Current issues of ACP Journal Club are published in Annals of Internal Medicine


Meta-analysis: Oral potassium reduces blood pressure in patients with hypertension

ACP J Club. 1991 Sept-Oct;115:46. doi:10.7326/ACPJC-1991-115-2-046

Source Citation

Cappuccio FP, MacGregor GA. Does potassium supplementation lower blood pressure? A meta-analysis of published trials. J Hypertens. 1991 May;9:465-73.



To clarify the results of studies on the effects of oral potassium supplementation in lowering blood pressure (BP).

Data sources

Clinical trials published between 1980 and 1989 on the effect of potassium supplementation on BP were identified through manual searches of Current Contents, Index Medicus, and references in pertinent articles and through computer searches of MEDLINE and databases of the Institute for Scientific Information.

Study selection

Of 19 studies giving information on the effect of potassium supplementation on BP, 10 used a crossover design, 4 were parallel-group studies, and 5 used a sequential design. 11 of the 19 used a randomized, double-blind design. Participants in 13 studies were patients with varying degrees of essential hypertension and in 6 studies participants had normal BP.

Data extraction

Duration of the trial, amount and formulation of potassium administered, numbers of participants, demographic characteristics, and baseline and endpoint systolic and diastolic BP levels were extracted from the articles or from correspondence with their authors.

Main results

Of the 586 participants who received potassium (mostly potassium chloride; median dose, 96 mmol/d; range, 48 to 140 mmol/d), 412 had essential hypertension; 240 participants received placebo. For the 13 studies including only patients with hypertension, pooled estimates indicated that the mean effect of potassium supplementation was to reduce supine systolic BP by 8.2 mm Hg (95% CI 7.3 to 9.1 mm Hg); to reduce supine diastolic BP by 4.5 mm Hg (CI 3.8 to 5.2 mm Hg); to reduce standing systolic BP by 11.9 mm Hg (CI 10.5 to 13.3); and to reduce standing diastolic BP by 5.4 mm Hg (CI 4.4 to 6.4 mm Hg). These reductions were greater than those calculated from all 19 trials including participants with normal BP (supine BP reduction 5.9/3.4 mm Hg, standing reduction 10.1/4.7 mm Hg). Median duration of potassium supplementation was 28 days (range, 5 to 112 d). A weighted regression analysis showed a statistically significant relation between greater BP fall and longer duration of supplementation. A separate analysis showed the higher the baseline BP, the greater the response to a potassium supplement.


Oral potassium supplementation resultes in a reduction in blood pressure in patients with varying degrees of essential hypertension. Based on between-study differences, the effect appears greater with a higher baseline blood pressure and with longer duration of supplementation.

Source of funding: Not stated.

Address for article reprint: Dr. F.P. Cappuccio, Blood Pressure Unit, Department of Physiological Medicine, St. George's Hospital Medical School, Cranmer Terrace, London SW17 ORE, England, UK.


Quantitative analysis of pooled data (meta-analysis) is useful when studies done previously too small or conflicting, or when large trials are too expensive, to improve estimates of effect size, or to determine sample sizes for future studies. This study satisfies many, but not all, of the criteria for a robust meta-analysis (1). It states fairly clearly where the authors looked for relevant articles but not how many articles were initially found, nor what the selection criteria were for those articles they reviewed. It is very clear about data abstracted from each study but less clear about assessment of study quality. The analysis assesses the consistency of results across studies and appears to be stable for different groups of patients and different types of study designs. But results by gender, race, diuretic therapy, or adverse effects of potassium supplementation could not be reported because of lack of consistent information in the articles reviewed, so generalizability may be limited.

How can we use this meta-analysis in clinical practice and to guide our reading and research? First, the analysis overall suggests that moderate potassium supplementation can reduce blood pressure. Second, the authors estimate that a well-designed crossover trial with less than 50 patients could confirm their findings with 90% power.

Potassium supplementation appears to be a promising additional treatment for essential hypertension. It could be used instead of, or in addition to, pharmacologic therapy, potentially reducing costs. Because hydrochlorothiazide is also inexpensive, a comparison of benefits, risks, compliance, and cost between potassium supplementation and hydrochlorothiazide would be useful.

Val Lawrence, MD
University of TexasSan Antonio, Texas, USA


1. Mulrow CD. The medical review article: state of the science. Ann Intern Med. 1987;106:485-8.