Introduction

The HIV epidemic in the Russian Federation (Russia) is among the most rapidly growing globally [1]. Over the past two decades, HIV prevalence in Russia has increased from occasional transfusion-related cases to over 1 million people living with HIV in the country [2], propelled in large measure among people who inject drugs (PWID) [3, 4]. In such concentrated epidemic where HIV transmission occurs primarily among key populations and their contacts, personal risk behaviors such as unsafe injection drug use can be in part driven by its context and structural environment. The latter includes legal and social issues [5] and conceivably also stigma and discrimination [6]. These contextual and structural factors produce environments that increase HIV risks for PWID and their sexual partners [7]. Stigma, defined as the dehumanization and social exclusion of individuals based on their undesirable social category [8], is an important part of the risk environment of PWID in Russia [9]. Stigma related to HIV infection is a well-documented phenomenon in populations where HIV is transmitted sexually [10], but it has been less studied among people with addiction and in settings where HIV is transmitted commonly through unsafe injections [11].

HIV-related stigma is common in Russia [12], decreases employment opportunities and quality of life [13,14,15,16,17], and is a significant barrier to HIV prevention and treatment [17, 18]. PWID experience stigmatizing attitudes related to HIV and other identities (e.g., substance user), from law enforcement officers, health care providers, families and communities. [19]. As a consequence of anticipated and experienced stigma, people living with HIV (PLWH) often choose not to disclose their status to others (e.g., to their employer, health service providers, family, or sex partners) [20, 21]. A UNAIDS report estimated that 10% of PLWH in Russia were denied health services, 5% denied family planning services, and 3% were denied employment [22], indicating that stigma might contribute to the marginalization of PLWH.

In addition to HIV stigma, unhealthy alcohol use is a personal risk factor that adds to the HIV risk environment [23, 24]. Alcohol consumption is high in Russia and associated with the highest rate of alcohol-attributable mortality in the world [25, 26]. A study conducted among PWID in St. Petersburg, Russia, found high prevalence of binge drinking (64%) and HIV risks (88% of documented sex unprotected) in this population [27]. Another Russian study, among PLWH, reported that hospitalized, alcohol-dependent patients consumed a mean of 28 g of alcohol per day (the equivalent of 2 standard drinks per NIAAA definition) in the past month [28]. Excessive alcohol consumption is widespread in Russia and is closely linked with excess mortality particularly among Russian men [29], but also among PWID [30]. We assume that even in an environment where drinking is normative, excessive alcohol use might be associated with self-stigma [31], especially when alcohol consumption is disorderly and apparent to others, such as among people with more severe alcohol use disorders [32].

Currently, limited literature describes the potential link between HIV-related stigma and unhealthy alcohol use among PLWH. In a study of HIV-positive mothers, those who perceive greater HIV-related stigma reported more alcohol use [33]. Since life stressors are associated with substance use and heavy drinking, an analogous association could be hypothesized between HIV-related stigma and unhealthy alcohol use [34].

Given the constellation of highly prevalent personal risk factors such as unhealthy alcohol use and structural risk factors such as HIV stigma in Russia, we sought to understand their relation in affected populations. This study’s aims were thus to [1] describe HIV stigma among people living with HIV who report a history of unhealthy alcohol use, and [2] examine associations between recent unhealthy alcohol use (i.e., alcohol dependence and drinking at risky levels) and HIV-related stigma in this Russian population.

Methods

We conducted a secondary analysis of data collected from 700 people with HIV infection and unhealthy alcohol use in St. Petersburg, Russia, enrolled in the HERMITAGE (HIV Evolution in Russia—Mitigating Infection Transmission and Alcoholism in a Growing Epidemic) study. This randomized controlled secondary HIV prevention trial implemented a behavioral intervention to reduce high-risk sexual activity among HIV-infected risky drinkers. We assessed HIV stigma and alcohol use at baseline and 12-month follow-up in all participants, i.e., in individuals receiving the intervention and those who were control subjects. The HERMITAGE Study recruited participants based on convenience sampling.

As previously reported [35], we recruited a cohort of HIV-infected risky drinkers, who had reported unprotected sex in the previous 6 months, from four HIV and addiction clinics, as well as at a needle exchange program which referred to the clinics. Research associates approached potential participants, assessed eligibility, offered participation, obtained informed consent and conducted the assessments.

Inclusion criteria were the following: age 18 years or older; HIV infection; reported unsafe sex (anal or vaginal sex without a condom) in the past 6 months; reported risky drinking levels in the past 6 months as defined by the U.S. National Institute on Alcohol Abuse and Alcoholism (NIAAA) (Men: >4 drinks/day or >14 drinks/week; Women: >3 drinks/day or >7 drinks/week); provision of contact information; a stable address within 150 km of St. Petersburg; and the ability to provide informed consent. Exclusion criteria were anticipated incarceration or the expressed intent to conceive a child.

Data Collection

We conducted all baseline and follow-up interviews in Russian. Baseline data were collected in three steps, a face-to-face interview with a research associate and a self-administered questionnaire for particularly sensitive questions (including stigma and depressive symptoms). Participants were compensated the equivalent of US $50 for the assessment and received 30 condoms.

The parent trial is registered at ClinicalTrials.gov as NCT00483483. Both the Institutional Review Boards of Boston Medical Center and St Petersburg Pavlov State Medical University approved this study.

Measures

We measured HIV stigma using the abbreviated Berger HIV stigma scale [36]. In the original sample of HIV-positive individuals, Berger and colleagues identified personalized stigma, disclosure concerns, negative self-image and concerns with public attitudes toward people with HIV [37]. Personalized Stigma identifies negative consequences of having HIV, such as experiences with rejection. This measure does not does not distinguish experienced from anticipated stigma. Disclosure Concerns are measured with items on control and sharing of information about one’s HIV status. Negative Self-Image relates to feeling unclean or inferior to others because of being HIV-positive. Concerns with Public Attitudes about People with HIV refer to perceptions of what other people typically think about a person who has HIV or what individuals with HIV can expect from others when the latter learn about their HIV-positive status. We used a validated, abbreviated version of the 40-item HIV Stigma Scale shortened to ten questions measuring these concepts with good internal consistency and validity [36], translated into Russian, in which participants responded to statements related to their experience of HIV-related stigma on a 4-point Likert scale capturing degrees of agreement and disagreement (Appendix Table 5).

The primary dependent variable was “Total Stigma Score”, which was computed by summation of participants’ responses to all questions. Secondary outcome variables were subscale scores for “Personalized Stigma” and “HIV Disclosure Stigma”. Three additional secondary outcomes were analyzed as binary variables: “Negative Self Image” and “Public Attitudes Stigma” were dichotomized at the median, and the response to the additional question “I worry that the people I drink or do drugs with will act negatively toward me if I tell them that I am HIV positive” was dichotomized as responses with “agree or strongly agree” vs. “disagree or strongly disagree”.

The main independent variable was alcohol dependence, evaluated using the Composite International Diagnostic Interview (CIDI) [38]. We measured alcohol dependence at baseline and 12 months. In secondary analyses, we evaluated risky alcohol use in the past 30 days, assessed by 30-day timeline followback [39] and defined by the above noted NIAAA criteria.

Covariates in these analyses to address potential confounders included the following: age; a composite variable combining gender and the variable men having sex with men, classifying participants into one of the following three categories: females, males who have sex with men, and males who do not have sex with men; marital status; education (up to nine versus ten or more years of education); employment status; mental health (mental health score, MCS) and physical health (physical health scores, PCS) using the SF 12 [40]; time since HIV diagnosis; depressive symptoms (Beck’s Depression Index) [41]; drug dependence as measured by the CIDI [38]; lifetime suicide attempts (dichotomous); and social support measured by the Social Support Scale [42].

Analysis

We included all study participants in this secondary analysis and generated descriptive statistics for all variables at baseline overall and stratified by alcohol dependence, the main independent variable. Those with and without alcohol dependence were compared using Chi square and Student’s t-tests as appropriate. We fit separate linear mixed effects models to evaluate the association between alcohol dependence and each of the following outcomes: total stigma score (primary outcome), personalized stigma score, and disclosure stigma score at each assessment (i.e., baseline and 12-months). Subject-specific random intercepts were used to account for the correlation due to having repeated observations. We used generalized estimating equations (GEE) logistic models to evaluate the associations between alcohol use (dependence and risky use) and the dichotomous outcomes for negative self-image stigma, public attitudes stigma and the response to “I worry that the people I drink or do drugs with will act negatively toward me if I tell them that I am HIV positive.” The GEE models were fit using an independence working correlation structure and robust standard errors are reported from the models. The longitudinal models accounted for the correlation from using repeated observations from subjects over time. Magnitudes of effect for those with alcohol dependence versus those without for each of the outcomes of interest were estimated using mean differences (mixed effects models) and odds ratios (logistic models), along with corresponding 95% confidence intervals. Prior to regression modeling, we calculated Spearman’s correlation coefficients between all independent variables and covariates and verified that no pair of variables in the same regression model had correlation greater than 0.40. Final models were adjusted for all covariates. We modeled the alcohol variables, depressive symptoms, social support, drug dependence, mental and physical health scores as time-dependent covariates. All other covariates were baseline values. We performed all analyses using two-sided tests and a significance level of 0.05, using SAS software (version 9.3; SAS Institute, NC, USA).

Results

Participants’ demographic and clinical characteristics are described in Table 1. Among the study sample, the mean participant age was 30.1 years (SD = 5.2), 41% were female, 36% were married or living with a partner, 78% reported having 10 or more years of education, and 73% were regularly employed.

Table 1 Baseline demographic and behavioral characteristics of a cohort of HIV-positive PLWH in Saint Petersburg, Russia (N = 700) overall and stratified by alcohol-dependence
Full size table

Unadjusted analyses of baseline characteristics for descriptive purposes suggested that alcohol dependent individuals had higher total stigma, personalized stigma and negative self-image scores than those without dependence (Table 1). Among people who reported risky drinking in the past 30 days, mean total HIV stigma scores and stigma subscale scores were similar to those who did not drink at-risk (Table 2).

Table 2 Mean and standard deviation (SD) for Baseline HIV stigma scores overall (N = 700) and stratified by alcohol dependent and not alcohol-dependent
Full size table

The longitudinal analysis included 1176 observations from 700 study participants. As detailed in Table 3, after adjusting for potential confounders, the total HIV stigma score was on average 0.60 (95% CI: 0.03, 1.17) points higher for participants with alcohol dependence than for those without alcohol dependence (p = 0.04). In addition, the covariates moderate to severe depressive symptoms (β 1.86, 95% CI (1.23, 2.49) and low social support (less than median, β 0.67, 95% CI 0.10, 1.24) appeared to be associated with higher total stigma score.

Table 3 Multivariable longitudinal regression models to assess association between alcohol dependence and primary and secondary outcomes
Full size table

In secondary analyses, those with alcohol dependence had a Personalized Stigma score (PSS) that was on average 0.34 (95% CI 0.11, 0.57, p < 0.01) higher than for those without dependence in adjusted analyses. It was also associated with the statement “I worry that the people I drink or do drugs with will act negatively toward me if I tell them that I am HIV positive” (AOR 1.44, 95% CI 1.10–1.89. p = 0.01) in adjusted models. We observed no significant association between alcohol dependence and the other secondary outcomes (HIV Disclosure Stigma Score, Negative Self Image, and Public Attitudes Stigma). No significant associations were observed between risky alcohol use in the past 30 days and any measure of HIV-related stigma (Table 4).

Table 4 Multivariable longitudinal regression models to assess association between risky alcohol use and primary and secondary outcomes
Full size table

Discussion

This study investigated the relationship between unhealthy alcohol use and HIV-related stigma in a cohort of PLWH in St. Petersburg, Russia, a setting where the prevalence of both HIV infection and unhealthy alcohol use are high. The results indicate that alcohol dependence is significantly associated with greater total HIV stigma. We did not detect associations for lesser degrees of alcohol problems (i.e., recent risky alcohol use).

The current literature examining the association between alcohol use and HIV stigma is limited, but seems to support an association between increased alcohol use and more HIV-related stigma. A study of HIV stigma and social support in a convenience sample of 283 African Americans in Los Angeles, USA, found that those who met the clinical criteria for either current alcohol abuse or dependence reported a higher level of perceived HIV-related stigma compared to those who did not meet the criteria [34]. Another study indicated a relationship between alcohol use and HIV stigma: a cross-sectional survey of 1230 men who have sex with men in Shandong province, China, revealed that frequent episodes of alcohol use were independently associated with higher levels of HIV-related stigma and discrimination, and that experiencing higher levels of HIV-related stigmatizing and discriminatory attitudes was independently associated with alcohol use [43].

Both alcohol dependence and HIV stigma have independent associations with HIV risks and therefore direct implications for HIV prevention. Risky alcohol use and alcohol dependence among PLWH are associated with various HIV risks [44]. Alcohol use can create risks for HIV transmission through adverse behavior by reducing drinkers’ inhibitions to engage in risky sexual behavior and through biological pathways promoting HIV transmission and progression [45, 46]. Meanwhile, alcohol consumption levels in Russia have been consistently high particularly among males [47], and hazardous alcohol consumption has been a major determinant of mortality [29, 48]. Like unhealthy alcohol use, HIV-related stigma carries significant consequences on HIV risks and health care of PLWH. Studies suggest that HIV-related stigma adversely affects access to medical care [49], HIV treatment adherence [49,50,51], and HIV serostatus disclosure [52].

The association between alcohol dependence and HIV stigma observed in this study might indicate that alcohol use disorders such as alcohol dependence, as a maladaptive reaction to HIV infection [53], potentiate the high stigma affecting Russians. Inversely, alcohol use disorders might precede HIV infection and be potentiated by HIV stigma. Thus, high HIV stigma might reduce people’s ability to cope with their HIV infection or address their alcohol use; or alternatively, alcohol use might further increases their vulnerability to HIV-related (and other forms of) stigma and discrimination.

In the absence of a report of clinical consequences of the Berger stigma scale, it is difficult to interpret to what extent the observed magnitude of effect (i.e., the 0.60 statistically significant difference in total stigma score), is clinically meaningful. However, given that these findings are statistically significant, providers who care for PLWH should not only evaluate and treat alcohol disorders, but also be aware of HIV stigma in this population. In addition, the apparent associations between depressive symptoms and low social support with stigma, while included as covariates and not primary independent variables in our analysis, suggest another reason why these two factors warrant attention in the clinical care of HIV-positive people and those with alcohol dependence. Assessment of and clinical interventions for HIV stigma, and possibly other forms of stigma, have not yet been established in clinical care and are an excellent subject of future research. Research measuring stigma related to alcohol use and other stigma sources is also necessary in order to understand the intersectionality between alcohol use and HIV-related stigma. Given that alcohol dependence is also associated specifically with the subscales of personal stigma and the worry of other people acting negative to HIV disclosure, interventions that address shame and build internal coping mechanism among affected people, such as acceptance and commitment therapy [54, 55], might warrant particular focus for clinical intervention research.

Finally, HIV stigma and alcohol use disorders are certainly only a part of the HIV prevention challenge in Russia. Measures aiming to reduce HIV transmission attributable to stigma and unhealthy alcohol use can ultimately only be effective when they are framed in a health system that also offers both effective HIV treatment and effective drug use disorder therapies including opioid agonists [56].

Limitations

This study has several limitations. Given its observational design, we cannot assign causality or ascertain the directionality of the observed association between alcohol dependence and HIV stigma. While alcohol dependence might increase HIV-related stigma and its adverse consequences, it is also conceivable that, inversely, people with higher HIV-stigma have less agency and use alcohol as unhealthy, maladaptive coping process. Further, the alcohol use measures (alcohol dependence and risky drinking) rely on self-report and were not confirmed by objective measurements. Moreover, this survey was subject to the usual recall and social desirability biases. In addition, stigma identities other than those related to HIV could not be measured by our survey instrument. Finally, 29.7% of participants had missed follow-up visits at 12 months. Among the baseline demographic and clinical characteristics presented in Table 1, most were similar for those who completed the 12 month visit compared to those who did not. However, two characteristics appeared to differ: moderate/severe depressive symptoms (less prevalent among completers) and men who have sex with men (more prevalent among completers). If there is non-ignorable missing data, the results of this study may be biased.

Conclusions

In this study, we observed an association between alcohol dependence and HIV stigma among Russian PLWH. This might indicate that unhealthy alcohol use is a maladaptive reaction to HIV infection, which further potentiates the already high stigmatization of affected PLWH. Inversely, HIV stigma might decrease their ability to cope with alcohol use. Clinicians caring for PLWH need to be aware that those with alcohol use disorders might be more likely to be stigmatized, adding the need for interventions to address the potentially multiple stigmatizations of HIV-positive people who also have alcohol-related and other problems.

Targeted HIV prevention efforts in Russia need to consider unhealthy alcohol use for its known related HIV risk behaviors and should address the potential adverse effects of HIV-related stigma among those with alcohol use disorders. Clinical and public health efforts need to develop and test stigma interventions in Russia, as those might not only reduce stigma, but also potentially contribute to limit alcohol use and its negative consequences on HIV risks among PLWH.