Prevalence of private and public health insurance among transgender and gender diverse adults

Abstract

Background

Little is known about how health insurance payer types differ between transgender and gender diverse (TGD) people and cisgender people. Much of what is known about insurance coverage among TGD adults has been based on research from claims and electronic health record data, which excludes individuals who have not accessed gender-affirming care. Research designed to understand how TGD populations pay for healthcare to best inform care interventions and public insurance policies is lacking.

Objectives

To examine differences in prevalence of public and private health insurance between transgender and cisgender adults.

Methods

Using data from the Behavioral Risk Factor Surveillance System, this study estimated prevalence of health insurance coverage among TGD and cisgender adults residing in 22 states that administered the Sexual Orientation and Gender Identity module and the Healthcare Access module from 2014 to 2019. This study estimated the odds of health insurance coverage (no insurance, private insurance, public insurance) among cisgender adults compared to TGD adults.

Results

TGD people had greater odds of being uninsured, compared to cisgender women. Among non-disabled, non-elderly respondents, TGD adults had lower odds of having private insurance and higher odds of public insurance compared to cisgender men. Among respondents who were likely Medicaid eligible, TGD respondents had lower odds of having public insurance and higher odds of being uninsured compared to cisgender women.

Conclusions

These findings provide foundational information about the payer mix among TGD people and provide insight into barriers to health insurance that TGD adults may face.

Introduction

Transgender and gender diverse (TGD) populations have poorer health compared to cisgender populations.¹ This may be exacerbated by lower rates of health insurance compared to cisgender people.² Adequate insurance coverage is critical to addressing health inequities among TGD populations,³ yet it is unknown whether differences in rates of uninsurance, or being uninsured, are driven by inequitable access to public, private, or both types of insurance.

TGD populations may have higher rates of uninsurance due to reduced access to employer-sponsored insurance (ESI). TGD adults face wide-spread employment discrimination, limiting access to jobs that provide ESI.⁴ Furthermore, TGD adults, who are more likely than cisgender adults to identify as sexual minorities and be in same-sex relationships, may have had less access to ESI prior to marriage equality in 2015.² While marriage equality had a substantial positive impact on ESI,⁵ disparities remain.⁶

Differences in uninsurance rates between TGD and cisgender adults may also be attributed to lower rates of enrollment in public insurance (including Medicare and Medicaid) among TGD adults. While recent evidence shows an increase in gender-affirming surgeries covered by Medicaid or Medicare,⁷ few studies have examined prevalence of public insurance among the broader TGD population (including those who have not accessed hormone therapy, surgeries, and other gender-affirming care). Despite experiencing high rates of poverty and disability,⁶ increasing Medicaid and Medicare eligibility, access to public insurance may be limited for TGD adults. Until May 2021, TGD adults were unprotected from discrimination based on gender identity by federal programs.⁸ The absence of protections for TGD adults may have resulted in reduced enrollment in public insurance.

Knowledge of which types of insurance TGD adults are enrolled in may help to explain why some TGD adults face inconsistent coverage for gender-affirming care. For example, ESI coverage of gender-affirming care is increasing³, yet TGD adults are less likely to have full-time employment than cisgender adults². Medicaid programs cover gender-affirming hormones in 34 states, and gender-affirming surgeries in 25 states,⁹ while Medicare covers gender-affirming care on a case-by-case basis.¹⁰

To reduce health inequalities and improve access to care, it is important to understand how TGD adults pay for health care. Limited research has explored differences in public and private insurance coverage between cisgender and TGD adults, irrespective of their gender-affirming care utilization. We address this using representative survey data to describe the payer mix among TGD individuals and examine differences in insurance coverage between TGD and cisgender adults.

This study hypothesized that TGD-identified adults will be more likely to be uninsured, more likely to have public insurance, and less likely to have private insurance compared to cisgender adults.

Methods

This study used data from the Behavioral Risk Factor Surveillance System (BRFSS) 2014, 2016, 2017, 2018, and 2019 surveys downloaded in December of 2020. The 2015 survey did not include insurance information and was therefore excluded. Individual states can elect to include specific modules, including gender identity and health insurance coverage modules. Data were collected from TGD-identified and cisgender adults residing in 22 states that included the Sexual Orientation and Gender Identity module and the Healthcare Access module during the study period (see Table, Supplemental Digital Content 1 for included states).

Measures.

Gender identity.

Respondents were asked: Do you consider yourself to be transgender? Those who responded yes were categorized as TGD-identified. Respondents who responded no were considered cisgender women if female sex, and cisgender men if male sex. Respondents were excluded if they were unsure or refused (n = 3,573).

Primary health insurance.

Respondents were asked: What is the primary source of your health care coverage? Those who had “a plan purchased through an employer or union,” “a plan that you or another family member buys on your own,” or “TRICARE, VA, or Military” were considered to have private insurance. Those with Medicare, Medicaid or other state programs were considered to have public insurance. Those with no coverage or Tribal Health Services were considered uninsured ¹¹.

Covariates.

Demographics included age; race/ethnicity (non-Hispanic white, non-Hispanic black, Hispanic, other/multiple); educational attainment (did not complete high school, high school graduate, some college/technical school, college/technical school graduate); percent of the federal poverty line (FPL) (calculated based on year and state thresholds using median household income and number of adults and children in the household); employment status (employed, unemployed, not looking, retired, unable to work); marital status; presence of children in the household; and residence in a Medicaid expansion state. Disability status was assessed based on the presence of at least one functional disability including disability in mobility, cognition, independent living, vision, or self-care.²

Statistical Analysis

Data were pooled and reweighted to be representative of each state in the sample. Survey weights in each year were multiplied by the number of all respondents that year divided by the number of respondents surveyed from 2014–2019.

The weighted prevalence of demographic and insurance variables was estimated for each group (TGD-identified, cisgender women, cisgender men); Pearson chi-square tests were used to compare weighted group frequencies.

Three multivariable-adjusted logit models assessed the relationship between insurance outcome (no insurance, private insurance, public insurance) and gender identity (cisgender women and men compared to TGD-identified respondents) controlling for all covariates listed above. To understand how insurance varied among subgroups, these same three models were run among three different subsets of respondents: 1) likely not Medicare-eligible (non-elderly, non-disabled); likely Medicare-eligible (65+ or at least 1 disability); and likely Medicaid-eligible (under 250% of the FPL). Analyses were conducted using StataMP 15 in December of 2020.

Results

Descriptive.

The sample consisted of 256,136 respondents (0.4% TGD-identified adults, 42.2% cisgender women, 57.4% cisgender men) (Table 1). Compared to cisgender women and men, a lower proportion of TGD-identified respondents were 65+ and retired, while a higher proportion were disabled, unable to work, and under 250% of the FPL. Employment was higher among cisgender men, compared to both cisgender women and TGD-identified respondents. A lower proportion of TGD-identified adults were married compared to cisgender respondents, yet a similar proportion of TGD-identified adults and cisgender men had a child in their household. All groups had similar rates of residence in a Medicaid expansion state.

Table 1.

Characteristics of BRFSS Respondents from 2014–2019 by Gender Identity

	Transgender Individuals		Cisgender Men		Cisgender Women
	n =1,106		n = 108,097		n = 146,933
	n	%	n	%	n	%
Age Group1
18–24	114	10.3	7033	6.5	6053	4.1
25–34	112	10.1	10899	10.1	12462	8.5
35–44	130	11.8	12635	11.7	16374	11.1
45–54	183	16.5	18229	16.9	23851	16.2
55–64	253	22.9	24517	22.7	33172	22.6
65+	314	28.4	34784	32.2	55021	37.4
Non-Hispanic white1	788	72.8	86833	81.7	117232	80.8
Employment Status 1
Employed	484	43.8	48122	56.0	64700	44.1
Unemployed	78	7.1	4786	4.4	5920	4.1
Retired	284	25.7	31555	29.2	46612	31.7
Unable to work	153	13.8	7553	7.0	12955	8.8
Disabled1	434	39.2	25518	23.6	44492	30.3
Married1	489	44.4	61244	56.9	71905	49.2
Child in Household1	240	21.9	26696	24.9	38292	26.2
Under 250% FPL1	522	58.2	33571	36.8	55196	46.3
Expansion state2	600	54.2	59615	55.1	81271	55.3

^1.Pearson chi-square test: p<0.001

^2.Pearson chi-square test: p=0.577

Prevalence of primary health insurance (unadjusted).

TGD-identified adults had the highest prevalence of uninsurance compared to cisgender men and women (23.5%, 16.1%, and 12.8% respectively); this finding was consistent across all subgroups (non-elderly/ non-disabled, likely Medicare-eligible, and likely Medicaid-eligible) (Table 2). The prevalence of public insurance among TGD-identified adults (24.6%) was higher than cisgender men (21.8%) but lower than cisgender women (29.3%).

Table 2.

Prevalence of Primary Health Insurance Type by Gender Identity

	Transgender Individuals		Cisgender Men		Cisgender Women
All Adults	n = 1,106		n = 108,097		n = 146,933
	n	%	n	%	n	%
No insurance	192	23.3 (18.5, 28.8)	12959	16.1 (15.6, 16.6)	14198	12.8 (12.4, 13.2)
Private insurance	500	48.7 (42.8, 54.6)	60895	59.5 (58.9, 60.1)	74988	55.7 (55.2, 56.2)
Public insurance	383	24.6 (20.0, 29.9)	32146	21.8 (21.4, 22.3)	55119	29.3 (28.8, 29.7)
Non-Disabled, Non-Elderly	n = 496		n = 58,534		n = 68,439
	n	%	n	%	n	%
No insurance	99	26.0 (19.1, 34.2)	8,411	17.4 (16.7, 18.0)	8,060	14.0 (13.5, 14.6)
Private insurance	317	57.8 (49.1, 66.1)	45,099	73.1 (72.4, 73.8)	52,652	72.7 (72.0, 73.4)
Public insurance	67	12.4 (7.1, 21.1)	3,958	7.1 (6.7, 7.6)	6,913	11.4 (10.9, 11.9)
Medicare Eligible	n = 610		n = 49,563		n =78,494
	n	%	n	%	n	%
No insurance	93	19.8 (14.0, 27.1)	4,548	13.6 (12.9, 14.4)	6,138	11.1 (10.6,11.6)
Private insurance	183	36.7 (29.1, 45.2)	15,796	34.2 (33.4, 35.1)	22,336	31.4 (30.7, 32.1)
Public insurance	316	40.4 (33.3, 47.9)	28,188	49.3 (48.4, 50.2)	48,205	54.7 (54.0, 55.4)
Under 250% FPL	n = 522		n = 33,571		n = 55,196
	n	%	n	%	n	%
No insurance	108	28.4 (21.1, 36.9)	6,793	26.1 (25.1, 27.1)	8,378	19.4 (18.6, 20.1)
Private insurance	175	38.5 (29.9, 47.9)	13,023	41.6 (40.5, 42.6)	18,188	37.7 (36.8, 38.6)
Public insurance	228	31.9 (24.3, 39.9)	13,016	29.8 (28.8, 30.7)	27,598	40.9 (40.0, 41.7)

Note: Unknown/refused categories are not included due to small cell sizes; thus number do not sum to total n. Data are weighted.

Differences in odds of primary health insurance.

Compared to TGD-identified adults, cisgender women had lower odds of uninsurance (AOR= 0.703, 95% CI= 0.580, 0.853) and cisgender men had marginally higher odds of having private insurance (AOR= 1.143, 95% CI= 0.975, 1.341) (Table 3). No differences were observed for public insurance. Differences in private insurance were assessed among those not likely Medicare insurance eligible. Compared to TGD-identified adults, cisgender women (AOR= 1.382, 95% CI= 1.081, 1.766), and cisgender men (AOR= 1.303, 95% CI = 1.019, 1.665) had higher odds of having private insurance. Third, the study looked at differences in public insurance among likely Medicare-eligible respondents and no differences were observed. Finally, differences in public insurance among those likely Medicaid-eligible (under 25% FPL) were estimated. Compared to TGD-identified adults, cisgender women had higher odds (AOR= 1.221, 95% CI= 1.006, 1.481), and cisgender men had marginally lower odds, of having public insurance (AOR= 0.840, 95% CI = 0.692, 1.019) of having public insurance.

Table 3.

Odds of insurance outcomes among cisgender women and cisgender men relative to TGD individuals

		No Insurance		Private		Public
		OR	95% CI	OR	95% CI	OR	95% CI
All1	Cisgender women	0.703**	0.580,0.853	1.122	0.957,1.315	1.095	0.922,1.299
	Cisgender men	0.924	0.762,1.121	1.143+	0.975,1.341	0.901	0.759,1.070
Non-Disabled, Non-Elderly2	Cisgender women	0.668**	0.509,0.876	1.382**	1.081,1.766	1.069	0.769,1.487
	Cisgender men	0.897	0.684,1.177	1.303*	1.019,1.665	0.731+	0.525,1.018
Medicare Eligible3	Cisgender women	0.748*	0.567,0.987	0.961	0.783,1.180	1.138	0.937,1.382
	Cisgender men	0.945	0.716,1.248	1.058	0.861,1.299	0.964	0.793,1.171
Under 250% FPL4	Cisgender women	0.794*	0.633,0.995	0.943	0.777,1.145	1.221*	1.006,1.481
	Cisgender men	1.033	0.824,1.295	1.124	0.926,1.365	0.840+	0.692,1.019

^**p< 0.01

^*p<0.05

⁺p<0.10;

reference group is TGD adults; OR (odds ratio) and CI (confidence interval)

¹Adjusted model controls for age, race, ethnicity, 1+ child in the house, any disability, under 250% FPL, residence in Medicaid expansion state, state, year

²Adjusted model controls for age, race, ethnicity, 1+ child in the house, under 250% FPL, residence in Medicaid expansion state, state, year

³Adjusted model controls for race, ethnicity, 1+ child in the house, any disability, under 250% FPL, residence in Medicaid expansion state, state, year

⁴Adjusted model controls for age, race, ethnicity, 1+ child in the house, any disability, residence in Medicaid expansion state, state, year

Discussion

This study is among the first to use a population-based sample to compare the payer mix for insurance coverage between TGD and cisgender adults. Importantly, this study includes TGD adults who have accessed gender-affirming healthcare, through insurance and those who have not.

Consistent with our hypotheses and supported by previous studies, TGD-identified adults were more likely to be uninsured compared to cisgender adults.² 11 TGD-identified adults have less access to resources and socioeconomic advantage,² likely creating barriers to insurance, particularly in states with more limited Medicaid eligibility. Without insurance coverage, gender-affirming care, and other health services, may be prohibitively expensive for many TGD adults.

Contrary to the study hypothesis, analyses adjusted for socioeconomic confounders showed that TGD-identified adults and cisgender men had similar odds of uninsurance. This is not surprising considering prior research that found that men are more likely to be uninsured than women¹³. Additionally, Medicaid eligibility via pregnancy may create more insurance opportunities for cisgender women. Future research should explore why TGD adults, despite additional, unique health needs,² face similar uninsurance rates to cisgender men, and more importantly, the specific barriers to insurance access for TGD adults.

This study also found that TGD-identified adults were less likely to have private insurance, among non-elderly, non-disabled adults. ESI has expanded to cover gender-affirming care,³ however, access to ESI remains restricted for TGD adults due to employment inequities and the previous absence of federal policies banning health insurance discrimination based on gender identity among private insurers. Many states do have laws in place banning private insurers from gender identity-based discrimination, however, even in these states some insurers are exempt from these laws.¹⁴ Previous work shows that TGD adults have lower rates of college education, and higher rates of poverty and workplace discrimination, all of which are tied to employment.⁶ Unit 2020, no federal policies protected TGD adults from workplace discrimination, though a patchwork of laws did exist in some states.¹⁴ Additionally, TGD adults may be less likely to have access to ESI through their partner if they are in a same-sex relationship, though marriage equality played an important role in expanding access to ESI for same-sex couples.⁵

Differences in private insurance for TGD adults may also result from lack of clarity surrounding access, as many TGD adults are uncertain about gender-affirming healthcare coverage.¹⁵ Future studies should examine whether uptake of private insurance among eligible TGD workers is hindered by lack of clear coverage guidelines. Examining access to coverage information may be a tangible way to begin addressing specific barriers for TGD adults seeking gender-affirming care.

Additionally, public insurance was similar across likely Medicare-eligible adults. However, TGD-identified respondents were more likely to be uninsured than cisgender women in this population. This is concerning considering TGD adults have higher rates of disability.^.16 Minority stress and systemic oppression can impact the health of TGD people¹⁷ and many TGD adults could benefit from Medicare coverage. It is important to explore factors that may prevent TGD adults with disabilities from accessing Medicare coverage. Navigating complex healthcare systems can be difficult for people with disabilities. This may be additionally challenging for disabled TGD adults.¹⁸

Finally, TGD-identified respondents who were Medicaid-eligible were less likely to have public insurance than cisgender women, but levels similar to cisgender men. This may be due, in part, to lower rates of children in the household that afford many cisgender women access to Medicaid ², or because less than half of state Medicaid programs cover gender-affirming care, impacting who enrolls in Medicaid programs.⁹ Future research should examine uptake of Medicaid among those eligible to understand TGD-specific barriers.

Limitations

In this study, the sample consisted of participants from only 22 states, potentially reducing generalizability. Second, small sample sizes prevented separation of results beyond private and public insurance to examine outcomes for Medicaid, Medicare, and ESI separately. Third, the study likely undercounts gender diverse respondents without utilizing the two-question gender identity assessment method¹⁹ and was unable to separate results by gender identity or other intersecting identities (including race) among TGD-identified participants due to small sample sizes.^{12 19} Fourth, use of probabilistic samples to study TGD populations before 2016 is controversial due to sex-specific ranking and the way sex was collected. ^{19 20} However, most of our data was collected after 2016. Finally, this study does not explicitly test which explanatory factors contribute to the observed differences in coverage.

Conclusion

This study provides some of the first representative estimates of prevalence of private and public insurance among TGD adults. These findings provide foundational information about the payer mix of the broader TGD population and can inform an understanding of TGD-specific barriers to insurance.

Supplementary Material

Supplemental Data File (.doc, .tif, pdf, etc.)

Table that provides information on each of the states included in this study.