Peer-reviewed article

Predicting poor health-related quality of life among Chinese adolescents using explainable machine learning: the role of school adjustment, family context, and lifestyle factors.

Zhao L, Gao Y, Liu C, Zhang Z, Yang Y et al. · 10 Jun 2026

Abstract

Background Adolescents' health-related quality of life (HRQoL) is influenced by multiple domains, including school, family, and individual lifestyle. However, evidence remains limited on accurately predicting adolescent HRQoL and identifying modifiable determinants. This study aimed to develop a machine-learning prediction model for adolescent HRQoL and to determine key contextual factors associated with low quality of life. Methods This cross-sectional study included 20,236 adolescents (51.71% female; mean age 15.97 years). HRQoL was assessed using the validated Quality of Life Scale for Children and Adolescents (QLSCA). Twenty predictors covering school adjustment, family characteristics, and health behaviours were entered into 9 machine-learning classifiers and logistic regression. Feature contributions were examined using permutation importance, partial dependence plots (PDPs), and SHapley Additive exPlanations (SHAP). Stratified analyses by sex and school stage were conducted to identify group-specific determinants. Results School adjustment was the primary predictor of HRQoL, with self-management and peer relations contributing most. Family factors ranked next, particularly more frequent exercise with parents, which was associated with higher HRQoL. Among lifestyle behaviours, higher physical activity and healthier dietary patterns were associated with better HRQoL, whereas excessive screen time was associated with poorer HRQoL. In stratified analyses, peer relations and physical activity were more influential among boys, while self-management, exercising with parents, and adequate sleep contributed more among girls. Middle school students showed more adverse associations with disruptive/antisocial behaviours and screen time, whereas high school students derived greater benefits from family interaction and health behaviours. Conclusions Explainable machine-learning analyses highlight school adjustment as a central correlate of adolescent HRQoL, alongside family characteristics and health behaviours. These findings provide multidimensional evidence to support targeted interventions in educational, family, and public health settings to improve adolescents' HRQoL.