Understanding HIV Sexual Protection and Its Association With Substance Use During Sex Among MSM in an Era of Multiple Primary Prevention Products.

Abstract

To the Editors: In an era of effective1 and increasingly accessible2 daily oral HIV pre-exposure prophylaxis (PrEP) in the United States, the understanding of sexual risk and protected behavior among men who have sex with men (MSM) is changing.3 Before PrEP, condoms were the only primary HIV prevention product available; sexual risk was defined as lack of condom use during anal sex, with encounter-specific behavior being most useful for understanding contextual risk (eg, substance use).4 Furthermore, decades of research have documented evidence of associations between substance use and condomless sex among MSM,4–7 with only sex-linked methamphetamine (stimulant) and binge alcohol use consistently associated with condomless anal sex in a prior review of the literature on sexual event-level substance use and HIV risk behavior.4 This association of risk is generally thought of as a direct relation between cognitive-behavioral state of substance-using sex (for some substances), and risk because of impaired judgement, limited physical coordination, competing emotional priorities, and other contextual explanations in the “hot” moment of a sexual encounter.8 As use of daily PrEP increases in MSM populations,2 a more complex assessment is possible and needed to measure “protected” and “unprotected” sex through use of 2 recommended primary HIV prevention products, condoms and daily PrEP.3 Advanced measurement and analyses may provide better understanding of the nature of the relationship between substance use and daily PrEP, a prevention method typically taken in “cold” moments outside of the sexual encounter,8 as well as the “hot” moment for condom use.8 Herein, we present analyses that address (1) condom and/or PrEP use as protected sex, and (2) the association of sexual event-level substance use with condom and/or PrEP use. Data were from the Mobile Messaging for Men (M-cubed) Study, methods for which have been previously reported.9 Briefly, we recruited a total of 841 HIV-negative MSM in Atlanta, Detroit, and New York City through venue-based and online recruitment. Eligible men were ≥18 years old, lived in one of the 3 city metropolitan statistical areas (MSAs), and indicated anal sex with a male partner in the prior 12 months and being HIV-negative. In the baseline survey, participants reported on behavior during anal sex with their most recent casual partner in the prior 3 months, including condom and PrEP use, and whether they were drunk or buzzed on alcohol or had taken recreational or illicit drugs within 2 hours before or during sex. The primary outcomes for this analysis were reporting condomless anal sex, anal sex not protected by daily oral PrEP (only 5% of current PrEP users in this study10 had ever used it “on demand”),11 and unprotected anal sex (ie, not protected by condoms or PrEP). Multivariable logistic regression models evaluated the association between the primary outcomes and being under the influence of alcohol or drugs during that sexual event, controlling for race/ethnicity, age group, education level and city/MSA of residence, including regressions models of the most commonly used individual drugs during sex: marijuana, poppers (amyl nitrite), and methamphetamine. We also evaluated the associations of alcohol or drug use and demographic variables. Of 782 MSM who reported being HIV-seronegative, 553 (71%) reported on having at least one casual anal sex partner 3 months before the baseline survey (Table 1). Of the 553 sexual events, 331 (60%) did not use condoms (222 did), 343 (62%) did not use PrEP (210 did), and 183 (33%) did not use condoms or PrEP; 84% of PrEP users took pills daily for the last 7 days, indicating good adherence among men reporting current PrEP use. Overall, 147 anal sex events (27%) occurred while the participant was drunk/buzzed on alcohol, and 85 events (15%) occurred when the participant was under the influence of drugs (54% of which involved marijuana, 42% poppers, 24% methamphetamine, 34% other). Sex not protected by PrEP or unprotected anal sex (condomless and PrEP-less) was more common among men with less education, and sex while under the influence of drugs was more commonly reported by younger men. TABLE 1. - Event-level Substance Use and Sexual Risk for HIV Infection Among HIV-negative MSM: Descriptive and Multivariable Analysis Results in 3 U.S. Cities, 2018 Characteristic n (%) Overall Respondent Behavior Related to Most Recent Casual Anal Sex Encounter No Condom No PrEP Use UA (No Condom or PrEP) Drunk/Buzzed Drug Use n (%) AOR (95% CI) n (%) AOR (95% CI) n (%) AOR (95% CI) n (%) AOR (95% CI) n (%) AOR (95% CI) Overall 553 (100) 331 (60) 343 (62) 183 (33) 147 (27) 85 (15) Substance use (before/during encounter) Drunk/buzzed, yes 147 (27) 101 (67)* 1.45 (0.96 to 2.19)† 46 (31)* 1.57 (1.03 to 2.40)* 62 (42)* 1.66 (1.09 to 2.51)* — — 40 (27)* — No (ref) 406 (73) 238 (57) — 164 (40) — 121 (29) — — — 48 (12) — Drug use, yes 85 (15) 63 (72)* 1.62 (0.95 to 2.76)† 31 (36) 0.94 (0.56 to 1.58) 37 (43)* 1.44 (0.86 to 2.41) 40 (45)* — — — No (ref) 468 (85) 276 (58) — 179 (38) — 146 (31) — 110 (23) — — — Race/ethnicity Black/African American 119 (21) 65 (55) 0.58 (0.36 to 0.92)* 83 (70) 1.34 (0.82 to 2.19) 41 (34) 0.83 (0.51 to 1.35) 29 (25) 0.73 (0.43 to 1.21) 15 (13) 0.77 (0.39 to 1.50) Hispanic/Latino 83 (15) 48 (58) 0.76 (0.45 to 1.30) 46 (55) 0.79 (0.46 to 1.35) 25 (30) 0.75 (0.42 to 1.33) 24 (29) 0.88 (0.50 to 1.55) 18 (22) 1.49 (0.76 to 2.93) Other/mixed 70 (13) 38 (54) 0.71 (0.41 to 1.24) 44 (63) 1.29 (0.73 to 2.28) 21 (30) 0.92 (0.51 to 1.68) 12 (17) 0.47 (0.24 to 0.93)* 13 (19) 1.39 (0.68 to 2.84) White (ref) 281 (51) 180 (64) — 170 (61) — 96 (34) — 82 (30) — 39 (14) — Age group 18–29 yrs 238 (43) 129 (54) 0.67 (0.42 to 1.05)† 147 (62) 0.62 (0.38 to 0.99)* 75 (32) 0.62 (0.39 to 0.99)* 69 (29) 1.54 (0.92 to 2.58)† 21 (9)* 0.35 (0.18 to 0.66)* 30–39 yrs 177 (32) 114 (64) 1.05 (0.65 to 1.69) 102 (58) 0.58 (0.36 to 0.94)* 54 (31) 0.63 (0.39 to 1.04)† 50 (29) 1.61 (0.94 to 2.76)† 38 (21) 1.11 (0.63 to 1.98) 40+ yr (ref) 138 (25) 88 (64) — 94 (68) — 54 (39) — 28 (21) — 26 (19) — Education <=post HS educ (ref) 207 (37) 135 (65) — 152 (73)* — 94 (45)* — 60 (30) — 38 (18) — 4-year college degree 199 (36) 111 (56) 0.61 (0.40 to 0.94)* 113 (57) 0.51 (0.33 to 0.79)* 51 (26) 0.41 (0.26 to 0.64)* 56 (28) 0.92 (0.58 to 1.45) 31 (16) 0.71 (0.41 to 1.25) >=post-grad educ 147 (27) 85 (58) 0.64 (0.40 to 1.03)† 78 (53) 0.42 (0.26 to 0.67)* 38 (26) 0.40 (0.25 to 0.66)* 31 (22) 0.65 (0.39 to 1.09) 16 (11) 0.41 (0.21 to 0.79)* City/MSA Atlanta (ref) 194 (35) 124 (64) — 119 (61) — 64 (33) — 48 (25) — 24 (12) — Detroit 154 (28) 94 (61) 0.84 (0.53 to 1.33) 107 (69) 1.57 (0.98 to 2.52)† 60 (39) 1.29 (0.81 to 2.07) 43 (29) 1.11 (0.68 to 1.83) 22 (14) 1.21 (0.63 to 2.32) New York 205 (37) 113 (55) 0.70 (0.46 to 1.08) 117 (57) 0.96 (0.63 to 1.48) 59 (29) 0.91 (0.58 to 1.44) 56 (28) 1.18 (0.73 to 1.89) 39 (19) 1.63 (0.90 to 2.93) *P < 0.05†P < 0.10—, not applicable/analyzed; HS, high school; MSA, metropolitan statistical area; UA, unprotected anal sex (no condom or PrEP use). In multivariable analyses (Table 1), being drunk/buzzed on alcohol during recent anal sex with a casual partner was associated with condomless (marginally, AOR = 1.45, 95% CI = 0.96 to 2.19), PrEP-less (AOR = 1.57, 95% CI = 1.03 to 2.40), and unprotected anal (condomless and PrEP-less; AOR = 1.66, 95% CI = 1.09 to 2.51) sex. Any drug use at last anal sex with a casual partner was marginally associated with condomless sex (AOR = 1.62, 95% CI = 0.95 to 2.76), but not associated with PrEP-less or unprotected anal sex. Having condomless anal sex was less commonly reported by black MSM (versus white), by MSM aged 18–29 years (versus ≥40 years), and by men with more education (versus high school education or less). Sex not protected by PrEP was more common in men ≥40 years old (versus <40 years) and for men in Detroit (versus Atlanta). Having anal sex while drunk/buzzed on alcohol was less common among racial/ethnic groups other than black or Hispanic/Latino (versus white non-Hispanic). Having anal sex under the influence of drugs was less common among men aged 18–29 years (versus ≥40 years) and men with some postgraduate study (versus high school education or less). Having anal sex under the influence of drugs was more common among men recruited in New York (versus those living in Atlanta). As mentioned, the most commonly used individual drugs reported during the last anal sex with a casual partner (not in Table 1) were marijuana (n = 46, 8.4% of sample), poppers (n = 36, 6.5%), and methamphetamine (n = 18, 3.3%). In multivariable regression models consistent with those presented in the first 3 columns of Table 1, marijuana use (modeled in place of any drug use) was not associated with condomless, PrEP-less, or unprotected anal sex risk behaviors, similar to earlier sexual event-level studies on marijuana and condom use.4 Poppers use was associated with PrEP-less sex in a protective direction (AOR = 0.33, 95% CI = 0.16 to 0.71), meaning poppers use during sex was associated with daily PrEP use. Alternatively, methamphetamine use was marginally associated with condomless anal (AOR = 4.45, 95% CI = 0.98 to 20.19, P = 0.053) and unprotected anal (AOR = 2.57, 95% CI = 0.90 to 7.29, P = 0.077) sex, consistent with prior studies on methamphetamine use and condom use among MSM.6,7 None of these 3 individual drugs were linked to PrEP-less sex. Thus, methamphetamine and drunk/buzzed alcohol use during anal sex are situationally linked to sexual risk in “hot” moments of sex, and only drunk/buzzed alcohol use was linked to “cold” moments of risk for not currently taking PrEP.8 Although the situational context of drunk/buzzed sex and condom use is more clearly defined,4 more research on larger substance using MSM populations is needed on the nature of drunk/buzzed sex association with daily PrEP use. For example, is it attributable to chronic alcohol use/abuse or another third variable linked to alcohol use during sex that accounts for the more distal association with daily PrEP? These results advance understanding of complex relationships between sex-linked substance use and use of primary HIV prevention products—condoms and daily PrEP—in 2 ways. First, our analysis considers primary protection of HIV during anal sex for HIV-negative men to be use of condoms and/or daily PrEP in an era of multiple options, and we describe frequencies of use of those methods for protection among MSM within a large subsample of PrEP-users in a multi-city/MSA sample. Second, we evaluated the association of substance use during recent anal sex with a casual partner with the 2 primary HIV prevention products for MSM, including one generally taken during hot moments of adherence (condom use) and one generally taken during cold moments of adherence (daily PrEP).8 Being buzzed or drunk on alcohol during anal sex was linked to both hot moment condom (marginally) and cold moment PrEP use, and methamphetamine use during sex was marginally linked to hot moment condom, but not cold moment PrEP use. In the final analysis of unprotected anal sex (condomless and PrEP-less), both drunk/buzzed alcohol and methamphetamine (marginally) use were linked to risk for HIV infection, whereas the other major substances used during sex (marijuana and poppers) were not. As more MSM use daily PrEP and new methods of PrEP become available, ongoing research is needed to better understand primary HIV protection in this highly affected group. Our study is limited in self-reported behavior and a focus on use of primary HIV prevention products that are currently recommended for use (condoms and daily oral PrEP), and do not include community adaptive behaviors such as HIV serosorting12 or secondary prevention through viral suppression.13 Nonetheless these findings underscore the importance of more research in the era of daily PrEP use. This includes valid and reliable measurement of increasingly complex sexual risk and hot and cold moment protective behaviors, and sexual event-level substance use linkages with use of an increasing number of effective biomedical methods to prevent HIV infection among MSM.