It is generally understood that the great majority of individuals with psychiatric disorders, including both mental and substance use disorders, do not receive treatment for them [1–3]. Many studies focusing on issues that pertain to unmet need for mental health treatment have found that underutilization of treatment is highest among those groups that are traditionally underserved, including the elderly, racial/ethnic minorities, those with low-incomes, the uninsured, and residents of rural areas [3, 4]. It is also well documented that utilization of substance abuse treatment services is higher among individuals who have co-occurring mental disorders . Further, treatment use varies by several key sociodemographic characteristics. For example, after controlling for number of disorders and other demographic characteristics, men with at least one past-year disorder had nearly twice the odds of having received substance abuse services, compared with women. In contrast, women were more likely to seek mental health treatment, after controlling for both the presence of psychiatric disorder and its severity . One group that has been identified as heavier users of mental health services is lesbians, gay men, and bisexual individuals , although the reasons for this are not well understood . This paper examines the relationship of gender and sexual orientation with treatment received for substance use or mental disorders in a population-based survey.
Prevalence of substance use and mental disorders among sexual minority groups
Prior epidemiological surveys, both population-based and respondent-driven, have shown that minority sexual orientation populations report higher rates of drug use and related problems than do others [9, 10]. Findings regarding alcohol use among sexual minorities are less consistent and often limited by the challenges of obtaining representative samples . Analyses conducted with national survey data have shown lower rates of alcohol abstention and higher rates of alcohol use and problem drinking among homosexually active women compared with heterosexually active women, but no difference between homosexually active and heterosexually active men, controlling for sociodemographic characteristics . In contrast, Hughes and colleagues found no differences between lesbians and heterosexual women in self-reported alcohol problems using national survey data . Other studies have found few differences in alcohol consumption or symptoms of alcohol dependence among men with same-sex partners compared to men with opposite-sex partners [14, 15]. However, there may be differences between gay men and lesbians in their patterns of substance use, with gay men having higher rates of inhalant and marijuana use compared with lesbians, and with older age associated with reduced marijuana use among lesbians, but not gay men .
Additional evidence comes from a population-based survey of women aged 18 to 29 in low-income neighborhoods in Northern California; women who reported having both male and female sexual partners had significantly higher rates of injection drug use compared with others . Similarly, a survey of women in California showed that homosexually experienced women, particularly those who had both male and female sexual partners, reported higher and riskier alcohol use compared with exclusively heterosexually experienced women . Stall and colleagues  surveyed men in 4 major urban areas who reported having male sex partners and found they had elevated levels of alcohol-related problems and recreational drug use. Moreover, their substance use was associated in complex ways with adverse early life circumstances, social and sexual practices, current mental health status, and degree of connection to gay male culture.
Elevated rates of some common mental disorders among sexual orientation minorities have also been demonstrated . Using the National Comorbidity Survey, Gilman and colleagues found that women with same-sex sexual partners had a significantly higher likelihood of having any psychiatric disorder in the past year, including major depression, simple phobia, and posttraumatic stress disorder, compared with women who had only male partners . Men reporting same-sex sexual partners were more likely than men reporting only opposite-sex partners to have an anxiety, mood, or substance use disorder. Cochran and colleagues  used national survey data to show that gay/bisexual men had a higher prevalence of depression, panic attacks, and psychological distress compared with heterosexual men, whereas lesbian/bisexual women had a greater prevalence of generalized anxiety disorder than heterosexual women. Last, a recent study showed higher rates of hazardous drinking, lifetime and current depression, and childhood sexual abuse among sexual minority women, compared with heterosexual women who were matched on demographics .
Several explanations have been posited for the generally higher prevalence of both substance use and mental health disorders among sexual minority populations. One study using national survey data showed that women who reported same-sex sexual partners spent more time in bars and party settings, and that these women consumed more alcohol in these settings, compared with exclusively heterosexual women . Although gay men spent more time in bars than bisexual and heterosexual men, rates of heavy drinking among men did not vary by sexual orientation across settings. Thus, for lesbians especially, the social context of bars and parties may promote increased alcohol consumption .
Others studies have documented a link between having a sexual orientation minority status and exposure to life stressors, often stemming from experiences of discrimination and stigma , antigay violence or harassment (among men) , relative lack of coping skills , childhood adversity and familial rejection , and lack of other resources . Indeed, the developmental challenges encountered by young gay/bisexual male youth often includes gay-related harassment and homophobic attacks, which have been associated with adverse health problems among adult gay men . Moreover, several studies have demonstrated higher rates of psychological distress among gay, lesbian, or bisexual men and women, or homosexually experienced heterosexuals, as compared with individuals who were exclusively heterosexual, after adjusting for other confounding factors [7, 24, 28]. According to the "stress and vulnerability" model  and the "minority stress" model , these disparities in health among sexual minorities may be attributed to their cumulative exposure to harassment, maltreatment, discrimination, and victimization stemming from a hostile and homophobic culture. Thus, mental health and substance use disorders are not intrinsic to sexual minority orientation, but most likely result from the greater exposure to stressors typically experienced by sexual minorities, coupled with other individual and environmental risk factors [7, 31, 32]
Treatment utilization among sexual minorities
Findings suggest that patterns of mental health and substance misuse treatment utilization among sexual minority groups differ from those of heterosexuals. For example, in a study using the 2000 National Alcohol Survey, Drabble and colleagues found that although lesbian and bisexual women had lower abstention rates overall, they were also more likely to report alcohol-related problems (e.g., being in fights or arguments, having conflicts with spouse/partner, losing time at work) and to have sought help for an alcohol problem . Hughes and colleagues found that lesbians were more likely to report being in recovery or having received treatment for alcohol-related problems, although they consumed less alcohol than a matched sample of heterosexual women [13, 33]. In a survey of over 2,000 lesbians and bisexual women recruited through multiple outreach strategies in California, only about two-fifths (41.5%) of respondents who reported impairment related to drug use had received lifetime professional help for a substance use problem and 16% wanted, but had not received, such assistance . In another study using a population-based sample of women in Los Angeles County, Diamant and colleagues  found that lesbians and bisexual women were more likely than heterosexual women to use tobacco and alcohol, and, among lesbians, to drink heavily, however, they were less likely than heterosexual women to have health insurance, more likely to have been uninsured for health care in the preceding year, and more likely to have had problems in obtaining needed medical services.
In contrast, studies of mental health services utilization have shown that lesbians tend to utilize mental health services at higher rates and for longer duration as compared with heterosexual women . One study showed that prior traumatic events, including childhood sexual abuse and physical abuse, were strongly associated with use of mental health services for lesbians, but were unrelated to treatment use for heterosexual women . Another study used national survey data to examine receipt of mental health/substance abuse services among both men and women, comparing those with same-gender sex partners and those who were exclusively heterosexual . Both men and women who had same-gender sex partners in the past year were more likely than their respective counterparts to have sought mental health/substance abuse services over the same period.
Taken collectively, these findings suggest that help seeking for mental health and substance abuse problems may be differentially influenced by sexual orientation status and gender. However, much of this work has been hampered by small sample sizes and limited assessment of clinical disorders. Further, sexual minorities who are also ethnic minorities face additional barriers to seeking health services and are less likely to receive care .
The goal of the present paper is to examine the relationship of gender and sexual orientation with treatment utilization for psychiatric problems, including both mental health (MH) and alcohol and drug (AOD) disorders. We apply the Gelberg-Andersen Behavioral Model for Vulnerable Populations . This model, a modified version of the original Andersen behavioral model of health services utilization [41, 42], posits a set of factors that influence services use. These include predisposing characteristics that exist prior to the perception of illness (e.g., race, education, age), resources that facilitate or, when lacking, impede health services utilization (e.g., income, health insurance, social support), and need variables that pertain to the type and severity of disorder(s). In addition to these domains, the expanded model for vulnerable populations takes into consideration other factors that may facilitate or impede services utilization among populations that encounter greater risks, such as residential instability, exposure to trauma and victimization, substance abuse and mental illness, and associated life stressors . We hypothesize that individuals with sexual minority orientations will be more likely than heterosexuals to participate in treatment due to higher levels of stress and vulnerability, after controlling for predisposing, enabling, and need-related variables. We also anticipate the highest rates of treatment use to be among lesbians and bisexual women, reflecting the dual stress of vulnerability from both minority sexual orientation and gender.