Abstract
BACKGROUND: University physical education (PE) class may promote healthier behaviors and better mental health, yet evidence on their impact in higher education is limited. This study examined associations between PE participation, mental health, health-related behaviors, and sleep quality among Chinese university students. METHODS: A cross-sectional survey was conducted among 2,117 s-year undergraduate students, using proportional stratified sampling in 2024. Validated instruments were used to collect data: demographics and anthropometrics (self-report height and weight to calculate BMI); International Physical Activity Questionnaire-Short Form (IPAQ-SF) for total weekly physical activity; General Health Questionnaire-12 (GHQ-12) for mental health; Pittsburgh Sleep Quality Index (PSQI) for sleep quality; and self-reported screen time. To objectively assess activity levels during PE lessons, ActiGraph wGT3X-BT accelerometers were used. Logistic and multiple linear regression analyses identified associations and predictors of mental health, health-related behaviors, and MVPA proportion during PE class. RESULTS: In linear models, total physical activity (β = 0.17, p = 0.003), greater MVPA during PE (β = 0.19, p = 0.003), and shorter screen time (β = -0.14, p = 0.009) predicted better mental health. On average, students spent 48.6% of PE lesson time in MVPA. A higher MVPA proportion was associated with lower odds of obesity (OR = 0.41, p < 0.001), physical inactivity (OR = 0.38, p < 0.001), and poor sleep quality (OR = 0.72, p = 0.004). Male students had greater odds of obesity (OR = 3.14, p < 0.001) and smoking (OR = 2.02, p = 0.001), while females reported poorer sleep. Predictors of MVPA proportion in PE included higher BMI (β = -0.11, p < 0.001), female gender (β = -0.14, p < 0.001), poorer sleep quality (β = -0.09, p = 0.003), and lower mental health (β = 0.11, p = 0.002). CONCLUSION: PE class may provide an important context for promoting health-related behaviors and mental health among university students. However, given the cross-sectional design, causal relationships cannot be inferred. Longitudinal and intervention studies are needed to confirm these findings and clarify underlying mechanisms.