Associations Between Having a Spouse/Partner with Alcohol Problems and One’s Own Risk of Mental Health and Substance Use Disorders by Sexual Orientation

Abstract

Introduction:

The association between an individual’s alcohol misuse and their spouse’s/partner’s behavioral health is well documented. Whether these associations vary by sexual orientation is unknown.

Methods:

We used data from the 2012–2013 National Epidemiologic Survey on Alcohol and Related Conditions (NESARC-III; n=36,309 U.S. adults ≥18 years). Using multiple logistic regression models, we examined the associations between sexual orientation (heterosexual-identified with concordant attraction/behavior, heterosexual-identified with discordant attraction/behavior, gay/lesbian, and bisexual) and having a lifetime history of spouse/partner with alcohol problems with lifetime DSM-5 alcohol use disorder, tobacco use disorder, any substance use disorder (SUD), anxiety disorder, mood disorder, and comorbid mental health and SUDs. We conducted two-way interaction tests between history of spouse/partner with alcohol problems and sexual orientation for each outcome.

Results:

The prevalence of a history of spouse/partner with problem drinking varied by sex and sexual orientation, ranging from 28.8% (bisexual females) to 9.1% (bisexual males). Having a history of a spouse/partner with alcohol problems was associated with increased odds of each outcome (adjusted odds ratio [AOR] range: 1.6 to 2.3, p<0.001). All sexual minority subgroups had increased odds of each outcome (AOR range: 1.3 to 3.2, p<0.001), relative to their heterosexual concordant counterparts. Our interaction tests were not significant at the p<0.001 level; the effect of having a spouse/partner with alcohol problems is similar across sexual orientation subgroups.

Conclusions:

Findings underscore the importance of sexual orientation and having a history of a spouse/partner with alcohol problems as part of the assessment and treatment of mental health and SUDs.

In the U.S., approximately one in six adults engage in binge drinking (i.e., ≥4 drinks for females and ≥5 drinks for males, in the same occasion; Azagba et al., 2020; National Institute on Alcohol Abuse and Alcoholism, 2004) and almost three in ten adults meet the criteria for lifetime prevalence of Diagnostic and Statistical Manual of Mental Disorder, Fifth Edition (DSM-5) alcohol use disorder (AUD; Grant, Goldstein, Saha, et al., 2015). AUD and alcohol misuse, including binge drinking and high intensity drinking (i.e., twice or more the sex-specific threshold for binge drinking [≥8/≥10 drinks for females/males on the same occasion, respectively; Linden-Carmichael et al., 2017; Patrick et al., 2016]), carry significant social and economic burden, such as high healthcare and criminal justice costs and decreased productivity (Fergusson et al., 2013; GBD 2016 Alcohol Collaborators, 2018; Karriker-Jaffe et al., 2018; Rehm et al., 2009, 2017; Rehm & Hingson, 2014). Alcohol use is also a leading cause of preventable mortality in the U.S., with approximately 95,000 deaths annually attributed to alcohol-related causes (White et al., 2022).

Sexual minority individuals (e.g., those who identify as lesbian, gay, or bisexual) are at greater risk of alcohol misuse, AUD, and severe AUD, relative to heterosexual individuals (Boyd et al., 2019; Evans-Polce et al., 2023; Hughes et al., 2020; McCabe et al., 2009, 2019). This increased risk is, at least in part, related to the discrimination and excess stress they experience associated with being members of a stigmatized group based on their sexual minority identity (McCabe et al., 2019; Meyer, 2003). Among sexual minority couples, these unique stressors can also negatively influence their relationship functioning, dynamics, and well-being (Cao et al., 2017). Within an environment already strained by stress, sexual minority relationships can be further complicated by having a spouse/partner who experiences a drinking problem (Cao et al., 2017).

Being married or in a committed romantic relationship is generally a protective factor against alcohol misuse and AUD (Evans-Polce et al., 2020; Grant, Goldstein, Saha, et al., 2015, p. 201; Kahle et al., 2020; Staff et al., 2014; Veldhuis et al., 2019). Most individuals select romantic partners with drinking or nondrinking levels similar to theirs, with the dyad often mutually reinforcing each other’s alcohol use behaviors (Bartel et al., 2017; Howe et al., 2019; Low et al., 2007; Wiersma-Mosley et al., 2016). Among spouses/partners, compatibility in drinking quantity and frequency may serve as a positive relational bond, while discrepant drinking patterns can become a stressor, which in turn, can result in lower relationship satisfaction and increased relationship dissolution (Fischer & Wiersma-Mosley, 2012; Homish & Leonard, 2007; Leonard & Eiden, 2007). Similarity in drinking patterns may also attract couples and reinforce their drinking behaviors resulting in positive relationship characteristics (e.g., intimacy and sexual enhancement) relative to discrepant drinking patterns (Fischer & Wiersma-Mosley, 2012).

While similar drinking patterns may have a positive impact on a couple’s relationship, alcohol misuse and AUD often have detrimental effects. For example, with respect to negative social consequences, research finds that alcohol misuse is associated with relationship breakdown, interpersonal violence, and risky sexual behavior (Boden et al., 2013; Fergusson et al., 2013; Fischer & Wiersma-Mosley, 2012; Karriker-Jaffe et al., 2018, p.). Additionally, being in a romantic relationship where only one person has a drinking problem is found to be associated with poorer physical and psychiatric outcomes for the other person, including increased risk of bodily injury from intimate partner violence (Caetano et al., 2005; Foran & O’Leary, 2008), mood and anxiety disorders, poorer general health, and poorer quality of life (Dawson et al., 2007). A systematic review on alcohol use and romantic relationships found that the vast majority of studies focus on heterosexual couples, suggesting the need for research that focuses on sexual minority relationships (Fischer & Wiersma-Mosley, 2012). Further, while prior work has demonstrated that the partner of an individual who misuses substances is vulnerable to psychosocial problems, especially if the partner identifies as a woman (Di Sarno et al., 2021; Maghsoudi et al., 2019), it is not yet known if or how these risks vary by sexual orientation among U.S. adults.

The National Academies of Sciences, Engineering, and Medicine (NASEM; 2020) has identified the need to examine how sexual minority spouses/partners may influence each other’s health and well-being. Our study addresses this research gap by examining the clinical profile of sexual minority individuals who have spouses/partners with a drinking problem. Understanding whether having a history of spouse/partner who experiences alcohol problems may be an important relationship risk factor associated with increased odds of experiencing mental health disorders and substance use disorders (SUDs), and how this association may vary among sexual orientation subgroups, has important implications for assessment and treatment. In addition, this study considers the multiple dimensions of sexual orientation consisting of identity, behavior, and attraction (NASEM, 2022). Prior work found a greater risk for alcohol misuse and AUD among individuals with identity-attraction or identity-behavior discordance (i.e., mismatch between sexual identity and sexual attraction or sexual behavior) versus concordance (Evans-Polce et al., 2023; McCabe et al., 2019, 2024; Talley et al., 2015). Additionally, bisexual-identified individuals bear elevated risk of mental health and SUD relative to other sexual minority subgroups. Understanding the differential risk among sexual minority populations can inform intervention strategies. Thus, we aimed to address the aforementioned gaps in the literature by:

1. Estimating the prevalence of lifetime history of having a spouse/partner who experiences alcohol problems, SUDs, mental health disorders, and comorbid mental health and SUDs, by sexual orientation (i.e., gay/lesbian, bisexual, heterosexual-identified with concordant attraction/behavior [heterosexual concordant], heterosexual-identified with discordant attraction/behavior [heterosexual discordant]) and sex.

2. Examining bivariate and multivariable associations of sexual orientation and lifetime history of having a spouse/partner with alcohol problems with lifetime (a) AUD, (b) tobacco use disorder (TUD), (c) any SUD, (d) anxiety disorder, (e) mood disorder, and (f) comorbid mental health and SUDs (based on DSM-5 criteria).

3. Testing for significant interactions between lifetime history of spouse/partner with alcohol problems and sexual orientation for each outcome to assess whether the associations differ by sexual orientation.

Materials and Methods

We analyzed data from the 2012–2013 National Epidemiologic Survey on Alcohol and Related Conditions (NESARC-III), which was collected using structured diagnostic face-to-face interviews with a large nationally representative sample of 36,309 U.S. adults aged ≥18 years. The NESARC-III included nuanced measures of sexual orientation domains and was also the primary source of information regarding lifetime DSM-5 SUDs (including alcohol, tobacco, and other drug use disorders) and mental disorders (e.g., anxiety and mood disorders) among the general civilian, non-institutionalized population of U.S. adults. The NESARC-III included the National Institute on Alcohol Abuse and Alcoholism’s Alcohol Use Disorder and Associated Disabilities Interview Schedule-5 (AUDADIS-5), which was a fully structured diagnostic interview of mental illness, substance use, and other adverse health conditions conducted with individuals in households. The overall response rate was 60.1%, which was calculated by taking the product of the household response rate (percentage of households that agreed to participate [72%]) and the person-level response rate (percentage of eligible individuals within participating households who agreed to be interviewed [84%]). The NESARC-III sample design, response rates, and weighting procedures are described elsewhere (Grant, Chu, et al., 2015). All NESARC-III study procedures received full human subjects review and Institutional Review Board (IRB) approval; the current study received IRB approval from the last author’s institution. The NESARC-III data is a controlled access dataset. Researchers must apply for access through the National Institute on Alcohol Abuse and Alcoholism (NIAAA) and complete a data use agreement before access can be granted.

Measures

Sexual orientation was assessed based on respondents’ self-reported sexual identity, sexual attraction, and sexual behavior, consistent with prior work (Kcomt et al., 2021; McCabe et al., 2022). Respondents were asked to select the sexual identity category that best described them (heterosexual [straight], gay or lesbian, bisexual, or not sure). Respondents who were unsure of or did not answer the question about their sexual identity (n=199, <0.5%) were excluded from our analyses due to this subgroup’s small sample size. Respondents were asked to describe their feelings of sexual attraction to other people (only attracted to females, mostly attracted to females, equally attracted to females and males, mostly attracted to males, or only attracted to males). Sexual behavior was assessed using two items: “Have you had sex in the last 12 months?” and “During the last 12 months, did you have sex with only males, only females, or both males and females?” We grouped responses into four sexual behavior categories: only same sex, only other sex, both sexes, and did not have sex. We categorized heterosexual-identified individuals who reported same-sex attraction and/or behavior as heterosexual discordant, and heterosexual-identified individuals without same-sex attraction nor behavior as heterosexual concordant. We defined sexual minority individuals as gay/lesbian-, bisexual-, or heterosexual discordant.

Lifetime DSM-5 AUD was assessed using the AUDADIS-5 and DSM-5 diagnostic criteria and was measured as a dichotomous variable (yes/no). AUD diagnosis was based on participant reports of at least two of the 11 DSM-5 criteria (e.g., craving, tolerance, withdrawal) at the same time (American Psychiatric Association, 2013). Reliability and validity of the DSM-5 AUD criteria scale have been examined previously and have demonstrated excellent reliability (ICC=0.9; Grant, Goldstein, Saha, et al., 2015).

Lifetime DSM-5 TUD was assessed similarly to AUD and based on participant reports of at least two of the 11 DSM-5 (e.g., craving, tolerance, withdrawal) criteria at the same time (American Psychiatric Association, 2013). Reliability and validity of the DSM-based diagnosis of TUD have been established in prior psychometric studies (Hasin, Greenstein, et al., 2015).

A diagnosis for any lifetime DSM-5 SUD was assessed as a dichotomous measure (yes/no) and included substance-specific diagnoses for 11 substances: alcohol, tobacco, cannabis, cocaine, heroin, hallucinogens, inhalants, prescription opioids, sedatives/tranquilizers, stimulants, and other drugs (e.g., ecstasy and ketamine). Each substance-specific diagnosis required two or more of the 11 DSM-5 criteria (e.g., craving, tolerance, withdrawal) at the same time (American Psychiatric Association, 2013). Previous research has found that the test-retest reliability and validity of each AUDADIS-5 DSM-5 SUD diagnoses ranged from fair to good, and dimensional criteria scales ranged from fair to excellent (Grant et al., 2016; Hasin, Greenstein, et al., 2015).

Mental disorders were assessed as dichotomous measures (yes/no) using the AUDADIS-5 and DSM-5 diagnostic criteria. Lifetime DSM-5 anxiety disorder included agoraphobia, generalized anxiety disorder, panic, social, and specific phobias. Lifetime DSM-5 mood disorder included bipolar, dysthymia, and major depressive disorders. Consistent with the DSM-5, these diagnoses excluded substance- and illness-induced disorders. Acceptable reliability and validity of the DSM-based AUDADIS-5 diagnoses of anxiety and mood disorders have been established previously (Grant, Goldstein, Smith, et al., 2015; Hasin, Shmulewitz, et al., 2015).

Lifetime comorbid mental health and SUDs (yes/no) was defined as having both a DSM-5 SUD and a mental disorder in the respondents’ lifetime.

Lifetime history of spouse/partner with alcohol problems (yes/no) was based on the respondents’ affirmative response to any of the following items: “Were you ever married to an alcoholic or problem drinker? Did you ever live as if married with someone who was an alcoholic or problem drinker?” The inclusive language in these measures accounted for respondents residing in states where same-sex marriage was not yet legalized.

Potential confounders included parental lifetime history of alcohol problems (yes/no) based on the respondents’ affirmative response to the following items: “Has your blood or natural father been an alcoholic or problem drinker at any time in his life? Has your blood or natural mother been an alcoholic or problem drinker at any time in her life?” Sociodemographic variables included the respondents’ sex (male, female), age (in years), race-ethnicity (White, African American, Hispanic, Asian/Native Hawaiian/Pacific Islander/American Indian/Alaskan Native/Another race or ethnicity), educational attainment (high school degree or less, some college, college degree or higher), urbanicity (urban, rural), and geographic region (Northeast, Midwest, South, West).

Data Analysis

All analyses in this study accounted for the complex sample design features of the NESARC-III (weights, stratum codes, and cluster codes) when computing weighted estimates for the target NESARC-III population and estimating the sampling variances of the population estimates (Heeringa et al., 2017). We began with descriptive analyses, estimating the prevalence of any history of having a spouse/partner with alcohol problems by sexual orientation and sex. We then examined bivariate associations between lifetime history of spouse/partner with alcohol problems and meeting criteria for DSM-5 disorders, stratified by sexual orientation. Tests of associations between the variables were conducted using the design-adjusted Rao-Scott F-test (Rao & Scott, 1984). Given the multiple comparisons in our analyses, p-values less than 0.001 were considered statistically significant (Benjamin et al., 2018). Next, we conducted multiple logistic regression analyses to estimate the odds of mental health and SUDs as a function of having a spouse/partner with alcohol problems and sexual orientation, controlling for sex, age, race-ethnicity, educational attainment, urbanicity, geographic region, and parental lifetime history of alcohol problems. We also tested for significant interactions between sexual orientation and lifetime history of spouse/partner with alcohol problems for each outcome. Each interaction was tested with a design-adjusted Wald F-test. Each model was also tested for goodness-of-fit using a design-adjusted Archer-Lemeshow goodness-of-fit test (Archer & Lemeshow, 2006). We conducted supplemental sex-stratified analyses to examine sex-specific multivariable associations of sexual orientation, lifetime history of spouse/partner with alcohol problems, and their interaction with meeting criteria for DSM-5 mental health and substance use disorders. All analyses were performed using the Stata software (Version 18).

Missing data were present in the lifetime history of spouse/partner with alcohol problems, parental lifetime history of alcohol problems, and sexual identity variables only, and the levels of missingness were <5%. Thus, imputation was not necessary based on the statistical literature (Jakobsen et al., 2017). For a bias check, we compared the levels of missing value for these variables by the DSM-5 variables and sociodemographic variables (sex, age, race, education), and found no systemic patterns of missingness for our variables. Missing responses were handled on a listwise basis. As a result of listwise deletion, 1.8% (n=647) of individuals were excluded from the analysis.

Results

Table 1 displays the estimated distributions on sociodemographic measures for the overall target population and stratified by sexual orientation. An estimated 90.8% were categorized as heterosexual concordant, 6.4% as heterosexual discordant, 1.5% as gay/lesbian, and 1.4% as bisexual. Almost two-thirds of the sample identified as White (66.2%), 14.7% Hispanic, 11.8% African American, and 7.3% Asian, Native Hawaiian or Pacific Islander, American Indian or Alaskan Native, or another race or ethnicity. Approximately 48.1% of the target population were males. The prevalence of having a spouse/partner with alcohol problems ranged from 23.2% among bisexual individuals to 15.6% among heterosexual concordant individuals. The prevalence of each respective outcome was also higher among sexual minority subgroups—particularly bisexual individuals—relative to the heterosexual concordant subgroup.

Table 1:

Estimated Distributions on Sociodemographic Measures for the 2012–2013 National Epidemiologic Survey on Alcohol and Related Conditions (NESARC-III) Target Population (N = 36,309)

	Overall	Heterosexual Concordanta n = 32,158 (90.8%)	Heterosexual Discordantb n = 2,396 (6.4%)	Gay/Lesbian n = 586 (1.5%)	Bisexual n = 566 (1.4%)
Variables	n (%)	n (%)	n (%)	n (%)	n (%)
Sex *
Male	15,862 (48.1%)	14,295 (49.0%)	872 (39.4%)	321 (58.2%)	144 (28.8%)
Female	20,447 (51.9%)	17,863 (51.0%)	1,524 (60.6%)	265 (41.8%)	422 (71.2%)
Age *
18–24	4,496 (13.1%)	3,822 (12.6%)	316 (13.5%)	110 (18.3%)	171 (36.3%)
25–44	13,765 (34.4%)	12,098 (34.1%)	967 (36.2%)	238 (39.0%)	266 (42.6%)
45–64	12,242 (35.0%)	10,967 (35.4%)	759 (32.8%)	200 (34.9%)	106 (17.9%)
≥65	5,806 (17.6%)	5,271 (17.9%)	354 (17.5%)	38 (7.8%)	23 (3.2%)
Race-Ethnicity
White	19,194 (66.2%)	16,994 (66.3%)	1,308 (66.7%)	333 (67.6%)	289 (64.9%)
African American	7,766 (11.8%)	6,852 (11.7%)	495 (11.5%)	116 (13.1%)	150 (15.6%)
Hispanic	7,037 (14.7%)	6,286 (14.8%)	413 (13.3%)	115 (15.4%)	101 (13.8%)
Asian/Native Hawaiian/Pacific Islander/American Indian/Alaskan Native/Another race or ethnicity	2,312 (7.3%)	2,026 (7.2%)	180 (8.5%)	22 (4.0%)	26 (5.7%)
Educational Attainment *
High school degree or less	15,289 (38.8%)	13,693 (39.0%)	922 (36.8%)	160 (28.1%)	230 (40.0%)
Some college	7,982 (21.6%)	6,966 (21.4%)	582 (23.9%)	145 (21.3%)	175 (30.8%)
College degree or higher	13,038 (39.6%)	11,499 (39.7%)	892 (39.3%)	281 (50.6%)	161 (29.2%)
Urbanicity *
Urban	30,193 (78.8%)	26,541 (78.1%)	2,081 (82.9%)	537 (88.6%)	507 (85.5%)
Rural	6,116 (21.3%)	5,617 (21.9%)	315 (17.1%)	49 (11.4%)	59 (14.5%)
Geographic Region *
Northeast	5,180 (18.2%)	4,531 (18.2%)	355 (17.8%)	110 (25.4%)	96 (21.2%)
Midwest	7,566 (21.5%)	6,739 (21.6%)	475 (20.1%)	91 (15.3%)	127 (22.5%)
South	14,532 (37.1%)	13,041 (37.5%)	849 (33.4%)	207 (31.8%)	201 (32.4%)
West	9,031 (23.2%)	7,847 (22.8%)	717 (28.7%)	178 (27.6%)	142 (23.9%)
Parental Lifetime History of Alcohol (yes)*	10,953 (30.0%)	9,480 (29.4%)	824 (33.5%)	229 (38.9%)	229 (39.7%)
Lifetime DSM-5 alcohol use disorder (AUD) (yes)*	10,001 (29.1%)	8,456 (28.1%)	847 (35.3%)	269 (45.9%)	281 (52.0%)
Lifetime DSM-5 tobacco use disorder (TUD) (yes)*	9,747 (27.9%)	8,343 (27.1%)	772 (32.6%)	223 (39.2%)	254 (45.0%)
Any lifetime DSM-5 substance use disorder (SUD) (yes)*	15,104 (43.1%)	12,909 (41.9%)	1,221 (51.5%)	373 (65.4%)	375 (65.9%)
Lifetime DSM-5 anxiety disorder (yes)*	5,991 (17.0%)	5,032 (16.1%)	519 (22.2%)	141 (26.4%)	176 (33.7%)
Lifetime DSM-5 mood disorder (yes)*	8,642 (23.9%)	7,235 (22.7%)	738 (31.1%)	237 (43.7%)	255 (46.7%)
Lifetime comorbid mental health and substance use disorders (yes)*	7,954 (22.3%)	6,513 (20.8%)	777 (32.0%)	244 (43.4%)	277 (50.2%)
Lifetime History of Spouse/Partner with Alcohol Problems (yes)*	6,240 (16.2%)	5,304 (15.6%)	570 (22.7%)	116 (19.1%)	142 (23.2%)

Notes: Missing data for individual variables (≤2%). All percentages are weighted. Asterisk denotes statistical significance based on differences in sexual orientation (*p < 0.001).

^aHeterosexual concordant refers to individuals who identified as heterosexual without any same-sex attraction or same-sex behavior.

^bHeterosexual discordant refers to individuals who identified as heterosexual with same-sex attraction or same-sex behavior.

Table 2 displays the prevalence of lifetime history of spouse/partner with alcohol problems, mental health disorders, and SUDs, stratified by sex and sexual orientation. Among females, the prevalence of lifetime history of having a spouse/partner with alcohol problems and each respective mental health disorder or SUD outcome was greater in magnitude among bisexual individuals relative to the other sexual orientation subgroups. Among males, an estimated 15.9% of heterosexual discordant individuals reported a lifetime history of having a spouse/partner with alcohol problems. The prevalence of AUD, TUD, and comorbid mental health and SUDs was greater in magnitude among bisexual males, whereas the prevalence of SUD, anxiety disorder, and mood disorder was greater in magnitude among gay males relative to the other sexual orientation subgroups.

Table 2.

Estimated Prevalence of Lifetime History of Spouse/Partner with Alcohol Problems and DSM-5 Mental Health and Substance Use Disorders among U.S. Adults by Sex and Sexual Orientation

	Females				Males
	Heterosexual Concordanta	Heterosexual Discordantb	Gay/Lesbian	Bisexual	Heterosexual Concordanta	Heterosexual Discordantb	Gay	Bisexual
	(n = 17,851)	(n = 1,520)	(n = 265)	(n = 422)	(n = 14,280)	(n = 870)	(n = 321)	(n = 144)
Lifetime History	% (95% CI)	% (95% CI)	% (95% CI)	% (95% CI)	% (95% CI)	% (95% CI)	% (95% CI)	% (95% CI)
Spouse/Partner with Alcohol Problems	21.7 (20.7, 22.6)	27.2 (24.2, 30.4)	25.7 (20.0, 32.4)	28.8 (23.5, 34.8)	9.2 (8.6, 9.8)	15.9 (13.3, 18.9)	14.4 (10.6, 19.2)	9.1 (5.3, 15.3)
DSM-5 Alcohol Use Disorder	20.6 (19.5, 21.6)	37.0 (33.8, 40.4)	45.8 (38.9, 52.9)	51.7 (44.6, 58.8)	36.0 (34.7, 37.2)	32.7 (28.9, 36.6)	46.0 (39.8, 52.3)	52.6 (41.9, 52.3)
DSM-5 Tobacco Use Disorder	22.5 (21.4, 23.5)	33.9 (30.9, 37.1)	37.1 (30.3, 44.4)	42.2 (35.7, 48.9)	31.9 (30.5, 33.4)	30.6 (27.2, 34.1)	40.7 (34.3, 47.3)	52.0 (41.6, 62.1)
Any DSM-5 Substance Use Disorder c	33.9 (32.8, 35.1)	53.7 (50.3, 57.2)	62.8 (55.8, 69.3)	66.5 (59.0, 73.3)	50.3 (48.9, 51.7)	47.9 (44.1, 51.9)	67.3 (61.8, 72.4)	64.3 (54.0, 73.5)
DSM-5 Anxiety Disorder d	20.5 (19.6, 21.5)	27.l (24.1, 30.3)	26.1 (20.5, 32.6)	36.9 (31.4, 42.7)	11.5 (10.8, 12.3)	14.7 (11.7, 18.1)	26.5 (19.8, 34.5)	25.9 (18.5, 35.1)
DSM-5 Mood Disorder e	28.0 (26.9, 29.1)	37.2 (33.8, 40.6)	45.7 (38.0, 53.6)	50.9 (44.8, 57.0)	17.1 (16.2, 18.0)	21.8 (18.8, 25.0)	42.2 (35.2, 49.6)	36.3 (25.6, 48.5)
Comorbid Mental Health and Substance Use Disorders f	20.4 (19.4, 21.4)	36.8 (33.6, 40.2)	45.0 (38.7, 51.5)	52.1 (44.5, 59.6)	21.2 (20.1, 22.4)	24.6 (21.2, 28.3)	42.3 (34.2, 50.7)	45.5 (34.5, 56.9)

Notes.

^aHeterosexual concordant refers to individuals who identified as heterosexual without any same-sex attraction or same-sex behavior.

^bHeterosexual discordant refers to individuals who identified as heterosexual with same-sex attraction or same-sex behavior.

^cSubstance use disorder includes at least one of the following: DSM-5 alcohol, tobacco, cannabis, inhalants/solvents, cocaine, hallucinogen, heroin, prescription opioid, sedative/tranquilizer, stimulant, club drug, or other drug use disorder.

^dAnxiety disorder consists at least one of the following lifetime DSM-5 anxiety disorders (ie, agoraphobia, generalized anxiety, panic, social and specific phobias).

^eMood disorder consists of at least one of the following lifetime DSM-5 mood disorders (ie, bipolar, dysthymia, or major depressive).

^fComorbid substance use and mental disorders is defined as at least one DSM-5 mental disorder (anxiety, mood, personality, PTSD, or eating disorder) and at least one DSM-5 substance use disorder.

Table 3 depicts the bivariate associations between lifetime history of a spouse/partner with alcohol problems and lifetime DSM-5 mental health and SUDs stratified by sexual orientation. Results show that lifetime history of a spouse/partner with alcohol problems is significantly associated with increased prevalence of lifetime AUD, TUD, SUD, and comorbid mental health and SUDs among heterosexual concordant, heterosexual discordant, and bisexual individuals. The prevalence of each outcome was greater in magnitude among bisexual adults with lifetime history of spouse/partner with alcohol problems relative to all other sexual orientation subgroups. Almost nine-in-ten bisexual adults (88.1%) who had a lifetime history of a spouse/partner with alcohol problems reported a lifetime SUD. Notably, most bisexual individuals who had lifetime history of a spouse/partner with alcohol problems also reported lifetime AUD (71.6% vs. 46.0% of those without a spouse/partner with alcohol problems), lifetime TUD (71.5% vs. 37.0% of those without a spouse/partner with alcohol problems), and comorbid mental health and SUDs (78.3% vs. 41.7% of those without a spouse/partner with alcohol problems).

Table 3:

Lifetime History of Spouse/Partner with Alcohol Problems Associated with Lifetime Individual and Comorbid DSM-5 Mental Health and Substance Use Disorders among U.S. Adults by Sexual Orientation (N = 35,673)

	Heterosexual Concordanta	Heterosexual Discordantb	Gay/Lesbian	Bisexual
	(n = 32,131)	(n = 2,390)	(n = 586)	(n = 566)
Lifetime	% (95% CI)	% (95% CI)	% (95% CI)	% (95% CI)
DSM-5 Alcohol Use Disorder
No History of Spouse/Partner with Alcohol Problems	25.7 (24.8 – 26.7)*	30.0 (27.2 – 33.0)*	43.2 (37.7 – 48.8)	46.0 (39.6 – 52.6)*
Lifetime History of Spouse/Partner with Alcohol Problems	41.0 (38.9 – 43.1)	53.8 (48.4 – 59.1)	57.6 (46.3 – 68.2)	71.6 (62.3 – 79.4)
DSM-5 Tobacco Use Disorder
No History of Spouse/Partner with Alcohol Problems	23.7 (22.7 – 24.8)*	27.3 (24.6 – 30.1)*	36.5 (31.2 – 42.1)	37.0 (31.1 – 43.4)*
Any History of Spouse/Partner with Alcohol Problems	45.6 (43.7 – 47.6)	50.7 (46.1 – 55.4)	50.5 (40.4 – 60.5)	71.5 (62.1 – 79.3)
Any DSM-5 Substance Use Disorder c
No History of Spouse/Partner with Alcohol Problems	38.6 (37.5 – 39.7)*	44.9 (41.8 – 47.9)*	64.0 (59.0 – 68.7)	59.2 (52.5 – 65.6)*
Lifetime History of Spouse/Partner with Alcohol Problems	60.2 (58.3 – 62.1)	74.2 (69.1 – 78.7)	71.7 (60.1 – 80.9)	88.1 (80.5 – 92.9)
DSM-5 Anxiety Disorder d
No History of Spouse/Partner with Alcohol Problems	13.9 (13.3 – 14.5)*	18.5 (16.4 – 20.8)*	24.7 (19.5 – 30.7)	31.2 (26.3 – 36.6)
Lifetime History of Spouse/Partner with Alcohol Problems	28.4 (26.8 – 30.0)	35.2 (29.8 – 40.9)	33.5 (23.1 – 45.7)	42.1 (31.6 – 53.4)
DSM-5 Mood Disorder e
No History of Spouse/Partner with Alcohol Problems	19.4 (18.6 – 20.3)*	27.2 (24.7 – 29.9)*	41.0 (35.2 – 47.1)	43.9 (37.2 – 50.9)
Lifetime History of Spouse/Partner with Alcohol Problems	40.3 (38.5 – 42.1)	44.6 (39.1 – 50.2)	55.0 (44.2 – 65.3)	55.9 (46.6 – 64.9)
Comorbid Mental Health and Substance Use Disorders f
No History of Spouse/Partner with Alcohol Problems	17.1 (16.3 – 18.0)*	25.6 (23.4 – 27.9)*	40.3 (34.0 – 46.9)	41.7 (35.3 – 48.5)*
Lifetime History of Spouse/Partner with Alcohol Problems	40.6 (38.6 – 42.7)	54.1 (49.1 – 59.0)	56.5 (45.8 – 66.6)	78.3 (69.1 – 85.3)

Notes. Asterisk denotes statistical significance in comparison between no history of spouse/partner with alcohol problems and having a lifetime history of spouse/partner with alcohol problems (*p < 0.001).

^aHeterosexual concordant refers to individuals who identified as heterosexual without any same-sex attraction or same-sex behavior.

^bHeterosexual discordant refers to individuals who identified as heterosexual with same-sex attraction or same-sex behavior.

^cSubstance use disorder includes at least one of the following: DSM-5 alcohol, tobacco, cannabis, inhalants/solvents, cocaine, hallucinogen, heroin, prescription opioid, sedative/tranquilizer, stimulant, club drug, or other drug use disorder.

^dAnxiety disorder consists at least one of the following lifetime DSM-5 anxiety disorders (ie, agoraphobia, generalized anxiety, panic, social and specific phobias).

^eMood disorder consists of at least one of the following lifetime DSM-5 mood disorders (ie, bipolar, dysthymia, or major depressive).

^fComorbid substance use and mental disorders is defined as at least one DSM-5 mental disorder (anxiety, mood, personality, PTSD, or eating disorder) and at least one DSM-5 substance use disorder.

Table 4 displays the multivariable associations of sexual orientation and lifetime history of spouse/partner with alcohol problems with lifetime AUD, TUD, SUD, anxiety disorder, mood disorder and comorbid mental health and SUDs, adjusting for sex, age, race-ethnicity, educational attainment, urbanicity, geographic region, and parental lifetime history of alcohol problems. Having a history of a spouse/partner with alcohol problems was associated with greater odds of each outcome (adjusted odds ratio [AOR] range 1.6–2.3, p<0.001). Sexual minority individuals—particularly bisexual adults—had significantly increased odds of each outcome, compared to heterosexual concordant adults (AOR range 1.3–3.2, p<0.001).

Table 4.

Estimated Multivariable Associations of Sexual Orientation and Lifetime History of Spouse/Partner with Alcohol Problems with Lifetime DSM-5 Substance Use and Mental Disorders (n = 35,662)

	Lifetime Substance Use and Mental Disorders
	Model 1 AUD	Model 2 TUD	Model 3 Any SUD	Model 4 Anxiety Disorder	Model 5 Mood Disorder	Model 6 Comorbid Mental Health and Substance Use Disordersa
	AOR (95% CI)	AOR (95% CI)	AOR (95% CI)	AOR (95% CI)	AOR (95% CI)	AOR (95% CI)
Sexual Orientation
Heterosexual Concordantb	Ref	Ref	Ref	Ref	Ref	Ref
Heterosexual Discordantc	1.4 (1.2–1.6)*	1.3 (1.1–1.5)*	1.5 (1.3–1.7)*	1.3 (1.2–1.5)*	1.4 (1.2–1.6)*	1.7 (1.4–1.9)*
Gay/Lesbian	1.7 (1.4–2.1)*	1.7 (1.3–2.1)*	2.3 (1.9–2.8)*	1.9 (1.4–2.5)*	2.6 (2.0–3.2)*	2.6 (2.0–3.4)*
Bisexual	2.4 (1.9–3.2)*	2.2 (1.8–2.9)*	2.5 (1.8–3.3)*	2.1 (1.7–2.6)*	2.3 (1.8–3.1)*	3.2 (2.4–4.3)*
Lifetime History of Spouse/Partner with Alcohol Problems
No	Ref	Ref	Ref	Ref	Ref	Ref
Yes	1.9 (1.7–2.1)*	2.1 (1.9–2.3)*	2.1 (2.0–2.3)*	1.6 (1.5–1.8)*	1.8 (1.7–2.0)*	2.3 (2.1–2.5)*
Design-Adjusted Goodness of Fit	P < 0.001	P = 0.029	P = 0.399	P = 0.087	P = 0.511	P = 0.243

Notes. AOR, adjusted odds ratio; AUD, DSM-5 alcohol use disorder; CI, confidence interval; SUD, DSM-5 substance use disorder; TUD, DSM-5 tobacco use disorder.

^*p < 0.001. All models controlled for sex, age, race-ethnicity, educational attainment, urbanicity, geographic region and parental lifetime history of alcohol problems.

^aComorbid substance use and mental disorders is defined as at least one DSM-5 mental disorder (anxiety, mood, personality, PTSD, or eating disorder) and at least one DSM-5 SUD (alcohol, tobacco, cannabis, inhalants/solvents, cocaine, hallucinogen, heroin, prescription opioid, sedative/tranquilizer, stimulant, club drug, or other drug use disorder).

^bHeterosexual concordant refers to individuals who identified as heterosexual without any same-sex attraction or same-sex behavior.

^cHeterosexual discordant refers to individuals who identified as heterosexual with same-sex attraction or same-sex behavior.

In our interaction tests (Table 5), we found no significant interactions when using the p < .001 criteria we set for this study to adjust for multiple testing, suggesting that the effect of having a spouse/partner with alcohol problems is similar across the sexual orientation subgroups. However, when using the less stringent criteria of p < .05, we found two marginally significant results. The association of having a history of spouse/partner with alcohol problems and lifetime mood disorder was marginally weaker for bisexual adults compared to heterosexual concordant adults (see Supplemental Figure 1). Additionally, the association of having a history of spouse/partner with alcohol problems and comorbid mental health and SUDs was marginally weaker for gay/lesbian adults compared to heterosexual concordant adults (see Supplemental Figure 2).

Table 5.

Estimated Multivariable Associations of Sexual Orientation, Lifetime History of Spouse/Partner with Alcohol Problems, and their Interaction with Lifetime DSM-5 Substance Use and Mental Disorders (n = 35,662)

	Lifetime Substance Use and Mental Disorders
	Model 1 AUD	Model 2 TUD	Model 3 Any SUD	Model 4 Anxiety Disorder	Model 5 Mood Disorder	Model 6 Comorbid Mental Health and Substance Use Disordersa
	AOR (95% CI)	AOR (95% CI)	AOR (95% CI)	AOR (95% CI)	AOR (95% CI)	AOR (95% CI)
Sexual Orientation
Heterosexual Concordantb	Ref	Ref	Ref	Ref	Ref	Ref
Heterosexual Discordantc	1.3 (1.1–1.5)*	1.3 (1.1–1.5)*	1.4 (1.2–1.6)*	1.3 (1.1–1.5)*	1.5 (1.3–1.7)*	1.7 (1.5–1.9)*
Gay/Lesbian	1.8 (1.4–2.3)*	1.8 (1.4–2.3)*	2.5 (2.0–3.2)*	2.1 (1.5–2.8)*	2.8 (2.1–3.7)*	2.9 (2.2–3.9)*
Bisexual	2.3 (1.7–3.1)*	2.1 (1.5–2.8)*	2.3 (1.6–3.1)*	2.4 (1.9–3.1)*	2.7 (2.0–3.7)*	3.0 (2.2–4.1)*
Lifetime History of Spouse/Partner with Alcohol Problems
No	Ref	Ref	Ref	Ref	Ref	Ref
Yes	1.9 (1.7–2.1)*	2.1 (1.9–2.3)*	2.1 (1.9–2.3)*	1.7 (1.5–1.8)*	1.9 (1.7–2.1)*	2.3 (2.1–2.5)*
Lifetime History of Spouse/Partner with Alcohol Problems X Sexual Orientation d
Yes x Heterosexual Discordantc	1.2 (0.4–1.4)	0.9 (0.7–1.2)	1.3 (1.0–1.8)	1.0 (0.7–1.3)	0.8 (0.6–1.0)	1.0 (0.8–1.3)
Yes x Gay/Lesbian	0.8 (0.4–1.4)	0.7 (0.4–1.1)	0.5 (0.3–1.0)	0.6 (0.4–1.2)	0.6 (0.4–1.0)	0.6 (0.3–0.9)†
Yes x Bisexual	1.4 (0.8–2.3)	1.5 (0.9–2.5)	1.9 (1.0–3.6)	0.6 (0.4–1.1)	0.5 (0.3–0.9)†	1.4 (0.8–2.5)
Adjusted Wald F-Test for Interaction	P = 0.30	P = 0.18	P = 0.004	P = 0.280	P = 0.012	P = 0.073
Design-Adjusted Goodness of Fit Test	P < 0.001	P = 0.021	P = 0.6988	P = 0.087	P = 0.745	P = 0.291

Notes. AOR, adjusted odds ratio; AUD, DSM-5 alcohol use disorder; CI, confidence interval; SUD, DSM-5 substance use disorder; TUD, DSM-5 tobacco use disorder. Asterisk denotes statistical significance (*p < 0.001). Cross denotes marginal significance (†p < 0.05). All models controlled for sex, age, race-ethnicity, educational attainment, urbanicity, geographic region and parental lifetime history of alcohol problems.

^aComorbid substance use and mental disorders is defined as at least one DSM-5 mental disorder (anxiety, mood, personality, PTSD, or eating disorder) and at least one DSM-5 SUD (alcohol, tobacco, cannabis, inhalants/solvents, cocaine, hallucinogen, heroin, prescription opioid, sedative/tranquilizer, stimulant, club drug, or other drug use disorder).

^bHeterosexual concordant refers to individuals who identified as heterosexual without any same-sex attraction or same-sex behavior.

^cHeterosexual discordant refers to individuals who identified as heterosexual with same-sex attraction or same-sex behavior.

^dReference category for interaction tests were heterosexual concordant individuals without a lifetime history of spouse/partner with alcohol problems.

We also conducted interaction tests in supplemental sex-stratified analyses (Supplemental Tables 1 and 2) and found no significant interactions when using the p < .001 criteria. We found that among females, the association of having a history of spouse/partner with alcohol problems and lifetime comorbid mental health and SUDs was marginally stronger for bisexual females compared to heterosexual concordant females when using the less stringent criteria of p < .05 (see Supplemental Figure 3). Additionally, the association of having a history of spouse/partner with alcohol problems and TUD (Supplemental Figure 4) and SUD (Supplemental Figure 5) was weaker for gay/lesbian females compared to heterosexual concordant females. Among males, the association of a history of spouse/partner with alcohol problems and TUD (Supplemental Figure 6) and SUD (Supplemental Figure 7) was stronger for bisexual compared to heterosexual concordant males. These interactions had p < .05, but this was greater than our p < .001 criterion accounting for the multiple testing.

Discussion

Our results showed that the prevalence of having a lifetime history of a spouse/partner with alcohol problems and each lifetime mental health and SUD outcome was significantly higher among sexual minority adults—especially bisexual adults—compared to heterosexual concordant adults. Having a lifetime history of a spouse/partner with alcohol problems is significantly associated with each mental health or SUD outcome in our study. Sexual minority adults had increased odds of each outcome compared to their heterosexual concordant counterparts. However, with a couple of marginally significant exceptions, the results of our interaction tests showed that the effect of having a lifetime history of spouse/partner with alcohol problems is generally similar across all sexual orientation subgroups. These findings suggest important practice implications for clinicians providing behavioral health treatment, family therapy, or couples therapy.

Relationship history with current or former partners with problematic alcohol use should be considered in the assessment and treatment of mental health and SUDs. A comprehensive assessment should include an examination of drinking patterns in their clients’ relationships. Exploring relationship dynamics and addressing problematic alcohol use among partners have implications for treatment, as this can influence their overall wellbeing, relationship satisfaction and stability, and marital and family development (Homish & Leonard, 2007; Howe et al., 2019; Leonard & Eiden, 2007; Low et al., 2007; Wiersma-Mosley et al., 2016). The spouses/partners of individuals with problem drinking often experience high levels of stress, anxiety, and depression resulting from the unpredictable behavior and emotional instability associated with alcohol misuse (Homish et al., 2006; Sharma et al., 2016). Individuals seeking substance use treatment may face greater challenges in maintaining their sobriety if they are in a relationship with a spouse/partner who actively misuses alcohol (Ellis et al., 2004). The spouse/partner’s alcohol misuse and the quality of spouse/partner relationship can influence an individual’s recovery from mental health or SUD.

Sexual orientation should also be considered in the assessment and treatment of mental health and SUDs. Bisexual individuals with a history of spouse/partner with alcohol problems consistently had the highest estimated prevalence of SUDs. This is consistent with other research that found greater severity of SUDs among sexual minority populations, especially bisexual-identified adults, relative to their heterosexual-identified counterparts (Boyd et al., 2019; McCabe et al., 2019). Adults with more severe SUD may require more intensive substance use treatment. In our study, almost three-quarters of bisexual adults with a history of spouse/partner with alcohol problems reported lifetime TUD, a persistent disorder that can be difficult to treat and remains one of the leading causes of preventable death. Tobacco cessation efforts can be integrated into couples’ therapy, particularly if a member of the romantic dyad identifies as bisexual. Notably, almost eight-in-ten bisexual individuals who had a spouse/partner with drinking problems met the criteria for comorbid mental health and SUDs, which requires specialized intervention from behavioral health providers who are competent in treating dual diagnoses. Although our findings are consistent with prior research showing the unique health risks associated with having a family member with alcohol dependence (Ray et al., 2009), our study highlights the heterogeneity of sexual minority populations and the need to consider sexual orientation as part of a holistic approach to assessment and treatment.

Many sexual minority individuals may be reluctant to disclose their sexual orientation to providers due to heteronormative assumptions and fear of further discrimination or harm (Brooks et al., 2018). Bisexual individuals, in particular, face unique stressors (e.g., rejection from both heterosexual and gay/lesbian communities) related to their bisexual identity (Balsam & Mohr, 2007; Bostwick, 2012). Bisexual stigma is evidenced through negative stereotypes and attitudes toward bisexuality and is associated with an increased risk of poor mental health and substance use among bisexual people (Balsam & Mohr, 2007; Bostwick, 2012; Feinstein & Dyar, 2017). Behavioral health providers should create welcoming and inclusive environments where clients may feel safe to disclose their sexual orientation. This requires providers to assess their own heteronormative assumptions and address any barriers in their organizational policies and practice.

Behavioral health providers should also be aware of the challenges of the compounded stigma that sexual minority clients may encounter. In the U.S., almost all individuals, regardless of sexual orientation, who seek treatment for mental health and SUDs usually experience behavioral health stigma (i.e., an internalized negative self-perception related to having an addiction or mental illness), which in large, part stems from societal beliefs and misconceptions regarding personal control, will power, and normalcy (Ahmedani, 2011; Shim & Rust, 2013). Sexual minority individuals receiving treatment for mental health and SUDs experience behavioral health stigma in addition to minority stress related to their sexual minority identity (Rojas et al., 2019). We emphasize the importance of reducing the shame of addiction or mental illness and creating supportive environments that reinforces sexual minority clients’ and their family’s recovery process. The findings also indicate the potential value in having mental health and SUD treatment available for spouses/partners and other family members, given that most of them have experienced a DSM-5 mental health or SUD.

Our study also highlights some sex differences among sexual orientation subgroups. In our sex-stratified analyses, we found that the having a history of spouse/partner with alcohol problems was more prevalent among females—particularly bisexual females—than among males. Our supplemental analyses also resulted in several marginally significant results among bisexual females and males that warrant additional research. We found that the association of having a history of spouse/partner with alcohol problems and lifetime comorbid mental health and SUDs was marginally stronger for bisexual females compared to their heterosexual concordant counterparts. The association of a history of spouse/partner with alcohol problems and TUD and SUD, respectively, was marginally stronger for bisexual males compared to their heterosexual concordant counterparts. Understanding clients’ sexual orientation and how experiences may differ by sex enables treatment specialists to consider potential risk factors in the social environment (e.g., spouse/partner’s alcohol use, stigma related to their clients’ sexual minority identity) that may be associated with poor mental health and substance use among specific subgroups.

Strengths, Limitations, and Future Research

The strengths of this study include the use of a large, nationally representative, probability-based sample of non-institutionalized U.S. sexual minority adults, with survey items that assessed multiple dimensions of sexual orientation, thus enabling us to examine heterosexual-identified sexual minorities. It also contained items assessing relationship history with a spouse/partner with alcohol problems and validated measures of mental health and SUDs based on DSM-5 criteria.

We also note several limitations to our study. Due to the cross-sectional design of this study, we lack information about the temporality between having a lifetime history of spouse/partner with alcohol problems and the presence of DSM-5 mental health and SUDs. The NESARC-III study did not assess the partners’ sociodemographic information, sexual orientation, or marital/relationship satisfaction and thus, we were unable to control for these variables in our analyses. The wording of items assessing a lifetime history of spouse/partner with alcohol problems inherently limits the sample to respondents who had been married or living as married at some point in their lives, thereby excluding those in long-term relationships or otherwise dating.

Future research should include more recent indicators (e.g., past-year, past-month) of respondents’ history of having a spouse/partner with alcohol problems and investigate how other mechanisms such as socialization (partner influencing one’s drinking behavior) or assortative mate selection (choosing a partner with similar drinking patterns as one’s own) may influence these outcomes, particularly among sexual minority populations. For race-ethnicity, individuals who identified as Asian, Native Hawaiian or other Pacific Islander, and American Indian or Alaska Native were collapsed together to ensure adequate power to detect group differences. We note that the sample size for bisexual males was small, which contributed to the wide confidence intervals in the sex-stratified analyses that produced a couple of marginally significant interaction test results. The NESARC-III study did not include items measuring gender identity and thus, we were unable to control for this in our analyses. Future work should address mental health and SUDs among gender minority individuals with a history of spouse/partner with alcohol problems. Finally, a more recent version of the NESARC study is not yet available. The data were collected over 12 years ago; sexual identity labels continue to evolve, and the operationalization of sexual minority individuals does not represent all identities that exist (e.g., demi-sexual, pansexual).

Conclusion

In summary, our study highlights the importance of understanding the unique needs of sexual minority individuals seeking treatment for mental disorders or SUDs, especially in the context of a past or current partner’s alcohol use. The prevalence of having a past or current spouse/partner with drinking problems varies by sexual orientation and sex, with those who identify as bisexual bearing the greatest burden. Moreover, the estimated prevalence of lifetime history of having a spouse/partner with alcohol problems was higher among females than males within each sexual orientation subgroup. Study findings suggest that for U.S. sexual minority adults, having a lifetime history of spouse/partner with alcohol problems substantially increases the odds of each mental disorder and SUD outcome. Clients’ relationship history with a spouse/partner with alcohol problems and sexual orientation should be considered as part of a comprehensive assessment and treatment process for psychiatric disorders. Our findings also highlight the importance of developing evidence-based resources and health professionals’ skills in advocating, promoting, and supporting sexual minority individuals and their spouses/partners/families struggling with mental illness and/or substance use.