Characterizing the Communication Networks of Spanish- and English-Speaking Latinx Sexual Minority Men and their Roles in Pre-exposure Prophylaxis Uptake

Abstract

While pre-exposure prophylaxis (PrEP) has demonstrated efficacy in preventing HIV transmission, disparities in access persist in the United States, especially among Hispanic/Latinx sexual minority men (SMM). Language barriers and differences in how Latinx SMM obtain information may impact access to PrEP and HIV prevention. This study used data from the 2021 American Men’s Internet Survey (AMIS) to examine differences in communication networks and PrEP use among Latinx SMM by primary language (Spanish vs. English). We examined the associations between Latinx SMM’s individual- and meso-level communication networks and PrEP-related outcomes using modified Poisson regression with robust variances. Spanish-speaking Latinx SMM in the study were less likely to test for HIV, be aware of PrEP, and use daily PrEP, compared to English-speaking participants. Sexuality disclosure to a healthcare provider was positively associated with PrEP uptake among all participants and predicted STI testing over the past 12 months among English-speaking Latinx SMM. Findings highlight disparities in PrEP awareness and uptake among Latinx SMM, especially among those whose primary language is Spanish. Addressing these disparities through targeted interventions, including improved communication with healthcare providers, may help facilitate PrEP access and use in this population.

Introduction

Pre-exposure prophylaxis (PrEP) is a highly effective intervention for preventing HIV transmission when taken as prescribed (Cohen et al., 2011; Eshleman et al., 2017). Despite PrEP’s demonstrated efficacy, disparities persist in its accessibility and use in the United States (U.S.) (Raifman et al., 2019; Sullivan et al., 2020). Among gay, bisexual, and other sexual minority men (SMM), a group at high risk for HIV, Hispanic/Latino/x SMM experience poorer outcomes across the PrEP continuum of care (i.e., PrEP awareness, discussing PrEP with a provider, PrEP use) than White non-Hispanic SMM (Finlayson et al., 2019; Kanny et al., 2019; Raifman et al., 2019). Only 14 percent of Latinos who could benefit from PrEP were prescribed it in 2019 (Centers for Disease Control and Prevention, 2021), demonstrating major gaps in PrEP access (Blashill et al., 2020; Nunn et al., 2017).

Multiple barriers, including lack of health insurance, stigma, financial hardship, and low health literacy can limit PrEP access (Page et al., 2017). Language, in particular, is critical in one’s ability to obtain, process, and exchange information and services related to PrEP and HIV prevention (Mogobe et al., 2016; Storey et al., 2014). Individuals in the U.S. whose primary language is not English may have difficulty communicating effectively with providers and utilizing resources for HIV prevention (Jacobs et al., 2004). The shortage of Spanish-speaking staff across the spectrum of services creates barriers to scheduling appointments and having clinical encounters (Hassan et al., 2023). While interpreters may reduce challenges related to language, miscommunication is common and can exacerbate the difficulties related to service utilization (Hassan, et al., 2023; Zapata, de St. Aubin, Rodriguez-Diaz, & Malave-Rivera, 2022).

As Spanish is the primary language of a significant proportion of Latinos in the U.S. (Batalova et al., 2016), Latinx SMM may be at a disadvantage for learning about and obtaining PrEP (Mansergh et al., 2019). Notably, Spanish language as the preferred language for survey completion has been associated with lower odds of PrEP awareness among Latinx SMM compared to those who prefer to complete surveys in English (Lee, Barry, Kerani, Sanchez, & Katz, 2023; Mansergh, Herbst, Holman, & Mimiaga, 2019). These studies demonstrate the potential importance of language-related factors in shaping PrEP access and use.

According to Communication Infrastructure Theory (CIT), a community’s communication system can determine the availability of communication resources that relay relevant information and stories to impact health behaviors and outcomes (Ball-Rokeach et al., 2001; Wilkin, 2013; Wilkin et al., 2010). The communication system includes storytellers, individuals and organizations who have the capacity to tell stories and contribute to health communication that influence behaviors (Babalola et al., 2017). Due to language barriers, Latinx SMM whose primary language is Spanish may engage different storytelling networks than those of Latinx SMM whose primary language is English (Wilkin & Ball-Rokeach, 2011). Language preferences, the presence of local storytellers, and the availability of communication resources in multiple languages can shape health-related attitudes and behaviors about PrEP and other HIV prevention strategies. Thus, PrEP-related information can be framed and distributed differently across distinct storytelling networks, facilitating varying levels of connection to health and prevention resources (Wilkin et al., 2015).

Storytelling networks consist of communication agents across the individual level (e.g., friends, family members) and the meso level (e.g., members of community-based organizations) (Wilkin, 2013). Given the disproportionate burden of HIV among SMM, SMM peers may be important storytellers about HIV prevention resources for Latinx SMM. While family, friends, and healthcare providers can be individual-level members of Latinx SMM’s storytelling networks, their willingness and ability to serve in this capacity related to PrEP and HIV prevention may be dependent on whether Latinx SMM disclose their sexual identity to these individuals (García et al., 2012; Marks et al., 1992). When healthcare providers share information or resources from a clinic or healthcare system, they may also act as meso-level storytelling agents.

The meso-level storytellers that provide health information related to HIV and PrEP are often comprised of social groups and organizations (Wilkin et al., 2010). For example, Latinx SMM who are engaged with churches, clubs, or cultural and community organizations may have an active storytelling network that can influence PrEP-related knowledge and behaviors (Wilkin et al., 2015). Notably, the contexts in which storytelling networks operate (e.g., region of residence, urbanicity), referred to as the communication action context, can also shape communication to impact PrEP use behaviors (Wilkin et al., 2010).

As outlined by Ball-Rokeach et al. (2001), these contexts extend beyond mere physical locations to encompass a complex network of associations and identities. The concept of a “storytelling neighborhood” involves the active construction of identity as a member of a residential community, highlighting the dynamic nature of community narratives and interactions. The communication action context is central to the concept of a storytelling neighborhood and is shaped by various features of residential environments. These features include physical attributes, psychological influences, sociocultural dynamics, and economic factors—all of which contribute to the enabling or constraining of communication behaviors (Ball-Rokeach, Kim, & Matei, 2001).

Within this framework, geographic regions characterized by varying sociocultural, economic, and psychological characteristics operate within distinct communication action contexts that ultimately impact if and how health-related information is disseminated. Rurality and urbanicity, for example, are characteristics of the communication action context that often determine what local resources are available and who has access to them. Such characteristics can either inhibit or promote community building and health promotion efforts. For example, Nah et al. (2021) observed a stronger positive relationship between community-oriented social media and civic participation in rural communities (Nah, Kwon, Liu, & McNealy, 2021). Further, prior research has documented differences in PrEP awareness based on urbanicity of residence, suggesting that resources and PrEP information may be more widely available in urban communities given the distinct sociocultural and economic features of their communication action contexts (Sullivan et al., 2020).

While several studies suggest that the creation of Spanish-language resources can improve access to PrEP-related information among Latinx SMM, there is limited understanding of the storytelling networks of Spanish- and English-speaking Latinx SMM that are vital to relaying such information in a culturally appropriate manner (García & Harris, 2017; Raifman et al., 2019). The availability of new advances in HIV prevention, highlights the need to be able to disseminate information rapidly and effectively to communities that will benefit most to facilitate uptake. Given the growing recognition of the importance of health communication for HIV prevention, focused attention on communication systems may point to potential avenues to strengthen connections between Latinx SMM and PrEP resources.

This study examined differences in the communication networks, contexts, and PrEP-related outcomes of Latinx SMM by primary language (Spanish vs. English). Informed by CIT, we assessed the associations between individual- and meso-level storytellers and PrEP-related outcomes (e.g., PrEP awareness, PrEP use, and sexual health outcomes) among Latinx SMM whose primary language is Spanish and those whose primary language is English. We hypothesized that the influential storytellers at the individual and meso levels would differ among Latinx SMM by primary language across different communication action contexts. Study findings can inform approaches to developing health communication interventions to improve PrEP outcomes among linguistically diverse SMM.

Methods

Study Design, Setting, and Population

Data used in this study were from the American Men’s Internet Survey (AMIS), which is an ongoing annual, cross-sectional, web-based behavioral survey among U.S. SMM. AMIS began in 2013 with surveys conducted in English. We used data from the 2021 cycle collected September 2021 through March 2022, when the survey was conducted in both English and Spanish. The English survey was translated by a bilingual translator proficient in both English and Spanish and was reviewed by two experts in the field to ensure accuracy and conceptual equivalence between the English and Spanish versions of the survey. Participants were recruited through convenience sampling via websites and social media platforms and were eligible to participate if they were at least 15 years of age, identified as a cisgender male, resided in the US, and reported oral or anal sex with a man at least once in the past. For the purposes of this study, we only included participants who identified as Hispanic or Latino/x/e in the survey and self-reported HIV negative status.

Study Procedures

Questions in AMIS ask about sociodemographic characteristics, sexual behaviors, substance use, HIV testing, HIV prevention, social capital, community engagement, and other topics related to sexual health. Surveys required approximately 30 minutes to complete. Additional information about AMIS procedures can be found elsewhere (Zlotorzynska et al., 2019). Participants provided informed consent and were not compensated. AMIS study procedures were reviewed and approved by Emory University’s Institutional Review Board.

Measures

PrEP-related outcomes

HIV and STI testing.

As HIV testing and testing for sexually transmitted infections (STIs) are important for HIV prevention and are required prior to PrEP use, we assessed whether participants had been tested for HIV in the past 12 months (Yes or No) and whether participants had tested for any bacterial STIs (i.e., gonorrhea, chlamydia, and syphilis) in the past 12 months.

PrEP awareness.

Participants were asked about their awareness of daily oral PrEP. (“Have you ever heard of people who do not have HIV taking PrEP, the antiretroviral medicine taken every day for months or years to reduce the risk of getting HIV?”) (Yes/No).

PrEP use.

We assessed PrEP use with a question that asked whether participants had taken PrEP within the past 12 months. Only participants who answered “Yes” to the PrEP awareness question were asked about PrEP use. Participants who reported being unaware of PrEP were coded as “No” for PrEP use.

Communication resources

Individual-level storytellers.

As the storytellers within the social networks of Latinx SMM have the capacity to tell stories that can shape behaviors, we assessed key potential storytellers at the individual level (Wilkin, 2013). As disclosure of sexual orientation is an indication of open communication (Suen et al., 2022), potential storytellers were identified by asking participants to respond to “Yes” or “No” questions about whether participants had ever disclosed their sexual orientation to the following groups of people: to 1) gay, lesbian, or bisexual (GLB) friends; 2) friends who are not gay, lesbian, or bisexual (non-GLB); 3) family members; and 4) healthcare providers. Disclosure to each of these groups of individuals were indicators that they were potential individual-level storytellers. As social networks and linkages with other SMM are primary sources of social support and information (Amirkhanian, 2014), we also assessed participants’ number of SMM friends.

Meso-level storytellers.

When individuals participate in activities or are members of organizations, these organizations may serve as potentially important sources of information related to health and prevention and can be considered meso-level storytellers. Hence, we determined potential meso-level storytellers that relay information to Latinx SMM by asking participants about their involvement in the following organization-based activities: (1) church activities (or other religious group); (2) cultural activities; and (3) activities in the community. Participants indicated their level of participation by choosing from the following options: “don’t participate”, “member”, “active member”, “group leader”, “prefer not to answer”, and “don’t know”. We categorized responses into two groups—those who did participate (members, active members, or group leaders) and those who did not participate for each of the organization-based activities.

Communication action context.

Where an individual lives and the urbanicity of the area in which one lives characterize their communication action contexts, which are influential characteristics that can affect the availability and accessibility of communication resources. Hence, we assessed participants’ potential communication action contexts by their Census region of residence and their urban-rural classification of county of residence based on their reported zip code, classifying contexts as large metro, medium metro, small metro, or micropolitan (National Center for Health Statistics, 2017; United States Census Bureau, 2021).

Sociodemographic characteristics.

We assessed participants’ sociodemographic characteristics including race, ethnicity, age, income, education, whether they had health insurance, and country of birth. Participants’ primary language was determined by their response to a question that asked about their primary language (English or Spanish). Additionally, eligibility for PrEP was determined among participants through self-reported behaviors if they (1) had a main male sex partner who is HIV positive, or (2) had 2 or more male sex partners in the past 12 months along with condomless sex or an STI diagnosis in the past 12 months.

Data Analyses

Chi-square tests or t-tests were used to examine the associations between Latinx SMM’s primary language and their sociodemographic characteristics, who they named as potential storytellers at the individual and meso-levels, and communication action context.

Poisson regression with robust variances was used to calculate prevalence ratios (PR) for the associations between the following set of predictors: (1) potential individual-level storytellers; (2) potential meso-level storytellers; and (3) communication action context characteristics and sexual health outcomes (i.e., HIV testing in the past 12 months, STI testing in past 12 months, awareness of daily oral PrEP, and oral PrEP use in the past 12 months). Each of the storytellers and communication action context variables were independent variables and the sexual health variables were dependent variables. All analyses were conducted separately for participants whose primary language was English and for those whose primary language was Spanish to identify potential differences in the communication networks of Latinx SMM in relation to PrEP outcomes. All models controlled for sociodemographic variables determined a priori as potential confounders (i.e., age, education, and health insurance) due to significant differences between English- and Spanish-speaking participants identified by t-tests or Chi-squared tests. The communication variables were then grouped and analyzed in a stepwise manner: Model 1 included potential individual-level storytellers, Model 2 included potential meso-level storytellers, and Model 3 included communication action context variables. We conducted a fourth model for sensitivity analyses, which included all predictors that were significant (p<.05) in the previous three models along with PrEP eligibility.

Results

Overall, there were 9,061 SMM who completed AMIS in 2021. Among the 1,186 SMM who identified as Hispanic/Latinx, 1,179 indicated a primary/preferred language (Spanish vs. English). Of these participants, 856 reported HIV-negative status and were the focus of our analyses. We excluded participants who reported HIV positive (n=153) or unknown status (n=170). The sociodemographic characteristics of Hispanic/Latinx SMM are presented in Table 1. Compared with individuals whose primary language was English (N = 564; 66.4%), herein referred to as English-speaking Latinx SMM, Hispanic/Latino SMM whose primary language was Spanish (N = 285; 33.6%), herein referred to as Spanish-speaking Latinx SMM, were older, earned lower incomes, reported lower education levels, were less likely to have health insurance, were more likely to be from the southern region of the U.S., were less likely to have been born in the U.S., and were less likely to be eligible for PrEP (p<.001) (Table 1).

Table 1.

Sociodemographic characteristics of Latinx sexual minority men by primary language, American Men’s Internet Survey, 2021^a

	English (n=564)	Spanish (n=285)	Total (N=849)	p value
Age				< 0.001
Mean (SD)	38.3 (13.1)	43.4 (12.1)	40.0 (13.0)
Range	18.0 – 78.0	20.0 – 77.0	18.0 – 78.0
Age Categories				< 0.001
40+	219 (38.8%)	163 (57.2%)	382 (45.0%)
30–39	180 (31.9%)	84 (29.5%)	264 (31.1%)
25–29	92 (16.3%)	30 (10.5%)	122 (14.4%)
15–24	73 (12.9%)	8 (2.8%)	81 (9.5%)
Income				< 0.001
$75,000+	248 (46.8%)	78 (31.1%)	326 (41.7%)
$40,000–74,999	76 (14.3%)	69 (27.5%)	145 (18.6%)
$20,000–39,999	157 (29.6%)	56 (22.3%)	213 (27.3%)
$0–19,999	49 (9.2%)	48 (19.1%)	97 (12.4%)
Missing	34	34	68
Education				< 0.001
High school or below	48 (8.5%)	56 (19.8%)	104 (12.3%)
Technical degree	169 (30.1%)	65 (23.0%)	234 (27.7%)
College or postgrad	345 (61.4%)	162 (57.2%)	507 (60.0%)
Missing	2	2	4
Insurance Type				< 0.001
No	51 (9.2%)	65 (23.7%)	116 (14.0%)
Yes	506 (90.8%)	209 (76.3%)	715 (86.0%)
Missing	7	11	18
Raceb				NA
White	156 (27.7%)	16 (5.6%)	172 (20.3%)
Asian	10 (1.8%)	2 (0.7%)	12 (1.4%)
American Indian	25 (4.4%)	2 (0.7%)	27 (3.2%)
Middle Eastern	9 (1.6%)	1 (0.4%)	10 (1.2%)
Native Hawaiian	3 (0.5%)	0 (0.0%)	3 (0.4%)
Other	2 (0.4%)	1(0.4%)	3 (0.4%)
Country of Birth c				< 0.001
United States	466 (84.4%)	52 (21.5%)	518 (65.2%)
Mexico	36 (6.5%)	82 (33.9%)	118 (14.9%)
Central America	16 (2.9%)	28 (11.6%)	44 (5.5%)
South America	29 (5.3%)	77 (31.8%)	106 (13.4%)
Other	5 (0.9%)	3 (1.2%)	8 (1.0%)
Missing	12	43	55
PrEP Eligibility				< 0.001
No	163 (28.9%)	126 (44.2%)
Yes	401 (71.1%)	159 (55.8%)

^aMissing data omitted from table

^bParticipants responded to separate questions for each race category. Therefore, participants were able to endorse more than one race category or not endorse any at all. Count and percentages of participants endorsing each race category were presented. Chi-square test was not conducted.

^cCentral American countries: Guatemala, Honduras, Nicaragua, El Salvador, Costa Rica, and Panama; South American countries: Argentina, Bolivia, Brazil, Chile, Colombia, Ecuador, Paraguay, Peru, and Venezuela; Other countries: Albania, Japan, Spain, Philippines, Belarus, Germany, and Paraguay

Storytelling networks of Latinx SMM

A greater proportion of English-speaking Latinx SMM (97.0%) reported having ever disclosed their sexual orientation to someone than Spanish-speaking SMM (89.7%, p<.001) (Table 2). English-speaking Latinx SMM were also more likely to have come out to their GLB friends (97.2% vs. 94.0%, p=.026) as well as to their family (83.7% vs. 77.1%, p=.027) and healthcare provider (83.1% vs. 72.9%, p<.001) than Spanish-speaking Latinx SMM. At the meso-level, a higher percentage of English-speaking participants were involved in cultural activities (30.1% vs. 22%, p=.001) and community-based activities (34.4% vs. 18.6%, p<.001) compared to Spanish-speaking participants (Table 2).

Table 2.

Differences in storytelling networks of English- vs. Spanish-speaking Latinx sexual minority men, American Men’s Internet Survey, 2021

	English (n=564)	Spanish (n=285)	Total (N=849)	p value
Individual-level storytellers
Number of MSM friends				0.306
Mean (SD)	22.6 (90.0)	15.9 (40.2)	20.5 (78.1)
Range	0.0 – 1500.0	0.0 – 500.0	0.0 – 1500.0
Missing	97	76	173
Sexual orientation disclosure
Out to: anyone				< 0.001
No	17 (3.0%)	29 (10.3%)	46 (5.5%)
Yes	543 (97.0%)	253 (89.7%)	796 (94.5%)
Missing	4	3	7
Out to: gay, lesbian, or bisexual (GLB) friends				0.026
No	15 (2.8%)	15 (6.0%)	30 (3.8%)
Yes	527 (97.2%)	234 (94.0%)	761 (96.2%)
Missing	22	36	58
Out to: non-GLB friends				0.054
No	37 (6.9%)	27 (10.9%)	64 (8.1%)
Yes	503 (93.1%)	221 (89.1%)	724 (91.9%)
Missing	24	37	61
Out to: family members				0.027
No	88 (16.3%)	57 (22.9%)	145 (18.4%)
Yes	451 (83.7%)	192 (77.1%)	643 (81.6%)
Missing	25	36	61
Out to: healthcare providers				< 0.001
No	91 (16.9%)	68 (27.1%)	159 (20.1%)
Yes	449 (83.1%)	183 (72.9%)	632 (79.9%)
Missing	91	68	159
Meso-level storytellers
Church activities				0.752
No	413 (79.0%)	201 (77.9%)	614 (78.6%)
Yes	110 (21.0%)	57 (22.1%)	167 (21.4%)
Missing	41	27	68
Cultural activities				0.001
No	350 (69.9%)	188 (78.0%)	538 (72.5%)
Yes	151 (30.1%)	53 (22.0%)	204 (27.5%)
Missing	63	44	107
Community activities				< 0.001
No	332 (65.6%)	206 (81.4%)	538 (70.9%)
Yes	174 (34.4%)	47 (18.6%)	221 (29.1%)
Missing	58	32	90
Communication action context
Census Region				< 0.001
Northeast	78 (13.8%)	49 (17.2%)	127 (15.0%)
Midwest	62 (11.0%)	32 (11.2%)	94 (11.1%)
South	194 (34.4%)	140 (49.1%)	334 (39.3%)
West	230 (40.8%)	64 (22.5%)	294 (34.6%)
NCHS rural-urban category				0.258
Large Metro	425 (75.5%)	213 (76.1%)	638 (75.7%)
Medium metro	87 (15.5%)	35 (12.5%)	122 (14.5%)
Small metro	26 (4.6%)	21 (7.5%)	47 (5.6%)
Micropolitan	25 (4.4%)	11 (3.9%)	36 (4.3%)
Missing	1	5	6

HIV testing, STI testing, and PrEP-related outcomes among Latinx SMM by language

English-speaking Latinx SMM were more likely to have tested for HIV in the past 12 months (77.8% vs. 70.5%, p=.020), to report being aware of daily oral PrEP (94.3% vs. 79.3%, p<.001), and to have used PrEP in the past 12 months (43.4% vs. 27.7%, p<.001), than Spanish-speaking Latino SMM (Table 3). Overall, the vast majority of Latinx SMM indicated that they were aware of PrEP (89.3%), although only about thirty-eight percent (38.2%) of all participants reported having taken PrEP in the past 12 months (Table 3). There were no differences in reports of STI testing in the past 12 months (Table 3).

Table 3.

HIV testing, STI testing, and PrEP-related outcomes among Latinx sexual minority men by primary language

	English (N=564)	Spanish (N=285)	Total (N=849)	p value
HIV test (past 12 months)				0.020
No	125 (22.2%)	84 (29.5%)	209 (24.6%)
Yes	439 (77.8%)	201 (70.5%)	640 (75.4%)
Any STI test (past 12 months)				0.128
No	244 (43.3%)	139 (48.8%)	383 (45.1%)
Yes	320 (56.7%)	146 (51.2%)	466 (54.9%)
Awareness of daily oral PrEP				< 0.001
No	32 (5.7%)	59 (20.7%)	91 (10.7%)
Yes	532 (94.3%)	226 (79.3%)	758 (89.3%)
PrEP use (past 12 months)				< 0.001
No	319 (56.6%)	206 (72.3%)	525 (61.8%)
Yes	245 (43.4%)	79 (27.7%)	324 (38.2%)

Communication networks and HIV testing and STI testing

None of the communication network models, which included individual-level and meso-level storytellers and communication action contexts, revealed significant associations with HIV testing for both English- and Spanish-speaking Latinx SMM (Table 4).

Table 4.

Associations between communication network factors and HIV and STI testing among Spanish- and English-speaking Latinx sexual minority men

	HIV Testing PR (95% CI)a			STI Testing PR (95% CI)
Spanish	Model 1b	Model 2b	Model 3b	Model 1	Model 2	Model 3
Intercept	0.60 (0.38, 0.97)	0.71 (0.64, 0.78)	0.87 (0.77, 0.99)	0.12 (0.04, 0.37)	0.53 (0.46, 0.61)	0.71 (0.59, 0.86)
Number of MSM friends	1 (1, 1)			1 (1, 1)
Out to: GLB friends	1.01 (0.63, 1.59)			1.60 (0.64, 3.95)
Out to: non-GLB friends	0.93 (0.65, 1.33)			2.22 (0.94, 5.27)
Out to: family	0.92 (0.73, 1.17)			1.01 (0.73, 1.40)
Out to: healthcare providers	1.49 (1.10, 2.01)			1.48 (0.97, 2.26)
Church activities		1.04 (0.85, 1.27)			1.12 (0.86, 1.45)
Cultural activities		0.95 (0.72, 1.27)			1.01 (0.71, 1.43)
Community activities		1.02 (0.76, 1.37)			1.10 (0.77, 1.57)
Census Region: Midwest			0.86 (0.67, 1.12)			0.64 (0.42, 0.99)
Census Region: South			0.80 (0.68, 0.95)			0.61 (0.47, 0.79)
Census Region: West			0.78 (0.64, 0.97)			0.75 (0.56, 1.00)
NCHS Rural-Urban: Medium metro			0.86 (0.66, 1.12)			1.07 (0.77, 1.48)
NCHS Rural-Urban: Micropolitan			0.73 (0.43, 1.26)			0.90 (0.47, 1.75)
NCHS Rural-Urban: Small metro			0.83 (0.58, 1.19)			0.94 (0.59, 1.51)
Observations	181	235	280	181	235	280
R 2	0.05	0.001	0.025	0.101	0.006	0.039
	HIV Testing			STI Testing
English	Model 1b	Model 2b	Model 3b	Model 1	Model 2	Model 3
Intercept	0.47 (0.30, 0.74)	0.76 (0.72, 0.82)	0.85 (0.77, 0.94)	0.27 (0.15, 0.47)	0.59 (0.54, 0.65)	0.74 (0.64, 0.85)
Number of MSM friends	1 (1, 1)			1 (1, 1)
Out to: GLB friends	1.08 (0.76, 1.54)			1.07 (0.71, 1.63)
Out to: non-GLB friends	1.26 (0.95, 1.68)			1.36 (0.89, 2.07)
Out to: family	1.00 (0.86, 1.17)			0.98 (0.77, 1.24)
Out to: healthcare providers	1.29 (1.06, 1.56)			1.80 (1.31, 2.49)
Church activities		0.99 (0.88, 1.11)			0.95 (0.79, 1.14)
Cultural activities		1.00 (0.89, 1.12)			1.08 (0.91, 1.29)
Community activities		1.06 (0.95, 1.19)			1.05 (0.88, 1.25)
Census Region: Midwest			1.04 (0.90, 1.21)			0.85 (0.67, 1.09)
Census Region: South			0.93 (0.82, 1.06)			0.80 (0.66, 0.97)
Census Region: West			0.92 (0.81, 1.04)			0.77 (0.64, 0.93)
NCHS Rural-Urban: Medium metro			0.84 (0.72, 0.98)			0.69 (0.53, 0.90)
NCHS Rural-Urban: Micropolitan			0.74 (0.54, 1.01)			0.94 (0.66, 1.33)
NCHS Rural-Urban: Small metro			0.90 (0.71, 1.15)			0.71 (0.45, 1.12)
Observations	441	486	563	441	486	563
R 2	0.034	0.002	0.021	0.064	0.003	0.033

^aPR: Prevalence Ratios

^bModel 1 included potential individual-level storytellers, Model 2 included potential meso-level storytellers, and Model 3 included communication action context variables.

Among English-speaking Latinx SMM, sexual orientation disclosure to a healthcare provider was associated with higher likelihood of STI testing over the past 12 months (PR=1.80, 95%CI: 1.21, 2.80). Residing in a medium-sized metro city (vs. large-sized metro city) was associated with lower likelihood of STI testing over the past 12 months (PR=0.69, 95%CI: 0.48, 0.96) (Table 4). After controlling for PrEP eligibility, disclosure to a healthcare provider remained significantly associated with greater likelihood of STI testing over the past 12 months (PR=1.71, 95%CI: 1.18, 2.55) (Supplemental Table 1).

Among Spanish-speaking participants, residing in the south (vs. northeast) (PR=0.61, 95%CI: 0.41, 0.94) was associated with lower likelihood of STI testing over the past 12 months (Table 4). Census region of residence was no longer significantly associated with STI testing after controlling for PrEP eligibility (Supplemental Table 1).

Communication networks and PrEP-related outcomes

Across all models, none of the communication network variables were significantly associated with PrEP awareness among English-speaking and Spanish-speaking Latinx SMM (Table 5). Among English-speaking Latinx SMM, disclosure of sexual orientation to a healthcare provider was also associated with higher likelihood of PrEP use in the past 12 months (PR=2.39, 95%CI: 1.44, 4.26). Participating in church was associated with lower likelihood of PrEP uptake in the past 12 months (PR=0.67, 95%CI: 0.45, 0.96). After controlling for PrEP eligibility, only disclosure to a healthcare provider remained a significant predictor of PrEP use in the past 12 months (PR=2.36, 95%CI: 1.45, 4.17) (Supplemental Table 1).

Table 5.

Associations between communication network factors and PrEP-related outcomes among Spanish- and English-speaking Latinx sexual minority men

	PrEP Awareness PR (95% CI)a			PrEP Use PR (95% CI)
Spanish	Model 1b	Model 2b	Model 3b	Model 1	Model 2	Model 3
Intercept	0.75 (0.52, 1.08)	0.82 (0.76, 0.88)	0.83 (0.72, 0.94)	0.06 (0.01, 0.31)	0.29 (0.23, 0.37)	0.43 (0.30, 0.62)
Number of MSM friends	1 (1, 1)			1 (0.99, 1)
Out to: GLB friends	0.9 (0.67, 1.21)			0.94 (0.31, 2.89)
Out to: non-GLB friends	1.07 (0.79, 1.46)			2.79 (0.77, 10.04)
Out to: family	0.88 (0.75, 1.03)			0.58 (0.38, 0.87)
Out to: healthcare providers	1.40 (1.11, 1.77)			3.78 (1.37, 10.47)
Church activities		0.79 (0.65, 0.96)			0.83 (0.49, 1.40)
Cultural activities		1.01 (0.85, 1.20)			0.95 (0.48, 1.88)
Community activities		1.14 (0.96, 1.35)			1.32 (0.68, 2.58)
Census Region: Midwest			1.22 (1.02, 1.45)			1 (0.56, 1.78)
Census Region: South			0.92 (0.78, 1.09)			0.5 (0.30, 0.83)
Census Region: West			1.02 (0.85, 1.21)			0.94 (0.58, 1.53)
NCHS Rural-Urban: Medium metro			0.97 (0.80, 1.17)			0.38 (0.15, 0.93)
NCHS Rural-Urban: Micropolitan		0.75 (0.49, 1.15)			0.55 (0.15, 2.01)
NCHS Rural-Urban: Small metro			0.74 (0.53, 1.03)			0.29 (0.08, 1.11)
Observations	181	235	280	181	235	280
R 2	0.067	0.026	0.035	0.153	0.007	0.12
	PrEP Awareness			PrEP Use
English	Model 1b	Model 2b	Model 3b	Model 1	Model 2	Model 3
Intercept	0.86 (0.71, 1.04)	0.94 (0.91, 0.97)	0.99 (0.96, 1.01)	0.11 (0.04, 0.28)	0.42 (0.37, 0.48)	0.48 (0.38, 0.61)
Number of MSM friends	1 (1, 1)			1 (1, 1)
Out to: GLB friends	1.03 (0.89, 1.21)			1.37 (0.64, 2.94)
Out to: non-GLB friends	1.06 (0.94, 1.21)			1.59 (0.91, 2.77)
Out to: family	0.97 (0.91, 1.03)			0.88 (0.66, 1.18)
Out to: healthcare providers	1.06 (0.99, 1.13)			2.39 (1.49, 3.82)
Church activities		0.95 (0.89, 1.01)			0.67 (0.50, 0.89)
Cultural activities		1.01 (0.97, 1.06)			1.05 (0.83, 1.34)
Community activities		1.05 (1.00, 1.09)			1.34 (1.06, 1.68)
Census Region: Midwest			1.01 (0.96, 1.05)			1.21 (0.86, 1.69)
Census Region: South			0.93 (0.89, 0.98)			0.84 (0.62, 1.14)
Census Region: West			0.95 (0.92, 1.00)			0.99 (0.75, 1.31)
NCHS Rural-Urban: Medium metro			1.01 (0.96, 1.07)			0.77 (0.56, 1.04)
NCHS Rural-Urban: Micropolitan			0.93 (0.81, 1.07)			0.86 (0.53, 1.41)
NCHS Rural-Urban: Small metro			0.94 (0.82, 1.08)			0.67 (0.38, 1.18)
Observations	441	486	563	441	486	563
R 2	0.009	0.008	0.011	0.084	0.032	0.024

^aPR: Prevalence Ratios

^bModel 1 included potential individual-level storytellers, Model 2 included potential meso-level storytellers, and Model 3 included communication action context variables.

^cRural-urban variable was removed due to empty cells that caused model instability in model estimation

For Spanish-speaking Latinx SMM, residing in the south (vs. northeast) (PR=0.50, 95%CI: 0.27, 0.93) was associated with lower likelihood of PrEP uptake in the past 12 months. Additionally, disclosure of sexual orientation to a healthcare provider was associated with higher likelihood of having used PrEP within the past 12 months (PR=3.78, 95%CI: 1.49, 12.88) among Spanish-speaking Latinx SMM. The association between disclosure to a healthcare provider and PrEP use in the past 12 months among Spanish-speaking Latinx SMM was stronger after controlling for PrEP eligibility (PR=4.10, 95%CI: 1.93, 10.62) (Table 5).

Discussion

This study examined the association between communication network-related factors and access to and use of HIV prevention resources among Latinx SMM. Specifically, we assessed differences in the communication networks of Latinx SMM whose primary language is Spanish and those whose primary language is English to examine disparities in HIV testing, STI testing, PrEP awareness, and PrEP use.

Results demonstrated that Spanish-speaking Latinx SMM experienced multiple structural disadvantages to accessing HIV-related services. As Spanish-speaking Latinx SMM in this study were generally less educated, earned lower incomes, less likely to have health insurance, and more likely to be born outside the U.S., our study underscores the structural barriers this group faces to accessing information and healthcare. This is consistent with the literature on immigration and acculturation, where immigrant groups are likely to be excluded from health care options due to immigration status and face barriers to accessing health care due to immigration-related stressors (Tanner et al., 2014). Sexual identity stigma may exacerbate these barriers for Latinx SMM, particularly for those who are new arrivals to the U.S. and may be unfamiliar with U.S. social and cultural norms (Jaiswal et al., 2018). This is consistent with the lower levels of sexual orientation disclosure we found among Spanish-speaking participants.

Individual- and meso-level storytellers are important channels from which individuals can obtain information related to HIV and prevention (Wilkin, 2013; Wilkin et al., 2010). However, we found that Spanish-speaking Latinx SMM were less likely to report access to both the individual- and meso-level storytellers that we assessed. Specifically, Spanish-speaking Latinx SMM reported lower rates of disclosure to family, GLB friends, healthcare providers, and lower involvement with community and cultural activities at the meso-level than their English-speaking counterparts. This might be due to different types of or smaller networks in the Latinx SMM community that emerge from cultural norms and expectations (Lee et. al, 2020; Sandfort et al., 2007; Steffens et al., 2015). Hence, Spanish-speaking Latinx SMM may strategically choose not to disclose to network members or engage in community activities for fear of negative social consequences (Chuang & Lacombe-Duncan, 2016; Wells et al., 2023; Zea et al., 2004). It is also possible that due to limited English proficiency, access to healthcare care and participation in organizational activities may be less accessible for Spanish-speaking Latinx SMM if English language communication is prioritized at those organizations (Nakamura et al., 2013; Wohl et al., 2009).

As community engagement not only widens one’s communication network but is positively associated with mental and physical health among marginalized populations and individuals living with HIV (Chuang & Lacombe-Duncan, 2016; Roy & Cain, 2001; Stahlman et al., 2016), greater understanding of organizational involvement among Spanish-speaking Latinx SMM in the U.S. may be needed to facilitate access to health-related information and HIV prevention.

At the individual level, our results indicated that disclosure to a healthcare provider was associated with increased likelihood of PrEP use among both English- and Spanish-speaking Latinx SMM. This finding highlights the importance of establishing a trusting provider-patient relationship, especially for Latinx SMM who face additional language-related barriers to obtaining HIV prevention information during healthcare visits. This highlights the importance of ensuring that culturally and linguistically appropriate health and prevention services are available. Language and stigma, among other factors, might hinder the disclosure of SMM identity to a healthcare provider (Durso & Meyer, 2013; Gessner et al., 2020). Further, language has implications for training and supporting providers in delivering culturally sensitive services for SMM and relaying information in inclusive and appropriate ways for diverse communities (Gessner et al., 2020; Steele et al., 2006). Inclusive environments where Latinx SMM can feel safe to disclose their sexual orientation may facilitate healthcare utilization and increase access to HIV prevention resources (Tanner et al., 2014). Such efforts may require attention to community-based services that include peer-based services or peer navigation interventions that can reduce stigma and offer services in culturally sensitive ways. Specifically, peer-based interventions can address potential cultural and language barriers that Latinx SMM may encounter given greater trust, access, and shared characteristics that peers have with the population (He et al., 2020).

Among meso-level storytellers, we found that involvement with church activities was associated with a lower likelihood of PrEP use among English-speaking Latinx SMM. This finding may be due to the association between religion and shame-related HIV stigma, where someone who uses PrEP is perceived as “promiscuous,” undermining individuals’ motivation to use PrEP (Calabrese & Underhill, 2015). Notably, this effect was no longer significant after controlling for disclosure to a doctor and PrEP eligibility among English-speaking Latinx SMM. Despite similar levels of church engagement between English-speaking and Spanish-speaking participants, involvement in church activities was not associated with PrEP use among Spanish-speaking Latinx SMM. In the same way that higher acculturation levels among English-speaking Latinx SMM might increase exposure to information perpetuating HIV and sexual identity stigma in mainstream U.S. culture, English-speaking Latinx SMM who are involved in church may be more likely to receive messaging that negatively impacts PrEP uptake (Chng et al., 2003). Our finding may also be due to lower levels of PrEP uptake among Spanish-speaking Latinx SMM in this study (27.7%) compared to the English-speaking Latinx SMM (43.4%) —hence, an association was less likely to be found due to the floor effect (Vosvick et al., 2016).

As for the communication action context, we found that Spanish-speaking Latinx SMM who resided in the southern region of the U.S. compared to those from the northeast were less likely to use PrEP and report STI testing over the past 12 months. Compared to other geographical regions in the U.S., individuals from the South might experience greater HIV stigma and negative attitudes towards sexual minorities (Baunach & Burgess, 2013; Berg & Ross, 2014; Darlington & Hutson, 2017). Notably, Latinx immigrants were more likely to reside in the South where they may face anti-immigrant policies and language barriers due to lack of Spanish-speaking providers (Adimora et al., 2014; Morales-Aleman & Sutton, 2014). As some of these states have not expanded Medicare as part of the Affordable Care Act, greater uninsurance rates in these states may also limit access to health and HIV care (Luque, Soulen, Davila, & Cartmell, 2018). The lack of access to HIV-related services is reflected by the high rates of HIV infections in the South compared to other regions in the US from 2015 to 2019 (CDC, 2022).

Among English-speaking Latinx SMM, those from medium-sized metro areas were less likely to have tested for any STI compared to those from large metro areas. Individuals residing in less urban areas can face unique barriers to accessing care, including lack of HIV providers and treatment sites, long distances to local HIV treatment sites, and concerns about privacy and confidentiality (Sutton et al., 2010; Tran et al., 2020; Trepka et al., 2014). Together, these findings helped identify the geographical location factors (communication action context) where increased HIV prevention efforts could be directed. However, this effect disappeared after controlling for disclosure to doctors and PrEP eligibility, suggesting that PrEP eligibility information and sexual orientation disclosure to doctors could protect individuals against geographic factors and enable individuals to engage in HIV and STI testing and PrEP-related services (Bernstein et al., 2008; Qiao et al., 2018).

Despite the valuable insights provided by this study, several limitations should be acknowledged. First, the cross-sectional design prevents us from drawing causal inferences, and longitudinal studies are necessary to better understand the directionality of the observed associations and inform intervention studies. Second, the self-reported nature of the data might be subject to social desirability and recall biases, which could potentially impact the accuracy of the findings. Third, our sample may not be representative of the broader Latinx SMM population, as the participants were recruited through convenience sampling methods, limiting the generalizability of the results. Additionally, the study did not account for the potential influence of other contextual factors, such as access to healthcare providers or regional policies related to health insurance or health provision, which might further contribute to the observed disparities. Although we focused on language as a proxy for cultural and structural barriers, it may not capture all the complexities associated with the experiences of Latinx SMM in accessing healthcare resources. Further, for participants who spoke Spanish and English equally, the survey did not permit selection of both languages as their primary language. Lastly, we assumed that individuals to whom Latinx SMM disclosed their sexual orientation were part of participants’ storytelling networks, but disclosure does not always lead to storytelling. We also did not assess the types of storytelling related to HIV that participants were exposed to. Additionally, we did not analyze other social ties, and it is possible that individuals who were not aware of participants’ sexual orientation may still be important storytellers who relay relevant information about HIV prevention. Further, the media are storytellers often included in CIT that was not accounted for in this study. Future research should address these limitations by employing longitudinal designs, using more representative sampling methods, and considering additional contextual and communication-related factors to provide a more comprehensive understanding of the barriers faced by Latinx SMM in accessing HIV prevention and care. Moreover, exploring alternative proxies for cultural and structural barriers may help to develop more effective interventions tailored to the unique needs of this population.

Conclusion

This study highlights the significant disparities between Spanish-speaking and English-speaking Latinx SMM in their communication resources and in accessing and utilizing HIV prevention services. Attention to salient storytellers across subgroups of Latinx SMM by language preference reveals potential opportunities for increasing storytelling and enhancing utilization of HIV prevention services. The findings demonstrate how individual- and meso-level communication networks can facilitate access to HIV prevention information and services. Establishing trusting provider-patient relationships and fostering inclusive healthcare environments are essential for overcoming barriers to health information and ensuring that health and HIV-related information reach Latinx SMM and are communicated effectively.

Interventions must also support these salient storytellers to relay information that will facilitate and not inhibit HIV service-seeking behaviors. Efforts will require a focus on the contextual characteristics and norms of the community to promote the tailoring and relevance of information in the preferred language. Enhancing community engagement and targeting specific subgroups, such as those from specific geographic regions, may improve access to care and overall health outcomes for Latinx SMM. Future research should focus on consolidating these findings and exploring interventions that address language barriers, increase communication resources, reduce stigma, and strengthen social and healthcare networks to promote equitable access to HIV prevention and care among Latinx SMM.

Availability of data and material:

The data used in this study were completely derived from the American Men’s Internet Survey conducted by PRISM Health at Emory University and are available by request.