The Impact of Mental Health Risks and Needs on Substance Use in Community Supervision: A Multilevel Model Examining Moderation By Age

Background and Purpose: Nearly 4 million adults are on probation or parole in the United States and rates of substance use disorders (SUDs) are disproportionately higher among all justice-involved adults. While SUDs tend to co-occur with mental illness, few studies have examined SUD/mental health co-occurrence (CODs) among adults on probation/parole and how these experiences vary across the life course. For younger adults, CODs may be attenuated due to earlier experiences with treatment. However, for older adults, co-occurrence may be more severe, given the prolonged time since symptom onset. To examine this topic, we examine a statewide sample of persons on probation and parole to examine the effects of SUD and mental health needs across the lifespan.

Methods: Administrative data, provided by the State of Georgia Department of Community Supervision included all adults entering probation or parole between June 2018 and May 2019 (N= 33,079). Mental health needs and substance use risks and needs were assessed through two scales based on Bonta and Andrews’ (2007) risk/needs/responsivity model. Additional needs involving employment/education, peer and family, and criminal thinking were assessed in a similar manner. Demographic measures included age, gender, and race/ethnicity. Also measured were dichotomous indictors for criminal justice involvement including revocations, reincarceration and reconviction while on probation/parole. We also examined a dichotomous indicator on whether probation or parole sentence was related to drug offenses. Given the statewide nature of the sample, multilevel modeling was used to examine associations with substance use risks and needs while controlling for variation across judicial circuits. This study was reviewed and approved by the lead author’s university IRB.

Results: The multilevel model showed that substance use risk and need significantly varied across judicial circuits (p<.01). Results also showed substance use risk and need was significantly associated with mental health (b=0.62, p<.01) and age (b=0.08, p<.01). Employment/education (b= 0.18, p<.01), peer and family (b= 0.2, p<.01), reincarceration (b= 0.75, p<.01) and reconviction (b= -0.19, p<.01), as well as being male (b= 0.89, p<.01) and non-White (b= -0.41, p<.01) were all significantly associated with substance use risk and need. Criminal thinking and revocations were not associated with substance use. Finally, results revealed a significant interaction term between age and mental health (b=.01, p<.001). Decomposing the interaction showed that older adults had a significantly stronger relationship between substance use risk and need and mental health risk and need, compared to younger adults.

Conclusions and Implications: Effects of mental health on substance use risks and needs among persons on probation or parole vary regionally, and older persons on probation or parole show the highest rates of mental health and substance use treatment needs. Social workers in criminal justice settings may benefit from these findings, which suggest a need for increased awareness of potential co-occurring needs among older adults on probation or parole, and available community-based interventions to meet these needs. Research to evaluate training and available resources in the delivery of care for those with CODs who are on probation or parole remains a serious need in criminal justice research.