Association between physical activity and mental health among Yi students aged 10-18 years in Southwest China: a school-based cross-sectional study.

Background: Physical activity is a modifiable behavior that may contribute to youth mental health, yet evidence from ethnic minority school populations in Southwest China remains limited. Methods: This school-based cross-sectional study analyzed questionnaire data from 671 Yi students aged 10-18 years. Physical activity was assessed using an 8-item PAQ-like measure. Mental health was operationalized through three adverse domains: anxiety, loneliness, and social anxiety. Age, sex, and school stage were included as covariates. Psychometric performance was examined using Cronbach's alpha, the Kaiser-Meyer-Olkin statistic, and Bartlett's test of sphericity. Primary associations were estimated with ordinary least squares regression using HC3 robust standard errors, followed by sex- and stage-stratified models, PAQ tertile analyses, linear trend tests, strict complete-item sensitivity analyses, and exploratory parenting extension models. Results: The final sample included 379 girls and 292 boys with a mean age of 13.87 years. Reliability was acceptable to excellent across the four core scales (alpha = 0.709-0.938), and factorability indices supported the use of scores in this sample. Physical activity was inversely correlated with loneliness ( Conclusion: In this minority school sample, the mental health pattern associated with physical activity was selective rather than global. Higher physical activity was most consistently linked to lower loneliness, while its associations with anxiety and social anxiety were not robust. These findings support a more contextualized public health interpretation of youth physical activity, in which relational connectedness and family climate may shape psychological benefits.