Mental Health Profiles and Recidivism in Juvenile Justice-Involved Youths
Background: Youths with mental health disorders are overrepresented in the juvenile justice system (JJS) and at a higher risk for future JJS involvement (recidivism), compared to youths without mental health disorders. Youths with high mental health symptomatology tend to have a more extensive history of JJS involvement, but the relationship between different mental health profiles and future recidivism has been minimally examined. The literature on JJS-involved youths with mental health or substance use disorders suggests that interventions targeting these needs may decrease the risk of recidivism, but research documenting the effectiveness of mental health interventions in reducing recidivism rates among youth with mental health disorders is scarce. Purpose: The research questions addressed in this study are: (1) Can the youths in this sample be classified into distinct groups, based on their mental health symptomatology? (2) Are there any demographic differences in the distribution of mental health profiles? (3) Does mental health group membership predict whether a youth recidivates and the type of recidivating offense? (4) Is the relationship between mental health group membership and recidivism moderated by the completion of a mental health program? Methods: The data used in the study consisted of records of 9,694 JJS-involved youths, collected between 2017 and 2022 at a large urban juvenile department in Texas. The youths were 75.5% boys, 43.6% Hispanic, 44.8% Black, and 11.6% non-Hispanic White. The measures used in this study were the Massachusetts Youth Screening Instrument-2 (MAYSI-2), two indicators of recidivism, one indicator of completion (or noncompletion) of a mental health program, and demographic information. Statistical analyses included (1) a latent profile analysis using the scores for the MAYSI-2 scales to identify groups of youths with similar patterns of mental health needs, (2) two logistic regression analyses predicting whether a youth recidivated and the type of recidivating offense, and (3) a logistic regression analysis examining the moderator role of completing a mental health program. Results: The results of the latent profile analysis showed that a model consisting of three classes of mental health needs fitted the data the best. The groups that emerged were a low (59.8%), moderate (26.6%), and high mental health needs group (13.7%). The youths in the moderate needs group were more likely to recidivate in general, compared to youths in the low and high needs groups and they were also more likely to recidivate with certain types of offenses (i.e., against person, drug, and weapon offenses). There was no association found between the completion of a mental health program and recidivism rates. Conclusion: The results supported an association between mental health needs and recidivism for the moderate needs group and revealed that this relationship is more nuanced for youths with high needs. The extant literature and present study have provided mixed or inconclusive results regarding the effectiveness of interventions in reducing mental health outcomes and recidivism rates for youths in the JJS.