Abstract
Purpose: Sexual minority individuals have heightened risk for substance use; however, previous studies have not assessed severity of alcohol use disorders (AUDs), tobacco use disorders (TUDs), and drug use disorders (DUDs) among lesbian/gay and bisexual individuals and those “not sure” of their sexual identity compared with heterosexual individuals. This study examined how three dimensions of sexual orientation (identity, attraction, and behavior) relate to severity of AUD, TUD, and DUD.
Methods: This study used cross-sectional national data (N = 36,309) from the National Epidemiologic Survey on Alcohol and Related Conditions-III, and well-validated alcohol, tobacco, and drug measures that align with Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition criteria. Data were collected through in-person interviews in 2012–2013.
Results: Sexual minority respondents, based on sexual identity, had higher odds of severe AUD or TUD than heterosexual respondents. Opposite- and same-sex behavior was a predictor of severe AUD (adjusted odds ratio [AOR] = 2.44; 95% confidence interval [CI] = 1.24–4.79) and TUD (AOR = 2.16; 95% CI = 1.19–3.93), but not DUD. Those “not sure” of their sexual identity had higher odds of severe AUD, TUD, and DUD: AUD (AOR = 5.05; 95% CI = 2.78–9.16), TUD (AOR = 4.18; 95% CI = 2.29–7.64), and DUD (AOR = 4.40; 95% CI = 1.72–11.2), than heterosexual respondents. There were few significant differences between “not sure” and bisexual respondents.
Conclusions: These findings provide strong evidence that bisexual and “not sure” U.S. adults are more likely to have a severe AUD and TUD. They also demonstrate the importance of treatment strategies that address sexual minority-specific risks, particularly for bisexual individuals and those “not sure” of their sexual identity.
Keywords: alcohol use, dimensions of sexual orientation, drug use, DSM-5 disorders, sexual orientation, tobacco use
Introduction
Research indicates that those with a sexual minority identity are at heightened risk for substance use.1–11 Although less attention has been paid to individuals who are unsure of their sexual identity, this group also appears to be at greater risk for negative substance use outcomes than heterosexual individuals.6,12 Indeed, sexual minority individuals are represented disproportionately among the 30 million U.S. adults who meet Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (DSM-5)13 criteria for substance use disorder. Most prevalent are alcohol use disorders (AUDs) and tobacco use disorders (TUDs), followed by other drug use disorders (DUDs).14–17
Hughes et al.6 examined a sample of heterosexual and sexual minority respondents in the National Epidemiologic Survey on Alcohol and Related Conditions (NESARC-II.) They found that odds of substance use disorders were generally higher among sexual minority respondents (n = 747) than among heterosexual respondents (n = 33,598). Using more recent data from the NESARC-III, Slater et al.12 examined substance use among 1351 respondents. More than a quarter (25.9%) of lesbian/gay, 30.2% of bisexual, and 25.3% of “not sure” respondents met criteria for DSM-5 AUD. DSM-5 TUD was seen in 28.9% of lesbian/gay, 37.6% of bisexual, and 31.2% of “not sure” respondents. Fewer respondents were found to have a DSM-5 DUD: 7.4% of lesbian/gay, 11.0% of bisexual, and 10.7% of “not sure” respondents.
Despite important insights into substance use among sexual minority individuals afforded by the aforementioned studies,6,12 there is a notable gap in national probability studies on substance use. None of the studies assessed severity of substance use disorders. The lack of attention to the severity of a substance use disorder obscures important information about impairment, which has implications for substance use treatment.
Studies on sexual orientation and substance use often assess only one dimension of sexual orientation (e.g., identity or behavior), or do not distinguish among the dimensions. Sexual identity is understood as the self-concept around which individuals organize their sexual attraction, often resulting in the self-label of lesbian, gay, bisexual, or heterosexual.18 Although identity and behavior or attraction are often assumed to be overlapping (e.g., that lesbian women only have sex with women), it is important to recognize that these dimensions do not always align. For instance, a man may identify himself as heterosexual, yet may also be sexually attracted to and sexually active with men. Thus, we chose to examine substance use across the three major sexual orientation dimensions.
Although most people indicate their sexual identity on drug and alcohol surveys, some are unsure. One advantage of the 2012–2013 NESARC-III is that in addition to the response options of lesbian/gay, bisexual, or heterosexual, respondents were given the option of indicating that they were unsure of their sexual identity.19 A respondent's choice of “not sure” can mean a number of things, including, but not limited to, not understanding the question, not feeling that they fit into the categories provided (for example, the absence of queer or other labels as response options), or actually being unsure of one's sexual identity.
Objective
We focused on DSM-5 severity of impairment because it allows us to target the health consequences of AUD, TUD, and DUD. We based our definition of severity on the 11 items that comprise the four impairment domains in the DSM-5. Drawing from current literature, we expected that bisexual and “not sure” respondents would have higher prevalence of severe AUD, TUD, and DUD than heterosexuals or lesbian/gay individuals.
Methods
There are few well-designed studies that include large enough subsamples of sexual minority inidviduals—including those “not sure” of their sexual identities—to assess the severity of substance use disorders meaningfully. One exception is the NESARC-III nationally representative household survey sponsored by the National Institutes of Health.14,15
Our study used NESARC-III data collected from April 2012 through June 2013. The primary source of information was the general U.S. civilian noninstitutionalized population of individuals 18 years of age and older. Data were collected through in-person interviews. Response rates were 72% (by household), 84% (by person), and 60% (overall). Respondents (N = 36,309) were given $90.00 for their participation in the survey. The NESARC-III sample design, response rates, and weighting procedures have been described elsewhere.20,21 All NESARC-III procedures received full human subjects review and institutional review board (IRB) approval before data collection. The IRB of the first author's institution determined that this secondary analytic study with deidentified data is exempt.
The NESARC-III includes reliable and validated measures that align with the DSM-513 criteria for substance use disorders and allows a determination of AUD, TUD, and DUD severity among lesbian, gay, and bisexual respondents according to all three major dimensions of sexual orientation (identity, attraction, and behavior).22 It uses the National Institute on Alcohol Abuse and Alcoholism's “Alcohol Use Disorder and Associated Disabilities Interview Schedule-5” (AUDADIS-5), a fully structured diagnostic interview. Depending on the substance, 38 to 40 items are included in the AUDADIS-5 that map onto the 11 DSM-5 symptom criteria. Those 11 criteria make up the four domains reflecting impairment: (1) impaired control, (2) social impairment, (3) risky use, and (4) tolerance/withdrawal. According to the DSM-5 and regardless of type of impairment, a mild disorder is indicated by endorsement of 2–3 criteria, moderate disorder is indicated by endorsement of 4–5 criteria, and severe disorder is indicated by endorsement of 6–11 criteria.
Past-year AUD, TUD, and DUD and severity level
For past-year AUD and TUD: (1) “any disorder” was defined as endorsing 2 or more of the 11 DSM-5 symptom criteria specific to alcohol or tobacco use and (2) “severe disorder” was defined as endorsing 6 or more of the 11 DSM-5 symptom criteria. DUD diagnosis was similar to that for AUD and TUD except that multiple drugs were included, and a diagnosis required endorsement of at least two symptoms for the same drug class (i.e., sedative/tranquilizer, cannabis, amphetamine, cocaine, nonheroin opioid, heroin, hallucinogen, club drugs, and solvents/inhalants). Test–retest reliability for DSM-5 AUD, TUD, and DUD diagnoses was fair and dimensional criteria scales were fair to excellent.15,23,24
Past-year substance use impairment based on the DSM-5
Eleven symptom criteria were mapped onto four domains of impairment and were assigned to only one domain per criterion. The DSM-5 domain assignments were (1) Impaired Control (four symptoms such as “had times when you ended up using a substance more, or longer, than you intended”); (2) Social Impairment (three symptoms such as “continued to use a substance even though it was causing trouble with family and friends”); (3) Risky Use (two symptoms such as “continued to use a substance even though it was adding to a health problem”); and (4) Tolerance/Withdrawal (two symptoms such as “had to use a substance more in order to get the desired effect”). If a respondent endorsed one symptom within an impairment domain, they were determined to have endorsed the domain.
Reliability and validity of the DSM-5-based diagnoses of substance use disorder have been established in prior psychometric studies.23,24
Sexual minority and sexual orientation
The variable “sexual minority” is a composite using the three dimensions of sexual orientation: identity (i.e., gay or lesbian, bisexual, or not sure), attraction (i.e., males and females “equally attracted to females and males”; males who were “only” or “mostly” attracted to males; females who were “only” or “mostly” attracted to females), and behavior (i.e., males and females who had sex with “both males and females”; males who had sex with “only males”; females who had sex with “only females”). Sexual identity was assessed by asking, “Which of the categories on the card best describes you: (1) heterosexual, (2) gay or lesbian, (3) bisexual, or (4) not sure?” Sexual attraction was assessed by asking, “People are different in their sexual attraction to other people. Which category on the card best describes your feelings: (1) only attracted to females, (2) mostly attracted to females, (3) equally attracted to females and males, (4) mostly attracted to males, or (5) only attracted to males?” Sexual attraction was recoded to reflect three attraction-based groups: (1) predominantly attracted to the opposite sex, (2) predominantly attracted to the same sex, and (3) attracted equally to opposite and same sex. Sexual behavior was assessed by asking all respondents whether they had ever had sex (lifetime). If they responded affirmatively to the question about lifetime sex, respondents were asked, “During the last 12 months, did you have sex with (1) only males, (2) only females, or (3) both males and females?” Sexual behavior in the past 12 months was recoded to reflect four behavior-based groups: same-sex behavior only, opposite-sex behavior only, same- and opposite-sex behavior, and no sexual behavior.
Sociodemographic characteristics
Sociodemographic characteristics included age (18–34, 35–54, and 55 years and older); sex (male and female); and race/ethnicity (White, African American, Hispanic, and other). These sociodemographic characteristics served as control variables in the analytic models outlined hereunder.
Data analyses
We examined the overall prevalence of any past-year AUDs, TUDs, and DUDs (e.g., two or more symptoms), severe substance use disorders, and the four impairment domains among heterosexual and sexual minority respondents. Second, we assessed differences in the AUD, TUD, and DUD outcomes separately based on identity, attraction, and behavior, controlling for age, sex, and race/ethnicity. We used binary logistic regression to assess the unique association within the three dimensions of sexual orientation. Third, we assessed differences in AUD, TUD, and DUD, incorporating each sexual orientation dimension simultaneously (controlling for age, sex, and race/ethnicity) to determine which dimension had the strongest overall associations. We do not present the results of the multiple logistic regression analyses assessing the four domains of substance use disorders due to notable similarities with the analytic models assessing past-year any and severe AUD, TUD, and DUD.
We used STATA 15.0 (Version 15.0; StataCorp LP, College Station, TX) to estimate the models outlined above. We controlled for potentially confounding factors (i.e., age, sex, and race/ethnicity) and report adjusted odds ratios (AORs) and 95% confidence intervals (CIs). Because the NESARC-III design included stratification and clustering of the target population, analytic techniques were design based using sampling weights to calculate estimates of population parameters and specialized variance estimation techniques to accommodate the complex design features of the sample when estimating standard errors. All estimates provided in the tables used the sampling weights and accounted for the complex sampling design. However, unweighted sample sizes are provided to show the actual number of respondents within each subpopulation included in the analyses. Because preliminary analyses showed no statistically significant differences between sexual minority men and women with respect to substance use disorders and to preserve power, we chose to conduct the analyses by controlling for rather than stratifying by sex. We used listwise deletion to handle the missing data given the small overall percentage of respondents (2.6%) who would be excluded based on missing data.
Results
Approximately 6.4% (n = 2335) of the sample was categorized as sexual minority. Table 1 summarizes the distribution of the three sexual orientation dimensions (identity, attraction, and behavior) and Table 2 summarizes the four domains of impairment (impaired control, social impairment, risky use, and tolerance/withdrawal), and severity of AUD, TUD, and DUD. In this study, all three dimensions of sexual orientation were separately predictive of severity of AUD, TUD, and DUD; however, when incorporated into the analytic models simultaneously, identity was the most strongly predictive of severity.
Table 1.
Sample Characteristics from the National Epidemiologic Survey on Alcohol and Related Conditions-III, United States, 2012–2013
| Sociodemographics | Total (N = 36,309) n (%) |
|---|---|
| Sexual identity | |
| Heterosexual | 34,644 (96.7) |
| Gay or lesbian | 586 (1.5) |
| Bisexual | 566 (1.3) |
| Not sure | 199 (0.5) |
| Sexual attraction | |
| Opposite-sex attraction | 34,114 (95.3) |
| Same-sex attraction | 1278 (3.3) |
| Opposite- and same-sex attraction | 582 (1.4) |
| Sexual behavior (past year) | |
| Opposite-sex behavior | 24,525 (71.4) |
| Same-sex behavior | 971 (2.5) |
| Opposite- and same-sex behavior | 227 (0.5) |
| No sex (past-year/lifetime) | 9812 (25.6) |
| Sex | |
| Male | 15,862 (48.1) |
| Female | 20,447 (51.9) |
| Race | |
| White | 19,194 (66.2) |
| African American | 7766 (11.8) |
| Hispanic | 7037 (14.7) |
| Other race | 2312 (7.3) |
| Age | |
| 18 to 34 | 11,755 (30.3) |
| 35 to 54 | 13,150 (35.7) |
| 55 and older | 11,404 (34.0) |
Sample sizes vary due to missing data. Unweighted sample sizes are provided.
Percentages incorporate survey weights provided by the NESARC-III.
Table 2.
Substance Use Outcomes Among Heterosexual and Sexual Minority Respondents
| Substance use outcomes | Total (N = 36,309) n (%) | Heterosexuala(n = 33,708)a, n (%) | Sexual minoritya(n = 2335)a, n (%) |
|---|---|---|---|
| AUD | |||
| Any AUD (mild, moderate, or severe) | 4957 (13.4) | 4397 (12.8) | 525 (21.5) |
| Severe AUD | 1232 (3.1) | 1044 (2.9) | 179 (7.2) |
| Impaired control | 6524 (17.7) | 5847 (17.2) | 631 (26.4) |
| Social impairment | 1972 (5.2) | 1713 (4.9) | 243 (9.6) |
| Risky use | 3893 (10.9) | 3444 (10.5) | 416 (17.1) |
| Tolerance/withdrawal | 3684 (9.7) | 3260 (9.3) | 396 (15.7) |
| TUD | |||
| Any TUD (mild, moderate, or severe) | 6260 (17.4) | 5664 (17.0) | 553 (23.6) |
| Severe TUD | 1749 (4.9) | 1546 (4.6) | 191 (8.1) |
| Impaired control | 7721 (21.1) | 7002 (20.7) | 661 (27.9) |
| Social impairment | 2371 (6.6) | 2135 (6.5) | 219 (9.2) |
| Risky use | 5180 (14.3) | 4710 (14.1) | 439 (18.4) |
| Tolerance/withdrawal | 2673 (7.4) | 2382 (7.1) | 273 (11.6) |
| DUD | |||
| Any DUD (mild, moderate, or severe) | 1547 (4.0) | 1354 (3.8) | 186 (7.7) |
| Severe DUD | 419 (1.1) | 363 (1.1) | 54 (2.0) |
| Impaired control | 1731 (4.4) | 1507 (4.1) | 215 (9.0) |
| Social impairment | 647 (1.7) | 560 (1.6) | 84 (3.0) |
| Risky use | 1328 (3.6) | 1173 (3.4) | 152 (6.6) |
| Tolerance/withdrawal | 1435 (3.8) | 1243 (3.5) | 185 (8.2) |
Unweighted sample sizes are provided. Percentages incorporate survey weights provided by the NESARC-III.
The bivariate analysis assesses the overall prevalence of past-year substance use disorders (e.g., two or more symptoms), severe substance use disorders, and impairment domains among heterosexual and sexual minority respondents based on a composite measure using the three dimensions of sexual orientation: identity (i.e., gay or lesbian, bisexual, or not sure), attraction (i.e., males and females “equally attracted to females and males”; males who were “only” or “mostly” attracted to males; females who were “only” or “mostly” attracted to females), and behavior (i.e., males and females who had sex with “both males and females”; males who had sex with “only males”; females who had sex with “only females”). Differences (using binary logistic regression) between heterosexual and sexual minority respondents were all significant at the 0.001 alpha level (controlling for sex, race/ethnicity, and age).
AUD, alcohol use disorder; DUD, drug use disorder; TUD, tobacco use disorder.
As shown in Table 2, not only was severity of AUD, TUD, and DUD higher among sexual minority respondents than among heterosexual respondents, but sexual minority respondents were also significantly more likely to meet criteria on the four impairment domains for alcohol, tobacco, and drugs. The most prevalent impairments for anyone with AUD and TUD were “impaired control” and “risky use,” but this was not true for those with DUD; “impaired control” and “tolerance/withdrawal” were the two most prevalent impairments for respondents with DUD.
We also examined sexual attraction, when controlling for sexual identity, sexual behavior, and other sociodemographic variables. We found that respondents who endorsed same-sex attraction had lower odds of any past-year TUD than opposite-sex attracted respondents (Table 3). Those who indicated both opposite- and same-sex behavior (i.e., bisexual behavior) compared with those who reported either exclusively heterosexual or exclusively same-sex behavior had higher odds of reporting any past-year AUD or TUD. Indeed, respondents who indicated both opposite- and same-sex behavior had almost two times the odds of any past-year AUD and TUD as respondents who reported either exclusively heterosexual or exclusively same-sex behavior. Notably, those who reported “no sex” had significantly lower odds of AUD than any other behavioral group (i.e., only opposite-sex behavior, only same-sex behavior, and both opposite- and same-sex behavior). Moreover, those individuals who reported “no sex” had lower odds for TUD and DUD than individuals who reported only opposite-sex behaviors, or who reported opposite- and same-sex behaviors.
Table 3.
Any Alcohol Use Disorder, Tobacco Use Disorder, and Drug Use Disorder by Dimensions of Sexual Orientation
| Any AUD (mild, moderate, or severe) n = 35,409a | Any TUD (mild, moderate, or severe) n = 35,356a | Any DUD (mild, moderate, or severe) n = 35,415a | ||||
|---|---|---|---|---|---|---|
| %b | AOR (95% CI) Model 1c | %b | AOR (95% CI) Model 2c | %b | AOR (95% CI) Model 3c | |
| Identity | ||||||
| Heterosexual | 12.9def | Referencedef | 17.0def | Referencedef | 3.8def | Referencef |
| Lesbian/gay | 25.4g | 2.03 (1.49–2.77)g | 24.6eg | 1.80 (1.27–2.56)g | 7.8g | 1.62 (0.880–2.99) |
| Bisexual | 29.9g | 1.94 (1.34–2.79)g | 33.4dg | 1.72 (1.22–2.43)g | 11.5g | 1.69 (0.932–3.07) |
| Not sure | 24.7g | 2.68 (1.62–4.40)g | 29.3g | 2.41 (1.59–3.65)g | 10.3g | 2.81 (1.50–5.25)g |
| Attraction | ||||||
| Opposite sex | 13.1de | Reference | 17.2e | Referenced | 3.9de | Reference |
| Same sex | 17.8eg | 0.857 (0.595–1.23) | 18.0e | 0.665 (0.497–0.889)eg | 5.8eg | 1.09 (0.625–1.92) |
| Opposite and same sex | 22.4dg | 0.975 (0.679–1.40) | 28.8dg | 1.21 (0.874–1.69)d | 9.7dg | 1.35 (0.819–2.24) |
| Behavior | ||||||
| Opposite sex | 15.6def | Referenceef | 18.8ef | Referenceef | 4.5def | Referencef |
| Same sex | 21.9efg | 1.08 (0.794–1.49)ef | 21.1ef | 1.12 (0.811–1.55)e | 6.7efg | 0.949 (0.586–1.53)e |
| Opposite and same sex | 43.2dfg | 2.00 (1.29–3.10)dfg | 44.2dfg | 1.98 (1.29–3.06)dfg | 17.9dfg | 1.86 (0.989–3.50)df |
| No sex | 5.9deg | 0.520 (0.460–0.588)deg | 12.9deg | 0.843 (0.773–0.918)eg | 2.1deg | 0.703 (0.562–0.879)eg |
Any disorder (i.e., 2+ symptoms).
Significant differences in reference to heterosexual (identity), opposite-sex only (attraction), and opposite-sex only (behavior) respondents are in bold for the full analytical models.
Sample sizes vary due to missing data. Unweighted sample sizes are provided.
The percentages provide the differences with respect to the bivariate associations (controlling for sex, race/ethnicity, and age).
Models 1 through 3 incorporate sexual identity, attraction, and behavior simultaneously to estimate AORs and 95% CIs (including controls for sex, race/ethnicity, and age); the AORs presented use (identity), opposite-sex only (attraction), and opposite-sex only (behavior) as the reference category.
p < 0.05 using lesbian/gay respondents as the reference group (identity); same-sex respondents as the reference group (attraction); same-sex respondents as the reference group (behavior).
p < 0.05 using bisexual respondents as the reference group (identity); opposite- and same-sex respondents as the reference group (attraction); opposite- and same-sex respondents as the reference group (behavior).
p < 0.05 using not sure as the reference group (identity); no sex respondents as the reference group (behavior).
p < 0.05 using heterosexual respondents as the reference group (identity); opposite-sex respondents as the reference group (attraction); opposite-sex respondents as the reference group (behavior).
AOR, adjusted odds ratio; CI, confidence interval.
Table 4 shows associations between the three dimensions of sexual orientation and past-year severe AUD, TUD, and DUD.
Table 4.
Severe Alcohol Use Disorder, Tobacco Use Disorder, and Drug Use Disorder by Dimensions of Sexual Orientation
| Severe AUD n = 35,409a | Severe TUD n = 35,356a | Severe DUD n = 35,415a | ||||
|---|---|---|---|---|---|---|
| %b | AOR (95% CI) Model 4c | %b | AOR (95% CI) Model 5c | %b | AOR (95% CI) Model 6c | |
| Identity | ||||||
| Heterosexual | 2.9def | Referencedef | 4.7ef | Referencedef | 1.1ef | Referencef |
| Lesbian/gay | 7.1fg | 2.36 (1.17–4.74)fg | 7.5ef | 2.22 (1.19–4.13)g | 1.4f | 0.704 (0.196–2.52)f |
| Bisexual | 11.4g | 3.08 (1.44–6.58)g | 13.3dg | 2.51 (1.29–4.88)g | 3.5g | 1.55 (0.538–4.49)f |
| Not sure | 11.5dg | 5.05 (2.78–9.16)dg | 14.4dg | 4.18 (2.29–7.64)g | 4.8dg | 4.40 (1.72–11.2)deg |
| Attraction | ||||||
| Opposite sex | 2.9de | Reference | 4.8e | Reference | 1.1e | Reference |
| Same sex | 4.6eg | 0.589 (0.269–1.29) | 5.1e | 0.625 (0.382–1.02) | 1.6 | 0.951 (0.320–2.82) |
| Opposite and same sex | 7.7dg | 0.777 (0.443–1.36) | 10.6dg | 0.926 (0.523–1.64) | 2.9g | 1.25 (0.393–4.01) |
| Behavior | ||||||
| Opposite sex | 3.5def | Referenceef | 5.2e | Referencee | 1.3ef | Referencef |
| Same sex | 6.3efg | 1.53 (0.866–2.70)f | 5.9e | 1.00 (0.666–1.52)e | 2.2f | 1.66 (0.752–3.69)f |
| Opposite and same sex | 18.5dfg | 2.44 (1.24–4.79)fg | 19.4dfg | 2.16 (1.19–3.93)dfg | 6.4fg | 2.18 (0.897–5.30)f |
| No sex | 1.5deg | 0.672 (0.538–.840)deg | 3.7e | 0.962 (0.827–1.11)e | 0.4deg | 0.453 (0.277–.741)deg |
Severe disorder (i.e., 6+ symptoms).
Significant differences in reference to heterosexual (identity), opposite-sex only (attraction), and opposite-sex only (behavior) respondents are in bold for the full analytical models.
Sample sizes vary due to missing data. Unweighted sample sizes are provided.
The percentages provide the differences with respect to the bivariate associations (controlling for sex, race/ethnicity, and age).
Models 4 through 6 incorporate sexual identity, attraction, and behavior simultaneously to estimate AORs and 95% CIs (including controls for sex, race/ethnicity, and age); the AORs use heterosexual (identity), opposite-sex only (attraction), and opposite-sex only (behavior) as the reference category.
p < 0.05 using lesbian/gay respondents as the reference group (identity); same-sex respondents as the reference group (attraction); same-sex respondents as the reference group (behavior).
p < 0.05 using bisexual respondents as the reference group (identity); opposite- and same-sex respondents as the reference group (attraction); opposite- and same-sex respondents as the reference group (behavior).
p < 0.05 using not sure as the reference group (identity); no sex respondents as the reference group (behavior).
p < 0.05 using heterosexual respondents as the reference group (identity); opposite-sex respondents as the reference group (attraction); opposite-sex respondents as the reference group (behavior).
Sexual minority respondents, based on sexual identity, had higher odds of severe AUD and TUD than heterosexual respondents. Opposite- and same-sex behaviors were predictive of severe AUD (AOR = 2.44; 95% CI = 1.24–4.79) and TUD (AOR = 2.16; 95% CI = 1.19–3.93), but not DUD. Respondents who were “not sure” of their sexual identity had higher odds for all severe substance use disorders (AUD [AOR = 5.05; 95% CI = 2.78–9.16], TUD [AOR = 4.18; 95% CI = 2.29–7.64], and DUD [AOR = 4.40; 95% CI = 1.72–11.2]) than heterosexual respondents. We found few significant differences between “not sure” and bisexual respondents. No significant differences for severity were seen between men and women when categorized by either sexual attraction or behavior.
Discussion
This study, using a national probability sample, is the first to examine the relationships among the dimensions of sexual orientation and AUD, TUD, and DUD severity. It is also the first study to examine the four DSM-5 domains of impairment across sexual identity, attraction, and behavior. As such, this study broadens our understanding of the severity and types of impairment experienced by sexual minority individuals.
Our results demonstrated that U.S. adults who identified as lesbian/gay and bisexual had greater odds of reporting AUD and TUD, but not DUD, than heterosexual respondents. This finding is consistent with that of Slater et al.12 Only respondents who identified as “not sure” had greater odds of any past-year DUD than heterosexual respondents. Furthermore, we found no statistically significant differences among the three dimensions of sexual orientation between men and women with respect to AUD, TUD, and DUD (data not shown, available on request).
We found differences among sexual minority respondents. Those who indicated both opposite- and same-sex behavior were found to have higher odds of reporting any past-year AUD and TUD than respondents with same-sex behavior only. In their review of the literature on substance use among lesbian, gay, and bisexual individuals, Green and Feinstein4 found that bisexual identity or behavior was related to increased risk for substance use. Other studies have found that bisexual women and men were more likely to report alcohol and drug use than both lesbian/gay and heterosexual identified respondents.9,25,26 Moreover, our study showed that not only was severity of AUD, TUD, and DUD higher for sexual minority respondents than for heterosexual respondents, indicators of impairment were also more prevalent.
We found some variation in the odds of severe AUD, TUD, and DUD across the three dimensions of sexual orientation. For instance, respondents who indicated that they were “not sure” of their sexual identity had significantly greater odds than lesbian/gay and bisexual respondents of meeting criteria for severe DUD. Bostwick and Hequembourg27 suggest that the higher rates of substance use among bisexual individuals may be related to biphobia or the stigma of not finding an identifiable community. This may also be true for those who are “unsure” of their identity.
Few researchers to date have included the response option “not sure” in their studies or they have excluded this group from their analyses. It is possible that individuals who are unsure of their sexual identity also face minority stress in the form of stigma and marginalization. It is also possible that individuals who chose “not sure” to describe their sexual identity may be grappling with their sexual minority identity and may use alcohol or other drugs to cope. It is important to understand whether respondents who indicate that they are unsure of their sexual identity either lack cognitive understanding of the question or are providing a more socially desirable response than lesbian/gay or bisexual. Additional research is needed to better understand the meaning of “not sure,” the characteristics of respondents who choose this response, and how being “not sure” is linked to substance use. Those who are either “unsure” or discordant in their orientation (e.g., their identity does not match their attraction or behavior) may experience social stress, which is, in turn, related to substance use.
Limitations
Despite numerous strengths of NESARC-III, there are some limitations that should be considered. The data are cross-sectional and only from U.S. respondents, limiting international generalizability and understanding of causality. Institutionalized respondents (e.g., incarcerated) were not included in the NESARC-III sample and this likely leads to underestimation of substance use. Although NESARC-III includes a large U.S. sample of sexual minority respondents, the sample size was not large enough to permit examination of multivariate relationships among the minority subgroups separately for men and women with respect to severe substance use disorders. This is an important limitation given findings that show substantial differences in risk for sexual minority men and women. For example, in contrast to findings from the general population, sexual minority women appear to be at substantially higher risk than sexual minority men for AUDs.28 Finally, NESARC-III did not assess sex assignment at birth nor gender identity. Sex (male or female) was assessed as binary. Questions regarding transgender identity, a group at particularly high risk for substance use disorders, were not included in NESARC-III. The exclusion of questions about transgender identity is a notable limitation. Future studies should endeavor to include the two-step question recommended by the Williams Institute that first asks about sex assigned at birth and then about current identity.29
Conclusion
Our findings provide strong evidence that a higher proportion of sexual minority individuals, particularly bisexual individuals and individuals who are “not sure” of their sexual identities, have severe AUDs and TUDs, and those “not sure” also have severe DUDs. Higher rates of severe substance use disorders among sexual minority individuals pose unique challenges for the treatment community and signify substantial health disparities. Our findings demonstrate the importance of designing treatment programs that target sexual minority individuals and their unique risks.
Acknowledgments
The development of this study was supported by grants DA043696, DA036541, CA212517, and AA025684 from the National Institutes of Health (NIH). This article used limited access data obtained from the National Institute on Alcohol Abuse and Alcoholism (NIAAA). The funders had no role in the design and conduct of the study nor the interpretation of the data. The authors take responsibility for the secondary analysis and interpretation of the data and for preparing the article.
Disclaimer
The content is the responsibility of the authors and does not necessarily represent the official views of the NIH, the National Cancer Institute, the NIAAA, the National Institute on Drug Abuse, or the U.S. Government.
Author Disclosure Statement
No competing financial interests exist.
References
- 1. Cochran SD, Mays VM: Estimating prevalence of mental and substance-use disorders among lesbians and gay men from existing national health data. In: Sexual Orientation and Mental Health: Examining Identity and Development in Lesbian, Gay, and Bisexual People. Edited by Omoto AM, Kurtzman HS. Washington, DC: American Psychological Association, 2006, pp 143–165 [Google Scholar]
- 2. Drabble L, Midanik LT, Trocki K: Reports of alcohol consumption and alcohol-related problems among homosexual, bisexual and heterosexual respondents: Results from the 2000 National Alcohol Survey. J Stud Alcohol 2005;66:111–120 [DOI] [PubMed] [Google Scholar]
- 3. Drabble L, Trocki KF, Hughes TL, et al. : Sexual orientation differences in the relationship between victimization and hazardous drinking among women in the National Alcohol Survey. Psychol Addict Behav 2013;27:639–648 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Green KE, Feinstein BA: Substance use in lesbian, gay, and bisexual populations: An update on empirical research and implications for treatment. Psychol Addict Behav 2012;26:265–278 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Hughes TL, Eliason M: Substance use and abuse in lesbian, gay, bisexual and transgender populations. J Prim Prev 2002;22:263–298 [Google Scholar]
- 6. Hughes T, McCabe SE, Wilsnack SC, et al. : Victimization and substance use disorders in a national sample of heterosexual and sexual minority women and men. Addiction 2010;105:2130–2140 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Hughes TL, Szalacha LA, Johnson TP, et al. : Sexual victimization and hazardous drinking among heterosexual and sexual minority women. Addict Behav 2010;35:1152–1156 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Institute of Medicine (US) Committee on Lesbian, Gay Bisexual, and Transgender Health Issues and Research Gaps and Opportunities: The Health of Lesbian, Gay, Bisexual, and Transgender People: Building a Foundation for Better Understanding. Washington, DC: National Academies Press, 2011 [PubMed] [Google Scholar]
- 9. McCabe SE, Hughes TL, Bostwick W, Boyd CJ: Assessment of difference in dimensions of sexual orientation: Implications for substance use research in a college-age population. J Stud Alcohol 2005;66:620–629 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. McCabe SE, Hughes TL, Bostwick WB, et al. : Sexual orientation, substance use behaviors and substance dependence in the United States. Addiction 2009;104:1333–1345 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. McCabe SE, Hughes TL, Boyd CJ: Substance use and misuse: Are bisexual women at greater risk? J Psychoactive Drugs 2004;36:217–225 [DOI] [PubMed] [Google Scholar]
- 12. Slater ME, Godette D, Huang B, et al. : Sexual orientation-based discrimination, excessive alcohol use, and substance use disorders among sexual minority adults. LGBT Health 2017;4:337–344 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. American Psychiatric Association: Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition. Arlington, VA: American Psychiatric Association, 2013 [Google Scholar]
- 14. Grant BF, Goldstein RB, Saha TD, et al. : Epidemiology of DSM-5 alcohol use disorder: Results from the National Epidemiologic Survey on Alcohol and Related Conditions III. JAMA Psychiatry 2015;72:757–766 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Grant BF, Saha TD, Ruan WJ, et al. : Epidemiology of DSM-5 drug use disorder: Results from the National Epidemiologic Survey on Alcohol and Related Conditions-III. JAMA Psychiatry 2016;73:39–47 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Thiede H, Valleroy LA, MacKellar DA, et al. : Regional patterns and correlates of substance use among young men who have sex with men in 7 US urban areas. Am J Public Health 2003;93:1915–1921 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Bostwick WB, Boyd CJ, Hughes TL, McCabe SE: Dimensions of sexual orientation and the prevalence of mood and anxiety disorders in the United States. Am J Public Health 2010;100:468–475 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Diamond LM: Sexual identity, attractions, and behavior among young sexual-minority women over a 2-year period. Dev Psychol 2000;36:241–250 [DOI] [PubMed] [Google Scholar]
- 19. McCabe SE, Hughes TL, Matthews AK, et al. : Sexual orientation discrimination and tobacco use disparities in the United States. Nicotine Tob Res 2017. [Epub ahead of print]; DOI: 10.1093/ntr/ntx283 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Kerridge BT, Pickering RP, Saha TD, et al. : Prevalence, sociodemographic correlates and DSM-5 substance use disorders and other psychiatric disorders among sexual minorities in the United States. Drug Alcohol Depend 2017;170:82–92 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Grant BF, Chu A, Sigman R, et al. : Source and Accuracy Statement for the National Epidemiologic Survey on Alcohol and Related Conditions-III (NESARC-III). Rockville, MD: National Institute on Alcohol Abuse and Alcoholism, 2015 [Google Scholar]
- 22. Ruan WJ, Goldstein RB, Chou SP, et al. : The Alcohol Use Disorder and Associated Disabilities Interview Schedule-IV (AUDADIS-IV): Reliability of new psychiatric diagnostic modules and risk factors in a general population sample. Drug Alcohol Depend 2008;92:27–36 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Grant BF, Goldstein RB, Smith SM, et al. : The Alcohol Use Disorder and Associated Disabilities Interview Schedule-5 (AUDADIS-5): Reliability of substance use and psychiatric disorder modules in a general population sample. Drug Alcohol Depend 2015;148:27–33 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Hasin DS, Greenstein E, Aivadyan C, et al. : The Alcohol Use Disorder and Associated Disabilities Interview Schedule-5 (AUDADIS-5): Procedural validity of substance use disorders modules through clinical re-appraisal in a general population sample. Drug Alcohol Depend 2015;148:40–46 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Tucker JS, Ellickson PL, Klein DJ: Understanding differences in substance use among bisexual and heterosexual young women. Womens Health Issues 2008;18:387–398 [DOI] [PubMed] [Google Scholar]
- 26. Wilsnack SC, Hughes TL, Johnson TP, et al. : Drinking and drinking-related problems among heterosexual and sexual minority women. J Stud Alcohol Drugs 2008;69:129–139 [DOI] [PubMed] [Google Scholar]
- 27. Bostwick W, Hequembourg A: ‘Just a little hint’: Bisexual-specific microaggressions and their connection to epistemic injustices. Cult Health Sex 2014;16:488–503 [DOI] [PubMed] [Google Scholar]
- 28. Hughes TL, Wilsnack SC, Kantor LW: The influence of gender and sexual orientation on alcohol use and alcohol-related problems: Toward a global perspective. Alcohol Res 2016;38:121–132 [PMC free article] [PubMed] [Google Scholar]
- 29. The GenIUSS Group: Best Practices for Asking Questions to Identify Transgender and Other Gender Minority Respondents on Population-Based Surveys. Edited by Herman JL. Los Angeles, CA: The Williams Institute, 2014 [Google Scholar]