Yiqiong Huang

Introduction This cross-sectional, exploratory study investigated the association between physical exercise (PE) and social anxiety (SA) among 862 left-behind adolescents in China. It examined emotion regulation abilities (ERA) and peer relationships (PR) as potential mediators and gender as a moderator in this association. Methods A cross-sectional exploratory design was adopted. Participants were 862 Chinese left-behind adolescents. Path analysis was used to test the hypothesized sequential mediation model and gender moderation effect. Results Path analysis revealed that PE was negatively associated with SA. Each separate path of the hypothesized sequential chain (PE → ERA, ERA → PR, PR → SA) was statistically significant, while the overall sequential mediation effect (PE → ERA → PR → SA) was non-significant (95% CI included zero). This indicated that the significant PE–ERA association had a small effect size and could not propagate fully through the chain to SA. The direct path from PE to PR was also non-significant. The data favoured independent mediation pathways rather than a sequential mediation process. Gender moderated the pathway: the positive PE–ERA association was significant for females but non-significant for males. Discussion The findings are hypothesis-generating rather than confirmatory and require replication via longitudinal research. For vulnerable adolescents, PE may help build emotional competence, which contributes to improved peer bonds. Interventions could integrate emotion-regulation skill training into physical activity programmes, particularly targeting female adolescents. Due to the cross-sectional nature of the study, causal conclusions cannot be drawn.