Effects of Different Modes of Exercise Training on Glucose Control and Risk Factors for Complications in Type 2 Diabetic Patients

We sought to meta-analyze the effects of different modes of exercise training on measures of glucose control and other risk factors for complications of diabetes. RESEARCH DESIGN AND METHODS: The 27 qualifying studies were controlled trials providing, for each measure, 4-18 estimates for the effect of aerobic training, 2-7 for resistance training, and 1-5 for combined training, with 1,003 type 2 diabetic patients (age 55 +/- 7 years [mean +/- between-study SD]) over 5-104 weeks. The meta-analytic mixed model included main-effect covariates to control for between-study differences in disease severity, sex, total training time, training intensity, and dietary cointervention (13 studies). To interpret magnitudes, effects were standardized after meta-analysis using composite baseline between-subject SD.

Differences among the effects of aerobic, resistance, and combined training on HbA(1c) (A1C) were trivial; for training lasting >/=12 weeks, the overall effect was a small beneficial reduction (A1C 0.8 +/- 0.3% [mean +/- 90% confidence limit]). There were generally small to moderate benefits for other measures of glucose control. For other risk factors, there were either small benefits or effects were trivial or unclear, although combined training was generally superior to aerobic and resistance training. Effects of covariates were generally trivial or unclear, but there were small additional benefits of exercise on glucose control with increased disease severity.