The impact of physical activity on social-emotional competence among female graduate students: a three-wave longitudinal study in Chinese Universities

Social-emotional competence (SEC) is a critical protective factor for mental health among female graduate students. While cross-sectional research suggests potential associations between physical activity (PA) and SEC development, longitudinal evidence specifically targeting highly educated adult females remains scarce, hindering the design of tailored interventions. This study investigated the longitudinal effects of PA on SEC among Chinese female graduate students and elucidated the underlying mechanisms. A three-wave longitudinal panel design with a 9-month follow-up was employed. Using cluster random sampling, female graduate students (master’s or doctoral level; ≥18 years) from 68 “Double First-Class” universities were recruited. The final sample consisted of 860 participants who completed all three waves. Assessment tools included the International Physical Activity Questionnaire-Short Form (IPAQ-SF), the Body Esteem Scale for Adolescents and Adults (BESAA), and the Delaware Social-Emotional Competency Scale-Short (DSECS-S). Cross-lagged panel modeling (CLPM) with bootstrapping was utilized to examine longitudinal relationships, mediation, and moderated mediation effects. PA exerted sustained positive predictive effects on SEC. Specifically, PA at T1 predicted SEC at T2 (β = 0.264, p < 0.001), and PA at T2 predicted SEC at T3 (β = 0.229, p < 0.001). Body esteem significantly mediated this relationship (indirect effect β = 0.074, 95% CI: 0.055–0.096). Furthermore, the upbringing environment (urban vs. rural) significantly moderated the mediation pathway; specifically, the beneficial effects were more pronounced for rural participants(β = 0.103) than for their urban counterparts(β = 0.057), with a moderated mediation index of 0.045 (95% CI: 0.005–0.087). Regular PA consistently promotes SEC development in female graduate students by enhancing body esteem, with stronger effects observed among those from rural backgrounds. These findings offer a scientific basis for designing differentiated campus-based PA interventions. Specifically, programs should integrate PA with body-esteem enhancement and implement stratified strategies tailored to upbringing environments to optimize mental health outcomes for this population.