Association of Thiazide-Type Diuretics With Glycemic Changes in Hypertensive Patients: A Systematic Review and Meta-Analysis of Randomized Controlled Clinical Trials

The effect of different classes of antihypertensive drugs on incident diabetes mellitus is controversial because traditional meta-analyses are hindered by heterogeneity across trials and the absence of trials comparing angiotensin-converting-enzyme (ACE) inhibitors with angiotensin-receptor blockers (ARB). We therefore undertook a network meta-analysis, which accounts for both direct and indirect comparisons to assess the effects of antihypertensive agents on incident diabetes. We undertook a systematic review up to Sept 15, 2006, and identified 48 randomised groups of 22 clinical trials with 143,153 participants who did not have diabetes at randomisation and so were eligible for inclusion in our analysis. 17 trials enrolled patients with hypertension, three enrolled high-risk patients, and one enrolled those with heart failure. The main outcome was the proportion of patients who developed diabetes. Initial drug therapy used in the trials (and the number of patients with diabetes of the total number at risk) included: an ARB (1189 of 14,185, or 8.38%), ACE inhibitor (1618 of 22,941, or 7.05%), calcium-channel blocker (CCB, 2791 of 38,607, or 7.23%), placebo (1686 of 24,767, or 6.81%), beta blocker (2705 of 35,745, or 7.57%), or diuretic (998 of 18,699, or 5.34%). With an initial diuretic as the standard of comparison (eight groups), the degree of incoherence (a measure of how closely the entire network fits together) was small (omega=0.000017, eight degrees of freedom). The odds ratios were: ARB (five groups) 0.57 (95% CI 0.46-0.72, p<0.0001); ACE inhibitor (eight groups) 0.67 (0.56-0.80, p<0.0001); CCB (nine groups): 0.75 (0.62-0.90, p=0.002); placebo (nine groups) 0.77 (0.63-0.94, p = 0.009); beta blocker (nine groups) 0.90 (0.75-1.09, p=0.30). These estimates changed little in many sensitivity analyses. The association of antihypertensive drugs with incident diabetes is therefore lowest for ARB and ACE inhibitors followed by CCB and placebo, beta blockers and diuretics in rank order.

It has been suggested that the metabolic side effects of antihypertensive drugs are responsible for their failure to reduce cardiovascular morbidity in patients with hypertension. Therefore, in 50 patients with essential hypertension, we performed a randomized, double-blind, crossover study comparing the effects of carbohydrate and lipid metabolism of captopril (mean [+/- SD] dose, 81 +/- 24 mg per day) and hydrochlorothiazide (40 +/- 12 mg per day) over two four-month treatment periods. Captopril increased the insulin-mediated disposal of glucose, as compared with placebo, from 5.7 +/- 2.4 to 6.3 +/- 2.5 mg per kilogram of body weight per minute (P less than 0.05), whereas hydrochlorothiazide caused a decrease from 6.4 +/- 2.0 to 5.7 +/- 1.9 (P less than 0.01). Captopril had no effect on the basal insulin concentration, but it decreased the late (30- to 90-minute) insulin response to glucose and increased the early (2- to 6-minute) insulin peak. Hydrochlorothiazide increased the basal insulin concentration and the late insulin response to glucose. These findings may be explained by an increase in insulin sensitivity with captopril and a decrease with hydrochlorothiazide. Little or no change was seen in serum lipid or lipoprotein levels during treatment with captopril, whereas hydrochlorothiazide caused significant increases in serum total (5 percent) and low-density lipoprotein (6 percent) cholesterol levels and total (15 percent) and very-low-density lipoprotein (25 percent) triglyceride levels, as compared with placebo (P less than 0.01 for all comparisons). We conclude that hydrochlorothiazide for the treatment of essential hypertension has adverse effects on glucose and lipid metabolism. It is possible, but not proved in this study, that these changes may contribute to the risk for diabetes mellitus and coronary heart disease. In contrast, captopril appears to have beneficial or no effects on glucose and lipid metabolism.

Diuretic-based antihypertensive therapy is associated with the development of diabetes but with improved clinical outcomes. It has been proposed that the duration of clinical trials has been too short to detect the adverse effects of diabetes. We assessed the long-term mortality rate of subjects in the Systolic Hypertension in the Elderly Program (n = 4,732) who were randomized to stepped-care therapy with 12.5 to 25.0 mg/day of chlorthalidone or matching placebo. If blood pressure remained above the goal, atenolol or matching placebo was added. At a mean follow-up of 14.3 years, cardiovascular (CV) mortality rate was significantly lower in the chlorthalidone group (19%) than in the placebo group (22%; adjusted hazard ratio [HR] 0.854, 95% confidence interval [CI] 0.751 to 0.972). Diabetes at baseline (n = 799) was associated with increased CV mortality rate (adjusted HR 1.659, 95% CI 1.413 to 1.949) and total mortality rate (adjusted HR 1.510, 95% CI 1.347 to 1.693). Diabetes that developed during the trial among subjects on placebo (n = 169) was also associated with increased CV adverse outcome (adjusted HR 1.562, 95% CI 1.117 to 2.184) and total mortality rate (adjusted HR 1.348, 95% CI 1.051 to 1.727). However, diabetes that developed among subjects during diuretic therapy (n = 258) did not have significant associations with CV mortality rate (adjusted HR 1.043, 95% CI 0.745 to 1.459) or total mortality rate (adjusted HR 1.151, 95% CI 0.925 to 1.433). Diuretic treatment in subjects who had diabetes was strongly associated with lower long-term CV mortality rate (adjusted HR 0.688, 95% CI 0.526 to 0.848) and total mortality rate (adjusted HR 0.805, 95% CI 0.680 to 0.952). Thus, chlorthalidone-based treatment improved long-term outcomes, especially among subjects who had diabetes. Subjects who had diabetes associated with chlorthalidone had no significant increase in CV events and had a better prognosis than did those who had preexisting diabetes.