Latent profile analysis of physical activity and mental health among chinese high school adolescents.

Adolescent mental health is an increasingly urgent public health issue, and physical activity is often promoted as a practical way to support psychological well-being. Yet the benefits of physical activity may not increase linearly, and average associations can mask meaningful differences between subgroups. Using questionnaire data from a nationally archived Chinese adolescent health database, we analyzed 13,494 students aged 14-18 years. Physical activity was assessed with the Physical Activity Questionnaire for Adolescents, and psychological symptoms were measured with the Symptom Checklist-90, including nine symptom domains and a total index. We evaluated the shape of the activity-symptom association across the full activity range and used latent profile analysis to identify subgroups defined by the joint distribution of activity and symptom burden. A four-profile model was supported by fit and classification indices, with AIC = 876,522, BIC = 892,341, and entropy = 0.89; the Lo-Mendell-Rubin test favored four over five profiles (p = 0.012), and the bootstrap likelihood ratio test was significant (p < 0.001). The profiles captured substantial heterogeneity, spanning a high-activity/low-symptom pattern, a low-activity/high-symptom pattern, and two intermediate profiles in which similar activity levels co-occurred with different levels of distress. At the population level, symptom burden was lowest at moderate activity, while both lower and higher activity levels were associated with higher symptom burden than the moderate group. These findings indicate that activity volume alone does not fully explain adolescent psychological distress and support profile-informed approaches that prioritize an attainable moderate-activity target and targeted support for higher-risk subgroups.