A Tale of Two Cities: Exploring the Role of Race/Ethnicity and Geographic Setting on PrEP Use Among Adolescent Cisgender MSM

Abstract

Although pre-exposure prophylaxis (PrEP) could substantially reduce the risk of HIV acquisition among adolescent cisgender men who have sex with men (cisMSM), various barriers faced by people of color, particularly within the southern region of the U.S., may lead to racial disparities in the utilization of PrEP. Few studies, however, have explored racial/ethnic differences in PrEP use by geographic setting among adolescent cisMSM. We conducted a cross-sectional analysis examining racial disparities in PrEP use among cisMSM ages 15–24 years in New Orleans, Louisiana, and Los Angeles, California recruited between May, 2017 and September, 2019. The odds of PrEP use among AA adolescents were considerably lower than White adolescents in New Orleans (OR (95% CI): 0.24 (0.10, 0.53)), although we did not find evidence of differences in Los Angeles. Our findings underscore the need for targeted interventions to promote PrEP use among adolescent MSM, particularly among AA adolescent cisMSM living in the southern region of U.S.

Introduction

Adolescents (aged 13–24) are among the highest risk groups for HIV infection, accounting for 21% of all new infections nationwide in 2018 [1]. Among those of greatest risk for HIV infection are African American (AA) and Latinx adolescent cisgender men who have sex with men (cisMSM); in 2017, approximately 83% new infections among youth in the U.S. were among gay and bisexual men, the majority of whom were AA (51%) or Latinx (25%) [2]. These disparities are likely due to a variety of structural factors, including racial and ethnic differences in socioeconomic status and access to and use of HIV-related treatment and prevention services, including pre-exposure prophylaxis (PrEP) for the prevention of HIV transmission [3-6].

Relative to other age groups, the gap between need for pre-exposure prophylaxis (PrEP) and its use is widest among adolescents [2]. Adolescents of color stand to benefit most from the use of PrEP, given their higher rates of HIV acquisition relative to their White counterparts [7], though they may face a host of barriers to PrEP use related to the delivery of the medicine in real-world community settings. These include barriers related to affordability, access, the need for regular clinic visits, stigma, daily adherence to PrEP, side effect concerns, provider training, and discontinuation of PrEP [8-13]. In addition, socioeconomic and sociocultural barriers faced by people of color and sexual and gender minority people may contribute to racial disparities in the utilization of PrEP among young cisMSM [14]. These socioeconomic and sociocultural barriers can be exacerbated in certain regions of the United States, including rural and Southern areas, where there are greater concentrations of people of color accompanied by lower economic status and mobility [15], lower access to PrEP providing clinics [16], and greater levels of racial discrimination and HIV-related stigma [17]. Indeed, there are marked racial and geographical inequities in reported PrEP use among adult cisMSM, with cisMSM who are AA and Latino or who reside in rural or Southern areas reporting much lower PrEP use than White cisMSM and those living in urban areas and outside the South [18, 19].

Although adolescent AA and Latinx cisMSM are among the highest risk groups for HIV acquisition [2], studies exploring regional and racial/ethnic disparities in PrEP use have largely focused on adult cisMSM populations [18-20]. Furthermore, few studies have quantitatively assessed whether potential racial and ethnic differences in PrEP use among adolescent cisMSM vary by geographic location, as would be expected given regional variation in barriers to care [15, 16], though a national sample of young (13–24) sexual minority men revealed that those who lived in the Northeast, Midwest, and West had higher odds of being a current PrEP user than those who lived in the South [21].

The objective of our analysis is to further understand of the role of race and place play on PrEP use by assessing the presence and extent of racial/ethnic disparities in these outcomes among adolescent cisMSM in two geographical settings, Los Angeles and New Orleans, both HIV epicenter cities located within the West and South regions of the U.S., respectively. Both cities maintain substantial racial disparities in wealth, though significantly larger poverty rates are observed among AA relative to other racial groups in New Orleans: nearly one third (32.4%) of AA live below the poverty level, compared to 21.2% and 13.0% of Latinx and White residents, respectively. In Los Angeles, by contrast, similarly high poverty rates are observed among AA and Latinx individuals: 24.5% and 23.7% of AA and Latinx residents, respectively, live below the poverty level, compared with only 10.6% of White residents [22]. Through subgroup and adjusted analyses, we sought to explore whether potential efforts to improve PrEP awareness and equalize socioeconomic variables and care access across racial/ethnic groups could mitigate potential disparities in PrEP use. We hypothesize that adolescent cisMSM of color will experience lower levels of PrEP use relative to their White counterparts, and that this racial/ethnic disparity in PrEP use will be greater in New Orleans relative to Los Angeles. We further hypothesize that equalizing socioeconomic variables and care access across racial/ethnic groups could mitigate these disparities in both settings.

Methods

We analyzed baseline data from a randomized controlled trial (RCT) within the NIH-funded Adolescent Medicine Trials Network (ATN) (Clinicaltrials.gov NCT03134833). The objective of the parent study was to assess the effect of several intervention conditions (automated text-messaging and monitoring [AMM] and repeat HIV/STI testing assessment procedures, online group peer support plus AMM and/or patient-centered coaching plus AMM) on uptake of HIV prevention continuum steps [linkage to health care, use of PrEP or postexposure prophylaxis (PEP), condoms, and prevention services] over 24 months among adolescents at high risk of HIV acquisition. Between May, 2017 and September, 2019, 1487 HIV uninfected adolescents aged between 14 and 24 years were recruited from community-based organizations and clinics serving gay, bisexual, and transgender youth, homeless youth, and post-incarcerated youth in Los Angeles County, California, and New Orleans, Louisiana. We also recruited on social media and through referrals.

Individuals we eligible if they were determined to be at high risk of HIV acquisition according to the following criteria: adolescents must have tested seronegative on a rapid point-of-care fourth-generation Alere test (Alere, Waltham, MA, USA) at screening and reported at least 3 of the following criteria: self-reporting as gay, bisexual, or transgender; African American or Latino race/ethnicity; having unprotected anal sex, sharing needles for injecting drugs, or an HIV-positive partner in the last 12 months; having been homeless (defined as not having a regular place to sleep for 3 or more months); illicit substance use (not including marijuana) in the last 12 months; having been hospitalized for a mental health disorder; having been in jail or on probation; having an STI in the last 12 months. These criteria were based on the 2017 guidelines for PrEP eligibility released by the Centers for Disease Control and Prevention (CDC) [23]. Written consent was provided by participants before enrollment, and the research was approved by the institutional review boards (IRB) at Tulane University and the University of California, Los Angeles (IRB#16-001372).

Measurement

At baseline, interviewers administered a questionnaire assessing various demographic and behavioral characteristics of participants, including: age, racial and ethnic identification; gender identity; employment status (employed, unemployed, student), housing status (stable/non-transient housing, unstable/transient housing); sexual behaviors (e.g., condom use behaviors, recent intercourse with individuals of unknown or known HIV status); and PrEP and post-exposure prophylaxis (PEP) use. PrEP awareness was assessed based on a participant’s response to the following question: ‘Before today, have you ever heard about PrEP?’. Current PrEP use was assessed using the following question: ‘Are you currently taking any PrEP medication, also known as Truvada, to reduce your risk of HIV transmission?’ Respondents could select from the following options: ‘Yes’, ‘No’ or ‘Refuse to Answer’. Respondents who selected ‘Refuse to answer’ were classified as missing in our analysis.

Statistical Analysis

We analyzed baseline data for adolescents identified as cisMSM (n = 729). Participants we categorized as cisMSM if they were assigned male at birth, self-identified as a man, and reported that they engaged in sexual intercourse with men. We excluded homosexual women (n = 89), transgender men (n = 50), transgender women (n = 74), gender non-conforming individuals (n = 73) and cisgender heterosexual men and women (n = 380) because we did not have adequate power to perform our analyses separately for these groups given limitations in sample size, and because they have different risk profiles for HIV acquisition [2]. Participants of racial/ethnic groups who were not White, Latinx or AA, (i.e., Asian, Native American, ‘Other’; n = 136) were also excluded as we did not have adequate power to perform comparisons for these groups.

We compared demographic and behavioral characteristics by race/ethnic group (White, Latinx and AA) using chi-square and Wilcoxon rank sum tests. We then ran logistic regression models to examine the association between race/ethnic group and current PrEP use. We ran separate unadjusted and adjusted regression models for each outcome, in the total cisMSM sample and stratified by geographical location (Los Angeles and New Orleans). We also ran a subgroup analysis assessing current PrEP use among cisMSM participants who reported being aware of PrEP.

We note that the adjusted models within our analysis were not performed to control for potential confounding of the effect of race/ethnicity; race and ethnic identification, being a (generally) inalterable characteristic, would not be affected by any potential variables which are associated with PrEP awareness and use [24]. Variables related to healthcare access and socioeconomic status would likely then be on the causal pathway between race and PrEP use, and would then be considered mediators, not confounders, of the relationship between race and these outcomes. Including these variables in the model would induce a scenario in which these characteristics were equalized across all racial/ethnic groups, say, through an intervention to reduce racial disparities in healthcare access and socioeconomic status [24]. We included variables in the adjusted model that were hypothesized a priori as variables on the causal pathway between race/ethnicity and PrEP use, and which could be altered under hypothetical interventions to reduce racial disparities in these outcomes [14]. These variables included housing status (unstable/transient, stable/permanent), employment status (employed or student, unemployed), and access to a regular healthcare provider (yes/no). Additionally, we included age (in years; at baseline) in the adjusted model.

Results

Population Characteristics

Of our sample of 729 adolescents who identified as cisMSM, approximately 23% (n = 166) were White, 46% (n = 334) were AA and 31% (n = 229) were Latinx. Over half (n = 406, 56%) of the participants in our sample lived in Los Angeles, while 44% (n = 323) lived in New Orleans (see Table 1).

Table 1.

Demographic and behavioral characteristics by racial/ethnic group among adolescent cisgender men who have sex with men (cisMSM) at high risk of HIV acquisition in New Orleans and Los Angeles

	Total (n = 729)		White (n = 166)		African American (n = 334)		Latinx (n = 229)		p-valuea
Age: median (IQR)	21 (19–23)		21 (20–23)		21 (19–23)		21 (19–23)		0.39
	N	%	N	%	N	%	N	%	p-valuea
Study location									<0.001
Los Angeles	406	55.7	89	53.6	113	33.8	204	89.1
New Orleans	323	44.3	77	46.4	221	66.2	25	10.9
Current housing									0.04
Stable/permanent	575	78.9	131	78.9	258	77.2	186	81.2
Unstable/transient	137	18.8	33	19.9	64	19.2	40	17.5
Missing	17	2.3	2	1.2	12	3.6	3	1.3
Employment status									0.77
Employed/student	575	77.8	139	76.8	282	76.8	199	79.9
Unemployed	160	20.1	40	22.1	73	19.9	47	18.9
Missing	17	2.1	2	1.1	12	3.3	3	1.2
Has regular medical provider									0.03
Yes	487	66.8	116	69.9	207	62	164	71.6
No	238	32.6	52	28.9	126	37.7	64	27.9
Missing	4	0.5	2	1.2	1	0.3	1	0.5

^ap-values based on Wilcoxon rank sum and chi-square tests for continuous and categorical variables, respectively

We did not find evidence of differences across race/ethnicity for age or employment status. There were racial/ethnic differences in the proportion of participants living in stable or permanent housing: approximately 81% of Latinx participants reporting that they lived in stable or permanent housing, compared to 79% and 77% of White and AA participants, respectively (p = 0.04). A lower proportion of AA adolescents reported that they had access to a medical provider (62%) compared to 72% of Latinx and 70% of White adolescents (p = 0.03). Additionally, there were substantial differences in the racial/ethnic composition of participants by geographic location: A greater proportion of AA participants were enrolled in New Orleans (66%), and a greater proportion of Latinx participants were enrolled in Los Angeles (89%; p < 0.001), reflecting the demographic differences in the two cities (see Table 1).

PrEP Use

The majority of participants in the total sample (n = 646, 89%) were not currently using PrEP. In both New Orleans and Los Angeles, the odds of PrEP use tended to be higher among those who had stable housing, were employed or a student, and had access to a medical provider, though the confidence intervals for the effect estimates were large (see Table 2). In New Orleans, approximately 9% (n = 29) of the sample were using PrEP, and PrEP use was highest among White adolescents (20%), followed by Latinx and AA adolescents (8% and 6%, respectively). In unadjusted analysis, the odds of PrEP use among AA adolescents was less than one quarter that of White adolescents (OR (95% CI): 0.24 [0.10, 0.53]). Results were virtually identical after adjusting for age, housing status, employment status, and access to a medical provider (OR_adj (95% CI): 0.24 (0.10, 0.59)). As for Latinx adolescents, we did not find evidence of a substantial difference in PrEP use between White and Latinx adolescents in unadjusted analysis (OR (95% CI): 0.37 (0.08, 1.75)) or adjusted analysis (OR_adj (95% CI): 0.32 (0.06. 1.75)), though the confidence intervals were large (see Table 2).

Table 2.

Associations between racial/ethnicity, structural covariates and pre-exposure prophylaxis (PrEP) use by city (New Orleans and Los Angeles) among adolescent cisgender men who have sex with men (cisMSM) at risk of HIV acquisition

	New Orleans (N = 343)a						Los Angeles (N = 405)b
	Uses PrEP (n = 29)		Does not use PrEP (n = 290)		OR (95% CI)	ORadjc (95% CI)	Uses PrEP (n = 54)		Does not use PrEP (n = 351)		OR (95% CI)	ORadjc (95% CI)
	N	%	N	%			N	%	N	%
Race
African American	12	5.5	207	94.5	0.24 (0.10, 0.53)	0.24 (0.10, 0.59)	13	11.5	100	88.5	0.58 (0.26, 1.30)	0.67 (0.29, 1.55)
Latinx	2	8.3	22	91.7	0.37 (0.08, 1.75)	0.32 (0.06, 1.68)	25	12.2	179	87.8	0.63 (0.32, 1.25)	0.56 (0.27, 1.15)
White	15	19.7	61	80.3	1 (ref)	1 (ref)	16	18.2	72	81.8	1 (ref)	1 (ref)
Current housing
Stable	26	10.7	218	89.3	2.10 (0.61, 7.22)	2.06 (0.54, 7.76)	43	16.1	224	83.9	2.22 (1.10, 4.45)	1.91 (0.87, 4.14)
Unstable	3	5.4	53	94.6	1 (ref)	1 (ref)	11	8	127	92	1 (ref)	1 (ref)
Missing	0		19				0		0
Employment status
Employed/student	28	10.6	235	89.4	-d	-d	44	14.2	265	85.8	1.59 (0.74, 3.38)	1.00 (0.44, 2.32)
Unemployed	0	0	42	100	1 (ref)	1 (ref)	9	9.5	86	90.5	1 (ref)	1 (ref)
Missing	0		13				1		0
Has regular medical provider
Yes	27	12.7	186	87.3	14.95 (2.00, 111.63)	15.09 (1.94, 117.70)	50	18.3	223	81.7	7.12 (2.51, 20.17)	6.90 (2.39, 19.90)
No	1	1	103	99	1 (ref)	1 (ref)	4	3.1	127	96.9	1 (ref)	1 (ref)
Missing	1		1				0		1

^an = 4 missing

^bn = 1 missing

^cAdjusted for age, housing status, employment status, and access to a medical provider

^dOdds ratios excluded as employment status perfectly predicted current PrEP use in this subgroup

In Los Angeles, approximately 13% (n = 54) of the sample were currently using PrEP. As with New Orleans, PrEP use was highest among White adolescents (18%), followed by Latinx and AA adolescents (12% and 11%, respectively). We did not find evidence of a difference in odds of PrEP use between White and AA adolescents, however, in either unadjusted analysis (OR (95% CI): 0.58 (0.26, 1.30)) or adjusted analysis (OR_adj (95% CI): 0.67 (0.29, 1.55)). Similarly, we did not find evidence of a substantial difference in odds of PrEP use between White and Latinx adolescents in unadjusted analysis (OR (95% CI): 0.63 (0.32, 1.25)) or adjusted analysis (OR_adj (95% CI): 0.56 (0.27, 1.15); see Table 2).

Sub-group Analysis: PrEP Use Among Adolescents Aware of PrEP

Most (n = 600, 82%) reported that they had heard of PrEP in their lifetime. In New Orleans, approximately 79% (n = 255) of adolescents had heard of PrEP, and awareness was highest among White adolescents (87%), followed by AA (77%) and Latinx (75%) adolescents. PrEP awareness was somewhat higher in In Los Angeles, where approximately 85% (n = 345) of adolescents had heard of PrEP. As in New Orleans, PrEP awareness in Los Angeles was highest among White adolescents (90%), followed by Latinx (87%) and AA (77%) adolescents.

Findings for PrEP use among the sub-group of adolescents who were aware of PrEP were comparable to those of the full sample. In New Orleans, the odds of PrEP use among AA adolescents who were aware of PrEP were 0.26 that of their White counterparts (OR (95 CI%): 0.26 (0.11, 0.59)), and similar results were observed after adjusting for age, housing status, employment status, and access to a medical provider (OR_adj (95% CI): 0.23 (0.09, 0.58)). We similarly did not find evidence of a difference in odds of PrEP use between White and Latinx adolescents who were aware of PrEP in either unadjusted analysis (OR (95% CI): 0.42 (0.09, 2.06)) or adjusted analysis (OR_adj (95% CI): 0.30 (0.06, 1.58)).

In Los Angeles, we did not observe evidence of differences in odds of PrEP use between AA and White adolescents who were aware of PrEP in either unadjusted analysis (OR (95% CI): 0.69 (0.31, 1.55)) or adjusted analysis (OR_adj (95% CI): 0.69 (0.31, 1.55), or between Latinx and White adolescents who were aware of PrEP in either unadjusted analysis (OR (95% CI): 0.23 (0.32, 1.29)) or adjusted analysis (OR_adj (95% CI): 0.58 (0.28. 1.21)), mirroring our findings from the full sample.

Discussion

In this sample of adolescent cisMSM, we observed that, while most adolescents were aware of PrEP, only about one in ten were currently using it. There were considerable racial/ethnic disparities as well, with AA adolescents exhibiting the lowest levels of PrEP use, followed by Latinx and then White adolescents. In New Orleans, the disparities in PrEP use between AA and White adolescents were particularly stark: the odds of PrEP use among AA adolescents was approximately one quarter that of White adolescents. In Los Angeles, by contrast, we did not observe evidence of substantial differences in the odds PrEP use across racial and ethnic groups.

PrEP awareness and use in this study are similar to those observed in a prior study conducted from 2013 to 2015 among slightly older young MSM (median age 24 vs. 21 in our sample) across the U.S., in which approximately one third were not aware of PrEP, and fewer than one in ten used it [25], compared to 18% and 11%, respectively, in this study. In another nationwide sample of young adolescent (ages 15–17) cisMSM, not knowing how to access PrEP was significantly associated with identifying as a racial/ethnic minority, though PrEP awareness did not differ by geographic location [21]. Among adult cisMSM, racial disparities in PrEP use parallel those observed in our own sample: in a study of 10,104 MSM across 23 U.S. urban areas in 2017, White cisMSM were significantly more likely than were Latinx and AA cisMSM to report PrEP awareness and use [20]. Furthermore, as with our own sample, there are considerable regional disparities in the use of PrEP, particularly when comparing populations in the South verses other U.S. regions: in 2016, only 30% of all PrEP users lived in the Southern U.S., though over half of new HIV infections occurred among individuals in this region [26].

Findings from our adjusted analysis suggest that interventions to equalize access to a medical provider and housing and employment status across racial/ethnic groups may be insufficient to address disparities between in PrEP use between AA and White adolescents in New Orleans, which is persistent at nearly identical levels after adjustment for these variables. Further, disparities in PrEP use between White and AA adolescents persisted when limiting the sample to those who were aware of PrEP, suggesting that these discrepancies in PrEP use are not solely the result of lower awareness among AA adolescents. Awareness of PrEP, in this case, may not necessarily be accompanied by an appreciation of its efficacy, minimal side effects, and modes of free access.

Sociocultural and structural barriers to PrEP use, and medical care in general, may also contribute racial and geographic disparities observed in our study. MSM of color can experience intersectional stigma through their embodiment of multiple stigmatized identities (e.g. race, sexual orientation, low economic status) [27]. Because of real or anticipated experiences with stigma in the clinical setting, people of color, and especially those who are sexual and gender minorities, could exhibit greater levels of medical mistrust and may find it difficult to discuss risk behaviors and HIV prevention with their provider [28, 29]. Furthermore, within the AA community, heightened heterosexism, masculinity, and gendered social norms may make it more difficult for AA sexual and gender minorities to be open about their sexual orientation or identity, which may inhibit their ability to seek PrEP [30]. Latinx MSM, too, also face internalized and within-community stigma associated with their sexual identity [31, 32], as well as language barriers and documentation status as further barriers to care.

Providers may also exhibit social biases that prevent them from prescribing PrEP to patients of color: a study among U.S. medical students revealed that, when presented with hypothetical vignettes of MSM patients seeking PrEP, students rated AA patients as more likely than White patients to engage in increased unprotected sex if prescribed PrEP [34]. Beyond the direct effects of stigma and bias, AA and Latinx adolescents may also face more structural and financial barriers to assessing care (e.g., affordability concerns over costs of care and clinic attendance), given the racial disparities in socioeconomic status throughout the U.S. [35].

The barriers to PrEP use may be amplified in the southern region of the U.S., which experiences greater levels of discrimination and HIV-related stigma relative to other regions in the U.S. [17]. Notably, in our study, the odds of PrEP use in Los Angeles was comparable across all racial/ethnic groups, despite disparities in PrEP awareness. These findings highlight the strong role of context or place in access to care and could suggest that adolescent cisMSM of color in Los Angeles may encounter fewer barriers to PrEP use than their counterparts in New Orleans.

In addition to the marked disparities in PrEP awareness and use, our findings illustrate the need to encourage the initiation of PrEP among all adolescent cisMSM at risk of HIV acquisition, regardless of race/ethnicity. Even among White adolescents in our sample, who exhibited the highest levels of PrEP use relative to other groups, fewer than one in five were currently using PrEP. Potential barriers to the use of PrEP include: fears of side effects [36, 37], concerns about PrEP effectiveness [36], stigma [14, 38, 39], lack of a consistent healthcare provider [40], low perceived risk of HIV acquisition [40], difficulty accessing PrEP and attending clinic appointments [39, 41], conspiracy beliefs concerning the use of PrEP [40, 42], and concerns about the affordability of PrEP [38, 39]. Beyond these barriers, adolescents may experience additional challenges to PrEP initiation, including concerns over confidentiality: many adolescents receive health coverage through their parents’ insurance, and providers have cited fears that discussing and prescribing PrEP could lead to an unwanted disclosure of sexual orientation or risk behavior to one’s parents [43]. Furthermore, adolescents, in general, are more likely to engage in impulsive decisions that favor immediate rewards over long term benefits [44], which could hinder their adoption of preventative care services such as PrEP.

Several interventions could help mitigate these barriers to improve access to and utilization of PrEP among eligible adolescents. At the patient level, interventions could address PrEP knowledge and myths, managing disclosure and HIV stigma, medical mistrust, and navigating access and payment issues. Currently implemented PrEP navigation interventions address these factors on a one-on-one level but broader education initiatives may be warranted, or more intensive and wholistic interventions for youth targeting multiple risk and developmental factors, and hierarchies of needs. Telehealth-based approaches, too, could mitigate barriers to PrEP use related to difficulties in accessing care [45]. Past studies demonstrated that a telehealth-based PrEP initiation and monitoring program can reach young cisMSM quickly and conveniently start users on PrEP, and transition participants to a sustainable PrEP provider [45]. We also note the importance of policy-level interventions which could mitigate financial barriers to accessing care among adolescent patients. For instance, California’s Office of AIDS has recently developed the PrEP Assistance Program (PREP-AP) to help low income (less than 5 times the Federal Poverty Level) Californians pay for PrEP-related medical services, such as doctor visits or laboratory tests [46]. Such initiatives could substantially expand PrEP access to individuals who are financially constrained and do not have adequate insurance coverage to afford these necessary services.

For cisMSM of color, targeted interventions to promote access and support adherence to PrEP are lacking. Provider-level approaches could potentially fill this gap: enhancing medical training to better understand and address the needs of racial minorities and sexual and gender minorities could mitigate social biases (e.g., racism, heterosexism) among providers which may lead of medical mistrust and feelings of discrimination and stigma among patients of color [47, 48]. Crucially, providers should consider adopting patient-centered approaches to care, which consider the broad spectrum of factors that affect young cisMSM of color’s decision making regarding the risks and benefits of PrEP treatment options, and prioritizes understanding patients’ needs for treatment rather than simply providing instructions for care [49]. This collaborative approach between the provider and the patient can produce an individualized treatment plan that is feasible and acceptable. Further, a provider with a nuanced understanding of PrEP barriers and obstacles for adolescents of color will be prepared to have nonjudgmental sex-positive communication and will create a safe space which promotes open communication about wellness as a whole.

We acknowledge several limitations of our study. First, upon stratifying the data by racial/ethnic group and city, the sample size for certain sub-groups was limited, particularly for Latinx adolescents in New Orleans. This limitation may jeopardize the statistical power necessary to detect differences in the outcomes between Latinx and White adolescents. Additionally, we note that that the levels of PrEP use in our sample may be influenced the timing of the FDA’s approval of Truvada for PrEP for individuals under the age of 18: approval was granted in May, 2018 [22], whereas enrollment for our study began in May, 2017. Thus, some younger individuals in our sample may not have used PrEP because their providers deemed them ineligible due to their age. Nonetheless, minor consent laws in the U.S. in 2017 did not expressly prohibit minor access to PrEP or other HIV prevention methods [50]. Furthermore, the vast majority (95%; n = 689) of our participants were above the age of 18 at enrollment. It is probable, then, that all or nearly all of our population sample could potentially access PrEP at the time of enrollment.

We note that variable included in our adjusted analysis were limited to certain structural factors influencing PrEP access, including housing status, employment status, and access to a medical provider. Other potential factors affecting PrEP access, including experienced stigma, heterosexism, and racism, were not measured at enrollment in this study. Future research assessing the potential effect of interventions addressing these inter- and intra-personal barriers should consider controlling for these factors when exploring racial/ethnic disparities in PrEP use. We also acknowledge that it is critically important to examine PrEP awareness and use among other sexual and gender minorities (i.e., gender non-conforming individuals, transgender men and women, homosexual women), racial/ethnic minorities (i.e., Native American, Asian American and “other”) and minority women, specifically, who may also be at heightened risk of HIV acquisition relative to the general population. Given limitations in our sample size for these groups, we did not have adequate power to perform our comparisons among these groups. Future research should examine PrEP eligibility, awareness and use among these populations.

Despite these limitations, our study is unique in that we examined PrEP use among high risk adolescents recruited from community-based organizations and explored the differences in these outcomes by both race and geographic location. The sizeable population of AA and Latinx cisMSM in our sample demonstrates our success in engaging vulnerable adolescents of color through community-based organizations and social media.

Conclusions

Our findings highlight concerning racial disparities that could jeopardize HIV prevention among individuals most vulnerable to infection. CisMSM of color continue to experience disproportionately high rates of HIV acquisition, especially within the southern region of the U.S., yet PrEP use remains comparatively low in this population [26]. Among adolescents, this discrepancy could translate to pervasive and expanding racial disparities in HIV incidence for generations to come. These findings underscore the need for further research to understand and address the underlying causes of regional and racial disparities in PrEP use, as well as targeted strategies to improve engagement and care access among young cisMSM of color.