Abstract
Background:
Substance use (SU) and misuse are disproportionately more common among sexual and gender minority (SGM) individuals compared to their heterosexual peers. Yet, little is known about regional and demographic differences in use and misuse among SGM. In this study, we investigated regional and demographic differences in SU and misuse in a U.S. national, internet-based cohort (n = 6,280) of men and trans persons who have sex with men.
Methods:
Data collected included the WHO ASSIST (substance) and AUDIT (alcohol) SU scales to estimate recent (≤ 3 months) non-problematic use (≤ 3 ASSIST, ≤ 10 AUDIT) and misuse (≥4 ASSIST, ≥11 AUDIT). We used bivariate and multivariable logistic models to examine demographic and regional factors associated with SU and misuse.
Results:
Participants reported using alcohol (85.6%), cannabis (53.9%), and inhalants (39.1%) in the past three months. More than one-third self-reported misuse of cannabis, Gamma-Hydroxybutyrate (GHB), inhalants, methamphetamines, and prescription sedatives. We observed regional differences in substance use for cannabis (Southeast aOR = 0.76, 95% CI: 0.63–0.93; West aOR = 1.27, 95% CI: 1.02–1.59, ref. Northeast) and prescription Stimulants (Midwest aOR = 1.39, 95% CI: 1.00–1.93), as well as for cannabis misuse (Southeast aOR = 0.83, 95% CI: 0.69–0.99). We also observed significant associations between socioeconomic factors with use and misuse.
Conclusions:
Findings suggest geographic differences in misuse of certain substances among men and trans persons who have sex with men in the US, and that socio-economic factors, also play a key role in indicating risk.
Keywords: substance use, substance misuse, geographic contexts, MSM, AUDIT, ASSIST
Introduction
The recent increases of drug-related deaths—e.g., opioid deaths (Hedegaard, Warner, & Miniño, 2018)—point to concerns about increases in substance use (SU) and misuse in the U.S., precursors to these fatal drug-related outcomes. Previous research has indicated that U.S. geography plays an important role in the prevalence of substance use (Hedegaard et al., 2018; Sean Esteban McCabe & West, 2017). Indeed, national reports illustrate regional variability in the prevalence of specific substances across the U.S. For example, methamphetamine is more readily available in the Western and Midwestern U.S. (National Institute on Drug Abuse, National Institutes of Health, & U.S. Department of Health and Human Services, 2019). In these regions, methamphetamine use is higher than in others (Hirshfield, Remien, Humberstone, Walavalkar, & Chiasson, 2004), and has been recognized by local law enforcement as the greatest drug threat in their areas (National Institute on Drug Abuse et al., 2019). In contrast, heroin is thought to be more localized in urban areas, although heroin use has begun to impact suburban communities surrounding metropolitan areas (National Institute on Drug Abuse, National Institutes of Health, & U.S. Department of Health and Human Services, 2018). Other studies have found evidence of substance use discrepancies between regions for injection stimulant use (Sullivan, Nakashima, Purcell, & Ward, 1998), and discrepancies in opioid use between states with large proportions of rural populations (Keyes, Cerdá, Brady, Havens, & Galea, 2014). Therefore, understanding key geographic differences in distribution of substance use and misuse can help narrow focus on key targets for intervention, yielding more effective results.
Beyond geographic variability, certain populations, particularly men (broadly) and sexual and gender minority (SGM) communities, are more likely to use and misuse alcohol and other substances than non-SGM U.S. populations (Boyd, West, & McCabe, 2018; Carrico et al., 2012; Cochran, Ackerman, Mays, & Ross, 2004; Fenton, Keyes, Martins, & Hasin, 2010; Hirshfield et al., 2004; Keuroghlian, Reisner, White, & Weiss, 2015; Sean Esteban McCabe & West, 2017; Stall et al., 2001). This is particularly concerning among men who have sex with men (MSM) and trans persons given the association of substance use with engaging in other HIV-related risk behaviors (Ciccarone & Bourgois, 2003; Hirshfield et al., 2004; Reback & Fletcher, 2014; Stall et al., 2003) and the barrier substance use or misuse may pose to HIV prevention (Closson, Mitty, Malone, Mayer, & Mimiaga, 2018). Substances commonly reported by MSM include cocaine, inhalants, marijuana, and alcohol, but other substances such as methamphetamines, Gamma-Hydroxybutyrate (GHB) and hallucinogens have also been reported (Cochran et al., 2004; Gattis, Sacco, & Cunningham-Williams, 2012; Halkitis, Green, & Mourgues, 2005; Hatzenbuehler, Corbin, & Fromme, 2008; Koblin et al., 2003; Marshal et al., 2008; Sean Esteban McCabe, Hughes, Bostwick, West, & Boyd, 2009; S. E. McCabe, West, Hughes, & Boyd, 2013; Ostrow et al., 1993; Pantalone, Bimbi, Holder, Golub, & Parsons, 2010). Even within MSM populations there appear to be variability by other demographic factors. Younger MSM experience high rates of substance use (Marshal et al., 2008; Marshal, Friedman, Stall, & Thompson, 2009; Schuler, Rice, Evans-Polce, & Collins, 2018), and, as with the general population, drug use among MSM appears to vary by geographic areas within the US (Hirshfield et al., 2004; Koblin et al., 2003). Often, geographic nearness or spatial clustering of drug/alcohol use and sexual risk behavior overlap indicating the importance of understanding linkages in environmental contexts with substance use and HIV risk behavior (Tobin, Yang, King, Latkin, & Curriero, 2016). Additionally, the types of substances used and the methods through which MSM use them vary by geographic location (Sullivan et al., 1998).
In addition to higher rates of substance use overall, SGM appear more susceptible than their non-SGM peers to developing problematic patterns of alcohol use and substance use (Bolton & Sareen, 2011; Gilman et al., 2001; S. E. McCabe et al., 2013), potentially because of the need to cope with the significant identity-based discrimination and its consequences termed sexual minority stress in the literature (Meyer, 2003). This sexual minority stress may be more acutely felt in smaller, rural, and Southern communities (Swank, Frost, & Fahs, 2012) exasperating and accentuating feelings of stigma and isolation that increase substance use and misuse. Studies investigating substance use and misuse, including alcohol, among MSM have indicated increased vulnerability to substance misuse and diagnosable substance use disorders (Evans-Polce, Veliz, Boyd, Hughes, & McCabe, 2019; Gilman et al., 2001; Kerridge et al., 2017; Sean Esteban McCabe et al., 2009; S. E. McCabe et al., 2013; Schuler et al., 2018).. The increased prevalence of alcohol and substance use and misuse as well as the association of these alcohol and substance using behaviors with risk behaviors associated with increased vulnerability for HIV indicate opportunities for multifaceted interventions.
Although smaller, geographically localized studies have shown higher rates of alcohol use, substance use, and substance misuse among MSM, relatively fewer studies look at substance use among geographically diverse, larger groups of MSM. Even fewer, to our knowledge, have examined misuse differences within U.S. regional contexts. In the present study, we examine data from an online, U.S. national cohort to address two aims: (1) to document the prevalence of substance use and misuse by geographic region in the US, and (2) to evaluate regional correlates of substance use and misuse among men, trans men, and trans women.
Material and methods
Study enrollment and recruitment
This analysis used data from the Together 5000 (T5K) study, a U.S. national, internet-based cohort study of men, trans men, and trans women who have sex with men (Figure 1) to identify modifiable individual and structural factors associated with HIV seroconversion and PrEP uptake. Enrolment began in October 2017 using ads on men-for-men geosocial networking phone applications (apps) and concluded in June 2018. The cohort and study procedures have been fully described elsewhere (Grov et al., 2019; Nash et al., 2019). Briefly, core eligibility criteria for enrollment specified that participants were aged 16 to 49; had at least two male sex partners in the past three months; were not currently participating in an HIV vaccine or PrEP clinical trial; were not currently on PrEP; lived in the U.S. or its territories; were not known to be HIV-positive; had a gender identity other than cisgender female; and met at least one other criteria indicating that they engaged in higher risk sexual behaviors which are listed elsewhere (Grov et al., 2019; Nash et al., 2019).
Figure 1.
Geographic distribution of Together 5,000 study participants who completed the psychosocial measurements survey, 2017–2018 (n = 6,280).
Participants clicking on one of our study ads were routed from geosocial apps to a secured informed consent and enrolment survey webpage that presented questions about demographic characteristics, sexual behavior, and substance use. Of those who completed the enrolment survey, 8,774 participants met eligibility criteria and provided contact information for later follow-up. Eligible participants who consented and completed the enrolment survey were later sent a link to complete a supplemental secondary survey. Of the 8,774 participants who enrolled, 6,280 (71.6%) completed the secondary survey and received a $15 incentive (Grov et al., 2019). The information used for this manuscript was collected as part of both the screening and secondary surveys; as such, the final sample used for this analysis is 6,280.
Surveys and measures
The main outcomes of interest for the current analyses were alcohol and substance use, collected using the widely used World Health Organization (WHO) Alcohol Use Disorders Identification Test (AUDIT) and the Alcohol, Smoking and Substance Involvement Screening Test (ASSIST) measures adapted for specific substances (Babor, Higgins-Biddle, Saunders, & Monteiro, 2001; Humeniuk et al., 2010). We investigated regional differences in both recent (≤ 3 months) substance use and substance misuse.
AUDIT
The AUDIT assessed frequency and quantity of alcohol consumption in the past 12 months. Alcohol misuse was measured by ascertaining the need to drink, and the impact of problem drinking on employment and personal relationships. Questions were scored from zero to four based on reported frequency, and a total score was calculated (range, 0–40). Recommended cut offs for non-problematic alcohol use were a total score less than or equal to 10 (for more specificity), and any scores of 11+ were labelled as alcohol misuse.
ASSIST
Using the ASSIST questions (Humeniuk et al., 2010), we evaluated cannabis, cocaine, prescription stimulants, methamphetamine, inhalants, prescription sedatives or sleeping pills, GHB, Ecstasy/MDMA/Molly, LSD/Acid/mushrooms/PCP (Angel Dust), Special K (Ketamine), street opioids (e.g., heroin, opium), and prescription opioids (e.g., morphine, codeine) use. The ASSIST asks about lifetime use, frequency of use (≤ 3 months), frequency of strong urges to use each substance (≤ 3 months), and how the use of each substance has impacted their health, social relationships, finances, and responsibilities as well as any legal trouble their substance use may have caused. Lifetime substance use determined which substances to ask follow-up questions about. Therefore, the final ASSIST scores represent recent (past 3 months) substance use only, and our analysis focuses only on recent use (vs. not recent use) and recent misuse. Substance-specific questions were used to calculate total sum scores (range, 0–39) for each substance. Recommended cut-offs for non-problematic substance use were a total score ≤ 3, and any scores 4+ were considered substance misuse. Of note, for prescription drug questions (stimulants, sedatives, and opioids), subsequent questions were asked to ascertain per-prescription usage. However, due to a programming error in the survey, prescription dosing information was not fully captured (see limitations) and, thus, opioid use variables in our analyses represent any use.
Covariates
Surveys assessed demographic, socioeconomic, and sexual health measures such as age, gender, race/ethnicity, sexual orientation, marital status, employment status, highest level of education, annual income, housing instability in the past 5 years, sex work in the past 3 months, lifetime incarceration status, and ZIP code of current residence—which we used to determine geographic region based on the U.S. Census Bureau regional definitions (Northeast, Southeast, Midwest, West, Pacific, and U.S. Territories). Due to limited numbers of participants residing in these areas, the Pacific regions and U.S. Territories (Puerto Rico and Guam) were combined.
Analyses
Categorical descriptive statistics (frequencies, percentages) were calculated for demographic and geographic characteristics for the analyzed sample. Additional descriptive statistics were calculated for each of the continuous alcohol and substance use scale total scores (e.g., means (), standard deviations (s)), and for the binary indicators of recent use and misuse (frequencies, percentages). We also calculated Cronbach’s alpha to measure internal consistency of each scale in our study population. Tests of differences (chi-squared analyses) between recent use and non-recent use (among those reporting any use of each substance) as well as non-problematic use and misuse were conducted for demographic and geographic factors with each substance individually. Factors considered for bivariate analyses were determined a priori from the literature.
We then used multivariable logistic regression to investigate demographic, behavioral, and geographic factors associations with substance use and misuse. Our two goals for multivariable models were to assess regional differences in patterns of substance use and misuse among our participants. Therefore, we used our geographical bivariate results to focus our multivariable models to substances indicating initial regional differences for recent use and misuse. Bivariate analyses indicated initial differences for substance misuse for cannabis, cocaine, prescription opioids, prescription stimulants, and prescriptions sedatives. Each of these substances were assessed in separate models. The variables that were included in these adjusted models were determined from the bivariate analyses results of substances and demographic, socioeconomic, and sexual health factors (Supplemental Tables 1 and 2). Final models were determined using changes in model fit criteria (e.g., AIC). As a sub-analysis, additional multivariable models were used to investigate potential moderating effects of region for all covariates of interest among those eight substances with initial use and misuse differences by region. These models consisted of variables previously identified in the final multivariable models through prior bivariate analyses and assessing model fit criteria. The significance level for all tests of associations was 0.05, and we report adjusted odds ratios (aORs) with 95% confidence intervals (95% CIs) for regression results. All analyses were conducted using SAS 9.4.
Results
Demographic characteristics of the sample
Most study participants identified as cisgender male (97.6%) and gay, queer, or homosexual (84.4%) (Table 1). Half were between 25–35 years old and just under half were persons of color (48.2%). In general, participants reported being well-educated with 45% reporting at least some college education and 38% reporting a college degree or higher. Over half (61.9%) reported having full-time employment, and 41.5% reported making an annual income ranging from $20,000 and $49,999. However, 20.9% reported experiencing housing instability in the past 5 years, 15.1% reported engaging in sex work in the past 3 months, 8% reported ever using injection drugs, and 14.5% reported that they had ever been incarcerated. Almost half (48.2%) reported living in the Southeastern U.S.
Table 1.
Demographic characteristics of participants completing the secondary survey for the total study population and by U.S. region, Together 5,000 study, 2017–2018, n = 6,280.
| U.S. Region | ||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Total | Northeast | Southeast | Midwest | West | Pacific/US Territories/Other | |||||||
| n = 6,280 | n = 910 | (15.0) | n = 2936 | (48.2) | n = 887 | (14.6) | n = 1306 | (21.5) | n = 48 | (0.8) | ||
| Characteristic | n | % | n | % | n | % | n | % | n | % | n | % |
| Age | ||||||||||||
| 16–24 years old | 1533 | (24.4) | 199 | (21.9) | 793 | (27.0) | 191 | (21.5) | 247 | (18.9) | 6 | (12.5) |
| 25–35 years old | 3150 | (50.2) | 507 | (55.7) | 1411 | (48.1) | 442 | (49.8) | 690 | (52.8) | 25 | (52.1) |
| 36–45 years old | 1263 | (20.1) | 162 | (17.8) | 579 | (19.7) | 198 | (22.3) | 296 | (22.7) | 12 | (25.0) |
| 46–55 years old | 334 | (5.3) | 42 | (4.6) | 153 | (5.2) | 56 | (6.3) | 73 | (5.6) | 5 | (10.4) |
| Gender | ||||||||||||
| Cisgender male | 6128 | (97.6) | 881 | (96.8) | 2868 | (97.7) | 871 | (98.2) | 1276 | (97.7) | 47 | (97.9) |
| Transgender female | 41 | (0.7) | 4 | (0.4) | 24 | (0.8) | 5 | (0.6) | 7 | (0.5) | -- | -- |
| Transgender male | 42 | (0.7) | 9 | (1.0) | 18 | (0.6) | 4 | (0.5) | 8 | (0.6) | -- | -- |
| Something Else | 69 | (1.1) | 16 | (1.8) | 26 | (0.9) | 7 | (0.8) | 15 | (1.2) | 1 | (2.1) |
| Race/ethnicity | ||||||||||||
| White | 3255 | (51.8) | 496 | (54.5) | 1423 | (48.5) | 632 | (71.3) | 621 | (47.6) | 14 | (29.2) |
| Black | 697 | (111) | 96 | (10.6) | 447 | (15.2) | 73 | (8.2) | 57 | (4.4) | 1 | (2.1) |
| Latino/Latinx | 1542 | (24.6) | 168 | (18.5) | 751 | (25.6) | 92 | (10.4) | 429 | (32.9) | 23 | (47.9) |
| Asian or Pacific Islander | 226 | (3.6) | 64 | (7.0) | 79 | (2.7) | 28 | (3.2) | 51 | (3.9) | -- | -- |
| Other or multi-racial/ethnic | 560 | (8.9) | 86 | (9.5) | 236 | (8.0) | 62 | (7.0) | 148 | (11.3) | 10 | (20.8) |
| Sexual orientation | ||||||||||||
| Gay, Queer, Homosexual | 5299 | (84.4) | 795 | (87.4) | 2451 | (83.5) | 764 | (86.1) | 1117 | (85.5) | 37 | (77.1) |
| Bisexual | 898 | (14.3) | 103 | (11.3) | 445 | (15.2) | 111 | (12.5) | 176 | (13.5) | 11 | (22.9) |
| Other | 83 | (1.3) | 12 | (1.3) | 40 | (1.4) | 12 | (1.4) | 13 | (1.0) | -- | -- |
| Marital status | ||||||||||||
| Yes | 787 | (12.5) | 111 | (12.2) | 348 | (11.9) | 113 | (12.7) | 194 | (14.9) | 7 | (14.6) |
| No | 5493 | (87.5) | 799 | (87.8) | 2588 | (88.2) | 774 | (87.3) | 1112 | (85.2) | 41 | (85.4) |
| Employment | ||||||||||||
| Full-time (40 hours per week) | 3890 | (61.9) | 576 | (63.3) | 1840 | (62.7) | 604 | (68.1) | 760 | (58.2) | 32 | (66.7) |
| Part-time (less than 40 hours per week) | 818 | (13.0) | 128 | (14.1) | 362 | (12.3) | 87 | (9.8) | 209 | (16.0) | 6 | (12.5) |
| Working or full-time student | 964 | (15.4) | 131 | (14.4) | 462 | (15.7) | 115 | (13.0) | 187 | (14.3) | 6 | (12.5) |
| Unemployed/Other | 608 | (9.7) | 75 | (8.2) | 272 | (9.3) | 81 | (9.1) | 150 | (11.5) | 4 | (8.3) |
| Education | ||||||||||||
| < High school diploma | 148 | (2.4) | 16 | (1.8) | 72 | (2.5) | 20 | (2.3) | 28 | (2.1) | 1 | (2.1) |
| High school diploma or GED | 906 | (14.4) | 80 | (8.8) | 462 | (15.7) | 123 | (13.9) | 191 | (14.6) | 6 | (12.5) |
| Some college or technical school training | 2826 | (45.0) | 341 | (37.5) | 1383 | (47.1) | 398 | (44.9) | 594 | (45.5) | 22 | (45.8) |
| College graduate + | 2400 | (38.2) | 473 | (52.0) | 1019 | (34.7) | 346 | (39.0) | 493 | (37.8) | 19 | (39.6) |
| Income | ||||||||||||
| Less than $20,000 | 2110 | (33.6) | 256 | (28.1) | 1044 | (35.6) | 250 | (28.2) | 434 | (33.2) | 15 | (31.3) |
| $20,000-$49,999 | 2603 | (41.5) | 333 | (36.6) | 1244 | (42.4) | 424 | (47.8) | 529 | (40.5) | 19 | (39.6) |
| 50,000+ | 1567 | (25.0) | 321 | (35.3) | 648 | (22.1) | 213 | (24.0) | 343 | (26.3) | 14 | (29.2) |
| Housing instability | ||||||||||||
| No/Not within last 5 years | 4966 | (79.1) | 751 | (82.5) | 2283 | (77.8) | 744 | (83.9) | 1009 | (77.3) | 39 | (81.3) |
| Yes, within last five years | 1314 | (20.9) | 159 | (17.5) | 653 | (22.2) | 143 | (16.1) | 297 | (22.7) | 9 | (18.8) |
| Sex work (past 3 months) | ||||||||||||
| No | 5333 | (84.9) | 779 | (85.6) | 2447 | (83.3) | 783 | (88.3) | 1130 | (86.5) | 39 | (81.3) |
| Yes | 947 | (15.1) | 131 | (14.4) | 489 | (16.7) | 104 | (11.7) | 176 | (13.5) | 9 | (18.8) |
| Incarcerated (ever) | ||||||||||||
| No | 5370 | (85.5) | 857 | (94.2) | 2443 | (83.2) | 766 | (86.4) | 1090 | (83.5) | 44 | (91.7) |
| Yes | 910 | (14.5) | 53 | (5.8) | 493 | (16.8) | 121 | (13.6) | 216 | (16.5) | 4 | (8.3) |
| Injection drug use (ever) | ||||||||||||
| No | 5778 | (92.0) | 856 | (94.1) | 2680 | (91.3) | 835 | (94.1) | 1183 | (90.6) | 44 | (91.7) |
| Yes | 502 | (8.0) | 54 | (5.9) | 256 | (8.7) | 52 | (5.9) | 123 | (9.4) | 4 | (8.3) |
| Recent Substance Use* | ||||||||||||
| Alcohol | 5374 | (85.6) | 800 | (87.9) | 2516 | (85.7) | 772 | (87.0) | 1088 | (83.3) | 45 | (93.8) |
| Cannabis | 3382 | (53.9) | 492 | (70.3) | 1482 | (66.3) | 469 | (66.7) | 808 | (74.3) | 26 | (72.2) |
| Cocaine | 1068 | (17.0) | 163 | (46.1) | 500 | (44.8) | 118 | (37.3) | 253 | (43.0) | 6 | (37.5) |
| GHB | 427 | (6.8) | 48 | (43.6) | 199 | (47.3) | 49 | (43.4) | 114 | (44.5) | 3 | (60.0) |
| Inhalants | 2455 | (39.1) | 382 | (70.0) | 1085 | (66.4) | 360 | (66.7) | 543 | (66.8) | 23 | (74.2) |
| Ketamine | 157 | (2.5) | 29 | (26.6) | 63 | (19.9) | 14 | (16.5) | 47 | (23.0) | -- | -- |
| LSD | 427 | (6.8) | 60 | (25.5) | 184 | (25.2) | 55 | (24.2) | 111 | (26.7) | 3 | (21.4) |
| MDMA | 623 | (9.9) | 87 | (27.3) | 295 | (30.3) | 66 | (25.3) | 148 | (27.1) | 6 | (40.0) |
| Methamphetamine | 807 | (12.9) | 82 | (50.3) | 391 | (57.0) | 90 | (50.6) | 214 | (54.7) | 5 | (71.4) |
| Opioids | 97 | (1.5) | 11 | (19.3) | 53 | (24.5) | 13 | (25.0) | 18 | (16.5) | -- | -- |
| Rx Opioids§ | 724 | (11.5) | 93 | (34.2) | 359 | (38.1) | 98 | (34.6) | 145 | (31.1) | 5 | (41.7) |
| Rx Stimulants¥ | 859 | (13.7) | 100 | (31.9) | 437 | (38.5) | 132 | (38.9) | 150 | (29.2) | 3 | (20.0) |
| Rx Sedatives£ | 1117 | (17.8) | 134 | (44.8) | 586 | (49.6) | 149 | (44.2) | 210 | (39.4) | 6 | (42.9) |
At least once in the past 3 months
All prescription substances include participants with valid and those
U.S. Census region was calculated from valid address information. N = 193 participants are missing region information.
5.9% used prescription opioids without a prescription
8% used prescription stimulants without a prescription
7.9% used precription sedatives without a prescription
Alcohol and substance use among participants
Reported substance use and AUDIT/ASSIST scores are presented in Tables 1 and 2. Most participants (85.6%, which includes 407 out of 540 participants under the legal drinking age of 21) reported drinking alcohol in the last 12 months, and the mean AUDIT score was 7.3 (s = 6.0). Nineteen percent of participants reported AUDIT scores indicative of alcohol misuse (≥ 11). After alcohol, the second most commonly, reported recently used substance was cannabis (53.9%, ASSIST = 7.19, s = 7.8), and the third was inhalants (39.1%, ASSIST = 4.7, s = 5.6). Seventeen percent of our sample reported recent cocaine use (ASSIST = 3.9, s = 6.0), and 17.8% reported recent sedatives use (7.9% without a prescription, ASSIST = 4.4, s = 6.6). More than ten percent of participants reported recent methamphetamine use (12.9%, ASSIST = 10.2, s = 11.7), prescription opioid use (11.5% overall recent use and 5.9% reported use without a prescription, ASSIST = 3.7, s = 6.7), and prescription stimulant use (13% overall recent use and 8% reported use without a prescription, ASSIST = 3.3, s = 5.7). For cannabis (55.8%) and methamphetamine use (52.6%), at least half of self-reported users received ASSIST scores indicative of substance misuse. Other drugs with more than one-third of participants who self-reported using received ASSIST scores indicating misuse were GHB (35.6%), inhalants (43.7%), and sedatives (37.6%). The Cronbach’s alpha scores (raw and standardized) for all substances were ≥ 0.85 indicating high internal reliability of the scales within our study population (Table 2).
Table 2.
Mean AUDIT and ASSIST scores, Together 5,000 study, 2017–2018, n = 6,280
| Continuous Scores | Categorical Scores | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| Non-problematic | Misuse | ||||||||||
| Mean AUDIT and ASSIST scores | n | Raw Cronbach’s Alpha | Std Cronbach’s Alpha | Mean () | Std. Dev. (s) | Median | IQR | n | % | n | % |
| AUDIT Score | 5374 | 0.88 | 0.89 | 7.34 | 6.0 | 5 | 3 -- 10 | 4191 | 78.0 | 1183 | 22.0 |
| ASSIST Cannabis Score | 4892 | 0.89 | 0.90 | 7.19 | 7.8 | 5 | 0 -- 12 | 2165 | 44.2 | 2727 | 55.8 |
| ASSIST Cocaine Score | 2447 | 0.88 | 0.89 | 3.87 | 6.0 | 2 | 0 -- 5 | 1671 | 68.2 | 776 | 31.8 |
| ASSIST GHB Score | 927 | 0.87 | 0.88 | 4.80 | 7.6 | 2 | 0 -- 6 | 598 | 64.4 | 329 | 35.6 |
| ASSIST Inhalants Score | 3649 | 0.87 | 0.89 | 4.65 | 5.6 | 2 | 0 -- 7 | 2053 | 56.3 | 1596 | 43.7 |
| ASSIST Ketamine Score | 733 | 0.87 | 0.88 | 1.87 | 4.6 | 0 | 0 -- 2 | 623 | 85.0 | 110 | 15.0 |
| ASSIST LSD Score | 1659 | 0.87 | 0.88 | 1.86 | 4.0 | 0 | 0 -- 2 | 1371 | 82.6 | 288 | 17.4 |
| ASSIST MDMA Score | 2165 | 0.87 | 0.88 | 2.25 | 4.5 | 0 | 0 -- 3 | 1724 | 79.6 | 441 | 20.4 |
| ASSIST Methamphetamine Score | 1460 | 0.90 | 0.90 | 10.23 | 11.7 | 5 | 0 -- 19 | 692 | 47.4 | 768 | 52.6 |
| ASSIST Opioids Score (Street) | 445 | 0.85 | 0.88 | 4.51 | 8.6 | 0 | 0 -- 6 | 323 | 72.6 | 122 | 27.4 |
| ASSIST Rx Opioids Score* | 2022 | 0.87 | 0.88 | 3.67 | 6.7 | 0 | 0 -- 5 | 1450 | 71.7 | 572 | 28.3 |
| ASSIST Rx Stimulants Score* | 2377 | 0.87 | 0.88 | 3.31 | 5.7 | 0 | 0 -- 5 | 1713 | 72.1 | 664 | 27.9 |
| ASSIST Rx Sedatives Score* | 2430 | 0.87 | 0.88 | 4.40 | 6.6 | 2 | 0 -- 6 | 1517 | 62.4 | 913 | 37.6 |
Scores include participants with and without valid prescriptions
Factors associated with any regional differences in substance use
The main goal of our analysis was to determine regional factors associated with substance use and misuse among our geographically diverse cohort. Bivariate differences (p-value ≤ 0.05) between recent substance use and no recent substance use by U.S. region of residence were identified for cannabis, prescription stimulants, and prescription sedatives while differences between problematic and non-problematic use by region were identified for cannabis, cocaine, prescription opioids, prescription stimulants, and prescription sedatives (results not presented) (Supplemental Tables 1 and 2). We then further investigated these substances using multivariable models to determine if differences by region were sustained after accounting for other factors (Table 3).
Table 3.
Associations of demographic and geographic characteristics with cannabis, cocaine, prescription opioid, and sedative recent use, Together 5,000 study, 2017–2018, n = 6,280.
| Cannabis | Rx Stimulants | Rx Sedatives | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 95% Confidence Interval | 95% Confidence Interval | 95% Confidence Interval | ||||||||||
| Characteristic | aOR | LL | UL | p-value | aOR | LL | UL | p-value | aOR | LL | UL | p-value |
| Age | ||||||||||||
| 16–24 years old | Ref | Ref | Ref | |||||||||
| 25–35 years old | 0.52 | 0.43 -- 0.63 | <.0001 | 0.63 | 0.50 -- 0.79 | <.0001 | 0.74 | 0.58 -- 0.94 | 0.01 | |||
| 36–45 years old | 0.31 | 0.25 -- 0.38 | <.0001 | 0.49 | 0.37 -- 0.65 | <.0001 | 0.89 | 0.68 -- 1.16 | 0.38 | |||
| 46–55 years old | 0.24 | 0.18 -- 0.34 | <.0001 | 0.56 | 0.35 -- 0.92 | 0.02 | 1.11 | 0.75 -- 1.62 | 0.61 | |||
| Gender | ||||||||||||
| Cisgender male | Ref | -- | -- | |||||||||
| Transgender female | 1.13 | 0.47 -- 2.68 | 0.78 | -- | -- | |||||||
| Transgender male | 1.26 | 0.55 -- 2.89 | 0.58 | -- | -- | |||||||
| Other gender identity | 2.85 | 1.20 -- 6.78 | 0.02 | -- | -- | |||||||
| Race/Ethnicity | ||||||||||||
| White | Ref | -- | -- | |||||||||
| Black or African American | 1.49 | 1.18 -- 1.88 | 0.001 | -- | -- | |||||||
| Latino/Latinx | 1.17 | 0.99 -- 1.37 | 0.06 | -- | -- | |||||||
| Asian or Pacific Islander | 1.66 | 1.10 -- 2.52 | 0.02 | -- | -- | |||||||
| Multiracial/Other | 1.16 | 0.91 -- 1.47 | 0.22 | -- | -- | |||||||
| Employment Status (current) | ||||||||||||
| Full-time | Ref | Ref | -- | |||||||||
| Part-time | 1.48 | 1.20 -- 1.82 | 0.0002 | 1.04 | 0.80 -- 1.36 | 0.74 | -- | |||||
| Working or full-time student | 1.08 | 0.87 -- 1.35 | 0.48 | 1.62 | 1.23 -- 2.13 | 0.001 | -- | |||||
| Unemployed/Other | 0.94 | 0.75 -- 1.17 | 0.58 | 1.07 | 0.80 -- 1.42 | 0.66 | -- | |||||
| Highest level of Education | ||||||||||||
| < High school diploma | 1.94 | 1.08 -- 3.50 | 0.03 | -- | 1.92 | 1.06 -- 3.47 | 0.03 | |||||
| High school diploma or GED | Ref | -- | Ref | |||||||||
| Some college or associates degree | 0.97 | 0.79 -- 1.19 | 0.78 | -- | 1.07 | 0.84 -- 1.37 | 0.58 | |||||
| College graduate or higher | 0.84 | 0.68 -- 1.04 | 0.12 | -- | 0.97 | 0.75 -- 1.26 | 0.81 | |||||
| Housing instability | ||||||||||||
| Yes, within the last 5 years | 1.30 | 1.10 -- 1.55 | 0.003 | -- | -- | |||||||
| No or not within the last 5 years | Ref | -- | -- | |||||||||
| Sex work in the past 3 months | ||||||||||||
| Yes | 1.31 | 1.08 -- 1.60 | 0.01 | 1.63 | 1.31 -- 2.01 | <.0001 | 1.56 | 1.27 -- 1.92 | <.0001 | |||
| No | Ref | Ref | Ref | |||||||||
| Region | ||||||||||||
| Northeast | Ref | Ref | Ref | |||||||||
| Southeast | 0.76 | 0.63 -- 0.93 | 0.01 | 1.25 | 0.95 -- 1.64 | 0.11 | 1.13 | 0.87 -- 1.47 | 0.35 | |||
| Midwest | 0.91 | 0.72 -- 1.16 | 0.46 | 1.39 | 1.00 -- 1.93 | 0.05 | 0.95 | 0.69 -- 1.30 | 0.74 | |||
| West | 1.27 | 1.02 -- 1.59 | 0.03 | 0.88 | 0.64 -- 1.20 | 0.41 | 0.79 | 0.59 -- 1.05 | 0.11 | |||
| Pacific/US Territories/Other | 1.17 | 0.54 -- 2.52 | 0.69 | 0.50 | 0.13 -- 1.86 | 0.30 | 0.80 | 0.27 -- 2.40 | 0.69 | |||
Factors associated with regional differences in cannabis use and misuse
In multivariable models, cannabis had lower odds of being used in the Southeast (aOR = 0.76, 95% CI: 0.63–0.93, ref. Northeast) and higher odds of being used in the West (aOR = 1.27, 95% CI: 1.02–1.59, ref. Northeast). We also found that participants who identified as not-cisgender, reported being Black, reported being Asian or Pacific Islander, working part-time, having less than a high school education, having experienced housing in stability in the last 5 years, or engaging in sex work in the past 3 months had a higher odds of reporting recent cannabis use. Participants who were older had a lower odds of reporting recent cannabis use. Looking further into cannabis use, participants who reported living in the Southeast had a lower odds (aOR = 0.83, 95% CI: 0.69–0.99) of having ASSIST scores classified as misuse compared to participants in the Northeast. We found similar associations of participants who identified as not being a cis-gendered male, reported being Black or African American, working part-time, having experience housing in stability in past 5 years, or engaging in sex work in the past 3 months had a higher odds of scoring in the misuse range. Additionally, participants who made less than $20,000 annually or made $20,000-$49,999 annually had higher odds of reporting cannabis misuse. As with recent use, older participants had a lower odds of reporting cannabis use, and, additionally, participants who reported having at least a 4-year college degree also had lower odds of reporting cannabis misuse.
Factors associated with regional differences in prescription stimulants use and misuse
Adjusted analyses indicate that participants who reported residing in the Midwest (aOR = 1.39, 95% CI: 1.00–1.93, ref. Northeast) had a higher odds of reporting recent prescription stimulant use. Further, participants who reported being a working or full-time student or having engaged in sex work in the past 3 months had a higher odds of recent prescription stimulant use. Participants who were older had a lower odds of reporting recent prescription stimulant use. There were no statistically significant differences observed between non-problematic use and misuse between regions found for prescription stimulants, however participants who reported being unmarried, working part-time, being a working or full-time student, or engaging in sex work in the past 3 months had a higher odds of misusing prescription stimulants.
Factors associated with regional differences in prescription sedatives use and misuse
We found no statistically significant differences between regions for recent use nor misuse among our participants. We did find that participants who reported less than high school education or engaged in sex work in the past 3 months had a higher odds of recently using prescription sedatives while participants 25–34 years old had a lower odds of recent sedatives use. Participants who reported less than high school education, engaged in sex work in the past 3 months, and were older than 35 years of age had higher odds of prescription sedatives misuse. Conversely, participants who identified as trans women or other/multiracial had lower odds of prescription sedatives use.
Other factors related to substance use and misuse
We only explored regional differences for cocaine and prescription opioid misuse per the initial bivariate analysis but found no sustained statistically significant differences between regions in multivariable analyses. We did find that participants who identified as Black/African American, identified as Latinx, reported being unmarried, engaged in sex work in the past 3 months, or reported ever being incarcerated had a higher odds of misusing cocaine while participants who were 46–55 years old had a lower odds of misuse. For prescription opioids, we found that participants who engaged in sex work in the past 3 months or reported ever being incarcerated had a higher odds of prescription opioids misuse, and participants who had at least a college degree were less likely to misuse opioids.
Regional differences in factors affecting substance use and misuse
Finally, as a sub-analyses, we investigated the potential that living in the four main regions of the U.S.—Northeast (NE), South (S), Midwest (MW), and West (W)—modifies the association between covariates and substance use (cannabis, prescription stimulants, prescription sedatives) and misuse (cannabis, cocaine, prescription opioids, prescription stimulants, and prescription sedatives). These sub-analyses help to better elucidate differences between each region and the factors impacting substance use and misuse. Across most regions, differential impacts were determined for the following substances and use patterns:
Older participants had a higher odds of using (NE, S, MW, W) and misusing (N, S, W) cannabis, misusing cocaine (NE), using prescription stimulants (NE, S, MW), and using (NE) prescription sedatives. However, in the South and Midwest older participants had a higher odds of misusing prescription sedatives.
Sex workers had a higher odds of using (MW) and misusing (MW, W) cannabis, misusing cocaine (S, W), misusing opioids (S, W), using (NE, S, MW) and misusing (NE, S, MW) prescription stimulants, and using (S, MW, W) and misusing (NE, S, MW) prescription sedatives.
Participants who were unemployed had a higher odds of misusing prescription opioids (NE); employed part-time had a higher odds of using (S, MW) and misusing (S) cannabis; and student who were employed had higher odds of using prescription stimulants (S, MW, W).
Finally, participants who had a college degree or higher had a lower odds of misusing prescription opioids (S, W) while participants with less than a high school education had a higher odds of using prescription sedatives (MW, W).
Beyond these characteristics, there were also statistically significant differences for other factors among the regions:
Participants who made lower annual incomes had higher odds of misusing cannabis (S, W).
Participants who reported being Black or African American had a higher odds of using (S) and misusing (S) cannabis, misusing cocaine (NE, S), but participants who identified as being multiracial/ethnic or some other racial/ethnic background had a lower odds of misusing prescription sedatives (S).
Participants who lived in the South and identified as non-binary gender had a higher odds of misusing cannabis and transgender men and transgender women had lower odds of misusing prescription sedatives.
Additionally, participants who lived in the South and were not married had a higher odds of misusing cocaine.
Participants who reported being ever incarcerated and who lived in the Midwest had a higher odds of misusing prescription opioids.
Finally, participants who reported being unstably housed in the past 5 years and who resided in the West had a higher odds of using and misusing cannabis.
Discussion
Our findings document geographic differences in substance use and misuse in the U.S. among a geographically diverse sample of SGM. These findings are novel in that they represent a U.S. geographically diverse sample of over 6,000 men, trans men, and trans women who are vulnerable for contracting HIV. Results from this study highlight the importance of continued multi-faceted interventions for HIV prevention addressing substance use and sexual behaviors simultaneously. Adjusted logistic regression models suggested regional differences for recent use for cannabis and prescription stimulants as well as regional differences in misuse for cannabis. Results also highlight the potential high prevalence of substance misuse among persons who are at higher risk for HIV, and, specifically, high proportions of drug misuse among participants.
First, this study contributes to the literature by examining potential differences in substance misuse by U.S. region for men, trans men, and trans women who have sex with men. Initial analyses indicated that recent substance use differed for cannabis, prescription stimulants, and prescription sedatives; and misuse differed by regions for cannabis, cocaine, prescription opioids, prescription stimulants, and prescription sedatives. After adjusting for other factors known to be associated with substance use, we found that participants living in the Southeast had lower odds of using and misusing cannabis, participants living in the Western U.S. had a higher odds of using cannabis, and participants living in the Midwest had higher odds of using prescription stimulants. The findings related to cannabis are likely due to the legal status of cannabis in the Southeast—medical or no legal standing common—and the West—medical and recreational use common—and differing access and availability (National Conference of State Legislatures, 2019). Recent polling shows similar, positive support for marijuana legalization across regions regardless of political affiliation (McCarthy, 2019), though other cultural and contextual factors may influence legislative progress in Southeastern states (Felson, Adamczyk, & Thomas, 2019) barring further medical and initial recreational cannabis use legalization or slowing acceptance and implementation.
Prior research has noted that methamphetamine use is higher in the Midwest (Hirshfield et al., 2004), and we expected to see this represented in our data as well. We did not find differences for methamphetamine use for our participants living in different U.S. regions. However, we did find that participants in the Midwest had a higher odds of using prescription stimulants. This is notable because other research has linked methamphetamine use and prescription stimulant misuse (Wu, Pilowsky, Schlenger, & Galvin, 2007). So, while our data does not support higher methamphetamine use in the Midwest, there may be indications of higher prescription stimulant use that would be indicative of future methamphetamine use among our participants. More research is needed to elucidate the relationship between prescription stimulant misuse and transitioning to or co-use with methamphetamine. Given our high proportion of methamphetamine use, our findings of increased odds of prescription stimulant use in the Midwest may suggest a key difference in stimulant misuse in the region. We did not find other differences in substance use and misuse between regions; however, this could be because region is too broad of a geographic location (Hedegaard et al., 2018). We may also have not found differences between regions due to the nature of our cohort. One goal of our cohort is to identify modifiable factors related to HIV seroconversion. As such, we successfully recruited, by design, a population that was engaging in multiple high-risk behaviors regardless of where they lived. Sub-analyses further highlight that each substance and each geographic location has their own unique set of demographic, socioeconomic, and behavioral factors associated with substance use and misuse. Our results indicate that interventions, policies, and broader health programming should be tailored to meet the unique substance use needs of the area.
This study also highlights the high prevalence of substance use and misuse among men, trans men, and trans women at high risk for contracting HIV across the U.S. Compared to 2017 national estimates, T5K study participants had higher recent (past 3 months) usage of cannabis (53.9% vs. 15.3%), cocaine (17% vs. 2.4%), LSD (6.8% vs. 0.8%), ecstasy (9.9% vs. 0.9%), inhalants (39.1% vs. 0.5%), and methamphetamine (12.9% vs. 0.6%) than annual use among a nationally representative sample (Center for Behavioral Health Statistics and Quality, 2018). We also found higher proportions of prescription drug misuse—defined for comparison purposes among T5K participants as use without a prescription—among our participants for both stimulants (8% vs. 2.2%) and sedatives (7.9% vs. 0.5%) compared to a nationally representative sample (Center for Behavioral Health Statistics and Quality, 2018). Attempting to discern differences in all opioid use between the general population and our cohort is difficult due to lacking comparable data on use. However, in 2017, 17% of Americans filled a prescription for opioids (Centers for Disease Control and Prevention, 2018) which is higher than use reported among our participants.
Among MSM, our results suggest continued high rates of cocaine, ecstasy, inhalants, methamphetamines, GHB, and Ketamine use (Pantalone et al., 2010) and suggest the continued need to monitor national substance use trends among large populations of SGM is necessary. In addition to high rates of any recent use, we found high proportions of participants who had scores indicating substance misuse. Over a third of our participants were scored as misusing cannabis, GHB, inhalants, methamphetamine and sedatives. The high proportion of GHB misuse is especially concerning, given the high rates of overdose associated with GHB use (Degenhardt, Darke, & Dillon, 2002). Our findings highlight the need to address GHB use and misuse among MSM at risk for HIV acquisition. Many of these substances have well-documented health risks (Fenton et al., 2010) and have been associated with sexual risk behaviors leading to HIV seroconversion (Chesney, Barrett, & Stall, 1998; Hirshfield et al., 2004; Javanbakht, Ragsdale, Shoptaw, & Gorbach, 2018; Koblin et al., 2003; Ostrow et al., 1993; Parsons, Grov, & Golub, 2012; Stall et al., 2003) highlighting the need for multifaceted strategies that can address both HIV risk in tandem with substance use and misuse.
Limitations
Our findings should be interpreted in light of study limitations. First, this study was mainly online and subject to fraudulent or duplicate participants; however, we took measures to prevent and remove such participants—such as blocking multiple submissions, recording IP address, requiring unique and valid mailing addresses. Next, this study primarily used self-reported data, and, therefore, data collected on sensitive topics, like substance use, may not accurately represent actual substance use due to response biases. Our study also recruited persons who were at high risk for contracting HIV. Of note, in addition to our core eligibility criteria, potential participants were enrolled into the study if they met at least one additional criterion indicating higher risk for HIV seroconversion. Among these criteria were sharing needles to use drugs in the past 12 months or using methamphetamine in the past 3 months. Although this could have led to an over-enrolment of participants who met these specific drug-related criteria, we note that 94% of methamphetamine users and 97% of those having shared needles otherwise also met inclusion criteria that made them eligible to enroll. A limitation of the measurement of substance use is with the recommended thresholds from the ASSIST scale in misuse of cannabis. With the legalization of cannabis in certain states and under certain conditions, scoring cannabis like all other illicit substances may not be appropriate. A limitation unique to the current study is the lack of additional information surrounding prescription drug use due to survey programming logic errors. Note, these questions did not impact the ASSIST scores used for this analysis for prescription drug use. Finally, while we have large proportions of substance users and misusers for certain substances, we have smaller frequencies for some (e.g., street opioids) and for certain geographic regions impacting the precision of our estimates.
Conclusion
Our results highlight persistent differences of substance use and misuse among men, trans men, and trans women compared to U.S. national estimates. Despite recent attention to high rates of opioid overdose and deaths across the U.S., the prescription opioid use appears to be proportionate to the general population among SGM. Rather, results suggest concerning differences in substance use and misuse of methamphetamine and GHB that have noted harmful health effects and would be better directions for future research and prevention among SGM. Our results reflect that considering place and substance type when developing and delivering public health interventions for men, trans men, and trans women who have sex with men may be critical to developing optimally effective, multifaceted health interventions and innovative delivery methods to reach geographically diverse persons who are at most risk for HIV seroconversion.
Supplementary Material
Table 4.
Associations of demographic and geographic characteristics with cannabis, cocaine, prescription opioid, and sedative misuse, Together 5,000 study, 2017–2018, n = 6,280.
| Cannabis | Cocaine | Rx Opioids | Rx Stimulants | Rx Sedatives | ||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 95% Confidence Interval | 95% Confidence Interval | 95% Confidence Interval | 95% Confidence Interval | 95% Confidence Interval | ||||||||||||||||
| Characteristic | aOR | LL | UL | p-value | aOR | LL | UL | p-value | aOR | LL | UL | p-value | aOR | LL | UL | p-value | aOR | LL | UL | p-value |
| Age | ||||||||||||||||||||
| 16–24 years old | Ref | -- | -- | -- | Ref | |||||||||||||||
| 25–35 years old | 0.60 | 0.50 -- 0.71 | <.0001 | 0.89 | 0.69 -- 1.14 | 0.36 | -- | -- | 0.97 | 0.74 -- 1.26 | 0.80 | |||||||||
| 36–45 years old | 0.44 | 0.36 -- 0.55 | <.0001 | 0.83 | 0.62 -- 1.11 | 0.21 | -- | -- | 1.41 | 1.04 -- 1.90 | 0.03 | |||||||||
| 46–55 years old | 0.35 | 0.26 -- 0.48 | <.0001 | 0.60 | 0.37 -- 0.96 | 0.03 | -- | -- | 1.78 | 1.18 -- 2.69 | 0.01 | |||||||||
| Gender | ||||||||||||||||||||
| Cisgender male | Ref | -- | -- | -- | Ref | |||||||||||||||
| Transgender female | 0.88 | 0.41 -- 1.90 | 0.74 | -- | -- | -- | 0.31 | 0.11 -- 0.85 | 0.02 | |||||||||||
| Transgender male | 2.17 | 0.95 -- 4.94 | 0.06 | -- | -- | -- | 2.13 | 0.87 -- 5.19 | 0.10 | |||||||||||
| Non-binary (male at birth) | 2.43 | 1.23 -- 4.79 | 0.01 | -- | -- | -- | 1.50 | 0.71 -- 3.17 | 0.29 | |||||||||||
| Race/Ethnicity | ||||||||||||||||||||
| White | Ref | Ref | -- | -- | Ref | |||||||||||||||
| Black or African American | 1.26 | 1.02 -- 1.56 | 0.03 | 1.62 | 1.15 -- 2.29 | 0.01 | -- | -- | 0.80 | 0.57 -- 1.13 | 0.21 | |||||||||
| Latino/Latinx | 1.02 | 0.88 -- 1.18 | 0.82 | 1.27 | 1.03 -- 1.58 | 0.03 | -- | -- | 0.97 | 0.78 -- 1.20 | 0.78 | |||||||||
| Asian or Pacific Islander | 0.80 | 0.56 -- 1.14 | 0.21 | 1.00 | 0.49 -- 2.04 | 1.00 | -- | -- | 0.95 | 0.47 -- 1.89 | 0.88 | |||||||||
| Multiracial/Other | 1.14 | 0.92 -- 1.42 | 0.24 | 0.93 | 0.68 -- 1.27 | 0.66 | -- | -- | 0.66 | 0.48 -- 0.90 | 0.01 | |||||||||
| Sexual orientation | ||||||||||||||||||||
| Gay, Queer, Homosexual | Ref | -- | Ref | -- | Ref | |||||||||||||||
| Bisexual | 1.11 | 0.93 -- 1.33 | 0.23 | -- | 1.08 | 0.81 -- 1.44 | 0.60 | -- | -- | |||||||||||
| Other | 1.49 | 0.81 -- 2.74 | 0.20 | -- | 1.55 | 0.73 -- 3.29 | 0.26 | -- | -- | |||||||||||
| Marital status | ||||||||||||||||||||
| No | -- | 1.38 | 1.05 -- 1.82 | 0.02 | -- | 1.37 | 1.01 -- 1.85 | 0.04 | -- | |||||||||||
| Married or civil union | Ref | -- | Ref | -- | Ref | |||||||||||||||
| Employment Status (current) | ||||||||||||||||||||
| Full-time | Ref | -- | Ref | -- | Ref | |||||||||||||||
| Part-time | 1.28 | 1.06 -- 1.55 | 0.01 | -- | 0.82 | 0.59 -- 1.15 | 0.25 | 1.31 | 1.00 -- 1.73 | 0.05 | 1.12 | 0.86 -- 1.46 | 0.41 | |||||||
| Working or full-time student | 0.90 | 0.73 -- 1.11 | 0.33 | -- | 0.89 | 0.60 -- 1.31 | 0.55 | 1.40 | 1.06 -- 1.84 | 0.02 | 1.14 | 0.83 -- 1.58 | 0.41 | |||||||
| Unemployed/Other | 0.93 | 0.74 -- 1.17 | 0.54 | -- | 1.37 | 0.98 -- 1.90 | 0.06 | 1.17 | 0.86 -- 1.58 | 0.31 | 1.12 | 0.83 -- 1.50 | 0.46 | |||||||
| Highest level of Education | ||||||||||||||||||||
| < High school diploma | 1.52 | 0.94 -- 2.46 | 0.08 | -- | 1.64 | 0.86 -- 3.14 | 0.13 | -- | 1.90 | 1.05 -- 3.41 | 0.03 | |||||||||
| High school diploma or GED | Ref | -- | Ref | -- | Ref | |||||||||||||||
| Some college or associates degree | 0.94 | 0.78 -- 1.13 | 0.49 | -- | 0.77 | 0.58 -- 1.03 | 0.07 | -- | 0.93 | 0.72 -- 1.20 | 0.56 | |||||||||
| College graduate or higher | 0.81 | 0.66 -- 0.99 | 0.04 | -- | 0.58 | 0.42 -- 0.81 | 0.001 | -- | 0.99 | 0.75 -- 1.32 | 0.96 | |||||||||
| Income | ||||||||||||||||||||
| Less than $20,000 | 1.24 | 1.01 -- 1.53 | 0.04 | 0.91 | 0.70 -- 1.19 | 0.50 | 1.10 | 0.77 -- 1.57 | 0.58 | -- | 1.12 | 0.83 -- 1.51 | 0.45 | |||||||
| $20,000-$49,999 | 1.21 | 1.03 -- 1.41 | 0.02 | 0.89 | 0.70 -- 1.13 | 0.33 | 1.10 | 0.83 -- 1.45 | 0.50 | -- | 1.01 | 0.80 -- 1.28 | 0.92 | |||||||
| $50,000 or more | Ref | -- | Ref | -- | Ref | |||||||||||||||
| Housing instability | ||||||||||||||||||||
| Yes, within the last 5 years | 1.33 | 1.14 -- 1.56 | 0.0004 | -- | 1.11 | 0.98 -- 1.25 | 0.10 | 1.02 | 0.92 -- 1.14 | 0.70 | 1.17 | 0.95 -- 1.44 | 0.14 | |||||||
| No or not within the last 5 years | Ref | -- | Ref | -- | Ref | |||||||||||||||
| Sex work in the past 3 months | ||||||||||||||||||||
| Yes | 1.40 | 1.17 -- 1.67 | 0.0002 | 1.70 | 1.37 -- 2.10 | <.0001 | 1.58 | 1.23 -- 2.03 | 0.0003 | 2.03 | 1.62 -- 2.56 | <.0001 | 1.73 | 1.39 -- 2.16 | <.0001 | |||||
| No | Ref | -- | Ref | -- | Ref | |||||||||||||||
| Incarcerated (ever) | ||||||||||||||||||||
| Yes | 1.20 | 1.01 -- 1.43 | 0.03 | 1.12 | 1.01 -- 1.25 | 0.03 | 1.13 | 1.00 -- 1.27 | 0.04 | -- | -- | |||||||||
| No | Ref | -- | Ref | -- | Ref | |||||||||||||||
| Region | ||||||||||||||||||||
| Northeast | Ref | -- | Ref | -- | Ref | |||||||||||||||
| Southeast | 0.83 | 0.69 -- 0.99 | 0.04 | 0.96 | 0.74 -- 1.25 | 0.76 | 1.23 | 0.89 -- 1.71 | 0.20 | 1.11 | 0.84 -- 1.49 | 0.46 | 1.13 | 0.86 -- 1.48 | 0.39 | |||||
| Midwest | 0.92 | 0.74 -- 1.15 | 0.46 | 0.72 | 0.51 -- 1.01 | 0.06 | 1.12 | 0.76 -- 1.66 | 0.57 | 1.18 | 0.83 -- 1.67 | 0.36 | 0.85 | 0.60 -- 1.18 | 0.33 | |||||
| West | 1.19 | 0.98 -- 1.46 | 0.08 | 0.81 | 0.60 -- 1.09 | 0.16 | 0.79 | 0.55 -- 1.14 | 0.21 | 0.82 | 0.59 -- 1.14 | 0.24 | 0.84 | 0.62 -- 1.14 | 0.26 | |||||
| Pacific/US Territories/Other | 0.70 | 0.35 -- 1.39 | 0.30 | 0.52 | 0.14 -- 1.89 | 0.32 | 0.86 | 0.22 -- 3.37 | 0.83 | 0.00 | 0.00 -- 1.00 | 1.06 | 0.35 -- 3.24 | 0.91 | ||||||
Acknowledgements:
Special thanks to additional members of the T5K study team: Sarit A. Golub, Viraj V. Patel, Gregorio Millett, Don Hoover, Sarah Kulkarni, Matthew Stief, Caitlin MacCrate, Chloe Mirzayi, Alexa D’Angelo, Corey Morrison, Javier Lopez-Rios, & Pedro B. Carneiro. Thank you to the program staff at NIH: Gerald Sharp, Sonia Lee, and Michael Stirratt. And thank you to the members of our Scientific Advisory Board: Michael Camacho, Demetre Daskalakis, Sabina Hirshfield, Jeremiah Johnson, Claude Mellins, and Milo Santos. Finally, a very special thanks to Meredith Ray for her coding and programming expertise. While the NIH financially supported this research, the content is the responsibility of the authors and does not necessarily reflect official views of the NIH. Parts of this paper were presented at the 2019 National LGBTQ Health Conference May 31-June 1 in Atlanta, GA.
Funding: This work was supported by the National Institutes of Health [UG3 AI 133675 - PI Grov]. Other forms of support include the CUNY Institute for Implementation Science in Population Health, the Einstein, Rockefeller, CUNY Center for AIDS Research (ERC CFAR, P30 AI 124414).
Footnotes
Disclosure statement: The authors declare no conflicts of interest.
References:
- Babor TF, Higgins-Biddle JC, Saunders JB, & Monteiro MG (2001). AUDIT: The alcohol use disorders identification test: Guidelines for use in primary care (2nd ed.). WHO [Google Scholar]
- Bolton SL, & Sareen J. (2011). Sexual orientation and its relation to mental disorders and suicide attempts: findings from a nationally representative sample. Can J Psychiatry, 56(1), 35–43. doi: 10.1177/070674371105600107 [DOI] [PubMed] [Google Scholar]
- Boyd C, West B, & McCabe S. (2018). Does misuse lead to a disorder? The misuse of prescription tranquilizer and sedative medications and subsequent substance use disorders in a US longitudinal sample. Addictive Behaviors, 79, 17–23. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Carrico AW, Pollack LM, Stall RD, Shade SB, Neilands TB, Rice TM, . . . Moskowitz JT (2012). Psychological processes and stimulant use among men who have sex with men. Drug and alcohol dependence, 123(1–3), 79–83. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Center for Behavioral Health Statistics and Quality. (2018). 2017 National Survey on Drug Use and Health: Detailed Tables. In. Rockville, Maryland: Substance Abuse and Mental Health Services Administration. [Google Scholar]
- Centers for Disease Control and Prevention. (2018). 2018 Annual Surveillance Report of Drug-Related Risks and Outcomes-United States. Retrieved from https://www.cdc.gov/drugoverdose/pdf/pubs/2018-cdc-drug-surveillance-report.pdf [Google Scholar]
- Chesney MA, Barrett DC, & Stall R. (1998). Histories of substance use and risk behavior: precursors to HIV seroconversion in homosexual men. Am J Public Health, 88(1), 113–116. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Ciccarone D, & Bourgois P. (2003). Explaining the geographical variation of HIV among injection drug users in the United States. Substance use & misuse, 38(14), 2049–2063. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Closson EF, Mitty JA, Malone J, Mayer KH, & Mimiaga MJ (2018). Exploring strategies for PrEP adherence and dosing preferences in the context of sexualized recreational drug use among MSM: a qualitative study. AIDS care, 30(2), 191–198. [DOI] [PubMed] [Google Scholar]
- Cochran SD, Ackerman D, Mays VM, & Ross MW (2004). Prevalence of non‐medical drug use and dependence among homosexually active men and women in the US population. Addiction, 99(8), 989–998. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Degenhardt L, Darke S, & Dillon P. (2002). GHB use among Australians: characteristics, use patterns and associated harm. Drug and alcohol dependence, 67(1), 89–94. [DOI] [PubMed] [Google Scholar]
- Evans-Polce RJ, Veliz PT, Boyd CJ, Hughes TL, & McCabe SE (2019). Associations between sexual orientation discrimination and substance use disorders: differences by age in US adults. Social psychiatry and psychiatric epidemiology, 1–10. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Felson J, Adamczyk A, & Thomas C. (2019). How and why have attitudes about cannabis legalization changed so much? Social Science Research, 78, 12–27. [DOI] [PubMed] [Google Scholar]
- Fenton MC, Keyes KM, Martins SS, & Hasin DS (2010). The role of a prescription in anxiety medication use, abuse, and dependence. American Journal of Psychiatry, 167(10), 1247–1253. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Gattis MN, Sacco P, & Cunningham-Williams RM (2012). Substance use and mental health disorders among heterosexual identified men and women who have same-sex partners or same-sex attraction: results from the national epidemiological survey on alcohol and related conditions. Archives of sexual behavior, 41(5), 1185–1197. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Gilman SE, Cochran SD, Mays VM, Hughes M, Ostrow D, & Kessler RC (2001). Risk of psychiatric disorders among individuals reporting same-sex sexual partners in the National Comorbidity Survey. Am J Public Health, 91(6), 933. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Grov C, Westmoreland DA, Carneiro PB, Stief M, MacCrate C, Mirzayi C, . . . Nash D. (2019). Recruiting vulnerable populations to participate in HIV prevention research: findings from the Together 5000 cohort study. Ann Epidemiol, 35, 4–11. doi: 10.1016/j.annepidem.2019.05.003 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Halkitis PN, Green KA, & Mourgues P. (2005). Longitudinal investigation of methamphetamine use among gay and bisexual men in New York City: findings from Project BUMPS. Journal of Urban Health, 82(1), i18–i25. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Hatzenbuehler ML, Corbin WR, & Fromme K. (2008). Trajectories and determinants of alcohol use among LGB young adults and their heterosexual peers: results from a prospective study. Developmental Psychology, 44(1), 81. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Hedegaard H, Warner M, & Miniño AM (2018). Drug overdose deaths in the United States, 1999–2017: US Department of Health and Human Services, Centers for Disease Control and; …. [Google Scholar]
- Hirshfield S, Remien RH, Humberstone M, Walavalkar I, & Chiasson MA (2004). Substance use and high-risk sex among men who have sex with men: a national online study in the USA. AIDS care, 16(8), 1036–1047. [DOI] [PubMed] [Google Scholar]
- Humeniuk R, Henry-Edwards S, Ali R, Poznyak V, Monteiro MG, & Organization WH (2010). The Alcohol, Smoking and Substance Involvement Screening Test (ASSIST): manual for use in primary care (9522727083). Retrieved from [Google Scholar]
- Javanbakht M, Ragsdale A, Shoptaw S, & Gorbach PM (2018). Transactional sex among men who have sex with men: differences by substance use and HIV status. Journal of Urban Health, 1–13. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Kerridge BT, Pickering RP, Saha TD, Ruan WJ, Chou SP, Zhang H, . . . Hasin DS (2017). Prevalence, sociodemographic correlates and DSM-5 substance use disorders and other psychiatric disorders among sexual minorities in the United States. Drug and alcohol dependence, 170, 82–92. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Keuroghlian AS, Reisner SL, White JM, & Weiss RD (2015). Substance use and treatment of substance use disorders in a community sample of transgender adults. Drug and alcohol dependence, 152, 139–146. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Keyes KM, Cerdá M, Brady JE, Havens JR, & Galea S. (2014). Understanding the rural–urban differences in nonmedical prescription opioid use and abuse in the United States. Am J Public Health, 104(2), e52–e59. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Koblin BA, Chesney MA, Husnik MJ, Bozeman S, Celum CL, Buchbinder S, . . . Huang Y. (2003). High-risk behaviors among men who have sex with men in 6 US cities: baseline data from the EXPLORE Study. Am J Public Health, 93(6), 926–932. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Marshal MP, Friedman MS, Stall R, King KM, Miles J, Gold MA, . . . Morse JQ (2008). Sexual orientation and adolescent substance use: a meta‐analysis and methodological review. Addiction, 103(4), 546–556. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Marshal MP, Friedman MS, Stall R, & Thompson AL (2009). Individual trajectories of substance use in lesbian, gay and bisexual youth and heterosexual youth. Addiction, 104(6), 974–981. [DOI] [PMC free article] [PubMed] [Google Scholar]
- McCabe SE, Hughes TL, Bostwick WB, West BT, & Boyd CJ (2009). Sexual orientation, substance use behaviors and substance dependence in the United States. Addiction, 104(8), 1333–1345. [DOI] [PMC free article] [PubMed] [Google Scholar]
- McCabe SE, & West BT (2017). The three-year course of multiple substance use disorders in the United States: a national longitudinal study. The Journal of clinical psychiatry, 78(5), e537. [DOI] [PMC free article] [PubMed] [Google Scholar]
- McCabe SE, West BT, Hughes TL, & Boyd CJ (2013). Sexual orientation and substance abuse treatment utilization in the United States: results from a national survey. J Subst Abuse Treat, 44(1), 4–12. doi: 10.1016/j.jsat.2012.01.007 [DOI] [PMC free article] [PubMed] [Google Scholar]
- McCarthy J. (2019). Two in Three Americans Now Support Legalizing Marijuana. Retrieved from Gallup https://news.gallup.com/poll/243908/two-three-americans-support-legalizing-marijuana.aspx
- Meyer IH (2003). Prejudice, social stress, and mental health in lesbian, gay, and bisexual populations: conceptual issues and research evidence. Psychological Bulletin, 129(5), 674–697. doi: 10.1037/0033-2909.129.5.674 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Nash D, Stief M, MacCrate C, Mirzayi C, Patel VV, Hoover D, . . . Grov C. (2019). A Web-Based Study of HIV Prevention in the Era of Pre-Exposure Prophylaxis Among Vulnerable HIV-Negative Gay and Bisexual Men, Transmen, and Transwomen Who Have Sex With Men: Protocol for an Observational Cohort Study. JMIR Res Protoc, 8(9), e13715. doi: 10.2196/13715 [DOI] [PMC free article] [PubMed] [Google Scholar]
- National Conference of State Legislatures. (2019, July 2, 2019). State Medical Marijuana Laws. Retrieved from http://www.ncsl.org/research/health/state-medical-marijuana-laws.aspx
- National Institute on Drug Abuse, National Institutes of Health, & U.S. Department of Health and Human Services. (2018).
- National Institute on Drug Abuse (NIDA): Heroin. National Institute on Drug Abuse, National Institutes of Health, & U.S. Department of Health and Human Services. (2019). National Institute on Drug Abuse (NIDA): Methamphetamine. Retrieved from [Google Scholar]
- Ostrow DG, Beltran ED, Joseph JG, DiFranceisco W, Wesch J, & Chmiel JS (1993). Recreational drugs and sexual behavior in the Chicago MACS/CCS cohort of homosexually active men. Chicago Multicenter AIDS Cohort Study (MACS)/Coping and Change Study. Journal of Substance Abuse, 5(4), 311–325. doi: 10.1016/0899-3289(93)90001-r [DOI] [PubMed] [Google Scholar]
- Pantalone DW, Bimbi DS, Holder CA, Golub SA, & Parsons JT (2010). Consistency and change in club drug use by sexual minority men in New York City, 2002 to 2007. Am J Public Health, 100(10), 1892–1895. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Parsons JT, Grov C, & Golub SA (2012). Sexual compulsivity, co-occurring psychosocial health problems, and HIV risk among gay and bisexual men: further evidence of a syndemic. Am J Public Health, 102(1), 156–162. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Reback CJ, & Fletcher JB (2014). HIV prevalence, substance use, and sexual risk behaviors among transgender women recruited through outreach. AIDS Behav, 18(7), 1359–1367. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Schuler MS, Rice CE, Evans-Polce RJ, & Collins RL (2018). Disparities in substance use behaviors and disorders among adult sexual minorities by age, gender, and sexual identity. Drug and alcohol dependence, 189, 139–146. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Stall R, Mills TC, Williamson J, Hart T, Greenwood G, Paul J, . . . Catania JA (2003). Association of co-occurring psychosocial health problems and increased vulnerability to HIV/AIDS among urban men who have sex with men. Am J Public Health, 93(6), 939–942. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Stall R, Paul JP, Greenwood G, Pollack LM, Bein E, Crosby GM, . . . Catania JA (2001). Alcohol use, drug use and alcohol‐related problems among men who have sex with men: the Urban Men’s Health Study. Addiction, 96(11), 1589–1601. [DOI] [PubMed] [Google Scholar]
- Sullivan PS, Nakashima AK, Purcell DW, & Ward JW (1998). Geographic differences in noninjection and injection substance use among HIV-seropositive men who have sex with men: western United States versus other regions. Supplement to HIV/AIDS Surveillance Study Group. Journal of acquired immune deficiency syndromes and human retrovirology: official publication of the International Retrovirology Association, 19(3), 266–273. [DOI] [PubMed] [Google Scholar]
- Swank E, Frost DM, & Fahs B. (2012). Rural location and exposure to minority stress among sexual minorities in the United States. Psychology & Sexuality, 3(3), 226–243. [Google Scholar]
- Tobin KE, Yang C, King K, Latkin CA, & Curriero FC (2016). Associations between drug and alcohol use patterns and sexual risk in a sample of African American men who have sex with men. AIDS Behav, 20(3), 590–599. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Wu L-T, Pilowsky DJ, Schlenger WE, & Galvin DM (2007). Misuse of methamphetamine and prescription stimulants among youths and young adults in the community. Drug and alcohol dependence, 89(2–3), 195–205. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.