Black sexually minoritized men (SMM) and transgender women (TW) in the USA continue to experience HIV-related health disparities (Centers for Disease Control & Prevention, 2021a, b; Hess et al., 2017). While new HIV infections significantly decreased among White SMM from 2010 to 2019, incidence did not decrease among Black SMM and TW during this same time period (Pitasi et al., 2021). Ample research, including Baral’s modified social ecological model, has suggested that disparities in HIV prevention uptake are not due to lack of uptake but multilevel barriers to care including individual and network factors (Baral et al., 2013). Millet et al. (2007a, b) and Maulsby et al. (2014) found that network-level factors (i.e., sexual assortativity or homo-phily) could begin to explain racial disparities in HIV vulnerability among SMM (Maulsby et al., 2014; Millett et al., 2007a, b). Black SMM were found to have sexual partners who were also Black at a threefold higher level than chance alone (Raymond & McFarland, 2009). In the context of this study, this means that Black SMM and TW may be more likely to have sex with other Black SMM and TW (relative to White SMM and TW for example), groups experiencing the highest HIV prevalence, thereby increasing vulnerability to HIV relative to sex with other racialized groups (Maulsby et al., 2014; Millett et al., 2007a, b).

Substance use and HIV are interrelated epidemics (Baral et al., 2013), or a syndemic (Singer & Clair, 2003; Singer et al., 2017). Black SMM and TW experience enhanced vulnerability to substance use during sex despite having lower rates of substance use and sexual behaviors with heightened HIV vulnerability (Callander et al., 2019). Not only are substance use and sex-drug use higher among White SMM, but also the consequences are greater for Black SMM and TW due to a lack of access to biomedical prevention options (Centers for Disease Control & Prevention, 2018), health disinvestment (Nuriddin et al., 2020), and within-group mixing (Maulsby et al., 2014; Millett et al., 2007a, b; Schneider et al., 2013). Sex-drug use could increase sexual behaviors with heightened vulnerability to HIV, such as condomless sex (Yu et al., 2015) or sex with someone living with HIV (LWH) or unknown serostatus (Sewell et al., 2017).

Social norms surrounding a behavior within a network have been shown to be associated with alcohol use (Knox et al., 2019), drug use (Schneider et al., 2013), and sexual behaviors with heightened vulnerability to HIV such as group sex and sex-drug use during sex (Shrader et al., 2023). Additionally, sex-drug use is associated with negative HIV outcomes among Black SMM (Morgan et al., 2016). Substance use plays a complex role in HIV acquisition: daily marijuana use (Knox et al., 2022) and stimulant use have been associated with HIV acquisition (Pakianathan et al., 2018) perhaps because SMM may use substances to cope with negative emotions, in part due to intersecting discrimination at societal levels (Duncan et al., 2020; English et al., 2020; Feinstein & Newcomb, 2016). Not only does sex-drug use enhance sexual arousal and pleasure during sex (Feinstein & Newcomb, 2016; Wray et al., 2020), but also alcohol or drug use may decrease internalized homophobia to make sex more pleasurable (Feinstein & Newcomb, 2016). Sex-drug use is also linked to behavioral disinhibition through the moderation of sensation seeking (Newcomb et al., 2011) and the impairment of judgement to reduce condom use (Allen et al., 2015; Steele & Josephs, 1990). Previous studies have found that 58% of Black SMM and TW report using drugs or alcohol before sex in the past 6 months (Dangerfield et al., 2021) and 33% of Black SMM reported stimulant use during sex at least monthly (Mimiaga et al., 2010); however, it is unclear whether these substances are used to enhance sex and what other factors at the individual level (i.e., ego) or sexual partner level (i.e., alter) influence condom use during sex.

There is also a gap in characterization of the specific drug combination during sex-drug use among Black SMM and TW which may enhance HIV vulnerability. Previous studies have focused on the utility of latent class analyses and found that SMM who use substances can be typologized into different patterns of substance use, group assignments which are associated with increased HIV risk behaviors such as condomless sex (McCarty-Caplan et al., 2014; Scholz-Hehn et al., 2022; Wong et al., 2020). For example, people who engage in sex-drug use (also known as “chemsex” in European and Asian contexts) have been grouped into labels such as “intense drug users” (i.e., extensive use of different types of drugs including poppers, methamphetamine, GHB, and erectile dysfunction drugs),“sex drug users” (i.e., use of methamphetamine, poppers, club drugs (MDMA, ecstasy, ketamine, GHB), and erectile dysfunction drugs) (McCarty-Caplan et al., 2014), and “polyvalent users” (i.e., use of a wide range of substances including cannabis, poppers, MDMA, cocaine, amphetamine, methamphetamine, ketamine, GHB) (Scholz-Hehn et al., 2022). However, these studies did not examine which drugs were specifically used together to enhance the sexual experience and instead focused on broader drug use. Exploration of a two-mode affiliation network, whereby modes are the drugs used during sex and the sexual encounters (Borgatti & Everett, 1997), could provide greater insight into which drugs are used together during sex among Black SMM and TW while accounting for other critical factors operating at relevant network levels. This understanding could refine intervention components to be inclusive and tailored to specific potential sex-polydrug use, or the use of 2 or more drugs, or one or more drugs and alcohol use during sex.

Previous studies have focused on general substance use and associations of condom use without examining condom use at the sexual partner level using an egocentric approach. Other studies have also focused on describing typologies of substance use without examining specific sexual partner-level substance use and associations of condomless sex. To provide a clearer understanding of factors associated with condomless sex at the sexual partner level, including sex-drug and sex-polydrug use, this study aims to examine the sexual network-level associations of condom use during anal sex among Black SMM and TW living in Chicago, Illinois, USA. We hypothesize that there are associations of individual respondent-level (i.e., age, gender, sexual orientation, housing stability, education, relationship status, income, employment status, and adherence to PrEP or viral suppression) and sexual partner-level (i.e., gender, Black racial identity, type of partner, HIV serostatus difference, communication frequency, sex-drug use) factors, on the outcome of condomless sex, a sexual behavior with heightened vulnerability to HIV.

This study utilized baseline data from Black SMM and TW from the Neighborhoods and Networks (N2) Cohort Study. Black SMM and TW were recruited using convenience sampling at a community health center and clinic in Chicago, Illinois, USA, and peer referral sampling, which has been successfully used in recruiting this population previously (Coombs et al., 2014; Schneider et al., 2017). Respondent inclusion criteria included (a) identifying as Black or African American, (b) being age 16 years or older, (c) being assigned male at birth, (d) residing in the Chicago Metropolitan Statistical Area (MSA), (e) reporting at least one sexual encounter with a cisgender man or transgender woman in the past year, (f) being willing to wear a GPS device, and (g) having no plans to move from the Chicago MSA during the course of the study. Respondents provided informed written consent, and then, data were collected by trained interviewers from January 2018 until December 2019 using interviewer-administered computer-assisted assessments in a private room at the study site. Assessments lasted 1–2 h in length and consisted of respondents (i.e., egos) completing a sociodemographic and behavioral assessment and an egocentric network inventory of sexual partners (i.e. alters) and providing blood for an HIV test and viral load count. Respondents were provided $150 for their time. Additional information about the study can be found elsewhere (Duncan et al., 2019). We used the STROBE guidelines to prepare our manuscript (Vandenbroucke et al., 2007).

A total of 412 respondents were enrolled into the N2 study, of which 379 Black SMM and TW provided information about 998 sexual partners. There were complete data available for 352 respondents and 933 sexual partners. Respondents reported being a median age of 25 years. The majority of respondents identified as cisgender men (89%), did not experience housing instability (70%), were high school educated (87%), and reported single relationship status (62%) and an income below $20,000 (66%). Of respondents, 21% reported currently using PrEP and 26% were virally suppressed. We found that of respondents, 61% reported sex-drug use, 40% reported sex-polydrug use, and 79% reported condomless sex. Additional information about respondents can be found in Table 1.

Respondents reported a mean of 2.6 sexual partners and described sexual partners as majority cisgender men (84%), Black (86%), casual sex partners (68%), and serostatus concordant (59%). Respondents reported sex-drug use to enhance sex with 45% of sexual partners; sex-polydrug use to enhance sex with 26% of respondents; exchanging sex for money, food, drugs, or shelter with 6% of sexual partners; and condomless sex with 56% of sexual partners. There were statistically significant differences between sexual partners with whom respondents had condomless sex with relative to whom they did not have condomless sex regarding gender (of sexual partners that respondents had condomless sex with, 80% were cisgender men, 13% were cisgender women, and 7% of transgender women, but among those sexual partners that respondents consistently used condoms with during sex, 89% were cisgender men, 7% were cisgender women and 3.5% were transgender women; p < 0.001), partner type (of sexual partners that respondents had condomless sex with, 35% were main partners, 61% were casual partners, and 4% were exchange partners, but among those sexual partners that respondents consistently used condoms with during sex, 17% were main partners, 77% were casual partners, and 6% were exchange partners; p < 0.001), and communication frequency (of sexual partners that respondents had condomless sex with, 39% communicated daily, 38% communicated weekly, 20% communicated monthly, and 20% communicated yearly, but among those sexual partners that respondents consistently used condoms with during sex, 25% communicated daily, 30% communicated weekly, 21% communicated monthly, and 24% communicated yearly; p < 0.001). In addition, there were statistically significant differences in condom use based on HIV homophily: of sexual partners that respondents had condomless sex with, 48% were homophilous on HIV-negative serostatus, 16% were homophilous on LWH, 20% had different HIV serostatus, and 17% of sexual partners’ HIV serostatus was unknown, but among those sexual partners that respondents consistently used condoms with during sex, 45% were homophilous on HIV-negative serostatus, 7% were homophilous on HIV-positive serostatus, 26% had different HIV serostatus, and 21% of sexual partners’ HIV serostatus was unknown (p < 0.001). Further, there were statistically significant differences in the use of drugs or alcohol to enhance sex (i.e., sex-drug use), with a higher percentage of sex-drug use being reported during condomless sex relative to consistent condom use during sex (49% vs 38%; p = 0.001). More specifically, sex-polydrug use (30% vs 20%; p = 0.002), mari-juana as a sex-drug (41% vs 26%; p < 0.001), and cocaine/crack as a sex-drug (6% vs 3%; p = 0.007) were higher during condomless sex relative to consistent condom use during sex. Table 2 provides additional information about sexual partners and sex-drug use, stratified by whether respondents ever had condomless sex with the sexual partner in the past 6 months.

We sought to identify sexual partner-level associations of condomless sex among Black SMM and TW and explore sex-polydrug use using two-mode network analyses. We found that the sexual partner’s gender and sex-polydrug use were positively associated with condomless sex, whereas type of partnership (i.e., casual or exchange partner relative to main partner), having an HIV-negative serostatus and having a partner with an HIV-negative serostatus, having a different HIV serostatus with sexual partners, not knowing a sexual partner’s HIV status, and having an income below $20,000 were negatively associated with condomless sex. Respondents who reported using marijuana or cocaine/crack were more likely to engage in condomless sex. Our findings provide greater insight into which drugs are used together during sex, and the association of sex-drug use and condomless sex. Our findings could contribute towards the refinement of substance use and HIV prevention intervention components to be inclusive and tailored to communities who engage in sex-polydrug use.

Importantly, we found that sex-polydrug use to enhance sex was found to be associated with condomless sex, but sex-drug use (i.e., sex-drug use with only one drug or alcohol) was not associated with condomless sex. To the best of our knowledge, this may be a novel contribution to the literature. The most used sex-drugs by Black SMM and TW included marijuana (34%), alcohol (31%), ecstasy (6%), and cocaine/crack (5%); however, only marijuana and cocaine/crack use were associated with condomless sex. Our findings support previous studies which establish that Black SMM and TW who reported marijuana use (Bustamante et al., 2022; Kelly et al., 2013; Morgan et al., 2016) or cocaine/crack use (Kelly et al., 2013; Tobin et al., 2011) during sex also report condomless sex. Contrary to previous studies, we did not find an association between alcohol use (Hess et al., 2015; Purcell et al., 2001) or methamphetamine use (Ober et al., 2009) with condomless sex in our unadjusted models. Methamphetamine use was low (2.7%) in our sample and lower than cocaine/crack use: a finding in alignment with the existing literature on Black SMM (Millett et al., 2007a, b; Ober et al., 2009). This could be due to the wider availability of marijuana and cocaine/crack in the Chicago area in 2019 (Drug Enforcement Administration, 2019) and the low cost of cocaine/crack relative to other illicit drugs (Palamar et al., 2015). However, as the availability and use of methamphetamine continues to increase among Black communities in Chicago, future interventions may need to redirect focus to also include methamphetamine use if usage increases (Drug Enforcement Administration, 2019). This supports the need for HIV prevention interventions to be tailored for the geographic region, including the unique substance use environment and availability. Future qualitative research is needed to understand the context and characteristics of substance use during sex and how substance use may be introduced into sex (or vice-versa).

Although only 45% of sexual relationships were contextualized by drug or alcohol use during sex, the majority (61%) of respondents reported drug or alcohol use to enhance sex in the past 6 months and 26% of Black SMM and TW reported engaging in sex-polydrug use. Previous research on sex-drug use has suggested that drug use can enhance sexual arousal and more pleasureful sex (Wray et al., 2020). Marijuana has been established previously as increasing sensitivity, pleasure, and emotional intensity of sex and orgasm (Palamar et al., 2018). Thus, Black SMM and TW may engage in condomless sex after or concurrent with the use of marijuana or cocaine/crack with the goal of engaging in pleasurable sex or “sensation seeking” (Kalichman & Rompa, 1995). It has been established that condom use during sex can diminish pleasure (Randolph et al., 2007; Siegler et al., 2014) or performance (Shrader et al., 2021; Siegler et al., 2014), be a “turn-off” (Crosby et al., 2008), and interfere with the expression of love, intimacy, and trust (Flowers et al., 1997; Mitchell, 2014; Mitchell et al., 2012; Worth et al., 2002). Our study found that Black SMM and TW reported condomless sex with 45% of their sexual partners, suggesting that additional HIV prevention efforts are needed, which should be love-, intimacy-, pleasure-, and trust-centered and account for HIV prevention strategy preferences.

In addition to condoms, biomedical prevention including ART and PrEP can be offered at supraoptimal levels (Fauci, 2016). As our sample had both PLWH- and HIV-negative people and less than half of respondents reported current PrEP use or were found to be virally suppressed (47%), this finding necessitates the importance of increasing the use of biomedical prevention strategies such as ART (Eisinger et al., 2019), oral PrEP (Liu et al., 2013), injectable PrEP (Network, 2017), and PrEP-on-demand (2-1-1 PrEP) (Molina et al., 2015) in addition to condoms among Black SMM and TW. Although PrEP or viral suppression (i.e., U = U) is not a replacement for condom use, it is sometimes perceived that PrEP could replace condoms (Mitchell et al., 2016) or that PrEP/ART could confer protection in instances when condom use was intended but did not occur (Bourne et al., 2017). Thus, HIV prevention interventions designed to reduce HIV transmission through sex could focus on initiation and adherence to long-acting biomedical prevention strategies (i.e., injectable PrEP, ART).

This study is not without limitations. First, there may be social desirability bias within our study as survey items were interviewer-administered. Because our data were selfreported and collected by an interview-administered egocentric inventory, our data may be subject to recall bias. Additionally, we used peer referral recruitment to recruit respondents and did not adjust for potential dependencies using an estimator. However, we conducted sensitivity analyses which controlled for possible dependencies related to peer referral recruitment. Further, substance use and condomless sex were assessed at the sexual partner level, and we did not account for specific acts of sex between respondents and their sexual partners because the data were unavailable, which may have provided additional clarity. In addition, because this is a cross-sectional study, we cannot make inferences of causality. Finally, it is important to note that only one respondent reported using heroin with a sexual partner: statistical associations with heroin must be interpreted with caution.

Although sex-drug use was not associated with condomless sex, sex-polydrug use was associated with condomless sex. Sex-polydrug use may have additive and interactive effects related to the substances being used which may further impair HIV risk reduction decision-making. HIV prevention interventions for Black SMM and TW designed to reduce HIV transmission should address sex-polydrug use through sex-positive and love-, intimacy-, trust-, and pleasure-centered harm reduction strategies. Additionally, there is not only a need to continue the provision but also the promotion of biomedical prevention and care options at supraoptimal levels.