Associations Among Depression, Stress, Resilience, and Alzheimer’s Disease Biomarkers in Older Adults Following a Major Societal Stressor

Abstract

AbstractBackgroundBetter predicting depression following a stressor like COVID‐19 would help identify vulnerable individuals who might benefit the most from early intervention. We examined change in depressive symptoms relative to the onset of COVID‐19 and longitudinal trajectories of depressive symptoms post‐pandemic‐onset. As a secondary objective, we examined associations of these changes with stress, resilience, and in vivo AD biomarkers.MethodData were analyzed from 101 older adults (cognitively unimpaired at baseline) from the Harvard Aging Brain Study, a longitudinal observational cohort study. Depressive symptoms are assessed annually with the 30‐item Geriatric Depression Scale (GDS), and for these analyses, participants had no/mild self‐reported depression (GDS<11) at the last pre‐COVID examination. GDS‐delta scores were computed using the mean of the post‐COVID‐onset scores subtracted by the last pre‐COVID examination for participants. Levels of resilience (Connor‐Davidson Resilience scale) and stress (Perceived Stress Scale) were assessed within a year after pandemic onset. Cortical amyloid (PiB) and entorhinal (EC) and inferior temporal (IT) tau (FTP) were analyzed using pre‐pandemic acquisition (mean 1.6 years to pandemic onset). We used linear models to predict GDS‐delta and linear mixed effects models to predict GDS trajectory post‐COVID onset from resilience, stress, and the interaction of each with AD biomarkers (PiB and FTP).ResultsGDS‐delta was predicted by resilience (t = ‐4.69, p<0.001) and stress (t = 3.51, p<0.001) such that those with lower resilience and higher stress experienced higher GDS‐delta. EC tau, but not IT tau or cortical amyloid, interacted with stress to predict GDS‐deltas (EC tau*stress: t = 2.04; p = 0.04) indicating that those with higher levels of EC tau and stress experienced higher GDS‐delta. For post‐pandemic‐onset GDS trajectory, no significant longitudinal associations were observed, but there was a marginal interaction between EC‐tau*stress in predicting greater GDS trajectories (t = ‐1.79, p = 0.08).ConclusionResults suggest that change in depressive symptoms in elders following a societal stressor is associated with lower resilience and higher stress; this may be exacerbated by concurrent EC tau pathology. Moreover, elevated depressive symptoms in the years following onset of a societal stressor can be partially predicted by increased EC tau and stress, highlighting a vulnerable population who may benefit from preventive interventions throughout a major stressor.