Double jeopardy: Minority stress and the influence of transgender identity and race/ethnicity

Abstract

Background: Prior research suggests transgender individuals with multiple minority statuses experience higher psychological stress compared to their singly disadvantaged counterparts, and both Black, Indigenous, People of Color (BIPOC), and transgender minorities experience more frequent and severe forms of discrimination than White and cisgender individuals.

Aims: This study aims to examine racial/ethnic differences in gender-related discrimination and psychological distress within a sample of transgender individuals.

Methods: Using data from a convenience sample of 99 self-identified transgender adults recruited through North American LGBTQ organizations, data were analyzed to determine the relationship between race/ethnicity, gender minority stress, and psychological distress.

Results: When White and BIPOC participants are compared, no significant group differences were found in levels of gender discrimination or victimization. However, some individual racial/ethnic groups reported significantly higher or lower scores and results indicate that changes in reported gender minority stress are in fact positively correlated with reported psychological distress.

Conclusion: This research highlights that BIPOC are a heterogeneous group; by solely examining race/ethnicity as a binary variable, studies mask potential important differences among different groups.

Introduction

The most recent statistics indicate that approximately 1.4 million people in the United States identify as transgender—individuals who often face discrimination because their gender identity does not match their sex assigned at birth (Flores et al., 2016).¹ Scholars continuously highlight the importance of other identity markers, such as race and ethnicity,² when researching the oppression and stigmatization of the trans community (Choo & Ferree, 2010; Schilt & Lagos, 2017; Vidal-Ortiz, 2008). Race/ethnicity, like gender identity, is typically conveyed through an individual’s outward appearance and information revealed through interactions (Roth, 2016; Westbrook & Schilt, 2014).

Given their multiple minority identities, trans Black, Indigenous, People of Color (BIPOC) may experience greater gender discrimination than White trans individuals, receive more or less acceptance and affirmation from families and cultural communities, and face racism within the LGBTQ community (Levitt & Ippolito, 2014). Gender inequalities do not exist in a vacuum—gender discrimination may affect the mental health of individuals who fall outside of the normative conceptions of gender categories differently depending on their race/ethnicity. We seek to answer the call of Tan et al. (2020), guided by theories of intersectionality and minority stress, this research aims to (1) understand how race/ethnicity affects experiences of discrimination based on trans identity and (2) examine if gender minority stress is associated with mental health. This study approached these questions by testing for racial/ethnic differences in gender minority stress and psychological distress within a sample of trans and nonbinary individuals.

Gender, race, and minority stress

Minority stress

Meyer’s (1995, 2003) theory of minority stress posits that stress associated with being a member of a marginalized group negatively affects well-being. This paradigm contends that all individuals experience stressors, but it is when the stress becomes overwhelming that one experiences stress (Tan et al., 2020). Minority groups face an additional stressor, called “minority stress,” that is not experienced by dominant groups and thus contributes to disparities in mental and physical health outcomes across social groups (e.g., race, gender, class, sexuality) (Lee et al., 2020; Meyer, 1995, 2003; Tan et al., 2020). Examples of minority stressors can include microaggressions, discrimination, and a hostile work or school environment. Prior research on identity and social stress indicates discrimination is positively associated with poor mental health and psychological distress (Pascoe & Richman, 2009; Perry et al., 2013; Thoits, 2010; Turner et al., 1995).

For trans individuals, findings of previous studies consistently show that gender-related discrimination and victimization contribute to psychological distress, internalized transphobia, expectations of rejection, poor self-esteem, suicidality, and depression symptoms (Austin & Goodman, 2017; Barboza et al., 2016; Breslow et al., 2015; Downing & Przedworski, 2018; Glick et al., 2020; Jefferson et al., 2013; Lee et al., 2020; McLemore, 2018; Nuttbrock et al., 2014; Tebbe & Moradi, 2016; Testa et al., 2017; Timmins et al., 2017; White-Hughto et al., 2015). These outcomes may vary by race/ethnicity given that transgender people who have multiple marginalized identities are more likely to report negative mental health symptoms than those with more privileged identities (Lytle et al., 2016).

Transgender-related discrimination

Transgender-related discrimination stems from ideologies that insist gender identity aligns with sex, and that sex and gender are binary concepts. The trans population contains considerable diversity. Some trans-identified people identify within the binary categories of men and women; however, many trans-identified people identify outside of the gender binary, with gender identities such as nonbinary, genderqueer, or agender (Lefevor, Boyd-Rogers, et al., 2019). This variation in identities is accompanied by variation in gender presentation. Depending on an individual’s gender presentation, they may face more or less prejudice and discrimination (Gagné et al., 1997; West & Zimmerman, 1987; Westbrook & Schilt, 2014). When compared to traditional Western notions of gender, being transgender, endorsing a non-normative gender identity, and displaying a non-normative gender presentation are seen as deviant. Even though being trans is no longer inherently considered a mental disorder since the replacement of “Gender Identity Disorder” with the diagnosis of “Gender Dysphoria” in the Diagnostic and Statistical Manual-5, trans people continue to be regarded as deviant and “other” by much of Western social institutions, such as government agencies and health care (Valdiserri et al., 2019; White-Hughto et al., 2015).

The sociological, medical, psychological, and historical literatures indicate individuals with non-normative gender identities and expressions report high levels of gender discrimination including exposure to violence, sexual assault, and harassment (Bauerband et al., 2019; Bockting et al., 2013; Clements-Nolle et al., 2006; Kammer-Kerwick et al., 2019; Lombardi et al., 2001; Schulman & Erickson-Schroth, 2019; Stotzer, 2009; Testa et al., 2012), employment-related discrimination (Glick et al., 2020), microaggressions (Galupo et al., 2014; Nadal et al., 2012; 2016; Parr & Howe, 2019), and rejection from social support (Bradford et al., 2013; Erich, Tittsworth, & Kersten, 2010). For trans people, navigating discrimination is often a fundamental aspect of life as they experience gender discrimination in various realms including health care, employment, and social services (De Vries, 2012; Erich, Tittsworth, Colton-Meier, et al., 2010; Glick et al., 2020; Kattari & Hasche, 2016).

Mental health

According to minority stress theory, psychological distress and poor mental health are consequences of transgender discrimination (Lee et al., 2020). While studies agree transgender individuals as a group experience worse mental health relative to cisgender individuals, results of the few studies focused on racial/ethnic disparities in mental health within the trans population have been inconsistent. In some studies, White participants reported worse mental health than racial/ethnic minority participants, which contradicts the expected findings based on minority stress theory (Barboza et al., 2016; Nuttbrock et al., 2010). Nuttbrock et al. (2010) found White participants reported higher levels of depression than African American or Black participants, and Barboza et al. (2016) found that White trans people had a higher risk of suicide than trans BIPOC.

Conversely, some research on race/ethnicity and mental health of trans individuals has supported minority stress theory. Recent studies found multiracial participants were more likely to report attempting suicide than White participants (Lytle et al., 2016; Miller & Grollman, 2015), while African American or Black, Latino or Hispanic, and multiracial participants reported worse depressive distress than White participants (Katz-Wise et al., 2017). Most recently, Adams and Vincent (2019) found that First Nations individuals reported the highest rate of lifetime suicide attempts, and BIPOC as a group reported higher rates of lifetime suicide attempts than White participants. Furthermore, two studies of trans BIPOC found experiences with gender and racial discrimination were significantly and positively associated with depression symptoms (Jefferson et al., 2013; Lefevor, Janis, et al., 2019). However, other studies did not find significant differences in depressive symptoms (Pflum et al., 2015), psychological distress (Hatchel et al., 2019), self-harm, or suicidal ideation (Lytle et al., 2016) between racial/ethnic groups.

Intersectionality

Intersectionality theory is a feminist framework for conceptualizing one’s experience with privilege and oppression (Crenshaw, 1991). Intersectionality theory posits that systems of oppression intersect to shape the experiences of individuals as opposed to an individual experiencing privilege or oppression compartmentally through individual social statuses. Experiences with oppression and privilege related to race and gender are not just additive, but interconnected and multiplicative (Crenshaw, 1991; Parent et al., 2013; Warner & Shields, 2013). Gender and race are typically visible master statuses that strongly influence how others perceive and treat an individual (Hughes, 1945). For trans BIPOC, multiple marginalized identities may interact to form experiences with oppression, such as racialized transphobia (Bowleg, 2008, 2013; Glick et al., 2020; Lytle et al., 2016; Meyer, 2012).

Previous research findings tend to support arguments informed by intersectionality theory; BIPOC report greater discrimination (Bradford et al., 2013; Kattari & Hasche, 2016; Reisner et al., 2016), physical assault (Herman, 2013; Nuttbrock et al., 2010), and feelings of stigma (Bockting et al., 2013) than White trans individuals. Furthermore, LGBTQ individuals of color report racism within the LGBTQ community, and heterosexism and transphobia in some racial/ethnic communities (Levitt & Ippolito, 2014). Discrimination in what are supposed to be safe spaces may cause strain on trans BIPOC and impede their sense of inclusion in the LGBTQ community and racial/ethnic communities (Bowleg, 2013; Narváez et al., 2009).

However, research findings have been inconsistent regarding disparities in experiences with discrimination between racial/ethnic groups beyond categories of White and BIPOC (Kattari et al., 2017; Lombardi, 2009; Miller & Grollman, 2015). In Miller and Grollman (2015) study of experiences with trans discrimination, Asian/Pacific Islander participants and African American or Black participants reported fewer discriminatory events than White participants, while multiracial, American Indian, and Latinx participants reported more discriminatory events than White participants. Conversely, several studies found similar experiences with transphobia (Herman, 2013; Lombardi, 2009; Nuttbrock et al., 2010) or racial and sexual violence (Testa et al., 2012) across racial/ethnic groups. The different dimensions of transphobia and discrimination assessed in previous studies, such as physical assault, verbal harassment, or institutional discrimination, may cause these inconsistent findings.

Current study

The literature on racial/ethnic differences in gender discrimination and psychological distress has several limitations. First, prior research on the trans population primarily focuses on White trans people and trans-identified people who identify as either trans men or trans women (Bockting et al., 2013; Budge et al., 2016; Miller & Grollman, 2015). Given the paucity of research that takes into account nonbinary identities, research efforts focused on the trans population should make deliberate efforts to obtain a racially and ethnically diverse sample of trans individuals with diverse identities (e.g., nonbinary, agender, genderqueer) (Fiani & Han, 2019). Second, several previous studies measure the effects of race/ethnicity on discrimination and victimization, but only a small portion analyze differences between racial/ethnic groups beyond only comparing White trans people and trans BIPOC (Dinno, 2017; Kattari et al., 2017; Miller & Grollman, 2015; Nuttbrock et al., 2010; Testa et al., 2012). This analysis may miss differences in experiences with discrimination and victimization between various racial/ethnic groups.

Third, very few quantitative studies evaluate racial/ethnic and gender differences in mental health and gender minority stress among the trans population. Research focused on racial/ethnic differences in discrimination and mental health within the trans population have been small, qualitative studies (De Vries, 2012; Levitt & Ippolito, 2014) or focus on subgroups of the trans population, such as sex workers (Nemoto et al., 2011) and youth (Singh, 2013). While these studies offer very rich descriptions of the experiences of trans people, they are limited in the extent to which statistically significant differences can be identified, and results from subgroups of the trans population do not necessarily reflect the typical trans experience.

This research aims to fill this gap in the literature by examining racial/ethnic differences in experiences with gender minority stress and psychological distress within a sample of trans adults. This study compares gender minority stress and psychological distress across racial/ethnic groups. Additionally, this study aims to include nonbinary participants and other participants who identify outside of binary categories of men and women.

To reach these goals, we propose the following hypotheses to guide these analyses: In a sample of trans individuals, BIPOC will report significantly higher levels of gender minority stress than White individuals (H1), and reported gender minority stress will significantly differ across racial and ethnic groups (H2). Additionally, in a sample of trans individuals, BIPOC will experience higher rates of psychological distress compared to White individuals (H3) and reported psychological distress will significantly differ across racial and ethnic groups (H4). Lastly, changes in minority stress will be positively correlated with reported psychological distress (H5).

Methods

Participants

Because trans people make up a small portion of the population, a convenience sampling technique was used to recruit participants for an online survey. The first author emailed the study information to 443 North American LGBTQ organizations (e.g., university organizations and LGBTQ community centers). Sixty-six organizations publicized the study on their websites, blogs, newsletters, listservs, or social media pages. Organizations were provided with flyers, a survey description, and the survey link. Since being trans is highly stigmatized and carries concerns related to privacy, distributing the survey online had the benefit of potentially reaching participants who are not openly out as trans or may not have connections to trans-specific organizations or information channels that reach the trans population (Compton, 2018; Rosser et al., 2007). Participant eligibility requirements included being at least 18 years of age and self-identifying as transgender.

Participants included 99 self-identified trans adults. Although the full sample included 215 participants, this analysis only utilized data from participants with complete data for all measures and controls examined. See Table 1 for complete descriptive statistics. The 116 respondents excluded from the analyses due to missing data were not significantly different from the analytic sample in terms of age, race, sex assigned at birth, or gender identity. They were, however, less likely to have been born outside the United States. Among those included in the study analyses, there were no differences in age, region, sex assigned at birth, or gender identity by race/ethnicity. Participants’ ages ranged from 18-year old to 63-year old. The average age was 28.43-year old (SD = 12.26).

Table 1.

Descriptive statistics (n = 99).

Variable	Frequency	Mean/Percent
Age		28.43
Gender Identity
Transfeminine	25	25.51
Transmasculine	26	26.53
Nonbinary	45	45.92
Other	2	2.04
Missing	1	1.01
Sex Assigned at Birth
Male	32	32.32
Female	64	64.65
vIntersex	3	3.03
Race/Ethnicity
White	63	63.64
Black or African American	4	4.04
Hispanic or Latino	7	7.07
Asian	6	6.06
Multiracial	15	15.15
Other	4	4.04
Region
Northeast	15	15.15
South	19	19.19
Midwest	23	23.23
Mountain West	8	8.08
Pacific	19	19.19
Alaska and Hawaii	2	2.02
Outside United States	13	13.13

Measures

To maximize the range of gender and racial/ethnic diversity in the data, data on participant gender identity were gathered with an open-ended response and data on race/ethnicity as a “select all that apply” question with a text box for participants to write-in their racial/ethnic identity if desired. Open-ended race responses were recoded into White (63.64%), African American or Black (4.04%), Hispanic or Latino (7.07%), Asian (6.06%), multiracial (15.15%), and “other” racial/ethnic group (4.04%) participants. “Other” racial/ethnic group participants included Native Hawaiian or other Pacific Islander, Japanese and Salvadorian, Arab, and one participant who identified as half-Middle Eastern.

As with race, open-ended participant reported gender identities were coded into four categories. “Transfeminine” included participants who identified as trans women, women, or female. “Transmasculine” included participants who self-identified as trans men, men, or male. “Nonbinary” included participants who self-identified as nonbinary, gender nonconforming, genderless, or agender. The majority of participants self-identified as nonbinary (45.92%), followed by transmasculine (26.53%), transfeminine (25.51%), and other gender labels (2.04%), such as “bigender.” The two "other" gender-identified participants were not included in the between-gender group analysis. One participant did not record their gender identity. Lastly, participants indicated their sex assigned at birth as either male, female, or intersex. The majority of participants were assigned female at birth (64.65%), followed by assigned male at birth (32.32%), and intersex (3.03%).

Gender minority stress

The gender minority stress scale includes two subscales regarding gender-related discrimination and gender-related victimization, based on Meyer’s (1995) theory of minority stress to assess minority stress and resilience strategies (Testa et al., 2015). The subscales include five and six statements, respectively. For the gender-related discrimination subscale, statements pertained to difficulty obtaining medical or mental health treatment, identity documents, a safe restroom, housing, employment, and promotion because of their gender identity or expression (α = 0.67). For the gender-related victimization subscale, statements pertained to verbal harassment, being outed or blackmailed, personal property damage, physical harm, and unwanted sexual content related to their gender identity or expression (α = 0.84) (Testa et al., 2015). Participants indicated their level of agreement to each statement on a 7-point scale ranging from “strongly disagree” to “strongly agree.” The gender minority stress combined scale refers to summation of the gender discrimination and victimization scales (α = 0.85). For each scale responses were summed and then divided by the number of questions in the scale, producing a range of 1 to 7. Higher scores indicate greater gender minority stress.

Psychological distress

To measure psychological distress, we use seven statements from The Hopkins Symptom Checklist-21 to capture general psychological distress. The Hopkins Symptom Checklist-21 is an abridged version of a widely used measure to assess symptoms of psychological distress in ethnically diverse populations with well-established reliability and validity (Cepede-Benito & Gleaves, 2000; Green et al., 1988). The subscale included seven statements pertaining to self-blame, feeling lonely, feeling down, being sensitive, feeling misunderstood, feelings that others do not like the person, and feelings of inferiority (α = 0.90) (Cepede-Benito & Gleaves, 2000; Green et al., 1988). Respondents were asked to answer how often they felt psychological distress symptoms in the previous seven days by selecting either “not at all” (1), “a little” (2), “quite a bit” (3), or “extremely” (4). Again, the items were summed and averaged such that scores could range from 1 to 4 where higher scores indicate higher psychological distress.

Controls

Due to limitations of the data, only four control variables were included: age, sex assigned at birth, gender identity, and region. Age is a continuous variable measured in years. Sex assigned at birth is a categorical variable with options male, female, or intersex. The coding of gender identity is described above. Region refers to the region the participant comes from within the United States, split up into Northeast, South, Midwest, Mountain West, Pacific, Alaska and Hawaii, and outside the United States.

Procedure

This study utilized data from a nationwide survey of trans identity development. The questionnaire was constructed using the Qualtrics research platform. Prior to the release of the official survey, a pilot survey was conducted to determine how the length of the questionnaire could affect the response rate. The full completion rate for the pilot study of six participants was 83%. The questionnaire was offered in both English and Spanish; however, there were no participants who chose to take the survey in Spanish.

Ethical research conduct and protecting the safety and rights of participants were priorities in carrying out this research. Ensuring this study was ethical and just was prioritized particularly because trans people—especially trans BIPOC—have been historically marginalized. The informed consent form at the beginning of the questionnaire gave a detailed description regarding the topics of interest of the study, the potential risks of participation, the steps taken by the researcher to preserve confidentiality and minimize risks, and how to access mental health resources that specialize in providing support to trans people. Participants were not compensated, nor were incentives provided to complete the survey. Participants gave their informed consent by checking a box before the beginning of the questionnaire.

Throughout the questionnaire, participants were notified about potentially sensitive topics before seeing these specific questions and reminded that they can skip any questions without penalty. At the end of the questionnaire, the debriefing form indicated how to access mental health resources that specialize in providing support to trans people. The debriefing form also contained the primary investigator and faculty advisor’s contact information. Participants were encouraged to contact the researchers with any questions or concerns. Before finalizing the questionnaire, a local university LGBTQ organization and 31 of the organizations that publicized this study reviewed the questionnaire to assess whether the language and measures were inclusive and respectful toward the trans community. This study was approved by the Institutional Review Board of Coastal Carolina University.

Data analysis

Data analysis was conducted using Stata version 16 software. For the first set of analyses, we examined bivariate statistics. H1 and H3 were first tested using two-tailed independent samples t-tests to test for significant differences in gender minority stress and psychological distress between White and BIPOC participants. To assess H2 and H4 we conducted one-way ANOVA to test for significant differences in gender minority stress and psychological distress across racial/ethnic groups. H5 was first tested though correlations between the dependent variables.

After the bivariate analyses, we used Ordinary Least Squares regression to again test each hypothesis, functioning as a robustness check. H1 and H3 were tested by regressing white versus BIPOC on the gender minority stress scales and psychological distress scale, respectively, and then adding control variables to the model. H2 and H4 were tested in the same fashion by first regressing only race/ethnicity on the gender minority stress scales and psychological distress scale, and then again with the addition of the control variables. Lastly H5 was tested, through regressions predicting psychological distress using the gender minority stress combined scale as an independent variable. Both race/ethnicity categories and White versus BIPOC were examined–with and without the control variables.

Results

Gender minority stress

The total sample had an average score of 3.66 out of 7 for the gender-related discrimination subscale. White respondents compared to BIPOC had virtually the same mean gender-related discrimination scores, leading us to reject H1. Broken down by race/ethnicity, African American or Black respondents and Hispanic or Latino respondents on average had higher scores than other racial and ethnic respondents, while Asian and “Other” race respondents on average had lower scores, indicating less discrimination. The average white and multiracial respondent had average gender-related discrimination scores that were similar to the overall mean. As shown in Tables 2 and 3, two-tailed independent tests and the ANOVA test suggest these bivariate differences were not statistically significant.

Table 2.

Subscale composite score ANOVA results – race/ethnicity.

	Mean	Std. Deviation	F.	Sig.
Gender-related discrimination
White (N = 63)	3.66	1.20	1.21	0.3093
African American Or Black (N = 4)	4.45	0.19
Hispanic Or Latino (N = 7)	4.29	0.55
Asian (N = 6)	3.00	0.54
Multiracial (N = 15)	3.51	1.51
Other (N = 4)	3.35	1.51
Total (n = 99)	3.66	1.20
Gender-related victimization
White	2.85	1.54	1.52	0.1924
African American Or Black	3.08	1.32
Hispanic Or Latino	4.10	0. 92
Asian	2.06	0.44
Multiracial	3.14	1.62
Other	2.25	2.28
Total (n = 99)	2.92	1.52
General psych distress
White	2.25	0.81	3.36	0.0078
African American Or Black	3.21	0.36
Hispanic Or Latino	2.23	0.42
Asian	2.00	1.00
Multiracial	2.99	0.76
Other	2.54	1.09
Total (n = 99)	2.40	0.84

Table 3.

Subscale composite score T-test results – race/ethnicity (n = 99).

	Mean	Std. Deviation	T-test for equality of means	Sig. (two-tailed)
Gender-related discrimination
White (N = 63)	3.66	1.20	0.00	0.9975
BIPOC (N = 36)	3.66	1.21
Gender-related victimization
White	2.85	1.54	0.60	0.5532
BIPOC	3.04	1.51
General psych distress
White	2.25	0.81	2.34	0.0213
BIPOC	2.65	0.84

However, as shown in Table 4, using regression, some significant racial/ethnic differences were found. Hispanic or Latino respondents had gender-related discrimination scores 1.07-points higher relative to White respondents (p < .05), holding all controls constant. Asian participants, on average, have scores 0.90 points lower on the gender-related discrimination scale, but this finding was only marginally significant (p < .10). In no other racial/ethnic groups, however, were significant differences in gender related discrimination found. Therefore, H2 can be partially accepted. We also found that those from the Pacific Northwest had lower levels of gender-based discrimination relative to those in the northeast, holding all else equal (p < .01). This is perhaps because the Pacific Northwest of the United states is known for liberal politics and a greater acceptance of LGBT individuals, broadly.

Table 4.

Coefficients of OLS regression of race/ethnicity on gender-based discrimination.

	Base model	+ controls	Base model	+ controls
BIPOC	–0.00	–0.14
	(–0.00)	(–0.50)
Race/ethnicity
Black			0.79	0.57
			(1.29)	(0.79)
Hispanic or Latino			0.63	1.07*
			(1.32)	(2.23)
Asian			–0.66	–0.90+
			(–1.30)	(–1.70)
Multiracial			–0.15	0.02
			(–0.45)	(0.05)
Other			–0.31	0.02
			(–0.51)	(0.04)
Age		–0.01		–0.01
		(–0.54)		(–0.89)
Sex assigned at birth
Female		–0.46		–0.41
		(–0.99)		(–0.89)
Intersex		0.90		0.96
		(1.15)		(1.26)
Gender identity
Trans Man/Male		0.64		0.53
		(1.14)		(0.96)
Nonbinary		0.37		0.31
		(0.74)		(0.62)
Other		0.34		0.51
		(0.31)		(0.47)
Region
South		–0.47		–0.44
		(–1.11)		(–1.06)
Midwest		0.03		0.08
		(0.06)		(0.20)
Mountain West		–0.49		–0.50
		(–0.94)		(–0.97)
Pacific		–1.16**		–1.23**
		(–2.79)		(–2.96)
Alaska and Hawaii		–0.11		0.32
		(–0.11)		(0.31)
Outside United States		–0.38		–0.35
		(–0.80)		(–0.73)
Constant	3.66***	4.25***	3.66***	4.23***
	(18.27)	(7.16)	(24.41)	(7.16)
Observations	99	98	99	98
BIC	324.64	358.19	336.77	365.68

Note: Estimates made using unstandardized dependent variables. T-statistic in parenthesis. White is the baseline category for both race/ethnicity measures. Other reference categories are male for sex assigned at birth, Trans-Female/Trans-Feminine for gender identity, Northeast for Region.

+p < .10,

*p < .05,

**p < .01,

***p < .001.

The sample had an average gender-based victimization score of 2.92 out of 7, indicating they somewhat disagree that they have experienced discrimination. White respondents, on average, had scores of 2.85 whereas BIPOC on average scored 3.04. ANOVA tests and multivariate analyses suggest this difference is insignificant, leading us to reject H3 (see Table 5). Looking at differences between ethnic group there appear to be some differences; Hispanic or Latinos had the highest gender-related victimization scores of any racial/ethnic group with an average of 4.10 and Asians had the lowest scores with an average of 2.06. Again, results of ANOVA and t-tests suggest these differences are not statistically significant. Through multivariate analysis, however, looking at gender-based victimization net of controls, significant differences were found among Hispanic or Latinos who, on average, scored 1.57 points higher on the gender-related victimization scale relative to Whites (p < .05). Thus, we partially accept H4.

Table 5.

Coefficients of OLS regression of race/ethnicity on gender-based victimization.

	Base model		+ controls	Base model	+ controls
White vs. BIPOC	–0.19		–0.28
	(–0.60)		(–0.78)
Race/Ethnicity
Black				0.23	–0.61
				(0.30)	(–0.65)
Hispanic or Latino				1.24*	1.57*
				(2.08)	(2.52)
Asian				–0.80	–1.00
				(–1.24)	(–1.47)
Multiracial				0.29	0.48
				(0.68)	(1.05)
Other				–0.60	–0.36
				(–0.78)	(–0.45)
Age			–0.01		–0.02
			(–0.73)		(–1.15)
Sex assigned at birth
Female			–1.27*		–1.35*
			(–2.07)		(–2.27)
Intersex			0.57		0.61
			(0.56)		(0.62)
Gender identity
Trans Man/Male			1.24+		1.26+
			(1.69)		(1.78)
Nonbinary			1.01		1.14+
			(1.52)		(1.75)
Other			0.51		1.04
			(0.35)		(0.74)
Region
South			–0.57		–0.73
			(–1.04)		(–1.35)
Midwest			–0.04		–0.16
			(–0.08)		(–0.31)
Mountain West			–0.70		–0.85
			(–1.03)		(–1.27)
Pacific			–0.55		–0.84
			(–1.01)		(–1.56)
2003 Alaska and Hawaii			0.40		0.51
			(0.30)		(0.39)
Outside United States			0.10		–0.09
			(0.17)		(–0.14)
Constant		3.04***	3.67***	2.85***	3.70***
		(11.95)	(4.73)	(15.07)	(4.83)
Observations		99	98	99	98
BIC		371.92	410.76	382.90	416.51

Note: Estimates made using unstandardized dependent variables. T-statistic in parenthesis. White is the base category for race/ethnicity. Other reference categories are male for sex assigned at birth, Trans-Female/Trans-Feminine for gender identity, Northeast for Region.

+p < .10,

*p < .05,

**p < .01,

***p < .001.

We also found significant differences in gender-based victimization scores by sex assigned at birth and gender identity. Results indicate that trans individuals assigned female at birth have gender-based victimization score 1.35 points lower on the scale relative to those assigned male at birth (p < .05). Furthermore, individuals identifying as trans masculine have 1.26 higher gender-minority related victimization relative to trans feminine respondents, and nonbinary individuals have scores 1.14 higher, holding all else equal (both p < .1).

Psychological distress

BIPOC reported significantly more psychological distress than white respondents. Across all racial/ethnic groups, the average psychological distress score was 2.40. On average BIPOC had general psychological distress scores of 2.65 compared to white participants who averaged 2.25. A two-tailed test showed this difference is indeed significant (p < .05). Regression results confirm these bivariate patterns (See Table 6). Without any controls in the model, a White respondent, on average, has a 0.40-point lower psychological distress score compared to a BIPOC respondent (p < .05). When controls are added to the model, however, this effect becomes only marginally significant (p < .10) indicating the controls at least partially explain the role of race/ethnicity. Based on this finding, we partially accept H3.

Table 6.

Coefficients of OLS regression of race/ethnicity on psychological distress.

	Base model	+ controls	Base model	+ controls
White vs. BIPOC	–0.40*	–0.30
	(–2.34)	(–1.59)
Race/Ethnicity
Black			0.96*	0.85+
			(2.36)	(1.72)
Hispanic or Latino			–0.02	0.08
			(–0.08)	(0.26)
Asian			–0.25	–0.60
			(–0.74)	(–1.65)
Multiracial			0.74**	0.70**
			(3.25)	(2.87)
Other			0.29	0.19
			(0.70)	(0.45)
Age		–0.02**		–0.02**
		(–2.88)		(–2.87)
Constant	2.65***	3.38***	2.25***	3.01***
	(19.38)	(8.23)	(22.51)	(7.44)
Observations	99	98	99	98
BIC	249.04	286.15	256.43	291.00

Note: Estimates made using unstandardized dependent variables. t statistic in parenthesis. White is the base category for race/ethnicity. Omitted controls are insignificant.

+p < .10,

*p < .05,

**p < .01,

***p < .001.

Black or African American respondents had the highest average psychological distress score of all racial/ethnic groups at 3.21, while Asians had the lowest average score of 2.00. ANOVA tests show that these racial/ethnic differences are significant (p < .01). Furthermore, regression results indicate being Black or African American is associated with a 0.85 point increase on the psychological distress scale (p < .1), and Multiracial respondents on average are associated with a 0.70 higher score (p < .01), when compared to White respondents, holding all else equal. This leads us to again partially accept H4, as there are only some significant differences in psychological distress across the comparisons of racial and ethnic groups.

Lastly, we find minority stress is positively associated with psychological distress, leading us to accept H5. We find psychological distress is significantly and positively correlated with both gender minority stress subscales and the combined scale. Additionally, Table 7 shows the results of OLS regressions further examining the relationship between gender minority stress and race/ethnicity on psychological distress. In all models, gender-minority stress is positively associated with psychological distress, such that more gender-minority related stress confers more psychological distress. When using a detailed measure of race/ethnicity, increased gender-minority stress, being Black (b = 0.87, p < .1), and being multiracial (b = 0.64, p < .01) are significantly associated with increased psychological distress scores, holding all controls constant.

Table 7.

Predicting psychological distress by minority stress.

	Base model	+ controls	Base model	+ controls
Gender minority stress	0.23***	0.24***	0.23***	0.23**
	(3.53)	(3.43)	(3.62)	(3.28)
White vs. BIPOC	–0.38*	–0.25
	(–2.33)	(–1.40)
Race/Ethnicity
Black			0.85*	0.87+
			(2.21)	(1.86)
Hispanic or Latino			–0.25	–0.23
			(–0.82)	(–0.71)
Asian			–0.08	–0.37
			(–0.24)	(–1.07)
Multiracial			0.72**	0.64**
			(3.36)	(2.75)
Other			0.40	0.23
			(1.03)	(0.59)
Age		–0.02**		–0.02*
		(–2.77)		(–2.57)
		(–0.96)		(–0.14)
Constant	1.89***	2.44***	1.50***	2.08***
	(7.56)	(5.16)	(6.56)	(4.38)
Observations	99	98	99	98
BIC	241.58	277.7	247.87	283.10

Note: Estimates made using unstandardized dependent variables. t statistic in parenthesis. White is the base category for race/ethnicity. Omitted controls are insignificant.

+p < .10,

*p < .05,

**p < .01,

***p < .001.

Discussion

Minority stress theory posits that individuals who experience discrimination and victimization have higher levels of stress than individuals with privileged social identities, which leads to compromised mental health. Intersectionality theory posits that individual experiences depend on the intersection of identity statuses. For individuals with multiple minority social statuses, minority stress may be greater due to possessing multiple marginalized identities. Our study on trans individuals, a dramatically marginalized group, finds only particular racial and ethnic groups within the trans BIPOC umbrella report greater discrimination and victimization, and higher psychological distress relative to white trans individuals.

When BIPOC participants are treated as a single group, they did not report significantly different levels of gender discrimination or victimization compared to their White counterparts. Additionally, gender discrimination was not significantly different based on gender identity or sex assigned at birth. Those who reported being assigned female at birth, however, did report less gender victimization relative to individual’s assigned male at birth. Additionally, transmasculine individuals had marginally higher levels of gender victimization relative to transfeminine. While these findings may seem at odds, it’s important to keep in mind that female at birth does not necessarily mean transmasculine; rather, these findings may more so indicate a stricter policing of masculinity relative to femininity (Koenig, 2018). We also find those who identify outside the gender-binary experience moderately higher levels of gender victimization relative to transfeminine individuals, likely due to societal discomfort and greater stigma associated with identifying outside of the male–female binary (Lefevor, Boyd-Rogers, et al., 2019; Miller & Grollman, 2015).

Certain individual racial/ethnic groups reported significantly different levels of gender-minority stress relative to white participants. Asian participants reported marginally less gender-related discrimination compared to White participants, and Hispanic or Latino participants reported significantly higher gender-related discrimination and gender-related victimization compared to White participants. It is perhaps because ethnic groups have different levels of minority stress that when we combine this group these differences are lost and the effects cancel each other out, giving us our observed effect that trans BIPOC as a group do not have levels of stress different from White trans people

These findings support intersectionality theory and adds to the growing body of research that attempts to understand the complex relationship between minority stress, psychological distress, and race among trans individuals. Previous research findings that some racial/ethnic minority groups report significantly greater victimization and discrimination than White trans individuals (Bradford et al., 2013; Kattari & Hasche, 2016; Kattari et al., 2017; Miller & Grollman, 2015; Nuttbrock et al., 2010; Reisner et al., 2016), but even these studies find inconsistent patterns within the BIPOC group, and other studies have found mixed or no significant differences in transphobia-related discrimination and victimization across racial/ethnic groups (Herman, 2013; Lombardi, 2009; Testa et al., 2012). For example, Miller and Grollman only find elevated experiences of major discrimination in Latinos, American Indians, and multiracial individuals while lower levels in Asians, compared to Whites. This parallels our finding that Hispanic/Latino participants had significantly higher scores for gender related discrimination compared to Whites, while Asian participants had marginally lower scores. This may be due to the compounding influence of racial discrimination that affects groups of color that may not be able to pass as White or are more heavily stigmatized than Asians, who are often conceived of as the “model minority” (Wong et al., 1998)—which may explain the lower mean scores reported by Asian participants.

Despite few differences in gender minority stress across ethnic/racial groups BIPOC on average reported significantly more psychological distress than White respondents. However, this difference diminished to marginal significance when more explanatory variables were added to the model, and when accounting for minority stress, became entirely insignificant. Specifically, Black participants and multiracial participants reported the highest rates of psychological distress. The increased distress of Black respondents remained consistent across models, although with the addition and controls and gender minority stress it became only marginally significant. The increased distress of multiracial respondents was consistently significant across all models. This increased distress among Black and multiracial respondents is consistent with research that has shown Black and multiracial individuals experience high levels of transgender stress (Lombardi, 2009) and multiracial transgender individuals are at an increased risk of suicide relative to White respondents (Miller & Grollman, 2015), which indicates psychological distress. These findings are in line with a generally mixed body of research on the relationship between racial/ethnic minorities and mental health (Adams & Vincent, 2019; Hatchel et al., 2019; Katz-Wise et al., 2017; Lytle et al., 2016; Pflum et al., 2015).

It is worth to note that in studies of cisgender individuals, Black people have been found to have better mental well-being compared to white respondents (Keyes, 2009). Although this is at odds with stress process frameworks, it is often attributed to a strong sense of self-worth, religion, and group-identification (Keyes, 2009), however, our finding that Black trans individuals experience more psychological distress suggests trans people may not be able to reap these same benefits of the collective. Although we did not find differences in minority stress by being Black, Black trans individuals, and particularly Black trans women experience an astronomically high rate of violence (Dinno, 2017; Human Rights Campaign, 2019). Perhaps because of this, we surmise Black trans individuals live in fear of violence and discrimination, regardless of having personally experienced such, which may be detrimental to psychological well-being.

In sum, the current investigation presents a complex picture. This research highlights that BIPOC make up a heterogeneous group; by solely examining race/ethnicity as a binary variable, studies mask potential important differences between groups. The lack of consistency in the relationships between race/ethnicity, gender minority stress, and psychological distress suggests that some other mediating or confounding variables, such as racial/ethnic discrimination, or measures of self-concept, might contribute to disparities in mental health across racial/ethnic groups. Thus, future research is necessary to further parse out these different results among racial and ethnic groups and determine the factors that contribute to disparities in psychological distress and mental health by race/ethnicity within the trans population.

Limitations and conclusions

While this study’s findings offer insight into differences in discrimination, victimization, and mental health among trans individuals, this study was not without limitations—including limitations characteristic of LGBTQ research more broadly. First, the analysis of discrimination, victimization, and psychological distress did not consider all other confounding variables that may contribute to these experiences. Previous studies document the effects of age, socioeconomic status, social support, sexual orientation, internalized transphobia, and education on experiences with discrimination, victimization, and individuals’ mental health.

Furthermore, while our results also show gender minority stress is consistently associated with psychological distress, we cannot establish causality due to the cross-sectional nature of our data. However, the finding that gender minority stress is correlated with and predicts psychological distress is consistent with the psychological mediation framework that posits stressors may make trans people more vulnerable to maladaptive behaviors and negative psychological processes (Hendricks & Testa, 2012; Tan et al., 2020). Recent research has also proposed additional potential mediators of the relationship between minority stress and mental health, such as psychological inflexibility among primarily white transgender British participants (Lloyd et al., 2019), and self-efficacy and self-esteem among African American and Caribbean Black cisgender adolescents (Smith & Nicholson, 2020), however, due to limitations of our data we were not able to test these hypotheses in our study population.

Second, the racial/ethnic categories used in this study do not fully capture the diversity of racial/ethnic groups, nor do they consider how an individual is connected to their racial/ethnic community, nor the visibility of an individual’s race/ethnicity. Methods of categorizing individuals’ race/ethnicity are contested, and categories are not discrete, but fluid. Likewise, racial/ethnic identity is, in many ways, subjective to self-identification (Brubaker, 2016). When coding the racial/ethnic data in this research, this study aimed to preserve racial/ethnic diversity while still employing broad enough categories for use in the statistical analysis with a relatively small sample size.

Lastly, this study is limited by its sample. The sample size is small, and a convenience sampling technique was used to reach participants. While a random sampling technique is typically superior in obtaining a representative sample, random sampling is often not feasible for studies of the LGBTQ population. Random sampling and population research with LGBTQ people are particularly challenging due to the various ways to define a LGBTQ identity and lack of sampling frame available to researchers (Compton, 2018). Additionally, the sample in this study is likely limited to those with internet access, social media access, and/or connections to LGBTQ-specific organizations. However, given the small size of the trans population and the privacy concerns associated with disclosing one’s trans identity, internet-based recruitment is one of the most widely-used participant recruitment techniques when studying LGBTQ-identified individuals (Rosser et al., 2007).

Moving forward, future research should continue to examine differences in minority stress across various social groups (especially harder-to-reach subgroups) within the trans population. This research may also benefit from paying particular attention to potential psychosocial pathways and how place influences experience of minority stress and distress among ethnic minorities, as it may be possible that the location interacts with ethnic identity, as regional cultural differences or the existence of “gayborhoods” in local contexts may influence the extent one experiences stressors and as a consequence distress (Gieryn, 2000; Stone, 2018). Through continued research on the intersections of identity statuses and experiences with discrimination affecting the trans population, researchers, clinicians, and advocates for the trans community can improve support services and advocacy efforts to address the harms associated with transphobia and poor mental health.

Declaration of interest

The authors declare that they have no conflict of interest.

Ethical approval

All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards.

Data availability statement

That the authors are willing to allow the journal to review their data if requested.