Left untreated, syphilis can result in serious health complications including increased risk of HIV infection (Abara et al., 2016; Kidd et al., 2018). Over the past two decades, rates of syphilis incidence in the United States have increased dramatically. From 2000 to 2017, there was a 413% increase in the number of newly diagnosed syphilis cases (Schmidt et al., 2019). This trend is largely attributed to increased cases among men, particularly gay, bisexual, and other sexual minority men (SMM; Centers for Disease Control and Prevention, CDC, 2019). In 2018, SMM accounted for 53.5% of all syphilis cases and 82% of male syphilis cases (CDC, 2019). SMM diagnosed with syphilis also experience high rates of co-infection with HIV (CDC, 2019; Schmidt et al., 2019). In 2018, of reported syphilis cases among SMM, 41.6% were among SMM with HIV (CDC, 2019).

Syphilis and HIV disparities exist among Black sexual minority men (BSMM). A comparison of national syphilis rates revealed that syphilis diagnoses among BSMM (583.9/100,000) were more than three times higher compared with White SMM (168.4/100,000; Grey et al., 2017). One study in New York City showed that between 2012 and 2016, syphilis rates increased by 81% among SMM, with the largest increase among non-Hispanic BSMM (Schillinger et al., 2018). Explanations for this increase include differences in social network structures (Fujimoto et al., 2018) and other factors such as younger age, history of sexually transmitted infections (STIs), and structural factors such as wealth inequality (Allan-Blitz et al., 2021; Owusu-Edusei et al., 2013). Syphilis incidence is often indicative of high transmission syphilis and HIV networks (Fujimoto et al., 2018; Jennings et al., 2020). In past years, many SMM have met sexual partners at various physical locations, which have transitioned to include online environments in more recent years; collectively, these locations helped form dense sexual networks (Fujimoto et al., 2018; Jennings et al., 2020). Participation in such networks by SMM increases the risk of syphilis transmission and HIV co-infection (Fujimoto et al., 2018; Jennings et al., 2020). A recent study found that sexual networks among BSMM in Baltimore are associated with increased rates of syphilis, HIV, and co-infection diagnoses (Jennings et al., 2020). Analyses of sexual networks among SMM in other areas with high syphilis and HIV burden show similar correlations and concentrations of syphilis and HIV co-infections among young BSMM (Fujimoto et al., 2018). Given this relationship, assessing social networks among BSMM may be useful in understanding syphilis transmission patterns.

Social networks influence health behaviors such as alcohol and tobacco consumption, drug use, and sexual practices (Desai et al., 2020; Gustafsson et al., 2021). A mechanism by which social networks influence health, including potentially syphilis infection, is social support. Social support can be categorized into (a) emotional support, involving the provision of empathy, love, trust, and caring; (b) instrumental support, involving the provision of tangible aid and services; (c) informational support, involving the provision of advice, suggestions, and information; and (d) appraisal support, involving the provision of information that is useful for self-evaluation (Glanz et al., 2008). Through increased social support, individuals can increase their self-efficacy and gain more confidence in their ability to perform health-promoting behaviors (Al-Dwaikat et al., 2021; Wu & Sheng, 2019). Research has shown that social networks that provide these forms of social support influence HIV risk and prevention behaviors (Hermanstyne et al., 2018).

Important characteristics of social networks include network size (number of individuals in network), density (connections between individuals in network), and homogeneity (similarity between individuals in network). The literature shows high homophily (contact between similar network members) and density within the social networks of men who have sex with men, including BSMM, result in high HIV and syphilis transmission rates among BSMM (Asikainen et al., 2020; Jennings et al., 2020).There are limited data exploring the relationships between social networks and syphilis testing among BSMM. To address this gap in the literature, we evaluated the impact of social networks on syphilis testing among BSMM in Baltimore City (Maryland, United States). Baltimore City has had one of the most severe syphilis epidemics in the United States. In 2018, the syphilis rate in Baltimore City (45.3/100,000; Maryland Department of Health, 2019) was 4.2 times higher than the national rate (10.8/100,000; CDC, 2019). From 2009 to 2015, 86.2% of syphilis cases were attributed to BSMM and 67.2% to SMM co-infected with HIV (Schumacher et al., 2018). Similarly, surveillance data indicate that syphilis and HIV are concentrated among young BSMM, between 20 to 30 years of age (Cooley et al., 2015; Schumacher et al., 2018). Although the CDC recommends frequent syphilis testing for high-risk groups such as SMM (Said et al., 2015; Schumacher et al., 2018), BSMM have low testing rates. Pooled data from the National HIV Behavioral Surveillance SMM cycles conducted in 2008, 2011, and 2014 showed that 26% of BSMM tested for syphilis in the past 12 months, similar to Hispanic SMM (26%) and lower than White SMM (41%; An et al., 2017).

The present analysis assessed the effects of social network members providing emotional, informational, and instrumental support on syphilis testing behavior in the past year among BSMM. We hypothesize that informational support, along with structural characteristics of social networks, would explain variance in syphilis testing. These data are derived from the Social Networks and Prevention (SNAP) study, a randomized controlled trial of a sexual health behavioral intervention for BSMM in Baltimore City (Dayton et al., 2020). Our aim was to identify individual and social network characteristics that influenced syphilis testing and whether social networks offer an effective pathway to increase syphilis testing among BSMM.

Methods

Participants, Design, and Setting

This is an analysis of baseline data from the SNAP randomized control trial (RCT) of a sexual health behavioral intervention for BSMM in Baltimore, Maryland (Dayton et al., 2020). Participants were recruited into the intervention on a rolling basis, with baseline data visits conducted from October 2012 to November 2015. Data collection procedures at the baseline visit for eligible participants included providing written informed consent, behavioral risk assessment, and a social network survey. Individuals were eligible if (a) they were 18 years of age or older; (b) self-reported engaging in sex with a male partner in the past 6 months; (c) were assigned biological male sex at birth; (d) identify as male gender; (e) identified as Black race; and (f) engaged in condomless sex with any partner in the past 6 months. We restricted analysis to individuals who were BSMM and were not missing the outcome measure. Of the 318 individuals enrolled in the parent RCT, 62 (19%) were missing outcome data yielding a sample of 256 included in this analysis. In a sensitivity analysis, those excluded had significantly higher mean numbers of alters (individuals in their network; p < .001) and alters with HIV (p =.05), compared with those with outcome measures, but did not significantly differ on all demographics and all other social network variables used in the present analysis. The Johns Hopkins Bloomberg School of Public Health Institutional Review Board (IRB #3617) approved all study procedures.

Measures

The outcome measure used in this study was self-reported testing for syphilis in the past year, which was treated as a dichotomous variable (yes/no).

Independent Variables

Individual Characteristics

Age, sexual orientation (heterosexual or straight; bisexual; queer; gay, or same gender-loving; not sure or questioning; and other), highest education level, current health insurance status (yes or no), diagnosis of an STI (gonorrhea, Chlamydia, or syphilis) in the past 6 months (yes or no), condom use in the past 90 days (always used condoms or inconsistently/never), and searching for information online about STIs were all self-reported. HIV status was measured with a rapid HIV test completed at the participant's baseline visit or documentation of a confirmed HIV infection.

Network Characteristics

We used an egocentric network inventory to collect data on the social network (Barrera et al., 1981). For the first section, the participant generated names of alters. For example, participants were asked to list individuals with whom they talked about personal issues, individuals who provided economic support, individuals who used alcohol or drugs with them within the past 6 months, and with whom they have had sex in the past 90 days. At the follow-up assessments, participants could rename individuals and add new names. In the second section, participants were asked to characterize each listed alter by demographics, substances the alter used, and communication frequency about health topics. For the analysis, we included the following variables: total network size, number of male sexual partners, and number of alters with HIV. Emotional support included the mean number of alters the participant talked to about private issues. Informational support was mean number of alters that could be asked for health advice. Instrumental support was the mean number of alters who could provide help if needed (e.g., run errands). Economic support included alters who lent money. Health communication with alters was assessed by number of alters the participant encouraged to get syphilis tested, talked to about HIV testing, and/or gave condoms to.

A descriptive norm, a measure of what people in a group think, was measured of SMM friends who think it is important to tell their health care providers they have sex with men. Responses were none (0%), a few (25%), half (50%), most (75%), and all (100%). We collapsed responses into a few to half (25-50%) and most to all (75-100%), with none (0%) as the reference category.

We also asked how comfortable the participant felt accompanying a friend to get tested for HIV or STIs as a 5-point Likert scale and transformed it to a dichotomous variable, collapsing very comfortable and comfortable compared with the reference uncomfortable (neutral, uncomfortable, or very uncomfortable).

Data Analysis

Bivariate associations were assessed using t-test for continuous variables and chi-square for dichotomous and categorical variables to compare the difference of individual and network characteristics between participants who have tested for syphilis during the past year and those who did not. We applied hierarchical logistic regression to assess associations of selected social network characteristics that were associated with syphilis testing and where variance was explained by social network characteristics. Model one only included individual variables. Network social support variables and network health communication variables were put into Model two. Finally, significant variables (p<0.05) from Models one and two were included in Model three. The parent study was powered on HIV and STI prevention behaviors for BSMM, particularly reporting always using condoms during sex. Post-hoc power analyses conducted in G*Power (Faul et al., 2007) with inputs of the final sample size, a two-sided alpha of 0.05, and demographic and network covariates indicated sufficient power (>=0.8) to conduct these regressions.

Results

Among the 256 included participants, 63% of them (n = 161) did not test for syphilis during the past year (Table 1). The mean age of the sample was 38 years, with 55% reporting homosexual, gay, or same gender-loving sexual identity. Only a small proportion of the overall population (21%) reported not completing high school or equivalent. About 78% of the sample had health insurance, 43% had HIV, and 40% reported a history of an STI. When analyzing the sociodemographic characteristics, health status, and behaviors of BSMM with different syphilis testing behavior, there were significant associations. Participants who were younger, more educated, and had health insurance were more likely to have tested for syphilis within a year (all p < .05). When looking at the health-related behaviors, having HIV and having a history of STIs increased the likelihood of receiving a syphilis test. Being tested for HIV and searching the Internet for information about STIs were significantly more common among those who were tested for syphilis in the past year.

Table 1. Personal and Network Characteristics of BSMM (N = 256)

Network characteristics also significantly differed between those who tested for syphilis and those who did not. Individuals who tested for syphilis in the past year had significantly more alters in their network, male partners, and alters with HIV. For informational support, those who tested for syphilis had more alters whom they could talk to about HIV testing and ask for health advice. This group also had a significantly higher number of alters whom they encouraged to get syphilis testing, could talk to about private matters, and who felt comfortable accompanying a friend to get tested for HIV or STIs. A greater proportion of individuals who tested for syphilis reported that more than 75% of their friends think it is important to tell health care providers that they have sex with men. Economic support did not differ by outcome.

Table 2 shows the three multivariable logistic regressions. We fit an adjusted model with participant's age, health insurance, HIV status, history of STIs, and a variable indicating whether the participant had searched the Internet for information about STIs. All variables, except searching for information about STIs, were significantly associated with testing for syphilis and were included in Model three. Younger age {adjusted odds ratio 0.97; 95% confidence interval (CI) [0.94-0.99], p = .017} was associated with testing for syphilis. Individuals with health insurance had 2.67 (95% CI [1.23-5.96]; p = .01) times higher adjusted odds of syphilis testing compared with those who were uninsured, and those with HIV had 2.1 (95% CI [1.20-3.86]; p = .01) times higher adjusted odds of testing compared with being without HIV. A history of STIs had the strongest association with syphilis testing. Having a history of STIs was associated with 3.7 (95% CI [2.07-6.64]; p < .01) times higher adjusted odds of syphilis testing in the past year compared with those who reported no history of STIs.

Table 2. Multivariable Regression Models for Syphilis Screening Among BSMM (N = 256).

Our examination of network characteristics in Model two found number of male partners in the network, number of alters participants encouraged to get a syphilis test, and feeling comfortable accompanying a friend to their HIV or STI test were all associated with participants testing for syphilis in the past year. These three variables were carried into our final model with the significant individual characteristics highlighted (Model three column). For every increased number of male partners in their network, participants had 33% (95% CI [1.03-1.72]; p = .03) higher adjusted odds of testing for syphilis in the past year. There was also 94% (95% CI [1.12-3.36]; p = .02) higher adjusted odds of testing for syphilis for every increase in alters whom the participant encouraged to get a syphilis check. Finally, those who reported feeling comfortable accompanying their alters to get an HIV or STI test were 2.83 times significantly more likely to test for syphilis compared with those who felt uncomfortable (95% CI [1.17-6.83]; p = .02).

All network characteristics in Model three remained significantly associated with syphilis testing in the past year. Younger age (95% CI [0.93-0.98]; p = .00) and a history of STIs (95% CI [2.15-7.31]; p < .00) had slightly stronger significant associations with syphilis testing.

Although attenuated, having health insurance (95% CI [1.18-6.04]; p = .02) and living with HIV (95% CI [1.05-3.59]; p = .03) were both significantly associated with the outcome in the final adjusted model. Increases in the number of male partners and alters whom the participant encouraged to get STI testing had 22% (95% CI [1.01-1.46]; p = .03) and 89% (95% CI [1.19-3.00]; p = .01) increased odds of testing for syphilis, respectively. Finally, feeling comfortable accompanying a member of their network to get HIV or STI testing was associated with syphilis testing (95% CI [1.00-6.07]; p = .05).

Discussion

The purpose of this analysis was to examine whether social network characteristics are associated with syphilis testing in a sample of sexually active Black sexual minority men. We found that nearly two-thirds of the sample reported not testing for syphilis in the past 12 months, similar to previous surveillance estimates conducted in Baltimore, MD (Said et al., 2015). Given the sustained increase in syphilis infections, especially among Black sexually active SMM and the low prevalence in this sample of always using condoms, this finding is concerning.

We found several structural and functional network characteristics were associated with testing. Structurally, more male sex partners were associated with higher odds of recent STI testing. This may be associated with increased risk of previous syphilis infection, given increased number of partners. Sexual networks with a high proportion of SMM and prevalence of HIV and syphilis have been linked as a salient factor in transmission among BSMM (Fujimoto et al., 2018; Jennings et al., 2020). However, studies have shown that other behaviors, including substance use before sex, high rates of condomless sex, and low knowledge of individuals' HIV serostatus in these networks, were also associated with HIV and STI transmission (Heidari et al., 2020; Tobin et al., 2016).

Functionally, several social network variables were significant in the multivariable model controlling for individual-level variables. Encouraging alters to get tested for syphilis was associated with a nearly 2-fold increase in individual STI testing. Cognitive dissonance theory posits that incongruence between what one promotes and individual behavior creates anxiety and either leads to cessation of promotion or alignment of behavior with what is being promoted (Festinger, 1957). This mechanism is used in peer-based interventions where an individual is trained to promote and diffuse health-related information and resources through their social networks. Peer education models behavior, reduces stigma of discussing the behavior, and heightens perceptions that the behavior is normative. The discussion may also provide information on how to overcome barriers to testing as well as enhancing self-efficacy. Numerous HIV peer-based interventions have demonstrated efficacy (Katz et al., 2021; Shangani et al., 2017). Encouraging testing in the egocentric network also operates to establish a positive social norm about testing in general and has been shown to be sustained in social networks (Yamanis et al., 2017; Zhao et al., 2018). Adding syphilis testing and general STI testing to the peer trainings may improve testing uptake. With the advent of home-based testing kits for syphilis and studies that have demonstrated acceptability of home-based testing (McRee et al., 2015), peers could be used to distribute tests with clinic information for confirmatory testing. Future research could also examine the best ways to bring up syphilis testing in conversations and what messages diffuse in networks.

Comfort accompanying a friend for HIV and/or STI testing was also associated with a 2.5-fold increase in personal syphilis testing. Willingness to provide emotional and instrumental support to a friend may be a proxy measure for presence of emotional and instrumental support in the egocentric network. Syphilis testing was associated with HIV status, which may reflect the clinical guidelines for STI testing at HIV care visits, because this population is more likely to have more health care visits with increased testing (Workowski et al., 2021). However, there may be a lack of attention to the need for BSMM to be tested for other STIs, including syphilis, at other primary care visits. Low disclosure to providers about same-sex behavior is associated with lower testing (Qiao et al., 2018). In bivariate analysis, we found that a greater proportion of those who had undergone syphilis testing significantly reported that disclosure of same-sex behavior to providers is highly important. Universal testing of sexually active individuals in primary care settings for all STIs could improve the identification of infection and prompt treatment. This is of particular importance because 45% of the study sample did not identify as gay or same gender-loving and may not have a strong social network to offer informational or emotional support. Unlike HIV rapid testing and urine-based testing for bacterial STIs, testing for syphilis requires a serum sample or dried blood spots, and therefore, many nongovernmental organization HIV testing sites are not able to provide access, beyond a referral. Although the only point-of-care syphilis test that provides a confirmatory determination of active infection was recently recalled by the FDA (Food and Drug Administration, 2021), availability of such community syphilis testing should increase access and rapid treatment. Research assessing acceptability and feasibility of home-based syphilis testing will be important to identify additional barriers to use, such as optimizing access to confirmatory testing, counseling, and treatment.

There are several limitations to this study that should be acknowledged. This study was conducted with a convenience sample of BSMM in an urban setting, which may not be generalizable to other locations. Syphilis testing data were self-reported and therefore could reflect reporting bias. Reasons for testing were not collected, including if the participant received testing through contact tracing, due to being symptomatic, or through peer influence. We did, however, control for a history of STIs and searched for information regarding STIs in our final model. There were 62 individuals enrolled in the parent RCT who were missing on the outcome measure and significantly more likely to have a higher mean number of alters and alters living with HIV, limiting the sample size. However, these two measures did not significantly predict the outcome and were not included in the final model. Finally, the network measure may not have captured key aspects of the network, such as transient relationship, and participants may have been reluctant to list other network members.

Notwithstanding these limitations, this study identifies both structural and functional characteristics of egocentric networks that are associated with recent testing for syphilis. Network approaches to testing and treatment can promote positive social norms around testing and leverage for hard-to-reach populations. Public health campaigns should include messaging about STI testing as well as PrEP and utilize widespread outreach to increase uptake and messaging about testing for STIs. Providers should equip patients with knowledge and training, so their patients can inform and encourage their social networks to seek HIV and STI testing. Through peer educator training, patients in PrEP care can serve as ambassadors to their networks about HIV and STI transmission, prevention, and treatment.

Disclosures

The authors report no real or perceived vested interests related to this article that could be construed as a conflict of interest. As with all peer-reviewed manuscripts published in JANAC, this article was reviewed by two impartial reviewers in a double-blind review process. The JANAC Editor-in-Chief handled the review process for the paper, and the Deputy Editor, Allison Webel, had no access to the paper in her role as Deputy Editor or reviewer.

Funding

This work was funded by the National Institutes of Health grant numbers R01DA031030 (PI Latkin), R01DA032217 (PI Latkin), and T32DA007292 (MPI Johnson and Maher). The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.

Author Contributions

All authors on this article meet the four criteria for authorship as identified by the International Committee of Medical Journal Editors; all authors have contributed to the conception and design of the study, drafted or have been involved in revising this manuscript, reviewed the final version of this manuscript before submission, and agree to be accountable for all aspects of the work. Specifically, using the CRediT taxonomy, the specific contributions of each author are as follows: conceptualization and methodology: O. Heidari, K. E. Tobin, S. Pollock, FY. Li, A. R. Webel, D. T. Duncan, and C. Latkin; formal analysis: O. Heidari, K. E. Tobin, and FY. Li; writing, original draft: O. Heidari, K. E. Tobin, and S. Pollock; and writing/revising: O. Heidari, K. E. Tobin, S. Pollock, FY. Li, A. R. Webel, D. T. Duncan, and C. Latkin.

Key Considerations

* Syphilis disproportionately affects Black sexual minority men (BSMM).

* Social network approaches are an effective means of health promotion and affecting behavior change.

* In analyses adjusting for salient individual characteristics, BSMM who had more sexual partners, and encouraged their social network to get a syphilis check, were significantly more likely to get tested for syphilis in the past year.

* Training individuals on sexual health topics and health access for their social network may lead to increased uptake of testing and behaviors that reduce risk.

* Providers should encourage their patients to engage their social networks to promote syphilis, HIV, and other STI testing

Acknowledgments

The authors thank their study participants and Roeina Love, Tonya Johnson, Denise Mitchell, Charles Moore, Marlesha Bates, and Joanne Jenkins for their assistance and support in data collection.

References