[Skip to Content]
[Skip to Content Landing]
February 26, 1992

Diuretics, Serum and Intracellular Electrolyte Levels, and Ventricular Arrhythmias in Hypertensive Men

Author Affiliations

From the Departments of Epidemiology and Biostatistics (Drs Siegel, Hulley, Black, and Hearst and Ms Seeley), Medicine (Drs Siegel, Hulley, Cheitlin, Sebastian, and Fine), and Family and Community Medicine (Dr Hearst), University of California, San Francisco.

JAMA. 1992;267(8):1083-1089. doi:10.1001/jama.1992.03480080053026

Objective.  —To investigate the patterns of electrolyte abnormalities resulting from thiazide administration and whether they cause ventricular arrhythmias, and to help resolve the controversy over whether clinicians should routinely prescribe potassium-conserving therapy to all patients treated with thiazides.

Design.  —Double-blind, randomized controlled trial.

Participants.  —A total of 233 hypertensive men aged 35 to 70 years.

Interventions.  —Participants were withdrawn from prior diuretic treatment and were replenished with oral potassium chloride and magnesium oxide. They were then randomized to 2 months of treatment with (1) hydrochlorothiazide; (2) hydrochlorothiazide with oral potassium; (3) hydrochlorothiazide with oral potassium and magnesium; (4) hydrochlorothiazide and triamterene; (5) chlorthalidone; or (6) placebo.

Main Outcome Measures.  —Ventricular arrhythmias on 24-hour Holter monitoring and serum and intracellular potassium and magnesium levels.

Results.  —Of the 233 participants, 212(91%) completed the study. Serum potassium levels were 0.4 mmol/L lower in the hydrochlorothiazide group than in the placebo group (P<.01), and this mean difference was not affected by supplementation with potassium, with potassium and magnesium, or with triamterene. However, the supplements did prevent the occasional occurrence of marked hypokalemia; all 12 of the men who developed serum potassium levels of 3.0 mmol/L or less were among the 90 who received diuretics without supplementation (P<.01). Similarly, the overall proportion of men with ventricular arrhythmias was not affected by randomized treatment, but there was a twofold increase in the proportion with arrhythmias among the 12 men with serum potassium levels of 3.0 mmol/L or less (P =.02). Serum magnesium and intracellular potassium and magnesium levels were not reduced by hydrochlorothiazide, nor were they related to ventricular arrhythmias.

Conclusions.  —In the majority of hypertensive patients, treatment with 50 mg/d of hydrochlorothiazide does not cause marked hypokalemia or ventricular arrhythmias. However, because some individuals will develop hypokalemia after starting diuretic therapy, serum potassium levels should be monitored and potassiumsparing strategies should be used when indicated.(JAMA. 1992;267:1083-1089)