Childhood trauma and non-suicidal self-injury in children and adolescents with depression: exploring the mediating role of shame and gender differences

Abstract

Background: Non-suicidal self-injury is highly prevalent among children and adolescents with depression, yet the underlying psychosocial mechanisms remain poorly understood. This study investigated the associations between childhood trauma, psychological symptoms, shame, and gender, and examined mediating and moderating pathways linking childhood trauma to NSSI in this population. Methods: In this cross-sectional study, we assessed NSSI, childhood trauma (CTQ), depressive symptoms (HAMD), anxiety symptoms (HAMA), and shame (SS) in children and adolescents diagnosed with depressive disorder. Correlation analyses, logistic regression, and moderated mediation models were applied to explore direct and indirect links between childhood trauma and NSSI, with shame as a mediator and gender as a moderator. Results: Among 177 children and adolescents with depression, 69.5% reported NSSI, with a significantly higher proportion of females in the NSSI group (90.2%) compared to the non-NSSI group (63.0%). Children and adolescents with NSSI had higher scores on childhood trauma, depressive symptoms, anxiety, and shame (p < 0.001 for all). Logistic regression showed male gender as a protective factor (OR = 0.18, p < 0.001), while childhood trauma (OR = 1.04), depressive symptoms (OR = 1.15), anxiety (OR = 1.17), and shame (OR = 1.06) were risk factors (p < 0.01 for all). Mediation analysis revealed that shame mediated the trauma-NSSI link (indirect effect = 0.016, 95% CI [0.004, 0.033]); This effect was statistically detectable in females (indirect effect = 0.02, 95% CI [0.01, 0.04]) but not in males (indirect effect = 0.00, 95% CI [− 0.02, 0.03]), consistent with limited male sample size (n = 32). Conclusions: Childhood trauma, psychological symptoms, and shame were associated with NSSI in children and adolescents with depression. Shame mediated the trauma–NSSI relationship in the full sample. This mediation effect was detectable in females but not in males, consistent with the greater statistical power afforded by the larger female subgroup. These findings support clinical prioritization of shame-reduction strategies for depressed youth presenting with trauma exposure and elevated shame.