Mental health among correctional populations: Predictors, mechanisms, and long-term trajectories of outcomes

Mental health problems are more common among correctional population relative to the general population, presenting an important relating to public health and safety. To have a comprehensive understanding towards mental health of correctional populations, this dissertation comprises three separate studies that investigated the predictors, mechanisms, and long-term trajectories of mental health outcomes.
Study 1 investigated the associations of different forms of trauma with mental disorders among prisoners and ex-prisoners. This meta-analysis identified studies published from 1998 to 2021 by searching PsycINFO, PubMed, Medline, and Web of Science. 62 studies were analyzed containing 15,115 prisoners (97.86%) and 330 ex-prisoners (2.14%) across 16 countries. The multilevel meta-analysis found that overall trauma was positively associated with more diagnoses or symptoms of mental disorders (Zr=0.198, 95% CI=[0.167, 0.229], pStudy 2 addressed the behavioral mechanisms for post-release mental health in the context of everyday life of ex-prisoners. Maladaptive adjustment to post-incarceration life is related to higher chances of common affective disorders and recidivism. Currently, little is known about post-release daily adaptation, not to mention valid and reliable instruments for assessing post-release daily routines pertinent to mental health. This study developed and validated a self-report instrument, hereafter referred to as Post Release Living Inventory for Ex-prisoners (PORLI-ex). Three separate samples of ex-prisoners were recruited to complete an online survey (N=1,277). The final model evidenced acceptable goodness-of-fit and consisted of 45 items on nine dimensions, which loaded on three second-order factors: Consolidation (three dimensions; e.g., Institutional Routines), Replacement (two dimensions; e.g., Maladaptive Behaviors), and Addition (four dimensions; e.g., Socializing with Exprisoner Friends) (α=.695–.915). Convergent validity, discriminant validity, criterion-related validity, and incremental validity were demonstrated.
Study 3 enriched the previous two studies by focusing on trajectories of vulnerability and resilience and multilevel predictors for the trajectories among juvenile delinquents following the conviction of serious crime. This study used Growth Mixture Modeling (GMM) to identify 7-year longitudinal trajectories of probable anxiety and probable depression among juvenile offenders who were first convicted for serious crimes and without detention history (N=574). Lasso logistic regression was adopted to select multilevel factors for predicting memberships of resilience (vs. non-resilience) and chronicity (vs. nonchronicity). Recovery refers to initially high but gradually declining symptoms and was combined into non-resilience group or non-chronicity group in Lasso logistic regression analysis. Three identical trajectories were found for both anxiety and depression: resilience (75.78% to 75.96%), chronicity (10.98% to 15.16%), and recovery (8.89% to 13.24%). Risk factors for resilience included: person-level factors (e.g., neuroticism and exposure to violence), relationship-level factors (peer antisocial behaviors and parental hostility), and context-level (e.g., chaotic neighborhood conditions and father’s remarriage). Resilience factors included self-identity, work efficacy, perceived opportunities for work, and community involvement. Predictors of the chronicity included neuroticism, drugs use in the past six months, and father’s remarriage. All rights reserved.