Body Mass Index and Mortality in Elderly Subjects from the Moli-Sani Study: A Possible Mediation by Low-Grade...

The association between obesity and mortality in the elderly remains controversial. To test the association between BMI and mortality, with the hypothesis of a mediation by low-grade inflammation (LGI), a prospective study design (median follow-up 7.8 years) was used on a sample of 4,970 elderly subjects (age ≥ 65 years) from the Moli-sani Study cohort. The association between BMI categories and overall or cause-specific mortality (hazard ratio, HR) was calculated by multivariable Cox regression. Dose-response relationship was tested using restricted cubic splines. Interaction between BMI and LGI, assessed through high-sensitivity C-reactive protein (hs-CRP) and INFLA-score, was also tested. In comparison with normal-weight, overweight was significantly associated with a 20% (adjusted HR = 0.80; 95%CI 0.67-0.95) reduced risk of total mortality, while severe obesity (BMI > 40) with an increased risk (HR = 1.81; 95%CI 1.13-2.93). Cubic spline curves showed a U-shaped relationship between BMI and total mortality (p value for nonlinear relationship = 0.001). Similar results were found for cardio-cerebrovascular and other causes mortality. Hs-CRP and INFLA-score were associated with an increased risk of total mortality in adjusted analyses. Mediation analysis did not show any effect of LGI on the association between BMI and mortality. However, after stratification for LGI under or below the population median, greater LGI increased the risk of mortality in obese elderly more than expected (p for interaction = 0.04). A U-shaped association between BMI and mortality was observed in Italian elderly subjects. While the association was independent of LGI levels, there was a significant interaction between BMI and LGI in increasing mortality risk in obese elderly individuals.