The Short-Term Course of Nonsuicidal Self-Injury Among Individuals Seeking Psychiatric Treatment
Glenn Kiekens,Laurence Claes,Evan M Kleiman,Koen Luyckx,Daniel D L Coppersmith,Rebecca G Fortgang,Inez Myin-Germeys,Matthew K Nock
DOI: https://doi.org/10.1001/jamanetworkopen.2024.40510
2024-10-01
Abstract:Importance: A major portion of adolescents and adults seeking psychiatric treatment report nonsuicidal self-injury (NSSI) within the past month, yet the short-term course of NSSI among these patients remains poorly understood. Objective: To advance the understanding of the short-term course of NSSI cognitions (ie, thoughts, urges, and self-efficacy to resist self-injury) and behavior. Design, setting, and participants: A cohort study was conducted using an intensive longitudinal design with ecological momentary assessment, including 6 daily surveys and event registrations of self-injury for 28 days. Data were collected from June 2021 to August 2023. Individuals using mental health services in the Flanders region in Belgium reporting past-month NSSI urges and/or behavior at intake were recruited by referral. Main outcomes and measures: Nonsuicidal self-injury thoughts, urges, self-efficacy, and behavior. Sociodemographic and clinical baseline characteristics served as between-person variables. Descriptive and variability statistics and dynamic structural equation modeling were used. Results: Participants completed a mean (SD) of 121 (34.5) surveys, totaling 15 098 longitudinal assessments (median adherence, 78.6%; IQR, 59.5%-88.7%). Among 125 patients (87.2% female; median age, 22.0 [range, 15-39] years; 52.8% heterosexual), NSSI thoughts and urges were present during most assessments but were low in intensity (individual means [SD] on a 0- to 6-point scale, 1.52 [1.13] for urges; 1.57 [1.18] for thoughts). The prevalence of NSSI behavior was 84.0% monthly, 49.90% weekly, and 18.19% daily. Between-patient variability was substantial (intraclass correlation coefficient, 0.43-0.47; range of individual means for cognitions, 0-6; individual frequency behavior, 0-103), with recency and frequency of NSSI thoughts and behavior at intake consistently associated with individual differences in the course of NSSI. The greatest variability was observed within patients (root mean square of successive differences from 1.31 for self-efficacy to 1.40 for instability of thoughts), characterized by changes in the intensity of cognitions by more than 1 within-person SD between assessments less than 2 hours apart in 1 of 5 instances. Nonsuicidal self-injury behavior is rare in the morning, increased in the afternoon, and most frequent in the evening. Nonsuicidal self-injury cognitions were contemporaneous and temporally associated with each other's course, with higher-than-usual thoughts and lower self-efficacy uniquely signaling heightened risk for NSSI behavior in the next 2 hours. Conclusions and relevance: In this cohort study of treatment-seeking individuals, NSSI cognitions and behavior appeared to be dynamic over the short term. These findings suggest the potential utility of self-monitoring outside the therapy setting and the need to focus assessment and interventions on the evening hours.