Diuretic compounds are generally well tolerated, and two large prospective controlled series reported only a 3% withdrawal rate for adverse effects in subjects randomized to diuretic therapy. However, questions have been raised by the observation that BP-lowering therapy in large clinical trials has significantly reduced pressure-related complications (stroke, congestive heart failure, renal impairment, and left ventricular hypertrophy). Although there were improvements in coronary heart disease outcomes (angina, fatal and nonfatal myocardial infarction), these were somewhat less than predicted by epidemiologic data. Some investigators believe that metabolic changes from diuretics may account for this shortfall. Several explanations have been advanced to account for this phenomenon.
First, atherosclerotic coronary heart disease is a longterm multifactorial process whose outcome may not be affected by short-term manipulation of a single risk factor such as high BP. It is possible that the short duration of the diuretic-based treatment trials compared to the longer duration of epidemiologic follow-up explains an apparent shortfall in coronary heart disease benefit. For instance, a 12/4–5 mmHg reduction in BP predicts a decrease of approximately 20 to 25% in coronary heart diseasen events over 10+ years.
In the three- to five-year clinical trials, a decrease in BP of this magnitude led to only16% fewer coronary events. Newer studies, however, indicate that longer-term BP reduction improves coronary event rates to a greater degree than that seen in shorter-term trials. In the ALLHAT (Anti-hypertensive and Lipid-Lowering Treatment to Prevent Heart Attack Trial) study, for example—which compared a diuretic, angiotensin-converting enzyme inhibitor, and calcium channel blocker—thiazide diuretics reduced coronary events to the same extent as competing agents.
The metabolic side effects of thiazide diuretics (electrolyte, lipid, and glucose changes) were originally advanced
as reasons for the less than dramatic reduction of CHD events in the early studies The claim was that these may negate some of the advantages of BP lowering. These questions have been addressed elsewhere. Specific components of the “metabolic debate” are reviewed in material following. In general, the results of recent trials such as ALLHAT have negated these speculations.
Diuretic-associated hypokalemia is dose related. Up to one-third of patients show a decrease in serum potassium, ranging from 0.5 to 0.8 mEq/L when given high doses of hydrochlorothiazide (50 to 100 mg daily). The use of an intermediate dose of 25 mg hydrochlorothiazide results in a decrease in potassium of 0.3 to 0.4 mEq/L, with minimal effects seen at 12.5 or 6.25 mg daily. Potassiumsparing diuretics may be advisable in patients who are especially sensitive to hypokalemia—particularly in older individuals, those taking digitalis preparations, or diabetic patients whose insulin utilization may be affected by hypokalemia.
Hypokalemia can also be minimized by using combination agents such as spironolactone/thiazide or an angiotensin converting enzyme inhibitor or angiotensin-2 receptor blocker with thiazide.The potassium-conserving tendencies of the former medications serve to offset potassium loss caused by the latter. However, diuretic-induced hypokalemia with presently recommended doses is not infrequently of clinical significance.
One report argued that thiazide-related hypokalemia could provoke increased ventricular ectopy, including
ventricular tachycardia and sudden death. In the Multiple Risk Factor Intervention Trial (MRFIT), subjects with
abnormal electrocardiograms in the special care higher dose diuretic group had a higher coronary heart disease
mortality compared to usual care patients who received lower doses of diuretics.This finding, however, was likely a statistical anomaly. Retrospective non-randomized case-control studies questioned the increased incidence of sudden death in diuretic-treated patients, yet comparative groups were not adequately matched in these series.