Differential effect of β-blocker therapy on insulin resistance as a function of insulin sensitizer use:: results from GEMINI

Aims To determine whether the beneficial effects of carvedilol on insulin resistance (IR) are affected by the concomitant use of insulin sensitizers [thiazolidinediones (TZDs) and metformin]. Methods Changes in HbA(1c) and homeostasis model assessment-insulin resistance (HOMA-IR) were assessed over 5 months, comparing carvedilol with metoprolol tartrate according to insulin sensitizer (TZDs and metformin) use. Results In TZD/metformin users, carvedilol patients showed a 5.4% decrease [95% confidence interval (CI) -11.9, 1.6; P = 0.13] and metoprolol tartrate patients showed a 2.8% decrease (95% CI -8.5, 3.2; P = 0.35) in HOMA-IR. The -2.6% difference between treatments was not significant (95% CI -10.7, 6.2; P = 0.55). In contrast, those not taking TZD/metformin experienced a 13.2% increase in HOMA-IR on metoprolol tartrate (95% CI 3.2, 24.1; P < 0.01) and a 4.8% decrease in HOMA-IR on carvedilol (95% CI -14.6, 6.0; P = 0.37), with a significant treatment difference of -15.9% favouring carvedilol (95% CI -26.6, -3.6; P = 0.01). There was no significant treatment interaction for the use of TZD/metformin and HbA(1c). A statistically significant treatment difference was observed for HbA(1c) after 5 months favouring carvedilol after adjusting for insulin sensitizer use (-0.11%, 95% CI -0.214, -0.009; P = 0.03). Conclusions In patients with diabetes and hypertension not taking insulin sensitizers, the use of metoprolol tartrate resulted in a worsening of insulin resistance, an effect not seen with carvedilol. However, in TZD/metformin users the difference between the beta-blockers was not statistically significant.