Are candidate intervention targets consistent across adolescents with problematic internet use: an empirical study integrating multiple methods

Problematic internet use (PIU) in adolescents is closely related to maladaptive cognitive processing and dysregulated emotional responses. Inspired by the cognitive-behavioral model, this study examined whether adolescents with PIU show heterogeneous anxiety-rumination profiles, distinct dimension-level network structures, and subgroup-specific candidate intervention targets. From June to August 2024, a total of 24,470 students in Grades 7–12 from 28 secondary schools in Wuyuan, Yantai, and Rizhao, China, were recruited using convenience sampling. PIU, anxiety, and rumination were assessed using Chinese versions of the Internet Addiction Test, Multidimensional Anxiety Scale for Children, and Ruminative Response Scale. Network analysis, latent profile analysis, and the NodeIdentifyR algorithm were integrated to compare network structures across subgroups and identify subgroup-specific candidate intervention targets. Among all participants, 4,350 adolescents (17.8%) met the operational criterion for PIU. Adolescents with PIU reported higher anxiety and rumination than those without PIU, and their dimension-level network structure differed significantly. Latent profile analysis identified two PIU subgroups: low and high anxiety-rumination co-occurrence. These subgroups differed significantly in global network strength and network structure. Network simulations identified Harm Avoidance as the candidate intervention target in the low co-occurrence subgroup and Symptom Rumination as the candidate intervention target in the high co-occurrence subgroup. Adolescents with PIU are psychologically heterogeneous. Subgroup-specific network simulations suggest that intervention targets may differ according to anxiety-rumination co-occurrence, although these model-based findings require validation in longitudinal and experimental studies. Not applicable.