Dynamic patterns of physical activity behavior, inflammatory trajectories, and risk of mental disorders in college students: a prospective cohort study using repeated longitudinal assessments

Mental disorders impose a substantial burden on college students, and prevention has become a public health priority. Regular physical activity is extensively recognized as advantageous for mental health, but current evidence primarily focuses on the dose of exercise while overlooking the role of exercise patterns (such as regularity) and their potential biological processes. The inflammatory hypothesis of mental disorders offers a theoretical framework for this, yet the temporal association between behavioral regularity, inflammatory dynamics, and psychological risk remains unknown. This study aimed to examine whether objectively measured exercise regularity, assessed using wearable devices, was associated with the risk of new-onset mental disorders in college students and whether systemic inflammation, indexed by hs-CRP and IL-6, partially mediated this association. Participants were recruited between January 1 and May 31, 2024, and followed for 18 months with one baseline assessment and six follow-up assessments conducted every 3 months, 4,415 undergraduate students without baseline mental disorders were recruited from Qufu Normal University and Shandong Sport University. Our study employed an 18-month longitudinal design with a total of 7 evaluations (baseline and every 3 months thereafter). The key exposure was exercise regularity, operationalized as the coefficient of variation (CV) of weekly moderate-to-vigorous physical activity (MVPA) minutes over the preceding 90 days based on POLO Watch data. A CV ≥ 25% was defined a priori as irregular exercise, indicating that week-to-week MVPA variation exceeded one quarter of the participant’s mean weekly MVPA. Venous blood was collected at each follow-up to detect inflammatory markers (hs-CRP, IL-6), and depressive and anxiety symptoms were assessed using the Patient Health Questionnaire-9 (PHQ-9) and Generalized Anxiety Disorder Scale-7 (GAD-7). Positive cases were confirmed by clinical interviews. The main outcome was confirmed new-onset depression or anxiety disorder. Time-dependent Cox proportional hazards models and longitudinal mediation models were used for statistical analysis. A total of 3,280 participants (retention rate of 74.3%) were included in the final analysis. During the 18-month follow-up, 217 new-onset mental disorder events were identified (total incidence rate of 6.6%). Time-dependent Cox models reported that in comparison with periods of regular exercise (CV < 25%), individuals in an irregular exercise state (CV ≥ 25%) had a significantly higher risk of new-onset mental disorders (adjusted hazard ratio aHR = 2.21, 95% CI: 1.68–2.90, p < 0.001). Longitudinal mediation analysis reported that irregular exercise significantly predicted an increase in serum hs-CRP levels 3 months later (β = 0.18, p < 0.001), and increased hs-CRP levels further significantly predicted an higher psychological risk 3 months thereafter (β = 0.12, p = 0.002). Bootstrap tests reported that hs-CRP played a significant mediating role in the correlation between irregular exercise and mental disorders, with a mediation proportion of 14.7% (95% CI: 5.3%–26.0%, p = 0.002). IL-6 also showed a similar, but marginally weaker mediating effect (mediation proportion of 9.5%, 95% CI: 2.1%–19.8%, p = 0.012). These findings suggest that campus mental health promotion strategies may benefit from emphasizing both stable exercise patterns and sufficient exercise volume.