Abstract
BACKGROUND: Internet addiction (IA) has become a growing concern, particularly among adolescents, due to its adverse effects on mental health, physical well-being, and future development. Adolescents with suicidal ideation (SI) are particularly vulnerable to IA, which may be associated with a higher risk of engaging in suicidal behaviors. However, the relationship and underlying mechanisms between SI and IA remain unclear. This study, grounded in the cognitive-behavioral model of pathological internet use, investigates the relationship and explores the roles of self-esteem (mediator) and school connectedness (moderator) in this association. METHODS: In this cross-sectional study, 462 Chinese adolescents with SI (79.0% female) were recruited from psychiatric outpatient clinics between June 2024 and September 2025. Validated instruments measured SI, self-esteem, school connectedness, and IA. Structural equation modeling with bootstrapping procedures was used to test the mediation effect of self-esteem on the relationship between SI and IA. The moderating role of school connectedness was examined using PROCESS Model 8. RESULTS: SI was positively associated with IA (β = 0.224, p < 0.001). SI was negatively associated with self-esteem (β = -0.464, p < 0.001), and self-esteem was further negatively associated with IA (β = -0.448, p < 0.001). Self-esteem partially mediated the relationship between SI and IA, with an indirect effect of 0.208 (95% CI: 0.154-0.271). School connectedness significantly moderated the direct association between SI and IA (β = -0.005, p = 0.001), but did not moderate the association between SI and the mediator, self-esteem (β = 0.004, p = 0.202). CONCLUSION: This study identifies a significant positive association between SI and IA among adolescents with SI, with self-esteem partially mediating this link. Furthermore, school connectedness showed a very weak buffering effect on the direct association between SI and IA, and it does not moderate the association between SI and self-esteem. These findings enhance our understanding of the mechanisms underlying IA in this vulnerable population and suggest potential targets for interventions.