Comparing sexual risks and patterns of alcohol and drug use between injection drug users (IDUs) and non-IDUs who report sexual partnerships with IDUs in St. Petersburg, Russia

To date the great majority of Russian HIV infections and AIDS cases have been diagnosed among injection drug users (IDUs) {Population Reference Bureau, 2007 #574; Shaboltas, 2006 #384; Kozlov, 2006 #385} and previous research indicates that drug injection is markedly more prevalent than a decade ago [4]. In Russia and elsewhere in the former Soviet states, the drugs most commonly injected are heroin and ephedrine derivates, including methamphetamine [5‐7]. Both categories of drugs have been independently associated with sexual practices that risk HIV transmission or acquisition [6, 8, 9]. In addition, alcohol consumption in Russia has substantially increased in the turbulent years following the collapse of the Soviet Union [10, 11] and is reportedly among the highest in the world [12]. Heavy alcohol consumption, which is prevalent among IDUs in Russia, has been associated with sexual risk-taking and injection practices that facilitate HIV transmission [13, 14].

Heavy use of illicit injectable drugs and alcohol has led to concerns about the potential for a sexual transmission of HIV into the general population. There are several factors that determine whether HIV transmission will be limited to the core population of injectors in which it is prevalent or if it will become a generalized epidemic [15]. Among these are the size and activity levels of the bridging population, which is comprised of those who form partnerships with both IDUs and with non-IDUs, and partnership formation patterns within the general population [16, 17]. Studies conducted among Russian non-drug using youth indicate that in recent years levels of unprotected sex have increased, age of sexual debut has decreased, rates of partner change have increased, and sexual activity under the influence of alcohol has become more prevalent among the general population, increasing the potential for the transmission of sexually transmitted infections (STIs) overall [18‐20]. While HIV transmission-related behavior is prevalent among Russian IDUs [21‐23], it is not known whether such behavior is more or less prevalent among their non-IDU partners. In addition, while a number of substance-use-related patterns have been associated with sexual risk among Russian IDUs, association between sexual risk and substance use has not been estimated for non-IDUs who have IDU sex partners.

The current analysis was conducted using data from a study in St. Petersburg that was part of a large multi-site project, the Sexual Acquisition and Transmission of HIV Cooperative Agreement Program (SATH-CAP) [24]. The goal of SATH-CAP was to collect and analyze data on sexual and drug use practices and other social, environmental and biological factors that may influence HIV transmission, from high-prevalence core populations, i.e. IDUs and men who have sex with men (MSM) to the greater general populations of which they are part. This analysis compared the sexual risk behaviors and patterns of substance use between heterosexual IDUs and non-IDUs who have IDU sexual partners with the following aims: (1) to investigate whether substance use measures in both groups differed in their correlations with sexual risks, and (2) to determine whether sexual risk behaviors are associated with global substance use levels and with substance use in sexual contexts.

Contrary to the initial hypothesis, IDUs were significantly less likely than non-IDUs who reported at least one IDU sex partner to have multiple sex partners and unprotected sex with casual partners. IDUs had fewer partners even after controlling for methamphetamine use, which was associated with higher number of sexual partners. This contrasts with findings from other studies that suggested that IDUs engaged in as much or more [9, 30] sexual risk behaviors compared to individuals who did not inject.

IDUs were less likely than non-IDUs to engage in risky drinking, and risky drinking was not independently associated with sexual risk among IDUs. These results contrast with studies showing alcohol consumption to be greater among narcotic [31] or methamphetamine users [8, 32] and to be associated with sexual risk taking among IDUs [33, 34]. In a previous study, IDUs in Russia reported consuming lesser quantities of alcohol than IDUs in the United States [13]. Given that drug dependence and alcoholism have been observed to be co-occurring events in Russia [14] it is unclear whether the lesser prevalence of risky drinking among IDUs is due to a less risky alcohol consumption pattern or if it results from our failure to identify some other type of hazardous drinking that may occur in Russia [12]. It could be argued that the lower levels of risky alcohol use among IDUs could be due to underreporting as a result of social desirability bias or stigma and discrimination against IDUs. However, alcohol use is widely accepted in Russia and studies have shown that drinking is often used as a facilitator of sex and as a symbol of masculinity in the Russian culture [35]. In addition, there was no significant difference in the reported frequency of alcohol use between IDUs and non-IDUs in this study, it was only the pattern of risky drinking that significantly differed between the two groups. A possible explanation for these results is that alcohol is often used to remove pressure, relax and reduce stress whereas opiates are much more powerful relaxants than alcohol [35‐39]. Since the majority of IDUs in this study injected heroin, they may have had a lesser need to resort to risky patterns of drinking in order to self-medicate [31] compared to non-IDUs. Another possible explanation for the lower levels of risky drinking among IDUs, comes from data showing that a majority 95% of IDUs in Russia are infected with hepatitis C virus (HCV) [40] and HCV infections often lead to liver diseases [41]. IDUs are also more likely to have HIV, to receive antiretroviral therapy or to have co-morbid conditions leading to poor health outcomes [42]. All of these factors could lead to a reduction of tolerance among IDUs for heavy patterns of drinking. The sedating effects of heroin and poor health conditions among IDUs could also account for the lower levels of sexual risk behaviors observed among IDUs compared to non-IDUs. In addition, the discrepancy between the two groups may be compounded by the increase in sexual risks that result from the greater risky alcohol use among non-IDUs [35].

Similar to previous findings, the present results show methamphetamine use to be the strongest predictor of having multiple sex partners and unprotected sex among this sample of heterosexual IDUs [6, 8, 43, 44], while among non-IDUs, those sexual behaviors were associated with risky alcohol use [20, 45].

The greater likelihood of reporting sex while high on drugs or alcohol among IDUs as well as among non-IDUs who had multiple sex partners probably reflects the greater amounts of substance use among IDUs and non-IDUs who had multiple sex partners. Only IDUs were more likely to forgo condom use when they had sex when high on drugs or alcohol. Similar to other studies of non-IDUs, the major risk factor associated with unprotected sex was the participant's general (i.e. monthly) pattern of risky drinking and not being drunk at the time of sex [46]. Thus, programs to increase condom use among IDUs need to address substance use patterns that occur in the context of sex, while among non-IDUs, general drinking patterns must be addressed [33].

No significant difference was observed in the sexual risk behaviors according to participants' gender, either overall or in the analysis stratified by drug use status suggesting greater similarities rather than differences between the sexual risks of men and women in this study. Given that previous studies have shown substance abusing women to be at greater risk for unsafe sex and STIs than substance abusing men [47‐49] and studies in Russia have shown sex risk behaviors to be more likely among substance abusing men than among substance abusing women [35], it is also possible that subtle gender differences exist but were not identified in this study. Future research that takes into consideration motivations for risk behaviors and self efficacy for risk reduction might be able to provide further insight into these results.

The prevalence of STIs other than HIV has important epidemiologic implications given the propensity of STIs to increase the likelihood of HIV transmission [50, 51]. Overall, STI prevalence among non-IDUs was comparable to the STI prevalence from the general population in St. Petersburg [52, 53], and STI prevalence among IDUs in this study was similar to the prevalence found among IDUs in other countries [54, 55]. The low STI prevalence might reflect a trend of reduced STIs being reported over the years in Russia [56‐58]. The results have implications for behavior intervention research. The low STI prevalence may suggest that it may not be feasible or cost effective to use STIs outcomes in projects to reduce risk behaviors among these groups. However, given that the STI prevalence among participants is actually higher because other common STIs were not tested in this study, given the great prevalence of sexual risk behaviors among non-IDUs, and given that participants are part of vulnerable populations, efforts to detect and treat STIs among these groups need to continue.

This study had several limitations. Participants recruited into this study were IDUs and non-IDUs who reported at least one IDU sexual partner; results may not apply to non-IDUs who have never had sex with an IDU. Because the capacity of RDS to produce probability samples is limited [59] the generalizability of these results even within the sampled populations may also be limited. Sex risk behaviors may relate to substance use patterns; therefore, behaviors might have been different if the proportions of heroin, methamphetamine and alcohol users had been different. This study may have recruited individuals who were more comfortable participating in research projects rather than IDUs who were harder to reach. Conversely, certain or all risk behaviors may have been underreported due to social desirability or other types of bias. Last, due to questionnaire design, we could not investigate whether having sex when "high" was a specific effect of alcohol or of another drug, nor could we assess whether it was the participant, his or her partner or both who were drunk or high when sex occurred, nor could we assess the level at which participant was "high". Missing of data resulted in fewer observations in multivariate models which may increase risk of type II error; further, if missing data did not occur at random the potential for bias in ORs exists.

Greater prevalence of sexual risk behaviors among non-IDUs who report partnerships with IDUs, compared to IDUs themselves, suggests a potential mechanism for spread of HIV beyond the IDU population. Interventions to reduce HIV transmission should address alcohol consumption among non-IDUs who report sexual partnerships with IDUs, and methamphetamine use and the propensity of having sex when high among IDUs.