Recurrent non-suicidal self-injury in depressed youth with mixed features: a 6-month prospective cohort study

Objective This study aimed to examine whether mixed features during a current major depressive episode (MDE) are associated with increased risk of non-suicidal self-injury (NSSI) among adolescents and young adults, focusing on both first-onset and recurrent NSSI during a 6-month follow-up period. Method A total of 713 individuals aged 13–25 years with current MDE were recruited, including 233 with mixed features. NSSI was assessed at baseline and at 1-, 3-, and 6-month follow-ups. Multiple imputation was used to handle missing data ( n = 626). Kaplan–Meier and Nelson–Aalen estimators were applied to visualize time-to-event and cumulative risk curves. Cox regression assessed first-onset NSSI, and Andersen–Gill models estimated the risk of repeated events. Rubin’s rules were used to pool estimates across imputed datasets. Sensitivity analyses were performed using complete-case data after multiple imputation, while subgroup analyses were conducted using stratified models. Results Participants with mixed features were more likely to be female and to report a shorter illness duration and aggression history. Mixed features were associated with earlier NSSI onset ( p = .010) and higher cumulative risk ( p < .001). Although no significant association was found with first-onset NSSI, mixed features significantly predicted recurrent NSSI in both imputed (HR = 1.35, p = .045) and complete-case models (HR = 1.58, p < .001). The effect was stronger among first-episode cases and those with illness duration < 6 months. Conclusion Mixed features in adolescent and young adult MDE may serve as a predictor of recurrent NSSI. Early identification and tailored monitoring strategies are warranted to reduce self-injury risk.