Trajectories of depressive symptoms in older adults: Correlates and consequences for mortality

INTRODUCTION: Depression in old age often has a poor clinical course, although there is substantial variability in depressive symptom trajectories. We aimed to characterise old-age depressive symptom trajectories, assess their multifactorial correlates, and their impact on mortality. METHODS: We used cohort data from 2118 dementia-free community-dwelling adults aged ≥60 years participating in the Swedish National study on Aging and Care in Kungsholmen (SNAC-K). Depressive symptoms were evaluated using the Montgomery-Åsberg Depression Rating Scale at baseline and 1-3 follow-ups over 9 years. Sociodemographic (age, sex, marital status, socioeconomic status), psychosocial (social connections and support), clinical (multimorbidity), and functional (gait speed; cognitive impairment) factors at baseline were considered as trajectory correlates. Generalised growth mixture models and multinomial logit models estimated depression trajectories and their correlates. Cox proportional hazard models estimated all-cause mortality risk. RESULTS: Three trajectory classes emerged: low, increasing, and U-shaped trajectories of depressive symptoms. Compared to the low trajectory, socioeconomic status involving high financial strain and poor social support was more common in the increasing and U-shaped trajectories. Slow gait speed was linked to higher odds of increasing depressive symptoms, while greater multimorbidity was associated with the U-shaped trajectory. The increasing and U-shaped trajectories were associated with higher mortality risk, and the association for increasing was robust to the adjustment of covariates. CONCLUSIONS: Depressive symptom trajectories in late life are heterogeneous and linked to diverse socio-economic, clinical, and functional factors, some of which are trajectory-specific. Given its association with mortality, older people should be carefully monitored for depressive symptomatology.