Sex differences in association between Internet addiction and aggression among adolescents aged 12 to 18 in mainland of China

Previous studies have reported that Internet addiction (IA) is an important predictor of aggression. However, little is known whether aggression is a risk factor of IA. This study aims to explore whether IA can predict total and sub-types of aggression, whether aggression also can predict IA, and the potential sex difference. A total of 15,977 adolescents aged 12 to 18 were recruited though a stratified cluster sampling across five representative provinces in mainland of China. Young's Internet Addiction Test (IAT) and Buss and Warren's Aggression Questionnaire (BWAQ) were used to measure IA and aggression, respectively. The prevalence of IA and total aggression was 16.8 % and 25.0 %, respectively. Binary logistic regression analysis indicated that IAT score was significantly associated with total and five sub-types of aggression (P < 0.001). However, scores of physical aggression, hostility, and indirect aggression were positively associated with IA (P < 0.001), while verbal aggression and anger were not significant (P > 0.05). Moreover, the interaction of sex and IA was significant for aggression. Specifically, females with IA had higher risks of anger and indirect aggression but had lower risks of physical and verbal aggression compared to males with IA. Cross-sectional design, self-reported information, and only Chinese rural adolescents. IA is a strong risk factor of total and five sub-types of aggression, while not all sub-types of aggression are predictors of IA. Prevention programs for adolescents' IA and aggression should consider different aggression sub-types and sex.