Prevalence and Co-occurrence of Alcohol, Nicotine, and Other Substance Use Disorder Diagnoses Among US Transgender and Cisgender Adults

Key Points Question What is the prevalence of substance use disorder diagnoses (SUDDs) among US transgender and cisgender adults? Findings In this study of 15 637 transgender and 46 911 cisgender adults, transgender adults had a significantly higher prevalence of nicotine, alcohol, and drug SUDDs than cisgender adults. Meaning In this study, transgender adults experienced elevated levels of SUDD relative to their cisgender peers, suggesting that effective, culturally tailored SUDD treatment interventions are warranted.


Introduction
Transgender people, whose gender identity differs from their assigned birth sex and who may access hormones or surgery to align their physical gender expression with their gender identity, are at risk of substance abuse and dependence. [1][2][3][4][5][6][7][8] Research suggests that substance misuse and related disorders are in part associated with some transgender people's reliance on substances to cope with the psychological toll of discrimination. [9][10][11] While research has captured the burden of substance misuse among transgender people, 1-7 estimates of substance use disorder (SUD) are limited and vary considerably (ie, 3.9% to 47.2%) depending on the sample and SUD type. 3,5,[12][13][14][15][16][17][18] Much of the research documenting the burden of substance use and related disorders among transgender individuals has relied on geographically limited, survey-based research that often focuses on high-risk subgroups, including younger transgender people and transfeminine (TF) people. [2][3][4][19][20][21] In addition to the frequent reliance on self-reported outcomes, survey-based research only collects data from individuals who self-identify as transgender and elect to participate in research, which raises concerns about the generalizability of findings. Conversely, health care administrative databases enable the identification of large, geographically representative cohorts of transgender individuals and allow for comparisons to be made between cisgender (ie, nontransgender) and transgender people. Despite the benefits of using administrative claims databases to study substance use disorder diagnoses (SUDDs) among transgender people, few studies use this approach. Furthermore, those that do typically focus on narrow populations (eg, transgender veterans), 17,18 fail to report the full range of SUDDs (eg, nicotine, alcohol, cannabis, opioid, cocaine), 12,13,[16][17][18] or fail to examine differences by gendered subgroups (eg, TF vs transmasculine [TM] people) 12,13,17,18 or across geographic regions. 12,13,[16][17][18] The present study aimed to fill these gaps by using a large, national administrative claims database to identify the distribution of SUDDs among transgender and cisgender adults across key subgroups defined by age, gender, and geographic region. The analyses were driven by the overarching goal of identifying subpopulations at greatest risk of SUDDs so that culturally tailored clinical interventions can be developed to treat SUDDs among the most at-risk segments of the transgender population.

Methods
We conducted a cross-sectional analysis of the OptumLabs Data Warehouse (OLDW), which includes deidentified claims data for commercially insured and Medicare Advantage enrollees. The patientlevel information in the OLDW comprises enrollment, medical claims, and pharmacy claims across care settings. Our study sample was drawn from approximately 74 million adults (ie, Ն18 years of age) enrolled in commercial or Medicare Advantage plans in 2017. The study was approved by the institutional review boards of Boston University and the RAND Corporation. Given that this is a secondary data analysis of a deidentified insurance claims dataset, collection of written informed consent was neither possible nor required. This study is reported in accordance with the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline.

Cohort Identification and Stratification
To minimize missing data, only individuals who were enrolled for all of 2017 and had at least 5 medical claims reported in 2017 were included in this analysis. Using our previously developed algorithm, 22 we identified a cohort of transgender adults using a combination of International Classification of Diseases, Ninth (ICD-9) and ICD-10 diagnostic codes specific to transgender individuals; Common Procedural Terminology codes for transgender-related surgical procedures; and the use of sex-discordant hormones. A cisgender cohort was matched 3:1 to the transgender cohort based on birth year and geographic region (ie, Northeast, South, Midwest, West).

Measures Outcomes
SUDDs were identified using ICD-10 codes from health care encounter claims in 2017 and included alcohol, nicotine, cannabis, cocaine, and opioids (see eAppendix in the Supplement). The ICD-10 also includes codes for other psychoactive substance-related diagnoses. Due to the low prevalence of sedative, stimulant, and hallucinogen SUDDs, individuals with diagnosis codes related to these disorders were included in the other SUDD group. To compare findings with the US general population, 23 nonalcohol and nonnicotine SUDDs were combined to create an indicator of drug SUDDs. The polysubstance SUDD variable included individuals with 2 or more SUDDs.

Demographic Characteristics
Age was divided into 9 categories, as follows: 18

Statistical Analysis
Analyses were performed using SAS version 9.4 (SAS Institute). Due to our cohort inclusion criteria (ie, full year of enrollment, minimum of 5 claims), there were no missing data. Period prevalence estimates for each SUDD were calculated for 2017 among transgender and cisgender individuals. We used 2-tailed χ 2 tests, with statistical significance set at P < .05, to assess within-group differences (among transgender people and among cisgender people) and between-group differences (across transgender and cisgender people) in SUDD by geographic region, age, and gender subgroup (TF vs TM; male vs female). Transgender individuals with an unknown gender were excluded from the within-group, gender-stratified analysis.

Results
In this US study of 15

Discussion
To our knowledge, this study represents the largest national study of SUDD disparities among transgender and cisgender adults to date. In 2017, the prevalence of polysubstance SUDD among transgender people in our sample was 4 times that of their cisgender peers. Moreover, the prevalence of any drug SUDD among transgender people was 3.6 times that of cisgender people in our sample and approximately 1.6 times that of the US general population in 2017. 23 Findings underscore the need for future research to examine the SUDD treatment experiences of transgender individuals so that effective clinical interventions can be developed to reduce the burden of SUDD among this population. When examining SUDDs separately, significant disparities were observed. Notably, nicotine SUDD was the most prevalent SUDD for both transgender and cisgender individuals, although the prevalence was significantly higher among transgender people (16.6%) than cisgender individuals   [24][25][26][27] Regarding cannabis, transgender individuals in this sample had more than 5 times the prevalence of cannabis SUDD than cisgender people and 1.5 times that of the 2017 US general population. 23 Similarly, transgender individuals in this sample had 4 times the prevalence of cocaine SUDD and 3 times the prevalence of opioid SUDD relative to cisgender people and 1.4 and 1.7 times the prevalence of cocaine and opioid SUDD, respectively, relative to the 2017 US general population. 23 Research has found that transgender individuals frequently use substances to cope with discrimination. [9][10][11] While the use of substances such as nicotine, cannabis, cocaine, and opioids may be an effective short-term strategy to cope with the stress of discrimination, these substances carry a myriad of health risks. 28-30 Transgender individuals who use substances to cope with discrimination may benefit from engagement in clinical interventions that promote positive coping strategies such as meditation, exercise, or peer support. 10,31 Much of the prior substance use research with transgender people is geographically limited, and the national research that does exist routinely fails to explore regional differences in SUDD. In this study, we found that transgender individuals in the Northeast, South, Midwest, and West consistently had higher rates of nicotine, alcohol, and all other SUDDs relative to cisgender individuals within the same region. We also found that transgender people in the Northeast had significantly higher prevalence of nicotine, alcohol, cannabis, cocaine, opioid, and other SUDDs relative to transgender individuals in other regions. The prevalence of alcohol and cocaine SUDD was similarly elevated among cisgender individuals in the Northeast, relative to those in other regions.
National 2017 data on US residents aged 12 years and older found that those living in the West had the highest prevalence of past-year drug misuse relative to those in other regions; however, the prevalence of substance use treatment was highest in the Northeast, relative to other regions. 32 Since health care engagement is required to receive an SUDD, it may be that cisgender and transgender people from the West misuse drugs more, but those in the Northeast have greater access and/or motivation to engage in treatment. Differences in the regional distribution of SUDDs in our study relative to the distribution of SUDDs among the US general population, aged 12 years and older, may be due to the fact that our study included people with insurance aged 18 years and older and older age and insurance may be particularly protective against having an SUDD for those living in the West vs other US regions. Additional research is needed to understand the mechanisms driving geographic disparities in SUDD among cisgender and transgender people in the United States.
When examining differences according to age, our study found that transgender adults aged 18 to 25 years had the highest burden of all SUDDs (except nicotine) relative to older transgender people and cisgender people of all ages. Our study extends findings from much smaller observational studies documenting the high prevalence of disordered alcohol and drug use behaviors among younger transgender people 2,3,5,20,33 and also aligns with 2017 survey data showing an elevated prevalence of alcohol and drug SUDDs among young adults (aged 18-25 years) relative to older members of the US general population. 23 Notably, however, the prevalence of nicotine SUDD in the present study doubled among transgender individuals between the ages of 18 and 25 years and those aged 61 years and older and more than tripled among cisgender individuals across age groups.
Although US general population data suggest greater nicotine use and dependence among younger adults, 34,35 concerns about the health risks of smoking and the desire to quit have been shown to increase with age, 36 and a significantly higher proportion of adults aged 65 years and older report being former smokers than those younger than 65 years. 35,37 Thus, the age-related disparities in nicotine SUDD documented among transgender and cisgender adults in our sample may reflect a greater recognition of problematic use and elevated treatment-seeking among adults in the older age groups relative to those in the younger age groups. Taken together, these findings highlight the need to detect and treat nicotine SUDD among US adults as they age and underscore the need for early

Limitations
Our analysis has several limitations. We used our previously developed algorithm 22  gender-affirming care had been included, with potentially wider SUDD disparities observed between the TF and TM subsamples. National data are needed to characterize SUDD disparities among TF, TM, and nonbinary people who have and have not accessed gender-affirming medical care.

Conclusions
To our knowledge, this is the largest national study to document within-group and between-group disparities in SUDDs among US transgender and cisgender adults. Transgender individuals in our study had significantly higher rates of SUDDs compared with cisgender individuals, a pattern that persisted when transgender and cisgender cohorts were compared across age groups and geographic areas. These findings highlight the need to ensure access to high-quality SUD treatment for transgender individuals as well as additional research to understand facilitators and barriers to SUD treatment engagement for this population. Such research can inform the development of novel public health interventions to prevent and treat SUD among transgender people in the United States.