Background: High rates of leisure activity have been associated with reduced risk of Alzheimer disease (AD).
Objective: To determine whether prediagnosis leisure activity modifies the rate of cognitive decline in patients with AD.
Design: Inception cohort followed up longitudinally for a mean of 5.3 years (up to 13.9 years).
Setting: Urban community.
Participants: A total of 283 patients with incident AD (mean age, 79 years; 56.2% Hispanic and 31.1% African American).
Main outcome measures: Change in a composite cognitive score from diagnosis on and during the entire study follow-up.
Results: In multivariate-adjusted generalized estimating equation models of postdiagnosis change (n = 133), each leisure activity was associated with an additional yearly decline of 0.005 of a z-score unit in cognitive score (P = .17). In models expanded to include cognitive change during study follow-up, including evaluations before and after diagnosis (n = 283), each activity was associated with an additional yearly decline of 0.005 of a z-score unit in cognitive score (P = .03). The association was strongest for intellectual activities.
Conclusions: Greater participation in prediagnosis leisure activities, especially intellectual activities, was associated with faster cognitive decline, supporting the hypothesis that the disease course in AD may vary as a function of cognitive reserve.