Factors associated with PrEP awareness and use among men who have sex with men who use drugs in the Southern United States

Abstract

Pre-exposure prophylaxis (PrEP) dramatically reduces HIV transmission risk. PrEP is underutilized among men who have sex with men who use substances (SU-MSM) in the Southern U.S., for whom there is limited research and high PrEP need. Using cross-sectional data from the National Institute on Drug Abuse (NIDA) Clinical Trials Network (CTN) 0082 study, we explored factors associated with PrEP awareness and use among 225 SU-MSM in the Southern U.S. Participants were recruited from community-based sexually transmitted infection clinics, syringe services programs and outpatient substance use treatment programs in eight cities across five Southern states with high HIV incidence. Multinomial logistic regressions examined PrEP awareness and use relative to sociodemographic factors, sexual behaviors and substance use. Results demonstrated overall high awareness, yet limited uptake of PrEP. Younger age, higher education, condomless anal sex and more frequent popper use were associated with greater odds of PrEP awareness. Higher education, condomless anal sex and more frequent popper use were associated with greater odds of PrEP use. Results highlight the need for innovative PrEP outreach to Southern SU-MSM that accounts for age, education and substances used.

Introduction

Southern U.S. states comprised 52% of new HIV diagnoses in 2021, with the Southern states of Florida, Louisiana, Georgia, Mississippi and Tennessee among the top seven states (ranging from 20.3/100,000 to 29.3/100,000 of the population) (Centers for Disease Control and Prevention, 2023b). The 2021 National HIV/AIDS Strategy has emphasized the importance of focusing on Southern U.S. states to address high HIV incidence in the region (The White House, 2021). Within Southern states (and throughout the U.S.), men who have sex with men (MSM) continue to face disproportionate HIV risk, with 67% of the HIV incidence in 2021 attributed to male-to-male sexual contact (Centers for Disease Control and Prevention, 2023b). Significant racial and ethnic disparities are evident in elevated HIV incidence among Black and Latino MSM, who respectively accounted for 37% and 33% of new HIV infections in 2021 (Centers for Disease Control and Prevention, 2023b). Such disparities can be understood in terms of intersectional stressors like poverty, structural racism, sexual minority discrimination, disenfranchisement and medical mistrust that limit access to and engagement in healthcare (Burns et al., 2021; Cahill et al., 2017; Millett, 2020; Millett et al., 2012).

Pre-exposure prophylaxis (PrEP) dramatically reduces HIV transmission risk in the context of sexual behavior (e.g., condomless anal sex; multiple sexual partners), injection drug use and sex in the context of substance use (i.e., chemsex), particularly among MSM (Grant et al., 2010; Smith et al., 2015; World Health Organization, 2014). Under the U.S. Affordable Care Act, private insurance and Medicaid plans are largely required to cover the cost of PrEP as well as associated medical visits and lab tests (Centers for Disease Control and Prevention, 2024). However, insurance coverage gaps and limited Medicaid expansion for lower-income people reduce access to PrEP in the Southern U.S. (Sullivan et al., 2019). PrEP assistance programs, such as those within pharmaceutical companies, provide free or lower-cost PrEP for individuals without insurance. State-based co-pay assistance programs also assist with PrEP costs (Centers for Disease Control and Prevention, 2024); yet, as of 2024, no Southern states offer such programs (HIV.gov, 2024).

Recent aggregate data from AIDSvu.org (2024b) and a systematic review (Kamitani et al., 2020) demonstrate that PrEP uptake in the South lags far behind need. Despite accounting for over half of HIV incidence, the Southern U.S. comprised only 38% of all people using PrEP (AIDSvu, 2024b). Racial disparities in PrEP uptake also persist in the South, with Black Southerners comprising 52% of HIV incidence but only 21% of those taking PrEP (AIDSvu, 2024b). Taken together, these data demonstrate a need for research exploring factors associated with PrEP awareness and uptake in the Southern U.S.

Not only is HIV incidence particularly high in the South, but regional data from the 2017 annual American Men’s Internet Survey (AMIS) also show considerable substance use among Southern MSM. According to the 2017 AMIS, 27.6% of Southern MSM reported using cannabis, 5.9% methamphetamines and 20.8% other illicit drugs (Zlotorzynska et al., 2020). Chemsex, particularly sex while under the influence of stimulants like methamphetamine, amplifies HIV transmission risk for some MSM (Volkow & Montaner, 2011). Despite elevated HIV risk factors among MSM in the South, the extant literature exploring PrEP use among Southern MSM has not focused on MSM who use substances (SU-MSM) (Jones et al., 2021), and available research on SU-MSM and PrEP use has not centered on SU-MSM in the Southern U.S. (Golub et al., 2019; Grant et al., 2014; Krakower et al., 2012; Morgan et al., 2018; Shover et al., 2018; van Dijk et al., 2021).

Given these gaps in the research, there is a dearth of information on how PrEP awareness, as well as behavioral and demographic factors, influence PrEP uptake among SU-MSM in the Southern U.S. Focused HIV prevention strategies like PrEP are critical to addressing enduring health disparities, particularly among MSM of color in the South, thus highlighting a critical need for research on factors associated with PrEP awareness and use among this population (Centers for Disease Control and Prevention, 2023a).

Materials and methods

Study design

The cross-sectional data for this implementation survey study were obtained from the National Institute on Drug Abuse (NIDA) Clinical Trials Network (CTN) 0082 study. Data were collected from both SU-MSM and people (of any gender) who use opioids (PWUO) between September 2020 and August 2021. Sites were located in five Southern U.S. states (Florida, Georgia, Louisiana, Mississippi and Tennessee) within eight cities with the highest HIV incidence at the time of study approval (Centers for Disease Control and Prevention, 2017a, 2017b). The University of Washington Human Subjects Division determined that this study met criteria for exemption from federal human subjects regulations. Given that MSM and PWUO live under different HIV risk conditions (e.g., condomless anal sex; injection drug use) (AIDSvu, 2024a) and differing PrEP campaign messages (e.g., campaigns often target MSM but overlook cisgender women and people who inject drugs) (e.g., Biello et al., 2021; Devlin et al., 2023), we chose to limit the current study to the SU-MSM sample (n = 225); results for PWUO participants are described elsewhere (Hatch et al., 2024). The PWUO and SU-MSM groups were mutually exclusive for the purpose of the current analyses; thus, there were no participants who were in both groups.

Recruitment and procedures

Potential sites were community-based sexually transmitted infection (STI) clinics, syringe services programs (SSPs) or outpatient substance use treatment programs (SUTPs), given that these organizations often serve large SU-MSM patient populations. Sites were identified via a systematic referral process from two CTN “Nodes” (i.e., addiction research centers) closest to our target cities; the regional Substance Abuse and Mental Health Services Administration (SAMHSA)-supported Addiction Technology Transfer Center (ATTC) and Health Resources and Services Administration (HRSA)-supported AIDS Education and Training Centers (AETC); and local public health professionals.

We contacted site directors, asked them to complete an eligibility survey, and interviewed them via Zoom. To be eligible, sites had to serve a substantial number of SU-MSM or PWUO. Because the overall objective of this survey study was to obtain data essential for the implementation of PrEP and opioid use-related services, additional site eligibility criteria included having not yet established on-site PrEP screening and prescribing programs for all clients and having not yet established on-site prescribing programs for medications for opioid use disorder (MOUD) for all clients who use opioids. Sixteen sites were selected based on these criteria; however, three sites dropped out due to COVID-related staffing challenges, resulting in 13 final sites (n = 6 STI clinics, n = 5 SSPs, n = 2 SUTPs).

To be eligible for participation, participants of any gender had to be at least 18 years old, report an HIV-negative or unknown HIV status, be able to speak and write in English, and not be incarcerated. Further, to be eligible for inclusion in what became the “SU-MSM” sample, participants also had to be assigned male at birth and identify as a cisgender man or a transgender woman, or be assigned female at birth and identify as a transgender man; have had sex in the past 12 months with at least one cisgender man or transgender woman; and report at least one binge drinking episode and/or any substance use (i.e., stimulants, opioids, club drugs [e.g., ecstasy, GHB], poppers, bath salts and/or steroids) in the past 12 months.

Figure 1 displays the study flow for the sample, starting with overall recruitment and then specifically for the current SU-MSM sample. A total of N = 919 participants were recruited for eligibility screening, and n = 580 met inclusion criteria for either the SU-MSM (n = 256) or PWUO (n = 324) group. Of the n = 256 SU-MSM, n = 248 completed the survey, and n = 225 identified as cisgender men (the final sample for the current study). For the purposes of the current analysis, we only included people who identified as cisgender men given the small number of participants who identified as transgender (n = 4 transgender men and n = 19 transgender women). All eligible participants provided verbal agreement prior to completing the eligibility screener and electronic agreement prior to completing the online survey. Remuneration included $5 for completing the brief eligibility screener and $45 for the online survey.

Figure 1.

Strobe flow diagram.

Measures

PrEP awareness and use

Before being asked questions about PrEP, participants were given a definition of the medication. Using measures adapted from Mimiaga et al. (2009), participants were then asked whether they had heard of PrEP (0 = No, 1 = Yes). If they had heard of it, participants were asked if they had ever taken PrEP (0 = No, 1 = Yes-in the past only, 2 = Yes-currently only, 3 = Yes-currently and in the past). For the current analyses, participants were collapsed into three PrEP use categories: 0 = Never heard of PrEP and, thus, never taken it (described henceforth as “Never heard of PrEP”), 1 = Heard of PrEP but never taken it, 2 = Heard of PrEP and taken it (past, current or both).

Sociodemographic characteristics

Sociodemographic characteristics were assessed using measures adapted from the PhenX demographics toolkit (Hamilton et al., 2011), and some categories were collapsed for analysis as follows: age (in years); race (1 = Black/African American, 2 = White, 3 = Multiracial/Another Race); Hispanic/Latinx ethnicity (0 = No, 1 = Yes); education (1 = Less than High School/ GED [< HS]; 2 = High School/GED [HS], 3 = More than High School/GED [> HS]); relationship status (0 = Single/not partnered, 1 = Married/partnered); sexual orientation (0 = Heterosexual/Bisexual [reflecting a category combining heterosexual, bisexual or other sexual identities among MSM], 1 = Gay); and health insurance coverage (0 = No, 1 = Yes).

Sexual behavior associated with HIV risk

Sexual behavior was assessed with three items adapted from AMIS (Emory University, 2017) and National HIV Behavior Surveillance survey (Centers for Disease Control and Prevention, 2017c). First, we summed the total number of main and/or casual sexual partners in the past 12 months. Due to extreme outliers, sums greater than 20 were capped at a maximum of 20 partners (n = 24 participants reported > 20 [range: 22–1,800] past-12-month partners). Second, exchange of sex for money, drugs, shelter or something else in the past 12 months was recoded as a dichotomous variable (0 = No, 1 = Yes) by combining four questions about whether participants had received and/or given money, drugs, shelter or something else in exchange for sex with main or casual partners. Third, a composite variable was created for anal sex that reflects participants’ engagement in condomless insertive and/or receptive anal sex (0 = No anal sex, 1 = Anal sex with a condom, 2 = Anal sex without a condom) during their last sexual encounter in the past 12 months.

Substance use frequency

Questions about substance use frequency in the past 12 months were adapted from the NIDA Risk Behavior Assessment (RBA) (Weatherby et al., 1994). Our response options differed from the RBA in that we did not ask about the number of days in the past 30 days that a participant used and/or injected a particular type of drug. Instead, participants were asked to report the frequency of their hazardous drinking and substance use. This enabled us to obtain separate frequency ratings that were treated as continuous variables for hazardous drinking, poppers, cannabis, non-injection stimulants, non-injection opioids and injection drugs that could be distinctly analyzed in our multivariable equations.

Frequency of hazardous drinking (i.e., 6 or more drinks on one occasion) was assessed with a 5-point scale (0 = Never or none during the past 12 months, 1 = Less than monthly, 2 = Monthly, 3 = Weekly and 4 = Daily or almost daily). Frequency of cannabis, non-injection stimulants, non-injection opioids, injection drugs, and popper use were measured using an 8-point scale (0 = Never or none, 1 = Less than once a month, 2 = Once a month, 3 = More than once a month, 4 = Once a week, 5 = More than once a week, 6 = Once a day and 7 = More than once a day). Non-injection opioids included painkillers, heroin and fentanyl. Non-injection stimulants included cocaine, crack, methamphetamine, ecstasy and other stimulants. Injection drug use included injected cocaine, injected methamphetamine, injected heroin, injected fentanyl, speed-ball (heroin and cocaine combined) and goofball (heroin and methamphetamine combined). A single, composite frequency was created for each drug category variable by using the highest frequency of use for each of the drugs within each category. For example, if a participant reported using non-injection painkillers once a month and using non-injection heroin more than once a day, the composite frequency for “non-injection opioids” variable was coded as more than once a day.

Data analyses

Descriptive statistics were conducted to describe PrEP awareness and use among the SU-MSM sample (Table 1), their sociodemographic characteristics (Table 2), sexual behavior associated with HIV risk (Table 3) and substance use frequency overall and by PrEP awareness and use (Table 4). Multinomial logistic regressions (Table 5) were conducted to examine associations between sociodemographic characteristics, sexual behavior associated with HIV risk, substance use frequency and PrEP awareness and use. Multinomial logistic regression was selected because PrEP awareness/use included three categories (i.e., never heard of PrEP; heard of PrEP but never taken it; heard of PrEP and taken it). Resulting coefficients provide odds ratios (ORs) that are interpreted as the odds of belonging in any category compared to a reference category.

Table 1.

PrEP awareness and use among SU-MSM (N = 225).

Variable	N	%
Have never heard of PrEP; thus, have never taken it	33	14.67
Have heard of PrEP and:
Have never taken it	102	45.33
Have taken it (currently or in the past)	90	40.00

Table 2.

SU-MSM sociodemographic characteristics overall and by PrEP awareness and use.

	Overall		Have never heard of, and thus, have never taken PrEP		Have heard of PrEP, and
					Have never taken		Have taken
	n = 225		n = 33		n = 102		n = 90
	M or N	SD or %	M or N	SD or %	M or N	SD or %	M or N	SD or %
Age (M, SD)	32.79	10.28	39.12	11.93	32.57	10.29	30.71	8.68
Race (N, %)a
Black	109	50.70	14	46.67	49	50.00	46	52.87
White	75	34.88	11	36.67	36	36.73	28	32.18
Multiracial/another race	31	14.42	5	16.67	13	13.27	13	14.94
Ethnicity (N, %)a
Non-Hispanic	180	80.72	28	84.85	82	80.39	70	79.55
Hispanic	43	19.28	5	15.15	20	19.61	18	20.45
Education (N, %)
Less than high school/GED (<HS)	31	13.78	11	33.33	15	14.71	5	5.56
High school/GED (HS)	66	29.33	15	45.45	32	31.37	19	21.11
More than high school/GED (>HS)	128	56.89	7	21.21	55	53.92	66	73.33
Health insurance (N, %)a
Have no insurance	94	42.15	19	57.58	43	43.00	32	35.56
Have insurance	129	57.85	14	42.42	57	57.00	58	64.44
Relationship status (N, %)a
Single	175	78.83	28	87.50	73	73.00	74	82.22
Married/partnered	47	21.17	4	12.50	27	27.00	16	17.78
Sexual orientation (N, %)a
Heterosexual/bisexualb	89	40.45	21	63.64	40	40.40	28	31.82
Gay	131	59.55	12	36.36	59	59.60	60	68.18

^aN < 225 due to missing responses.

^bIncludes MSM who identified as bisexual, heterosexual or another identity.

Table 3.

SU-MSM sexual behavior associated with HIV risk overall and by PrEP awareness and use.

	Overall		Have never heard of, and thus, have never taken PrEP		Have heard of PrEP, and
					Have never taken		Have taken
	n = 225		n = 33		n = 102		n = 90
	M or N	SD or %	M or N	SD or %	M or N	SD or %	M or N	SD or %
Number sexual partners (0–20) (M, SD)a	7.53	6.20	7.27	6.42	6.65	5.63	8.62	6.63
Exchange of sex for money or drugs (N, %)
No sex exchange	167	74.22	16	48.48	80	78.43	71	78.89
Sex exchange	58	25.78	17	51.52	22	21.57	19	21.11
Anal sex (N, %)
No anal sex	57	25.33	16	48.48	30	29.41	11	12.22
Anal sex with condom	46	20.44	7	21.21	21	20.59	18	20.00
Condomless anal sex	122	54.22	10	30.30	51	50.00	61	67.78

^aMaximum range set at 20 to address extreme outlier responses. n = 5 participants indicated having had sex in the past 12 months but did not consider those partners to be main and/or casual.

Table 4.

SU-MSM substance use frequency overall and by PrEP awareness and use.

	Overall		Have never heard of, and thus, have never taken PrEP		Have heard of PrEP, and
					Have never taken		Have taken
	n = 225		n = 33		n = 102		n = 90
	M	SD	M	SD	M	SD	M	SD
Hazardous drinking (0–4)	1.45	1.25	1.82	1.45	1.47	1.26	1.30	1.13
Poppers (0–7)a	1.09	1.85	0.58	1.30	0.88	1.69	1.51	2.11
Cannabis (0–7)a	2.56	2.78	2.76	2.84	2.65	2.78	2.39	2.76
Non-injection stimulants (0–7)	1.91	2.65	3.73	2.91	1.80	2.55	1.37	2.38
Non-injection opioids (0–7)	1.01	2.11	2.64	2.87	1.03	2.02	0.39	1.51
Injection drugs (0–7)	1.20	2.45	3.03	3.13	1.34	2.54	0.38	1.53

^aN < 225 due to missing responses.

Table 5.

Multinomial logistic regression results: relationship between sociodemographic characteristics, sexual behavior, substance use and PrEP use among SU-MSM (N = 203).^a

	Have heard of but never taken relative to have never heard of PrEP		Have heard of and taken relative to have never heard of PrEP		Have heard of and taken relative to have heard of but never taken PrEP
Sociodemographic variables	OR (95% CI)		OR (95% CI)		OR (95% CI)
Age	0.94	(0.89–0.99) *	0.95	(0.90–1.01)	1.01	(0.97–1.05)
Race (reference: White)
Black	0.78	(0.21–2.88)	1.31	(0.31–5.52)	1.68	(0.73–3.87)
Multiracial/other	0.85	(0.17–4.28)	1.03	(0.17–6.38)	1.20	(0.37–3.89)
Ethnicity
Hispanic/Latinx (reference: non-Hispanic/Latinx)	0.50	(0.08–2.94)	0.42	(0.06–2.74)	0.84	(0.30–2.32)
Education (ref: >HS)
Less than high school/GED (<HS)	0.45	(0.09–2.13)	0.10	(0.01-0.77)*	0.23	(0.05–1.12)
High school/GED (HS)	0.26	(0.07-0.96)*	0.12	(0.03-0.48)**	0.46	(0.20–1.03)
Health insurance
Have insurance (ref: no insurance)	1.79	(0.57–5.61)	1.84	(0.53–6.40)	1.03	(0.48–2.18)
Relationship status
Married/partnered (ref: single)	2.52	(0.58–10.90)	1.48	(0.30–7.25)	0.59	(0.24–1.42)
Sexual orientation
Gay (ref:) Heterosexual/bisexualb	1.58	(0.51–4.85)	1.59	(0.47–5.40)	1.01	(0.47–2.14)
Sexual behavior associated with HIV risk
Number of sexual partners (0–20)c	1.01	(0.91–1.11)	1.07	(0.97–1.19)	1.06	(0.99–1.13)
Exchange of Sex
Sex exchange (ref: no sex exchange)	0.58	(0.16–2.14)	0.46	(0.10–2.04)	0.79	(0.28–2.24)
Anal sex behavior (ref: No anal sex)
Anal sex with condom	0.63	(0.17–2.35)	1.00	(0.21–4.70)	1.60	(0.52–4.91)
Condomless anal sex	2.66	(0.75–9.47)	7.25	(1.69-31.08)**	2.72	(1.04-7.16)*
Substance use frequency
Hazardous drinking (0–4)	1.18	(0.79–1.79)	0.89	(0.56–1.42)	0.75	(0.55–1.03)
Poppers (0–7)a	1.07	(0.75–1.53)	1.48	(1.01-2.17)*	1.38	(1.11-1.72)**
Cannabis (0–7)a	0.96	(0.78–1.18)	0.86	(0.69–1.07)	0.90	(0.79–1.03)
Non-injection stimulants (0–7)	0.98	(0.76–1.26)	0.98	(0.73–1.32)	1.00	(0.81–1.24)
Non-injection opioids (0–7)	0.84	(0.66–1.07)	0.86	(0.62–1.19)	1.03	(0.78–1.36)
Injection drugs (0–7)	0.93	(0.72–1.20)	0.75	(0.54–1.05)	0.80	(0.61–1.06)

^aN < 225 for multinomial regression due to missing responses.

^bMSM who identified as bisexual, heterosexual or another identity.

^cMaximum range set at 20 to address extreme outlier responses.

^*p < .05.

^**p < .01.

We used two reference categories: “never heard of PrEP” and “heard of PrEP but never taken it”. Thus, we are comparing: (1) heard of PrEP but never taken it relative to never heard of PrEP; (2) heard of PrEP and taken it relative to never heard of PrEP; and (3) heard of PrEP and taken it relative to heard of PrEP but never taken it. We considered including site ID and agency type (STI, SSP, SUTP) in the multinomial logistic regression. However, crosstabs between site ID, agency type and PrEP use showed several empty cells for one or more PrEP use categories. Thus, both site ID and agency type were omitted from the analyses. All analyses were conducted using SAS version 9.4 (SAS Institute, 2016).

Results

PrEP awareness and use

Only 15% of participants had never heard of PrEP prior to the survey, 45% had heard of PrEP but had never taken it, and 40% had heard of PrEP and taken it (Table 1).

Sociodemographic characteristics

Overall, the mean age of participants was 33 years (range = 18–63), and median age was 31 years (SD = 10; interquartile range = 14). Half identified as Black (51%), 35% as white and 14% as multiracial or another race. The majority (81%) identified as non-Hispanic/Latinx and reported their relationship status as “single” (79%). Education level varied, with 57% reporting > HS, 29% HS, and 14% < HS. Slightly more than half of participants had health insurance (58%) and identified as gay (60%) (Table 2).

Sexual behavior associated with HIV risk

The overall mean number of main and/or casual sexual partners in the past 12 months was 7.53 (range = 0–20; n = 5 participants had indicated having had sex in the past 12 months but did not consider those partners to be main and/or casual). Most (74%) SU-MSM had not exchanged sex for money, drugs or housing. Slightly more than half (54%) had engaged in condomless anal sex during their last sexual encounter (Table 3).

Substance use frequency

The overall mean frequency varied based on the type of substance. The mean frequency of hazardous drinking was 1.45, popper use was 1.09, cannabis use was 2.56, non-injection stimulant use was 1.91, non-injection opioid use was 1.01 and injection drug use was 1.20 (Table 4).

Multinomial logistic regression

As shown in Table 5, age, education, condomless anal sex and popper use were associated with PrEP awareness and/or use. In contrast, none of the following characteristics or behaviors were associated with PrEP awareness or use: race; ethnicity; health insurance; relationship status; sexual orientation; number of sexual partners; exchange of sex for money, drugs, or housing; hazardous drinking; cannabis use; non-injection stimulant use; non-injection opioid use; or injection drug use (see Table 5). Factors associated with the three different PrEP awareness/use categories are presented below.

Heard of but never taken PrEP relative to never heard of it: Only age and education were significantly associated with having heard of but never taken PrEP relative to having never heard of it. Older participants had lower odds of having heard of but never taken PrEP (OR = 0.94, CI = 0.89–0.99). Compared to participants with > HS, those with a HS level of education had lower odds of having heard of but never taken PrEP (OR = 0.26, CI = 0.07–0.96). In other words, older compared to younger participants and those with a HS degree compared to > HS had greater odds of having never heard of PrEP.

Heard of and taken PrEP relative to never heard of it: Education, condomless anal sex and popper use frequency were significantly associated with having heard of and taken PrEP relative to never heard of it. Compared to participants with > HS, those with < HS and those with HS levels of education had lower odds of having heard of and taken PrEP (OR = 0.10, CI = 0.01–0.77; OR = 0.12, CI = 0.03–0.48). Further, participants who had condomless anal sex compared with those who had no anal sex had greater odds of having heard of and taken PrEP (OR = 7.25, CI = 1.69–31.08). Participants who used poppers more frequently had greater odds of taking PrEP (OR = 1.48, CI = 1.01–2.17).

Among those who had heard of PrEP – Taken relative to never taken PrEP: Only condomless anal sex and popper use frequency were significantly associated with having heard of and taken PrEP, relative to having heard of but never taken PrEP (see Table 5). Participants who had condomless anal sex compared with those who had no anal sex had greater odds of taking PrEP (OR = 2.72, CI = 1.04–7.16). Participants who used poppers more frequently had greater odds of taking PrEP (OR = 1.38, CI = 1.11–1.72). In summary, participants who had greater odds of taking PrEP included those reporting condomless anal sex and more frequent popper use.

Discussion

The current study presents findings on PrEP awareness and use among SU-MSM in the Southern U.S., a population for whom there is limited research (Golub et al., 2019; Grant et al., 2014; Jones et al., 2021; Krakower et al., 2012; Morgan et al., 2018; Shover et al., 2018; van Dijk et al., 2021) despite large HIV disparities and an urgent need for PrEP (AIDSvu, 2024b). Most participants (85%) had heard of PrEP prior to study engagement, yet this represented a slightly smaller proportion than the 89% of MSM who reported prior awareness of PrEP in the National HIV Behavioral Surveillance (NHSB) system (Jones et al., 2021). The site eligibility criteria of having not yet established on-site PrEP screening and prescribing programs for all clients could have contributed to 15% of participants having never heard of PrEP. Despite high awareness of PrEP and high HIV risk factors (e.g., condomless anal sex; poly substance use; injection drug use), less than half of participants in the current study had ever taken PrEP, further demonstrating gaps in PrEP use in the Southern U.S. (AIDSvu, 2024b) and suggesting important areas for future research and implementation strategies.

Factors associated with having heard of PrEP

In this study, we found that younger (relative to older) participants had greater awareness of PrEP while controlling for other demographic factors, substance use frequency and sexual behavior associated with HIV risk. Previous research on age as a factor related to PrEP awareness has been mixed, with one study demonstrating more PrEP awareness among older (i.e., age 50 and above) compared to younger (i.e., 18–39 years old) MSM (Klein & Washington, 2020) and another study focused on Black MSM reporting greater awareness among younger compared to older participants (mean age = 31.6; range = 19–76) (Maksut et al., 2021). However, of those two studies, only the study of Black MSM both took place in the Southern U.S. and accounted for substance use – and of note, this study was limited to only one city (i.e., Atlanta, GA) (Maksut et al., 2021). Researchers have suggested a need for targeted PrEP messaging that accounts for age (Klein & Washington, 2020), but future studies should examine how such messaging may need to be additionally tailored for SU-MSM in the Southern U.S.

Our findings also demonstrated that those with > HS (relative to HS or less) had greater awareness of PrEP, while controlling for other demographics, substance use frequency and sexual behavior associated with HIV risk. Previous research aligns with the current finding that higher education is a factor in greater PrEP awareness (Bauermeister et al., 2013; Fallon et al., 2017; Maksut et al., 2021). However, a recent study with young SU-MSM did not observe an association between education and PrEP awareness (Gebru et al., 2022). Prior PrEP outreach strategies have often targeted MSM with higher levels of education (Hosek et al., 2017; Liu et al., 2016; Maksut et al., 2021), thus suggesting a need to also focus on MSM with lower levels of education (Maksut et al., 2021) to ensure that knowledge of PrEP extends to all who could benefit from it.

In the current study, we also found that participants who had had condomless anal sex in the past 12 months (relative to those who had no anal sex) had greater awareness of PrEP. This finding reiterates prior research, as condomless anal sex is a common factor related to PrEP awareness (Brooks et al., 2020; Callander et al., 2019; Maksut et al., 2021). Findings are also encouraging in that they suggest that PrEP messages are reaching the people who may need them most.

Factors associated with having never taken PrEP

We found that less education was associated with lower odds of PrEP use among those who had heard of PrEP, echoing prior research conducted with MSM and transgender people in Los Angeles, CA (Shover et al., 2018). As noted above, PrEP campaigns have typically centered on those with higher levels of education (Hosek et al., 2017; Liu et al., 2016; Maksut et al., 2021). In a systematic review of implementation approaches to increase PrEP use in the Southern U.S., Sullivan et al. (2019) highlighted the promise of tailored campaigns based on, for example, messaging feedback from target groups like MSM, as well as implementing messaging approaches that leverage technology (e.g., online videos; apps) rather than relying on printed materials. Such strategies may make PrEP information more readily available for individuals with less education and associated low health literacy for whom print materials may be less accessible (Sullivan et al., 2019).

Previous research with MSM in Boston, MA underscores the suggested implementation strategies in the above-mentioned systematic review. Participants in that study identified media, as well as participation in HIV research and community outreach and education, as three of the most frequent sources of PrEP education (Mimiaga et al., 2009). In fact, it is conceivable that participants in the current study may have gained PrEP knowledge and engaged in subsequent PrEP use due to their involvement in the research, which specifically targeted community-based organizations that did not yet have an established on-site PrEP program for all clients.

An additional PrEP uptake strategy for SU-MSM with lower levels of education in the South may be to involve peer navigators (Sullivan et al., 2019). Peers whom clients can trust, particularly in an environment where same-gender sex and drug use are often stigmatized (Scott, 2022), could potentially guide SU-MSM through the complexities of the healthcare system (e.g., insurance coverage; PrEP assistance programs) and PrEP treatment (e.g., side effects; follow-up requirements). An additional strategy could be to involve peer navigators and clients in developing PrEP promotion materials using messaging and values familiar to the intended audience. Findings from the current study further highlight that a variety of additional strategies are needed to increase PrEP uptake regardless of education level (Maksut et al., 2021), with a particular focus on SU-MSM in the Southern U.S.

Factors associated with having taken PrEP

The current study found that among those who had heard of PrEP prior to the survey, condomless anal sex was associated with higher odds of PrEP use. Findings from the current study support those within the extant literature in regions outside the Southern U.S. (Golub et al., 2019; Shover et al., 2018). Results suggest that participants who engaged in sexual behaviors associated with increased HIV transmission risk were aware of and used PrEP to mitigate such risk. Results further imply that PrEP messaging may be effectively reaching people at greater risk for HIV transmission.

In the current study, the only substance associated with higher odds of PrEP use was poppers. Given that MSM typically use poppers to enhance sexual experiences (Vaccher et al., 2020), intentionality to use poppers may be linked to intentionality to use PrEP. Prior research on PrEP uptake among MSM who use substances has been mixed. Some studies have found that substance use is not associated with PrEP uptake among MSM after adjusting for demographic and other HIV risk factors (Golub et al., 2019; Grant et al., 2014; Krakower et al., 2012; Morgan et al., 2018; van Dijk et al., 2021), while other studies have documented that drugs typically used in the context of sex – such as stimulants and g-hydroxybutyric acid (GHB), in addition to poppers – are associated with higher odds of PrEP use (Shover et al., 2018). Findings in the current study could indicate that participants who used other drugs commonly linked to HIV transmission risk (e.g., stimulants; drugs via injection) were no more or less likely to use PrEP than other participants or could suggest that additional strategies are needed to increase PrEP use among these individuals. Future research could potentially explore factors that contribute to increased PrEP use among MSM who use poppers and leverage those findings to increase PrEP uptake among others at potentially higher vulnerability to HIV transmission risk (e.g., people who use stimulants and/or inject drugs).

Finally, in the current study, neither race nor ethnicity were associated with PrEP use, contrary to recent data demonstrating racial and ethnic disparities in PrEP use in the Southern U.S. (AIDSvu, 2024b). Our findings could indicate that study participants were receiving care from supportive organizations where they encountered less stigma and experienced lower levels of medical mistrust than may be present in other settings (Cahill et al., 2017).

Limitations

Although the current study offers data on an often-overlooked population in the HIV prevention literature, it has limitations. First, this study involved participants recruited from community-based organizations in urban settings within the Southern U.S.; thus, findings may not generalize to SU-MSM in rural Southern locations. Additionally, surveys collected self-reported data, which may have been vulnerable to social desirability bias. The current study only accounted for sociodemographic, sexual risk and substance use variables in regression analyses; future papers should further explore perceived barriers and facilitators (e.g., stigma, medical mistrust, PrEP peer norms, PrEP access, agency type) to PrEP use. Further, this study involved a one-time survey, thus precluding our ability to examine longitudinal associations or detect causal predictors of PrEP use.

Finally, although our eligibility criteria included transgender individuals, we were regrettably unable to recruit a large enough sample of transgender participants to have enough power for inclusion in the current analysis. Transgender individuals experience higher rates of HIV compared to cisgender people, largely due to societal stigma that poses barriers to employment and healthcare (Hughto et al., 2015; Van Gerwen et al., 2020). Thus, future research is needed to understand factors associated with PrEP awareness and uptake among transgender people who use drugs in the Southern U.S., as these are populations about whom there is a dearth of research and who could also significantly benefit from PrEP.

Conclusion

This study focused on PrEP use among SU-MSM receiving care at community-based organizations in the Southern U.S. Results documented that despite having heard of PrEP and having significant HIV transmission risk factors, less than half of participants had used PrEP. Findings highlight the need for ongoing and innovative PrEP outreach strategies (especially those that account for age, education level and substances used) with diverse communities of SU-MSM in the Southern U.S., where HIV incidence and substance use are high.

Data availability statement

The data that support the findings of this study are available from Mary Hatch, PhD, the lead investigator of this study (via e-mail: hatch@uw.edu) upon reasonable request.