Impact of Gender and Sex Work on Sexual and Injecting Risk Behaviors and Their Association With HIV Positivity Among Injecting Drug Users in an HIV Epidemic in Togliatti City, Russian Federation

Abstract

Objectives: 

Togliatti City is witness to a large epidemic of human immunodeficiency virus (HIV) associated with injecting drug use (IDU).

Goal: 

This study sought to examine whether risk behaviors and risk factors associated with HIV differed across a sample IDUs by gender and sex work.

Study: 

A sample of IDU (n = 423) comprising female sex workers (SWs) (n = 66), non–sex workers (nonSWs) (n = 89) and men (n = 268) were recruited by field workers in community settings. Behavioral and HIV prevalence data were collected.

Results: 

HIV prevalence did not differ across the groups (∼56%), but gender adversely affected some risk factors. A comparison of risk behaviors indicated that SWs were more likely to engage in risky injecting behaviors than either men or nonSWs. They were also more likely to report a history of sexually transmitted infections.

Conclusions: 

IDUs involved in sex work and IDU nonSWs require specific and targeted interventions to facilitate safer injecting and sexual behaviors.

In Brief

A study of injecting drug users recruited from community settings found that risk factors associated with human immunodeficiency virus positivity differed by gender, and gender and sex work adversely affected injecting and sexual risk behaviors.

THERE IS STRONG EVIDENCE that the Russian Federation has been experiencing recent epidemics of injecting drug use (IDU) and IDU-associated human immunodeficiency virus (HIV) infection over the past decade.¹ The number of people attending state-provided drug-treatment clinics in the Russian Federation has increased ninefold over the 1990s with nearly 350,000 problematic drug users registered in 2001, of which 86% were injecting opiate users (301,082).² Evidence also suggests that routine surveillance systems underestimate the number of young people using drugs by a ratio of 1:3 and that the true number of IDUs may be as high as 2 million.^2,3 There are currently 222,510 HIV cases registered in Russia, of which 90% are attributed to IDU according to official reports from the Federal Aids Centre. High unemployment, a growth in informal economies, and underfunded education and social services have increased young people’s vulnerability to drug use and to sex work.⁴ Research suggests a close relationship between drug use and sex work among women, with reports that between 22% and 82% of sex workers (SWs) are injecting drug users.^5–7 Studies among female injecting drug users suggest that between 15% and 66% of female IDUs are involved in sex work.^8,9

Togliatti City is the second city of Samara Oblast situated approximately 1000 km from Moscow. In 2001 the registered population was 740,636, though local key informants estimate that the actual population is nearer 1,000,000. Togliatti is a relatively young city, with 50% of the total population between the ages 15 and 44 years and 23% younger than 17 years.¹⁰ The HIV epidemic has occurred recently and rapidly in Togliatti, with nearly 8000 people currently registered as HIV-positive, 93% of whom are reported to have acquired HIV through IDU, and with estimates of HIV prevalence among IDUs estimated at over 50%.¹¹ Syphilis tests among IDUs at the city drug treatment services indicated a prevalence of 3.5% in 2001, 2.0% in 1998, and 2.7% in 2000. Prevalence among women was 8.7% in comparison to 2.4% among men in 2001.¹⁰ There is also a high prevalence of IDU among the 15- to 44-year-old age group in the city. A recent study estimated the number of IDUs to be 20,226 (16,971–24,749) of which 15,037 (12,696–18,515) are men and 5189 (4275–6234) women, a prevalence of 2.9% among the female and 7.8% among the male population aged 15 to 44 years.¹² The Department of Internal Affairs unofficially estimated there are approximately 2200 SWs in Togliatti, of which 50% work on the street.¹⁰

Gender significantly affects risk of HIV infection through biologic factors, as well as economic and social factors that enhance either susceptibility or exposure to risk.^13,14 Recent evidence indicates that gender also affects risk of infection within subpopulations such as IDUs, with disparities in social and economic situation influencing women’s capacity to negotiate safer sexual and injecting practices.¹⁵ There is evidence of higher rates of HIV infection among female IDUs than among male IDUs.^16,17 Female IDUs tend to be significantly younger than their male counterparts, they engage in higher rates of needle and syringe sharing, and are more likely to share their sex partners’ injecting equipment and engage in riskier sexual practices.^18–26 The evidence indicates that women who inject drugs are more likely to have IDUs as regular sex partners than men who inject drugs.^18,20 There are contradictory reports as to whether this acts as a protective factor for female IDUs, reducing the likelihood of them engaging in risky sexual and injecting behaviors outside of their partnership or whether it increases their risk of sexual transmission.^17,27,28 Further research is needed to examine the effect of having a regular IDU sex partner on female IDUs’ sexual and injecting risk behaviors. The evidence suggests the primary route of HIV transmission for IDU SWs remains drug injecting and not sexual transmission. IDU SWs have a greater risk of infection related to injection than women not currently engaging in sex work (nonSWs) due to higher rates of borrowing injecting equipment and injecting in higher risk environments.^29,30 However, there is also emerging evidence indicating that sexual behavior is associated with HIV-1 seroconversion in street-recruited IDU SWs.²⁸

Given the multiple vulnerabilities associated with drug injecting and sex work, magnified in a risk environment characterized by rapid social and economic transition, this paper focuses on the relationships between gender and sex work among a sample of injecting drug users in Togliatti City in the Russian Federation, a city characterized by a high prevalent HIV epidemic associated with drug injecting.¹¹ Recently published studies among SWs in the Russian Federation have highlighted the potential for sexual transmission of HIV from IDUs to the wider population.^31,32 Considering this risk, this paper specifically examines the relationship between gender and sexual and injecting risk behaviors among IDUs, investigating to what extent gender and sex work are associated with increased risk behavior and risk of HIV infection.

Methods and Materials

This paper draws on findings from a cross-sectional survey of injecting drug users (n = 426) conducted in September 2001, recruited through a harm reduction outreach team working as indigenous field workers in the community. Methods used in this study are described in detail elsewhere.¹¹ Oral fluid samples were obtained for unlinked anonymous HIV testing and demographic and behavioral data collected through a structured questionnaire. Case definition for IDU was defined as injecting within the last 4 weeks and, for SWs, as having exchanged anal, vaginal, or oral sex for money or goods within the last 4 weeks.

This paper summarizes risk factor analyses comparing men, nonSWs and SWs. Risk behavior is compared across the 3 groups to assess whether risk increases incrementally between men, nonSWs, and SWs. Additional analyses examine risk factors associated with an involvement among IDUs in sex work. Univariate and multivariate analysis examines risk factors for HIV within each of the 3 groups.

Data Analysis

Risk behaviors across the 3 groups were compared using χ² tests for categorical variables and Bartlett test for equal variance for continuous variables. Analyses were performed using the statistical package Stata version 7 (StataCorp, College Station, TX) and significance was set at the 5% level. Associations between the outcome variable (antibody positivity) and covariates were explored using the likelihood ratio statistic. Multivariable analysis was conducted for each group by examining risk factors together in one model and assessing their effect on the model through likelihood ratio statistics. Risk factors used in the multivariable model were selected when differences were found in the risk behavior analysis. Some risk factors were excluded if they contained insufficient values.

Results

A total of 423 IDUs were recruited, of whom 37% (155) were female. Among female IDUs, 57% (89) had never exchanged sex or had not done so in the last 4 weeks, 50% (77) had ever exchanged sex for goods or money, and 43% (66) had exchanged sex in the last 4 weeks. Two men reported having exchanged sex for goods or money but were not included in the analysis. HIV prevalence was 56% in the total sample and did not differ significantly across the 3 groups. All results are summarized in Table 1.

TABLE 1: Comparison of Demographic and Risk Behavior Indicators Among Injecting Drug Users in Togliatti City by Gender and Sex Work

Female IDUs were on average younger than male IDUs (26.7 years vs. 23.8 years, P <0.001) though there was no difference between SWs and nonSWs. Slightly more nonSWs reported completing higher education (73.0%) compared to men (66.8%) and SWs (56.1%), though this was only just significant (χ² = 2, DF = 4.9, P = 0.09).

History of Injecting Behaviors

On average, men and nonSWs started injecting at a similar age (18.9 and 18.2 years), but SWs reported first injection significantly earlier (17.4, P = 0.01, F = 2, DF = 4.7). A similar proportion of both men (5.3%) and nonSWs (5.7%) reported ever having injected with a needle/syringe previously used by someone known to be HIV-positive but a significantly higher proportion of SWs (16.9%, χ² = 2, DF = 10.8, P = 0.004). Similarly a significantly higher proportion of SWs reported ever injecting with a needle/syringe previously used by someone known to be HCV positive (45.2%) than either nonSWs (27.1%) or men (30%, χ² = 2, DF = 6.4, P = 0.04).

Current Injecting Risk Behaviors

Proportionally more SWs (86.4%) reported injecting daily than nonSWs (54.6%) and men (51.3%, χ² = 2, DF = 27.0, P <0.001). More SWs (51.5%) reported injecting with needles or syringes previously used by someone else in the last 4 weeks than both nonSWs (41.6%) and men (30.2%, χ² = 1, DF = 12.0, P = 0.002). Similarly, more SWs (81.8%) reported sharing paraphernalia in the last 4 weeks than men (67.2%) and nonSWs (65.2%, χ² = 2, DF = 6.1, P = 0.05). Injecting with the same needle or syringe more than once was reported more often by SWs (80.0%) than nonSWs (70.0%) and men (60.4%, χ² = 2, DF = 8.0, P = 0.02).

Sexual Risk Behavior

Sexual risk behaviors also varied across the groups. A significantly higher proportion of nonSWs (45.4%) reported having had sex with a regular partner in the last 4 weeks without using a condom than either men (27.9%) or SWs (12.5%, χ² = 2, DF = 17.4, P <0.001). NonSWs (70.2%) were more likely to report their regular partners to be IDUs than either SWs (68.4%) or men (40.8%, χ² = 2, DF = 11.9, P <0.001). Men (6.4%) were more likely to report having anal or vaginal sex with a casual partner with inconsistent condom use than either nonSWs (2.3%) or SWs (1.6%) though this was not significant (χ² = 2, DF = 2.8, P = 0.13). There was a significant difference in reported history of sexually transmitted infections, 57.1% of SWs reported ever having an STI compared to men (38.1%) and nonSWs (21.2%, χ² = 2, DF = 20.1, P <0.001).

Environmental Indicators

Significantly more men (34.2%) reported ever being in prison than SWs (22.7%) or nonSWs (18.6%, χ² = 2, DF = 11.5 P <0.001). More SWs (30.3%) reported having their needles or syringes confiscated by the police in the last 4 weeks than men (13.8%) or nonSWs (9.1%, χ² = 2, DF = 16.3, P <0.001). More nonSWs (44.3%) reported ever being in treatment than men (32.5%) or SWs (33.3%, χ² = 2, DF = 4.2, P = 0.12) though this was not significant. A higher proportion of SWs (63.6%) and nonSWs (60.9%) reported ever having lost consciousness (overdosed) due to drugs than men (48.1%, χ² = 2, DF = 7.7, P <0.02).

Multivariate Analysis

In the univariate analysis, male IDUs who had injected with used paraphernalia in the last 4 weeks had twice the odds of being HIV-positive than those who had not (95% CI, 1.26–3.6). Similarly those who reported ever injecting with someone known to have HCV were twice as likely to be HIV-positive (95% CI, 1.48–4.55). Those who reported having had anal/vaginal sex with a casual partner and inconsistent condom use in the last 4 weeks also had 3 times the odds of being HIV-positive (95% CI, 0.97–9.67). In the multivariate model, these risk factors remained significant, and the odds of being HIV-positive became more strongly associated for those reporting having anal or vaginal sex with a casual partner (OR, 5.3; 95% CI, 1.33–20.8). Interactions were explored between the casual sex partner variable and other variables. No interactions were found, but 80% (8/13) of HIV-positive male IDUs who reported having sex with a casual partner and inconsistent use of condoms also reported injecting with used needles and syringes in the last 4 weeks, and 81% (9/13) also reported injecting with used paraphernalia. Neither duration of injection nor injection with used needles/syringes in the last 4 weeks was associated with HIV univariately or in the multivariate model for male IDUs. These results are presented in Table 2.

TABLE 2: Univariate and Multivariate Risk Factors for Antibodies to HIV Among Male IDUs in Togliatti City, 2001

Similarly for nonSWs, duration of injection was not associated with HIV in the univariate analysis. Injection with a used needle or syringe in the last 4 weeks was the only risk factor associated with increased odds of having antibodies to HIV in the univariate analysis. This remained associated in the multivariate model, nonSWs who reported injecting with a used needle or syringe in the last 4 weeks were 4 times more likely to be HIV-positive (96% CI, 1.01–16.6). In the multivariate model, one protective factor was found, those who reported having been in treatment had slightly reduced odds of being HIV-positive (OR, 0.3; 95% CI, 0.01–1.12). These results are presented in Table 3.

TABLE 3: Univariate and Multivariate Risk Factors for Antibodies to HIV Among Non–Sex-Working Female IDUs in Togliatti City, 2001

No significant risk factors associated with HIV positivity were found for SWs either univariately or in the multivariate model. However, in the multivariate model, increased odds were found among those reporting ever injecting homemade drugs, those who had been in treatment, and those who had been arrested. These results are presented in Table 4.

TABLE 4: Univariate and Multivariate Risk Factors for Antibodies to HIV Among Sex-Working Female IDUs in Togliatti City, 2001

Behavior of IDU SW with Respect to Commercial Sex

The age of first sex work ranged from 13 to 30 years (mean = 21, SD = 4.4). On average, this occurred 3.6 years after age of first injection (range 6–17 years, SD = 3.9). The majority of SWs reported sex work as their main source of income, and the average reported length of time working was 2.8 years (range 0–16 years, SD = 2.9). SWs reported high rates of partner change among clients and a high turnover of new clients. One hundred percent of the sample reported exchanging penetrative sex in the last week with clients, and 97% reported exchanging oral sex. Eighty-six percent of SWs reported consistent condom use with clients. These data are presented in Table 5.

TABLE 5: Characteristics of Female IDU Sex Workers

Discussion

While HIV infection did not differ across the 3 groups, findings suggest some increased risk in relation to injecting and sexual risk behavior associated with gender and particularly sex work. Different risk factors were associated with HIV among nonSWs and men. For nonSWs, increased odds were associated with injecting with used needles or syringes and for men with injecting with used paraphernalia and anal/vaginal sex with a casual partner. Studies have shown that transmission of HIV via needle or syringe sharing is relatively efficient, suggesting that IDUs who are nonSWs in Togliatti remain highly vulnerable.³³ Further, our findings highlight that SWs are highly vulnerable to IDU-associated HIV. Comparisons suggested that risk behaviors varied between men, nonSWs, and SWs. This implies that the current transmission dynamics of the epidemic in Russia could change unless interventions are put in place that are sensitive to gender differences and sex work.

Our findings indicate that gender affects female IDUs’ susceptibility to HIV infection in several ways. Although demographic indicators, such as educational levels, do not differ between men and nonSWs, they are younger and their reported risk behaviors suggest that their lives are more entrenched in IDU networks. Our findings indicate that nonSWs are more likely than men to have regular sex partners who are also IDUs and their regular partner to be their main source of shared needles and syringes in the last 4 weeks. A similar pattern has been reported elsewhere, indicating that women may be more likely to share needles with individuals with whom they have some degree of an established relationship, such as family, sexual partner, or close friend.^13,34,35 Intuitively, sharing injecting equipment with an injector with whom there is a close or social relationship may be perceived to be safer than sharing with an acquaintance or stranger. However, research also highlights that such a practice pertains as a means of demonstrating trust and intimacy and consequently may be far more resistant to behavior change.²⁶ Evidently, this is an important issue as the odds of being HIV-positive were 4 times higher among nonSWs who reported injecting with previously used needles and syringes than those who did not.

This trend is repeated with respect to sexual risk behavior. A higher proportion of nonSWs reported having sex with a regular partner in the last 4 weeks than either men or SWs, and they were less likely to report always using a condom than SWs. This may also be borne out of a desire to demonstrate or realize trust within the context of primary or regular relationships.^13,36

We found that sex work was an additional risk factor for HIV transmission among women. Our findings indicate that SWs in Togliatti are a very young and highly marginalized group. On average, reported age of first sex work followed age of first injection, suggesting that sex work may provide a means to support an established drug habit. Injecting risk behavior increased among SWs in comparison to nonSWs and men. They were more likely to inject on a daily basis, inject with a previously used needle or syringe, inject with a needle or syringe previously used by someone known to be HIV or HCV positive, and have had a drug-related overdose than either nonSWs or men. These behaviors are corroborated by international evidence suggesting that SWs are more vulnerable to injecting in riskier environments.^18,37,38

Turnover of new clients appears to be very high, with SWs reporting an average of 35.8 new clients a week and 100% reporting having sold penetrative sex in the last 4 weeks. No data were collected on the profile of clients, although a previous qualitative study in Togliatti indicated that most clients of SWs are not IDUs.³⁹ The most compelling indicator of sexual risk behaviors across the 3 groups was a significant difference in the proportion of reported history of STI, which tripled between SWs and nonSWs, suggesting a high risk of sexual acquisition and transmission of STIs associated with sex work. It also suggests that the conditions exist for SWs to potentially be acting as a bridge passing STI/HIV from the IDU population to their non-IDU clients. A similar pattern has been shown elsewhere in the Russian Federation.^9,30,41 One study in St. Petersburg found that prevalence of syphilis was 9.5 times higher in female IDUs than male, with 66% of the female sample reported ever having exchanged sex for money or drugs.³²

The sexual behavior data highlighted another group in Togliatti not directly represented in the survey but potentially at risk. This group is female non-IDU sex partners of male IDUs. Over half of men reported their regular sex partner to be non-IDU and a third reported having sex with their regular partner without a condom. Three quarters reported having sex with a casual partner with inconsistent condom use. This is of particular relevance as in the multivariate analysis increased odds of HIV were associated with anal/vaginal sex with a casual partner for men. Given the wide confidence interval of this odds ratio and the small sample size, this finding should be interpreted with caution; however, it does further support the evidence that interventions targeting sexual risk behaviors of IDUs need to be considered. Evidence has shown that in many situations the majority of cases of heterosexual transmission of HIV to noninjecting women involve an IDU source.⁴¹ Further research to collect biologic indicators and assess the evidence of sexual transmission via both IDU SWs to their clients and from male IDUs to their non IDU sex partners is needed.

Some behaviors indicative of risk reduction were reported. High proportions of nonSWs reported accessing drug treatment services in comparison to men and in the multivariate analysis some association was found between attendance at drug treatment services and reduced odds of infection for nonSWs. Similar proportions of SWs and nonSWs reported accessing drug treatment services. It is important to note that lack of confidentially at needle and syringe exchange programmes and at STI clinics in Russia has been reported to be a factor discouraging people from using such services.^42,43

Limits

As in most studies of injecting drug users, reported behavioral findings are drawn from self-reports. The fact that behavioral data were collected via a fieldwork team involving current or former drug users may in some way limit potential bias associated with socially desirable responses. A limitation arising from the sampling method of recruiting IDUs in the community meant that there was no clear sampling frame, though there are advantages in potentially minimizing bias in community-recruited studies of IDU compared to treatment-based samples, especially if multisite and multinetwork recruitment strategies are used as in this study.⁴⁶

Conclusions

International evidence indicates that HIV transmission in IDU populations is influenced by the pattern and rates of sharing among IDU populations, as well as the structure of sharing networks.⁴¹ This study highlights the impact of gender and sex work on sexual and injecting risk behaviors among injecting drug users. This is particularly pertinent in the context of the Russian Federation, as surveillance data show that in locations where the IDU epidemic is maturing, the proportion of newly detected HIV cases with heterosexual transmission as the risk factor is increasing.⁷ Studies have suggested that interventions targeting IDUs have had greater success in fostering behavior changes in injecting risk behavior than sexual behavior.⁴¹ Interventions specifically targeting sexual behavior change would appear to be less developed in Russia, with the greatest weight to date placed on the need to maximize syringe distribution coverage and exchange.¹ Patterns of sexual behavior among IDU are also often not dissimilar to general population samples,⁴⁵ and this points to the importance of interventions with the capacity to foster normative social changes at the community level.^46,47 Our findings emphasize the need for sexual risk reduction as an integral feature of HIV prevention programming for IDUs in Russia and for interventions that can take account of how HIV vulnerability among IDUs intersects with gender and sex work.