Association between physical activity and adolescent pathological internet use: a chain mediation model of sleep problems and negative emotions, with multi-group SEM analysis by gender

In the contemporary digital era, adolescent Pathological Internet Use (PIU) has emerged as a significant global public health concern. Physical activity (PA) is widely recognized as a behavioral correlate of lower internet addiction risk, yet the complex psychophysiological interrelations potentially involved in this association remain insufficiently elucidated. Grounded in the Compensatory Internet Use Theory, this study aimed to examine the chain mediating roles of sleep problems and negative emotions in the association between physical activity and PIU, alongside exploring potential gender differences in these structural paths. A cross-sectional survey utilizing a stratified cluster sampling design was conducted among 1,245 adolescents (aged 12–18 years) recruited from four public secondary schools in a single major metropolitan area in China; findings are most directly applicable to the four participating schools, and should not be assumed to generalize even to other schools in the same city, to other metropolitan areas, or to Chinese adolescents in general, without further replication. Participants completed a battery of standardized instruments with published Chinese-language validation studies, though the validation populations differ by instrument and are not all adolescent samples (see § 2.2 for details): the International Physical Activity Questionnaire-Short Form (IPAQ-SF; Chinese version validated in college students, Qu & Li, 2004), the Pittsburgh Sleep Quality Index (PSQI; Chinese version validated in a mixed adult/college-student/psychiatric-patient sample, Liu et al., 1996), the Depression Anxiety Stress Scales-21 (DASS-21; Chinese version validated in college students, Gong et al., 2010), and the Compulsive Internet Use Scale (CIUS; Chinese-language version validated specifically in Chinese/Taiwanese high-school adolescents, Dhir et al., 2015). Structural Equation Modeling (SEM) and a 5,000-resample bootstrapping approach were employed to estimate indirect associations, preceded by tests of measurement invariance (ΔCFI = − 0.002, ΔRMSEA = − 0.001), followed by a multi-group SEM analysis. Physical activity was negatively associated with PIU. Furthermore, sleep problems and negative emotions each showed a statistically significant indirect association with PIU, and sequentially formed a statistically significant chain of indirect associations from physical activity to PIU. Multi-group analysis, conducted after establishing measurement invariance across gender groups, revealed significant gender heterogeneity in two specific structural paths: the association between physical activity and sleep problems was significantly stronger in females (β = −0.38 vs. −0.22, z = 2.94, p < .01), whereas the association between negative emotions and PIU was significantly stronger in males (β = 0.51 vs. 0.34, z = 3.12, p < .01). These cross-sectional findings suggest that physical activity, sleep, and emotional regulation are interrelated correlates of adolescent PIU. Comprehensive strategies integrating physical exercise, sleep hygiene, and gender-sensitive emotional regulation may inform the design of future intervention research; longitudinal and experimental studies are needed to establish the temporal and causal structure of these associations.