The association of age with rate of cognitive decline in elderly individuals residing in supporting care facilities

Objectives: This study examines the effect of age on rate of cognitive decline in different stages of dementia, of nursing home and assisted-living residents. Methods: In this longitudinal study, the Mini Mental State Examination (MMSE) was used to measure rate of cognitive decline in subjects who were nondemented [Clinical Dementia Rating (CDR)=0; n=353], questionably demented (CDR=0.5; n=121), or frankly demented (CDR1; n=213) at baseline. RESULTS: A generalized estimating equation was used to model the MMSE scores over time (mean follow-up 2.9±2.0 y). The generalized estimating equation model had the MMSE scores at successive follow-up time points as dependent variables and had linear and quadratic age, follow-up time from baseline, CDR at baseline, and all the interactions among them as independent variables, controlling for MMSE at baseline, sex, race, and education. The mean age of the entire sample was 85.2±7.4 years at baseline. There were no significant interactions of linear age effects with rate of cognitive decline. The analysis of interaction of quadratic age with rate of cognitive decline showed complex relationships: in the nondemented group, there was no substantial quadratic association of age with the rate of cognitive decline (P=0.13); in the questionable demented group, the oldest subjects declined relatively faster (P=0.02); and in the demented group, the youngest and oldest subjects tended to decline relatively less than subjects in the intermediate ages (P=0.07). Conclusions: This study adds an additional aspect to the complexity of the association between age and rate of cognitive decline, showing that the direction and amplitude of this effect differs according to the stage along the course of cognitive decline.