Results:

Annual incidence rates for type 2 diabetes mellitus were 2.1% in the pioglitazone group and 7.6% in the placebo group, and the hazard ratio for conversion to diabetes in the pioglitazone group was 0.28 (95% confidence interval, 0.16-0.49; p < 0.001). Conversion to normal glucose tolerance occurred in 48% of the patients in the pioglitazone group and 28% of those in the placebo group (p < 0.001). Treatment with pioglitazone as compared with placebo was associated with significantly reduced levels of fasting glucose (a decrease of 11.7 mg/dl vs. 8.1 mg/dl [0.7 mmol/L vs. 0.5 mmol/L], p < 0.001), 2-hour glucose (a decrease of 30.5 mg/dl vs. 15.6 mg/dl [1.6 mmol/L vs. 0.9 mmol/L], p < 0.001), and glycated hemoglobin (a decrease of 0.04 percentage points vs. an increase of 0.20 percentage points, p < 0.001). Pioglitazone therapy was also associated with a decrease in diastolic blood pressure (by 2.0 mm Hg vs. 0.0 mm Hg, p = 0.03), a reduced rate of carotid intima-media thickening (31.5%, p = 0.047), and a greater increase in the level of high-density lipoprotein cholesterol (by 7.35 mg/dl vs. 4.5 mg/dl [0.4 mmol/L vs. 0.3 mmol/L], p = 0.008). Weight gain was greater with pioglitazone than with placebo (3.9 kg vs. 0.77 kg, p < 0.001), and edema was more frequent (12.9% vs. 6.4%, p = 0.007).