Internet use for learning and insomnia symptoms among high school and university students in China: a two-wave network analysis

Background Insomnia symptoms and internet use are both common among adolescents, but existing studies often rely on broad measures of screen exposure or global sleep outcomes. It remains unclear whether specific internet-use activities are associated with individual insomnia symptoms and whether these associations differ by educational context. This study aims to explore the association between internet use activity and insomnia symptoms among Chinese high school and university students. Methods Two-wave prospective longitudinal data were collected between 2024 and 2025 from high school and university students in Sichuan and Chongqing, China. Six internet-use activities were assessed: communication, entertainment, gaming, shopping, self-directed learning, and browsing/news use. Insomnia symptoms were measured with the Insomnia Severity Index (ISI). Longitudinal models were then fitted in the full sample and stratified by high school and university students. Residualised longitudinal networks were compared between high school and university students. Results Among 4226 students (age: 18.34 ± 1.33, 62.3% female), gaming was positively associated with five ISI items at baseline, whereas self-directed learning was negatively associated with the same items at baseline. Follow-up cross-sectional associations were similar but weaker. No association between Internet-use activities and insomnia symptoms was retained in pooled longitudinal models after Holm adjustment. Four associations were retained among high school students, whereas none was retained among university students. Conclusions Internet-use activities showed different cross-sectional associations and longitudinal associations at item level with insomnia symptoms. Future research should consider activity type and educational context when studying student internet use and insomnia.