Developing a predictive model for classifying high-risk groups of post-stroke depressive symptoms based on the health...

Depressive symptoms affect a significant proportion of stroke survivors, negatively affecting quality of life and functional recovery. This study was guided by the health ecology model (HEM). It aimed to build a predictive model for identifying high-risk individuals with post-stroke depression symptoms (PSDS), providing theoretical support for prevention strategies. Data were extracted from the CHARLS. Depressive symptoms were measured using the CESD-10 scale. Guided by the HEM, influencing factors were identified and stratified. Binary logistic regression was used to analyze determinants of PSDS, while the Harvard Cancer Index was used to assess the risk of depressive symptoms among stroke survivors. A total of 54.13% of participants met CESD-10 criteria for depression. Multivariate analysis identified self-rated health status, activities of daily living, pain, drinking, night sleep time, marital status, life satisfaction, employment, and medical insurance as factors significantly associated with PSDS. Notably, risk stratification via the Harvard Cancer Index revealed a clear ordered trend with PSDS prevalence increasing progressively alongside higher risk categories (χ² = 41.395, p < 0.001). Significant differences existed between each consecutive risk level (χ² = 69.132, p < 0.001). PSDS is determined by multiple factors. The Harvard Cancer Index effectively stratifies PSDS risk in stroke survivors with distinct prevalence differences across ordered risk grades. This index provides a practical tool for identifying high-risk individuals and directly supports the development of targeted and efficient intervention strategies.