An Application of Syndemic Theory to Identify Drivers of the Syphilis Epidemic Among Gay, Bisexual, and Other Men Who Have Sex With Men

Abstract

Background 

We applied syndemic theory to explore the degree to which syndemic conditions explain the syphilis epidemic affecting Canadian gay and bisexual men who have sex with men (GBMSM).

Methods 

Data from a national survey comprising 7872 GBMSM were analyzed using multivariable logistic regression to measure associations between recent syphilis diagnosis (RSD; in previous 12 months) and the following variables: (1) sociodemographic information (sexuality, HIV status, age, income, ethnicity, relationship status), (2) antigay stigma (bullying, physical violence, sexual violence, career discrimination, health care discrimination), (3) syndemic conditions (suicidality, intimate partner violence, depression, illicit substance use, binge drinking), (4) sexual behaviors, (5) health care discrimination, and (6) the cumulative count of antigay experiences and syndemic conditions.

Results 

Three percent (n = 235) of GBMSM surveyed reported an RSD. Men were more likely to report an RSD if they were HIV positive (adjusted odds ratio [AOR], 6.27; 95% confidence interval [CI], 4.66–8.43). Recent syphilis diagnosis was also positively associated with career discrimination, health care discrimination, substance use, and intimate partner violence. Furthermore, prevalence of RSD increased with each additional form of stigma or syndemic condition. The odds of reporting RSD was 5.2 (95% CI, 1.0–25.9) times higher for men who reported experiencing all 4 forms of antigay stigma compared with those who reported no stigma, after adjusting for sociodemographics. Similarly, the adjusted odds of reporting RSD was 12.2 (95% CI, 2.0%–74.8%) times higher for GBMSM experiencing 5 syndemic conditions compared with those reporting no syndemic conditions.

Conclusions 

Evidence from this large cross-sectional study suggests that the Canadian syphilis epidemic among GBMSM is being driven by a syndemic constituted by multiple social and psychological conditions. Interventions addressing specific psychosocial health outcomes that increase the risk for syphilis should be developed and integrated within targeted sexual health services and syphilis prevention initiatives.

In Brief

This study used syndemic theory to investigate associations between sexual stigma, syndemic conditions, and recent syphilis diagnoses among Canadian gay, bisexual, and other men who have sex with men.

After years of decline, syphilis rates have risen in many North American cities since the turn of the century and continue to increase significantly in many jurisdictions.^1,2 The syphilis epidemic in urban North America is concentrated among gay, bisexual, and other men who have sex with men (GBMSM), with GBMSM living with HIV disproportionately impacted.^1,2 Efforts to address the syphilis epidemic have largely focused on extending testing and treatment, enhancing partner notification, and encouraging GBMSM to modify their sexual behaviors.³ Despite these efforts, syphilis incidence continues to increase.

One potential way to expand syphilis prevention programs beyond biomedical and behavioral approaches is to address the social conditions and co-occurring health issues that may contribute to the syphilis epidemic.⁴ Syndemic theory offers a cohesive approach to investigate overlapping epidemics, whereby health problems, such as sexually transmitted infections (STIs), HIV, and mental illness, tend to co-occur and create mutually reinforcing clusters of epidemics among marginalized populations.⁵ These synergistic epidemics, or syndemics, are theorized to be the result of social inequities (e.g., such as those related to class, gender, sexuality) and unfavorable structural factors (e.g., inadequate health care resources, unemployment, or other poor institutional supports).⁵ Syndemic theory was first applied to GBMSM by Stall and colleagues⁶ in the early 2000s. In a large sample of urban GBMSM in the United States, investigators found that common psychosocial risk factors for HIV—polydrug use, depression, childhood sexual abuse, and intimate partner violence—were highly correlated and additively increased HIV vulnerability.⁶ By revealing the relationships between these syndemic conditions and elucidating their role in determining the distribution of HIV infection within populations, the work of Stall and colleagues challenged the way in which epidemiologists, medical anthropologists, and health researchers considered these factors in isolation.

Since the original study by Stall and colleagues, syndemic conditions, with a focus on HIV or HIV-related behavioral outcomes, have been examined in a variety of GBMSM populations with relatively consistent findings: GBMSM experience multiple psychosocial issues related to social and structural conditions that have a cumulative effect on HIV transmission risk.^7–13 Little attention, however, has been given to integrating syndemic theory in the study of other STIs, including syphilis. Some researchers have noted that syphilis overlaps with other epidemics and social problems among GBMSM,^14–16 suggesting that syndemic theory may inform novel public health approaches to end this epidemic.

The objective of the current study was to apply syndemic theory to explore the degree to which syndemic conditions explain the syphilis epidemic affecting Canadian GBMSM. Our goal is to stimulate discussion about developing innovative and targeted interventions for effectively addressing this epidemic. We draw on data from a large cross-sectional survey of Canadian GBMSM to specifically examine relevant factors suggested by syndemic theory in relation to recent self-reported syphilis diagnosis: (1) sociodemographic characteristics, (2) experiences of sexual stigma, (3) syndemic conditions such as mental health difficulties and substance misuse, (4) sexual behaviors, and (5) health care discrimination.

METHODS

Data for this study were derived from Sex Now, a community-based serial cross-sectional survey of Canadian GBMSM stewarded by the Community-Based Research Centre for Gay Men's Health. Gay, bisexual, and other men who have sex with men were recruited through promotion within sex-seeking Web sites and applications, community groups, gay media, social media, and e-mail to participants from previous cycles of the survey. Surveys were completed anonymously, online, in one of Canada's 2 official languages (English or French). The survey and its methodology have been described in detail elsewhere.⁷ Briefly, the survey collected data on a broad set of community-identified issues, including the following: sexual behaviors, sexual health, relationships, health care experiences, working conditions, community participation, social support, mental health, and experiences of sexual stigma and discrimination. The present article reports data from Sex Now 2014–2015, which were collected from November 2014 to April 2015.

Conceptual Model

The research team met to review seminal and emerging theoretical writings on syndemic theory^5,17 as well as current empirical evidence^7–13 to develop a conceptual model reflecting possible syndemic-related drivers of the syphilis epidemic among GBMSM (Fig. 1). Consistent with syndemic theory,^5,17 the study team theorized that the accumulation of sexual stigma stressors lead to the development of syndemic conditions and that these conditions are mutually reinforcing, that is, working together to increase risk for syphilis among GBMSM. Experiences of health care discrimination have not been included in previous syndemic studies; however, it is hypothesized that barriers to health care access, including discrimination, can exacerbate existing health conditions and prevent GBMSM from accessing necessary health services.¹⁸ Finally, we hypothesized that GBMSM experiences of syndemic conditions are also influenced by their social positions and characteristics (e.g., related to race, education, income, and geography).

Figure 1: Conceptual framework for a study of associations between antigay stigma, syndemic conditions, and syphilis infections among Canadian gay and bisexual men.

Measures

Recent History of Syphilis Diagnosis. The main outcome (i.e., dependent variable) used for this analysis was self-reported syphilis diagnosis in the previous 12 months.

Participant Sociodemographics. Respondents indicated their age, sexual orientation, ethnicity, highest education achieved, income, and current community environment. In addition, respondents reported the results of their last HIV test result (positive, negative, or unknown for those untested), which was recoded as HIV positive or negative/unknown (Table 1).

TABLE 1: Sociodemographic Factors and Associations With Recent Syphilis Diagnosis (Last 12 Months), Among 7872 Canadian Gay and Bisexual Men

Sexual Stigma. Respondents were asked to report if they experienced the following within the past 12 months related to their sexuality: (a) verbal harassment, (b) cyber bullying, (c) physical violence, and (d) career discrimination. These measures have been found to be associated with syndemic conditions in previous iterations of the Sex Now survey.⁷

Syndemic Conditions. Respondents were asked to report if they experienced the following in the last 12 months: (a) use of one or more party drugs: cocaine, crystal meth, ecstasy, gamma-hydroxybutyrate, ketamine, or crack, hereafter referred to as party drug use; (b) being threatened, emotionally abused, or physically assaulted by a sexual partner, boyfriend, or partner, hereafter referred to as intimate partner violence; (c) drinking 5 or more alcoholic drinks in one sitting, more than 1 time per week, every single week, hereafter referred to as frequent binge drinking; (d) suicide thoughts or attempts, hereafter referred to as suicidality; and (e) talking to a health care provider about depression or anxiety, hereafter referred to as depression or anxiety.

Sexual Risk Behaviors. Respondents reported whether they had any condomless anal intercourse in the past 12 months. In addition, respondents were asked to report the number of sexual partners (with whom they had any type of sex) in the last 12 months as a continuous variable, which was recoded to a binary variable for our analysis (<20 sex partners and ≥20 sex partners).

Health Care Discrimination. Respondents were asked to report whether they encountered antigay discrimination within the health care system (i.e., clinic, laboratory, hospital) within the last 12 months.

Analysis

Analysis was guided by a conceptual model (Fig. 1) and included the following series of bivariate and multivariable logistic regression models, using recent syphilis diagnosis as the outcome. First, we estimated crude bivariate associations between recent syphilis diagnosis and individual variables within the following categories: (1) sociodemographics, (2) stigma measures, (3) syndemic conditions, (4) sexual behaviors, and (5) health care discrimination.

Second, separate multivariable models were estimated for each of the 5 categories mentioned previously. Multivariable models included only one category at a time because these categories were hypothesized to be sequential in a casual pathway, as described by the theory of syndemic production¹⁷ and shown in our conceptual model (Fig. 1).

Third, the number of sexual stigma indicators and the number of syndemic conditions reported by each participant were summed to form a count of sexual stigma (ranging from 0 to 4) and count of syndemic conditions (ranging from 0 to 5), given that sexual stigma and syndemic conditions were hypothesized to have cumulative effects. Logistic regression was again used to measure the relationship between counts of sexual stigma and syndemic conditions and recent syphilis diagnosis.

Fourth, we examined statistical evidence of additive interaction between syndemic conditions. Additive scale interaction suggests a departure from additivity of individual risk factors or conditions and is generally considered the most appropriate scale in public health contexts.¹⁹ In the present study, evidence of additive interaction suggests that the combination of multiple syndemic conditions is associated with a higher prevalence of recent syphilis diagnosis than would be expected in the absence of an interaction. We assessed interaction in 2 ways. First, we calculated the observed and expected prevalence of recent syphilis diagnosis by count of syndemic conditions, where expected prevalence was estimated for each count at least 2 using the following formula:

where PD is prevalence difference and z is count of syndemic conditions. Second, we estimated the relative risk due to interaction (RERI), with 95% confidence interval (CI), using the methods described by Knol and VanderWeele.¹⁹

All multivariable models were adjusted for the participants' sociodemographics: age, sexual orientation, ethnicity, highest education achieved, income, living environment, and HIV status. Associations were considered statistically significant if the 95% CI did not include the null association odds ratio of 1; RERI was considered statistically significant if the 95% CI excluded 0. Analyses were conducted using IBM SPPS version 23.

Ethics

The independent ethics committee of the Community-Based Research Centre for Gay Men's Health reviewed and approved this project.

RESULTS

Most of the 7872 participants were white, 30 years or older, educated, living in an urban setting, and gay (Table 1). Three percent (n = 235) of participants reported a syphilis diagnosis within the previous 12 months.

The largest association between sociodemographic variables and recent syphilis diagnosis was observed for HIV status (Table 1). With the exception of verbal violence, all sexual stigma variables were positively associated in bivariate analysis, but only career discrimination remained significant after adjusting for other sexual stigma (Table 2, model A). Those who reported experiencing 3 or 4 forms of stigma had significantly higher odds of syphilis compared with those with no stigma (Table 2, model B). All syndemic conditions were positively associated with syphilis in bivariate analysis, with the exception of binge drinking. Only party drug use and intimate partner violence remained significantly related to syphilis after adjusting for other syndemic conditions (Table 2, model C). Those with 2 or more syndemic conditions were at significantly increased odds of reporting a syphilis diagnosis compared with those reporting no syndemic conditions (Table 2, model D). Condomless anal intercourse was significantly associated with syphilis, as was having 20 or more sexual partners (Table 2, model E). Last, health care discrimination within the past 12 months and recent syphilis diagnosis were significantly associated (Table 2, model F). Evidence of public health interaction was detected on the additive scale at counts of syndemic conditions of 3 or more (Fig. 2). The RERI for cumulative count of syndemic conditions of 3 or more was 2.31 (95% CI, 0.55–4.07), further supporting the statistical significance of syndemic interaction.

TABLE 2: Associations Between Sexual Stigma, Psychosocial Issues, Health Care Discrimination, and Recent Syphilis Diagnoses, Among 7872 Canadian Gay and Bisexual Men

Figure 2: Observed versus expected prevalence of syphilis diagnosis (past 12 months) by number of co-occurring psychosocial problems (syndemic conditions).

DISCUSSION

Our study provides quantitative evidence to support that syndemic conditions are associated with syphilis diagnosis among GBMSM and that these conditions interact on an additive scale to increase the risk for syphilis beyond the rate expected if the conditions were not synergistic. Specifically, being HIV positive, experiencing career discrimination, substance abuse, intimate partner violence, and experiencing health care discrimination in the past 12 months were all significantly associated with syphilis (Tables 1 and 2). Notably, the greatest associations with syphilis diagnosis were found with cumulative counts of sexual stigma and syndemic conditions, further supporting the additive nature of these factors in relation to syphilis risk (Table 2). These results illuminate the potential for syndemic theory to transform the way public health researchers and practitioners theorize, study, and intervene to interrupt the syphilis epidemic.

The current study offers the following specific insights into the relations between an array of social and psychological factors and recent syphilis diagnoses among Canadian GBMSM. First, the finding that a cumulative count of experiences of antigay stigma was associated with recent syphilis diagnosis suggests that interventions to reduce exposure to sexual minority stress may also help to respond to concentrated epidemics of STIs, including syphilis. These interventions have predominantly been focused on schools (e.g., antibullying policies and implementation of gay-straight alliances)²⁰; however, given that most syphilis occurs among adults, strategies to address cumulative experiences of antigay/sexual stigma in clinical and community settings serving adults are needed.² How stigma and social inequities are operating is likely to be complex, especially if sexual stigma intersects with other forms of stigma (e.g., racial, age, and education), and in need of further investigation.

We also found evidence that substance use and intimate partner violence were associated with syphilis. Other researchers have found that illicit substance use is associated with sexual risk behaviors²¹ and syphilis²²; our results support this association but suggest that the impact of illicit drugs can be amplified by co-occurring syndemic factors and as such should not be studied in isolation. The association between intimate partner violence and syphilis has been studied in the context of pregnant women²³; however, it has rarely (if ever) been studied among GBMSM. One hypothesis for this association is the lack of services addressing intimate partner violence in the GBMSM community, which may disempower GBMSM victims, ultimately making them reluctant to access sexual health and health services. This latter hypothesis is reinforced by our finding that experiences of health care discrimination were positively associated with syphilis.

Those who experienced health care discrimination had more than 4 times the odds of reporting a recent syphilis, after accounting for sociodemographic characteristics. Although this study did not collect details about these experiences (e.g., specific health care settings), other researchers have noted that discrimination occurs within some sexual health and primary care clinics where health care providers draw on negative stereotypes when evaluating the sexual health history of GBMSM.²⁴ Discrimination within the health care system may be explained by the lack of educational supports available to ensure that health care providers understand how best to provide culturally competent care to sexual minorities. Dedicated education modules on sexual minority health are rare within North American medical and nursing school settings²⁵ and a recent survey of medical students in Canada reported that instructors often expressed negatives biases toward GBMSM.²⁶ Thus, improved cultural competency supported through the provision of explicitly included educational content is required to ensure that sexual minorities are not deterred from accessing health care services—including STI and sexual health resources—when they are needed.

Our results suggest a few specific targets for syphilis prevention and treatment among GBMSM, while generally supporting holistic approaches to specialized sexual and psychosocial health care for GBMSM. Most approaches to GBMSM syphilis prevention to date have sought to raise awareness among GBMSM about the rise in syphilis in isolation from other STI and psychological and social health concerns.³ However, our results show that syphilis is consistently patterned in association with multiple health issues, suggesting that GBMSM may themselves not prioritize syphilis as the predominating health “threat,” particularly because syphilis is curable. In addition to campaigns to raise awareness, we suggest that health care providers be supported in better identifying and attending to within-group risk for syphilis, in association with the syndemic conditions highlighted in this study: depression, anxiety, suicidality, substance use, and intimate partner violence. These issues should be discussed using supportive, gay/bisexual-positive, and nonpathologizing approaches. One approach, supported by the associations between sexual stigma and syphilis diagnosis observed in our study, is for health care providers working with GBMSM to acknowledge the high rates of both historical and contemporary experiences of antigay/bisexual stigma and to express to patients that they care about their health and well-being in this context.

Sexually transmitted infection clinics are specific environments that integrate culturally sensitive and responsive psychosocial health care for GBMSM at risk for syphilis and other STI. Specializing in GBMSM sexual health care—and often regarded as “safe” places for GBMSM to discuss their sexuality—these services offer low-barrier access wherein clients are not required to provide identification, thus allaying privacy and confidentiality concerns. Other studies have recommended this type of service integration within STI clinics, noting that STI clinic clients frequently report mental health, substance use, and other social or general health concerns when presenting for STI screening or treatment, although few of these studies have specifically examined the needs of GBMSM clients.^27,28 Several STI clinics targeting GBMSM have integrated social and mental health services. These expansions require additional resources and training for clinic staff to effectively deal with specific domains including mental health assessment and care; however, our results suggest that such investments might yield lower STI rates by reducing reinfection and increasing primary prevention in ways that advance men's sexual and mental health.

The present findings should be considered bearing in mind the following limitations. First, the online survey's Internet recruitment strategies may have oversampled GBMSM who are connected to the GBMSM community, educated and affluent, potentially leading to underreporting of syndemic conditions. Second, the study relies on self-reporting of syphilis, and as such, we may have missed some cases, particularly men who have syphilis but have yet to get tested. Third, our ability to infer a direct causal association between syndemics and syphilis is limited by the cross-sectional nature of the data.

Overall, the present study found evidence to support that syndemics is a useful theory to advance knowledge of the syphilis epidemic affecting GBMSM. Future work is needed to confirm the associations revealed in this study and to better determine how these syndemic conditions are produced. Particularly, intervention research is needed to evaluate the most effective and culturally appropriate policies and services to address these syndemic conditions. As other syndemics researchers have pointed out,²⁹ innovative solutions may be illuminated if researchers shift their gaze from GBMSM problems to their assets and resiliencies. That the resilience of GBMSM is a relatively “untapped resource” in public health also demands attention.²⁹ As such, the authors propose that future application of syndemic theory to syphilis intervention and prevention efforts integrate resiliency in thoughtfully considering how those community-based resources might best be mobilized.