Introduction

Black gay, bisexual, and other sexual minority men (SMM) in the United States (US) continue to experience a disproportionate burden of HIV [1]. In 2017, the rate of new HIV diagnoses among Black adults and adolescents was eight times that of white and more than twice that of Latinx individuals [2]. In 2018, SMM made up 92% of new HIV cases among adolescents and young adults aged 13 to 24, of whom 51% were among young Black SMM (YBSMM) [3]. Black transgender women (TW) also face significant disparities in HIV; among all trans women, HIV prevalence is 44.2% among Black TW, compared to 6.7% among white TW [4].

HIV pre-exposure prophylaxis (PrEP) is a key aspect of HIV prevention. In areas of high PrEP uptake, HIV incidence is significantly reduced, even when controlling for viral suppression [5]. Yet, PrEP has been slow to reach communities in the US that experience a disproportionate burden of HIV [6,7,8,9]. While evidence indicates that PrEP use may be increasing, [10] persistent racial disparities remain. Although PrEP awareness among Black TW is high, PrEP initiation and consistent use remain low [11, 12]. Between 2014 and 2016, Black individuals accounted for approximately 44% of those indicated for PrEP, yet just 11% of PrEP users during that same time period were Black [10]. Current estimates indicate that about 10% of PrEP users are Black SMM [13,14,15]. Furthermore, data indicate that only between 12 and 20% of Black SMM on PrEP achieve protective levels of PrEP [16,17,18].

Prior research has examined myriad individual, social, and structural factors, and their association with disparities in PrEP uptake [19]. One factor thought to contribute to the slow and uneven rollout of PrEP among Black SMM and TW communities is stigma and discrimination. There is evidence, for example, that unconscious bias among medical providers may lead to reduced willingness to prescribe PrEP to Black patients [20]. Additionally, experienced and anticipated racism and homonegativity in healthcare settings may contribute to difficulty receiving PrEP prescriptions among YBSMM [21] and TW [22]. Although sexuality disclosure to healthcare providers is positively associated with HIV prevention services among YBSMM, [23] individuals may be hesitant to disclose their sexuality if they worry about mistreatment from providers. There is also evidence that structural-level stigma may contribute to PrEP disparities. In states with nondiscrimination laws that protect sexual and gender minorities, there are higher levels of PrEP awareness and uptake [24]. Similarly, SMM who reside in states with high LGBTQ equity (as measured by state-level laws and policies affecting LGBTQ individuals), have significantly higher odds of PrEP use, compared to low equality states [25].

Influenced by intersectionality scholarship, research on the effects of stigma and discrimination on HIV disparities has largely shifted from examining a singular type of stigma toward examining the interlocking nature of various forms of stigma, oppression, and discrimination. Intersectionality is rooted in Black feminist thought [26,27,28] and highlights the ways that various forms of oppression (e.g. racism, heterosexism, sexism) are mutually reinforcing and should be considered simultaneously.[29, 30] Intersectionality scholarship suggests that marginalized social identities intersect and reflect multiple systems of societal oppression and power. [26, 30] Intersectional discrimination [31] is the process by which some individuals are exposed to multiple forms of oppression, prejudice, and discrimination.

There is some evidence that intersectional stigma and discrimination influence health outcomes among Black SMM and TW, [32, 33] although little research has examined the relationship between intersectional discrimination and PrEP use. In research with Black SMM living with HIV, Bogart and colleagues found that discrimination based on race, sexual orientation, and HIV interacted to predict higher levels of depressive symptoms, although none of the main effects from these forms of discrimination were individually associated with depressive symptoms. [34] More recent research has examined the interaction between racial discrimination and gay rejection sensitivity and found that the interaction of these constructs contributed to higher levels of emotion regulation difficulties, and subsequently, to heavy drinking among Black, Latinx, and multiracial SMM. [35] Intersectional stigma can also contribute to negative self-concept, negative future orientation, and reduced sense of social connectedness, which can place individuals at greater vulnerability to HIV [36, 37] and limit engagement in HIV prevention interventions, including PrEP. Qualitative research has begun to investigate how experiences of stigma and discrimination may influence PrEP use. [38, 39] For example, the intersections of racism and homonegativity can negatively affect YBSMM’s interactions with the healthcare system and contribute to medical mistrust, disengagement from the healthcare system, and skepticism surrounding PrEP [21].

Increasingly, resilience has been incorporated into intersectional stigma research to understand how various resilience resources may help racial minority SMM and TW cope with and respond to stigma and discrimination [40,41,42]. Resilience is a dynamic process wherein individuals are able to positively adapt or succeed within the context of adversity. [43, 44] Positive adaptation occurs via protective internal assets (individual-level resilience) or external resources (supportive social environments) that can help facilitate positive health outcomes,[45] such as social support [46,47,48] and connection to an affirming LGBTQ community.[49, 50] Individuals who face stigma and discrimination may seek out supportive social environments that offer a sense of safety and protection and build community in response to discrimination.

The psychological mediation framework posits that stigma-related stressors render sexual and gender minorities vulnerable to psychological processes that predict mental health outcomes. [51] Stigma and discrimination-related stressors can shape individual coping processes, including the development of resilience, building of social support, and stronger ties to the LGBTQ community, which in turn may mediate the process between stigma or discrimination and PrEP use outcomes. Research has demonstrated support for this model in examining resilient coping and social support as mediators of the relationship between sexual minority stigma and depression. [52,53,54] Prior research with gay men has also shown that, in response to discrimination, men developed resilience and built supportive social networks to overcome daily adversity. [55] This aligns with stress inoculation theories, which suggest that early experiences of racism may help racial minority LGBTQ individuals develop resilience processes. [56] YBSMM and TW who experience discrimination may receive support and learn coping skills for dealing with discrimination and may demonstrate greater resilience in the face of stigma due to stress-inoculation processes [57].

However, there is little research on the relationship between intersectional discrimination and PrEP use, and what role resilience factors may play. This is a critical gap in our understanding of how intersectional discrimination influences HIV prevention efforts and may help inform interventions to increase access to and usage of PrEP among YBSMM, TW, and others who contend with multiple societal stigmas. To that end, this study examined the relationship between intersectional discrimination and PrEP use among YBSMM and TW, and the role of resilience, social support, and Black LGBTQ community connectedness as potential mediators. Research has demonstrated that there are likely multiple mediating and moderating pathways of resilience factors. [58] The models tested in this paper are rooted in the psychological mediation framework, as we examine resilience factors that may be activated in response to the stigma experienced.[59] We hypothesized that individuals who have experienced higher levels of intersectional discrimination (that is, discrimination based on their marginalized intersectional identities) would be less likely to currently use PrEP and less likely to use PrEP in the future than those who have experienced less discrimination. However, given evidence that discrimination may also contribute to resilience, we also hypothesized that resilience factors would mediate associations between intersectional stigma and PrEP use. Specifically, we predicted that intersectional discrimination would be positively associated with resilience, social support, and Black LGBTQ community support, which in turn would be positively associated with PrEP use.

Methods

Between 2018 and 2020, we recruited 283 YBSMM to participate in this study. Eligibility criteria included: (1) self-identifying as Black or African American, (2) assigned male sex at birth, (3) being between the ages of 16 and 25, (4) identifying as gay, bisexual, or other sexual minority status, (5) residing in Milwaukee, WI or Cleveland, OH, and (6) reporting a negative or unknown HIV status.

Cleveland and Milwaukee are two mid-size midwestern cities with significant racial disparities in HIV and low uptake of PrEP. Geographically, both Cuyahoga and Milwaukee counties in Ohio and Wisconsin, respectively, have a disproportionate share of the states’ HIV incidence.[60, 61] In both cities, young Black SMM and TW are disproportionately affected by HIV.[60, 61] In 2018, 65% of new infections among African Americans in in Ohio were among SMM.[60] Similarly, in Wisconsin in 2018, racial and ethnic minorities made up just 18% of Wisconsin’s population yet consisted of 66% of new HIV diagnoses.

Participants were recruited through outreach activities led by experienced research associates in both cities. Outreach efforts included in-person recruitment at community organizations and social venues frequented by YBSMM, social media postings, and paid advertising on the dating/hookup smartphone apps Jack’d and Grindr. Interested individuals were screened for eligibility by study staff and eligible participants scheduled a time to complete the self-administered Qualtrics assessment in person in a community-based setting. Beginning in mid-2020, due to the COVID-19 pandemic, enrollment shifted to allow eligible participants to be sent a personalized link to the assessment to complete at home. Given the stigmatized nature of HIV, PrEP and sexuality, we received a waiver of parental/guardian consent for participants under the age of 18 and a waiver of documented consent for all participants. Prior to accessing the survey, participants were given an informational letter outlining the study procedures and risks and benefits and providing contact information for study team members and the PI. The median survey completion time was 32 min. Participants received $50 for completing the survey, along with a list of community resources. All study procedures were reviewed and approved by the Institutional Review Board at [Blinded for Review].

Measures

Assessment content was informed by focus groups conducted in 2017 and 2018 with YBSMM that explored intersectional stigma, HIV prevention, and PrEP use [Blinded for Review].

Demographics and other covariates. For city, we examined whether participants were enrolled in Cleveland or Milwaukee. We coded whether participants had enrolled in the study following COVID-19 shutdowns in March 2020. Participants self-reported their age, gender identity, ethnicity, sexual orientation, and highest level of education. A dummy variable was created indicating transgender, non-binary, or another gender identity; male gender identity served as the reference category. For sexual orientation, dummy variables were created indicating bisexual sexual orientation and straight/heterosexual or another sexual orientation; gay sexual orientation served as the reference category. A dummy variable indicated Latinx ethnicity; non-Latinx ethnicity was the reference category. Education was coded as never graduated high school (0), graduated high school or GED (1), some college (2), or graduated college or graduate schooling (3). To assess economic hardship, participants responded to 3 items (α = 0.81, e.g., “In the last 12 months, how often did you run out of money for your basic necessities?” [62]) using a scale of never (0) to many times (3). Items were averaged, with higher scores indicating greater economic hardship. Participants reported whether they currently had health insurance (0 = no, 1 = yes). An item assessed the number of sexual partners in the past 30 days (0 = 0, 1 = 1, 2 = 2–3, 4 = 4 or more). Finally, participants responded to 10 items assessing psychological distress (α = 0.95, e.g., “During the last 30 days, about how often did you feel depressed?” [63]) using a scale of none of the time (1) to all of the time (5). Items were averaged, with higher scores indicating greater distress.

Current PrEP use and likelihood of future PrEP use. We considered two separate outcome variables. First, participants reported whether they were currently taking PrEP (0 = no, 1 = yes). Second, those who were not currently taking PrEP reported their likelihood of taking PrEP in the future [64] (“How likely would you be to use PrEP in the future?”), with response options being not at all likely (1), probably not likely (2), neutral (3), probably likely (4), and definitely likely (5). Those who reported that they had never heard of PrEP did not respond to questions about current or future use and were coded as not currently taking PrEP (0) and not at all likely to take it in the future (1).

Stigma-related constructs

Intersectional discrimination. The Intersectional Discrimination Index (InDI) developed by Scheim and Bauer [65] was used to measure intersectional anticipated discrimination, day-to-day discrimination, and major discrimination (lifetime). The InDI was explicitly developed to assess discrimination across a range of intersectional marginalized social identities and positions by asking participants to endorse discriminatory experiences “because of who I am,” rather than attributing an experience of discrimination to a single axis of one’s identity (e.g., discrimination due to race). Prior research has shown that attributions in discrimination measures are difficult and even impossible for some research participants to answer. [28] For anticipated discrimination, participants responded to 9 items (e.g., “Because of who I am, a doctor or nurse or other health care provider might treat me poorly”) on a scale from strongly disagree (1) to strongly agree (5). Items were averaged, with higher scores indicating greater anticipated discrimination (α = 0.91). For day-to-day discrimination, participants indicated how frequently they had experienced 9 types of discrimination (e.g., “Been called names or heard/saw your identity used as an insult”); for each item, frequency was coded as never or not in the past year (0), once or twice in the past year (1), or many times in the past year (2). Items were averaged, with higher scores indicating greater day-to-day discrimination (α = 0.91). Finally, for major discrimination, participants indicated how frequently they had experienced 14 types of discrimination in their lifetime (e.g., “Because of who you are, has a health care provider ever refused you care?”), with response options being never (0), once (1), or more than once (2). Items were averaged (α = 0.89), with higher scores indicating more major discrimination experiences. Reliability of these three subscales was higher in this study (α = 0.89–91) than in the initial scale development study (α = 0.70–72) [65].

Interpersonal homophobia and racism. In addition to intersectional discrimination, we also measured experiences of interpersonal homophobia and racism with 9 items from Jeffries et al. (2013) that assessed the number of times in the last 12 months participants experienced a specific event, (e.g., “Felt that white gay men are uncomfortable around me because of my race or ethnicity”) using the response options never (0), once (1), and 2 + times (2). Items were averaged (α = 0.92), with higher scores indicating more experiences of interpersonal homophobia and racism [66].

Microaggressions. The 18-item LGBT People of Color Microaggressions scale [67] assessed whether individuals had ever experienced various microaggressions (e.g., “Being told that race isn’t important by white LGBT people”) and how much those experiences bothered participants. Responses were on a scale from it did not happen/is not applicable to me (0) to it happened and it bothered me EXTREMELY (5). Items were averaged (α = 0.96), with higher scores indicating more microaggressions.

PrEP social concerns. PrEP social concerns were assessed with 5 items developed for this study based on findings from our preliminary qualitative research [21] and prior measures of PrEP stigma [68, 69] (e.g., “I would be concerned if my friends found out I was taking PrEP”). Responses were on a scale from strongly disagree (1) to strongly agree (5). Items were averaged (α = 0.90), with higher scores indicating more social concerns related to using PrEP.

Resilience-related constructs

Resilience. Resilience was assessed with 10 items from the Connor-Davidson Resilience scale, [70] (e.g., “Having to cope with stress can make me stronger”). Responses were on a scale from not at all true (0) to true nearly all the time (4). Items were averaged (α = 0.97), with higher scores indicating greater resilience.

Social support. Social support was assessed with 12 items from the Multidimensional Scale of Perceived Support [71] that asked about support available from family, friends, and significant others (e.g., “I can count on my friends when things go wrong”). Responses were on a scale from very strongly disagree (1) to very strongly agree (7). Items were averaged (α = 0.97), with higher scores indicating greater social support.

Black LGBT community connectedness. Black LGBT community connectedness was assessed with 8 items adapted from Frost and Meyer [72]. Items were adapted to incorporate race (e.g., “I feel I am part of my community’s Black LGBT community”). Responses were on a scale from strongly disagree (1) to strongly agree (4). Items were averaged (α = 0.94), with higher scores indicating greater connection to the Black LGBT community.

Data Analysis

Missing data was relatively rare (3% of all data was missing). To address missing data, we used multiple imputation (MI), a modern method for dealing with missing data which allowed us to maintain the maximum sample size and avoid biases associated with complete case analysis or single imputations All study variables were included when imputing 100 datasets in Mplus 8. [73] Analyses were conducted with all datasets, and parameter estimates were pooled using the imputation algorithms in Mplus 8.

Path models in Mplus 8 were used to test associations between intersectional discrimination, resilience, and PrEP use and intentions. The model focused on likelihood of future PrEP use, excluding those who reported being current PrEP users. In both models, directional paths led from intersectional stigma constructs to resilience constructs and PrEP outcomes, and from resilience constructs to PrEP outcomes, in line with our hypotheses. Additionally, paths led from demographic covariates to all constructs. Resilience constructs (resilience, social support, and Black LGBTQ community connectedness) were allowed to correlate with one another.

We fit models using a full information maximum likelihood estimator robust to non-normality (the MLR estimator). Coefficients for predictors that were not associated with outcomes (p > .20) were constrained to zero to increase model parsimony and stabilize estimates. [74] Given this approach, our primary focus was on path coefficients and estimates of indirect effects rather than overall model fit. When testing mediation, 95% confidence intervals (CIs) were calculated using the Monte Carlo method in RMediation. [75] This is recommended given the non-normal distribution of indirect effects. We report unstandardized factor loadings, beta coefficients, and odds ratios (ORs). For indirect effects, we also report 95% CIs.

Results

Descriptive information

Based on eligibility criteria, all 283 participants were Black/African American; 8% were multiracial and 7% were Latinx. Most participants (87%) identified as male; the remainder identified as transgender women (11%) or another gender identity (2%). The average age was 22 (SD = 2.75, range = 16–25). With regard to PrEP use, 13% were currently using PrEP and 8% had previously used PrEP. Slightly more than half (56%) of participants enrolled in Milwaukee and the rest in Cleveland (44%). Descriptive statistics related to demographic covariates, stigma, resilience, and likelihood of taking PrEP are summarized in Table 1. Correlations between all study variables are included in Table 2.

Table 1 Descriptive Characteristics (N = 283)
Full size table
Table 2 Correlations Among Study Variables
Full size table

Current PrEP use: path model

The path model focused on current PrEP use (Fig. 1) showed that participants with higher levels of anticipated discrimination had lower odds of being current PrEP users, OR = 0.59, p = .013. Participants who experienced more day-to-day stigma had higher levels of resilience, B = 0.38 (0.10), p < .001, higher levels of social support, B = 0.37 (0.16), p = .019, and greater connection to the Black LGBTQ community, B = 0.19 (0.07), p = .006. Additionally, participants with more social concerns about PrEP use reported lower levels of resilience, B = -0.19 (0.07), p = .004, and a lower connection to the Black LGBTQ community, B = -0.11 (0.05), p = .013. Participants with more connection the Black LGBTQ community had higher odds of current PrEP use, OR = 2.25, p = .003. There was a significant and positive indirect effect of day-to-day discrimination on current PrEP use via Black LGBTQ community connectedness, B = 0.15 (0.08), 95% CI [0.03, 0.32], p < .01, such that community connectedness fully mediated the association between day-to-day discrimination and PrEP use. Additionally, there was a significant and negative indirect effect of PrEP social concerns on current PrEP use via Black LGBTQ community connectedness, B = -0.09 (0.05), 95% CI [-0.20, -0.01], p < .05, such that community connectedness fully mediated the effects of PrEP social concerns on current PrEP use.

Fig. 1
figure 1

Path model showing associations between intersectional stigma, resilience, and current PrEP use in a sample of young sexual minority men in the Midwest (N = 283). Unstandardized linear regression coefficients (Bs) and odds ratios (ORs) are presented. The model also included correlations between resilience, social support, and Black LGBT community connectedness (paths not shown); there were moderate correlations between the constructs, rs = 0.29–0.40, ps < 0.001. Demographic covariates (age, gender identity, ethnicity, sexual orientation, education, economic hardship, insurance status, number of sexual partners in the past 30 days, psychological distress, enrollment city, and enrollment following the COVID-19 shutdowns) were also included in the model (paths not shown). *** p < .001 ** p < .01 * p < .05

Full size image

In terms of demographic covariates, participants from Cleveland had lower levels of resilience than those from Milwaukee, B = -0.34 (0.14), p = .018, while those recruited to the study following COVID-19 shutdowns had higher levels of resilience, B = 0.30 (0.16), p = .018. Those with more education and current insurance had higher levels of resilience, B = 0.22 (0.07), p = .002 and B = 0.27 (0.13), p = .039, respectively. Those with higher levels of economic hardship had higher levels of resilience, B = 0.16 (0.07), p = .019. Transgender women and non-binary participants reported lower levels of social support than cis male participants, B = -0.76 (0.32), p = .017, while those with current insurance reported higher levels of social support, B = 0.55 (0.26), p = .032. No demographic covariates were significantly associated with Black LGBTQ community connectedness or with current PrEP use.

Because this path model was fit using the MLR estimator (allowing the report of odds ratios) with mediation and a binary outcome variable, no standard fit indices were available. The model explained 23% of the variance in current PrEP use (p = .006).

Future PrEP use: path model

The path model focused on intentions to use PrEP in the future (Fig. 2), which included only those participants not currently using PrEP (n = 247), showed that individuals who experienced more day-to-day discrimination were more likely to use PrEP in the future, B = 0.48 (0.16), p = .002. Consistent with the current use model, participants who experienced more day-to-day discrimination had higher levels of resilience, B = 0.35 (0.11), p < .001, higher levels of social support, B = 0.44 (0.17), p = .011, and greater connection to the Black LGBTQ community, B = 0.21 (0.07), p = .005. Participants with more social concerns about PrEP use reported a lower connection to the Black LGBT community, B = -0.11 (0.06), p = .037, and were less likely to use PrEP in the future, B = -0.23 (0.09), p = .013. Participants with higher levels of social support were more likely to use PrEP in the future, B = 0.15 (0.05), p = .005. There was a significant and positive indirect effect of day-to-day discrimination on likelihood of future PrEP use via social support, B = 0.07 (0.04), 95% CI [0.01, 0.15], p < .05. This indirect effect indicated that social support partially mediated the effect of day-to-day discrimination on likelihood of future PrEP use.

Fig. 2
figure 2

Path model showing associations between intersectional stigma, resilience, and likelihood of future PrEP use in a sample of young sexual minority men in the Midwest who are not current PrEP users (N = 247). Unstandardized linear regression coefficients (Bs) are presented. The model also included correlations between resilience, social support, and Black LGBT community connectedness (paths not shown); there were moderate correlations between the constructs, rs = 0.30–0.39, ps < 0.001. Demographic covariates (age, gender identity, ethnicity, sexual orientation, education, economic hardship, insurance status, number of sexual partners in the past 30 days, psychological distress, enrollment city, and enrollment following the COVID-19 shutdowns) were also included in the model (paths not shown). *** p < .001 ** p < .01 * p < .05

Full size image

In terms of demographic covariates, participants recruited to the study following COVID-19 shutdowns reported a lower likelihood of using PrEP in the future compared to those enrolling before the shutdowns, B = -0.60 (0.24), p = .012. Participants who identified as bisexual or as having another sexual orientation reported a lower likelihood of future use compared to gay individuals, B = -0.57 (0.21), p = .006 and B = -0.92 (0.24), p < .001, respectively. Those with higher levels of economic hardship and those with current insurance (vs. no insurance) were more likely to use PrEP in the future, B = 0.21 (0.10), p = .042 and B = 0.73 (0.19), p < .001, respectively. Participants from Cleveland had lower levels of resilience than those from Milwaukee, B = -0.33 (0.14), p = .019. Latinx participants had higher levels of resilience than non-Latinx participants, B = 0.50 (0.17), p = .004. Those with more education, higher levels of economic hardship, and current insurance (vs. no insurance) had higher levels of resilience, B = 0.19 (0.07), p = .007, B = 0.16 (0.07), p = .031, and B = 0.35 (0.13), p = .008, respectively. Those with other sexual orientations had lower levels of resilience compared to those identifying as gay, B = -0.39 (0.18), p = .026. Transgender women and non-binary participants reported lower levels of social support than cis male participants, B = -0.80 (0.34), p = .017, while Latinx participants (vs. non-Latinx participants) and those with current insurance (vs. those without insurance) reported higher levels of social support, B = 0.78 (0.29), p = .008 and B = 0.59 (0.25), p = .021, respectively. Finally, those with sexual orientations other than gay reported lower levels of Black LGBT community connectedness, B = -0.57 (0.13), p < .001.

The model was a good fit to the data, χ2(35, N = 247) = 22.41, RMSEA = < 0.001, CFI = 1.00, TLI = 1.12, and explained 27% of the variance in likelihood of future PrEP use (p < .001).

Discussion

Despite increasing attention to the effects of intersectional stigma and discrimination on the health of Black SMM and TW, [33, 39, 76, 77] little work has examined how these experiences influence PrEP outcomes. The present study examined the relationship between intersectional discrimination and (1) current PrEP use, and (2) the likelihood of future PrEP use among a sample of YBSMM and TW in two mid-sized Midwestern cities. Further, we examined the potential mediating roles of resilience, social support, and connection to the Black LGBTQ community. While our findings showed some unanticipated relationships, they provide insight into how discrimination and stigma may influence PrEP use and disparities.

Individuals who reported greater anticipated discrimination had lower odds of current PrEP use. Unexpectedly, our results also showed that while major discrimination was not associated with PrEP outcomes, daily experiences of intersectional discrimination in the past year were positively associated with likelihood of future PrEP use. This was surprising and stands in contrast to our qualitative research, which indicated that intersectional racism and homophobia can impede PrEP use among YBSMM. [21, 39] However, this finding does align with other recent studies. For example, research on the relationship between racism and PrEP use found that individuals who reported higher levels of internalized racism were more likely to report PrEP use. [78] Similarly, researchers have found higher levels of internalized homophobia among Black MSM who use PrEP compared to those who do not use PrEP. [79] Sexuality-related discrimination has also been found to be positively associated with PrEP use. [80] Collectively, this research highlights the complexity of stigma and discrimination and requires we move beyond basic assumptions that greater stigma or discrimination necessarily contribute to worse HIV prevention outcomes.

In both models, various aspects of discrimination were also associated with resilience factors. Individuals who experienced more day-to-day discrimination had higher levels of resilience, social support, and connection to the Black LGBTQ community. Individuals who experience greater discrimination may seek out social support and LGBTQ community connections in response to these experiences, which may, in turn, positively impact PrEP outcomes. These social connections can also increase positive perceptions of PrEP and likelihood of current and future use, [81] while mitigating the negative effects of discrimination. Previous research has found that SMM with stronger connections to gay communities may be more likely to know others on PrEP and may also have access to a greater variety of affirming PrEP messages and access. [50]

However, PrEP social concerns, a component of PrEP stigma, was negatively associated with likelihood of future PrEP use. This reflects prior research on PrEP stigma, wherein some individuals associate PrEP with being gay or sexually promiscuous, [39] which may be a deterrent to PrEP use. PrEP social concerns was also negatively associated with Black LGBTQ community connectedness. Higher levels of stigma may weaken connections to LGBTQ communities and in turn limit resources, access, and willingness to use PrEP. Our findings align with prior research highlighting the complexity of connection and attachment to the LGBTQ community, which has been identified to be both a source of risk and resilience with regard to sexual health. [50, 82, 83]

Our results highlight the importance of the Black LGBTQ community in supporting PrEP use for Black SMM and TW. We found that Black LGBTQ community connectedness mediates the effects of day-to-day discrimination and PrEP social concerns on current PrEP use. Connections to Black LGBTQ spaces that support and promote PrEP may increase comfort with the idea of PrEP and help reduce PrEP concerns. Among Black SMM, social support is a common strategy used to cope with racial discrimination. [40, 42] Researchers have found that social support may also be critical in helping sexual and gender minority individuals cope with HIV- and sexuality-related stigma, [84] and that connectedness to the gay community may reduce minority stress and mental distress symptoms. [72] Our findings highlight the importance of supporting and developing LGBTQ affirming spaces for YBSMM and TW to connect, build community, and support one another. Efforts to increase PrEP use among YBSMM and TW should build on existing community resources and networks, particularly Black LGBTQ spaces. Funding initiatives that enhance existing support networks and provide needed support for coping with stigma and discrimination may also support PrEP use among Black sexual and gender minority individuals.

Given the importance of social support and community connectedness identified in this study, it is concerning that trans and gender nonbinary individuals reported lower levels of social support. Similarly, individuals who identified as bisexual or a sexual identity other than gay had lower levels of connectedness to the Black LGBTQ community compared to those who were gay, and they also had a lower likelihood of future PrEP use. This supports prior research demonstrating lower levels of PrEP use among bisexual men compared to gay men. [85, 86] This may also be related to LGBTQ community connectedness and the challenges some bisexual and non-gay-identifying individuals face in navigating gay centric social spaces and obtaining gay community support. [87] These findings highlight the need to support gender and sexually diverse individuals, who may not be receiving support in currently available LGBTQ spaces and communities.

Additional research is needed to better understand the role of resilience, social support, and community connectedness in supporting PrEP use. For example, understanding how Black LGBTQ community connectedness supports PrEP use can provide needed information to grow and support Black LGBTQ community resources. Additionally, we did not find an association between resilience and PrEP outcomes, and future research may help identify aspects of resilience that are unmeasured in this study. Current measures of resilience may be insufficient at capturing all of the experiences of YBSMM and TW, particularly in response to intersectional discrimination, or may not reflect how YBSMM and TW conceptualize resilience. Individuals may elicit greater resilience from social support and community connectedness rather than “coping with stress,” “bouncing back,” or being “able to handle” adversity, the individual-level resilience concepts measured by common resilience measures. [70] Future qualitative research is needed to understand YBSMM’s conceptualizations of and experiences with resilience and coping with discrimination and stigma in order to inform the development of resilience-based interventions.

This study has limitations. First, recruitment extended from 2018 to 2020, spanning a changing political, social, and health care landscape, which may have influenced use and perceptions of PrEP over time. Enrollment was temporarily suspended from March-July 2020 due to the COVID-19 pandemic. However, our analyses did account for when participants enrolled in the study. Second, the study was dependent on self-report measures, and responses may have been influenced by recall bias or social desirability bias. Related, current and prior PrEP-use were not verified by medical records, and we did not capture length of time on PrEP or reasons for discontinuation. We examined participants’ self-reported likelihood to use PrEP in the future, yet this does not necessarily translate into using PrEP. Additionally, there was no timeframe included in the question about future PrEP use, so it is unclear whether participants were interested in using PrEP in the short- or long-term future. Third, this study’s cross-sectional design limits our ability to draw causal inferences and limits our interpretation of results and ability to determine directionality. It is possible, for example, that limited connection to Black LGBTQ communities contributes to greater PrEP stigma, rather than PrEP stigma weakening connection to Black LGBTQ communities, as identified in this study. Longitudinal research is needed to further test the associations identified in this study and understand how intersectional discrimination influences PrEP uptake and persistence over time. Finally, there remains a lack of consensus about the best ways to quantitatively measure intersectional discrimination. [32, 88] In this study, we used the intersectional discrimination index, given its alignment with intersectionality and focus on any aspects of a participants’ identity. [65] That said, participants may not have been considering their sexual and racial identities when answering the questions.

Importantly, this is one of few studies that have examined how intersectional discrimination relates to PrEP use among young Black sexual minority individuals, and our findings highlight the need for more research on this topic. Specifically, greater attention is needed to the complexities of the relationships between intersectional discrimination, resilience, social support resources, and HIV prevention efforts including PrEP. Finally, an intersectional approach to understanding stigma and discrimination has been critical in advancing our understanding of the effects of discrimination on HIV outcomes for marginalized populations. Stigma scholarship must continue to incorporate intersectional approaches and contribute to the development of intersectional interventions that address multiple mechanisms of stigma.

Citation diversity statement

Recently, researchers have shed light on the racial, ethnic, and gender inequities in citation practices in numerous academic fields. [89,90,91] Research has demonstrated bias in citation practices, such that papers authored by women and racial and ethnic minorities are under cited relative to the number of those papers in the field. Accordingly, during our writing process, we interrogated the literature we read and cited, and aimed to cite research that reflects the diversity of the field in thought, method, and gender, race, and ethnicity of authors. Although this approach is insufficient for addressing these biases, we aim to raise awareness about these inequities and encourage other scholars to examine their citation decisions.