Prevalence and Risk Factors of Nonmedical Prescription Opioid Use Among Transgender Girls and Young Women

IMPORTANCE Nonmedical prescription opioid use is a pressing public health issue in the United States. Transgender youth, including adolescent girls and young women who were assigned male at birth and currently identify as women, female, transgender women, or another diverse gender identity along the transfeminine gender spectrum, are more likely than their cisgender peers to report illicit substance use and meet diagnostic criteria for substance use disorders. However, relatively little is known about the experiences of these populations in the current era of opioid addiction and misuse. OBJECTIVE To report the prevalence of and risk factors associated with lifetime nonmedical prescription opioid use in a high-risk community sample of transgender adolescent girls and young women who are sexually active. DESIGN, SETTING, AND PARTICIPANTS This cross-sectional study used 2012 to 2015 baseline data from Project LifeSkills, a randomized clinical trial of a behavioral intervention to reduce the risk of HIV acquisition and transmission among a diverse sample of transgender adolescent girls and young womenrecruitedfromBoston,Massachusetts,andChicago,Illinois.Atotalof297transgendergirls and women aged 16 to 29 years who were sexually active were included in this analysis. Data were analyzed from June 2019 to August 2019. RESULTS Among 297 transgender adolescent girls and young women (mean [SD] age, 23.4 [3.5] years), 145 (48.8%) identified as non-Hispanic/Latinx black, 76 (25.6%) identified as non-Hispanic/ Latinx white, 37 (12.5%) identified as Hispanic/Latinx, 7 (2.4%) identified as non-Hispanic/Latinx health surveillance studies and evidence-based interventions to be comprehensively tailored to examine and respond to specific trends of substance use, particularly opioid use disorder, among transgender populations. t tests for continuous and χ 2 tests for categorical variables. Factors associated with using logistic In in unadjusted All analyses were conducted in SAS statistical (SAS values were and statistical significance was set at were analyzed from to


Introduction
Deaths attributable to accidental drug overdose, particularly those involving opioids, have reached epidemic levels in the United States, with more than 350 000 documented opioid overdose deaths between 2000 and 2015. 1,2 The opioid overdose epidemic has been described as a triple-wave epidemic, with a first wave of opioid overdoses associated with prescription opioids beginning in the late 1990s and early 2000s, a second wave of opioid overdoses associated with heroin beginning the middle of the 2000s, and a third wave of opioid overdoses associated with fentanyl, fentanyl analogues, and other synthetic opioids beginning in the early 2010s. 3 There is significant regional variation in the burden of opioid addiction and misuse, with the highest age-adjusted death rates due to opioid overdose observed in New England and the Midwest. 1,2 Nonmedical prescription opioid use is a pressing public health issue in the United States. 4 Given the dramatic increases in availability of prescription opioids in the United States during the past 20 years, results from a 2015 national study 4 suggest that the most recent generations of young adults have experienced earlier initiation of nonmedical prescription opioid use and have been at higher risk for transitioning to heroin use than previous generations. As a result, young adults engaging in nonmedical prescription opioid use are at increasingly high risk of adverse outcomes associated with drug use, such as increasing rates of unintentional overdose and acquisition of bloodborne infections. [5][6][7] As of 2017, 1 in 8 individuals aged 18 to 35 years report nonmedical prescription opioid use in their lifetime. 8 Thus, documenting the prevalence of nonmedical prescription opioid use among young adults at increased risk and identifying factors associated with use may help inform programs and policies designed to reduce opioid use and drug-related harms in this population.
Transgender individuals are more likely to report illicit substance use and meet diagnostic criteria for substance use disorders compared with their cisgender peers, [9][10][11][12][13] but relatively little is known about the experiences of these populations in the current era of opioid addiction and misuse.
A recent analysis of the National Survey of Drug Use and Health found that gay, lesbian, and bisexual individuals were more likely than their heterosexual counterparts to report nonmedical prescription opioid use in their lifetimes (20.2% vs 10.2%), in the past 12 months (9.0% vs 4.0%), and in the past 30 days (2.7% vs 1.2%). 14 However, to our knowledge, no national probability-based surveys of substance use and related outcomes currently include measures of both nonmedical prescription opioid use and gender identity, precluding any ability to understand the burden of opioid misuse in gender minority populations relative to cisgender groups. To begin to fill in this gap in the literature and explore this behavioral outcome in an understudied community, we estimated the prevalence of nonmedical prescription opioid use among a convenience sample of young transgender women in 2 urban centers heavily affected by opioid addiction and misuse and identify factors associated with this use.

Methods
This cross-sectional study leverages baseline data from Project LifeSkills, 15 a randomized clinical trial of a behavioral intervention to reduce the risk of HIV acquisition and transmission among transgender adolescent girls and young women. The study protocols and related methods have been described in detail elsewhere. 16 In brief, participants were recruited between 2012 and 2015 in JAMA Network Open | Substance Use and Addiction

Statistical Analysis
The associations of lifetime nonmedical opioid use with participant characteristics were compared using t tests for continuous variables and χ 2 tests for categorical variables. Factors associated with lifetime nonmedical opioid use were identified using logistic regression procedures. In the final adjusted model, the inclusion criteria for including variables were all demographic characteristics as well as all other variables found to be associated with lifetime nonmedical prescription opioid use in unadjusted analyses. All analyses were conducted in SAS statistical software version 9.4 (SAS Institute). P values were 2-tailed, and statistical significance was set at .05. Data were analyzed from June 2019 to August 2019.

Factors Associated With Nonmedical Prescription Opioid Use
The results of the unadjusted and adjusted logistic regressions are presented in

Discussion
To our knowledge, this is the first empirical study that examined self-reported lifetime nonmedical prescription opioid use among a community-recruited sample of transgender adolescent girls and  young women in 2 cosmopolitan cities in the United States. Our findings provide insights for this group's nonmedical use of prescription opioids at a time and context when the United States is experiencing an escalating opioid epidemic. 18 Results of this study suggest similar lifetime nonmedical prescription opioid use among this group compared with the national prevalence. 8 Specifically, transgender adolescent girls and young women who are current cigarette smokers and transgender adolescent girls and young women who identified as a sexual orientation other than heterosexual, gay, lesbian, or bisexual had significantly higher odds of lifetime nonmedical prescription opioid use than their counterparts.

JAMA Network Open | Substance Use and Addiction
We found that 11.8% of transgender adolescent girls and young women in our sample reported lifetime nonmedical prescription opioid use, compared with a national prevalence of 12.5% lifetime nonmedical prescription opioid use among their cisgender counterparts. 8  Furthermore, it has been suggested that rates of substance use among sexual and gender minority populations can vary by subgroup and the type of drug used. 9,14 In addition to addressing nonmedical prescription opioid use, there is a need to understand and characterize trends of heroin and fentanyl use among transgender populations, given how these 2 substances have also been associated with opioid-related overdoses and deaths in the United States. 18,20 As such, given that this study specifically only asked about lifetime nonmedical prescription opioid use between 2012 to 2015, there is a need for current or future surveillance studies to examine patterns of current or recent lifetime nonmedical prescription opioid use, types of opioids used, and polysubstance use among this group to inform evidence-based substance use-related interventions for transgender adolescent girls and young women. Moreover, given that the history of opioid use among this sample was similar to the current national prevalence, it is important for national surveillance systems to begin collecting data specific to gender identity to understand trends for this group's opioid-related behavior over time.
There are several possible explanations for why girls and women in our sample who reported having sexual orientation other than heterosexual, gay, lesbian, or bisexual had higher nonmedical prescription opioid use. The first reason could be owing to experiences of minority stress. 21,22 The minority stress model asserts that individuals who identify with and belong to minority groups (eg, gender minorities, such as transgender people) are likely to experience greater levels of stress due to several factors, including interpersonal, social, and structural forms of stigma and discrimination.
Greater levels of stress are likely to place individuals in minority groups at elevated rates of coping behaviors, including substance use and smoking. 23,24 Minority stress is particularly applicable to transgender adolescent girls and young women who reported having other sexual orientation types (eg, pansexual, asexual), in part because they may not fully belong to or identify with other more common sexual orientations (eg, heterosexual, gay, lesbian) that may have community-centered support that can buffer against minority stress. 25 As such, they may experience anxiety related to identifying with other sexual orientation types, which can influence stress levels as well as increase behaviors related to substance use, including cigarette smoking and nonmedical prescription opioid use. Future research should further examine this postulation, particularly regarding the role of social support, which has been posited to help alleviate stress and reduce substance use in the context of the minority stress model. 25 Social participation and cohesion theories 26-28 may also help explain why transgender adolescent girls and young women with other sexual orientation types reported higher nonmedical prescription opioid use. These theories highlight the role of community engagement to increase mutual aid, trust, and solidarity among group or community members and reduce negative health behaviors, like substance use, 29,30 which could also help explain why differences were not observed among gay, lesbian, or bisexual transgender adolescent girls and young women. Given that gay, lesbian, and bisexual groups in the United States have historically and relatively been more formalized and mobilized as communities or networks 31,32 compared with those who belong to other kinds of sexual orientations, it is likely that such social cohesiveness and participation among gay and lesbian transgender adolescent girls and young women may be associated with lower nonmedical prescription opioid use. Future research should also further investigate individuals whose sexual orientation is different from heterosexual, gay, lesbian, or bisexual.

Strengths and Limitations
This study has several strengths. One major strength was the characterization of risk factors of lifetime nonmedical prescription opioid use specifically among transgender adolescent girls and young women. Studies among transgender populations have often considered substance use as a factor associated with HIV risks, which limits our understanding about disparities related to substance use and factors that are associated with substance use among this population. 19 In contrast to this predominant approach in transgender research, the analysis approach of this study specifically explored risk factors associated with lifetime nonmedical prescription opioid use.
Our analysis also has several limitations. First, we used a non-probability sampling approach to recruit participants for the study, which may have resulted in selection bias. Second, the sample was drawn from a randomized clinical trial that had strict eligibility criteria, which might make the results not generalizable to the population of interest. Third, the original study was not powered to detect possible differences in nonmedical prescription opioid use; thus, it may be underpowered to detect meaningful differences in some outcomes. Fourth, the cross-sectional nature of the study precludes us from any conclusions regarding causality. The sample represents a subset of transgender people and is therefore not representative of the entire transgender community, many of whom have gender expansive or gender nonconforming identities. Additionally, the self-reported nature of the data collection may bias directionality of the results such that participants may have felt stigmatized to disclose or reveal their opioid use and other personal behaviors and experiences. Future research should include biomarkers to determine actual prevalence.

Conclusions
This cross-sectional study provides insights into the prevalence of lifetime nonmedical prescription opioid use among transgender adolescent girls and young women living in 2 metropolitan US cities.
Evidence-based services for substance use that specifically target opioid use disorder can and should be comprehensively tailored to transgender populations, particularly transgender adolescent girls and young women. Integration of such services with identity-affirming mental health services could also be valuable to reaching and targeting this population. The findings of this study may serve as a warranted call-to-action for public health surveillance studies to specifically examine trends of substance use, particularly opioid use disorder, among transgender populations.