A comparative study of network structures between different sedentary behaviors and depressive symptoms in adolescents with and without non-suicidal self-injury

This study aimed to examine association networks between sedentary behaviors and depressive symptoms among adolescents with and without non-suicidal self-injury (NSSI), compare network structures and global connectivity between the two groups, and identify key bridge nodes within each network. A total of 8,102 Chinese adolescents were included, comprising 1,996 in the NSSI group and 6,106 in the Non-NSSI group. Nine types of sedentary behaviors and 20 depressive symptoms from the Center for Epidemiologic Studies Depression Scale (CES-D) were assessed using self-reported questionnaires. Based on a mixed correlation matrix, sedentary behavior–depressive symptom networks were estimated separately for the two groups using the extended Bayesian information criterion graphical LASSO (EBICglasso), and bridge strength and one-step bridge expected influence were computed. A network comparison test was conducted to compare network structure and global strength between groups, along with covariate-adjusted sensitivity analyses. Significant differences in network structure were observed between the two groups (M = 0.114, P < 0.001), indicating distinct conditional association patterns between sedentary behaviors and depressive symptoms in adolescents with and without NSSI. The global strength was 11.913 for the NSSI group and 11.476 for the Non-NSSI group, with a non-significant difference of 0.437 (3.8%, P = 0.636). In the NSSI group, mobile phone/tablet use, offline classes/tutoring, and sitting and chatting showed higher bridge expected influence, while difficulty concentrating, reduced speech, and restless sleep emerged as prominent bridge symptom nodes. After covariate adjustment, the network structure differences remained statistically significant, whereas differences in global strength remained non-significant; core bridge nodes were largely stable preserved in both groups. Although differences were observed in the specific association structures of sedentary behavior–depressive symptom networks between adolescents with and without NSSI, there was no evidence that the NSSI group exhibited significantly higher overall network connectivity. Mobile phone/tablet use, study-related sedentary behaviors, sitting and chatting, and symptoms related to attention, sleep, and social withdrawal may represent important candidate nodes for future longitudinal studies and risk screening; however, their causal roles and potential intervention value require further investigation.