Effects of antihypertensive therapy on glucose and insulin metabolism and on left ventricular mass: A randomized, double-blind, controlled study of 21 obese hypertensives.

ABSTRACT

BACKGROUND: Glucose and insulin levels are associated with left ventricular mass (LVM) in insulin-resistant individuals. Antihypertensive drugs have different effects on glucose and insulin metabolism (GIM) and on LVM. To evaluate whether the effects of antihypertensive therapy on LVM are associated with its effects on GIM, we compared the effects of atenolol and perindopril on these parameters in a group of insulin-resistant, obese hypertensives. METHODS AND RESULTS: A total of 21 obese, nondiabetic hypertensives who were aged 55+/-12 years, had a body mass index of 32.8+/-5.0 kg/m(2), were free of coronary or valvular heart disease, and had normal LV function were randomized to treatment with atenolol (n=11) or perindopril (n=10). Echocardiographic LVM corrected for height (LVM/height) and GIM (3-hour intravenous glucose tolerance test) were measured after 4 to 6 weeks of washout and 6 months of treatment. Baseline characteristics were similar in both groups. Atenolol and perindopril effectively reduced blood pressure (from 149+/-13/98+/-4 to 127+/-8/82+/-6 mm Hg and from 148+/-9/98+/-4 to 129+/-9/82+/-6 mm Hg, respectively, for the atenolol and perindopril groups; P=0.002). Atenolol significantly worsened GIM parameters, fasting glucose levels (5.3+/-0.9 to 6.0+/-1.5 mmol/L; P=0.003), fasting insulin levels (121+/-121 to 189+/-228 pmol/L; P=0.03), and most other relevant metabolic measures (P<0.05 for all). Perindopril did not affect GIM. Atenolol did not affect LVM/height (119+/-12 to 120+/-17 g/m; P=0.8), whereas perindopril significantly reduced LVM/height (120+/-13 to 111+/-19 g/m; P=0.04). CONCLUSIONS: In obese, hypertensive individuals, adequate and similar blood pressure control was achieved with perindopril and atenolol. However, perindopril but not atenolol was associated with a more favorable GIM profile and led to a significant regression of LVM.