The association of weight change in Alzheimer's disease with severity of disease and mortality: a longitudinal analysis

Abstract

Objective: To describe weight loss in relation to the severity of Alzheimer's disease (AD), to identify confounders that might account for weight loss, and to investigate the association of weight change with mortality.

Methods: This study included 666 subjects with AD from the Consortium to Establish a Registry for Alzheimer's Disease, a longitudinal study of subjects with AD studied for < or = 6 years. Body weight was measured on entry and at annual follow-up examinations as part of standardized clinical assessments. Hierarchical linear models were used to estimate weight change per change in stage of AD, as measured by the Clinical Dementia Rating scale, and to assess the impact of possible confounders on the tendency for weight loss. Linear correlation of the trajectories of weight and CDR over time were used to determine the association between weight change and stage change. The association between weight change and mortality was modeled using the Cox proportional hazards model.

Results: Each change in stage of AD was associated with an estimated average weight loss of nearly 2 lb (P = .0001). The correlation between change in stage of AD and weight change was statistically significant (r = -0.09, P = .005), which indicates a greater tendency toward weight loss with progression of AD. Weight loss of > or = 5% in any year before death was a significant predictor of mortality. When weight loss in the year before death or censoring was analyzed as a continuous variable, it was a significant predictor of mortality. Additionally, the relative risk of death increased as weight loss increased. In contrast, weight gain decreased the risk of mortality.

Conclusion: The risk of weight loss tends to increase with severity and progression of AD. Weight loss is a predictor of mortality among subjects with AD, whereas weight gain appears to have a protective effect.