Thongtang et al compared the effect of atorvastatin to rosuvastatin at maximum equivalent doses and demonstrated a significant increase in insulin levels (5.2% and 8.7% from baseline, respectively), although these did not differ between groups after 6 weeks. The 2 drugs had different effects on glycated albumin: an increase (0.8%) after atorvastatin and a decrease (0.7%) after rosuvastatin (p = 0.002).

We compared the effect of rosuvastatin 10 mg (n = 18) to atorvastatin 20 mg (n = 18) on glucose metabolism in nondiabetic patients with dyslipidemia. Rosuvastatin caused a significant decrease in insulin levels and homeostasis model assessment–insulin resistance (HOMA-IR) levels and a significant increase in the quantitative insulin-sensitivity check index at 12 weeks of treatment compared to baseline and to atorvastatin. This difference was not seen at 4 weeks. Rosuvastatin compared to atorvastatin decreased insulin by −35% versus +17% (p = 0.032), decreased HOMA-IR by −33% versus +17% (p = 0.008), and increased the quantitative insulin sensitivity check index by +11% versus 6% (p = 0.015) at 12 weeks of treatment. Differences in insulin and HOMA-IR were significant (p = 0.012 and 0.007, respectively) after adjustment for age, gender, and change in waist circumference, adiponectin, leptin, and triglyceride levels. No difference was observed in fasting plasma glucose, hemoglobin A _1c , leptin, and adiponectin levels after treatment in either group.

A main difference between our study and that of Thongtang et al is that we assessed the effect of statins on insulin resistance indexes (not only insulin levels) at 2 intervals. The effect of rosuvastatin on glucose homeostasis was evident only at 12 weeks. Another study compared the effects of these 2 statins at similar doses on glucose metabolism and showed no change in HOMA-IR after 8 weeks of treatment with either statin. Atorvastatin led to a slight but significant increase in hemoglobin A _1c levels (p = 0.03).

The impact of statins on glucose metabolism and risk of new-onset diabetes remains controversial. Two recent meta-analyses evaluating the risk of new-onset diabetes related to statin treatment yielded conflicting results. Data regarding the effect of statins on insulin sensitivity are also inconclusive. We speculate that the different doses of statins, small number of patients, heterogeneity of these studies, and their short durations are the main reasons for these disagreements. Clearly, larger trials are needed.