Depressive Symptoms Profiles and Cognitive Outcomes After Stroke

Post-stroke depressive symptoms are heterogeneous and variably associated with other psycho-cognitive features. We employed cluster analysis to identify distinct profiles of post-stroke depressive symptomatology and their association with cognitive performance. We included consecutive patients undergoing neuropsychiatric evaluation 6 months after stroke. Cluster analysis incorporated the Center for Epidemiologic Studies Depression Scale, along with the apathy and anxiety items from the Neuropsychiatric Inventory questionnaire. Baseline clinical/neuroimaging variables and 6-months cognitive outcomes were compared across profiles. We included 189 patients with acute cerebrovascular events (median age 75.4 years, 62% male, 80% ischemic strokes). Three profiles emerged: (A) low-depressive symptoms (n = 108), (B) moderate-depressive symptoms plus anxiety (n = 41), (C) high-depressive symptoms plus apathy (n = 40). Regarding baseline predictors of 6-month depressive symptoms profiles, patients with high-depressive symptoms plus apathy exhibited lower Montreal Cognitive Assessment scores at baseline (16.0 vs. 21.5; adjusted odds ratio [adj.OR] per 1-point increase 0.91, 95% confidence interval [95% CI] 0.83-0.99) compared to patients with low-depressive symptoms; moderate-depressive symptoms plus anxiety patients had less cortical atrophy compared to both low-depressive symptoms (adj.OR 0.92, 95% CI 0.86-0.99) and high-depressive symptoms plus apathy (adj.OR 0.89, 95% CI 0.81-0.97) profiles. Regarding 6-month cognitive performance, high-depressive symptoms plus apathy patients showed higher rates of post-stroke dementia and attention/executive function impairment compared with the two other groups (both p < 0.05), and higher rates of language impairment compared with low-depressive symptoms profile (p < 0.05). By integrating apathy and anxiety in our model, depressive symptoms after stroke emerged as heterogeneous neuropsychiatric syndromes, showing different baseline predictors and distinctive cognitive patterns.