Longitudinal associations between multiple health-risk indicators and psychological distress among university students: a three-wave cross-lagged study

Abstract

Background Multiple health-risk indicators (MHRIs) and psychological distress may accumulate together among university students, yet longitudinal evidence regarding their developmental trajectories and bidirectional associations remains limited. This study examined changes in MHRIs and psychological distress and their bidirectional longitudinal associations to provide evidence for integrated risk identification and intervention strategies among university students. Methods A three-wave longitudinal study was conducted among university students from nine universities in Guangxi, Guizhou, and Yunnan provinces in April 2025, October 2025, and April 2026. The final analytic sample comprised 1,762 students who completed all three waves and had consistent anonymous codes. MHRIs count was calculated by summing nine behavioral, psychosocial, and health-related risk indicators. Psychological distress was assessed using the Depression Anxiety Stress Scales-21 (DASS-21). Latent growth models (LGMs) were used to examine developmental trajectories and individual differences in initial levels and rates of change, and a cross-lagged panel model (CLPM) was applied to assess bidirectional longitudinal associations after adjusting for baseline covariates, including sex, household registration, campus residence status, only-child status, and adverse childhood experiences. Results MHRIs count and psychological distress both showed increasing trends across the three measurement occasions. Unconditional LGMs indicated significant increases over time with individual differences in initial levels and rates of change, corresponding to average increases of 0.10 MHRIs and 1.55 psychological distress points per 6-month interval. The CLPM demonstrated acceptable-to-good model fit. After adjustment for baseline covariates, higher psychological distress predicted a greater number of subsequent MHRIs at both follow-up intervals (T1–T2: β = 0.167, p < 0.001; T2–T3: β = 0.141, p < 0.001). Conversely, higher MHRIs also predicted subsequent increases in psychological distress (T1–T2: β = 0.132, p < 0.001; T2–T3: β = 0.119, p < 0.001). Conclusion MHRIs and psychological distress increased over the 1-year follow-up period and demonstrated a stable positive bidirectional longitudinal association. Integrating psychological distress screening with assessment of multiple health risks, continuous monitoring, and coordinated health promotion strategies may help reduce the accumulation of psychological and health-related risks among university students.