Baiyi Yang, Xiaodi Yang, Zhengyang Fan, Chang Liu, Zhanfei Zheng
Background This study aimed to explore the mechanisms through which physical activity affects social anxiety in college students, with a specific focus on the sequential mediating roles of family support and self-efficacy. Methods A total of 391 valid responses were analyzed using the Physical Activity Rating Scale (PARS-3), the Interaction Anxiousness Scale (IAS). Family Support Scale (PSS-Fa), and the General Self-Efficacy Scale (GSES). Mediation analyses were conducted with structural equation modeling (SEM) and bootstrapping procedures (5,000 resamples). Results Physical activity significantly predicted lower social anxiety (β = − 0.187, p < .001). Both family support (β = − 0.309, p < .001) and self-efficacy (β = − 0.390, p < .001) mediated this association, with the sequential pathway (β = − 0.073, p < .001) accounting for 13.35% of the total effect. Conclusion Physical activity was associated with lower social anxiety indirectly via higher family support and self-efficacy, while the direct effect was not statistically significant in the full mediation model. These findings provide empirical support for incorporating physical activity into university-based mental health strategies aimed at alleviating social anxiety.