Network analysis of comorbid internet addiction, anxiety, and depression symptoms among Chinese junior high school students

Background Internet addiction (IA) frequently co-occurs with anxiety and depression during adolescence. However, traditional latent variable models based on total scores often mask specific symptom-level interactions. This study utilized network analysis to elucidate the fine-grained associations between IA, anxiety, and depression symptoms among Chinese adolescents and to explore potential gender differences in these network structures. Methods A cross-sectional study was conducted among 10,491 junior high school students (mean age = 13.88 years; 51.98% male) in Jingzhou City, China. Participants completed the Internet Addiction Test (IAT), the Generalized Anxiety Disorder-7 (GAD-7), and the Patient Health Questionnaire-9 (PHQ-9). A Gaussian Graphical Model (GGM) was estimated to visualize the symptom network. Centrality indices and bridge strength were calculated to identify key symptoms. The Network Comparison Test (NCT) was employed to assess gender differences in global strength and network structure. Results The network analysis revealed a stable structure driven primarily by symptoms reflecting escapism and social withdrawal. Specifically, “Preferring online interaction to real-world intimacy” (IAT19) and “Using the internet to block disturbing thoughts” (IAT10) exhibited the highest bridge strength, directly connecting to core depressive symptoms such as anhedonia (PHQ1) and depressed mood (PHQ2). The NCT indicated that while the global strength of the network was invariant across genders ( p > 0.05), the network structure differed significantly between males and females ( p = 0.01), suggesting distinct pathways of symptom interaction. Conclusion Escapism and the displacement of offline social interactions are critical mechanisms maintaining the comorbidity of IA and emotional distress. Interventions should prioritize targeting these bridge symptoms—specifically by improving emotion regulation strategies and enhancing real-world social connections—to disrupt the feedback loop between addiction and internalizing problems. The structural differences observed suggest that gender-sensitive approaches may be required for effective intervention.