Abstract
BACKGROUND: Non-suicidal self-injury (NSSI) has become increasingly prevalent, with its impact growing more severe among adolescents. Addictive NSSI typically manifests through higher frequencies, more severe injuries, and a broader range of affected body areas, potentially leading to trauma, disability, or even suicide. This study aims to explore the complex network relationships among adolescent attachment, anxiety, and NSSI addiction, offering new insights into the mechanisms underlying NSSI addiction in adolescents. METHODS: A total of 1169 adolescent patients with NSSI were enrolled. Demographic questionnaires, the Ottawa Self-Injury Inventory, the Inventory of Parent and Peer Attachment, and the Multidimensional Anxiety Scale for Children were administered for assessment. The complex network relationships among symptoms were analyzed using undirected network analysis and directed Bayesian network analysis, followed by causal inference. RESULTS: The core symptom nodes in the network model included four dimensions of anxiety symptoms: MASC2 (harm avoidance), MASC3 (social anxiety), MASC4 (separation anxiety), and MASC1 (physical symptoms), as well as Peer3 (peer alienation) related to attachment relationships. Undirected network analysis indicated that the key bridging nodes for NSSI addiction were Peer3 (peer alienation) and MASC1 (physical symptoms). Directed acyclic graph (DAG) analysis further confirmed this relationship, demonstrating that these two key bridging nodes directly influence NSSI addiction. Additionally, DAG analysis revealed that MASC3 (social anxiety) indirectly influences NSSI addiction by affecting Peer3 (peer alienation) and MASC1 (physical symptoms). CONCLUSION: Physical anxiety symptoms and peer alienation directly influence NSSI addiction among adolescents. Additionally, social anxiety indirectly influences NSSI addiction by affecting physical anxiety and peer alienation.