Community and Social Support Buffers Suicide Risk: Moderated Mediation of ADHD Tendencies and Suicidality Among Korean School-Dropout Adolescents

Background and Purpose: This study examined whether community and social support moderate indirect pathways from ADHD tendencies to suicidal ideation and planning through depression, anxiety, and stress among Korean school-dropout adolescents. ADHD tendencies impair attention and emotion regulation, increasing risks for school disengagement and internalizing problems, which further deepen social isolation and suicide risk. Despite these risks, prior research has examined ADHD-suicidality and dropout-suicidality separately. Guided by ecological systems, interpersonal suicide, and stress-buffering theories, we hypothesized that these mediated pathways would be attenuated at higher levels of community and social support.

Method: This study used data from the 2021 Korean National Adolescent Mental Health Survey, a cross-sectional sample of 752 adolescents from 220 Ggoomdream Centers nationwide. ADHD tendencies were measured via the CASS:S, a 27-item scale reflecting self-reported tendencies, not clinical diagnoses. Depression, anxiety, and stress served as distinct mediators, scored on 5-point Likert averages. Although inter-correlations were high(r = .54–.90), mediators entered models separately to ensure discriminability and prevent multicollinearity. Community and social support—perceived availability of trusted persons, crisis helpers, and accessible institutions beyond family and Ggoomdream Centers—were the b-path moderator, measured as a 3-item 4-point Likert average. Suicidal ideation and planning were assessed via single items. Gender, SES, and age-group were controlled. Six PROCESS Model 14 models with 5,000 bootstrap samples were estimated across mediators and outcomes, interpreted considering multiple testing.

Results: ADHD tendencies significantly predicted all three mediators(all p < .001): depression(B = .041, R² = .36), anxiety(B = .045, R² = .35), and stress(B = .022, R² = .17). All interaction terms and Index of Moderated Mediation values were significant, with bootstrap 95% CIs excluding zero, confirming moderated mediation. As community and social support increased from the 16th to 84th percentile, depression-mediated indirect effects declined by approximately 40% and anxiety-mediated effects by 43–52%, both remaining significant at high support. Stress-mediated effects declined by 88% for suicidal ideation and 96% for suicidal planning, becoming non-significant at high support, though this largely reflects the stress pathway's smaller a-path coefficient(B = .022 vs. .041–.045). Regarding mediation type, the depression model indicated indirect-only mediation for suicidal ideation(direct effect p= .580, ns); depression–planning and anxiety–ideation models indicated near-complete mediation(p= .096, .063); remaining models demonstrated partial mediation with significant direct effects.

Conclusions and Implications: Findings suggest that community and social support substantially buffer suicide risk among school-dropout adolescents with ADHD tendencies. Stress-mediated pathways were nearly fully attenuated at higher support levels, suggesting that accessible public support systems can serve as a baseline safety-net. However, depression- and anxiety-mediated risk was not fully buffered, underscoring the necessity of developing targeted clinical interventions beyond community-level care. Socio-emotional programs and peer networks may further strengthen such support. Social workers in community centers should screen for ADHD tendencies and depressive symptoms upon dropout and deliver informed crisis intervention for this underserved population. Although cross-sectional design limits causal inference and longitudinal research is needed, these findings highlight that community support and targeted mental health services should be integrated for high-risk adolescents.