Association Between Physical Activity and Risk of All-Cause Mortality and Cardiovascular Disease in Patients With Diabetes: A meta-analysis.

OBJECTIVE The association between habitual physical activity (PA) and lowered risk of all-cause mortality (ACM) and cardiovascular disease (CVD) has been suggested in patients with diabetes. This meta-analysis summarizes the risk reduction in relation to PA, focusing on clarifying dose-response associations. RESEARCH DESIGN AND METHODS Electronic literature searches were conducted for cohort studies that examined relative risk (RR) of ACM or CVD in relation to PA in patients with diabetes. For the qualitative assessment, RR for the highest versus the lowest PA category in each study was pooled with a random-effects model. We added linear and spline regression analyses to assess the quantitative relationship between increases in PA and ACM and CVD risk. RESULTS There were 17 eligible studies. Qualitatively, the highest PA category had a lower RR [95% CI] for ACM (0.61 [0.52-0.70]) and CVD (0.71 [0.60-0.84]) than the lowest PA category. The linear regression model indicated a high goodness of fit for the risk of ACM (adjusted R(2) = 0.44, P = 0.001) and CVD (adjusted R(2) = 0.51, P = 0.001), with the result that a 1 MET-h/day incrementally higher PA was associated with 9.5% (5.0-13.8%) and 7.9% (4.3-11.4%) reductions in ACM and CVD risk, respectively. The spline regression model was not significantly different from the linear model in goodness of fit (P = 0.14 for ACM risk; P = 0.60 for CVD risk). CONCLUSIONS More PA was associated with a larger reduction in future ACM and CVD risk in patients with diabetes. Nevertheless, any amount of habitual PA was better than inactivity.