Psychological-behavioral risk profiles of adolescents and their associations with suicidal behaviors.

OBJECTIVE

Adolescent suicidal behaviors pose a critical public health challenge. This study examined the co-occurrence patterns of psychological and behavioral risk factors and evaluated their associations with suicidal ideation (SI) and suicide attempts (SA).

METHODS

A multicenter adolescent survey conducted in 2023-2024 (n = 19,312) collected sociodemographic, psychological, and behavioral data. Complex survey logistic regression was applied to identify factors associated with SI and SA. Latent class analysis (LCA) was used to derive psychological-behavioral risk profiles, and their discriminative performance was evaluated using logistic regression with 10-fold cross-validation.

RESULTS

The prevalence of SI and SA was 25.41% and 8.03%, respectively. Fourteen factors were associated with SI, including resilience, social capital, hostility, depression, anxiety, maladjustment, emotional instability, psychological disequilibrium, virtual world integration, adverse childhood experiences, negative life events, sleep quality, internet addiction and alcohol use. For SA, academic stress and smoking were additionally retained, whereas anxiety, emotional instability, psychological disequilibrium and internet addiction were not retained. Four distinct profiles were identified: Low-risk/High-protection, Moderate distress, Psychological-problem dominant and High-risk/Multi-problem. Compared with the Low-risk group, adolescents in the High-risk group demonstrated substantially higher odds of SI (aOR = 15.87, 95% CI: 13.96-18.03) and SA (aOR = 11.72, 95% CI: 10.11-13.60). The LCA-based model showed good discrimination (AUC = 0.796 for SI; 0.772 for SA), outperforming single-variable models.

CONCLUSION

Psychological and behavioral factors among adolescents cluster into distinct risk profiles that are associated with different levels of suicidal behaviors. These findings provide a person-centered, multidimensional perspective for understanding heterogeneity in adolescent suicide-related risk.