A population-based cross-lagged panel network analysis of multidimensional mental health symptoms among Chinese secondary school students

Adolescent mental health issues are increasingly severe, with the prevalence of depression reaching 26.17%. Existing studies still face issues such as a narrow range of symptoms, cross-sectional designs, and small sample sizes. To address these gaps, the present study applied cross-lagged panel network (CLPN) analysis to a population-based sample of 17,236 secondary school students in Guangzhou, China (aged 12-18; 46.32% female; 66.91% junior high, 33.09% senior high). Participants were assessed on ten symptom dimensions at two time points approximately six months apart, during September-October 2023 (T1) and February-April 2024 (T2). Subgroup comparisons were conducted by gender and educational stage. The findings revealed: (1) cross-sectional networks exhibited robust and stable associations, with depression and anxiety forming the strongest link; (2) longitudinal analyses identified depression, anxiety, and emotional disturbance as central driving nodes that predicted and propagated changes in other mental health symptoms over time, while paranoid ideation functioned primarily as a bridging symptom; and (3) subgroup analyses showed largely consistent structures across gender and educational stages, though subtle differences emerged. Males demonstrated slightly stronger connectivity, with depression and emotional disturbance as core drivers, whereas hostility and interpersonal sensitivity were more central in females. Across educational stages, the "depression → anxiety" pathway remained stable, with emotional disturbance more prominent in junior high and hostility gaining influence in senior high. Overall, these results highlight the central role of anxiety in adolescent symptom networks, reflecting the pervasive influence of China's high-pressure educational environment on students' psychological functioning. They underscore the importance of disrupting the depression-anxiety cycle and suggest that tailored, stage- and gender-specific interventions targeting emotional symptoms may enhance the effectiveness of school-based mental health promotion.