Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2019 Nov 1.
Published in final edited form as: Psychol Violence. 2018 Nov;8(6):669–679. doi: 10.1037/vio0000218

The Effects of Intersecting Stigma: A Longitudinal Examination of Minority Stress, Mental Health, and Substance Use among Black, Latino, and Multiracial Gay and Bisexual Men

Devin English 1, H Jonathon Rendina 1,2,3,*, Jeffrey T Parsons 1,2,3
PMCID: PMC6415673  NIHMSID: NIHMS1015898  PMID: 30881729

Abstract

Objective:

Although Black, Latino, and multiracial gay and bisexual men (GBM) are disproportionately affected by health inequities facing GBM more broadly in the United States (CDC, 2017), there is a dearth of research examining how intersectional stigma affects psychological and behavioral outcomes such as depressive and anxiety symptoms and substance use. Based in minority stress and intersectionality theories, this study examined the main and intersectional effects of racial discrimination and gay rejection sensitivity on emotion regulation difficulties, depressive and anxiety symptoms, and later drug use and heavy drinking.

Method:

We collected longitudinal data from 170 GBM of Black, Latino, or multiracial descent. Measurements included baseline racial discrimination, gay rejection sensitivity, and emotion regulation difficulties, 6 month depressive and anxiety symptoms, and baseline to 12 month heavy drinking and drug use. We analyzed data using longitudinal structural equation models.

Results:

Our results indicated that racial discrimination and its interaction with gay rejection sensitivity were significantly associated with higher levels of emotion regulation difficulties, which predicted higher levels of depressive and anxiety symptoms at 6 months, which, in turn, predicted higher levels of heavy drinking, but not drug use, at 12 months. Moreover, the total indirect effect from the stigma variables to heavy drinking was statistically significant.

Conclusions:

These findings indicate that it is critical for researchers and clinicians to consider the effects of intersecting racial and sexual minority stress on emotion regulation in the persistence of psychological and behavioral health inequities facing Black, Latino, and multiracial GBM.

Keywords: Gay and Bisexual Men, Racial Discrimination, Sexual Minority Stigma, Minority Stress, Intersectionality

Black, Latino, and multiracial lesbian, gay, bisexual, transgender, queer, and 2 spirit (LGBTQ2S) people have been disproportionately affected by recent spikes in hate crimes and violence targeting LGBTQ2S communities (Waters, Pham, Convery, & Yacka-Bible, 2018). Among the community members who have been most affected by identity-related violence are Black, Latino, and multiracial gay and bisexual men (GBM), who are also at highest risk for health outcomes such as mood and anxiety disorders (Bostwick, Boyd, Hughes, & McCabe, 2010), substance use (Talley, Hughes, Aranda, Birkett, & Marshal, 2014), and several other serious and/or chronic illnesses (e.g., CDC, 2017). Despite the evidence of co-occurring victimization and persistent health inequities, there is a dearth of research examining how stigma based on marginalized identities among Black, Latino, and multiracial GBM contributes to these trends (Earnshaw, Bogart, Dovidio, & Williams, 2013). As a result, intervention programs that serve to promote equitable health outcomes often do not reach these men (Bush et al., 2016). In order to design effective intervention programs to stem these alarming trends, it is critical that research identifies factors such as violence expressed through intersecting racism and homophobia that inequitably affect Black, Latino, and multiracial GBM and lead to risk for internalizing symptoms and substance use (Meyer, 2012). Thus, with the present study we sought to examine the main and intersectional effects of racial and sexual minority stigma on emotion regulation difficulties, depressive and anxiety symptoms, and later drug use and heavy drinking.

A substantial body of research has utilized minority stress theory to examine stigma as a form of violence and source of inequities in psychological health and substance use among GBM broadly. Minority stress theory (Meyer, 2003) posits that several stress processes occur as a result of minority status (e.g., sexual orientation), including distal minority stressors in the form of discrimination and violence as well as proximal stressors such as expectations of rejection that predict negative health outcomes. Hatzenbuehler (2009) described how this stress process manifests among sexual minority individuals by detailing potential psychological pathways through which stigma-related stress leads to psychological and behavioral outcomes. This framework posits that the distal stigma-related stressors detailed in Meyer (2003) lead to increased experiences of proximal stressors (e.g., rejection sensitivity), which are linked to psychological processes such as emotion regulation difficulties that, in turn, lead to depressive symptoms, anxiety, and substance use (Pachankis et al., 2015).

There is a body of research that provides evidence for this stress process that shows sexual minority stigma is associated with emotion regulation difficulties (Pachankis et al., 2015; Rendina et al., 2017), depressive and anxiety symptoms (Lewis, Derlega, Griffin, & Krowinski, 2003), substance use (Hatzenbuehler, Nolen-Hoeksema, & Erickson, 2008), post-traumatic stress (Herek et al., 1999), suicidality (Hershberger & D’Augelli, 1995), and lower self-esteem (Huebner, Rebchook, & Kegeles, 2004). In addition, there is an emerging literature suggesting that racial discrimination experienced by Black, Latino, and multiracial GBM is associated with the negative psychological and behavioral health outcomes (e.g., Bogart et al., 2011; Bostwick et al., 2014; Yoshikawa, Wilson, Chae, & Cheng, 2004). Moreover, research also shows that these men are at unique risk for experiencing minority stress because they face sexual minority discrimination within their racial group and racial discrimination within their sexual minority group (Choi, Paul, Ayala, Boylan, & Gregorich, 2013).

This empirical work notwithstanding, there has been relatively less empirical inquiry regarding how racial and sexual minority stigma interact to affect the stress process that puts Black, Latino, and multiracial GBM at unique risk for inequities in psychological symptoms and substance use. Intersectionality theory provides an ideal lens through which to consider this process as it asserts that multiple social identities such as race, gender, socioeconomic status (SES), sexual orientation, and/or disability intersect at the individual level and reflect multiple and reciprocal systems of societal oppression (e.g., racism, heterosexism) that affect Black, Latino, and multiracial GBM (Collins, 1998; Crenshaw, 1991). Indeed, the intersectionality framework posits that social identities are not isolated from one another and simply additive, but rather are interdependent and mutually constitutive (Collins, 2016). As such, rather than conceptualizing stigma-related stress in a way that may assume minority stress experiences can be broken down as a sum of stress associated with an individual’s privileged or disadvantaged identities (e.g., racial stress + sexual minority stress), an intersectional approach suggests that different forms of stigma should be tested as multiplicative, such that the effects of one form of minority stress are dependent on that of another form of minority stress (Bowleg, 2008).

Extant, albeit limited, research suggests that racial and sexual minority stigma may interact to predict psychological and behavioral outcomes among Black, Latino, and multiracial GBM. For example, in a qualitative study of 71 Black GBM with HIV, HIV-related stigma, homophobia, and racism were interrelated and associated with participants’ sexual risk behavior, reluctance to obtain HIV testing or care, lower adherence to treatment, and less disclosure of a positive HIV status to sexual partners (Arnold, Rebchook, & Kegeles, 2014). In addition, results from Bostwick and colleagues (2014) and Grollman (2014) suggest that discrimination based on sexual orientation, gender, and race are most impactful when experienced in the presence of one or several other forms of discrimination, though their hypotheses were tested in an additive manner (i.e., the sum of different types of discrimination predicting outcomes). In an intersectional quantitative study of 181 Black GBM with HIV, researchers found that racial, sexual orientation, and HIV discrimination interacted to predict higher levels of depressive symptoms, though none of the individual main effects from these forms of discrimination were associated with depressive symptoms (Bogart et al., 2011). Although these studies provide important evidence linking the intersection of multiple forms of sigma facing Black, Latino, and multiracial GBM, we are unaware of any studies that have measured multiple forms of minority stress and their interaction longitudinally while also modelling potential mediators of their effects on relevant health outcomes as detailed in theories of minority stress and intersectionality.

We conducted the present study to inform intervention efforts to stem the negative effects of racist and homophobic violence by deepening the understanding of how racial and sexual minority stigma interact to contribute to health inequities for Black, Latino, and multiracial GBM. Specifically, we examined the main and interaction effects of racial discrimination—an enacted form of racial stigma—and gay rejection sensitivity—an anticipated, proximal form of sexual minority stigma (Pachankis, Goldfried, & Ramrattan, 2008)—within an intersectional, minority stress model. We focused on the extent to which these stressors are associated with emotion regulation difficulties and internalizing symptoms (i.e., depressive and anxiety symptoms) that, in turn, are associated with increased substance use over time. We took an intersectional quantitative approach by estimating an interaction between racial discrimination and gay rejection sensitivity (Bowleg & Bauer, 2016). In doing so, the present study combines minority stress theory and intersectionality theory to detail a stress process experienced by Black, Latino, and multiracial GBM. We hypothesized that: 1) both gay rejection sensitivity and racial discrimination would have positive main effects on emotion regulation difficulties which would predict higher levels of depressive and anxiety symptoms that would, in turn, predict increases in drug and alcohol use; and 2) the interaction between gay rejection sensitivity and racial discrimination would be associated with emotion regulation difficulties such that, at higher levels of each stressor there would be multiplicatively higher rates of emotion regulation difficulties that would predict internalizing symptoms and increases in drug and alcohol use.

Methods

We conducted the analyses for this paper using baseline, 6-month, and 12-month follow-up data from Pillow Talk, a longitudinal study that focused on issues related to sexual compulsivity among highly sexually active GBM in New York City. The primary goal of the study was to enroll sexual minority men who varied based on the amount of problems caused by high amounts of sexual activity. Although the study recruited sexual minority men generally, all of the men who enrolled identified as gay/queer or bisexual. Of the 376 men enrolled in the study, 170 (45.2%) identified as Black, Latino, or multiracial—of them, 76 identified as Black (44.7%), 51 as Latino (30.0%), and 43 as multiracial (25.3%). These 170 participants provided at least baseline data and were the sample of interest for our analyses.

Participants and Procedures

We recruited participants through the following strategies: (1) respondent-driven sampling, (2) Internet-based advertisements on social and sexual networking websites, (3) email blasts through New York City gay sex party listservs, and (4) active recruitment in New York City venues such as gay bars/clubs and ongoing gay community events. Participants completed a brief interview screener over the phone to confirm eligibility. Criteria were: (1) at least 18 years of age, (2) biologically and self-identified as male, (3) at least nine male sexual partners in the prior 90 days (i.e., highly sexually active; Pachankis et al., 2015), (4) self-identification as gay, bisexual, or another non-heterosexual identity, (5) able to complete assessment in English, and (6) daily access to the Internet to complete web-based study modules. We confirmed eligibility criteria at the in-person baseline appointment.

We excluded participants from the study if we found evidence of significant cognitive impairment, as indicated by a score of 23 or lower on the Mini-Mental Status Examination (MMSE; Folstein, Folstein, & McHugh, 1975), or found evidence of active symptoms of suicidality or psychosis on the Structured Clinical Interview for the DSM-IV-IR (SCID; First, Spitzer, Gibbon, & Williams, 2002), as these symptoms would have interfered with their participation and may have limited their ability to provide informed consent. Participants completed both web-based assessments and in-office assessments. Study staff confirmed participant eligibility over the phone through a short series of questions and obtained the participants’ informed consent. Overall, 79% of participants who completed the phone prescreening were eligible for the study, and of those screened, 62% enrolled in the study.

Once consented over the phone, a staff member sent the participant a link to complete an initial 1-hour, web-based survey. Study staff re-consented participants during follow-up, in-person assessments in the research office. This manuscript focuses on self-report data collected during the baseline and 6-month at-home surveys as well as the baseline and 12-month TLFB data collected in-office. We compensated participants $50 for the baseline assessment, $55 for the 6 month assessment, and $60 for the 12 month assessment. The Institutional Review Board of the City University of New York (CUNY) reviewed and approved all study procedures.

Measures

Demographics.

Participants reported several demographic characteristics including age, race/ethnicity, sexual orientation, relationship status, and employment status. Participants also documented their HIV+ status or had an HIV test to confirm negative status. We referred three participants who received a preliminary HIV positive result to confirmatory testing and linkage to care a local HIV clinic. Data from these three participants are not part of the present analyses.

Racial Discrimination.

We used the race version of the Everyday Discrimination Scale (Williams, Yan, Jackson, & Anderson, 1997), which asks about experiences in daily life that occur as a result of the participant’s race. The scale consists of nine items (e.g., “you are treated with less courtesy than other people”) that are rated on a scale from 1 (never) to 6 (almost every day). Thus, frequency index scores range from 9 to 54. The items ask about general experiences of racial discrimination that are not specific to stereotypes associated with a given racial group in the U.S. The scale has shown validity in its association with psychological and physical health outcomes among a multiracial sample (Williams et al., 1997). Items served as indicators of a racial discrimination latent variable in the models described below. In the present study, this scale demonstrated good internal consistency (α=0.95) at baseline among the analytic sample.

Gay Rejection Sensitivity.

We administered the Gay Rejection Sensitivity scale, a measure used to assess internalized stigma among gay and bisexual men (Pachankis, Goldfried, & Ramrattan, 2008). The measure consists of 14 vignettes (e.g., “You are in a locker room in a straight gym. One guy nearby moves to another area to change clothes.”) to which participants respond regarding how concerned they would be if the situation occurred as a result of their sexual orientation from 1 (very unconcerned) to 6 (very concerned) and how likely it would be that the situation occurred because of their sexual orientation from 1 (very unlikely) to 6 (very likely). Consistent with the rejection sensitivity paradigm (Downey & Feldman, 1996), we multiplied concern and likelihood responses for each vignette. These scores served as indicators of a rejection sensitivity latent variable in the models described below. We excluded item 1 (“You bring a male partner to a family reunion. Two of your old-fashioned aunts don’t come talk to you even though they see you”), item 9 (“Only you and a group of macho men are on a subway train late at night. They look in your direction and laugh”), item 5 (“You are going to have surgery, and the doctor tells you that he would like to give you an HIV test”), and item 3 (“You’ve been dating someone for a few years now, and you receive a wedding invitation to a straight friend’s wedding. The invite was addressed only to you, not you and a guest.”) because they had factor loadings below our set threshold of 0.60. Several of these items had relatively low factor loadings in the development and validation of this measure. The overall measure showed both convergant and discriminant validity with parental rejection, internalized homophobia, and unassertive interpersonal behavior among gay men (see Panchankis et al., 2008). In the present study, the scale showed good internal consistency (α=0.91) at baseline.

Emotion regulation difficulties.

Participants completed the 36-item Emotion Regulation Difficulties Scale (DERS; Gratz & Roemer, 2004), which measures general problems in regulating emotions. Participants responded on a scale from 1 (almost never [0–10 %]) to 5 (almost always [91–100 %]). Subscales for the DERS include the non-acceptance subscale (e.g., “when I’m upset, I feel like I am weak”), the goals subscale (e.g., “When I’m upset, I have difficulty getting work done”), the impulse subscale (e.g., “When I’m upset, I feel out of control”), the awareness subscale (e.g., “I am attentive to my feelings”), the strategies subscale (e.g., “When I’m upset, I believe that wallowing in it is all I can do”), and the clarity subscale (e.g., “I am confused about how I feel”). The scale has shown construct validity and predictive validity with other emotion regulation measures and behavioral outcome scales, respectively (Gratz & Roemer, 2004). We excluded the awareness subscale in this study because it had a low factor loading (0.25), which is consistent with past studies of the DERS factor structure in which the awareness subscale showed weak latent factor intercorrelations and it was recommended to be removed from the full scale DERS (Bardeen, Fergus, & Orcutt, 2012). Subscale scores served as indicators of an emotion regulation difficulties latent variable in the models described below. In the present study, this scale demonstrated good internal consistency (α=0.90) at baseline.

Internalizing Symptoms.

Participants completed the anxiety and depression items of the Brief Symptom Inventory (BSI; Derogatis & Melisaratos, 2009). Each of the subscales consists of six items that measure symptoms of depression (e.g., “feeling worthlessness”) or anxiety (e.g., “feeling tense or keyed up”) in the prior week. Response options range from 0 (not at all) to 4 (extremely), and each subscale score is calculated by averaging across its six items. The BSI has shown good convergent and construct validity (Derogatis & Melisaratos, 2009). Subscale scores served as indicators of an internalizing symptoms latent variable in the present study. In the present study, the depression (α=0.89) and anxiety (α=0.88) subscales demonstrated good internal consistency at baseline.

Drug Use and Heavy Drinking.

Participants completed a 42-day timeline follow-back (TLFB) interview of their drug and alcohol use (Sobell & Sobell, 1996). Using a calendar, a research assistant (RA) coded whether any substance use occurred on a given day. On days when substance use occurred, RAs coded the presence of heavy drinking (five or more alcoholic drinks) and/or the type of drug used (cocaine, crack, ecstasy, GHB, hallucinogens, heroin/other opiates, ketamine, marijuana, or methamphetamine). For the purposes of this manuscript, we created dichotomous indicators of whether the participant had engaged in heavy drinking and whether they had taken any of the nine drugs on a day. As such, heavy drinking and drug use days were count variables that ranged from 0–42 days for each participant. Past research suggests that this TLFB method provides an accurate picture of substance use across 42 days, even in comparison to daily diary methods (Rendina, Moody, Ventuneac, Grov, & Parsons, 2015).

Analysis Plan

Prior to testing the hypothesized model, we ran confirmatory factor analyses (CFA) on each of the variables we specified as latent, namely racial discrimination, gay rejection sensitivity, emotion regulation difficulties and internalizing symptoms. If needed, we ran several models to ensure factor loadings above 0.60 and acceptable model fit based on accepted indices (Hu & Bentler, 1999). We tested study hypotheses that racial discrimination, gay rejection sensitivity, and their interaction would predict increases in heavy drinking and drug use through positive associations with emotion regulation difficulties and internalizing symptoms within a longitudinal structural equation model (SEM) analyzed using Mplus 7.31 (Muthén & Muthén, 1998–2012). In this model, we tested main effects of a latent racial discrimination variable at baseline and a latent gay rejection sensitivity factor at baseline on a latent emotion regulation difficulties variable at baseline. Additionally, we used the XWITH function within a random effects model to construct a latent variable interaction between racial discrimination and gay rejection sensitivity and regressed emotion regulation difficulties on this interaction variable. The model included pathways from emotion regulation difficulties at baseline to a latent internalizing symptoms variable at 6 months and from internalizing symptoms at 6 months to the heavy drinking and drug use count variables at 12 months (See figure 1). We regressed heavy drinking and drug use on their baseline measurement to isolate change in the variables across time and so the mediation pathway predicted unique variance in the 12-month measurement. Within this model we included conceptually relevant pathways from control variables including baseline measurements of race/ethnicity, age, HIV status, and sexual orientation to emotion regulation difficulties, racial discrimination, and gay rejection sensitivity. We tested for the indirect effects of racial discrimination, gay rejection sensitivity, and the latent variable interaction using the Model Constraint command since Model Indirect is not available in random effects models.

Figure 1.

Figure 1.

Open in a new tab

Mediation Model Results ✝p≤ .10; *p ≤ .05; **p ≤ .01; ***p ≤ .001.

Note: Substance use outcomes are Negative Binomial distributed counts. Notation for the latent variable interaction is line with Muthén, B. (2012). For simplicity, indicators are not displayed for latent variables. BL= Baseline, 6M= 6 Months, 12M= 12 Months. This model is adjusted for baseline measurements of race/ethnicity, age, HIV status, and sexual orientation.

We specified baseline and 12-month heavy drinking and drug use as count variables with negative binomial distributions, which account for over-dispersion in the distribution that arises from a high range of responses. We set the variance of all latent variables to 1. There was significant residual covariance between some adjacent items in the racial discrimination measure (items 1–2, 5–6, and 8–9) and gay rejection sensitivity measure (items 11–12), as identified using the MODINDICES function in Mplus, which we freed the model to estimate.

After running the hypothesized model, we also ran two models that examined the direct effects from racial discrimination, gay rejection sensitivity, and the interaction variable to drug use, heavy drinking days, and internalizing symptoms. Since there are no fit indices for models with count outcomes, these analyses allowed us to determine the appropriateness of excluding these direct paths in the aforementioned model. We ran a model that included the direct effects on the internalizing symptoms latent variable at 6 months and then a second that included direct effects to the substance use variables at 12 months.

Missing data represents a concern across all longitudinal studies. For each of the variables, the following were the rates of complete data: Baseline racial discrimination (100%), baseline gay rejection sensitivity (100%), baseline emotion dysregulation (100%), baseline heavy drinking (100%), baseline drug use (100%), 6 month anxiety symptoms (85%), 6 month depressive symptoms (85%), 12 month heavy drinking (82%), and 12 month drug use (82%). In the case of missing data, Mplus 7.31 software uses a full information maximum likelihood estimator under the assumption that the data are missing at random (MAR; Arbuckle, 1996), which is a widely accepted way of handling missing data (Schafer & Graham, 2002). We found this assumption to be acceptable since there were no significant differences between participants with missing data and those without missing data across any variable included in this study. Thus, we handled missing data in line with the Mplus default, using all available data to estimate the model (Muthén & Muthén, 1998–2012) along with Monte Carlo Integration.

Results

Table 1 contains the demographic characteristics of the sample. The majority of the participants were Black/African American, gay-identified, single, and HIV-positive. The sample was 35 years old on average and ranged from 18 to 63 years old. The average number of heavy drinking days was 6.09 (Mdn = 3) at baseline and 5.25 (Mdn = 3) at 12-month follow-up while drug use days was 11.74 (Mdn = 6) at baseline and 12.21 (Mdn = 5) at 12-month follow-up.

Table 1.

Demographic Characteristics of the Sample

n %
Race/Ethnicity
 Black 76 42.7
 Latino 51 30.0
 Multiracial 43 25.3
Sexual Orientation
 Gay, queer, or homosexual 145 85.3
 Bisexual 25 14.7
 Heterosexual/straight 0 0
 Other 0 0
Employment Status
 Full-time 45 26.5
 Part-time 33 19.4
 Student (unemployed) 20 11.8
 Unemployed 45 26.5
 On disability 27 15.9
Highest Educational Attainment
 High school diploma/GED or less 29 17.1
 Some college or Associate’s degree 73 42.9
 Bachelor’s or other 4-year degree 43 25.3
 Graduate degree 25 14.7
HIV Status
 Positive 97 57.1
 Negative 73 42.9
Relationship Status
 Single 114 67.1
 Partnered 56 32.9
M SD
Age (Range: 18–63, Mdn= 33) 35.45 10.45
Number of Drug Use Days Baseline (Mdn= 6) 11.74 14.06
Number of Drug Use Days 12 months (Mdn= 5) 12.21 15.32
Number of Heavy Drinking Days Baseline (Mdn= 3) 6.09 8.71
Number of Heavy Drinking Days 12 Months (Mdn= 3) 5.25 8.15
Open in a new tab

Table 2 displays the descriptive statistics and bivariate correlations for each of the observed variables of interest. Racial discrimination was significantly positively correlated with gay rejection sensitivity, emotion regulation difficulties, and heavy drinking days at baseline. Racial discrimination was also significantly correlated with race/ethnicity and sexual orientation such that gay and Black participants were more likely to report racial discrimination. Gay rejection sensitivity was significantly positively correlated with emotion regulation difficulties. Emotion regulation difficulties was significantly positively correlated with anxiety symptoms, depressive symptoms, and drug use days at baseline and 12 months. Emotion regulation difficulties was also significantly correlated with age and race/ethnicity such that Latino and younger participants reported more emotion regulation difficulties. Additional correlations between study variables are provided in Table 2.

Table 2.

Correlations among Minority Stress and Internalizing Variables and Demographic Covariates

Variable 1 2 3 4 5 6 7 8 9
1. Racial Discrimination (BL) 1.0
2. Gay rejection sensitivity (BL) 0.19* 1.0
3. Emotion regulation difficulties (BL) 0.20** 0.15* 1.0
4. Anxiety Symptoms (6M) 0.09 0.15 0.30*** 1.0
5. Depressive Symptoms (6M) 0.09 0.15 0.37*** 0.79*** 1.0
6. Heavy Drinking (12M) 0.13 0.03 0.15 0.00 0.11 1.0
7. Drug Use (12m) −0.03 −0.16 0.24** −0.11 −0.01 0.18* 1.0
8. Heavy Drinking (BL) 0.16* −0.04 0.04 −0.08 −0.07 0.58*** 0.20* 1.0
9. Drug Use (BL) −0.08 −0.14 0.21** −0.04 0.01 0.14 0.75*** 0.29*** 1.0
10. Age 0.02 −0.02 −0.22** −0.24** −0.25** −0.18* −0.11 0.01 0.03
11. Black (1=yes) 0.15* 0.00 −0.15* −0.15 −0.17* −0.05 0.09 0.06 −0.03
12. Latino (1=yes) −0.06 0.01 0.16* −0.06 −0.07 −0.06 −0.10 −0.12 −0.05
13. Sexual Orientation (1=Bisexual) −0.17* −0.13 −0.10 0.01 0.01 −0.17* −0.11 −0.03 0.02
14. HIV Status (1=Positive) −0.04 0.04 0.07 −0.10 −0.08 −0.13 0.14 −0.01 0.26***
Mean 23.05 13.80 2.34 0.66 0.89 5.25 12.21 6.09 11.74
SD 11.06 8.45 0.77 0.77 0.82 8.15 15.32 8.71 14.06
Cronbach’s Alpha 0.95 0.92 0.90 0.89 0.88 -- -- -- --
Open in a new tab
*

p ≤ 0.05;

**

p ≤ 0.01;

***

p ≤ 0.001.

Note: All correlations are with observed scores. The estimates presented are Pearson’s r with the exception of those with heavy drinking variables (6 and 7) for which Spearman’s ρ is utilized due to the nonparametric nature of the count data.

BL= Baseline, 6M= 6 Months, 12M= 12 Months.

Regarding specification of latent variables, a CFA indicated that a nine-item one-factor specification of the racial discrimination variable fit the data adequately, χ2(24)=65.66, p<0.001; CFI=0.97, TLI=0.96, RMSEA=0.10. Standardized factor loadings across the racial discrimination items ranged from 0.64 to 0.90 in the CFA, and unstandardized factor loadings across these items ranged from 1.39 to 0.79 in the full model depicted in Figure 1. A CFA indicated that a 10-item, one-factor rejection sensitivity variable showed adequate model fit, χ2(34)=70.46, p<0.001; CFI=0.96, TLI=0.94, RMSEA=0.08. Standardized factor loadings across the gay rejection sensitivity items ranged from 0.62 to 0.77, and unstandardized factor loadings across these items ranged from 8.96 to 6.63 within the full model. Another CFA indicated that a 5-indicator, one-factor specification of the emotion regulation difficulties variable fit the data adequately, χ2(5)=28.74, p<0.01; CFI=0.96, TLI=0.92, RMSEA=0.16. Standardized factor loadings across all emotion regulation difficulty items ranged from 0.70 to 0.97, and unstandardized factor loadings across these items ranged from 0.44 to 0.81 within the full model. For the internalizing symptoms variable specification, although fit statistics cannot be produced in an under-identified model (i.e., a latent variable with 2 indicators), standardized factor loadings were high at 0.90 (anxiety subscale) and 0.87 (depression subscale) within the CFA, and unstandardized factor loadings were 0.67 and 0.58, respectively, within the full model.

Figure 1 depicts the mediation model in which racial discrimination, gay rejection sensitivity, and their interaction predicted heavy drinking and drug use through emotion regulation difficulties and internalizing symptoms. Baseline racial discrimination (β=0.27, S.E.=0.11, p=0.02) and the interaction between baseline gay rejection sensitivity and racial discrimination (β=0.21, S.E.=0.09, p=0.03) were both significantly and positively associated with baseline emotion regulation difficulties. The pattern of the interaction is depicted in Figure 2 and indicates that emotion regulation difficulties was highest at higher levels of racial discrimination and gay rejection sensitivity. The association between baseline gay rejection sensitivity (β=0.14, S.E.=0.09, p=0.13) and emotion regulation difficulties was not significant. Emotion regulation difficulties (β=0.40, S.E.=0.09, p<0.001) was significantly associated with internalizing symptoms at 6 months which, in turn, predicted heavy drinking at 12 months (b=0.36, S.E.=0.13, p=0.004). Internalizing symptoms at 6 months did not predict increases in drug use at 12 months (b=0.03, S.E.=0.14, p=0.85). Regarding covariates, the emotion regulation difficulties variable was significantly positively associated with HIV status (b=0.46, S.E.=0.18, p=0.01) and negatively associated with age (b=−0.03, S.E.=0.18, p=0.001). Racial discrimination and gay rejection sensitivity were not associated with any covariates.

Figure 2.

Figure 2.

Open in a new tab

Interaction of Racial Discrimination and Gay Rejection Sensitivity on Emotion Regulation Note: Graph set at the mean of emotion regulation, though it is measured as a latent variable. The rejection sensitivity variable is arbitrarily set as the moderator in this graph since the interaction variable is conceptualized as a having equal contributions of each variable.

The indirect effect results revealed that the indirect pathway from the interaction (b=0.03, S.E.=0.02, p=0.10), racial discrimination (b=0.04, S.E.=0.02, p=0.09), and gay rejection sensitivity (b=0.02, S.E.=0.02, p=0.20) were not significant. However, the total indirect effect from these three predictors combined was statistically significant (b=0.09, S.E.=0.04, p<0.05).

Exploratory Analysis of Direct Effects

Since the count outcomes of drug use and heavy drinking made it impossible to produce model fit statistics, we sought to explore whether the model was misspecified in leaving out the direct pathways from stigma to internalizing and substance use (i.e., constraining them to be 0). We did this by iteratively testing direct effects from racial discrimination, gay rejection sensitivity, and their interaction to internalizing symptoms, drug use, and heavy drinking. Due to the large number of paths in the model, the results of this model are displayed in Table 3 rather than in a figure. We ran the model with the same specifications as the model in Figure 1 with the exception that it included direct effects from the aforementioned baseline predictor variables to the 6-month latent internalizing symptoms variable (first model) and the 12-month substance use variables (second model). None of the direct effects were significant, suggesting that the original model was well-specified and that the assumption of a lack of direct effects was adequate.

Table 3.

Exploratory structural equation model with direct effects of stigma variables on internalizing and substance use variables

Internalizing Symptoms (6M) Heavy Drinking (12M) Drug Use (12M)
β S.E. p b S.E. p b S.E. p
Racial Discrimination (BL) 0.08 0.10 0.39 −0.05 0.13 0.73 0.27 0.16 0.09
Gay rejection sensitivity (BL) −0.01 0.11 0.91 0.11 0.14 0.42 −0.18 0.14 0.19
Interaction (BL) −0.10 0.11 0.34 −0.13 0.13 0.34 0.29 0.16 0.08
Open in a new tab
*

p ≤ .05;

**

p ≤ .01;

***

p ≤ .001.

Note: BL= Baseline.

Discussion

Despite the persistence of interpersonal violence and health inequities facing Black, Latino, and multiracial GBM in the United States, few studies have examined the connection between intersecting stigma and critical health outcomes for these men. With the present study we sought to add to the understanding of how the intersection of racial and sexual minority stigma lead to the stress process that predicts substance use through its effects on emotion regulation difficulties and psychological symptoms among Black, Latino, and multiracial GBM. Results from this study showed that racial discrimination was individually positively associated with emotion regulation difficulties, which predicted later internalizing symptoms that, in turn, predicted later heavy drinking among Black, Latino, and multiracial GBM. Gay rejection sensitivity was not individually significantly associated with emotion regulation difficulties, though we found there was a significant interaction between gay rejection sensitivity and racial discrimination such that, at higher levels of each construct, there were multiplicatively higher levels of emotion regulation difficulties. Moreover, the total indirect effect from gay rejection sensitivity, racial discrimination, and their interaction to heavy drinking through emotion regulation and internalizing symptoms was significant and positive. Taken together, these results suggest, although racial and sexual minority stigma are often examined in isolation, the psychological and behavioral effects of each are not mutually exclusive for Black, Latino, and multiracial GBM and, thus, are better understood when examined in the context of the other.

Our results did not support the hypothesis that the stigma-related stress process would result in greater levels of drug use. One possible explanation for this is that the drug use variable included several so-called “club drugs” (ketamine, MDMA/ecstasy, GHB, and methamphetamine). Club drugs are most commonly used in social situations and it is possible that with higher levels of anxiety and depressive symptoms among the participants that they were more socially isolated and, as a result, less likely to engage in use where club drugs are the most common: bars, nightclubs, concerts, and parties (National Institute on Drug Abuse, 2017).

Strengths and Limitations

The present study had several strengths worth noting that support the validity and impact of our results. First, we utilized longitudinal data with three time-points to model a stress process that extended minority stress theory. Indeed, it is one of the few projects that have blended intersectionality and minority stress theories as it considered racial stress in addition to, and in combination with, the more commonly-studied sexual minority stress for Black, Latino, and multiracial GBM, the groups at highest risk for several severe chronic illnesses (e.g., CDC, 2017). This study also presented a relatively strong statistical test of our conceptual model given we measured each of the independent variables as latent variables, which reduces measurement error inherent in path models with observed variables.

In addition to these strengths, this study had limitations that are important to note. First, although this study is one of relatively few that has focused specifically on Black, Latino, and multiracial GBM, and the sample size was adequate for analyses, a larger sample would provide the ability to examine heterogeneity within the sample. Indeed, future research can build on the present study to consider how this stress model varies among subgroups (e.g., Black GBM, Latino GBM) given there is considerable variance in sexual minority experiences and cultural narratives within these broad categories. In addition, it will be important to examine this model with a general sample, in addition to this highly sexually active sample, since heavy drinking and drug use tend to be more common among highly sexually active GBM (Rendina et al., 2017). The current study was also limited because, although we engaged in intersectional data analyses and interpretation, we did not measure intersectional experiences of discrimination. That is, we did not ask about stress associated with being a gay or bisexual man of color, including specific within-group cultural narratives (see Balsam, Molina, Beadnell, Simoni, & Walters, 2011; Bowleg et al., 2016), but asked about discrimination based on identities as if they were separate. In addition, although we incorporated several of the aspects of the minority stress model, we measured proximal (gay rejection sensitivity) and distal (racial discrimination) stigma variables and emotion regulation difficulties concurrently rather than testing for temporal precedence. Additional participants and follow-up data would allow for the testing of a more rigorous model such as a longitudinal autoregressive mediation model (e.g., Maxwell & Cole, 2007) that tests reciprocal associations between proximal and distal stigma variables as well as links between emotion regulation, psychological symptoms, and substance use. These limitations notwithstanding, there are several important implications for our findings in this study.

Research Implications

The finding that racial discrimination was associated with difficulties in emotion regulation, internalizing symptoms, and heavy drinking supports major tenants of the minority stress model (Meyer 2003; Hatzenbuehler, 2009) and past studies linking emotion regulation difficulties to internalizing symptoms and alcohol use (see Berking & Wupperman, 2012). Critically, the present results extend the minority stress model to include the intersection of racial and sexual minority stigma as predictors of this stress process. Indeed, our finding that the effects of both racial discrimination and gay rejection sensitivity were more impactful in the presence of the other provide support for a central tenet of intersectionality theory that the experience of different social positions are inextricably intertwined and should be considered together (Bowleg, 2008). Indeed, although racial discrimination did, and gay rejection sensitivity did not, have a significant main effect on emotion regulation difficulties, they showed a significant combined effect on heavy drinking through emotion regulation and internalizing symptoms. As such, it is critical that researchers and practitioners alike not only work to understand the effects of racial and sexual minority stigma for Black, Latino, and multiracial GBM, but also consider the interaction effects of these stressors.

Clinical and Policy Implications

These results also have important implications for intervention. First, it is critical to acknowledge that an unmeasured, yet driving force of the stress process modelled in this manuscript is historically-rooted and contemporaneously-perpetuated structural racism and heterosexism that are implicated in cultural narratives that promote prejudice and discrimination and systematically distribute power and privilege to White and heterosexual men in the U.S. (Kimmel & Ferber, 2016). As such, structural interventions that incorporate a critical equity stance on research, public policy, and institutional reform and aim to reduce and/or eliminate the structural inequities that disproportionately affect Black, Latino, and multiracial GBM will be the most powerful forms of prevention when considering the stress process we model in the present study (Bowleg, 2017). Such interventions may involve training educators to lead discussions and activities to aid students in reflection and social action to dismantle privilege in U.S. society (see Case, 2013), and could include reducing prejudice through extended contact within desegregated peer environments within these schools (Zhou, Page-Gould, Aron, Moyer, & Hewstone, 2018).

As researchers, political activists, community members, policy makers, administrators, and others work to address these structural inequities, the present results indicate areas of intervention to reduce the effects of persistent sources of stigma. Our findings suggest that interventions that help to empower Black, Latino, and multiracial GBM to understand and process emotions (i.e., emotionally regulate) following experiences of stigma may be helpful to reduce the psychological impact of the intersectional stigma examined in this study. Psychotherapeutic approaches that incorporate techniques for promoting emotional exploration (e.g., Pachankis, Hatzenbeuhler, Rendina, Safren & Parsons, 2015) such as those in Dialectical Behavioral Therapy (Linehan et al., 2006) and take a culturally competent, strengths-based, client-centered approach (see Paniagua, 2013) may be particularly beneficial for improving psychological outcomes among Black, Latino, and multiracial GBM. In addition, clinical researchers seeking to inform the development of interventions will benefit from considering the positive aspects of intersecting identities in addition to experiences of stigma, as these aspects likely serve as buffers for several of the associations in the present model (Bowleg et al., 2016).

Conclusion

Overall, the present study suggests that intersecting stigma may be an important mechanism causing psychological and behavioral health inequities facing Black, Latino, and multiracial GBM in the U.S. (CDC, 2017). Specifically, our findings suggest that taking an intersectional approach to examining minority stress for Black, Latino, and multiracial GBM is critical for understanding and intervening upon the stress process initiated by racial and sexual minority stigma that leads to emotion regulation difficulties, internalizing symptoms, and heavy drinking among these men. The development, application, and/or adaptation of techniques for emotion regulation-focused and culturally competent therapy for Black, Latino, and multiracial GBM will be important while structural interventions can, and should, play the central role in addressing current health inequities for Black, Latino, and multiracial GBM.

References

  1. Arbuckle JL (1996). Full information estimation in the presence of incomplete data. Advanced structural equation modeling: Issues and techniques, 243–277. [Google Scholar]
  2. Arnold EA, Rebchook GM, & Kegeles SM (2014). ‘Triply cursed’: Racism, homophobia and HIV-related stigma are barriers to regular HIV testing, treatment adherence and disclosure among young Black gay men. Culture, Health & Sexuality, 16(6), 710–722. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Balsam KF, Molina Y, Beadnell B, Simoni J, & Walters K (2011). Measuring multiple minority stress: the LGBT People of Color Microaggressions Scale. Cultural Diversity and Ethnic Minority Psychology, 17(2), 163–174. doi: 10.1037/a0023244 [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Bardeen JR, Fergus TA, & Orcutt HK (2012). An examination of the latent structure of the Difficulties in Emotion Regulation Scale. Journal of Psychopathology and Behavioral Assessment, 34(3), 382–392. doi: 10.1007/s10862-012-9280-y [DOI] [Google Scholar]
  5. Berking M, & Wupperman P (2012). Emotion regulation and mental health: Recent findings, current challenges, and future directions. Current Opinion in Psychiatry, 25(2), 128–134. [DOI] [PubMed] [Google Scholar]
  6. Bogart LM, Wagner GJ, Galvan FH, Landrine H, Klein DJ, & Sticklor LA (2011). Perceived discrimination and mental health symptoms among Black men with HIV. Cultural Diversity and Ethnic Minority Psychology, 17(3), 295–302. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Bostwick WB, Boyd CJ, Hughes TL, & McCabe SE (2010). Dimensions of sexual orientation and the prevalence of mood and anxiety disorders in the United States. American Journal of Public Health, 100(3), 468–475. doi: 10.2105/AJPH.2008.152942 [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Bostwick WB, Boyd CJ, Hughes TL, West BT, & McCabe SE (2014). Discrimination and mental health among lesbian, gay, and bisexual adults in the United States. American Journal of Orthopsychiatry, 84(1), 35–45. doi: 10.1037/h0098851 [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Bowleg L (2008). When Black + lesbian + woman ≠ Black lesbian woman: The methodological challenges of qualitative and quantitative intersectionality research. Sex Roles, 59(5), 312–325. doi: 10.1007/s11199-008-9400-z [DOI] [Google Scholar]
  10. Bowleg L (2017). Towards a critical health equity research stance: why epistemology and methodology matter more than qualitative methods. Health Education & Behavior, 44(5), 677–684. [DOI] [PubMed] [Google Scholar]
  11. Bowleg L, & Bauer G (2016). Invited reflection: Quantifying intersectionality. Psychology of Women Quarterly, 40(3), 337–341. doi: 10.1177/0361684316654282 [DOI] [Google Scholar]
  12. Bowleg L, English D, del Rio-Gonzalez AM, Burkholder GJ, Teti M, & Tschann JM (2016). Measuring the pros and cons of what it means to be a Black man: Development and validation of the Black Men’s Experiences Scale (BMES). Psychology of Men & Masculinity, 17(2), 177–188. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Bush S, Magnuson D, Rawlings MK, Hawkins T, McCallister S, & Mera Giler R (2016). Racial characteristics of FTC/TDF for pre-exposure prophylaxis (PrEP) users in the US. Paper presented at the ASM Microbe Conference. [Google Scholar]
  14. Case K (2013). Deconstructing privilege: Teaching and learning as allies in the classroom. Routledge. [DOI] [PubMed] [Google Scholar]
  15. Centers for Disease Control and Prevention (2017). Monitoring selected national HIV prevention and care objectives by using HIV surveillance data—United States and 6 dependent areas, 2015. HIV Surveillance Supplemental Report 22(2). [Google Scholar]
  16. Choi K-H, Paul J, Ayala G, Boylan R, & Gregorich SE (2013). Experiences of discrimination and their impact on the mental health among African American, Asian and Pacific Islander, and Latino men who have sex with men. American Journal of Public Health, 103(5), 868–874. doi: 10.2105/AJPH.2012.301052 [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Collins PH (1998). It’s all In the family: Intersections of gender, race, and nation. Hypatia, 13(3), 62–82. doi: 10.1111/j.1527-2001.1998.tb01370.x [DOI] [Google Scholar]
  18. Collins PH, & Bilge S (2016). Intersectionality: John Wiley & Sons. [Google Scholar]
  19. Crenshaw K (1991). Mapping the margins: Intersectionality, identity politics, and violence against women of color. Stanford Law Review, 43(6), 1241–1299. doi: 10.2307/1229039 [DOI] [Google Scholar]
  20. Derogatis LR, & Melisaratos N (2009). The Brief Symptom Inventory: An introductory report. Psychological Medicine, 13(3), 595–605. doi: 10.1017/S0033291700048017 [DOI] [PubMed] [Google Scholar]
  21. Downey G, & Feldman SI (1996). Implications of rejection sensitivity for intimate relationships. Journal of Personality and Social Psychology, 70(6), 1327–1343. [DOI] [PubMed] [Google Scholar]
  22. First MB, Spitzer RL, Gibbon M, & Williams JBW (2002). Structured clinical interview for DSM-IV-TR axis I disorders, research version, patient edition (SCID-I/P). New York: Biometrics Research. [Google Scholar]
  23. Folstein MF, Folstein SE, & McHugh PR (1975). “Mini-mental state”: A practical method for grading the cognitive state of patients for the clinician. Journal of Psychiatric Research, 12(3), 189–198. [DOI] [PubMed] [Google Scholar]
  24. Gratz KL, & Roemer L (2004). Multidimensional assessment of emotion regulation and dysregulation: Development, factor structure, and initial validation of the Difficulties in Emotion Regulation scale. Journal of Psychopathology and Behavioral Assessment, 26(1), 41–54. doi: 10.1023/B:JOBA.0000007455.08539.94 [DOI] [Google Scholar]
  25. Grollman EA (2014). Multiple disadvantaged statuses and health: The role of multiple forms of discrimination. Journal of Health and Social Behavior, 55(1), 3–19. doi: 10.1177/0022146514521215 [DOI] [PubMed] [Google Scholar]
  26. Hatzenbuehler ML (2009). How does stigma “get under the skin”?: The mediating role of emotion regulation. Psychological Science, 20(10), 1282–1289. doi: 10.1111/j.1467-9280.2009.02441.x [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Hatzenbuehler ML, Nolen-Hoeksema S, & Erickson SJ (2008). Minority stress predictors of HIV risk behavior, substance use, and depressive symptoms: Results from a prospective study of bereaved gay men. Health Psychology, 27(4), 455–462. [DOI] [PubMed] [Google Scholar]
  28. Herek GM, Gillis JR, & Cogan JC (1999). Psychological sequelae of hate-crime victimization among lesbian, gay, and bisexual adults. Journal of Consulting and Clinical Psychology, 67, 945–951. [DOI] [PubMed] [Google Scholar]
  29. Hershberger SL, & D’Augelli AR (1995). The impact of victimization on the mental health and suicidality of lesbian, gay, and bisexual youths. Developmental Psychology, 31(1), 65–74. doi: 10.1037/0012-1649.31.1.65 [DOI] [Google Scholar]
  30. Hu L, Bentler PM (1999). Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Structural Equation Modeling. 6, 1–55. [Google Scholar]
  31. Huebner DM, Rebchook GM, & Kegeles SM (2004). Experiences of harassment, discrimination, and physical violence among young gay and bisexual men. American Journal of Public Health, 94(7), 1200–1203. doi: 10.2105/ajph.94.7.1200 [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Kimmel MS, & Ferber AL (2016). Privilege: A reader: Westview Press. [Google Scholar]
  33. Lewis RJ, Derlega VJ, Griffin JL, & Krowinski AC (2003). Stressors for gay men and lesbians: Life stress, gay-related stress, stigma consciousness, and depressive symptoms. Journal of Social and Clinical Psychology, 22(6), 716–729. [Google Scholar]
  34. Linehan MM, Comtois KA, Murray AM, Brown MZ, Gallop RJ, Heard HL, … & Lindenboim N (2006). Two-year randomized controlled trial and follow-up of dialectical behavior therapy vs therapy by experts for suicidal behaviors and borderline personality disorder. Archives of General Psychiatry, 63(7), 757–766. [DOI] [PubMed] [Google Scholar]
  35. Maxwell SE, & Cole DA (2007). Bias in cross-sectional analyses of longitudinal mediation. Psychological Methods, 12(1), 23–44. doi: 10.1037/1082-989X.12.1.23 [DOI] [PubMed] [Google Scholar]
  36. Meyer IH (2003). Prejudice, social stress, and mental health in lesbian, gay, and bisexual populations: Conceptual issues and research evidence. Psychological Bulletin, 129(5), 674–697. doi: 10.1037/0033-2909.129.5.674 [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Meyer D (2012). An intersectional analysis of lesbian, gay, bisexual, and transgender (LGBT) people’s evaluations of anti-queer violence. Gender & Society, 26(6), 849–873. [Google Scholar]
  38. Muthén B (2012). Latent variable interactions. Unpublished manuscript. Retrieved. http://statmodel2.com/download/Latent%20variable%20interactions.pdf.
  39. Muthén L, & Muthén B (1998–2012). Mplus User’s Guide (7th ed.). Los Angeles: Muthén & Muthén. [Google Scholar]
  40. National Institute on Drug Abuse. (2017). Club Drugs. Retrieved from https://www.drugabuse.gov/drugs-abuse/club-drugs on 2017, October 30. [Google Scholar]
  41. Pachankis JE, Goldfried MR, & Ramrattan ME (2008). Extension of the rejection sensitivity construct to the interpersonal functioning of gay men. Journal of Consulting and Clinical Psychology, 76(2), 306–317. doi: 10.1037/0022-006X.76.2.306 [DOI] [PubMed] [Google Scholar]
  42. Pachankis JE, Hatzenbuehler ML, Rendina HJ, Safren SA, & Parsons JT (2015). LGB-affirmative cognitive-behavioral therapy for young adult gay and bisexual men: A randomized controlled trial of a transdiagnostic minority stress approach. Journal of Consulting and Clinical Psychology, 83(5), 875–889. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Pachankis JE, Rendina HJ, Restar A, Ventuneac A, Grov C, & Parsons JT (2015). A minority stress—emotion regulation model of sexual compulsivity among highly sexually active gay and bisexual men. Health Psychology, 34(8), 829–840. [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Paniagua FA (2013). Assessing and treating culturally diverse clients: A practical guide. Sage Publications. [Google Scholar]
  45. Rendina HJ, Gamarel KE, Pachankis JE, Ventuneac A, Grov C, & Parsons JT (2017). Extending the minority stress model to incorporate HIV-positive gay and bisexual men’s experiences: A longitudinal examination of mental health and sexual risk behavior. Annals of Behavioral Medicine, 51(2), 147–158. doi: 10.1007/s12160-016-9822-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Rendina HJ, Moody RL, Ventuneac A, Grov C, & Parsons JT (2015). Aggregate and event-level associations between substance use and sexual behavior among gay and bisexual men: comparing retrospective and prospective data. Drug & Alcohol Dependence, 154, 199–207. [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Schafer JL, & Graham JW (2002). Missing data: Our view of the state of the art. Psychological Methods, 7(2), 147–177. doi: 10.1037/1082-989X.7.2.147 [DOI] [PubMed] [Google Scholar]
  48. Sobell MB, & Sobell LC (1996). Problem drinkers: Guided self-change treatment: Guilford Press. [DOI] [PubMed] [Google Scholar]
  49. Talley AE, Hughes TL, Aranda F, Birkett M, & Marshal MP (2014). Exploring alcohol-use behaviors among heterosexual and sexual minority adolescents: Intersections with sex, age, and race/ethnicity. American Journal of Public Health, 104(2), 295–303. [DOI] [PMC free article] [PubMed] [Google Scholar]
  50. Vial AC, Starks TJ, & Parsons JT (2014). Finding and recruiting the highest risk HIV-negative men who have sex with men. AIDS Education and Prevention, 26(1), 56–67. [DOI] [PMC free article] [PubMed] [Google Scholar]
  51. Waters E, Pham L, Convery C, & Yacka-Bible S (2018). A crisis of hate a report on lesbian, gay, bisexual, transgender and queer hate violence homicides in 2017. Retrieved from https://avp.org/wp-content/uploads/2018/01/a-crisis-of-hate-january-release.pdf
  52. Williams DR, Yan Y, Jackson JS, & Anderson NB (1997). Racial differences in physical and mental health: Socio-economic status, stress and discrimination. Journal of Health Psychology, 2(3), 335–351. doi: 10.1177/135910539700200305 [DOI] [PubMed] [Google Scholar]
  53. Yoshikawa H, Wilson PA, Chae DH, & Cheng JF (2004). Do family and friendship networks protect against the influence of discrimination on mental health and HIV risk among Asian and Pacific Islander gay men? AIDS Education and Prevention, 16(1), 84–100. doi: 10.1521/aeap.16.1.84.27719 [DOI] [PubMed] [Google Scholar]
  54. Zhou S, Page-Gould E, Aron A, Moyer A, & Hewstone M (2018). The Extended contact hypothesis: A meta-analysis on 20 years of research. Personality and Social Psychology Review, No pagination specified. doi: 10.1177/1088868318762647 [DOI] [PubMed] [Google Scholar]

RESOURCES