Medication Use Among Sexual-Minority Populations for Self-Reported Feelings of Depression and Anxiety

This study used data on adults ages 18 years and older from the 2013–2017 National Health Interview Survey (NHIS), a nationally representative health survey of the civilian, noninstitutionalized population. Conducted annually by the National Center for Health Statistics (NCHS) at the Centers for Disease Control and Prevention, the NHIS provides timely and comprehensive data used to monitor the nation’s health (31). The family core questionnaire records basic demographic, health, and disability information for each household member. A single adult in each household is selected at random for a detailed interview on more specific health information that includes sexual-orientation identity, health conditions, and health services utilization. Participants are asked which of the following categories best represents how they think of themselves: lesbian or gay; straight, that is, not lesbian or gay; bisexual; something else; I don’t know the answer; or refuse.

Our study sample was drawn from the sample adult component of the 2013–2017 NHIS, which was accessed through the University of Minnesota’s Integrated Public Use Microdata Series Health Series, a harmonized and publicly available version of the NHIS (32). Beginning in 2013, approximately one-half of all adults in the sample were randomly selected to participate in the Adult Functioning and Disability (AFD) supplement, which is a multinational project designed to develop population-based measures of wellness, quality of life, and functioning (33). Adults receiving the AFD supplement were asked detailed questions about their activity limitations, cognitive disabilities, and mental health. More specifically, participants were asked, “How often [do you] feel worried, nervous, or anxious?” and “How often do you feel depressed?” Respondents could answer with the following options: daily, weekly, monthly, a few times a year, and never. First, we classified participants as having frequent feelings of depression if they said they felt depressed daily or weekly. Next, we classified participants as having frequent feelings of anxiety if they indicated feeling anxious daily or weekly. All respondents were also asked the following follow-up questions, “Do you take medication for depression?” and “Do you take medication for these [worried, nervous, or anxious] feelings?” Medication use was dichotomized as either yes or no.

We classified sexual orientation among AFD respondents as heterosexual (N=77,163), gay or lesbian (N=1,385), bisexual (N=737), or other (N=257). We excluded AFD respondents who did not know the answer (N=491), refused to answer the sexual-orientation question (N=492), or had incomplete data on activities of daily living (ADLs) (N=8).

We used descriptive statistics and chi-square tests to characterize the study sample by sexual orientation. We examined differences in the following sociodemographic variables: age category in years (18–24, 25–34, 35–44, 45–54, 55–64, ≥65), race-ethnicity (white, black, Hispanic, other or multiple races, missing data), citizenship (U.S. citizen or not, missing data), educational attainment (less than high school, high school graduate, some college, bachelor’s degree or higher, missing data), relationship status (married or living with a partner, formerly married, never married, missing data), children ages ≤18 in the household, family income relative to the federal poverty guidelines (FPG) (≤100%, 100%−199%, 200%−399%, ≥400%, missing data), current cigarette smoker (smoking more than 100 cigarettes in an individual’s lifetime and currently smoking every day or some days), self-rated health status (excellent/very good/good, poor/fair, missing data), needing help with ADLs or instrumental ADLs (IADLs), health insurance status (private, public, uninsured, missing data), office visit with a provider in the prior 12 months, and unmet medical care needs due to cost in the prior 12 months.

We estimated multivariable logistic regression models to compare the odds of self-reporting feelings of frequent depression or frequent anxiety by sexual orientation. Our fully adjusted models controlled for age, sex, race-ethnicity, relationship status, the presence of a child in the household, educational attainment, family income relative to FPG, current smoking status, self-rated health status, needing help with ADLs or IADLs, recent office visit with a provider, unmet medical care due to cost, U.S. Census region, and survey year. Finally, using fully adjusted logistic regression models, we compared the odds of taking medication for depression and anxiety—first for the entire sample (controlling for self-reported feelings of depression or anxiety) and then for the subset of individuals who indicated having frequent feelings of depression or frequent feelings of anxiety.

Results from the logistic regression models are presented as adjusted odds ratios (AORs) with 95% confidence intervals (CIs). We conducted all analyses in Stata, version 14, using survey weights for the AFD supplement and the svy command to adjust standard errors for the complex survey design of the NHIS. Because data were collected from publicly available and deidentified sources, the Vanderbilt University Institutional Review Board deemed this analysis exempt from review.

Table 1 presents sociodemographic and health characteristics of adults in the United States by sexual orientation. Adults who identified as gay, lesbian, bisexual, or other sexual orientation were more likely to be younger compared with heterosexual adults. The racial-ethnic composition of the sample was similar across sexual-orientation groups. Gay and lesbian adults were more likely than the other sexual-orientation groups to have a college degree and a family income equal to or above 400% of FPG. Adults who identified as gay, lesbian, bisexual, or other sexual orientation were less likely than heterosexual adults to be married or living with a partner and were more likely to be a current smoker and have unmet medical care needs due to cost.

Table 2 presents the prevalence and AORs (from multivariable logistic regression models) for frequent feelings of depression and frequent feelings of anxiety by selected characteristics. Approximately 8.4% of all adults reported frequent feelings of depression, and 19.5% of all adults reported frequent feelings of anxiety (data not shown). After the analyses controlled for sociodemographic characteristics, adults who identified as gay and lesbian (OR=1.88, 95% CI=1.51–2.34), bisexual (OR=3.00, 95% CI=2.31–3.91), or other sexual orientation (OR=2.28, 95% CI=1.35–3.85) were more likely than heterosexual adults to report frequent feelings of depression. Other demographic characteristics (independent of sexual orientation) were associated with increased odds of depression. For instance, frequent feelings of depression were more likely to be reported by whites versus Hispanics, by females, by adults who were formerly married or never married versus adults who were currently married, by adults with incomes under 400% of FPG, and by current cigarette smokers, adults with public versus private health insurance, adults needing help with ADLs or IADLs, and adults with unmet medical care needs due to cost.

Table 2 also presents logistic regression results for frequent feelings of anxiety. After the analyses controlled for sociodemographic variables, adults who identified as gay and lesbian (AOR=1.69, 95% CI=1.43–1.98), bisexual (AOR=2.75, 95% CI=2.18–3.47), or other sexual orientation (AOR=2.83, 95% CI=1.96–4.09) were more likely than heterosexual adults to report frequent feelings of anxiety. Other characteristics (independent of sexual orientation) that were associated with having frequent feelings of anxiety included being younger than 45, white versus Hispanic, female, having some college education or a college degree versus a high school education, being formerly married or never married versus currently married, low income (<400% of FPG), poor/fair versus excellent/very good/good health, needing help with ADLs or IADLs, and having unmet medical care needs due to cost.

Table 3 presents the prevalence and AORs of taking medication for depression and anxiety for all adults ages 18 and older. Overall, about 8.1% of all adults took medication for depressive feelings, and 8.7% of adults took medication for anxious feelings (data not shown). After the analyses adjusted for sociodemographic characteristics, gay and lesbian adults (AOR=2.07, 95% CI=1.64–2.62) and bisexual adults (AOR=1.56, 95% CI=1.11–2.18) exhibited greater odds of taking medication for feelings of depression compared with heterosexual adults. Having self-reported feelings of depression (AOR=11.21, 95% CI=10.26–12.25) was the strongest predictor of taking medication for depressive feelings. The odds of taking medication for depressive feelings were higher among white versus Hispanic adults, females, U.S. citizens, adults who were formerly married (e.g., separated, divorced, or widowed) versus currently married, adults with incomes between 200% and 399% of FPG versus <400% of FPG, current smokers, adults with poor/fair health versus excellent/very good/good health, and adults with public health insurance.

Table 3 also presents the logistic-regression results for taking medication for anxious feelings. Gay and lesbian adults (AOR=1.85, 95% CI=1.46–2.34) were more likely than heterosexual adults to take medication for frequent feelings of anxiety. Self-reporting frequent feelings of anxiety was the strongest predictor of taking medication for anxious feelings (AOR=7.43, 95% CI=6.88–8.03). The odds of taking medication for frequent feelings of anxiety were higher among white versus Hispanic adults, females, adults who were formerly married or never married versus currently married, current cigarette smokers, adults with poor/fair health versus excellent/very good/good health, adults needing help with ADLs or IADLs, and adults with public health insurance.

Table 4 presents the prevalence and AORs of taking medication for depression and anxiety among the subset of respondents with frequent feelings of depression and anxiety, respectively. Approximately 44% of adults with frequent feelings of depression took medication for depression, and 28.4% of adults with frequent feelings of anxiety took medication for anxiety (data not shown). After the analyses controlled for sociodemographic characteristics, there were no statistically significant differences in taking medication for depression between the sexual-minority and heterosexual adults who reported frequent feelings of depression. Among all adults with frequent feelings of depression, the likelihood of taking medication for depression was greater among females, adults with college degrees versus a high school education, adults with poor/fair health versus excellent/very good/good health, adults needing help with ADLs or IADLs, and adults with public health insurance.

Table 4 also compares the prevalence of taking medication for anxiety among the respondents reporting frequent feelings of anxiety. After the analyses adjusted for sociodemographic characteristics, gay and lesbian adults who reported frequent feelings of anxiety were more likely than heterosexual adults who reported frequent feelings of anxiety to take medication for anxiety (AOR=1.69, 95% CI=1.26–2.26). Among all respondents with frequent feelings of anxiety, the likelihood of taking medication for anxiety was greater among whites versus Hispanics, females, adults who were formerly married versus currently married, current cigarette smokers, adults with poor/fair health versus excellent/very good/good health, adults needing help with ADLs or IADLS, and adults with public health insurance.

This study used representative data from the NHIS and found that adults who self-identified as a sexual minority were more likely to report frequent feelings of depression and anxiety compared with their heterosexual peers. This study suggests that public health initiatives should focus on mental health promotion, disease prevention, and treatment of depression and anxiety among sexual-minority populations.

Preventing depression and anxiety among persons who identify as a sexual minority may require multifaceted and multisector approaches. Mental health promotion and disease prevention should be included in the national dialogue about combatting stigma around mental health. Mental health promotion efforts should incorporate the social determinants of health as a framework for addressing the root causes of psychological distress and mental illness among persons with sexual-minority status in the United States. For example, public policy should ensure a welcoming and safe environment for individuals who identify as a sexual minority—which may require legal protections from discrimination in all facets of society, including health care, employment, housing, and education.

At the time of this writing, persons who identify as a sexual minority can be the victim of a hate crime in 20 states without legal repercussions and can lose their jobs because of their sexual orientation in 28 states (34). Preventing violence and traumatic experiences against persons with a sexual-minority orientation may help reduce the prevalence of distress and anxiety in the sexual-minority population in the long term. Recent research suggests that some legal protections (e.g., same-sex marriage laws) may be associated with reductions in implicit and explicit bias toward the sexual-minority community (35) and improved mental health and well-being as well (36, 37).

Meanwhile, health care facilities and providers can also play an important role in detecting, preventing, and treating depression and anxiety among sexual-minority populations. Not only should providers conduct routine screening for mood disorders among patients who identify as a sexual minority, but health care facilities and providers should ensure that their practices are uniformly welcoming and affirming, regardless of patients’ sexual orientation. Subtle cues can be used in clinical settings to communicate that the setting is a welcoming and safe environment for patients with a sexual-minority orientation, such as collecting sexual-orientation and gender identity data in health records, displaying health information featuring same-sex couples, and using gender-neutral language that does not assume that the patient engages in heteronormative romantic relationships (38). Mental health providers should also be mindful of the clinical guidelines for treating patients with a sexual-minority orientation, which include affirming identification as a sexual minority as a normal variation in human sexuality that does not need to be changed and recognizing discrimination and stigma against sexual-minority status as a risk factor for mental illness (39–41).

We also found that after the analysis controlled for sociodemographic characteristics and health status, some adults who identified as a sexual minority were more likely than their heterosexual peers to take medication for depression and anxiety. Even among adults with frequent anxiety, gay and lesbian adults were more likely than heterosexual adults to take medication for anxiety. These are interesting findings that warrant further research, given that they suggest that persons who identify as a sexual minority may be more likely to seek and receive (or possibly misuse) mental health services, treatment, and medication. It is possible that stigma surrounding mental health may be less of a concern for those who identify as a sexual minority than for heterosexual adults. Indeed, one study with a sample of college students showed that gay, lesbian, and bisexual students reported less personal stigma about mental illness compared with heterosexual students (42). Although this analysis cannot explore the relationships between attitudes about mental health care and the receipt of medication, future research should explore whether and how mental health stigma varies by sexual orientation. Additional research should continue to monitor and evaluate the effectiveness of various mental health treatments among sexual-minority populations, especially as more data on sexual orientation and gender identity are collected in health surveys and electronic health records. There were several limitations to using data from the NHIS for this study. First, all responses to the NHIS are self-reported and may suffer from selection issues and recall bias. Respondents who choose to disclose their sexual-minority status may be different than those who do not accurately disclose their sexual-minority orientation. Some research suggests that adults who do not answer the sexual-orientation question are more likely to be a member of a racial-ethnic minority group, a non-English speaker, and a resident of an urban community (43). However, an increasing proportion of adults are willing to report their sexual orientation in health surveys (44). Other responses may be subject to social-desirability bias. For instance, some participants may have been reluctant to accurately describe their mental health and treatment to avoid being viewed negatively.

We were also limited by the small number of questions on depression and anxiety in the AFD supplement to the NHIS. Although the participants were asked how frequently they felt depressive or anxious feelings, they were not asked whether they had ever received a diagnosis of depression or anxiety from a health care professional. Meanwhile, the questions on depression and anxiety may not be as valid as formal diagnostic evaluations, such as the PHQ-9—the nine questions in the Patient Health Questionnaire that ascertain the severity of depression. More research should compare the validity, sensitivity, and specificity of the AFD questions for depression and anxiety. Relatedly, participants were asked only about whether they were taking any medications for their feelings. The wording of this question does not specifically ask about prescription medications, so respondents may include prescribed medications, over-the-counter medications, and possibly vitamins or herbal supplements. Meanwhile, some participants living with depression and anxiety may not take any medications but may be receiving psychotherapy, participating in support groups, or using complementary and alternative medicine for their mood. Given the elevated attention to mental health in the United States, the NCHS should incorporate additional questions related to mental health surveillance and mental health services utilization into the adult questionnaire for the main sample of the NHIS.

This study compared the prevalence of self-reported feelings of depression and anxiety, as well as the odds of taking medication for these feelings, among persons with a heterosexual or sexual-minority orientation. Consistent with previous research, we found that persons who self-identified as a sexual minority were more likely than heterosexual adults to report frequent feelings of depression and anxiety. We also found that persons with a sexual-minority orientation were more likely than heterosexuals to take medication for depression and anxiety. Given the elevated prevalence of anxiety and depression in sexual-minority populations, we recommend using a multifaceted approach to address mental health disparities among these groups. This approach should include key roles for policy makers, public health researchers, and health care providers and emphasize mental health promotion and disease prevention. Addressing discrimination in health care and public policy toward persons who self-identify as a sexual minority may help alleviate depression and anxiety among lesbians, gay men, bisexual individuals, and other non-heterosexual populations in the United States.