Gender differences and dynamic mechanisms in depressive symptom trajectories among middle-aged and older adults: a study based on latent trajectory modeling and cross-lagged network analysis.

objectiveThis study aimed to investigate the long-term developmental trajectories of depressive symptoms and their gender differences among middle-aged and older Chinese adults, utilizing latent trajectory modeling and cross-lagged network analysis to reveal the dynamic interactions among symptoms.

methodsBased on nationally representative data from the China Health and Retirement Longitudinal Study (CHARLS) spanning 2011–2020, a total of 4,285 adults aged 45 and above were included. Group-Based Trajectory Modeling (GBTM) was employed to identify heterogeneous developmental trajectories of depressive symptoms. Multinomial logistic regression was used to examine the associations of sociodemographic, health, and psychological characteristics with different trajectory classes. Gender-stratified temporal symptom networks were constructed using the multilevel Vector Autoregressive (mlVAR) model to analyze dynamic inter-symptom interactions and identify central driving nodes.

results1. Trajectory Characteristics: Significant heterogeneity in depressive symptoms trajectories was observed between genders. The optimal model for men identified four distinct classes (Persistently Low, Stable Moderate, Moderate-High Fluctuating, High-Increasing), while a five-class model was optimal for women (Persistently Low, Stable Moderate, High-Slightly Increasing, Moderately High-Progressive Increasing, High-Declining then Stabilizing). The proportion of women in moderate-to-high or high depressive symptoms trajectories (approximately 56%) was significantly greater than that of men (approximately 49%). 2. Risk Factors: Rural residence, a higher number of chronic conditions, presence of pain, shorter sleep duration, lower cognitive function, and lower life satisfaction were common risk factors associated with more severe depressive symptoms trajectories in both genders. Gender-specific risks were identified: smoking was a significant risk factor for men, whereas alcohol use, limitations in activities of daily living (ADL), and frequent physical labor were more closely associated with high-risk trajectories in women. 3. Network Structure: The overall temporal symptom network revealed a multidimensional feedback structure connecting emotional, cognitive, and behavioral dimensions. “Feeling unable to go on” and “difficulty concentrating” emerged as key driver nodes, while “feeling fearful” and “restless sleep” primarily functioned as outcome nodes. 4. Gender Differences: The male symptom network was relatively sparse and linear, dominated by cognitive-behavioral symptoms (depressed mood, difficulty concentrating) forming a “motivation-attention-sleep” pathway. In contrast, the female network was denser and more complex, centered on emotional-social symptoms (feeling unable to go on, anhedonia, loneliness), forming a core “emotion-loneliness-hopelessness” feedback loop.

conclusionThe development of depressive symptoms in middle-aged and older adults exhibits significant trajectory heterogeneity and gender-specific dynamic mechanisms. Depressive symptoms manifested primarily through a cognition-behavior-oriented pathway in men, versus an emotion-social-oriented pathway in women. These findings provide empirical evidence for developing gender-sensitive dynamic monitoring and precision interventions for depressive symptoms, advocating a paradigm shift from static description to the investigation of dynamic mechanisms.