Sexual Orientation, Mental Illness, and Substance Use Disorders Among Criminal Legal System–Involved Individuals

Approximately 6.8 million individuals (2.8% of the adult population) are under correctional supervision on any given day in the United States, including those under supervision in jails and prisons, on parole, or on probation (1–3). Researchers estimate that 70% of inmates experience chronic disabilities, mental health problems, or substance use disorders, with diabetes, tuberculosis, HIV/AIDS, and depression being present at higher levels among inmates than among individuals in the general population (4, 5). However, poor access to health care is a pervasive problem for many legal system–involved individuals who experience challenges navigating the health care system, disruption of medications during incarceration, and barriers to behavioral health services in community-based settings (6–8). Given the closures of mental health institutions and the decrease in psychiatric inpatient beds during the past 50 years, many jails and prisons have become the primary mental health system for many under- or uninsured people with co-occurring mental health and behavioral health problems (9, 10).

Lesbian, gay, bisexual, transgender, and queer or questioning (LGBTQ) adults represent approximately 4%–5% of adults nationwide, according to population-based and self-reported survey data (11). Findings from previous research suggest that individuals from sexual minority groups are more likely to have substance use and mental disorders because of exposure to structural and interpersonal discrimination (12–15). A study conducted in 2015 by the Substance Abuse and Mental Health Services Administration (SAMHSA) with data from the National Survey on Drug Use and Health (NSDUH), a cross-sectional and nationally representative annual survey of the noninstitutionalized civilian population, found that, compared with their heterosexual counterparts, individuals from sexual minority groups were more likely to be current cigarette smokers or alcohol consumers, to have a substance use disorder (related to alcohol, marijuana, or pain medications), and to have a need for substance use treatment (16). Previous research also found that, compared with their heterosexual counterparts, people belonging to sexual minority groups were more likely to have serious mental illnesses and major depressive episodes and were therefore more likely to receive mental health services in the past 12 months (16). Yet seeking treatment for behavioral health conditions has also been shown to place LGBTQ individuals at risk for experiencing provider-based discrimination or stigma, which can exacerbate existing conditions (16).

Public health research at the intersection of criminal legal involvement and LGBTQ identity is scarce. Results from previous studies suggest that adults identifying within the LGBTQ spectrum are incarcerated at higher rates than their heterosexual and cisgender peers. For example, data from the National Inmate Survey found that the incarceration rate of self-identified lesbian, gay, and bisexual people was 1,882 per 100,000 individuals, more than three times that of the general U.S. adult population (17). Additionally, 33.3% and 26.4% of women in prisons and jails, respectively, identified as lesbian or bisexual, which is eight to 10 times larger than the proportion of women who identify as lesbian or bisexual in the general adult population (17). To the best of our knowledge, no comprehensive, population-based study has analyzed mental health and substance use disorder disparities in the sexual minority population with previous criminal legal involvement. Our study fills important research gaps by using nationally representative data to examine the prevalence of mental health conditions, substance use disorders, and behavioral health treatment for legal system–involved lesbian, gay, and bisexual adults. Knowing the health needs of this population will better inform public health and clinical practice to improve support for legally involved LGBTQ people.

Methods

Data Source and Study Sample

This study used data from the 2015–2019 NSDUH. The survey is sponsored by SAMHSA, with a response rate of >70%. The NSDUH is routinely used to measure and monitor the prevalence of substance use and mental disorders in the United States (18). We restricted our sample to individuals ages ≥18 years because they were asked specific questions about criminal legal history, sexual orientation, and behavioral health outcomes (19).

Starting in 2015, the NSDUH questionnaire added a question about sexual orientation identity. Respondents were asked which of the following categories best represents how they identify themselves: heterosexual, that is, straight; lesbian or gay; or bisexual (19). Sexual minority individuals were defined as those who reported identifying as gay, lesbian, or bisexual. Although the NSDUH asks for further information about sexual attraction, we chose to focus this analysis on differences by sexual orientation identity among legally involved individuals. Individuals involved with the criminal legal system were defined as those who reported being arrested or booked, paroled, or on probation in the 12 months before the survey interview. The NSDUH defines “booked” as “taken into custody and processed by the police or by someone connected with the courts, even if you were then released.” Consistent with previous research (8, 20), and to maintain a reference time frame consistent with all outcomes examined in this study, we did not take into account whether a person had an interaction with the criminal legal system before the 12-month reference window. Our analysis excluded individuals who did not answer questions related to criminal legal involvement (N=2,080) and individuals who did not answer questions related to sexual orientation identity because they did not know (N=1,364), refused to answer (N=2,060), or did not answer for other reasons (N=119). Our final sample (N=195,239 heterosexual adults and N=14,995 sexual minority adults; total N=210,234) included 187,250 heterosexual adults with no legal involvement in the previous 12 months, 14,066 sexual minority adults with no legal involvement in the previous 12 months, 7,989 legally involved heterosexual adults; and 929 legally involved sexual minority adults.

Mental Illness, Substance Use, and Treatment Outcomes

We examined outcomes that represent the wide spectrum of mental illness and substance use. All variables and outcomes were based on self-reports. Any substance use or dependence was a composite indicator of abuse of or dependence on cocaine, heroin, nonmedical prescription opioids, or stimulants in the past 12 months. Nine substance use disorders (i.e., alcohol, cocaine, hallucinogen, inhalant, pain reliever, heroin, sedatives, marijuana, and tranquilizers) were defined by the NSDUH and were assessed according to whether the participant met the DSM-IV criteria for substance use disorder within the past year (3, 18, 19). Suicidal ideation was defined as seriously thinking about killing oneself in the past 12 months, suicidal plan was defined as making plans to kill oneself in the past 12 months, and suicide attempt was defined as trying to kill oneself in the past 12 months. Major depression was defined as having a major depressive episode in the past year, as established by the assessments and diagnostic criteria of the DSM-IV (21, 22). Mental illness severity was based on a prediction model used by SAMHSA to measure severe mental illness prevalence with Kessler’s six-item instrument to measure psychological distress in the previous 30 days and the World Health Organization’s Disability Assessment Schedule (19). Finally, treatment for substance use or mental illness included any outpatient, inpatient, or prescription medication–based treatment in the past 12 months.

Statistical Analysis

We used descriptive statistics to characterize the study sample and to compare the prevalence of each outcome of interest by sexual minority status and criminal legal involvement status. Next, we estimated multivariable logistic regression models to compare the odds of each outcome between individuals who identified as belonging to a sexual minority group and their heterosexual peers, by criminal legal–involvement status, while controlling for sociodemographic characteristics. All logistic regression models controlled for age in years (18–25, 26–34, 35–49, ≥50), race-ethnicity (non-Hispanic White, non-Hispanic Black, Hispanic, and other or multiple races), relationship status (married; divorced, separated, or widowed; never married), the presence of children in the household (yes or no), educational attainment (less than high school, high school graduate, some college, college graduate), employment status (employed full-time, employed part-time, unemployed, other), household income (<$20,000, $20,000–$49,999, $50,000–$74,999, ≥$75,000), self-rated health status (excellent, very good, or good vs. fair or poor), and health insurance status (insured or uninsured). Results from the logistic regression models are presented as adjusted odds ratios (AORs) with 95% confidence intervals. We conducted all analyses in Stata, version 16.1, by using survey weights and the svy command to adjust standard errors for the complex survey design of the NSDUH. This study was deemed exempt from review by the Vanderbilt University Institutional Review Board because data were obtained from deidentified, publicly available, and secondary sources.

Results

Sociodemographic Characteristics

After we applied survey weights, approximately 2.9% (N=7,989) of heterosexual and 5.4% (N=929) of sexual minority individuals were found to have criminal legal involvement in the 2015–2019 NSDUH. Approximately 9% (N=929) of legal system–involved individuals considered themselves gay, lesbian, or bisexual; 91% (N=7,989) considered themselves heterosexual (data are weighted and available on request). Table 1 presents the sample’s sociodemographic characteristics by criminal legal involvement and sexual minority status. Most legally involved heterosexual adults were male (73.3%), 18–34 years old (50.9%), non-Hispanic White (54.0%), never married (52.7%), and in families earning <$50,000 annually (69.3%). Most legally involved sexual minority individuals were bisexual (68.5%), and fewer were lesbian or gay (31.5%). The demographic data showed several trends (p values indicating statistical significance for percentage comparisons are available on request). Compared with their legally involved heterosexual peers, legally involved sexual minority individuals tended to be younger (26–34 years old, 38.5%), from a racial-ethnic minority group (non-Hispanic Black, 22.6%; Hispanic, 19.9%), never married (67.4%), highly educated (graduated from college, 10.3%), unemployed (17.5%), in families earning <$20,000 per year (41.0%), uninsured (21.8%), and in worse overall health (fair or poor health status, 21.5%).

Sexual minority individuals and heterosexuals with no criminal legal involvement shared similar patterns, with some notable exceptions. Most sexual minority individuals with no criminal legal involvement were bisexual (61.5%), but more than one-third (38.5%) were gay or lesbian. Among individuals with no criminal legal involvement and compared with their heterosexual counterparts, sexual minority individuals tended to be younger (18–25 years old, 30.6%), from a racial-ethnic minority group (non-Hispanic Black, 13.0%; Hispanic, 18.1%), never married (62.3%), less likely to have a child (32.5%), highly educated (some college or higher education, 67.3%), unemployed (7.1%), in families earning <$20,000 a year (22.2%), and in worse health (fair or poor health status, 15.2%).

Substance Use Outcomes

Table 2 presents prevalence estimates and multivariable logistic regression results for substance use disorders and mental illness by criminal legal involvement and sexual minority status. Approximately 41% of legally involved sexual minority individuals had a substance use disorder in the past year, with alcohol use disorder being the most prevalent (25.6%), followed by marijuana use disorder (10.7%). After controlling for sociodemographic characteristics, we detected no significant differences between legally involved heterosexual individuals and sexual minority individuals for sedative use disorder, tranquilizer use disorder, heroin use disorder, or pain medication use disorder. However, legally involved sexual minority individuals were marginally (p<0.100) more likely to report alcohol use disorder (AOR=1.30, p=0.050) and marijuana use disorder (AOR=1.37, p=0.050) and significantly (p<0.050) more likely to report cocaine use disorder (AOR=1.93, p=0.020), hallucinogen use disorder (AOR=2.49, p=0.030), and inhalant use disorder (AOR=5.50, p=0.002) compared with their legally involved heterosexual peers. Moreover, among adults without any criminal legal involvement in the previous 12 months, sexual minority individuals were more likely than their heterosexual peers to report disorders for all substances, except for heroin.

Mental Health and Illness Outcomes

Table 2 also presents comparisons of mental illness by legal involvement and sexual minority status. Approximately 50% of criminal legal–involved sexual minority individuals and 33% of criminal legal–involved heterosexual individuals reported indicators for mental illness. The prevalence of suicidal ideation (21.1% vs. 9.2%), suicidal plans (11.2% vs. 3.8%), and suicide attempts (6.0% vs. 1.9%) in the legally involved sexual minority population were higher than in the legally involved heterosexual population. Approximately one-fifth (22.0%) of legally involved sexual minority individuals had a serious mental illness, and approximately one-quarter (25.2%) indicated having depressive thoughts in the past year.

After we controlled for sociodemographic factors in logistic regression models, legally involved sexual minority individuals were more likely to report any mental illness (AOR=1.63, p<0.001), including serious mental illness (AOR=1.95, p<0.001), moderate mental illness (AOR=1.44, p=0.007), and depressive thoughts (AOR=1.95, p<0.001), compared with legally involved heterosexual individuals. In adjusted regression models, legally involved sexual minority individuals were more likely to report suicidal ideation (AOR=1.94, p<0.001), suicidal plans (AOR=2.48, p<0.001), and suicide attempts (AOR=2.31, p=0.003) than their legally involved heterosexual peers. Moreover, sexual minority individuals with no criminal legal involvement had similar patterns and were more likely to have mild mental illness (AOR=1.39, p<0.001) compared with their heterosexual counterparts with no criminal legal involvement.

Treatment for Mental Illness and Substance Use Disorders

Table 3 presents prevalence estimates and AORs comparing the utilization of mental health and substance use treatments between sexual minority and heterosexual adults, by legal involvement status. The prevalence of treatment among legally involved sexual minority individuals was higher than among their legally involved heterosexual counterparts for substance use disorder (22.7% vs. 18.4%), inpatient mental health (8.6% vs. 4.2%), outpatient mental health (19.6% vs. 11.8%), and prescription medication–based mental health treatment (29.8% vs. 17.1%). Nearly half (45.0%) of legally involved sexual minority individuals received treatment for substance use disorder or mental disorder in the past year, compared with one-third (33.4%) of their heterosexual peers. After we controlled for sociodemographic characteristics in regression models, legally involved sexual minority individuals were found to be more likely than their heterosexual counterparts to receive treatment for substance use disorder (AOR=1.29, p=0.030), any mental health treatment (AOR=1.58, p=0.001), inpatient mental health treatment (AOR=1.67, p=0.010), outpatient mental health treatment (AOR=1.43, p=0.030), prescription medication–based mental health treatment (AOR=1.64, p<0.001), and overall treatment either for substance use disorder or for mental illness (AOR=1.40, p=0.010). Sexual minority individuals with no criminal legal involvement shared similar patterns with their heterosexual peers. However, compared with legally involved sexual minority individuals, sexual minority individuals with no legal involvement generally reported lower levels of treatment for substance use (2.2% vs. 22.7%) or any mental disorders (29.3% vs. 35.7%).

Discussion

This study is one of the first to examine the prevalence of substance use disorder, mental illness, and access to treatment by sexual minority status and criminal legal involvement among a representative sample of adults across the United States. We found that individuals who identified as belonging to a sexual minority group had a higher prevalence of substance use and mental illness compared with their heterosexual counterparts, and this increased prevalence was especially prominent among those with past-year criminal legal involvement. Previous research has shown that persons from sexual minority groups have increased help-seeking behaviors (23, 24) and higher rates of treatment for mental and substance use disorders than their heterosexual peers (25–27). In the present study, we observed elevated levels of mental illness, substance use, and receipt of behavioral health treatment among both sexual minority individuals with and without criminal legal involvement compared with their heterosexual peers.

The findings of our study suggest that health care professionals and policy makers should focus on multiple avenues, beyond traditional treatment, to address the higher prevalence of mental and substance use disorders among legal system–involved sexual minority populations. These approaches may include public health initiatives, ranging from culturally competent clinical interventions to community-based programs providing sociopsychological resources to address behavioral health disorders. Other avenues to explore include establishing reentry programs that may provide better outreach at the time of release from correctional facilities to transitional support into the community, providing greater reimbursement for mental health and substance use treatment as individuals reenter the community, and improving educational resources for mental and substance use disorders in correctional facilities (3, 8, 16, 20). Other programs may include peer recovery, navigator-based interventions, and mobile health units, which have been previously shown to effectively support legally involved individuals with HIV, hepatitis C, and opioid use disorder (28–30).

More research at the intersection of sexual minority status and criminal legal involvement is needed. Future studies should evaluate differences in mental illness, substance use, and treatment among subcategories of sexual and gender minority groups, such as lesbian women, gay men, bisexual individuals, queer persons, and transgender individuals. Moreover, future research should analyze whether and how public policies across different states and municipalities affect sexual minority groups and their behavioral health as they transition into and out of the criminal legal system. Having more data is necessary for identifying best practices that address behavioral health during incarceration and after reentry into the community for legally involved sexual minority individuals.

This study had several limitations. All responses were self-reported, which can lead to response bias and social desirability bias. Moreover, reporting of sexual orientation may be subject to selection bias. Our sample of sexual minority adults included only those who were comfortable disclosing their sexual orientation in the NSDUH questionnaire. The survey also excluded homeless persons who do not use shelters, military personnel on active duty, and residents of institutional group quarters such as jails and hospitals, which have been previously shown to have a high proportion of individuals who are from a sexual minority group or have a history of criminal legal involvement (19, 31, 32).

This study focused on sexual orientation identity and did not consider sexual attraction or behavior; it thus excluded adults who are sexually active with or attracted to people of the same sex but do not identify as lesbian, gay, or bisexual. Furthermore, the sample of legally involved sexual minority individuals was relatively small (N=929) compared with the sample of their heterosexual peers (N=7,989). Although there may be important differences among various subgroups (e.g., among lesbian women, gay men, and bisexual individuals), we were not confident in having a sufficiently large sample size or statistical power to analyze these distinctions. Future studies should evaluate differences in mental health, substance use, and treatment utilization among subcategories of sexual and gender minority groups, including subpopulations from vulnerable socioeconomic statuses.

Finally, the NSDUH is a cross-sectional survey, and we were unable to longitudinally analyze behavioral health outcomes over time. Relatedly, we could not definitively establish the causal pathways for the observed associations because cross-sectional studies are prone to temporal and omitted variable bias (18). For example, we were unable to conclude whether poor mental health status led to criminal legal involvement or legal involvement led to poor mental health status. Moreover, missing unmeasured variables, such as exposure to discrimination or nondisclosure of sexual orientation to family, friends, and providers, may provide an alternative explanation for the relationship between sexual orientation and behavioral health. More nationwide health surveys and administrative data collection efforts should incorporate sexual orientation and previous criminal legal involvement data to facilitate a broader and more robust examination of legally involved sexual minority populations.

Conclusions

The findings from this nationally representative study indicate substantial disparities in mental illness and substance use among legal system–involved sexual minority populations, compared with their legally involved heterosexual peers. Public health practitioners, the criminal legal workforce, and health care professionals should be informed of the behavioral health needs of individuals in this population. Developing and implementing targeted approaches to mental health care for sexual minority groups in the adult criminal legal system may be one important step toward improving behavioral health outcomes among the most vulnerable sexual minority individuals. Future studies should continue to test, evaluate, and recommend best practices for the treatment of legally involved sexual minority individuals. Other public health initiatives and data collection efforts should continue to ascertain sexual orientation identity and criminal legal involvement for all persons in order to monitor progress toward achieving health equity.