Predictors of depression and anxiety in community dwelling stroke survivors: a cohort study

Purpose: Few longitudinal studies explore post-stroke patterns of psychological morbidity and factors contributing to their change over time. We aimed to explore predictors of post-stroke depression (PSD) and post-stroke anxiety over a 12-month period. Methods: A prospective cohort study. Consecutively recruited stroke patients (n - 134) participated in face-to-face interviews at baseline, 3, 6, 9, and 12 months. Primary outcome measures were depression and anxiety (measured via Hospital Anxiety and Depression Scale). Independent variables included disability (Modified Rankin Scale), Quality-of-life (Assessment Quality-of-life), social support (Multi-dimensional Scale Perceived Social Support) and community participation (Adelaide Activities Profile (AAP)). Secondary outcomes were predictors of resolution and development of PSD and anxiety. Results: Anxiety (47%) was more common than depression (22%) at baseline. Anxiety (but not depression) scores improved over time. Anxiety post-stroke was positively associated with baseline PSD (p<0.0001), baseline anxiety (p<0.0001) and less disability (p = 0.042). PSD was associated with baseline anxiety (p<0.0001), baseline depression (p = 0.0057), low social support (p = 0.0161) and low community participation (p<0.0001). The only baseline factor predicting the resolution of PSD (if depressed at baseline) was increased social support (p = 0.0421). Factors that predicted the onset of depression (if not depressed at baseline) were low community participation (p = 0.0015) and higher disability (p = 0.0057). Conclusion: While more common than depression immediately post-stroke, anxiety attenuates while the burden of depression persists over 12 months. Clinical programs should assess anxiety and depression, provide treatment pathways for those identified, and address modifiable risk factors, especially social support and social engagement.