Distinct functional recovery trajectories and their multidimensional determinants among stroke survivors: Evidence from a national longitudinal cohort

BACKGROUND: Post-stroke functional recovery is highly heterogeneous. However, evidence on long-term recovery trajectories and the role of multidimensional baseline factors in the Chinese population remains limited. This study identified recovery trajectories in activities of daily living (ADL) among Chinese stroke survivors and examined their associations with cognitive, physical, psychological, and social characteristics. METHODS: We analyzed four waves (2011-2018) of CHARLS data, including 851 stroke survivors with ≥ two follow-up ADL assessments. Latent class growth analysis (LCGA) identified recovery trajectories. Baseline variables were selected via Bootstrap-LASSO, and their associations with trajectory membership were assessed using multivariate logistic regression. RESULTS: Latent class growth analysis identified three distinct ADL-based functional recovery trajectories:Class 1, Moderate-Improving (16.8%); Class 2, Severe-Improving (18.0%); and Class 3, Mild-Worsening (65.2%). Patients in Classes 1 and 2 had greater functional impairment at baseline compared with Class 3, but showed significant improvements in ADL over time (slopes: -0.317, P = 0.011; -0.716, P < 0.001, respectively). In contrast, patients in Class 3 had relatively better baseline function but experienced gradual functional decline (slope = 0.382, P < 0.001). Multivariable analysis showed that cognitive function, self-rated health, depressive symptoms, physical function, social participation, and employment status were significant predictors of membership in the functional recovery trajectories. CONCLUSIONS: Long-term ADL-based functional recovery after stroke in China is highly heterogeneous and shaped by multidimensional baseline characteristics, including cognitive, physical, psychological, and social factors. Trajectory analysis of nationally representative longitudinal data can facilitate early identification of high-risk populations and provide evidence for risk-stratified management and precision rehabilitation strategies.