Zi-Ye Huang^1,2†, Qian-Nan Ruan^1,2†, Yawen Zheng³, Heng Miao², Yu-Wei Wu^4* and Wen-Jing Yan^2,5,6*

Abstract

Background Suicidal ideation (SI) is increasingly prevalent among adolescents, often arising from depression and linked with non-suicidal self-injury (NSSI). Previous studies have noted significant sex differences in the manifestation and predictors of SI, depression, and NSSI.

Aim This study aims to analyze and compare the relationships between SI, depression, and NSSI among male and female adolescents, examining whether these associations differ based on sex.

Methods A total of 368 adolescents (M = 15.43, SD = 1.22, about 56.2% female participants), both from clinical and school settings, were assessed for SI, depression, NSSI, and other related variables. Network analysis was utilized to explore the interconnections among these variables, focusing on identifying sex-specific patterns. Logistic regression was used to confirm the findings from the network analysis.

Results The network analysis revealed significant sex differences in the relationships between SI, depression, and NSSI. In the female network, the edge weights between SI and NSSI (0.93) and between SI and depression (0.31) were much higher compared to the male network (0.29 and 0, respectively). Centrality indices (strength, betweenness, closeness, and expected influence) for SI, NSSI, and depression were also higher in the female network. Logistic regression confirmed these findings, with depression being a potential predictor of SI only in females (OR = 1.349, p = 0.001) and NSSI having a stronger influence on SI in females (OR = 13.673, p < 0.001) than in males (OR = 2.752, p = 0.037).

Conclusion The findings underscore the necessity of considering sex differences when predicting suicidal ideation from depression and NSSI in adolescents. Intervention and prevention strategies should be tailored to address these distinct patterns in male and female adolescents.

Keywords Suicidal ideation, Depression, Non-suicidal self-injury, Adolescents, Sex differences, Network analysis, Logistic regression

GKB-NON-2024-00443