Trajectories of Proactive Health Behaviors Among Chinese Middle-Aged and Older Adults with Multimorbidity: A Cohort Study Using Group-Based Trajectory Modeling

(1) Background: Proactive health behaviors are key to reducing their burden and supporting healthy aging. (2) Methods: We analyzed five waves (2011–2020) of CHARLS data from 1343 middle-aged and older adults (≥45 years) with multimorbidity. An entropy weight method was used to create a composite score for proactive health behaviors, and group-based trajectory modeling identified behavioral trajectories. Multivariate logistic regression and Shapley value decomposition assessed determinants and their relative contributions. Generalized structural equation modeling and latent class analysis were applied to estimate direct and indirect effects across the full sample and key multimorbidity subgroups. (3) Results: Two trajectories emerged: a “declining group” (91.44%) and an “improving group” (8.56%). The improving group was more likely to include younger, urban individuals with higher education, retired status, smaller family size, and lower depression levels. Education (40.67%) and depressive symptoms (31.22%) were the strongest determinants of trajectory. Path analysis showed that higher education and retirement indirectly supported sustained proactive health behaviors by reducing depression. The direct and indirect effects varied across subgroups. (4) Conclusion: The proactive health behaviors of middle-aged and elderly patients with multimorbidity exhibit a declining trend. Future health policies and interventions should prioritize mental health.