Intersectional Minority Stress and Intimate Partner Violence: The Effects of Enacted Stigma on Sexual and Racial Minority Youth Assigned Female at Birth

Abstract

Many sexual minority youth of color experience enacted stigma based on each of their minority identities. We examined whether experiences of racist discrimination and heterosexist microaggressions were associated with intimate partner violence (IPV) among female-assigned at birth (FAB) sexual minority youth of color. Data were drawn from a larger study of FAB sexual and gender minority youth (FAB400; N = 488). We selected racial/ethnic minority participants who reported a sexual minority identity and reported a romantic relationship in the previous 6 months (N = 249). Negative binomial models were used to test for associations between enacted stigma (racial discrimination and heterosexist microaggressions) and IPV (psychological, physical, sexual, and sexual minority-specific) perpetration and victimization. When considered separately, both forms of enacted stigma was positively associated with perpetration and victimization across all four types of IPV. In multivariate models, racial discrimination and heterosexist microaggressions both had unique, additive effects on psychological IPV perpetration and physical- and sexual minority-specific IPV victimization. Only racial discrimination was uniquely associated with physical perpetration and psychological victimization. Only heterosexist microaggressions was uniquely associated with sexual minority-specific perpetration and sexual IPV perpetration and victimization. Findings illustrate how enacted stigma based on each minority identity intersect to raise risk for IPV among sexual minority youth of color.

INTRODUCTION

Although intimate partner violence (IPV) is prevalent among all U.S. adolescents and young adults (e.g., Breiding, Chen, & Black, 2014; Kann et al., 2016), there is growing recognition that those who identify as a sexual minority person are at greater risk of experiencing IPV than their heterosexual counterparts, including being twice as likely to experience psychological IPV, three times as likely to experience physical IPV, and five times as likely to experience sexual IPV (Kann, 2011; Porter & Williams, 2011). Recent research further suggests that, among sexual minority youth, those who also identify with a minority racial group are at greater risk of experiencing IPV than those who identify as White (Reuter, Newcomb, Whitton, & Mustanski, 2017; Steele, Everett, & Hughes, 2020; Whitton, Dyar, Mustanski, & Newcomb, 2018).

These findings highlight the importance of identifying potential mechanisms contributing to the higher risk of IPV experienced by sexual minority youth of color. Experiences of enacted stigma, which are defined as external experiences ranging from experiences of discrimination and overt victimization to experiences of covert acts of aggression based on one’s minority status, have been linked with a wide range of negative health effects on racial and sexual minority people (Birkett, Newcomb, & Mustanski, 2015; Carter, Lau, Johnson, & Kirkinis, 2017; Hurd, Varner, Caldwell, & Zimmerman, 2014; Lowry, Johns, Robin, & Kann, 2017). However, few studies have explored the effects of enacted stigma on IPV among either sexual or racial minorities, and even fewer have simultaneously examined the effects of enacted stigma based on both race and sexual orientation.

Minority Stress and Intimate Partner Violence Risk

Minority stress theory describes how members of minority groups experience unique stressors that arise from the stigmatization of their marginalized identities (Clark, Anderson, Clark, & Williams, 1999; Hatzenbuehler, 2009; Meyer, 2003). These “minority stressors” include experiences of enacted stigma such as discrimination, acts of violence, and microaggressions committed by other people. Research suggests that minority stress is associated with negative physical and mental health outcomes in sexual minority youth (e.g., Birkett et al., 2015) and racial minority youth (e.g., Carter et al., 2017) and can largely explain the health inequities faced by minority populations (Hatzenbuehler, Phelan, & Link, 2013).

Theories of IPV that emphasize how violence between romantic partners mirrors the cultural violence toward and oppression of certain groups (Almeida, Woods, Messineo, Font, & Heer, 1994) suggest that societal heterosexism and racism contribute to IPV among sexual and racial minorities by placing additional stress on their relationships that can foster dysfunctional relationship dynamics. This is in accordance with LeBlanc, Frost, and Wight’s (2015) extension of minority stress theory that describes how minority stressors may contribute, not only to negative individual outcomes, but also poor relationship outcomes, including couple distress, conflict, and breakup. Stress, more broadly defined, is well-known to adversely impact couple relationships (Karney & Bradbury, 1995; Neff & Karney, 2004), at least partially because it hinders adaptive couple processes such as constructive conflict resolution (Neff & Karney, 2009). It seems likely, then, that minority stressors might also raise individual risk for IPV by intensifying maladaptive couple processes.

In accordance with these theories, existing studies have demonstrated that experiences of enacted stigma based on race are associated with IPV among heterosexual people of color. Asian American women who report experiencing racial discrimination were more likely to be victims of IPV compared to Asian American women who did not report discrimination (Cho, 2012). Similarly, in a study of Black women, those who reported experiences of perceived discrimination were more likely to have experienced either psychological or physical/sexual IPV compared to those who did not experience discrimination (71% vs. 44%) (Waltermaurer, Watson, & McNutt, 2006). The association between racial discrimination and IPV appears to extend to younger women as well; in a sample of Black and Latina women between the ages of 19 and 20 years, experiences of discrimination were positively associated with psychological and physical IPV, both victimization and perpetration (Stueve & O’Donnell, 2008).

Similarly, the increased risk for IPV among sexual minority people has been linked to the enacted stigma they experience due to their sexual orientation. This has been true for sexual minority college students, who have expressed that stigma, at the societal and individual level, is central to their IPV experiences (Gillum & DiFulvio, 2012) and who have reported that sexual orientation-based victimization was associated with psychological, but not physical or sexual, IPV perpetration from their current same-sex partner (Edwards & Sylaska, 2013). Sexual orientation based discrimination has also been associated with lifetime experiences of physical and psychological IPV perpetration and victimization in lesbian, gay, and bisexual women (Balsam & Szymanski, 2005) and reports of physical and sexual IPV perpetration and victimization in the previous year in cisgender gay and bisexual men (Finneran & Stephenson, 2013).

Most of the existing literature on sexual minority experiences of IPV have not assessed gender identity (Martin-Storey, 2015; Messinger, 2011) or have been conducted exclusively with cisgender women (Balsam & Szymanski, 2005; Freedner, Freed, Yang, & Austin, 2002). This has been true despite transgender and non-binary people making up a sizable proportion of the sexual minority population (Kuper, Nussbaum, & Mustanski, 2012). Amongst youth between the ages of 18 to 34 years, up to 40% of LGBTQ individuals identify as both sexual and gender minorities (GLAAD, 2017). The exclusion of gender minority groups in prior research risks limiting our understanding of how IPV functions in sexual minority populations.

Intersectionality and Intimate Partner Violence

Intersectionality theory states that for people of multiple minority statuses, individual experiences are shaped by the convergence of their multiple identities, which results in experiences uniquely different from people who do not share that combination of identities (Bauer, 2014; McCall, 2005). In particular, multiple forms of oppression based in race and sexual orientation are posited to intersect to affect individuals’ experiences of victimization. Intersectionality theory encompasses multiple theoretical perspectives. Else-Quest and Hyde (2016) argued that quantitative approaches for exploring intersectionality should include analysis of multiple main effects, interaction effects, and intersectional effects. Research of multiple main effects is most akin to the greater risk perspective which posits that people with intersectional minority identities are made more vulnerable by each oppressive system to which they are exposed (Bowleg, Huang, Brooks, Black, & Burkholder, 2003; Greene, 1996; King, 2016). According to this perspective, sexual minority youth of color who experience enacted stigma based on both their race/ethnicity and sexual orientation would be the most vulnerable to negative outcomes because of the increased exposure they experience. The additive effects of each form of stigma may heighten vulnerability to maladaptive couple processes such as IPV.

Else-Quest and Hyde’s (2016) interest in interaction effects highlight the possibility that experiencing enacted stigma based on one identity will exacerbate the negative effects of stigma experiences based in the other identity, so that sexual minority youth of color who experience high levels of both racial and heterosexist stigma may be at particularly high risk for IPV. It is also consistent with the resilience perspective which describes an attenuating effect of exposure to multiple avenues of oppression (Meyer, 2010; Moradi et al., 2010): exposure to one form of stigma is proposed to facilitate coping mechanisms and skills that can be applied to experiences of another form of stigma and create resilience.

Finally, the last type of effect that Else-Quest and Hyde (2016) discussed in the context of applying quantitative methods to intersectionality research was intersectional effects. These effects refer to experiences unique to individuals at a specific intersection of identities. For instance, racism that sexual minority youth of color experience from other sexual minorities would be considered an intersectional effect. Debate exists within the literature on intersectional methods as to what can be considered a true intersectional approach (Bowleg & Bauer, 2016), with some arguing that additive effects and, to a lesser extent, interaction effects, cannot be considered intersectional. From this perspective, the analysis of multiple main effects as an element of intersectional research remains controversial.

Although previous studies have documented the associations between enacted stigma and IPV within racial minority (e.g., Cho, 2012; Stueve & O’Donnell, 2008) and sexual minority samples (e.g., Balsam & Szymanski, 2005; Gillum & DiFulvio, 2012), very little attention has been given to the potential intersectional effects of experiencing enacted stigma based on both race and sexual orientation in sexual minority youth of color. The only two studies we could locate exploring these issues did so in samples of cisgender male sexual minority populations; it was found that heterosexist discrimination and racial discrimination were each uniquely associated with an increased likelihood of experiencing physical and sexual IPV victimization, but only sexual orientation based discrimination was uniquely associated with increased risk of perpetrating either physical or sexual IPV (Finneran & Stephenson, 2013; Stephenson & Finneran, 2017). These findings suggest that experiences of enacted stigma based on both race and sexual orientation may contribute additively to IPV risk in sexual minority people of color. However, these findings require replication in other samples to clarify the role of intersecting identities in IPV within sexual-minority communities. It is particularly important to explore these issues in female assigned at birth (FAB) sexual minority youth, who experience higher rates of IPV than sexual minority youth who are assigned male at birth (Reuter et al., 2017; Whitton, Newcomb, Messinger, Byck, & Mustanski, 2019).

The Current Study

The goal of the current study was to test the influences of race-based and sexual orientation-based enacted stigma on IPV victimization and perpetration in a sample of sexual and gender minority (SGM) youth of color who were FAB. We focus on enacted stigma, rather than other minority stressors such as internalized stigma and expectations of rejection, based on evidence that external experiences of stigma are stronger predictors of IPV than internalized heterosexism (Balsam & Szymanski, 2005) and therefore particularly important to understanding IPV among FAB sexual minority youth. In accordance with previous research, we used racial discrimination as our measure of race-based enacted stigma. For enacted stigma based on sexual orientation, we used a measure of heterosexist microaggressions, which are subtle or covert acts of discrimination toward an individual based on their sexual orientation. We chose to focus on microaggressions because 1) sexual minority samples report experiencing microaggressions much more commonly than acute violent events (Kosciw, Greytak, Giga, Villenas, & Danischewski, 2016), 2) there is a greater similarity between discrimination and microaggressions compared to discrimination and overt acts of violence, 3) and heterosexist microaggressions have been associated with negative outcomes such as mental health problems (e.g., Swann, Minshew, Newcomb, and Mustanski, 2016). For our outcomes, we focused on four types of IPV: physical (i.e., slapping, hitting, etc.), psychological (i.e., name calling), sexual (i.e., forced or unwanted sex), and SGM-specific (i.e., threatening to “out” partner). Very few studies have reported on such a wide breadth of IPV measures, especially with FAB SGM youth.

Based on the greater risk perspective (Greene, 1996; King, 2016) and minority stress theory (Clark et al., 1999; Hatzenbuehler, 2009; Meyer, 2003), we hypothesized that enacted stigma based on each minority identity (i.e., racial discrimination and heterosexist microaggressions) would have bivariate associations with higher risk for IPV victimization and perpetration. We also hypothesized that, when considered together, each type of enacted stigma would contribute a unique, additive effect to the prediction of IPV. Finally, we explored whether racial discrimination and heterosexist microaggressions would interact to either exacerbate the effects of each other, such that people who experience high levels of both forms of enacted stigma will be at highest risk for IPV victimization and perpetration, or to attenuate the effects of each other, consistent with the resilience perspective (Meyer, 2010).

METHOD

Participants and Procedure

Data were drawn from FAB 400, an ongoing cohort study of 488 sexual and gender minority adolescents and young adults assigned female at birth (FAB SGM). The primary objectives of FAB 400 include describing the prevalence, risks, and protective factors of IPV among FAB SGM adolescents and young adults. FAB SGM initially completed a baseline assessment, with follow-up assessments every 6 months. Data from the present analyses come from the baseline visit.

FAB 400 consists of two merged cohorts: (1) 400 FAB SGM adolescents that were recruited directly into FAB 400 in 2016–2017, (2) and a subset of FAB SGM young adults (n = 88) recruited for Project Q2 in 2007 (Mustanski, Garofalo, & Emerson, 2010), who were invited to participate in FAB 400 in 2016–2017. At the time of enrollment into their initial cohorts, FAB SGM met the following criteria: 16–20 years old, English-speaking, and identified as a sexual or gender minority or reported same-gender attractions and/or sexual behavior. In addition, FAB SGM in FAB 400 were required to be assigned female at birth. Each cohort was recruited through a combination of venue-based, online, and peer incentivized recruitment. To determine if it was necessary to account for clustering due to recruitment chain, we calculated design effects, which quantify the extent to which the sampling error deviates from what would be expected if individuals were randomly assigned to clusters. Because the design effect for each IPV variable was less than the recommended cutoff of 2.0 (Muthen & Satorra, 1995), indicating that the small amount of non-independence present within recruitment chains would have a negligible effect on the Type I error rate, we did not account for clustering in analyses.

From November 2016 through December 2017, FAB SGM completed the FAB 400 baseline interview assessment: a psychosocial self-reported survey that included measures of individual and background characteristics, health behaviors, intimate relationships, mental health, and other related factors. FAB SGM were paid $50 for their time and all study components were approved by the Institutional Review Board at Northwestern University.

As the current study was concerned with the intersection of racial and heterosexist enacted stigma in regards to IPV, we excluded: 127 FAB SGM who identified as non-Hispanic White; six who identified as heterosexual/straight; and 90 FAB SGM who reported no romantic relationship in the past 6 months. Finally, although we did not specifically recruit couples into the cohort, there were 16 romantic dyads embedded within the cohort at baseline. We excluded one randomly selected partner from each of these dyads who reported on the same relationship. This resulted in a final analytic sample of 249 FAB SGM youth of color (see Table 1 for a detailed description of sample characteristics).

Table 1.

Participant Descriptives

	Full Sample	2007 Cohort	2016 Cohort
	N (%)	N (%)	N (%)
Race/Ethnicity
Black	125 (50.2)	50 (74.6)	75 (41.2)
Latinx	84 (33.7)	10 (14.9)	74 (40.7)
Asian	7 (2.8)	0 (0.0)	7 (3.8)
Multiracial	28 (11.2)	3 (4.5)	25 (13.7)
Other	5 (2.0)	4 (6.0)	1 (0.5)
Sexual Orientation
Lesbian	65 (26.1)	35 (52.2)	30 (16.5)
Bisexual	100 (40.2)	22 (32.8)	78 (42.9)
Pansexual	39 (15.7)	1 (1.5)	38 (20.9)
Queer	20 (8.0)	5 (7.5)	15 (8.2)
Gay	8 (3.2)	2 (3.0)	6 (3.2)
Unsure/Questioning	10 (4.0)	1 (1.5)	9 (4.9)
Asexual	2 (0.8)	0 (0.0)	2 (1.1)
Not Listed	5 (2.0)	1 (1.5)	4 (2.2)
Gender Identity
Female	198 (79.5)	60 (89.6)	137 (75.3)
Gender Non-conforming	13 (5.2)	2 (3.0)	11 (6.0)
Genderqueer	12 (4.8)	1 (1.5)	12 (6.6)
Non-binary	6 (2.4)	0 (0.0)	6 (3.3)
Transgender	8 (3.2)	1 (1.5)	7 (3.8)
Male	8 (3.2)	2 (3.0)	6 (3.3)
Not Listed	4 (1.6)	1 (1.5)	3 (1.6)
Partner Gender Identity
Cisgender Female	119 (43.1)	42 (59.2)	77 (37.6)
Cisgender Male	120 (43.5)	24 (33.8)	96 (46.8)
FAB Trans/GNC	25 (9.1)	4 (5.6)	21(10.2)
MAB Trans/GNC	10 (2.0)	1 (1.4)	9 (4.4)
Not Listed	1 (0.4)	0 (0.0)	1 (0.5)
Did Not Know	1 (0.4)	0 (0.0)	1 (0.5)
	Mean (SD)	Mean (SD)	Mean (SD)
Age	20.84 (4.07)	27.13 (1.64)	18.52 (1.28)
Psychological IPV Perpetration (range: 0–19)	1.46 (2.99)	1.82 (3.99)	1.32 (2.52)
Psychological IPV Victimization (range: 0–17)	1.42 (3.09)	2.21 (4.56)	1.13 (2.28)
Physical IPV Perpetration (range: 0–27)	0.44 (2.08)	0.87 (3.45)	0.28 (1.21)
Physical IPV Victimization (range: 0–27)	0.64 (2.52)	1.37 (4.10)	0.37 (1.52)
SGM-Specific IPV Perpetration (range: 0–21)	0.21 (1.48)	0.34 (2.57)	0.16 (0.76)
SGM-Specific IPV Victimization (range: 0–20)	0.46 (1.78)	0.52 (2.60)	0.43 (1.36)
Sexual IPV Perpetration (range: 0–21)	0.41 (1.68)	0.79 (2.81)	0.27 (0.95)
Sexual IPV Victimization (range: 0–19)	0.65 (1.91)	0.97 (2.84)	0.54 (1.41)

Notes: FAB = Female-assigned at birth. MAB = Male-assigned at birth. Trans = Transgender. GNC = Gender non-conforming.

Measures

Sexual Orientation-based Enacted Stigma

Heterosexist enacted stigma was measured using the Sexual Orientation Microaggressions Inventory (SOMI), a 19-item scale measuring frequency of experiences of anti-gay attitudes and expressions, denial of sexual identity, heterosexism, and societal disapproval that occurred within the last month (Swann et al., 2016). Example items include: “You were told your sexual orientation is just a phase” and “A heterosexual person didn’t believe that gay, lesbian, and bisexual people face discrimination.” Each item was measured on a five-point scale, from “not at all” (0) to “about every day” (4). As recommended by Swann et al. (2016), a bi-factor model was fit to the data using MPlus and the general factor from that model was exported and utilized as our measure of microaggression experiences.

Race/ethnicity-based Enacted Stigma

Racial and ethnic enacted stigma was assessed using the brief Perceived Ethnic Discrimination Questionnaire-Community Version (PEDQ-CV) (Brondolo et al., 2005). Twelve items measured frequency of perceived racial/ethnic discrimination experiences within the preceding six months (e.g., “Have others made you feel like an outsider who doesn’t fit in because of your dress, speech, or other characteristics related to your ethnicity?”). The PEDQ-CV is measured on a 5-point scale (0=Never Happened to 4=Happened Very Often). Racial discrimination scores represented the mean of all 12 items (α = .90).

Intimate Partner Violence

FAB SGM completed a battery of IPV measures about their most significant partner in the previous six months. The Sexual and Gender Minorities Conflict Tactics Scale (SGM-CTS2) is a newly developed version of the CTS2 (Straus, Hamby, Boney-McCoy, & Sugarman, 1996), modified to be culturally appropriate for the experiences of SGM samples (Dyar et al., 2019). In a psychometric evaluation of the SGM-CTS2, it showed the same factor structure as the CTS2, along with evidence of reliability and validity. Each item, which describes a specific IPV behavior or event, was asked first for perpetration (e.g., “I slapped [partner name]”) and then for victimization (e.g., “[Partner name] slapped me”). For each item, FAB SGM indicated how frequently the given event had occurred on a scale ranging from 0 (never) to 6 (more than 20 times). FAB SGM could also respond that an event had occurred but not in the previous six months. These responses were recoded as 0.

The current study used three SGM-CTS2 subscales: severe psychological violence, severe physical violence, and sexual IPV. For all of our IPV measures, we have included both Cronbach’s alpha and McDonald’s omega to assess internal consistency. McDonald’s omega is a more accurate measure of reliability with skewed data compared to Cronbach’s alpha (Dunn, Baguley, & Brunsden, 2014). Four pairs of questions measured severe psychological violence, which was characterized by acts of verbal abuse (e.g.: “I accused [partner name] of being a lousy partner”; α = .66 and ω = .83 for perpetration and α = .69 and ω = .88 for victimization). Severe physical violence was comprised of seven pairs of items of major physical assault such as “I choked [partner name]” or “I used a knife or gun on [partner name]” (Perpetration subscale: α = .86 and ω = .94; Victimization subscale: α = .83 and ω = .96). Five pairs of items captured sexual IPV (e.g.: “I used force (like hitting, holding down, or using a weapon) to make [partner name] have sex”; Perpetration subscale: α = .60 and ω = .78; Victimization subscale: α = .51 and ω = .85). Separate scores for victimization and perpetration were calculated for each subscale by taking the sum of all items.

The SGM-specific IPV Tactics Scale (Dyar et al., 2019) was developed to measure forms of IPV unique to sexual and gender minorities, which involve leveraging the societal stigma against SGM to control or cause harm to an intimate partner (Whitton et al., 2018). Five pairs of questions assessed SGM-specific IPV (e.g., “[Partner name] threatened to out me to my friends, family, or other people in my life if I didn’t do what they wanted.”). These items were measured on the same scale as the CTS-2 and scores for SGM-specific IPV perpetration and victimization were calculated by summing the five relevant items. This measure has demonstrated evidence of reliability and validity (α = .63 and ω = .91 for victimization and .85 and ω = .96 for perpetration; Dyar et al., 2019).

Analytic Plan

The associations between the two enacted stigma variables and both perpetration and victimization IPV measures were tested using negative binomial models in SPSS 25 to account for the positive skew in the distribution of the IPV measures. Age, race/ethnicity, sexual orientation, and gender identity were included in all of our models as control variables. Race/ethnicity, sexual orientation, and gender identity were dummy coded (see Table 1 for categories). We first ran separate models for each enacted stigma measure (sexual orientation-based and race/ethnicity-based) with each of the IPV outcome measures (perpetration and victimization of psychological IPV, physical IPV, SGM-specific IPV, and sexual IPV). Next, we ran multivariate models that simultaneously included both enacted stigma measures and the interaction between the two as predictors of each type of IPV. These models evaluated whether each type of enacted stigma (i.e., that based on sexual orientation and that based on racial/ethnic identity) contributed uniquely to IPV experiences when the other was taken into account, and if experiencing high levels of one type of stigma (e.g., based on sexual orientation) exacerbated or buffered the effects of experiencing the other type of stigma (e.g., based on racial/ethnic identity). In cases where the interaction effect was not significant, we excluded the interaction term from the final model. In all models, the enacted stigma predictors were standardized to facilitate comparisons of their effects and interpretation. Incident rate ratios (IRRs), reported for each predictor in the negative binomial models, represent the rate of change in the outcome for each one standard deviation change in the enacted stigma variable. Finally, we reran the multivariate models separately for cisgender and trans/GNC FAB of color and made note of any instances where the trends between the two subsamples differed.

RESULTS

Preliminary Results

Racist and heterosexist enacted stigma were moderately correlated (r = .45, df = 249, p < .001). The mean across the 12 racist discrimination items was 0.68 (SD = 0.66) for the sample, indicating that FAB SGM youth of color, on average, experienced each of the 12 specific forms of racial discrimination between “never” and “rarely” in the last 6 months. There was 16.1% (N = 40) of the sample that reported no racial discrimination in the previous 6 months. The mean for heterosexist microaggressions across the 19 items was 0.79 (SD = 0.63). That indicates that FAB SGM youth of color, on average, experienced each of the specific microaggressions between “not at all” and “1–3 times” in the previous one month. Only 2.8% (N = 7) of the sample reported experiencing no heterosexist microaggressions in the previous month.

Bivariate Associations between Enacted Stigma and Intimate Partner Violence

Our first hypothesis was that there would be significant bivariate associations, while controlling for demographic differences, between the enacted stigma measures and each of the IPV outcomes. We found in these analyses that stigma was significantly associated with almost all forms of IPV perpetration. This was true for heterosexist enacted stigma and psychological (IRR = 1.58, p = .001), SGM-specific (IRR = 3.47, p < .001), and sexual IPV perpetration (IRR = 2.08, p = .003). It was also true for racist enacted stigma and psychological (IRR = 1.42, p = .001), physical (IRR = 1.69, p = .002), SGM-specific (IRR = 2.04, p = .017), and sexual IPV perpetration (IRR = 1.75, p = .015). The association between heterosexist stigma and physical IPV perpetration did not reach significance (IRR = 1.54, p = .58). As hypothesized, for each type of enacted stigma, more stigma experiences were associated with more perpetration of IPV with the FAB SGM youth of color’s primary partner.

Bivariate comparisons between enacted stigma measures and IPV victimization outcomes, controlling for demographic covariates, were also significant. Heterosexist enacted stigma was significantly associated with psychological (IRR = 1.52, p = .002), physical (IRR = 1.97, p < .001), SGM-specific (IRR = 2.57, p < .001), and sexual IPV victimization (IRR = 2.11, p < .001). Racist enacted stigma was significantly associated with psychological (IRR = 1.57, p < .001), physical (IRR = 2.00, p < .001), SGM-specific (IRR = 2.96, p < .001), and sexual IPV victimization (IRR = 1.75, p = .001). For all associations, more experiences of enacted stigma were associated with higher IPV victimization.

Multivariate Associations between Enacted Stigma and Intimate Partner Violence

Our second hypothesis was that each type of enacted stigma would contribute a unique additive effect when included in the same multivariate model. We also had the exploratory goal of testing for significant interaction effects of the two forms of stigma to determine if there was an exacerbating or attenuating effect of experiencing both forms of stigma on IPV. Results from multivariate models, which included both enacted stigma measures in the same model as predictors of IPV perpetration, are shown in Table 2. In the model where severe psychological IPV was the outcome, there was a significant interaction between heterosexist enacted stigma and racist enacted stigma (IRR = 0.81, p = .036). Simple slopes analysis revealed that higher levels of racist enacted stigma were associated with significantly higher perpetration of psychological IPV for FAB SGM of color at −1 SD (IRR = 1.65, p = .003) and the mean (IRR = 1.34, p = .013) on experiences of heterosexist enacted stigma. However, for FAB SGM youth at +1 SD on heterosexist enacted stigma, more experiences of race/ethnicity-based enacted stigma were not associated with an increase in IPV perpetration (IRR = 1.09, p = .52). That is, consistent with the resilience perspective, for those experiencing the highest levels of one form of stigma, experiencing more of the other form of stigma was not associated with additional increases in perpetration (see Fig. 1).

Table 2.

Multivariate Models for Enacted Stigma on Perpetration of IPV

	Psychological IPV			Physical IPV			SGM-Specific IPV			Sexual IPV
	IRR	p-value	VIF	IRR	p-value	VIF	IRR	p-value	VIF	IRR	p-value	VIF
Heterosexist Enacted Stigma	1.51	.007	1.46	1.15	.579	1.34	3.02	<.001	1.34	1.78	.032	1.34
Racist Enacted Stigma	1.39	.009	1.40	1.60	.010	1.27	1.25	.356	1.27	1.25	.370	1.27
Heterosexist Enacted Stigma*Racist Enacted Stigma	0.81	.036	1.33	-	-	-	-	-	-	-	-	-
Age	1.05	.232	1.42	1.07	.248	1.38	0.98	.847	1.33	1.13	.035	1.38
Race/Ethnicity
Black (referent)	-	-	-	-	-	-	-	-	-	-	-	-
Latinx	0.60	.040	1.25	0.26	.001	1.19	1.87	.265	1.19	0.89	.809	1.19
Asian	0.17	.031	1.13	-a	-a	-a	-a	-a	-a	-a	-a	-a
Other	1.59	.488	1.08	0.49	.371	1.08	-b	-b	-b	0.54	.525	1.08
Multiracial	0.47	.091	1.13	0.37	.409	1.10	1.40	.725	1.10	0.50	.331	1.10
Sexual Orientation
Lesbian/Gay (referent)	-	-	-	-	-	-	-	-	-	-	-	-
Bisexual/Pansexual	0.78	.403	1.49	0.58	.266	1.49	0.26	.057	1.49	1.16	.780	1.49
Queer	0.31	.012	1.39	0.31	.383	1.38	0.10	.066	1.38	0.68	.705	1.38
Other	1.00	.994	1.30	0.83	.810	1.28	2.03	.469	1.27	1.78	.484	1.28
Gender Identity
Transgender/GNC (referent)	-	-	-	-	-	-	-	-	-	-	-	-
Cisgender	1.21	.445	1.21	4.97	.004	1.20	0.19	.002	1.19	0.97	.956	1.20

Notes: IRR = Incident rate ratio. VIF = Variance inflation factor. IPV = Intimate partner violence. GNC = Gender non-conforming. Significant enacted stigma effects in bold.

^aNo Asian-identifying participants reported physical, SGM-specific, or sexual IPV.

^bNo participants who reported their race/ethnicity as “other” reported SGM-specific IPV.

Figure 1.

Interaction of Racial/Ethnic and SGM Enacted Stigma on Perpetration of Psychological Intimate Partner Violence. SD = Standard Deviation.

In the models for physical, SGM-specific, and sexual IPV perpetration outcomes, the interaction terms between the two forms of enacted stigma were not significant. Further, the results did not support unique additive effects of each form of stigma. In the model for physical IPV perpetration, only racist enacted stigma was significantly associated, indicating that heterosexist stigma did not have a unique association with perpetrating physical IPV once experiences of racist stigma were controlled. The observed IRR indicated that FAB SGM youth who were a SD above the mean on race/ethnicity-based stigma reported 60% more physical IPV perpetration than FAB SGM youth who were at the mean. In contrast, in models of SGM-specific and sexual IPV perpetration, only heterosexist enacted stigma, and not racist stigma, was significant. FAB SGM youth of color who were one SD above the mean on heterosexist stigma reported 202% more perpetration of SGM-specific IPV and 78% higher sexual IPV perpetration compared to FAB SGM youth at the mean.

Multivariate models that included the associations of both forms of enacted stigma with IPV victimization are shown in Table 3. Interaction terms were not significant for any of the victimization models and were dropped. Unique additive effects of both forms of enacted stigma were observed for physical and SGM-specific IPV victimization only. FAB SGM youth of color who experienced high levels of heterosexist stigma (+1SD) reported 63% more physical IPV victimization and 59% more SGM-specific IPV victimization compared to those who experienced average levels of heterosexist stigma. Further, controlling for heterosexist stigma, FAB SGM youth who were a standard deviation above the mean on racist stigma reported 68% more physical IPV victimization and 128% more SGM-specific IPV victimization compared to FAB SGM youth experiencing average levels of racist stigma. In contrast, only racist enacted stigma was uniquely associated with psychological IPV victimization when both forms of enacted stigma were included in the model. FAB SGM youth who were one SD above the mean on experiences of racist enacted stigma reported 48% more psychological IPV victimization compared to FAB SGM youth at the mean. Sexual IPV victimization was only associated with heterosexist enacted stigma in the multivariate model. FAB SGM youth who were a SD above the mean on heterosexist stigma experiences were 74% higher on sexual IPV victimization compared to FAB SGM youth at the mean.

Table 3.

Multivariate Models for Enacted Stigma on Experiencing IPV Victimization

	Psychological IPV			Physical IPV			SGM-Specific IPV				Sexual IPV
	IRR	p-value	VIF	IRR	p-value	VIF	IRR	p-value	VIF	IRR		p-value	VIF
Heterosexist Enacted Stigma	1.17	.238	1.36	1.63	.009	1.34	1.59	.036	1.36	1.74		.010	1.36
Racist Enacted Stigma	1.48	.001	1.29	1.68	.004	1.27	2.28	<.001	1.29	1.35		.081	1.29
Age	1.08	.045	1.41	1.22	.001	1.38	1.03	.672	1.34	1.12		.024	1.39
Race/Ethnicity
Black (referent)	-	-	-	-	-	-	-	-	-	-		-	-
Latinx	0.65	.107	1.25	0.35	.008	1.19	2.20	.082	1.24	0.68		.340	1.25
Asian	0.22	.072	1.12	-a	-a	-a	1.74	.572	1.11	0.33		.427	1.12
Other	1.74	.351	1.08	0.94	.942	1.08	-b	-b	-b	0.35		.197	1.08
Multiracial	0.73	.558	1.13	1.24	.830	1.10	2.12	.153	1.12	1.06		.921	1.13
Sexual Orientation
Lesbian/Gay (referent)	-	-	-	-	-	-	-	-	-	-		-	-
Bisexual/Pansexual	0.60	.093	1.49	0.64	.353	1.49	4.09	.005	1.25	2.13		.076	1.25
Queer	0.69	.533	1.39	0.81	.833	1.38	-c	-c	-c	-c		-c	-c
Other	0.92	.863	1.30	0.96	.962	1.28	2.95	.161	1.19	1.29		.677	1.20
Gender Identity
Transgender/GNC (referent)	-	-	-	-	-	-	-	-	-	-		-	-
Cisgender	1.05	.865	1.21	2.51	.128	1.20	1.67	.242	1.06	0.82		.557	1.07

Notes: IRR = Incident rate ratio. VIF = Variance inflation factor. IPV = Intimate partner violence. GNC = gender non-conforming. Significant enacted stigma effects in bold.

^aNo Asian-identifying participants reported physical IPV.

^bNo participants who reported their race/ethnicity as “other” reported SGM-specific IPV.

^cNo queer-identifying participants reported SGM-specific or sexual IPV.

When multivariate models were run separately for cisgender and trans/GNC FAB, trends remained consistent between the two subsamples with two exceptions: 1) the association between heterosexist stigma and sexual IPV perpetration (full sample IRR = 1.78, p = .032; trans/GNC IRR = .82, p = .811; cisgender IRR = 1.76, p = .009) and 2) the association between racist stigma and SGM-specific IPV victimization (full sample IRR = 2.28, p < .001; trans/GNC IRR = .66, p = .552; cisgender IRR = 2.43, p < .001). In both cases, results suggest that the significant associations found in the full sample are accurate only for cisgender FAB. The consistency between the two subsamples across the remaining significant effects suggest that the experiences between cisgender and trans/GNC FAB for these associations are similar.

DISCUSSION

In the present study, we examined the effects of self-reported heterosexist and racist enacted stigma on engagement in IPV perpetration and experiences of IPV victimization among racial/ethnic minority FAB SGM youth. By examining enacted stigma based on both minority identities simultaneously, we were able to evaluate the unique effects of each form of stigma on IPV, test for additive effects of experiencing both, and explore whether experiencing enacted stigma based on one identity either exacerbated or attenuated the effects of stigma experiences based on the other identity.

Consistent with minority stress theory (Clark et al., 1999; Hatzenbuehler, 2009; Meyer, 2003), our results confirmed that when each form of stigma was examined separately, experiencing more enacted stigma based on either identity was associated with more IPV perpetration and victimization in all four of the IPV domains that we measured: psychological, physical, sexual, and SGM-specific. When both types of enacted stigma were considered together, however, there was not a consistent pattern regarding their effects on IPV. The differences in how heterosexist and racist enacted stigma were operationalized, as microaggressions and discrimination, respectively, lend further caution to how the two forms of stigma are compared, and we remain cognizant of that throughout this discussion.

Enacted stigma based in race/ethnicity was associated with all forms of IPV in univariate models, and showed unique associations with five of the eight IPV outcomes when controlling for heterosexist stigma experiences. These findings extend previous evidence that racial discrimination is associated with more psychological and physical IPV perpetration and victimization for heterosexual women of color (Cho, 2012; Stueve & O’Donnell, 2008; Waltermaurer et al., 2006) to sexual minority youth of color. The significance of race-based enacted stigma, for both victimization and perpetration of IPV, may be a result of the wide-ranging negative consequences of experiencing racial discrimination, including less job security (Pager & Shepherd, 2008), poorer health (Nazroo, 2003), and substance use problems (Gerrard et al., 2012), among other negative outcomes, all of which might exacerbate physically and psychologically abusive relationship dynamics. It is also possible that the internalized feelings of stigma and stress generated from these experiences (Clark et al., 1999) negatively affect relationship processes (LeBlanc et al., 2015), including by reducing couples’ capacities to constructively and non-violently manage relationship conflicts (Neff & Karney, 2009).

Similarly, heterosexist microaggressions were associated with all forms of IPV in univariate models, replicating previous research indicating that heterosexist stigma experiences raise risk for psychological, physical, and sexual IPV victimization and perpetration (Balsam & Szymanski, 2005; Edwards & Sylaska, 2013; Finneran & Stephenson, 2013). We extended these findings by showing that heterosexist stigma was uniquely associated with SGM-specific IPV perpetration and sexual IPV victimization and perpetration among SGM of color when racist discrimination experiences were controlled. Like racial discrimination, enacted stigma based on sexual orientation has a wide range of effects that potentially harm relationship functioning, including poorer mental health (Feinstein, Goldfried, & Davila, 2012; Russell, Ryan, Toomey, Diaz, & Sanchez, 2011; Swann, Stephens, Newcomb, & Whitton, 2020) and more substance use problems (Huebner, Thoma, & Neilands, 2015; Mereish, O’Cleirigh, & Bradford, 2014). Particularly noteworthy was the strong association between heterosexist enacted stigma and SGM-specific IPV perpetration. The SGM-specific IPV scale was newly developed for FAB 400 and, to our knowledge, enacted stigma has not been tested with any similar measures in previous research. This finding suggests that experiencing more enacted stigma associated with one’s sexual orientation might make one more likely to pass similar negative interactions onto one’s romantic partner. Alternatively, living in environments where heterosexist stigma is more common might make SGM-specific IPV behaviors, like threats of outing one’s partner, more effective tools to manipulate a romantic partner and encourage their continued use.

The main contribution of this study, however, was its simultaneous examination of racist and heterosexist enacted stigma. Only one type of enacted stigma was uniquely associated with many of the forms of IPV in multivariate models. The two forms of stigma were moderately correlated and that may have contributed to why we saw fewer additive effects. The relatively low variance inflation factors we saw within our models suggest that this may be a minor concern, but the conceptual similarities between racist and heterosexist enacted stigma can also potentially bias the multivariate results we found (Darlington, 1968; Vatcheva, Lee, McCormick, & Rahbar, 2016). However, it is also conceivable that some forms of stigma are more strongly associated with particular forms of IPV. For instance, heterosexist stigma may have been more reliably associated with SGM-specific and sexual IPV because they are both associated with sexuality through conceptions of sexual identity and enacted sexual behavior, respectively. Heterosexist stigma may have a unique effect on internalized homophobia that in turn degrades relationship dynamics specific to sexuality. The lack of previous research comparing the effects of racist and heterosexist stigma on IPV and the correlated and conceptual similarity of these measures suggests that these results should be replicated and expanded upon in future research in order to better interpret the specific pattern of findings and to make more concrete claims of their relative importance to one another in their association with IPV.

Results supported an additive model of the impact of racist and heterosexist enacted stigma on physical and SGM-specific IPV victimization. For these outcomes both individual and environmental factors could increase risk for experiencing violent relationships for FAB sexual minority youth of color who experience enacted stigma. For example, people who live in places where experiences of enacted stigma are common may have less access to resources and fewer avenues of support to help them when they experience IPV victimization (Guruge & Humphreys, 2009; Hausmann et al., 2008). Greater exposure to enacted stigma might also result in more mental health problems (Carter et al., 2017; Kulick et al., 2017; Swann, Forscher, Bettin, Newcomb, & Mustanski, 2019) that increase risk of IPV victimization, either by making it more difficult to extricate oneself from a violent relationship dynamic or making FAB sexual minority youth of color more vulnerable targets for victimization. The unique associations with both forms of stigma illustrate the multiple potential pathways FAB sexual minority youth of color have toward negative IPV outcomes.

Contrary to predictions, we found very little evidence that experiences of racist and heterosexist enacted stigma affect each other’s associations with IPV. In fact, psychological IPV perpetration was the single outcome for which there was a significant interaction between racist and heterosexist enacted stigma. Although those FAB SGM youth who reported high levels of both types of enacted stigma reported the highest levels of psychological IPV perpetration, decomposition of this interaction revealed an attenuating effect of experiencing high levels of both forms of stigma that resulted in lower psychological IPV perpetration than would be expected from the observed main effects alone.

This may be an indication of a resilience effect (Meyer, 2010; Moradi et al., 2010), such that FAB SGM youth who experience high levels of racist and heterosexist stigma have become less vulnerable to their negative impact through development of effective coping strategies that generalize across the two domains of oppression. It might also reflect a ceiling effect for these FAB SGM youth, in which more experiences of enacted stigma may reach the maximum limit at which they can increase risk for psychological IPV perpetration. We would caution that with a p-value of .036, this interaction effect may not have reached significance if we had adjusted for family-wise error and should be interpreted with caution. Additional research is needed to replicate this finding, explore links with coping and resilience, and identify levels at which more experiences of stigma are no longer associated with the same increases in psychological IPV.

Understanding how people of multiple minority statuses uniquely experience minority stressors is integral to an intersectional understanding of how they are negatively impacted by those stressors (Bauer, 2014; McCall, 2005). Our results suggest that heterosexist and racist stigma may affect the occurrence of IPV differently, but experiencing both increases the risk that FAB SGM youth of color are under. Our results only supported the greater risk perspective (Bowleg et al., 2003) for specific outcomes, but suggest overall that experiencing stigma based on multiple minority statuses comes with a wide range of heightened risk for IPV. At the same time, while we analyzed main effects and interaction effects in accordance with Else-Quest and Hyde’s (2016) suggestions for applying quantitative methods to the study of intersectionality, we did not analyze associations with uniquely intersectional effects and IPV, so there is a limit to the conclusions we can draw about intersectional experiences of stigma on IPV. There is an important contingent within the literature on intersectionality that would argue that the inclusion of these effects is necessary for research to be considered truly intersectional (Bowleg & Bauer, 2016). In the present study, we identified potential pathways by which SGM at the intersection of multiple minority identities may be vulnerable to the negative effects of multiple forms of stigma, but we did not include the effects of enacted stigma that may be unique to their intersectional identity. Future research on the effects of enacted stigma on IPV in FAB SGM of color should incorporate these intersectional effects to create a fuller picture of the experiences of FAB SGM youth of color.

Study limitations include the cross-sectional nature of the data, which prohibits conclusions regarding direction of effects. It is not clear whether experiences of enacted stigma predated IPV perpetration and victimization or if IPV experiences may raise perceptions of racial and sexual orientation-based stigma. Because sexual orientation-based enacted stigma was operationalized as heterosexist microaggressions in the previous month, whereas race/ethnicity-based enacted stigma was operationalized as perceived racist discrimination in the previous six months, observed differences between the associations of these variables with IPV may reflect differing effects of microaggressions versus discrimination or the time frame each was measured, rather than stigma due to sexual orientation versus race. We did not account for differences in socioeconomic status or family composition in our analyses. These factors may be particularly important for contextualizing the effects of racist stigma (Williams, 1996). Also, despite the representation of multiple racial and sexual identities in our sample, we did not have power to explore whether associations between enacted stigma and IPV differ between particular combinations of identities (i.e., the experiences of lesbian-identifying cisgender Asian women versus pansexual-identifying Black transmen). There are likely important nuances in how specific subgroups experience IPV that we have missed by collapsing across them. Future research should not only identify differences in IPV experiences between these subgroups, but should focus on the degree to which differences in experiences of multiple forms of enacted stigma mediate their differing experiences of IPV within their romantic relationships.

Despite these limitations, the present findings highlight the importance of attending to the intersectional influences of multiple minority identities when studying IPV in sexual minority populations. Among the SGM youth of color in this sample, enacted stigma based on each of their identities was associated with IPV, suggesting that experiences of racism and heterosexism may contribute to the elevated rates of IPV observed in this population (Reuter et al., 2017; Whitton et al., 2018). Future research should expand on the current study by exploring the specific mechanisms through which exposure to enacted stigma could contribute to experiences of IPV victimization and perpetration, including internalized stigma (Clark et al., 1999; Hatzenbuehler, 2009; Meyer, 2003), mental health (Carter et al., 2017; Espelage, Aragon, Birkett, & Koenig, 2008; Swann et al., 2020), substance use (Clark, 2014; Duncan, Hatzenbuehler, & Johnson, 2014; Unger, Schwartz, Huh, Soto, & Baezconde-Garbanati, 2014), and restricted access to resources (Barnett & Berchick, 2017; Hausmann et al., 2008). It will also be important to identify protective factors that might mitigate pathways between enacted stigma and IPV, to inform intervention and prevention resources aimed at reducing IPV for FAB sexual minority youth of color. The development of these efforts should account for the complex experiences of people with multiple minority identities and the unique exposure to enacted stigma that they experience.