lti va ri ab le m o d e l 2 § 0 .9 9 ( 0 .8 6 -1 .1 4) 1 .0 0 ( re f) 0 .9 8 ( 0 .8 5 -1 .1 3) 0. 51 0. 4 9 H az ar d r ati o s w e re d e ri ve d f ro m C o x p ro p o rti o n al h az ar d s m o d e ls . * D e ri ve d f ro m a C o x p ro p o rti o n al h az ar d s m o d e l b y u si n g u ri n ar y p o ta ss iu m e xc re ti o n a s a co n ti n u o u s l in e ar t e rm . † D e ri ve d b y u si n g t h e l ik e lih o o d r ati o t e st , c o m p ar in g n e st e d C o x p ro p o rti o n al h az ar d s r e gr e ss io n m o d e ls w it h a l in e ar o r l in e ar a n d c u b ic sp lin e t e rm s. A b b re vi ati o n s: P R E V EN D , P re ve n ti o n o f R e n al a n d V as cu la r E n d -S ta ge D is e as e. ‡ M u lti va ri ab le m o d e l 1 w as a n a ge -a d ju st e d m o d e l a n d w as a d d iti o n -al ly a d ju st e d f o r s e x, b o d y m as s i n d e x, s m o ki n g s ta tu s, a lc o h o l c o n su m p ti o n a n d p ar e n ta l h is to ry o f h yp e rt e n si o n . § M u lti va ri ab le m o d e l 2 w as a d ju st e d a s m o d e l 1 p lu s a d ju st e d f o r e d u ca ti o n a n d u ri n ar y m ag n e si u m a n d c al ci u m e xc re ti o n .
Supplemental Table 2. Systolic blood pressure, diastolic blood pressure and pulse
pressure at time of follow-up.
Time of follow-up Systolic blood pressure
(mm Hg)
Diastolic blood pres-sure (mm Hg) Pulse pressure (mm Hg) Baseline (N=5,511) 119 ± 11 70 ± 7 49 ± 8 Screening two (N=4,546) 119 ± 13 71 ± 8 48 ± 9 Screening three (N=3,928) 120 ± 14 72 ± 8 49 ± 10 Screening four (N=3,528) 122 ± 15 73 ± 9 49 ± 11 Values presented as mean ± SD.
Supplemental Figure 1. Association between low urinary potassium excretion (men: <68 mmol/24-hour;
women: <58 mmol/24-hour) and risk of hypertension in the overall population and stratified by selected characteristics (N=number of subjects, n=number of cases). Multivariable-adjusted hazard ratios (95% con-fidence intervals) for risk of hypertension for the lowest tertile compared with the upper 2 tertiles. A hazard ratio higher than 1 indicates that the lowest tertile of urinary potassium excretion is associated in the direc-tion of a higher risk for developing hypertension. Hazard ratios were derived from Cox propordirec-tional hazards regression models with time-varying covariates and adjusted for age, sex, body mass index, smoking status, parental history of hypertension, alcohol consumption, education, and urinary excretion of sodium, calcium, and magnesium. The P-values denote the P for interaction.
