Joint analysis of growth curve model in depressive symptoms and mortality: application to an elderly study

The primary aim in this study was to use a joint analysis approach to examine the association between longitudinal change of depressive symptoms and mortality in older Mexican Americans. The joint model was applied to data from the Hispanic Established Population for Epidemiological Study of the Elderly (HEPESE), a seven year longitudinal study of communitydwelling elderly Mexican-Americans. Depressive symptoms were measured by the Center of Epidemiological Studies Depression Scale (CES-D). The trajectories from two stage model and joint model using change in CES-D, modeled by longitudinal linear, and nonlinear curvatures including quadratic and exponential growth curve, were used as predictors to fit the survival curve. The joint analysis indicated that CES-D score at baseline and exponential curve were significant predictive factors of mortality and trajectories of linear and quadratic change in CES-D score were not associated with mortality in older Mexican-Americans.