Abstract
Background
Limited evidence suggests that higher body weight may be associated with improved survival among individuals with dementia relative to those in the normal weight range. The mechanisms underlying this association are uncertain and may include both physiological and methodological factors.
Objective
To examine the association between weight status at incident dementia and mortality and to assess potential sources of bias.
Methods
We examined the association between weight status at incident dementia and mortality in a large, U.S. population-based sample with extended follow-up. We also evaluated potential sources of bias affecting this association, including confounding due to weight loss, collider stratification bias, and survivor bias.
Results
Compared with the normal weight category, class I obesity at incident dementia was associated with a 27% lower risk of mortality (HR 0.73, 95% CI: 0.62–0.86). Overweight and class II/III obesity were also associated with lower mortality. Analyses evaluating potential sources of bias did not support these mechanisms as major contributors to the observed associations.
Conclusions
In a large, U.S. population-based sample with extended follow-up, overweight and obesity at incident dementia were associated with lower mortality relative to normal weight. While the underlying mechanisms are not fully understood, these findings suggest that weight status at diagnosis may serve as a useful prognostic indicator alongside other factors.