This nationally representative sample of U.S. adults with diabetes, 500 respondents (weighted to represent about 1 million adults) had a comorbid substance use disorder. Moreover, the presence of both conditions is highly concentrated among those with prior CJI. Adults with diabetes and prior CJI were disproportionately members of minority groups and of lower socioeconomic status and were more likely to have Medicaid or be uninsured. Finally, our multivariate regression analyses found that prior CJI in adults with diabetes is highly associated with multiple forms of substance use disorder, findings that persist after controlling for sociodemographic and clinical confounders.

Our findings contextualize the small amount of existing literature on comorbid diabetes and substance use disorder. Prevalence studies have generally found that substance use is lower among those with diabetes compared with the general population, but that it undermines treatment when present.³ Only one other nationally representative study has examined the rate of alcohol use disorder among patients with diabetes; it found a prevalence rate of 1% (compared to a 3.3% prevalence in our study).¹⁹ To our knowledge, this is the first study describing prevalence of opioid use disorder in U.S. adults with diabetes. That substance use prevalence is particularly concentrated among those with prior CJI is not surprising (CJI is known to be strongly associated with substance use⁷), but these findings suggest that those with diabetes and CJI experience substance use disorders at rates higher than the general population.

Moreover, these findings taken in sum with prior data showing the harmful effects of substance use on diabetes outcomes shed light into potential areas for effective intervention.

Compared to other substance use disorders, alcohol and opioid use disorder (the majority of cases in this study) have the best evidence base for effective treatment regimens; however, there is a paucity of literature examining interventions which effectively treat these substance use disorders along with comorbid diabetes.²⁰ One study examined contingency management, the provision of tangible positive reinforcement for abstinence from substances, in alcohol use disorder and found that patients with diabetes were more likely to achieve abstinence but did not examine diabetes outcomes.²¹ To our knowledge, there are no studies examining the effect of evidenced-based pharmacotherapy for alcohol use disorder such as naltrexone, disulfiram, or acamprosate in patients with diabetes.⁶

One small study in Canadian First Nations people with diabetes and opioid use disorder treated with buprenorphine-naloxone, a partial opioid agonist used to treat opioid use disorder (OUD), found substantial improvement in glycosylated hemoglobin in the treatment group.²² However, this study is not generalizable and only included 66 participants. Preclinical data suggest that methadone, a full opioid agonist used to treat OUD, may decrease glucose tolerance.⁶ Therefore, buprenorphine may be the preferable treatment option in candidates for opioid agonist therapy with diabetes, although additional clinical trials are warranted.

Our study shows that considering CJI as a social risk factor may facilitate treatment delivery in this population. However, current substance use treatment approaches for those currently involved in the legal system rarely align with best practices; there is no evidence finding that contact with the criminal justice system leads to improved access to effective, evidenced based care. For example, a recent study examining survey of court-mandated addiction treatment programs found that less than 10% record meaningful health-related outcomes.²³ Another study found that only 5% of adults in the system are offered the first-line treatment for opioid use disorder, opioid agonist therapy.²⁴ Most importantly, evidence suggests that a treatment-based approach is likely safer and more cost-effective than criminalization for managing substance use disorder.²⁵

Our findings taken in sum with what was previously known about the negative impact of substance use on diabetic control underscore an urgency in prioritizing treatment for substance use disorder for those with comorbid chronic disease such as diabetes within criminal justice settings.

Although no studies have specifically examined interventions for those with diabetes and addiction after exposure to the justice system, a basis of evidence suggests that a primary care-based model may hold promise. Interventions for justice-involved patients reentering into the community, such as the primary care-based Transitions Clinic Network,²⁶ may improve outcomes for those with both medical and behavioral health conditions. Similar interdisciplinary models, not specifically for the justice-involved, have been shown to improve outcomes when treating patients with several comorbid conditions.²⁷ Based on our results, development of these interventions should consider how to incorporate factors related to prior CJI, and how this influences diabetes self-care behaviors and outcomes.

We acknowledge several limitations of our study. First, in this cross-sectional sample of data, causal association cannot be established. Second, because we used self-reported data on diabetes status to create our sample, those with undiagnosed diabetes are not included in the sample, and given the lack of health care access for CJI populations may suggest that CJI was overrepresented in these missed cases.²⁸ However, this bias would render our findings more conservative. Third, the CIs were wide for some estimates, due to the small sample size; therefore, these estimates should be interpreted with caution. Finally, the CJI measure is also self-reported and may be underreported; however, other studies have shown this to be a reliable method for attaining such information.²⁹

In conclusion, over one million U.S. adults are living with both diabetes and substance use disorder, and coprevalence is concentrated among those with CJI. CJI was strongly and independently associated with substance use outcomes even after controlling for multiple sociodemographic and medical covariates. Leveraging resources to improve access to both SUD and diabetes treatment is needed to improve health outcomes in the justice-involved population.