Sex-Specific Classes of Self-Harm and Their Association with Suicide Risk In Adolescents: A Nationally Representative Study

Background and Purpose: Suicide remains a leading cause of preventable mortality among adolescents, and self-harm is one of its most robust predictors. Yet self-harm is frequently treated as a unitary construct, obscuring meaningful heterogeneity in behavioral patterns and their differential links to suicide risk. Most prior studies have pooled male and female adolescents or treated sex as a covariate, limiting insight into sex-specific self-harm configurations. This study aimed to (1) identify sex-specific latent profiles of self-harm in a nationally representative Korean adolescent sample, and (2) examine how these profiles differ in associations with suicidal ideation, planning, and attempts.

Methods: Data were drawn from the 2021 Korean National Youth Policy Institute National Survey on Mental Health of Youth (N = 6,689), collected via stratified cluster sampling. Self-harm was assessed using five dichotomous behavioral indicators. Suicidality was measured using the Mental Health Screening for Suicide Risk (MHS:S). Sex-stratified latent class analysis (LCA) derived typologies separately for males and females. Associations were estimated using negative binomial regression (IRRs) and multinomial logistic regression (RRRs), adjusting for demographic, psychological, family, school, and community covariates.

Results: LCA identified three classes among males: Low Self-Harm (80.7%), Self-Hitting (11.9%), and Multiple Self-Harm (7.4%). Among females, four classes emerged: Low Self-Harm (80.7%), Self-Hitting (5.8%), Body-Focused Injury (8.4%), and Multiple Self-Harm (4.9%), with the Body-Focused Injury class—characterized by cutting and scratching to the point of injury—identified exclusively among females. Regression analyses revealed a critical sex difference in how self-harm typologies relate to suicide risk. Among males, Self-Hitting was associated with elevated suicidality (adjusted IRR = 2.14); however, its RRR was non-significant after adjustment, indicating that membership in the Self-Hitting class did not independently elevate relative suicide risk above the reference group. This pattern suggests that elevated suicidality among male self-hitters may reflect pre-existing psychosocial burden rather than self-hitting itself driving suicidal risk—consistent with self-hitting functioning as an expression of diffuse psychological distress. Among females, by contrast, all self-harm classes—including Self-Hitting—showed significantly elevated RRRs, indicating that class membership itself independently conferred elevated suicide risk regardless of typology. The Body-Focused Injury class exhibited the highest odds of suicide attempts (adjusted OR = 5.97, 95% CI [4.05, 8.80]), followed by Multiple Self-Harm (adjusted OR = 4.97).

Conclusions and Implications: These findings reveal fundamentally different functional meanings of self-harm by sex. In male adolescents, self-hitting may serve as a behavioral signal of accumulated psychosocial distress—including family conflict, school violence exposure, and suppressed help-seeking—rather than an independent marker of escalating suicidal risk, underscoring the need for comprehensive psychosocial assessment over automatic risk escalation protocols. In female adolescents, any self-harm disclosure independently signals elevated suicide risk, warranting immediate structured assessment, with body-focused injury representing the highest-priority clinical indicator. Within multi-tiered systems of support, these sex-specific typologies should function as actionable decision points. Social workers are uniquely positioned to embed method-specific, sex-sensitive self-harm inquiry into screening protocols, dismantle gendered barriers to disclosure, and train practitioners to interpret typological risk—enabling more equitable and precise early intervention for the highest-risk youth.