Overall Sample for Analysis
The 28 articles identified by the search process were generated from 18 unique studies. Authors published 12 articles that contained only qualitative findings, 12 articles that contained only quantitative results, and 4 articles that contained both qualitative and quantitative findings. Sources consisted of characteristics of social networks (n=15), psychosocial factors affecting vulnerability for HIV transmission (n=13), and HIV prevention intervention studies (n=4) (Figure 2).
Social Network Characteristics of GFNs
Designs and Methods
The publications that were categorized in this theme were based on 7 cross-sectional quantitative analyses (Cahill et al., 2018; Holloway et al., 2014; Hosek et al., 2019; Murrill et al., 2008; Sanchez et al., 2010; Young et al., 2017; M. C. Zarwell & W. T. Robinson, 2018), 4 qualitative analyses (Arnold & Bailey, 2009; Dickson-Gomez et al., 2014; Galindo, 2013; Kubicek, Beyer, et al., 2013), and three mixed-method analyses (Holloway et al., 2012; Kipke et al., 2013; Kubicek, McNeeley, et al., 2013). Detailed information on the study design and methods is included in Table 2.
Sample Characteristics
Out of the 14 publications that were categorized as Social Network Characteristics of GFNs, 11 were based on data collected exclusively from HBC (Arnold & Bailey, 2009; Cahill et al., 2018; Galindo, 2013; Holloway et al., 2014; Holloway et al., 2012; Hosek et al., 2019; Kipke et al., 2013; Kubicek, Beyer, et al., 2013; Kubicek, McNeeley, et al., 2013; Murrill et al., 2008; Sanchez et al., 2010). Three studies were based on data collected from members of gay families and/or ballroom houses (Young et al., 2017), and constructed families (Dickson-Gomez et al., 2014; M. C. Zarwell & W. T. Robinson, 2018). Most publications (n=9) were based on studies conducted in Los Angeles (Holloway et al., 2014; Holloway et al., 2012; Kipke et al., 2013; Kubicek, Beyer, et al., 2013; Kubicek, McNeeley, et al., 2013), and New York City (Cahill et al., 2018; Galindo, 2013; Murrill et al., 2008; Sanchez et al., 2010). Young et al.'s study included the largest sample size (N=618) (Young et al., 2017) and Galindo’s study included the smallest sample size (N=20) (Galindo, 2013).
Study Purposes
Primary topics of investigation included risk and protective factors for sexual health, roles taken up amongst GFN members, social inequities along lines of race and culture, as well as differences in social versus sexual network members and GFN members versus non-GFN members in risk and protective factors (Arnold & Bailey, 2009; Dickson-Gomez et al., 2014; Holloway et al., 2014; Kubicek, Beyer, et al., 2013; Sanchez et al., 2010; Young et al., 2017; M. C. Zarwell & W. T. Robinson, 2018). Authors identified HIV prevention activities that occurred within the Los Angeles HBC and gauged HBC leaders' thoughts on culturally relevant approaches to HIV prevention (Holloway et al., 2012). Another study indicated various domains of resiliency and assets among GFNs that could be leveraged for intervention development (Kubicek, McNeeley, et al., 2013). Authors also investigated HIV risk behaviors and HIV prevalence (Hosek et al., 2019; Murrill et al., 2008), HIV-related stigma within the context of HBC (Galindo, 2013), rates of sexual and gender identity disclosure to healthcare providers, and healthcare access (Cahill et al., 2018). A more recent publication was based on baseline findings from a hybrid-design clinical trial to test the efficiency of a community-based sexual health intervention designed for HBC members (Hosek et al., 2019).
Major Findings
Individuals living with HIV were evenly spread out across GFNs and not clustered in specific networks as previously expected (Young et al., 2017). However, Black MSM who were members of ballroom houses seemed to have a high prevalence of HIV, being unaware of HIV status, high rates of CAI, lack of health insurance (approximately 45% of participants in Young et al’s Chicago-based sample)(Young et al., 2017), substance use before or during sex, and higher rates of exchange sex (defined as sex in exchange for money and/or goods)(Hosek et al., 2019; Kipke et al., 2013; Murrill et al., 2008; Sanchez et al., 2010; Young et al., 2017). Black MSM in HBC specifically who reported a negative or unknown HIV status and were not tested for HIV infection within the past year had increased odds of being diagnosed with HIV while enrolled in the study (Murrill et al., 2008).
Protective features of GFNs have also been reported in the literature, including fewer risk behaviors on average among members of GFNs and frequent HIV/STI testing. Many participants reported at least one healthcare provider whom they saw on a regular basis, participation in at least one HIV prevention program, and high rates of lifetime HIV and STI testing (Cahill et al., 2018; Holloway et al., 2012; Young et al., 2017; M. C. Zarwell & W. T. Robinson, 2018).
Qualitative studies revealed that GFNs, primarily ballroom houses, had a longstanding history of conducting organically fostered methods of HIV prevention, termed intravention, by Bailey (Arnold & Bailey, 2009). Intravention is a term used to describe prevention and intervention activities organically taken up by communities encountering any given health disparity (Arnold & Bailey, 2009; Friedman et al., 2004). Intraventions were first introduced by the ballroom house, House of Latex in New York City during the early 90’s and subsequently adopted by the House of Omni in Chicago (Phillips et al., 2011). HBC and GFN members partake in the processes of family making, reconstitution of gender/gender roles, and competition as a method to withstand and creatively respond to socioeconomic marginalization (Bailey, 2009). Arnold and Bailey contextualized the importance of “nurturing house mothers,” who often provided safe sex advice based on life experience. Equal importance was placed on “guiding house fathers,” who held members accountable for bettering themselves economically as a means of addressing socioeconomic and systemic influences of HIV vulnerability (Arnold & Bailey, 2009). Thus, it is plausible that emotional support from house parents and members may mitigate the impact of various socioeconomic stressors, including homophobia, transphobia, and HIV-related stigma, among other forms of stress. HIV-related stigma within HBC houses was found to be the root of HIV concealment from fellow HBC members and, at times, subsequent withdrawal from the community (Galindo, 2013).
In addition to serving as structures for the deployment of HIV intraventions, GFNs are social homes and sources of identity affirmation for lesbian, gay, bisexual, transgender, and queer youth of color (Arnold & Bailey, 2009; Dickson-Gomez et al., 2014; Holloway et al., 2014; Kubicek, Beyer, et al., 2013; Kubicek, McNeeley, et al., 2013) and other non-heteronormative individuals. Since these networks act as social homes, the roles adopted by members often include the provision of various forms of social support to other members, making GFNs ripe for culturally-tailored HIV prevention interventions (Arnold & Bailey, 2009; Dickson-Gomez et al., 2014; Kubicek, Beyer, et al., 2013).
Psychosocial Predictors of HIV Risk
Designs and Methods
Researchers conducted 11 distinct studies, which resulted in 13 published works. Seven articles were based on 5 unique quantitative analyses (Arnold et al., 2018; Cahill et al., 2018; Holloway et al., 2014; Hotton et al., 2020; Sanchez et al., 2010; Schrager et al., 2014; M. C. Zarwell & W. T. Robinson, 2018), and 5 articles based on 5 unique qualitative analyses (Alio et al., 2020; Castillo et al., 2012; Horne et al., 2015; Lemos et al., 2015; Levitt et al., 2017; Telander et al., 2017). One study conducted a mixed-method analysis based on qualitative interviews and self-report quantitative survey measures (Alio et al., 2020). Two of the 5 quantitative papers and 4 of the qualitative papers were based on studies that incorporated community-based participatory research principles (Alio et al., 2020; Castillo et al., 2012; Holloway et al., 2014; Lemos et al., 2015; Schrager et al., 2014; Telander et al., 2017). All results were based on cross-sectional observations (Table 2). Authors of quantitative publications reported utilizing Computer-Assisted Self Interviewing (CASI) to collect data (Arnold et al., 2018; Cahill et al., 2018; Holloway et al., 2014; Hotton et al., 2020; Sanchez et al., 2010; Schrager et al., 2014; M. C. Zarwell & W. T. Robinson, 2018). Grounded theory was commonly used to guide much of the qualitative analyses (Alio et al., 2020; Horne et al., 2015; Levitt et al., 2017).
Sample Characteristics
Authors identified a variety of psychosocial predictors of behaviors commonly associated with increased vulnerability for HIV transmission among GFNs. Samples consisted primarily of HBC parents and participants (Alio et al., 2020; Arnold et al., 2018; Cahill et al., 2018; Castillo et al., 2012; Holloway et al., 2014; Hotton et al., 2020; Lemos et al., 2015; Sanchez et al., 2010; Schrager et al., 2014; Telander et al., 2017). Two articles were based on samples of individuals who belonged to GFNs largely, some of whom were HBC-affiliated families (houses) while others were not (Levitt et al., 2017; M. C. Zarwell & W. T. Robinson, 2018). One article was based on a sample of individuals seeking to join or already belonging to gay families (Horne et al., 2015). Zarwell and Robinson (M. C. Zarwell & W. T. Robinson, 2018) and Levitt et al. (Levitt et al., 2017) focused on “constructed families,” which was an encompassing term for ballroom houses, gay families, pageant houses and other forms of participant-reported named chosen kinship groups. Information on geographical locations of data collection can be found in Table 2.
Zarwell and Robinson's publication had the largest number of participants (N=533), which was based on data collected in New Orleans, Louisiana (M. C. Zarwell & W. T. Robinson, 2018). Studies conducted in New York City (Cahill et al., 2018; Sanchez et al., 2010) and Los Angeles (Arnold et al., 2018; Holloway et al., 2014; Schrager et al., 2014) had moderate samples (N=209 to 367). Horne et al.'s qualitative analysis based on data collected in Memphis, Tennessee had the lowest number of participants (N=10) (Horne et al., 2015), followed by Alio et al’s qualitative analysis which consisted of 14 HBC leaders in Western New York (Alio et al., 2020).
Study Purposes
Publications that were categorized in this theme covered a variety of purposes related to sexual health and risk (Arnold et al., 2018; Castillo et al., 2012; Holloway et al., 2014; Horne et al., 2015; Sanchez et al., 2010; Schrager et al., 2014; M. C. Zarwell & W. T. Robinson, 2018). Some authors incorporated measures of social support (e.g. emotional and tangible support received from the fellow house and/or gay family members), connectedness to social networks, and how these concepts correlated to measures of sexual risk behaviors and substance use (Arnold et al., 2018; Holloway et al., 2014; Schrager et al., 2014). Lemos et al. incorporated the Diffusion of Innovation framework to explore social norms around HIV risk behaviors and how HIV prevention services can be modified to be more culturally responsive to the needs of HBC participants (Lemos et al., 2015). Alio et al. explored efforts by HBC leaders in Western New York (Rochester and Buffalo, NY) to address HIV and social issues among participants, as well as HBC leaders’ knowledge of HIV prevention strategies, including HIV vaccine clinical trials (Alio et al., 2020).
Major Findings
Factors associated with decreased risk of HIV included behavioral factors (STI testing, reduced instances of CAI, decreased lifetime substance use) and psychosocial factors (having sexual partners with a high level of influence on the participant’s life, being a member of a house, increased number of supportive social network members (Arnold et al., 2018; Holloway et al., 2014; Schrager et al., 2014; M. C. Zarwell & W. T. Robinson, 2018). In addition, participants belonging to networks that provided high levels of HIV-related support, as well as networks that had high levels of sexual and gender identity homophily, had significantly reduced instances of CAI and were more likely to have been tested for HIV and STIs in the past 6 months than those who had less social support (Arnold et al., 2018). Among MSM with greater numbers of sexual partners, having sexual partners who were highly influential (such as house parents) was associated with less risky sexual behavior (Schrager et al., 2014). Lastly, belonging to a named GFN, regardless of HBC affiliation, was also associated with less risky sexual behavior in comparison to participants who did not belong to such networks (Holloway et al., 2014; M. C. Zarwell & W. T. Robinson, 2018).
Quantitatively-measured factors that were associated with increased HIV risk included a previous diagnosis of an STI, increased number of sexual partners, multi- and white ethnicity, alcohol and illicit substance use, and younger age (Schrager et al., 2014; M. C. Zarwell & W. T. Robinson, 2018). Hotton et al., published findings that illuminate a plausible mechanism for increased CAI among HBC participants (Hotton et al., 2020); namely, that depressive symptoms are positively associated with socio-structural stressors (intimate partner violence, perceived HIV stigma, Post-Traumatic Stress Syndrome symptoms, socio-economic marginalization [history of homelessness and sex in exchange for shelter, money or drugs]) and that both depressive symptoms and socio-structural stressors are positively associated with CAI (Hotton et al., 2020).
In our review of qualitative studies that were categorized in the Psychosocial Predictors of HIV Risk theme, we found that GFNs provide coping strategies to their members that can contribute to increased or decreased HIV-risk behaviors, particularly within the face of intersectional marginalization (Horne et al., 2015; Levitt et al., 2017). More specifically, the quality of relationships moderate protective factors against HIV among family networks; however, stigma can act as a barrier against discussing HIV (Horne et al., 2015; Lemos et al., 2015; Levitt et al., 2017). For example, HIV stigma may prevent some members from disclosing their HIV status to fellow GFN members. On the other hand, GFN members’ demonstration of care regardless of HIV status is something that can reduce stigma among the family members. Additionally, values instilled within GFNs such as promoting high self-esteem and protection of family members, may serve as additional protective factors against HIV risk (Levitt et al., 2017; Telander et al., 2017). In contrast, some GFNs may promote sex work as a survival tool or encourage a risky sexual environment, thus promoting HIV risk factors (Castillo et al., 2012; Levitt et al., 2017). Additionally, in response to stigma, decisions to celebrate and self-define sexuality within affirming contexts may undermine HIV risk reduction efforts (Levitt et al., 2017).
HIV Prevention Intervention Studies
Designs and Methods
A total of 4 articles were categorized as HIV Prevention Intervention studies. Articles were categorized as HIV prevention intervention studies if they met the following criteria: 1) the article was focused on testing an intervention and 2) the article contained post-intervention findings. Two studies assessed community-informed interventions using quasi-experimental research designs, neither of which included control groups (Alio et al., 2014; Hosek et al., 2015). One intervention was developed based on data collected from in-depth interviews and collaborative workshops (Arnold et al., 2020). Finally, Castillo et al. identified risk factors that placed Young HBC Black MSM at increased risk for HIV and subsequently implemented structural processes to mitigate the impacts of such risk (Castillo et al., 2012). HIV prevention interventions were aimed at increasing social support, -HIV-related knowledge, access to health-specific resources, and willingness to participate in HIV vaccine research while decreasing HIV risk behavior and structural barriers to HIV testing (Alio et al., 2014; Arnold et al., 2020; Castillo et al., 2012; Hosek et al., 2015). Findings were primarily reported quantitatively (Alio et al., 2014; Hosek et al., 2015), whereas Arnold et al. (Arnold et al., 2020) shared only qualitative results, and Castillo et al. shared both qualitative and quantitative findings (Castillo et al., 2012)
Sample Characteristics
The four articles in this theme encompassed 4 unique HIV prevention intervention studies composed primarily of Black MSM and transwomen. All participants were members of HBC. One publication, by Alio et al., was based on Project VOGUE, conducted in Western New York State from 2009 to 2011 (Alio et al., 2014). The POSSE Project was piloted in Chicago, Illinois prior to 2015 by Hosek et al. (Hosek et al., 2015). Castillo et al. conducted a root cause analysis involving local HBC youth ages 13-24 from Philadelphia, but did not indicate the data collection period (Castillo et al., 2012). Arnold et al., collected data between August 2016 – December 2017 in Oakland, California (Arnold et al., 2020). Sample sizes varied across the studies, with 14 participants included in Project VOGUE and 406 participants included in the Chicago-based POSSE Project pilot.
Study Purposes
Castillo’s root cause analysis identified factors that placed HBC youth at an increased risk for HIV and implemented structural changes to increase HIV prevention and education (Castillo et al., 2012). Project VOGUE was designed to examine whether a culturally-tailored educational intervention increased willingness to participate in HIV vaccine research (Alio et al., 2014). The POSSE Project was conducted to test the preliminary feasibility and efficacy of a culturally-tailored HIV prevention intervention (Hosek et al., 2015). Arnold et al., worked with HBC members to co-create a mobile app aimed at promoting HIV testing, PrEP uptake, and linkage to HIV/STI care (Arnold et al., 2020).
Major Findings
The four HIV prevention interventions reviewed here were the only to be conducted with GFNs broadly, and were all conducted with the HBC (Alio et al., 2014; Arnold et al., 2020; Castillo et al., 2012; Hosek et al., 2015). Overall, these interventions addressed key drivers of vulnerability for HIV transmission such as CAI with multiple partners (of both known and unknown serostatus), self-reported HIV stigma, and HIV/STI testing (Alio et al., 2014; Castillo et al., 2012; Hosek et al., 2015). Increases in protective factors were noted, including the average amount of safer-sex conversations with GFN members across observation time points (Hosek et al., 2015), and HIV-related knowledge (Alio et al., 2014). Castillo et al’s intervention led to the implementation of several strategies to address HIV risk among young Black HBC MSM: prevention themed balls, where HBC community members were expected to incorporate HIV prevention messaging into their competition; offering incentives for HBC members to get tested at balls; holding local houses accountable for educating their youth on HIV prevention and treatment; the addition of a HBC-specific workshop for Philadelphia’s annual AIDS Education Month to discuss best practices and promising approaches. The HIV prevention intervention led by Arnold et al. identified features to make an mHealth app for HIV education, prevention, and care engaging. The mHealth app featured: a minimalist, modern design; relatable and personable faces visible on the app; positive and fun themes; community-generated content; and a safe space for sharing vital information (Arnold et al., 2020). Although some of the outcomes from the aforementioned interventions are promising with regards to reducing vulnerability to HIV transmission for Black HBC MSM, authors also detailed findings that point to a dire need for additional work. Collectively, the authors noted the need for multi-tiered approaches that address individual HIV prevention and treatment needs, while addressing the macro context of structural violence in which such prevention and treatment needs arise.