The Antihypertensive and Lipid-Lowering Treatment to Prevent Heart Attack Trial (ALLHAT) was a strongly powered, randomized, double-blind, multicenter clinical trial comparing the efficacies of various antihypertensive treatments in decreasing rates of coronary heart disease (CHD). The anithypertensives used in the trial included amlodipine, a calcium-channel blocker; lisinopril, an angiotensin-converting enzyme (ACE) inhibitor; doxazosin, an alpha blocker; and chlorthalidone, a thiazide diuretic. The study showed similar mortality rates and no improvement in CHD risk in all treatment groups. Chlorthalidone was found to be superior in preventing one or more major forms of cardiovascular disease, including stroke and heart failure. These findings resulted in the authors recommending that diuretics be the initial drug of choice and a part of any multidrug antihypertensive regimen. However, many questions have been raised about the study design and the interpretation of the results. Davis and associates reviewed the results of the ALLHAT study and the questions that have arisen.

The authors respond to questions about the ALLHAT study design by pointing out the open forum in which the design was developed. The use of intent-to-treat analyses supports the study’s conclusions and shows the lack of impact of drop-outs. Clinical outcomes were carefully evaluated by a central subcommittee, and all study activities were double-blinded.

Although the rates of primary outcomes, including fatal CHD and nonfatal myocardial infarction, were the same in the chlorthalidone, amlodipine, and lisinopril groups, the study’s secondary endpoints such as heart failure showed the superiority of diuretics, specifically chlorthalidone. The diuretic users had a small increase in serum glucose levels, but this increase did not affect adverse outcomes. Thiazide-associated diabetes also has been shown to be reversible with good potassium balance, weight control, and increased physical activity. Thiazide-based treatments for hypertension are less expensive, and the additional laboratory monitoring for hypokalemia or hypoglycemia is probably no more than that required during ACE-inhibitor administration.

Results from the Second Australian National Blood Pressure Study (ANBP2), which compared diuretic-based antihypertension therapy with ACE-inhibitor–based regimens, showed slightly more benefit with ACE inhibitors in the reduction of all cardiovascular events and all-cause mortality. This seeming contradiction with ALLHAT results may be explained by the open-labeled nature of the ANBP2, which may have resulted in bias and a lower frequency of cardiovascular endpoints in that study. Even though the conclusions are different, the upper confidence limit for relative risks in ANBP2 allow for the possibility that ALLHAT results are valid.

The authors conclude that the agent used to lower blood pressure appears to matter because of blood pressure–independent effects that may be unique to each class of antihypertensive agents. Although ALLHAT has legitimate limitations, its conclusions appear valid (see accompanying table). The newer antihypertensive agents tested are not superior to diuretics in the prevention of cardiovascular disease. These newer drugs are more expensive and appear to be less effective in preventing heart failure. Diuretics appear to be the preferred first-step drug and an important part of any multidrug regimen for the treatment of hypertension.