Cognitive dispersion and depressive symptoms in aging: distinct contributions to longitudinal decline

Depression is a noted risk factor for general cognitive decline and Alzheimer's disease and related dementias. Evidence suggests that this may be due to its association with greater variability in performance across cognitive tasks, known as cognitive dispersion. The current study examines how depression and dispersion uniquely predict cognitive decline in a longitudinal study of cognitively normal (CN) older adults.

Data were drawn from the Harvard Aging Brain Study. Participants were 287 individuals who were CN at baseline, contributing 1,547 observations across six longitudinal waves over 8.68 years. Depression was assessed using the Geriatric Depression Scale. Cognitive status was assessed via the Preclinical Alzheimer's Cognitive Composite. Cognitive dispersion was defined as the intraindividual standard deviation of scores across five cognitive tasks that represent complementary dimensions of executive functioning and cognitive control.

Cross-sectional mixed-effects models revealed that higher depression was associated with lower baseline cognitive performance (β = -0.02, SE = 0.01, t(1482) = -4.01, p < .001), but depression was not related to dispersion. In combined longitudinal models including both depression and dispersion, higher average dispersion (between-persons) predicted lower cognitive performance (β = -0.38, SE = 0.11, t(272) = -3.45, p < .001) and faster decline (β = -0.07, SE = 0.03, t(258)=-2.34, p = .02). Depression was not a significant predictor of cognitive decline when modeled with dispersion.

Our findings highlight the importance of monitoring depression during clinic visits, as it is significantly related to concurrent cognitive performance in aging populations. Since dispersion and not depression predicted decline over time, cognitive dispersion may serve as a specific marker of future impairment risk.