Patterns of Mental Health Care Utilization Among Sexual Orientation Minority Groups

ABSTRACT

Prior studies of the utilization of mental health professionals by sexual minority populations have relied on data that are now dated or not nationally representative. These studies have also provided mixed findings regarding gender differences in the utilization of mental health professionals among sexual minority individuals. Using data from the 2013–2015 National Health Interview Surveys, this study investigates (1) how sexual minority individuals compare to heterosexual participants in their utilization of mental health professionals; and (2) gender differences in that utilization. The results indicate sexual minority individuals utilize mental health care professionals at higher rates than heterosexual individuals even after controlling for measures of mental health and other demographic characteristics; this is true for both men and women. However, gender moderates the sexual minority effect on utilization rates. Sexual minority men utilize mental health professionals at a high rate, such that their utilization rates are similar to sexual minority women, contrary to the gender gap seen among heterosexuals.

KEYWORDS:

• Gay men
• lesbian women
• LGBT health care
• LGBT mental health
• mental health utilization
• National Health Interview Surveys
• sexual minority
• sexual minority populations

Determining mental health care utilization rates across different client populations can help identify if certain marginalized groups are underutilizing or overutilizing services. This knowledge can inform health outreach, practice, and training. The focus of the current research is to examine the utilization rates of mental health professionals by sexual orientation minority groups compared to heterosexual individuals, using a nationally representative sample that includes a sizable number of sexual orientation minority individuals. Although analyses on this issue have been conducted, prior research has used data that are now two decades old or data that cannot be generalized to sexual minority populations across the United States. Past research has also produced mixed findings in regard to gender differences among sexual minority individuals in the utilization of mental health professionals.

There are contrasting arguments for why sexual minority individuals may or may not utilize a mental health care provider. For example, the option to seek services may be a difficult one for some within the sexual minority community as the fields of psychology and psychiatry have a long, unfortunate history of pathologizing homosexuality (Drescher & Merlino, 2007). Although the American Psychological Association has now strongly condemned treatments such as conversion therapy or other sexual orientation change efforts (Anton, 2010) and endorses that same-sex attraction is not a disorder, there are likely many individuals who remember this history and may feel hesitant to seek out care for fear of seeing a nonaffirmative clinician (Spengler & Ægisdóttir, 2015). Unfortunately, this reluctance may be well founded as research continues to show that clinicians struggle with providing culturally appropriate care at times, including perpetrating microaggressions (Shelton & Delgado-Romero, 2011; Sue, 2010), clinician over-pathologizing in assessment and diagnosis (Bieschke, Paul, & Blasko, 2007), and some clinicians still conducting conversion-oriented therapy (Nicolosi & Nicolosi, 2002), despite condemnations from the American Psychological Association (Anton, 2010).

However, there is also evidence to suggest that sexual minority individuals may seek mental health care at higher rates compared to the heterosexuals despite the potential stigma. Much of this research has come from various probability and nonprobability samples drawn from California. In one of the earliest studies, Bell and Weinberg (1978) collected a nonprobability sample of sexual minority men and women in the San Francisco area and found higher rates of mental health service utilization compared to the heterosexual individuals in the sample. More recently, Grella, Greenwell, Mays, and Cochran (2009) analyzed the 2004–2005 California Quality of Life Survey and found that sexual orientation minority men and women were more likely than their heterosexual counterparts to have received treatment in the prior year for an emotional, mental, or substance-related problem. Using the same data, Grella, Cochran, Greenwell, and Mays (2011) also found that sexual minority men and women utilized services at a higher rate than heterosexual individuals. In a related finding, Wallace, Cochran, Durazo, and Ford (2011) examined the 2007 California Health Interview Study and found that older sexual minority individuals are more likely to seek mental health services when they feel they have a need for support compared to older heterosexual individuals. Although suggestive, caution must be taken in extrapolating too much from data drawn from a single state to sexual minority populations across the United States. There are, however, examples of studies examining sexual minority utilization using nationally representative data.

Two studies done by Cochran and colleagues provide the best evidence thus far of mental health utilization among sexual minority individuals. The first study (Cochran, Sullivan, & Mays, 2003) used the 1995 National Survey of Midlife Development and found that those who identified as a sexual minority had higher odds of seeing some type of mental health–related care provider compared to their heterosexual counterparts. The second study (Cochran & Mays, 2000) used data from the 1996 National Household Survey of Drug Abuse and found that sexual minority individuals were more likely to have received mental health or substance abuse treatment in the past 12 months. Although these national studies of sexual minority mental health care utilization are informative and provide preliminary evidence of higher utilization rates, they come with some important limitations. The data used by Cochran and colleagues (2000, 2003) are now two decades old, having been collected in 1996 and 1995, respectively. Further, the sample size of individuals who identified as a sexual minority in both studies were relatively small, with fewer than 200 in both samples (Cochran & Mays, 2000; Cochran et al., 2003).

Public attitudes and clinical training concerning sexual minority individuals have changed considerably in the last 20 years, making it important to reexamine whether this population continues to utilize mental health professionals at a higher rate. As Cochran (2001) and Herek and Garnets (2007) discussed, there is a paucity of national probability samples that have specifically measured sexual orientation and included measures of mental health service utilization. This evolution in attitudes and training combined with a scarcity of data on the topic indicates an important need for current research on the mental health utilization of sexual minority individuals. The present study is innovative in this regard, as the data are recent (2013, 2014, and 2015 National Health Interview Surveys) with a large, nationally representative sample of sexual minority individuals, using specific measures of mental health and service utilization.

Gender differences in sexual minority mental health care utilization

The current analysis will also look for gender utilization differences within the sexual minority sample. Prior research on the differences in mental health service utilization comparing sexual minority men and women has been limited for a number of reasons, including small samples, lack of comparison groups, and nonprobability samples. This has produced mixed findings in regard to gender differences among sexual minority individuals in the utilization of mental health professionals.

Some evidence suggests that sexual minority men and women seek mental health care at approximately the same rate, and at a higher rate than heterosexuals. For instance, Cochran and Mays (2000) found similar odds ratios between sexual minority men (OR = 3.10) and women (OR = 2.90) in likelihood of receiving “mental health or substance abuse services” in the last 12 months. Although these findings resulted from an analysis of a national probability sample, the data are now somewhat dated. More recently, Grella et al. (2011) analyzed a sample drawn from California and found that sexual minority men and women had similar rates of mental health and substance abuse treatment. This finding was consistent for those with or without a psychological disorder. However, as discussed previously, a California-based sample may not represent sexual minority populations in the United States more generally.

Other research has suggested that sexual minority women are more likely to utilize mental health professionals. Such a pattern would mirror the higher utilization rates of heterosexual women relative to heterosexual men. Cochran et al. (2003) found that 19.4% of gay-bisexual men and 33.0% of lesbian-bisexual women had seen a mental health professional in the past 12 months. Similarly, Grella et al. (2009) found lesbian and bisexual women had the highest rates of treatment participation regardless of whether or not they had a diagnosable mental health condition compared to gay and heterosexual men.

Unfortunately, much of the other research on gender-related factors has focused specifically on sexual minority women either alone or in comparison to heterosexual women, without a male comparison group. For example, Bradford, Ryan, and Rothblum (1994) found about 73% of lesbian women in the National Lesbian Health Care Survey, collected in 1984–1985, have received mental health services at some point in their lifetime. However, these data did not provide comparison data to heterosexual women or any category of men. Similarly, Koh and Ross (2013) examined survey data from an availability sample of women at health care centers and found that lesbian women used psychotherapy for depression more commonly than heterosexual or bisexual women, but there was not a comparison to men of any category. Similarly, Matthews, Hughes, Johnson, Razzano, and Cassidy (2002) found that 78% of lesbian women in their large convenience sample from three states had received mental health care in their lifetime. This rate was significantly higher than the heterosexual women in the sample. These results are consistent with Sorensen and Roberts (1997), who also found that lesbian women seek services at high rates—around 80% in their national convenience sample. However, this study again sampled only lesbian-identified women, thus limiting any comparisons to heterosexual women or men of any category.

Research on the rate of utilization of mental health professionals by sexual minority men is rather limited. In one small convenience sample of gay men with HIV/AIDS, King and Orel (2012) found that around 35% had sought services. In another small study examining help seeking attitudes about couples counseling, 40% of the gay male couples indicated an unwillingness to seek help compared to only 14% of lesbian women couples (Modrcin & Wyers, 1990). Although it is well documented that heterosexual men seek mental health services at lower rates compared to heterosexual women (Kessler et al., 2005), the research on gay men’s utilization is somewhat incomplete since more attention has focused on understanding the mental health needs of sexual minority, especially lesbian women. Given these limitations in prior research on gender, the current analysis sought to examine gender-related differences in mental health utilization, again using a large national sample. The goal of the current analysis was to examine, using recent data, whether the prior findings of equal utilization (Cochran & Mays, 2000) across sexual minority men and women would still be present today in a probability sample. The aim was also to address the limited available data on rates of utilization for sexual minority men and women compared to heterosexual respondents.

Using data from the 2013, 2014, and 2015 National Health Interview Surveys, the current study had the following research questions.

1. How do sexual minority and heterosexual participants compare on utilization of mental health care professionals?

2. Among sexual minority participants, are utilization patterns of mental health care professionals different across men and women?

Method

Participants and data

The data for this research come from the Adult Core Samples of the 2013, 2014, and 2015 National Health Interview Surveys (NHIS). The NHIS is collected annually by the U.S. Census Bureau and is sponsored by the Center for Disease Control and Prevention’s National Center for Health Statistics (CDC, 2016). The NHIS uses a stratified multistage sample design and collects data using face-to-face interviewing. One adult is randomly selected from each family identified through the sampling process to produce the data in the Sample Adult Core data. These data offer nationally representative measures of the health, health care utilization, and health-related behaviors of the U.S. resident civilian adult population (CDC, 2016).

Data from a total of n = 99,091 participants were analyzed in this research. Using the final weighted data, the mean age of the sample was 46.9 years of age (Lin. SE = .131). The sample was 48.2% male and 51.8% female. The racial distribution of the sample is 79.8% White, 12.0% Black or African American, 0.9% American Indian or Alaskan Native, 5.7% Asian, and 1.63% Multiracial. Fifteen percent of the respondents report a Hispanic ethnicity. In terms of citizenship, 91.6% of the participants were citizens of the United States, and 8.4% indicated they were not a citizen of the United States.

The distribution of educational attainment among the participants was: Less than a high school degree (13.1%), High school degree or GED (25.5%), Some college (31.0%), Bachelor’s degree (19.4%), Graduate degree (11.1%). Participants’ current marital status was as follows: Married, spouse in household (52.3%), Married, spouse not in household (1.1%), Widowed (5.9%), Divorced (9.4%), Separated (2.0%), Never married (22.1%), Living with partner (7.3%). Region of residence for the sample was: Northeast (17.4%), Midwest (22.7%), South (37.2%), and West (22.8%).

Like many other national surveys, the NHIS has only recently begun to include questions regarding sexual orientation (CDC, 2016). The first year that such a question was included in the NHIS was 2013. Many national surveys lack the overall sample size to generate a large enough number of sexual minority participants. The NHIS, however, collects over 30,000 responses for its Sample Adult Core survey. This provides many more sexual minority participants than the small numbers found in the typical national survey sample. In the sample used for this research, 97.0% (n = 95,903) of participants identified as “straight, that is, not gay,” which is the terminology used in the NHIS (CDC, 2016). In terms of sexual minority participants, 3.0% (n = 3,188) of the sample identified as a category other than “straight” with the options of gay/lesbian (n = 1,674), bisexual (n = 776), something else (n = 259), and don’t know (n = 479). With n = 3,188 total sexual minority participants, this is a robust, representative probability sample of the population in the United States. This nationally representative sample allows for a current examination of professional mental health care utilization among various sexual minority groups. These analyses build on the prior work of Cochran and colleagues (2000, 2003) and address the limitations of their earlier findings using one of the largest probability samples of sexual minorities to date.

Measures

To explore the research questions for this study, NHIS measures of sexual orientation, mental health utilization of professionals, and other demographic and mental health variables were used in the analyses.

Sexual orientation measure

The NHIS introduced questions about sexual orientation in 2013 (CDC, 2016). The question on the survey asks, “Which of the following best represents how you think of yourself?” The question response options varied slightly depending on whether the participant was male or female. For male participants the possible responses were (1) gay, (2) straight, that is, not gay, (3) bisexual, (4) something else, and (5) I don’t know the answer. For female participants the possible responses were (1) lesbian or gay, (2) straight, that is, not lesbian or gay, (3) bisexual, (4) something else, and (5) I don’t know the answer.

Individuals who provide the “something else” or “I don’t know the answer” responses are asked follow-up questions to specify the nature of their response. An analysis of these responses provided by the NHIS found that a significant number of these individuals could be considered to have a sexual orientation other than heterosexual or straight (National Health Interview Survey, 2014). For example, around 20% of the “something else” respondents stated in the follow-up questions that they “do not think of [them]selves as having sexuality,” 5% stated that they are “in the process of figuring out sexuality,” and 4% stated that they “identify with another label such has queer, trisexual, omnisexual, or pansexual.” Similarly, of those that originally responded, “I don’t know the answer,” 30% stated in the follow-up questions that they “are in the process of figuring out sexuality.” Given that these categories include some sexual minority individuals broadly defined, we include the “something else” (0.2%) and the “I don’t know the answer” (0.4%) responses in the analysis alongside the straight (97.0%), gay/lesbian (1.6%), and bisexual participants (0.7%).

Mental health care utilization measure

The primary question used as an outcome in this research comes from a series of questions asked in the NHIS concerning the different types of health care providers the participant has interacted with in the past 12 months. Specifically, these questions asked, “During the past 12 months, have you seen or talked to any of the following health care providers about your own health?” Following this lead-in, nine types of health care providers were offered to participants. One of these was, “A mental health professional such as a psychiatrist, psychologist, psychiatric nurse, or clinical social worker.” In the total sample, participants indicated, “Yes” (7.5%) or “No” (92.5%) to this prompt.

Mental health measures: Recent mental distress and mental health limitation

Prior research has consistently shown sexual minority individuals report higher levels of mental health issues as compared to heterosexual individuals (Cochran, 2001; Cochran et al., 2003; Meyer, 2003), which could theoretically explain higher rates of mental health care utilization. To account for this expected differential, two measures from the NHIS were included to assess the mental health of participants. The first measure is the Kessler Psychological Distress K6 Scale (Kessler et al., 2002) an additive scale of six items asking, “During the past 30 days, how often did you feel… (1) “…worthless?” (2) “…that everything was an effort?” (3) “…hopeless?” (4) “…restless or fidgety?” (5) “…nervous?” (6) “…so sad that nothing could cheer you up?” For each item participants could say (0) None of the time, (1) A little of the time, (2) Some of the time, (3) Most of the time, or (4) All of the time. Each participant’s score on these six items was totaled to create overall measure that labeled here as “recent mental distress.” The Cronbach’s alpha for this scale is .86 (M = 2.57, Lin. SE = .02).

The second measure of participant mental health comes from a series of questions asking whether the participant has difficulty doing a variety of daily activities (e.g., “Go out to things like shopping, movies, or sporting events”) due to a health problem. If a participant reports that they have even a little difficulty with any of these activities, they are then asked to identity what health problems produce this limitation. One response coded by the NHIS is if the participant refers to “Depression/anxiety/emotional problems” that produce the limitation. A dichotomous indicator was created for whether the individual is coded as having given this response (2.2% of the sample) and is referred to as “mental health limitation.”

Other control measures

A number of factors beyond sexual orientation could also likely influence the utilization of mental health providers and any sexual orientation–related associations could be explained away by such factors. As such, a number of other control measures were included in the current research. These measures include family income, education, age, whether a person has health insurance, gender, race, region of residence, marital status, number of children in the home, and citizenship status.

Results

Analyses were conducted using Stata 13.1. All analyses utilized the software’s survey design commands to account for the complex sample design and weighting of the NHIS data in accordance with the guidance provided by the NHIS (National Health Interview Survey, 2016).

Mental health care utilization

The analyses addressing the first research question of the study, “How do sexual minority and heterosexual participants compare on utilization of mental health care professionals?” yielded a number of important findings. Table 1 shows the percentage of respondents by each sexual orientation category who stated that they “have seen or talked to a mental health professional in the past 12 months.”

Table 1. Utilization of mental health professionals in past 12 months by gender and sexual orientation.

NHIS Question	Sexual Orientation
During the past 12 months, have you seen or talked to…A mental health professional such as a psychiatrist, psychologist, psychiatric nurse, or clinical social worker?—Yes	Straight	Gay/Lesbian	Bisexual	Something Else	Don’t Know	Overall
Men [95% CI]	5.88% [5.59–6.20]	19.84%* [16.06–24.25]	25.29%* [18.15–34.06]	14.26%* [8.84–22.19]	12.50%* [7.81–19.42]	6.27% [5.96–6.60]
Women [95% CI]	8.33% [7.95–8.73]	17.81%* [14.48–21.72]	26.24%* [21.80–31.23]	18.58%* [12.42–26.87]	13.67% [7.78–22.94]	8.70% [8.30–9.10]
Overall [95% CI]	7.15% [6.90–7.41]	18.91%* [16.21–21.95]	25.97%* [22.20–30.12]	16.52%* [12.13–22.09]	13.15%* [8.92–18.98]	7.53% [7.26–7.80]
N	95,903	1,674	776	259	479	99,091

Note. National Health Interview Surveys, 2013–2015, Sample Adult Data; 95% confidence intervals show in brackets; N = 99,091;

*Percentage is significantly different (p < .05) from Straight percentage within the same row.

Overall, 7.53% of all participants reported seeing a mental health professional in the past 12 months. Comparing different sexual identity groups shows that gay/lesbian (18.91%), bisexual (25.97%), those who said they were “something else” (16.52%), and those that said they “don’t know” (13.15%) are all significantly (p < .05) more likely than heterosexual respondents (7.15%) to have seen a mental health professional in the past 12 months. In sum, based on the point estimates, sexual minority individuals are about two to four times as likely as heterosexual individuals to have seen a mental health professional in the last 12 months.

This difference in utilization could be a function of differences in mental health needs, demographic, or socioeconomic differences between these groups. To account for the influence of other factors, Table 2 shows the results of a logistic regression analysis predicting whether the respondent has seen a mental health professional in the past 12 months. This table displays odds ratios, so that numbers above 1 indicate that a particular measure is associated with an increase in the odds of having seen a mental health professional, while numbers below 1 represent a decrease in the odds of having seen a mental health professional. The first model, which includes the sexual orientation indicators, mirrors the findings of Table 1. Compared to heterosexual respondents, all four other sexual identity groups have significantly higher odds of having seen a mental health professional in the past 12 months.

Table 2. Logistic regression analysis predicting whether respondent has seen a mental health professional in previous 12 months.

Odds ratios shown >1 = increased odds, <1 = reduced odds of having seen mental health professional	Model 1 Sexual Orientation Only	Model 2 + Mental Health	Model 3 + Other Controls	Model 4 + Moderating Effects of Sex
Sexual Orientation
Straight (ref.)	–	–	–	–
Gay/Lesbian	3.02**	2.80**	2.33**	2.82**
Bisexual	4.55**	2.96**	2.16**	3.87**
Something Else	2.56**	1.23	.95	.78
Don’t Know	1.96**	1.47	1.56	1.79*
Mental Health Limitation	–	8.55**	9.22**	9.26**
Recent Mental Distress Scale	–	1.15**	1.16**	1.16**
Family Income	–	–	1.00	1.00
Education	–	–	1.25**	1.25**
Insured	–	–	2.15**	2.15**
Age	–	–	.98**	.98**
Female	–	–	1.26**	1.29**
Race
White (ref.)	–	–	–	–
Black/African American	–	–	.78**	.78**
American Indian/Alaskan Native	–	–	.93	.94
Asian	–	–	.38**	.38**
Multiple Races	–	–	.76*	.77*
Hispanic Ethnicity	–	–	.73**	.73**
U.S. Citizen	–	–	1.85**	1.85**
Marital Status
Married, spouse in household (ref.)	–	–	–	–
Married, spouse not in household	–	–	1.85**	1.84**
Widowed	–	–	1.05	1.04
Divorced	–	–	1.69**	1.69**
Separated	–	–	1.98**	1.98**
Never Married	–	–	1.43**	1.42**
Living with Partner	–	–	1.11	1.12
Region
Northeast (ref.)	–	–	–	–
Midwest	–	–	.87^†	.87^†
South	–	–	.73**	.73**
West	–	–	.93	.93
Sexual Orientation × Female
Straight × Female (ref.)	–	–	–	–
Gay/Lesbian × Female	–	–	–	.65^†
Bisexual × Female	–	–	–	.44**
Something Else × Female	–	–	–	1.40
Don’t Know × Female	–	–	–	.80

Note. 2014 National Health Interview Survey, Sample Adult Data; N = 99,091.

†p < .06 *p < .05. **p < .01.

Model 2 in Table 2 enters the two measures of respondents’ mental health. As would be expected, both of these are associated with a significant increase in the odds of the respondent having seen a mental health professional in the past 12 months. However, the differences between heterosexual respondents and some of the other sexual orientation groups remain significant. Specifically, gay/lesbian and bisexual individuals are more likely than heterosexuals to have seen a mental health professional independent of their scores on the two mental health measures. After controlling for the two mental health measures, the sexual orientation category of “something else” and “don’t know” are no longer significantly different from heterosexuals, meaning that their initial higher likelihood of having seen a mental health professional as seen in Model 1 is a function of this group’s higher scores on these measures. Whereas, for the other sexual orientation categories, Model 2 indicates that those individuals, regardless of their mental health scores have a higher likelihood of having seen a mental health professional in the last 12 months as compared to heterosexual participants.

Model 3 in Table 2 enters all of the other demographic and control measures into the model. The gay/lesbian and bisexual categories remain significantly higher in utilization as compared to heterosexuals. Respondent’s education is positively associated with odds of having seen a mental health professional, as is having health insurance. Family income is not significantly related to mental health professional utilization net of the other measures included in the model. Older respondents have lower odds of having utilized mental health professionals. Black, Asian, and multiracial respondents have lower odds of having seen a mental health professional as compared to White respondents. Similarly, Hispanic respondents all have lower odds of having seen a mental health professional in the past 12 months compared to non-Hispanic respondents. U.S. citizens have higher odds compared to noncitizens. Relative to those who are married with their spouse in the household, those who are married but not living with their spouse along with those who are divorced, separated, and never married all have higher odds of having seen a mental health professional. Finally, respondents living in the South have lower odds of using mental health professionals relative to respondents living in the Northeast. The utilization rate in the Midwest also approaches being significantly lower than in the Northeast.

Gender differences in sexual minority individuals’ mental health care utilization

The analyses for the second research question of the study, “Among sexual minority participants, are utilization patterns of mental health care different across men and women?” yielded a number of important gender-related findings within the sexual minority sample. As seen in Table 1, among all respondents, women are significantly more likely to say they have seen a mental health professional compared to men (8.70% to 6.27%). However, this gender effect appears to differ across the sexual orientation groups. Among heterosexual individuals, heterosexual women are significantly more likely to have utilized a mental health professional in the past 12 months (8.33% compared to heterosexual men at 5.88%). Among gay and lesbian respondents, men have a higher utilization rate (19.74%) than women (17.81%), although this difference is not statistically significant. Among bisexuals, the rate of utilization is also statistically equal between men (25.29%) and women (25.24%). In sum, while women have higher utilization rates than men among heterosexual respondents, this does not appear to be the case among sexual minority respondents.

To look at this descriptive finding further, Model 4 of the logistic regressions (Table 2) analyzed whether the effects of sexual orientation on mental health utilization might differ by the sex of the individual. This model includes interaction terms between gender and sexual orientation. The addition of these terms means that the odds ratio for gender now represents the gender effect for only heterosexual respondents, while the sexual orientation odds ratios signify the effects for men. The interaction terms represent the multipliers for adjusting the men’s sexual orientation effects to represent odds ratios for women.

The model confirms the pattern seen in Table 1. Among men, gay (odds ratio = 2.82, p < .01), bisexual (odds ratio = 3.87, p < .01), and men that identified as “don’t know” (odds ratio = 1.79, p < .05) have significantly higher odds of seeing mental health professionals compared to heterosexual men. If we look at the interaction terms in this model, we see that some of the sexual minority effects for women are weaker compared to those for men. Specifically, the gap in utilization between bisexual women and heterosexual women is significantly smaller than the equivalent gap between bisexual men and heterosexual men. The interaction term for gay or lesbian women is also very close to the p < .05 significance level (p = .052), which suggests that this gap might also be smaller than it is between gay men and heterosexual men.

Table 3 summarizes sexual minority effects for men and women. The odds ratio for gay or lesbian women is 1.85 compared to 2.82 for gay men. The difference between these odds ratios is very close (p = .052) to the common standard for statistical significance. The odds ratios for bisexual women is 1.72 compared to 3.87 for bisexual men. The difference between these odds ratios is statistically significant at the p < .01 level.

Table 3. Summary of sexual minority effects by gender on the utilization of mental health professionals.

	Odds Ratios for Men Compared to Straight	Odds Ratios for Women Compared to Straight	Difference between Odds Ratios
Gay or Lesbian	2.82**	1.85**	†
Bisexual	3.87**	1.72**	**
Something Else	.78	1.10
Don’t Know	1.79*	1.44

†p < .06 *p < .05. **p < .01.

Note. 2014 National Health Interview Survey, Sample Adult Data; N = 99,091.

Table 4 summarizes the gender effects for each sexual orientation. The odds ratio for heterosexual women is 1.29, which is significant at the p < .01 level. This is, of course, the same effect seen in Model 4 of Table 2. There are no significant gender differences in any of the sexual minority categories, however. This finding indicates that although heterosexual men utilize therapy at lower rates than heterosexual women, this disparity is not present between sexual minority men and women. The higher the utilization of mental health professionals by sexual minority men not only exceeds heterosexual men, but also expands to the level of sexual minority women.

Table 4. Summary of gender effects by sexual orientation on the utilization of mental health professionals.

	Odds Ratio for Women Compared to Men
Straight	1.29**
Gay or Lesbian	.88
Bisexual	.66
Something Else	1.32
Don’t Know	.78

†p < .06 *p < .05. **p < .01.

Note. 2014 National Health Interview Survey, Sample Adult Data; N = 99,091.

Discussion

The analyses indicate that sexual minority individuals utilize mental health professionals at higher rates than do heterosexual individuals. The gender difference seen among heterosexuals, with women having higher utilization rates than men, is not present among sexual minority individuals. In turn, these two findings mean that the sexual minority effect on utilization is larger for men than women. In other words, the gap in utilization between heterosexual women and sexual minority women is smaller than the gap between heterosexual men and sexual minority men. These are important findings about the relationship between sexual orientation identification, gender, and the utilization of mental health care professionals.

The finding that sexual minority individuals utilize mental health providers at a higher rate than heterosexual individuals is consistent with the prior research on this topic. The current findings build on the earlier work of Cochran et al. (2003) and Cochran and Mays (2000) by examining sexual minority utilization rates in the current population as of 2015, whereas the original findings by Cochran and colleagues are now around 20 years old. It is particularly informative that, like the work of Cochran and colleagues, in a national probability sample sexual minority populations still utilize mental health care at higher rates than their heterosexual counterparts. The current findings are also consistent with the past state-level research, which also found that sexual minority individuals utilize mental health care at higher rates (Grella et al., 2009; 2011). However, the current findings build on that work because the sample is representative of the U.S. population instead of one region or state, providing a more comprehensive understanding of the sexual minority population.

Mays and Cochran (2001) originally theorized that higher utilization rates could likely stem from the need for support and the mental health difficulties that come from the significant minority stress, oppression, and stigma sexual minority individuals face. Despite improvements in public attitudes and a reduction in the stigma associated with being a sexual minority over the last two decades, this population still accesses mental health professionals at higher rates than their heterosexual counterparts. Although impossible to conclude definitively from the current analyses, possible explanations for higher utilization rates from prior research include topics such as family and relationship difficulties (Bradford et al., 1994; Sorensen & Roberts, 1997), positive community norms regarding therapy (Bradford et al., 1994), and issues related to sexual orientation other than mental health (Hughes, Haas, & Avery, 1997). It is important to highlight that the sexual minority effect on utilization remained even after accounting for two measures of mental health distress and mental health limitations. This suggests that sexual minority individuals have needs from mental health professionals that are not fully accounted for by such measures. Working to more fully understand the complex set of factors involved in these utilization rates will be important in future research.

The second research question examined gender differences in sexual minority utilization of mental health professionals. The current analysis showed similar rates of utilization between sexual minority men and women. However, given that there is a gender gap of utilization among heterosexuals, this finding means that the gap between sexual minority populations and heterosexuals is smaller among women than among men. This is seen in the significant interaction effects in Table 2. In other words, the sexual minority effect on utilization is larger among sexual minority men. In practice, this means sexual minority men are not only seeking services at higher rates than heterosexual men, but they are doing so at a rate that closes the gender gap seen in comparative heterosexual populations.

Minority stress may be a factor that explains why the gender gap closes among sexual minority individuals (Herek, 2000; Meyer, 1995). Sexual minority men have unique cultural experiences as the intersection of gender and sexual orientation is often particularly salient for this group. Traditional gender roles dictate that men show little intimacy, especially toward other men, out of fear of seeming gay or too feminine. There are many implicit cultural norms that enforce the taboo of homosexuality in men (Solebello & Elliott, 2011). Therefore, when a man does have same-sex attraction and identifies as a sexual minority, he may face considerable social backlash, sometimes even violent reactions (D’haese, Dewaele, & Van Houtte, 2016). This minority stress may be a contributing factor in seeking services, especially for sexual minority men (Herek & Garnets, 2007). For example, Meyer (1995) found minority stress in gay men was predictive of mental health distress in a large community sample in New York City. This additional need may increase the utilization of sexual minority men so that it not only exceeds heterosexual men, but also rises to the level of use by sexual minority women.

Another explanation could be that sexual minority men do not experience the same cultural pressure and gender norms against seeking mental health services that heterosexual men experience. Given that their sexual orientation already violates the norms of masculinity in some ways (Solebello & Elliott, 2011), the additional stigma to seeking mental health services may be less relevant. This explanation assumes that there should be equal utilization between heterosexual men and women and that the gap is a matter of heterosexual men suppressing their utilization for gender role conformity.

However, other prior research on rates of utilization of mental health services by sexual minority men is rather limited. Although there have been numerous national studies on women (lesbians) (Bradford et al., 1994; Koh & Ross, 2013; Sorensen & Roberts, 1997), the knowledge on men (gay, bisexual, other) is less robust. While it is well documented that heterosexual men seek mental health services at lower rates as compared to heterosexual women (Chandra & Minkovitz, 2006; Kessler et al., 2005; Leong & Zacher, 1999), the research on mental health utilization by sexual minority men needs more attention in future research to help better understand the current findings.

Limitations and future directions

As with any data analysis, particularly secondary data, there are several important limitations to consider. First, the current analyses and results are based on the limitations of how the original questions were asked of participants, particularly the mental health and utilization measures. Although the prompts provided offer valuable information, they do not allow participants to expand on answers or allow for comprehensively understanding these findings. The literature has recognized the need for more qualitative research into sexual minority counseling and calls for greater comprehension of this community’s particular needs as well (Singh & Shelton, 2011). It would be worthwhile to explore the use of services among sexual minority individuals further, to gain a better understanding of the potential causal or correlational factors that help explain the current results. There are also limitations in regard to how the NHIS survey defines sexual orientation categories. For the purposes of these analyses, all non-heterosexual respondents were included as sexual minority individuals. This is the commonly accepted definition in most lesbian, gay, bisexual, transgender (LGBT) research. It is acknowledged that this categorization may not fully capture the range of sexual behaviors and labels that may exist among these populations.

Relatedly, the NHIS questions did not inquire into the participant’s experience with the quality or quantity of services they are accessing. The prompt asks only a dichotomous yes/no question if the respondent has seen a mental health provider in the last 12 months. With these data it is not possible to know the quality and quantity of services sexual minority individuals are accessing. It would be beneficial in future research to investigate this question further. Since sexual minority clients, especially men, are utilizing services at a high rate, it will be important to assess if clients are experiencing multiculturally competent care in response to their cultural needs. Past research, using smaller convenience samples, has shown such competence is often lacking with serious negative consequences for treatment outcome (Pachankis & Goldfried, 2004). If similar findings are found in larger, representative samples, there could be many implications for training of clinicians who provide mental health care.

Even with the limitations of the questions asked, the implications of the current findings point to a need for training and awareness of sexually diverse clients, particularly among sexual minority men. Mental health providers should also be sensitive to the unique cultural factors sexual minority clients are facing and be able to address those in the counseling setting. Multicultural competence in clinical work has had more emphasis in training in the last decade, with more attention in recent years on the needs of sexually diverse clients (Bieschke et al., 2007). The current finding that sexual minority individuals seek mental health care at significantly higher rates demonstrates this emphasis on multicultural competence in training and practice is still quite important, as this client population is accessing services often. A clinician should expect that a portion of the average caseload is going to have sexual minority individuals, and should be prepared to work effectively with issues of minority stress, family concerns, and other culturally specific concerns that may be present (Sue & Sue, 2016). Additionally, unlike racial/ethnic background, sexual orientation is not necessarily a visible identity and can at times be fluid. Clinicians must recognize that there could be clients on a caseload that are still exploring their sexual identity, may not initially disclose their sexual orientation, or are presenting with less visible, but still pressing, cultural concerns. In short, the high utilization rates in the current study emphasize the need for clinicians to be multiculturally competent with this population.

In sum, the findings of the current analyses indicate sexual minority individuals utilize mental health care at higher rates than heterosexual individuals. Although there is a well-known gender gap in the utilization of mental health professionals among heterosexuals, this gap does not exist for sexual minority individuals. This also means that the utilization gap between heterosexual men and sexual minority men is larger than the gap between heterosexual women and sexual minority women. Given this, the mental health fields should continue to strive to provide the best quality of care for sexual minority clients and seek to understand the factors that drive high utilization rates.