Network structure of non-suicidal self-injury and multidomain psychiatric symptoms among adolescents with elevated depressive symptoms: a cross-sectional study
Year: 2026
Type: preprint
Study design: Observational Study
Abstract: Background Non-suicidal self-injury (NSSI) commonly co-occurs with depressive symptoms during adolescence, but its position within broader symptom-domain systems among adolescents with elevated depressive symptoms remains unclear. Examining this PHQ-defined subgroup may clarify whether NSSI is primarily linked with depressive symptoms or also connected with other domains relevant to follow-up assessment. This study examined how NSSI was embedded within a multidomain psychiatric symptom network among school-based adolescents with Patient Health Questionnaire (PHQ) module scores of 5 or higher. Methods This cross-sectional school-based screening study included 1,146 adolescents who met the Patient Health Questionnaire (PHQ) ≥ 5 inclusion criterion and had complete data on 10 symptom-domain nodes. A Spearman correlation matrix and EBICglasso were used to estimate the primary network. Signed-spinglass community detection with repeated seeds and 1,000 case-bootstrap networks was used to identify and validate data-driven communities. Node predictability, centrality, bridge expected influence, bootstrap stability, and relative importance analyses characterized symptom-domain organization and explained-variance patterns. Results The final network contained 43 nonzero edges among 45 possible edges. Signed-spinglass community detection identified a reproducible two-community organization: a core symptom community containing Ma, Pa, PTSD, GAD, PHQ, OCD, and NSSI, and a separation/social/school anxiety community containing SeP, SoP, and ScP. The two-community solution occurred in 999 of 1,000 case-bootstrap networks. Depressive symptoms showed the highest strength, one-step expected influence, two-step expected influence, and predictability. Bridge expected influence was highest for separation anxiety, school anxiety, and social anxiety symptoms, followed by depressive and obsessive-compulsive symptoms. NSSI showed moderate predictability (R² = 0.246), and depressive symptoms made the largest relative-importance contribution to NSSI explained variance. Conclusions Among adolescents with elevated depressive symptoms, NSSI was linked most closely with depressive, trauma-related, obsessive-compulsive, and generalized anxiety symptom domains. The findings suggest that NSSI in this subgroup was associated with depressive symptoms but was not limited to mood-related distress. Screening and follow-up assessment for adolescents with elevated depressive symptoms may benefit from considering trauma-related, repetitive/intrusive, and worry-related symptoms alongside depressive severity and NSSI. Because the data were cross-sectional and based on screening-domain scores, the findings should be interpreted as descriptive symptom-domain associations rather than causal pathways.
Source: Research Square
Institution: Wenzhou Medical University
Language: English
Subfield: Experimental and Cognitive Psychology
Field: Psychology
Domain: Social Sciences
Open Access status: green
OpenAlex ID: W7225059288