Predicting adult mental health from non-suicidal self-injury in adolescence: a prospective study spanning 2007-2023

This study examined associations between adolescent non-suicidal self-injury (NSSI) and adult mental health within a cohort of school students who answered questionnaires in 2007 (T1: N = 992, M_age = 13.7, 50.1% girls) and 2008 (T2: N = 987, M_age = 14.9, 50.9% girls), with follow-ups in 2017 (T3: N = 557, M_age = 25.3, 58.7% women) and 2023 (T4: N = 386, M_age = 29.9, 63.0% women). Attrition was accounted for with sample weighting and mixed modelling, and robust estimates balance for residual non-normality. Significant declines in any NSSI (≥ 1 instance) and repetitive NSSI (≥ 5 instances) were observed between T2-T3 (any NSSI: 43-19%; repetitive NSSI: 21-10%; p <.001), but not T3-T4 (any NSSI: 19% to around 14%; repetitive NSSI: 10-5%; p ≥.078). Compared to reporting no NSSI in adolescence, stable repetitive NSSI (≥ 5 instances at T1 and T2), unstable repetitive NSSI (≥ 5 instances at T1 or T2), and infrequent NSSI (1-4 instances at T1 and/or T2) were significantly associated with any NSSI engagement (OR = 2.63-6.67) and elevated depression, anxiety, and stress levels at T3/T4 (β_robust = 0.14-0.39), while controlling for gender and adolescent psychological difficulties. For stable repetitive NSSI, this applied to emotion dysregulation as well (β_robust = 0.44). Additionally, stable/unstable repetitive NSSI was significantly associated with lower well-being when not controlling for gender and psychological difficulties (β_robust =-0.32--0.41). The findings indicate that adolescent NSSI, especially when repetitive and stable, predicts poorer health across the first decade of adulthood.