Individuals who identify as sexual and gender minorities are at elevated risk of a range of mental health disparities, including higher rates of mental illness symptoms and mental health care needs (
1,
2). Individuals who identify as transgender and gender diverse (TGD) (neither male nor female) in particular have been found to have elevated frequency and intensity of suicidal thoughts and behaviors as well as symptoms of depression and anxiety (
3–
6). People who identify as TGD are significantly more likely to have unmet mental health needs compared with people who identify as heterosexual and people whose gender identity and birth sex align (cisgender) (
2,
7), although more than 90% have reported wanting to receive or are currently receiving mental health counseling (
8). Leading models of sexual- and gender-minority health disparities emphasize the ways in which structural factors such as laws and policies can contribute to disparities in mental health symptoms and related care (
9,
10).
Inequalities in health insurance coverage are a major structural contributor to disparities in health symptoms and treatment (
11,
12). People who identify as TGD have higher rates of being uninsured (14%–19%) than the general public (11%–17%;
8,
13,
14), which, in turn, is associated with disparities in access to care (
15). Individuals who identify as TGD have also qualitatively described the lack of health insurance as a notable barrier to utilizing mental health care (
16,
17).
Even among insured individuals, the type of health insurance may influence utilization of care. Among the general population, public health insurance such as Medicaid has historically been associated with increased access to mental health services compared with private insurance (
12). However, individuals who identify as TGD may diverge from this pattern. Among persons who identify as TGD, those with public health insurance have reported more frequent occurrences of discrimination by medical providers and postponing care because of costs compared with those with private insurance (
8). Among the privately insured, there may also be differences in access to mental health services depending on the insurance type. Individuals with privately purchased insurance are more likely to be underinsured compared with those with employer-based health care (
18).
To date, few studies have assessed associations between health insurance and health disparities with TGD. Findings primarily established that individuals who identify as TGD are more likely to be uninsured than cisgender individuals and the general population (
8,
13–
15). However, these studies were not able to indicate how insurance influenced subsequent utilization of care. In the largest survey of TGD experiences (N=27,715), the respondents again had higher rates of being uninsured than the general public (
13). Additionally, treatment utilization varied by insurance type. For instance, individuals with Medicare were significantly less likely to be denied gender-affirming care (i.e., hormone therapy, surgery) than those with other insurance types, suggesting that not all insurance types offer equitable access to care. Despite this disparity in health insurance coverage, no known research has assessed how insurance coverage may influence disparities in mental health treatment among persons who identify as TGD.
The current study sought to better understand the role of health insurance in mental health care utilization among individuals who identify as TGD. Given that previous research has shown disparities in health care access among those without insurance, it was hypothesized that uninsured individuals would have a higher likelihood of having never seen a therapist or psychiatric provider than those with health insurance. Exploratory analyses then considered how mental health care utilization differed among individuals on the basis of insurance type.
Insurance type has often been grouped on the basis of whether the insurer is private or public (
12). However, the Patient Protection and Affordable Care Act of 2010 (ACA) has led to an increase in the proportion of individuals who receive privately purchased insurance, which often provides significantly less coverage than employer-based insurance (
18). Thus, employer-based and privately purchased insurance were assessed separately. Additionally, given that individuals who identify as TGD are overrepresented in military and veteran populations and are receiving care from the Veterans Health Administration (VHA) at rapidly increasing rates, we assessed public and military/VHA insurance separately (
19). Thus differences in mental health treatment utilization were assessed on the basis of whether an individual was insured through employer, privately purchased, public, military, VHA, or other insurance.
Methods
Participants and Procedures
The current study utilized data collected in the 2017 Trans Lifeline Mental Health Survey, a collaboration between the National Lesbian, Gay, Bisexual, Transgender, and Queer (LGBTQ) Task Force and the Trans Lifeline (
20). Participants were primarily recruited online via listservs and social media to complete an online survey. Individuals were eligible if they self-identified as TGD, were older than age 18, and lived in the United States. Participants totaled 4,350 individuals who varied in their gender identity: 24% (N=1,027) masculine, 33% (N=1,442) feminine, 36% (N=1,552) nonbinary, and 6% (N=269) unspecified. Most participants self-identified as white (N=3,237, 74%). Although the rest of the sample identified themselves across more than 20 racial-ethnic minority groups, broadly speaking, 10% (N=421) identified as Hispanic (e.g., Latinx), 7% (N=322) as Native American (e.g., indigenous, American Indian), 6% (N=246) as Asian (e.g., Korean, Chinese), 6% (N=270) as multiracial (e.g., biracial, Afro-Latinx), and 5% (N=201) as black (e.g., African American, Caribbean). The most common education level was a high school degree (41%, N=1,781), with 5% (N=202) reporting less than a high school degree; 10% (N=416), an associate or technical school degree; 29% (N=1,261), a bachelor’s degree; and 16% (N=681), a graduate degree. When asked to identify all employment descriptions that applied, most participants were working at least part-time (64%, N=2,780), 11% (N= 478) were unemployed and actively looking, 24% (N=1,073) were students, 8% (N= 367) were unable to work, and 6% (N=253) were engaged in unpaid work (e.g., homemaker, retired). Participants had a mean±SD age of 32.36±13.29. Sixteen participants did not endorse an insurance category, resulting in a sample of 4,334 participants for the current study. Use of the survey data was approved as an exempt study by the University of Kentucky review board. Additional information about the sample can be found in a previous publication (
20).
Measures
Sociodemographic information included gender identity, ethnic identity, yearly household income, disability identity, and sexual orientation. Questions about gender identity and sexual orientation were open ended, with the responses qualitatively coded. Gender identity was coded as masculine expression, feminine expression, nonbinary expression, or unspecified. Sexual orientation was coded as heterosexual, gay, androphilic, gynephilic, bisexual or pansexual, asexual or aromantic spectrum, or an additional or unspecified category. Racial-ethnic identity was assessed with 29 possible options (e.g., black, Afro Latinx, indigenous, Japanese, Korean); participants were able to select all that applied. Because the sample sizes in the racial-ethnic categories were not sufficient to permit analysis on the basis of racial-ethnic identity, the variable was dichotomized such that if any non-white identity was selected, the participant was categorized as a person of color.
Insurance status was a self-report question with 10 possible options that were combined into six categories: uninsured, employer insurance, purchased insurance, public insurance, military or VHA insurance, and other insurance. Insurance through a current or former employer, or someone else’s employer, was grouped under employer insurance; insurance purchased by the participant or family member and student insurance were grouped under purchased insurance; Medicare, Medicaid, state public insurance, and other public insurance were grouped under public insurance.
Mental health care utilization was assessed with two dichotomous questions. Participants were asked whether they currently or in the past had a therapist (e.g., psychologist, social worker, counselor). A second question asked whether they currently or in the past had a psychiatric provider (e.g., medication management, such as a psychiatric provider, psychiatric nurse practitioner, psychiatric physician assistant).
Analytical Strategy
Chi-square tests assessed for differences in sociodemographic characteristics of individuals on the basis of insurance status; Bonferroni-corrected, post hoc comparisons compared sociodemographic characteristics among each of the six groups. Logistic regression analyses assessed the association of insurance status with lifetime utilization of a mental health therapist or psychiatric provider while adjusting for sociodemographic characteristics (gender identity, person of color status, household income, disability, sexual orientation). In all categories, the identity with the greatest societal privilege was considered the reference (e.g., no disability, highest income bracket).
Results
Overall, the majority of the sample reported currently or previously seeing a therapist (89%, N=3,266) or psychiatric provider (52%, N=2,255). Additionally, the majority of the sample reported being insured through an employer, either their own or someone else’s (53%). The next most common insurance type was public insurance (23%), followed by purchased insurance (13%), military or VHA insurance (2%), and other insurance (1%). Of the sample, 8% reported having no insurance. Uninsured individuals were significantly more likely than those with employer-based insurance to be a person of color and to have an income of less than or equal to $19,999 (
Table 1). Compared with all other insurance types, uninsured individuals were also less likely to identify as gay and more likely to identify as bisexual/pansexual or heterosexual. Those who were uninsured did not significantly differ from those with purchased insurance on the majority of demographic factors.
The crude prevalence rates of seeing a therapist and psychiatric provider were lowest among the uninsured individuals. In logistic regressions adjusting for sociodemographic information, individuals without insurance were significantly less likely than those insured by an employer to have seen a therapist or psychiatric provider (
Table 2).
Crude rates of having seen a therapist or psychiatric provider significantly differed among all insurance types, with the highest rates among those with military or VHA insurance (
Table 1). Those with military or VHA insurance were also significantly more likely to have feminine gender expression and were least likely to identify as nonbinary or as a person of color (
Table 1). The majority of individuals with military or VHA insurance reported having seen a therapist in their lifetime (69%), and nearly all reported having seen a psychiatric provider (95%). In logistic regressions adjusting for demographic factors, there were no significant differences in rates of having seen a therapist or psychiatric provider among individuals with employer-provided, purchased, or public insurance (
Table 2). These analyses showed that only individuals with insurance through the military or VHA had significantly greater odds than individuals with employer insurance of seeing a psychiatric provider. Individuals with insurance through the military or VHA were 2.18 times more likely to have seen a psychiatric provider than those with employer insurance.
Discussion
The current study assessed the association between health insurance and lifetime mental health care utilization among individuals who identified as TGD. In line with previous studies, the majority of individuals had health insurance, with most reporting employer or public health insurance (
13). Notably, only 8% of the sample were uninsured, which is substantially lower than previous studies that found that 13%–19% of individuals who identified as TGD were uninsured (
8,
13–
15). Given that the current data were collected in 2017, which is 2 to 10 years after previous studies assessed insurance status (
8,
13–
15), the lower rates of being uninsured may reflect an actual reduction in the number of uninsured individuals who identify as TGD. Supporting this point, the percentage of individuals in this sample who did not have insurance mirrors the most recent data on uninsurance rates in the broader U.S. population (9%), which has seen a significant reduction since the enactment of the ACA (
21). Alternatively, it should be considered that the current sample may be distinct from the larger studies that identified higher rates of having no insurance (
8,
14). The current study relied on Internet-based recruitment, which may have unintentionally been a barrier for those more likely to be uninsured. For instance, someone experiencing homelessness may be both uninsured and unable to complete the survey because of a lack of Internet access.
Demographic differences among individuals who identified as TGD emerged on the basis of insurance. Unsurprisingly, persons who were uninsured or had public insurance had the highest rates of incomes of less than or equal to $19,999. Notably, rates of identifying as a person of color were highest among those who were uninsured or utilizing public, purchased, or other insurance. By comparison, there were significantly fewer people of color among those with employer- or military- or VHA-provided insurance, which were the only insurance types to be provided directly through employment. Given that unemployment is elevated among people of color who identify as TGD (
22), this finding may point toward unemployment as a potential driver in disparities in health insurance coverage among people of color who identify as TGD. Future research would benefit from a more thorough assessment of how engagement with employment influences mental health utilization and relates to intersecting identities.
Health insurance had a notable association with mental health care utilization among individuals who identified as TGD. Individuals with no insurance had significantly lower rates of seeing a therapist or psychiatric provider in their lifetime compared with individuals with all other insurance types. This disparity persisted even when adjusting for demographic factors; the odds of seeing a therapist (AOR=0.38) or psychiatric provider (AOR=0.48) were significantly less for those with no insurance compared with those with employer-based insurance. These findings are consistent with broader research suggesting that being uninsured is a major barrier to receiving mental health care (
11). Notably, a disparity in mental health care utilization may have unique consequences for individuals who identify as TGD. Seeing a mental health professional is often a requirement for receiving gender-affirming medical treatments such as hormone-replacement treatment or gender-affirming surgeries (
23). Thus, uninsured individuals who have lower rates of mental health care utilization may have additional barriers to receiving appropriate gender-affirming medical care, perpetuating high-risk practices such as taking nonprescribed hormones and self-performed surgeries (
24,
25).
Striking differences emerged between individuals receiving military or VHA insurance and those receiving all other insurance types. Individuals with military or VHA insurance were more likely than those with any other insurance status to have reported seeing a therapist (69% compared with 35%–59%) or a psychiatric provider (95% compared with 74%–91%) in their lifetime. Even after adjusting for demographic factors, those with military or VHA insurance were twice as likely as those with employer health insurance to have seen a psychiatric provider. Although those with military or VHA insurance were also 2.47 times more likely than those with employer insurance to have seen a therapist, the finding was not significant. Given the relatively small number of individuals in the sample with military or VHA insurance, the analysis may have been underpowered. Taken together, findings point toward extraordinary utilization of mental health care among the individuals who were currently or had previously served in the military. This finding is particularly notable given recent policies disallowing transgender people to serve in the military (
26), despite the fact that transgender individuals are two times more likely than cisgender individuals to serve in the military (
27).
Although military and VHA insurance were assessed with a single variable, experiences likely differed between active duty service members and discharged veterans. Although many service members who identify as TGD receive support from command and fellow service members, there are also high rates of military sexual trauma, stigmatization, and forced discharge that may influence higher rates of mental health care utilization (
27–
29). Additionally, the recent reinstatement of policies that ban transgender individuals from serving in the military may exacerbate these concerns given the psychological distress, discrimination, and victimization that occurred among sexual-minority service members serving under the similar policy of “Don’t ask, don’t tell” (
29,
30).
For those who are no longer serving, higher rates of utilization may relate to the rapidly expanding availability of VHA services over the past decade for veterans who identify as TGD (
31). Veterans who identify as TGD have reported being treated respectfully by VHA providers and have described being highly satisfied with their care (
27,
32). It should also be considered that the higher rates of mental health care utilization may relate to structural differences in military and VHA mental health care as opposed to differences in distress among service members and veterans who identify as TGD. For instance, higher mental health care utilization may reflect the requirement that service members receive regular mental health evaluations to assess for deployment readiness.
Additionally, mental health utilization may simply be encouraged through VHA’s use of an integrated health care model, which has been associated with easier access and higher utilization of mental health services (
33). It will be critical to further assess the availability and quality of VHA care in the coming years for veterans who identify as TGD. The number of individuals who identify as TGD who are served by the VHA has dramatically risen in the past 15 years (
19). Rising rates may continue because recent policy changes banning TGD people from serving in the military may increase both the number of discharged service members who identify as TGD and the number who keep their identity secret until after discharge (
26).
Findings should be considered within the context of several limitations. First, mental health care utilization was measured as whether individuals ever saw a psychiatric provider or therapist at any point in their lifetime. As a result, we were unable to assess when care occurred. Participants may have seen a provider in the past when their insurance status was different than their current insurance status. Similarly, only current gender identity was assessed, which cannot speak to how mental health utilization may differ on the basis of gender-identity fluidity or a transition in gender expression.
Additionally, our measure of mental health care utilization cannot speak to the quality of care. Given that the current study found a striking increase compared with earlier studies in the number of individuals who identify as TGD and who are covered by health insurance, future research must consider the range of barriers that may interfere with receipt of adequate mental health care for insured individuals who identify as TGD, such as providers who are uninformed, are discriminatory, or deny treatment (
8,
17). Future research would benefit from a consideration of how insurance coverage may relate to the receipt of high-quality mental health care from informed providers.
Additionally, our data did not allow us to determine whether there were differences in the findings on the basis of intersecting marginalized identities. For instance, given that people of color had elevated rates of using public insurance, are there differences in mental health care utilization among people of color who identify as TGD and their white counterparts, even when both groups use public health care? It should also be considered that the current data were collected through a survey conducted by the Trans Lifeline, which may have attracted participants who were affected by mental health concerns.