From the American Society of Hypertension
Mark C. Houston, MD
NEW YORK CITY?Drug-induced metabolic changes are known to occur with certain antihypertensive agents (ie, diuretics, beta-blockers), but experts disagree on the clinical implications of these changes.
In one camp are those who argue that these metabolic consequences should be taken into account when choosing initial antihypertensive therapy. In the other camp are those who recommend relying only on clinical outcomes data to make rational choices.
These 2 viewpoints were debated at the 21st Annual Scientific Meeting of the American Society of Hypertension.
According to Mark C. Houston, MD, director of the Hypertension Institute, St. Thomas Hospital, and associate clinical professor of medicine, Vanderbilt University, both in Nashville, the metabolic consequences of diuretics and of nonselective and selective beta-blockers should preclude their use as initial therapy in certain patient populations, such as those at high risk for diabetes.
Barry R. Davis, MD, PhD
"Drug-induced diabetes and native diabetes represent a similar clinical entity and thus carry the same cardiovascular and total mortality risk," Dr Houston observed.
Impaired glucose tolerance, hypokalemia, azotemia, and hyperuricemia are among the potential adverse metabolic effects of thiazide-type diuretics.
Long-term treatment with diuretics (or beta-blockers) can impair glucose tolerance and precipitate overt diabetes mellitus. Evidence based on the Systolic Hypertension in the Elderly Program (SHEP) trial showed that over a 14-year follow-up, the increase in serum glucose with diuretic therapy was substantial and significant, he said.
"When diuretics were stopped, 40% of cases of diabetes mellitus did not revert back to normal, indicating persistence of adverse effects," Dr Houston said. "Long-term treatment with diuretics or beta-blockers can impair glucose tolerance and precipitate overt diabetes mellitus."
A continuum of risk is linked to increasing serum glucose levels between 75 and 110 mg/dL. Over this range, the risk of a cardiovascular (CV) event increases by 33%. Each 1-mg increase in fasting blood glucose is associated with a 1% increase in CV events.
In the Antihypertensive and Lipid-Lowering Treatment to Prevent Heart Attack Trial (ALLHAT), those randomized to the diuretic chlorthalidone (Thalitone) had the highest incidence of new-onset diabetes at 4 years' follow-up. The diuretic was also associated with adverse changes in serum potassium and total cholesterol levels.
Arch Intern Med
In addition, the diabetes induced by diuretic therapy may shorten lives, said Dr Houston. Among 996 patients aged <30 years who had type 1 diabetes and hypertension, those treated with diuretics for 6 years had a 4-fold increase in mortality compared with those who did not receive diuretics (. 1989;149:266-272). And among 1330 patients with type 1 or type 2 diabetes diagnosed after age 30, diuretic therapy increased mortality by 70%.
According to findings from the PIUMA (Progetto Ipertensione Umbria Monitoraggio Ambulatoriale) study, among treated hypertensives, "the occurrence of new-onset type 2 diabetes portends a risk for subsequent cardiovascular disease that is similar to pre-existing diabetes," Dr Houston said.
Other data show that drug-induced hypokalemia increases the likelihood of cardiac arrhythmias, cerebrovascular accidents, thrombosis, and other CV events, Dr Houston argued, and that diuretic-induced hyperuricemia has been identified as an independent risk factor for CV disease.
Finally, he noted, diuretics can accelerate the onset of renal failure in patients with hypertension plus type 1 or type 2 diabetes. Even with successful blood pressure lowering, 15% to 30% of hypertensive patients treated with diuretics and beta-blockers progress to renal insufficiency.
Barry R. Davis, MD, PhD, focused on outcomes generated from controlled, randomized clinical trials, arguing that the relatively small detrimental metabolic effects of thiazide-type diuretics should not affect their role as a preferred agent for the management of hypertension.
"Comparative metabolic consequences have not yet been shown to be important, while we are certain of the importance of clinical events and blood pressure control," said Dr Davis, director, Division of Biostatistics, and director, Coordinating Center for Clinical Trials, University of Texas School of Public Health, Houston.
Using data from ALLHAT, he showed that despite metabolic changes, chlorthalidone reduced the risk of heart failure by 38% compared with amlodipine (Norvasc) and by 19% compared with lisinopril (Prinivil). The incidence of stroke was also significantly reduced (by 15%) with chlorthalidone com-pared with the metabolically neutral lisinopril.
The main outcome of total mortality was similar in the chlorthalidone, amlodipine, and lisinopril groups.
"Overall, metabolic differences did not translate into more adverse cardiovascular events or into higher all-cause mortality with chlorthalidone," said Dr Davis.
Among participants in the SHEP trial who were followed for >14 years, active treatment with a diuretic decreased risk of total and CV-related mortality compared with placebo.
Total mortality and CV mortality rates were similar in the active treatment patients who did and did not develop diabetes during the extension phase of SHEP. "The milder long-term course of diabetes that occurred during diuretic therapy is likely related to a lesser degree of metabolic disturbance," Dr Davis said. "Medication-induced metabolic changes may not have the same implications as naturally occurring metabolic changes."
Dr Houston countered that the follow-up periods in clinical trials have not been adequate enough to demonstrate differences in event rates between patients treated with agents that cause metabolic disturbances and those treated with metabolically neutral drugs.
The decision to use diuretics and beta-blockers for hypertension is a subject of heated debate because of their associated metabolic changes.
Thiazide-type diuretics have been linked to impaired glucose tolerance, hypokalemia, azotemia, and hyperuricemia.
Some experts believe that these drug classes should not be used in patients at high risk for diabetes.
Others counter that thiazide-type diuretics have relatively minor detrimental metabolic effects that should not affect their status as a preferred treatment.