Review: Potassium supplementation reduces blood pressure
ACP J Club. 1997 Nov-Dec;127:72. doi:10.7326/ACPJC-1997-127-3-072
Whelton PK, He J, Cutler JA, et al. Effects of oral potassium on blood pressure. Meta-analysis of randomized controlled clinical trials. JAMA. 1997 May 28;277:1624-32.
To determine the effectiveness of supplementation with oral potassium on blood pressure by using meta-analysis.
Studies were identified by searching MEDLINE using the terms blood pressure, dietary potassium, potassium, and potassium chloride; bibliographies of studies and review articles were reviewed; and reprint files were searched.
Studies were selected if they were randomized controlled trials of potassium supplementation published before 1995 and reported mean changes in systolic and diastolic blood pressure. Study groups had to have no intervention difference except for the potassium supplementation. The groups included normotensive, hypertensive, treated, and untreated patients.
Data were extracted on patient characteristics; study design; study duration; presence of hypertension and use of antihypertensive medications; method of blood pressure measurement; dose and formulation of potassium supplementation and control treatment; mean baseline blood pressure; mean baseline 24-hour urinary excretion of sodium and potassium; mean change from baseline in urinary excretion, body weight, and blood pressure; and mean 24-hour urinary excretion during follow-up. Data were extracted independently by 2 reviewers.
33 trials (2609 patients) met the inclusion criteria. Potassium supplementation was associated with a mean net change in blood pressure varying from -41.0 to 2.8 mm Hg for systolic blood pressure (SBP) and from -17.0 to 4.8 mm Hg for diastolic blood pressure (DBP). An overall pooled net decrease in SBP (P < 0.001) and in DBP (P < 0.01) was found (Table). When an outlier trial that had a mean change in SBP of -41.0 and in DBP of -17.0 mm Hg was excluded, the pooled net changes in SBP and DBP still reached statistical significance (P < 0.001 and P < 0.01, respectively) (Table). Effect sizes were greater in 28 studies in which no antihyper-tensive medications were given (net SBP change -4.85, CI -2.74 to -6.95, P < 0.001; net DBP change -2.71, CI -0.80 to -4.61, P < 0.01).
Potassium supplementation lowers systolic and diastolic blood pressure.
Source of funding: National Institutes of Health.
For article reprint: Dr. P.K. Whelton, Office of the Dean, Tulane University, 17th Floor, 1502 Canal Street, New Orleans, LA 70112, USA. FAX 504-588-5718.
Table. Potassium supplementation vs control*
|Variable||Systolic blood pressure net change in mm Hg (95% CI)||Diastolic blood pressure net change in mm Hg (CI)|
|All trials||-4.44 (-2.53 to -6.36)||-2.45 (-0.74 to -4.16)|
|All trials without outlier||-3.11 (-1.91 to -4.31)||-1.97 (-0.52 to -3.42)|
*CI defined in Glossary.
Whelton and colleagues have done an exemplary overview to address the issue of potassium supplementation and its effect on blood pressure. Careful attention to the selection of articles, rigorous exclusion criteria, independent review by several authors, the use of tests of homogeneity, and even estimates of the effect of publication bias surpass most reviewers' methodologic criteria for overviews. In addition, the topic lends itself to this approach. Many relatively small trials have been done, but no conclusive evidence of an independent effect was found in most of them.
The study concludes that potassium supplementation "should be considered as part of recommendations for prevention and treatment of hypertension." The following issues are appropriate for the clinician to consider. As with other nonpharmacologic interventions, the reduction in blood pressure is modest (Table) in comparison with the use of antihypertensive drugs (1, 2). The actual effect on the risk for a cardiovascular event in any individual patient is likely to be small. Additional costs, including monitoring, ensue if potassium supplementation is prescribed. A slight chance of hyperkalemia exists if supplements are given with an-giotensin-converting enzyme inhibitors, especially in older patients who may have unrecognized renal insufficiency. Further, supplementation may not be as effective in patients who are taking antihypertensive medications (only 4 such studies were included in the meta-analysis, and the effects were minimal). Patient subgroups with volume-responsive hypertension (such as African Americans) may be more responsive to this intervention.
This overview shows that potassium supplementation may lower blood pressure. Most dramatic effects are likely to occur in patients with volume-responsive hypertension who are not receiving antihypertensive medications. Supplementing the diet is probably as effective, cheaper, and safer than prescribing potassium (3). The effect on long-term outcomes remains unproven.
Richard A. Davidson, MD, MPH
University of FloridaGainesville, Florida, USA