Introduction

Depression and other mental health disorders represent a large proportion of the global burden of disease and are the second leading cause of global disability [1, 2]. In the Russian Federation, this trend is consistent, if not magnified, among men. Here, the prevalence of depressive symptoms in men residing in urban areas has been reported as high as 23% in 1999–2000, as defined by a score of ≥16 on the Center for Epidemiological Studies Depression (CES-D) scale [3]. As of 2011, the suicide rate among Russian men was estimated at 35.1 per 100,000 persons—one of the highest in the world. [4] Global reports suggest that men who have sex with men (MSM) often have higher prevalence of depression than heterosexual men [5], and the little research that has been conducted in Russia on the topic suggests that such patterns may persist among MSM [6]. Analysis of HIV-infected MSM found that over one-third have reported symptomology indicative of probable depression and more than half of MSM in Moscow have reported hazardous or harmful levels of alcohol abuse that may be related to depression [7].

Across settings, research among MSM has shown that homophobia and social stigma may contribute to depression [8]. While less is known from the Russian Federation, research from similar epidemiologic settings in Central Asia has provided some insight, demonstrating that even where homosexuality has been decriminalized, stigmatization of homosexuality persists [9], and MSM who regularly practice the receptive role during anal intercourse experience increased levels of stigma [10]. Social stigma in Central Asia has been associated with reduced access to health and social services among MSM [9, 11], including key HIV prevention and treatment services [9]. Depressive symptoms among MSM in Central Asia has been explained in part by high prevalence of community-level stigma [9]. Research from other countries with similar contexts in which homosexuality is decriminalized but where there is no legal protection for MSM against discrimination [12] have demonstrated the association between increased levels of perceived and experienced stigma and depression among MSM [13, 14] and inconsistent use of condoms and depression among MSM [15]. Social support [16] and social cohesion [13] within the MSM community, however, appear to have a protective effect against depression among MSM in these countries.

Recent events within Russia provide context and rationale for understanding the factors associated with poor mental health outcomes among MSM. In June 2013, the Russian government issued the “anti-gay propaganda law,” a national ban on “propaganda for non-traditional sexual relations” that criminalized the sharing of information related to same sex practices. The law effectively legitimized public stigma and violence against lesbian, gay, bisexual, and transgender people (LGBT) in Russia [17, 18]. This came on the heels of similar laws in the City of St. Petersburg (March 2012) that would fine individuals up to 500,000 rubles (approximately $17,000 USD) for spreading “homosexual propaganda,” as well as other related legislation in multiple oblasts (administrative divisions) across the Russian Federation, and a widely praised draft law in neighboring Ukraine (October 2012) that would impose up to 5 years imprisonment for spreading “propaganda of homosexuality [19].” While Russia’s national anti-gay propaganda law was not passed until 2013, the St. Petersburg, other provincial laws, and Ukraine bill (subsequently revoked in 2015 after the conclusion of this study) [20] received heavy media attention at the time, prompted considerable political and social discussion around the acceptance of homosexuality in Russia, and built substantial support for a possible federal law in Russia. Subsequently, a poll conducted by the Pew Research Center in 2013 found that nearly three in four Russians believed that society should not accept homosexuality [21]. Political analysts have since suggested that Russian President Vladimir Putin’s stance that homosexuality is a threat to Russian society [17] is part of a political narrative that deliberately links non-traditional sexualities to an imperialistic West [18]. Given related literature [22] and current events in Russia, this research sought to assess the level of depression among MSM in Moscow, Russia, explore the relationship between sexual identity and depressive symptoms among MSM in Russia, and explore the role of environmental changes in acceptance of homosexuality in the Russian Federation as it relates to depression.

Methods

The data for this analysis was gathered through a cross-sectional, NIH-funded study conducted in Moscow, Russia, between October 2010 and April 2013. The research was a collaborative effort conducted jointly by Johns Hopkins University and AIDS Infoshare, a local non-governmental organization that provides anonymous HIV testing and counseling and STI screening and treatment in an LGBT-friendly context in Moscow. All sampling, data collection, and biologic testing methods have been previously described [7, 23].

Participants

Inclusion criteria required participants to be fluent in Russian, assigned male at birth, aged 18 years or older, currently reside in Moscow, report anal sex with a man in the past 12 months, and provide verbal informed consent. Participants were recruited via respondent-driven sampling (RDS), a chain recruitment method that is used to reach hidden or hard-to-reach populations [24]. Participants were provided 1000 rubles (approximately $40 USD) as reimbursement for their participation in the study and 500 rubles ($20 USD) for each peer recruited into the study. Thirty-one waves of recruitment produced a sample of 1367 participants for the analysis [23].

Measures

Informed by formative research [25], cognitive testing [26], and pilot testing methods, the structured survey was designed to gather information on sociodemographic characteristics, sexual practices and identity, substance use, experienced and perceived stigma, and mental health symptoms. Depression symptoms were the outcome of interest and sexual identity, stigma, and other known correlates [13, 15, 27–30] of depression were covariates of interest for this analysis. Measures of sexual identity and practices were adapted from the US CDC’s National Health Behavior Survey [31] and included questions about sexual identity (“heterosexual,” “homosexual,” “bisexual,” or “other”) and about the sex of the respondent’s sexual partners in the last 12 months (male or female).

The CES-D scale [32] was used as a screening tool for current (past 7 days) depressive symptomology (Cronbach’s α 0.889). The CES-D score was stratified into three categories: “unlikely to be depressed” (CES-D <16), “possibly depressed” (CES-D 16–22), and “probably depressed” (CES-D ≥23) [32]. Given that 36.7% endorsed a score of probably depressed (n = 505) and 11% of the sample scored in the possibly depressed range (n = 121), coupled with our interest in understanding the factors that were correlated with greater probability of depression [33], scores were transformed to a binary variable: “unlikely or possibly depressed” (CES-D <23) and probably depressed (CES-D ≥23).

A five-item scale was used to identify recent (last 12 months) experienced stigma (Cronbach’s α 0.713). This was a subscale of a larger stigma scale that had been developed from qualitative research [25], underwent cognitive testing [26] among MSM participants for further refinement, and were reviewed by MSM participants to assess face validity. The experienced stigma subscale included experience of public mocking, poor treatment by family members, experience of physical assault, being forced to move accommodation, and experiencing poor treatment within healthcare facilities due to sexual preferences. A positive indication for any one of these five variables was defined as a recent experience of stigma. The alcohol use disorders identification test (AUDIT), a validated scale in Russia, was included in the survey for identification of alcohol use in the last 12 months (Cronbach’s α = 0.80) [34]. AUDIT scores were stratified into three groups: a score <8 representing abstinence or low use, 8–15 “hazardous use,” and ≥16 “harmful use,” including those with potential alcohol dependency (score of 20 or higher), according to WHO categories [35].

Statistical Analysis

Survey data were analyzed using the STATA version 14 statistical software (Stata Corporation, College Station, USA). Depression, measured by the CES-D score, was the primary outcome of interest. Descriptive analysis was performed using Pearson’s chi-squared test. Sensitivity analyses were performed first using binary logistic analysis with unlikely to be depressed and a combined “possibly or probably depressed” and then using ordered logistic analysis with all three categories. A final sensitivity analysis was performed to test whether the model results changed with the inclusion of bisexual behavior (reporting both male and female sex partners) in place of bisexual identity.

Data collected via RDS recruitment are often weighted during statistical analysis to provide population prevalence estimates. Population estimates for the population characteristics and depression were calculated using the RDS-II estimator (Volz-Heckathorn) and bootstrapping with 1000 iterations to produce 95% confidence intervals (95%CI) [36]. Both crude sample proportions and RDS-weighted population prevalence estimates are provided in Appendix.

On exploratory analysis, the linear relationship between CES-D scores and time in which participants were surveyed was evaluated given increasing homophobia and the ultimate passage of the propaganda law. This relationship was significant, thus, we included the date of when participants completed the survey as a covariate in the model. This was transformed to a binary variable with March 7, 2012 as a breakpoint: the date in which the governor of St. Petersburg signed a law banning the sharing of homosexual propaganda [37]. While our study took place in Moscow City, political and media attention to the St. Petersburg law created an atmosphere of increased violence and discrimination against the LGBT population across Russia. The category for participation in the survey after the passage of the law is also inclusive of other key time periods that signify the deteriorating climate for LGBT populations, including the passage of the draft Ukraine law that received wide support and discussion in Russia (October 2012) and the date when the anti-gay propaganda law was introduced to the Russian Parliament (January 2013) [38]. Our model tests for an increase in the odds of probable depressive symptoms before and after the St. Petersburg law as a binary variable and as an interaction term with experienced stigma.

Variables known to be associated with depression symptoms (income, sexual identity, experience of stigma, social connectedness, alcohol use) [13, 15, 27–30] or with a p value of p < 0.10 in bivariate analysis were included in the multivariable model. The Furnival-Wilson leaps-and-bounds algorithm was used to perform variable selection on a wide array of models with depression as the outcome of interest and socioeconomic, demographic, experiential, and behavioral factors as predictor variables. Akaike’s information criteria were used to select the best model, and multiple logistic regression was used to obtain adjusted odds ratios (AOR) for the variables in the selected model. Collinearity for the model was checked using variance inflation factors (VIFs). While the VIF for the interaction between survey date and experience of stigma was 2.88, all other VIFs were well under 2.0, indicating that collinearity is not of concern. Pearson’s goodness-of-fit test was used to check model fit.

Ethical Review

The study was approved by both the Ethics Committee of the State Medical University, IP Pavlov, St. Petersburg, Russia, and the Johns Hopkins Bloomberg School of Public Health Institutional Review Board, Baltimore, Maryland.

Results

A total of 1376 MSM with a median age of 30 years [IQR 24–36] were enrolled in the study, 85.7% of whom were born within Russia. Over half (55%) identified as homosexual, 42.9% identified as bisexual, and 2.1% identified as other sexual identities, including heterosexual. By contrast, 64.6% reported only male sexual partners, while 35.4% reported both male and female partners in the last 12 months. Thirty-six percent reported experiencing stigma due to their sexual preferences. Over one-third (36.7%) qualified as probably depressed (n = 505), 11% as possibly depressed (n = 151), and 52.3% unlikely to be depressed according to the CES-D (n = 720).

Table 1 presents the participant demographic and behavioral characteristics stratified by depression symptomatology. Appendix Table 1 presents the RDS-weighted estimates stratified by depression symptomatology. Those with symptoms indicating probable depression (CESD ≥23) were more likely to have full-time employment (59.7%) than those with symptoms indicating unlikely or possible depression (CESD <23; 51.4%, p = 0.03). MSM in the probable depression group were more likely to identify as homosexual (64.7% vs. 49.5, p < 0.001) and less likely to report feeling comfortable with their sexual orientation (86.5 vs. 93.2%; p value <0.001) than those in the unlikely/possible depression group. Those with probable depression scores reported higher levels of experienced stigma (62.7 vs. 39.4%; p value <0.001) than those who were possibly/unlikely depressed. A larger proportion of those with probable depression had been detained in jail or prison (8.6%) than those with unlikely/possible depression (4.6%; p value <0.01). A greater proportion of those in the group with probable depression participated in the survey after the “propaganda law” was signed into law in St. Petersburg (61.4%) compared to those in the unlikely/possibly depressed group (48.7%; p value <0.001).

Table 1 Demographic and behavioral characteristics of MSM in Moscow, Russia, by depression symptomatology from 2010 to 13 (N = 1376)
Full size table

Table 2 displays the bivariate and multivariable logistic regression analyses of sexual identity, stigma, passage of the anti-gay propaganda laws, and other correlates with probable depression. After controlling for other confounders, the odds of probable depression was 29% lower among those identifying as bisexual (AOR 0.71 95%CI 0.52–0.97; p = 0.03) as compared to those reporting homosexual identity. The adjusted odds of depression was nearly twice as high for those who reported experiencing stigma in the last 12 months (ref: little/no stigma; AOR 1.75; 95%CI 1.20–2.56; p < 0.01). Those who reported discomfort with their sexual identity had almost sixfold increased odds of depression relative to those who reported feeling comfortable with their sexual identity (AOR 5.85; 95%CI 2.71–12.64; p < 0.001). A history of jail or detention was associated with a nearly twofold increased odds of probable depression (AOR 1.78; 95%CI 1.00–3.17; p = 0.05) compared to those who did not have a history of detention. Probable depression was also 35% lower for those with five or more sexual partners in the last 6 months (ref: ≤1 partner; AOR 0.54; 95%CI 0.37–0.81; p < 0.01).

Table 2 Crude and adjusted relative odds of depression (N = 1376)
Full size table

Prior to including the interaction between time and experienced stigma in the multivariable model, the adjusted odds of depression for those who were surveyed after the signing of the St. Petersburg propaganda law was 1.5-fold greater than that of participants who were surveyed before the approval of the law (AOR 1.65, 95%CI 1.23–2.22; p = 0.001). After including the interaction between time and experience of stigma in the final model, the adjusted odds of depression associated with the survey date was no longer significant (AOR 1.34, 95%CI 0.95–1.91; p = 0.10). The interaction itself, however, was significant, and computed estimates for the combination with the interaction were highly significant. In this case, among people with experienced stigma, depression increased twofold after the passage of the St. Petersburg law (ref: participation prior to the law; AOR 2.24; 95%CI 1.49–3.39; p < 0.001). Finally, in the period after the law was passed, the odds of probable depression among people with experienced stigma were almost threefold that of those without experienced stigma (AOR 2.92; 95%CI 2.02–4.24; p < 0.001).

A sensitivity analysis was performed with the same multivariable model but including the variable measuring behavioral bisexuality in terms of the gender of reported sex partners in the last 12 months (in lieu of bisexual identity). Here, the association became stronger as the adjusted odds ratio dropped from 0.71 (95%CI 0.52–0.97, p = 0.03) in the previous model to 0.58 (95%CI 0.42–0.81; p < 0.01).

Discussion

Depression is common in the Russian Federation, and particularly so for MSM. Over a third of MSM in this sample qualified with “probable depression” according to the CES-D scale, and nearly half scored high enough to qualify as at least “possibly depressed.” These estimates for MSM in Russia exceed the general estimates of 23% for possible depression (CES-D cutoff of 16) in men in urban areas of Russia [3]. While methodological differences and temporal effects may play a role, differences may also be explained, in part, by the high prevalence of experienced stigma and its strong relationship with depressive symptoms.

Evidence has shown that sexual identity may be a correlate of depression and other mental health outcomes. In this study, bisexual identity appears to provide a protective effect against depression among MSM in Russia. Discomfort with sexual identity was also strongly associated with depressive symptoms. The protective relationship between bisexual identity and depression contradicts the general body of research, which predominantly originates from the USA, that has demonstrated increased risk for poor mental health outcomes with bisexual identity and similarly high risk for poor social wellbeing compared to those who identify as homosexual or heterosexual [22].

One reason for this surprising discrepancy may be the context of increasing and state-sanctioned homophobia in Russia. In a country that has criminalized information related to same sex behaviors, and ultimately criminalized LGBT visibility, the political and social environment has become increasingly hostile towards MSM. The discourse around the anti-gay movement has promoted traditional morals, masculinity, and traditional sexual relations and has been associated with increased violence towards LGBT populations and may decrease individual comfort with one’s sexual identity [18]. It is possible that endorsing a level of opposite-sex attraction in addition to same-sex attraction allows a degree of psychological refuge from feelings of exclusion, self-stigma, and experiences of violence. Discomfort with sexual identity may further be measuring underlying, internalized homophobia, which has demonstrated significant impact on mental health, including anxiety and depression [39].

The experience of detention in jail or prison also had a strong relationship with depressive symptoms. While we did not collect information about the reason for imprisonment, it would be interesting to know if sexual identity played either a real or perceived role in the justification for imprisonment.

Several factors limit the inferences of this study. The study was cross-sectional, limiting our ability to determine the temporal relationship between sexual identity and depression and prevents causal association. However, the use of recall periods for experiences of stigma (past 12 months) and depression (past 7 days) allows some interpretation of temporality. We utilized a tool to measure depression symptoms rather than depression diagnosed by a mental health professional, which may potentially misclassify participants; however, the CES-D has been well-established and validated in the Russian context. Enrollment and testing methods, criteria for study inclusion, and financial incentives may have influenced our sample. It is also possible that RDS did not reach particular subgroups within the MSM community. However, this study is unusual in its success of recruiting a large number of MSM (31 waves and over 1300 MSM) in a highly stigmatized setting. Despite potential limitations, RDS seems to have been one of the most productive pathways of recruiting MSM in this context [23].

Our results show that homosexual identity, experience of stigma, and experience of imprisonment are all associated with depressive symptoms in MSM in Russia. The passing of anti-gay propaganda laws in the region in 2012 and in the Russian Federation in 2013 [40], which effectively criminalizes information related to homosexual and bisexual identity, further modifies the relationship between stigma and depression among MSM. Because social support has been shown to be protective against depression in MSM in this context [41], interventions that build social support in the MSM community in Russia and enable access to non-stigmatizing mental health services may help to combat the negative effects of stigma and help protect against depression in this population.