Post-attempt NSSI patterns and timing of reattempt: Descriptive trends and adjusted hazard over 12 months

BACKGROUND: Non-suicidal self-injury (NSSI) in adults after a suicide attempt may follow distinct courses with different clinical profiles and reattempt risk. AIMS: Examine NSSI patterns, correlates, and associations with suicide reattempts over 12 months. METHOD: Adults (n = 685; 72.4% female; mean age 42 [SD 15.45]) were assessed ≤15 days post-index attempt and followed 12 months. NSSI was assessed using Columbia-Suicide Severity Rating Scale (C-SSRS). Baseline assessments captured psychiatric diagnoses, global psychological distress, impulsivity, fearlessness about death, and childhood trauma (assessed with the MINI, BSI, BIS-11, ACSS-FAD, and CTQ-SF, respectively), and suicide-related variables; reattempts were tracked throughout follow-up. RESULTS: Four a priori patterns: NN (No NSSI→No NSSI; 53.4%), Onset (No→Yes; 5.7%), Remission (Yes→No; 23.9%), and Persistence (Yes→Yes; 16.9%). Persistence was associated with younger age and higher baseline impulsivity, higher baseline global psychological severity, and greater childhood trauma severity along with substance use before the attempt and higher acquired capability. PTSD predicted reattempt within Persistence. Onset was linked to earlier reattempts, narrowing the intervention window, whereas Persistence showed a modest descriptive delay, consistent with short-term regulation (without implying protection). During follow-up, 22.2% reattempted; median time-to-reattempt 138 days (IQR 55-222; mean 149). In multivariable Cox regression, older age showed lower hazard (HR 0.98, 95% CI 0.96-0.99), and a greater number of prior suicide attempts (HR 1.02, 1.01-1.04). Relative to NN, Remission (HR 0.56, 0.32-0.96) and Persistence (HR 0.58, 0.35-0.96) were associated with reduced 12-month hazard. CONCLUSIONS: While NSSI may briefly down-regulate severe affect, Persistence signals elevated clinical risk and acquired capability; Onset appears linked to earlier reattempts.