Obesity is associated with lower mortality risk in elderly diabetic subjects: the Casale Monferrato study

The relationship between obesity and mortality in people with type 2 diabetes has not been definitely assessed. We have examined this issue in a well-characterized population-based cohort of Mediterranean diabetic people. Standardized anthropometric data from the population-based Casale Monferrato Study have been prospectively analyzed. The cohort included 1,475 people (62.6% aged a parts per thousand yen65 years) who had been recruited in 1991 and followed-up to December 31, 2006. Cox proportional hazards modeling was employed to estimate the independent associations between all-cause and cardiovascular mortality and BMI. Out of 1,475 people, 972 deaths occurred during a 15-year follow-up. Cox regression analyses showed that with respect to BMI < 24.2 kg/m(2), values of 30.0 kg/m(2) and over were associated with lower all-cause and cardiovascular mortality risk (HR = 0.68, 95% CI 0.56-0.85, P for trend = 0.001; HR = 0.59, 0.44-0.80, P for trend = 0.002), independently of classical and new risk factors. As interaction between age and BMI was significant, we performed a stratified analysis by age, providing evidence that our finding was entirely due to a significant protective effect of BMI of 30.0 kg/m(2) and over in the elderly (all-cause mortality HR = 0.75, 95% CI 0.58-0.96; cardiovascular mortality HR = 0.67, 95% CI 0.45-0.95). In contrast, obesity was not significantly associated with mortality risk in diabetic subjects aged < 65 years. Results were confirmed even excluding from the analysis individuals who died within 2 years of follow-up, smokers and those with CHD. In Mediterranean diabetic people aged a parts per thousand yen65 years, obesity is significantly associated with lower 15-year mortality risk. In contrast, it was not significantly associated with mortality risk in diabetic subjects aged < 65 years. As more than two-thirds of people with type 2 diabetes are elderly, our findings, if confirmed, could have clinical implications.