Variations in sexual risks in drug users: Emerging themes in a behavioral context

Abstract

Drug users are an especially complex population among those studied in HIV-risk behavior research. While injection drug use accounts for over one-third of cumulative HIV transmission in the U.S., the scope of the direct and indirect impact of all drug use is difficult to quantify, especially in relation to attributing HIV to either drug use directly via parenteral exposures or indirectly, through unsafe sex. Important behavioral issues such as social and drug network overlaps, partner selection and the combinations of illicit drugs with erectile dysfunction medications have added complexity to the studying of sexual behavior in drug users. This review covers recent substantive research in the U.S. and Canada on current themes in sexual risk behavior in injection drug and non-injection drug users. We address gender, situational and sexual preference factors that may influence sexual behaviors affecting HIV risk by class of drug and route of administration. Special attention is paid to minority populations, both sexual and race/ethnicity, as their marginalized role in contemporary society places special barriers for risk reduction.

Introduction

In the early years of the HIV epidemic, the sexual risks experienced by drug users were ignored. Only more recently have researchers noted that drug users may acquire HIV sexually. Because the primary route of HIV transmission among IDUS was parenteral, the research community failed to investigate the role of sexual behavior in HIV transmission among injectors. The advent of crack use in the early 1990s changed this perspective, with the recognition that women were rapidly acquiring HIV due to transactional sex.

Sexual risks associated with drug use vary by type of drug used and route(s) of administration, gender and types of sexual interactions. In a recent cross-sectional, event-based study of drug users, those who smoked or injected amphetamines or heroin were less likely to use condoms at their last sexual encounter while no associations were observed for cocaine, marijuana and amphetamine by ingestion (1). Drug use during a recent sexual encounter was associated with decreased condom use in men but not women. It is evident that “drug” use is far too generic a term to fully address the complexities of “sexual risks”.

Until recently, it has been difficult to ascertain whether risky sex was antecedent to or a result of drug use, or if both behaviors are concomitantly associated with other factors. This is, in part, due to research methods used (largely cross-sectional designs) and the complexity in studying these dynamic associations. Researchers have begun to disentangle these relationships, taking into account confounding and temporal considerations using longitudinal cohort studies. Several key studies of injection drug users (IDUs) have reported factors related to sexual risk rather than drug use practices as the leading predictors of HIV seroconversion (2, 3). This review covers recent North American research by addressing current themes of sexual risk behavior in drug users; within types of drug administration, gender, situational factors and sexual preference are addressed that may influence sexual behavior.

IDUs

Male and female injection drug users (IDUs) represent an estimated 17% and 26%, of males and females living today with HIV/AIDS in the United States, respectively (4). While parenteral frequency is the primary driver of HIV acquisition, IDUs acquire and transmit HIV and sexually transmitted infections (STI) sexually as well. Holmberg et al. suggest that IDUs account for over half of all HIV seroconversions each year due to direct and indirect transmission (2, 5). Despite knowledge of their HIV-positive status, one-quarter (26%) of IDUs reported unprotected sex partially attributed to HAART initiation. In addition, longitudinal studies of IDUs have demonstrated associations between HIV incidence and sexual risk behaviors such as male same-sex contact, history of STIs (2, 6) and sex with an IDU (7).

Heterosexual Male IDUs

Inconsistent condom use is common among heterosexual male IDUs, which serves as a bridge for transmission to non-IDU sexual partners. Among a sample of heterosexual IDUs, only 12% of men with a main partner and 17% of men with multiple partners reported consistent condom use (8). Male IDU characteristics associated with consistent condom use vary by sexual partner type, but having peers engaged in sexual risk reduction, involvement in IDU risk reduction (9), and perceived support from partners (8) have been described. Inverse associations have also been observed between condom use and needle-sharing with sexual partners and intimate partner violence (8) among IDUs.

Numbers of sexual partners may also alter drug-related HIV risk. Younger age, injecting daily, less education, shooting gallery attendance, injecting cocaine and same sex activity were predictive of HIV seroconversion in a large prospective cohort study (2). Those with one partner or more had a reduced risk of HIV seroconversion, a finding that could be due to greater drug dependence of those without sexual partners and the social support conferred by having a sexual partner (2). Alternatively, those with only one sexual partner could potentially share injection equipment with a smaller number of people (7); note that differential risk for those with one or multiple partners could not be determined.

Among newly initiated male IDUs, engaging in sex with other men, being African American, younger age, and sharing needles in the past 6 months were associated with HIV infection (10). Recently initiated IDUs had, expectedly, a higher prevalence of HCV than non-IDU (55.23% vs. 3.04%); however, a higher prevalence of syphilis antibodies was observed in non-IDU (4.38%) compared to IDU (1.48%), results indicative of sexual transmission (10). New IDUs are often initiated into injection by older IDUs who have a higher prevalence of STIs, and sexual mixing of younger IDUs with older IDUs (11, 12) has also been implicated in seroconversion and as a bridge for sexual transmission to non-IDUs (11).

Male sex workers comprise a population whose risk is a neglected area in HIV-risk research. In a longitudinal study of male IDUs in Canada, 11% of men reported sex trade at baseline which was associated with HIV serostatus, injecting cocaine, smoking crack, inconsistent condom use with casual partners and frequent reports of borrowing syringes (13).

MSM IDUs

In 2008, CDC reported that 46% of the new reports of HIV/AIDS diagnoses reported between 2001 and 2006 were MSM and 4% were MSM engaging in IDU (14). Overall, 6% of MSM reported any injection for nonmedical purposes and 2% reported injection in the prior 12 months (15). IDU MSM visit shooting galleries (16) and share needles (17) more frequently than non-MSM. Incident HIV is 2.5 times greater for male IDUs having MSM (2) and MSM were 8.8 times as likely as heterosexual men to seroconvert in a case-control study in San Francisco (3). HIV prevalence is markedly higher among MSM IDU than male IDUs who do not report sex with men (47% vs. 8%) (10), suggesting an important epidemiologic risk factor.

Female IDUs

Female IDUs acquire HIV from both risky sex and drug use. Female IDUs had 13% higher incidence of syphilis, gonorrhea and Chlamydia, STIs not directly transmitted by injection drug use than non-IDUs in a sample of drug users (18). Recent cohort studies have refined our understanding of the relationship between risky sexual behavior and drug use, particularly among women. After accounting for confounding, Strathdee et al. demonstrated that sexual risk, not drug-use behavior, was an independent predictor of HIV incidence in female IDUs (10). A recent STI, an indicator of risky sexual behavior, more than doubled HIV seroconversion in IDUs (Rate Ratio = 2.52, 95% Confidence Interval (CI): 1.39, 4.58) while in men, drug-related behavior and having sex with other men were important risk factors for HIV seroconversion (2), highlighting the different mechanisms by which men (via drug and male-to-male sex risks) and women (via heterosexual sex risks) acquire HIV. Among recently initiated IDUs, HIV prevalence was higher in females than males (12% vs. 6%), where HIV prevalence was associated with having an IDU sex partner, having had an STI, and initiating sex before age 15 in women but not men (10).

In a nested case-control study of IDUs, Kral and colleagues found HIV risk in women was primarily attributable to sexual risks, as sex trade in the previous year (Adjusted Odds Ratio (AOR) = 5.10, 95% CI: 1.90–13.70) was the strongest predictor of HIV seroconversion (3). Having a steady sex partner who also injected drugs was protective against HIV seroconversion (AOR = .32, 95% CI: .11–.92), suggesting that sexual transmission among closed networks reduced risk. Multiple sex partners may confer a risk for female IDUs above and beyond risk associated with IDU history of their sexual partners. Female IDUs with steady sexual partnerships may be less likely to have other risky sexual or drug-related behavior (3). Moreover, high-frequency sex trade (exchanging money or drugs with 50 or more partners in the past 10 years) was associated with HIV seropositivity in a longitudinal study of female IDUs after adjusting for other drug use behaviors (19).

Women Who Have Sex with Women (WSW) IDUs

Of the 246,461 women with HIV up to the end of 2004, only 3.0% reported having sex with a woman, most of whom reported injection drug or sex with men use as well, leaving only 534 women (0.2%) with infection attributable to sex with women only (WSWO) (20). Very little research exists to clarify the risks of women who have sex with women (WSW) and even less data are available for WSWO. Limited reports, riddled with measurement and sampling issues, do suggest negligible HIV risk for female-to-female transmission (20). A sample of women (69.3% IDUs) who did and did not initiate injection reported that sex trade (AOR = 4.02, 95% CI: 1.67–9.68) and living without one s parents as a child (AOR = 3.05, 95% CI: 1.07–9.59) predicted recent WSW status while sex trade predicted former WSW status (AOR = 3.97, 95% CI: 1.65–9.59) when compared to never WSW. The three groups were significantly different on a number of sexual behaviors including age at sexual debut, proportion having more than one steady male partner and having had a partner with an STI as well as their own STI history, all of which were greater in former and current WSW than never WSW.

Similarly, in a Canadian study of female IDUs, WSW-IDUs were more likely to engage in sex work, although consistent condom use was similar for regular, casual and male sex-trade clients were similar (22). The limited number of women exclusively having sex with women in most samples does not allow for drawing strong conclusions about this population. However, it is clear that drug-using WSW have a number of sexual risk factors that should not be ignored.

Non-Injection Drug Users

Heterosexual Male Non-IDUs

Heterosexually identified men have the lowest frequency of condom use and safer sex compared to MSM, whom have the highest frequency of condom use, and bisexual men (23). Some drug use may precede inconsistent condom use and is often associated with prolonged intercourse. While cocaine and heroin have been reported to increase as well as depress sexual arousal, with either effect often coupled with a decrease in performance (18, 23), methamphetamine has been reported to increase libido without hindering sexual performance, prolonging stamina(18). Methamphetamine use in non-IDU has been associated with a number of HIV-risk behaviors, some of which include sex with multiple partners and decreased condom use, as well as trauma associated with prolonged intercourse (24).

Concurrent sexual relationships have emerged as an important topic in HIV and STI behavioral research. In a national survey, 11% of U.S. men reported having multiple sex partners concurrently (25), and risk included being unmarried, non-Hispanic Black or Hispanic, incarcerated in the previous year, had sex while drunk or high, had non-monogamous female partners, and ever engaged in sex with a man (25). Given the multiple STI-risk factors in this population, men with concurrent sexual partners present a significant risk to their female sex partners. Adimora and colleagues appropriately suggest that concurrent sexual relationships may contribute to the rapid dissemination of STI infection within networks. Among African Americans, concurrent sexual relationships have been reported as an independent risk factor for HIV infection (26).

Some evidence also exists for the importance of partner selection in the transmission of HIV and STIs. Although results failed to attain statistical significance, a trend was observed for urban young men diagnosed with asymptomatic Chlamydia or gonorrhea to select partners within the social network to be subsequently associated with an increased risk for STD reinfection (27). When placed in the context of the high levels of within-race partner selection observed in the African American community (28), and the disproportionate rate of incarceration which is 6 times more likely in Black than White men in the U.S.(29), this evidence implicates sexual networks as a significant factor explaining the higher rates of HIV infection observed in African American communities. Such network-level variables may be particularly relevant in drug-using populations due to their isolation and marginalization.

Non-IDU MSM

Young minority MSM have come into the forefront in the past decade as an especially high-risk population of interest in STI and HIV research (30). Many studies have reported higher prevalence and incidence of STIs (31) and HIV (32) in black MSM compared to non-Black MSM. In one multicenter longitudinal study of young men in five U.S. metropolitan areas, almost half of all Black MSM were HIV-positive and over two-thirds of Black MSM were unaware of their HIV status (33). Based on data from the Young Men s Health Survey sites in Baltimore and New York, the adjusted odds of prevalent HIV was 12.5 times greater in Black MSM and 8.6 times greater in MSM identifying as “Other/Mixed” than White MSM. Racially/ethnically disparate HIV rates are not likely due to differences in HIV risk behaviors such as having unprotected anal intercourse (32), needle sharing, frequency of risky sexual behavior or number of partners (31, 34), but reflect patterns of sexual partner selection that are limited to partners who may already be HIV infected.

Three hypotheses have been proposed to explain the higher rates of HIV in Black MSM for which empirical support exists. Higher rates of STI acquisition in Black MSM facilitate the disproportionate HIV seroconversion while others suggest that while Black MSM are just as likely to get tested for HIV, they may test less frequently and present at later stages of HIV than other MSM, thereby inadvertently transmitting HIV to their sex partners due to their higher HIV viral load (31). In addition, evidence suggests that because non-Hispanic Black MSM are more likely to have non-Hispanic Black sexual partners, sexual networks (32) and partner selection (35) may play a significant role in the HIV prevalence and incidence disparities. Among drug-using MSM, Rhodes et al. report a “distinct preference for having sex when high” and the experience in which “drug use rather than sexual orientation formed the core of personal identity” (36, p. 629). Based on focus groups and individual structured interviews, the researchers described the association of drug-using MSM with primarily other drug users, often MSM as well, limiting “contact with people who did not use drugs and the mainstream gay community” (36)(p. 629). These data and others have suggested that some MSM use drugs to replace intercourse and the “gay scene” (23, 36). However, among MSM, drug and alcohol use are strongly associated with being sexually active (37).

Circuit parties are often viewed as an escape from stigma and the HIV epidemic, associated with promotion of gay self-identity and self-expression (18). Crystal methamphetamine entered the circuit party scene in the 1990s following the earlier rampant use of cocaine, and amplified the unsafe sex due to the enhancement of libido, loss of time reference, and decreases in inhibition and control associated with methamphetamine use, often leading to prolonged sexual behavior with multiple partners. In the context of circuit parties where such disinhibition (sexual and otherwise) is encouraged, the risk of unsafe sex is high. Mattison et al. reported that 50% of MSM used alcohol and club drugs (e.g., ecstasy, ketamine, etc.) at circuit parties (38). Reported reasons for attending parties (such as “to have sex,” “to look and feel good,” and “to be uninhibited and wild”) suggest that sexual expectations that proceed unsafe sex may have more to do with subsequent risky behavior than the concurrent drug use and decreased inhibition (23, 38). Longitudinal studies of sexual behavior among men participating in this activity are needed to confidently determine the true risks involved.

Sildenafil, commonly prescribed to alleviate erectile dysfunction, is used illicitly, abused as a prescribed drug and sometimes taken in combination with club drugs to counteract the depressive effects of drug use on sexual behavior (39). Sildenafil has been associated with amphetamine use before or during sex in a sample of men (40), resulting in prolonged and sometimes aggressive sexual encounters. In addition, among MSM, Sildenafil is often taken in combination with amyl or butyl nitrates (poppers) and/or methamphetamine which are thought to relax sphincter muscles (41) and aid in achieving and maintaining an erection (42). Methamphetamine use, particularly among MSM, may be “emerging as a response to and a facilitator of HIV risk” (43).

Unpredictable combinations of drugs and intentions to “get out of it” (23, p. 298) have also been associated with unsafe sex at circuit parties. Twenty-eight percent of MSM engaged in unprotected anal sex during circuit parties which increased with the number of drugs used (39).

Heterosexual Female Non-IDUs

Illicit drugs have been associated with risky sexual behavior in women. Females injecting methamphetamine were more likely than non-methamphetamine injectors to engage in anal sex, to report sex with multiple partners, to engage in receptive needle-sharing and to share with more than one partner after adjusting for other factors (44). In a cross-sectional study of drug users in Long Beach, California, methamphetamine use was one of the strongest predictors of heterosexual anal sex (HAS) in women and those that engaged in HAS were more likely to have STIs than those that did not (45). Increased sexual activity and sex trade have also been reported for women using crack cocaine (23). Using heroin with sex partners is associated with higher acquisition for non-IDU women (10), which may be partially explained by the overlapping nature of female drug and sexual networks (16, 46). Alcohol intake has also been associated with elevated sexual HIV risk among women (47).

For women who do not inject drugs, their injecting partners pose a significant risk for acquiring HIV. In one national survey study, the effects of having an IDU partner on health and sexual risk behaviors varied by the nature of the sexual relationship (i.e., whether the woman had a single IDU partner, multiple partners or reported trading sex with multiple partners for drugs or money). Women with a single IDU partner were the least likely of the three groups to use a condom during sexual activity (48), as has been reported for monogamous men and drug using women.

Trading sex for money or drugs is also not uncommon among drug users, particularly among women using crack cocaine. In a prospective study of drug users, participants who smoked crack were most likely to be involved in trading sex for drugs or money and to have multiple sex partners (49). Similarly, stimulants are associated with an increased number of partners, decreased condom use, and more commonly reported exchange of sex for drugs or money (50). However, some evidence indicates that female drug users trading sex may use condoms with sex-trading clients but not with regular sex partners who may, in fact, impose as great a risk as paying clients (10).

Negotiating safe sex requires that a woman surmount gender roles imposed by society, her culture and in the relationship. In one study of low income, sexually-active women attending family planning and STD clinics in Miami, Hispanic women were significantly less comfortable than African American or White women talking to their partners about sex and condoms (51). In addition, women who did not participate in financial decision-making in the relationship were 90% less likely to use condoms consistently than women who made such financial decisions independently. For sex workers, particularly those selling sex for drugs, the challenge of negotiating condom use is even greater. Power issues, drug cravings, possible cognitive impairment, and lack of inhibition are serious barriers to condom negotiation in women trading sex for crack (52).

Managing sexual risk presents special hurdles for minority women, as high incarceration rates of minority men have lead to limited partner choices, especially in the inner-city environment, and may compromise acceptable levels of sexual risk (28). African American females also typically report having older sexual partners, partners whom are associated with greater odds of unprotected sex (53) and hence increased risk of STI and HIV transmission.

Non-IDU WSW

The majority of research on WSW focuses on IDUs. A recent investigation of the HIV risk behaviors of drug-using WSW indicates that the risk profiles of WSW-IDU and WSW may be similar (54). In one study, WSW were more likely than WSMO to trade sex, have an early (before age 15) sexual debut, and have multiple sexual partners. Other differences included a greater cocaine and crack dependence and greater likelihood of sex with MSM and sex with Hepatitis B or HIV-positive partners for females having sex with women. While risks for HIV acquisition would therefore appear to be increased among WSW relative to WSMO, there are no data that show a higher burden of HIV attributable to their sexual activity.

Conclusions

Epidemiologically, it is clear that drug use increases the risk of HIV and STI transmission. Parenteral drug use is much more efficient than indirect transmission in establishing viral infections, particularly HIV, among drug users. However, indirect transmission accounts for a substantial proportion of the risk in IDU populations and even more so among non-injection drug using populations through sexual risks. It is evident in reviewing the literature on drug use and HIV and STI that the assumptions researchers bring to their investigations (e.g., the lack of recognition among the research community that IDUs have a risk of sexual HIV acquisition) and the research methods and designs they employ to study these associations limit our knowledge. Researchers investigating sexual behavior in drug-using populations must continue to focus on traditional themes of elevated rates of unprotected sex, concurrency and partner selection factors, while incorporating emerging concepts of sexual risk outlined in this review.

Of importance, factors associated with differential sexual partner selection associated with overlaps among social and drug networks may lead to either risk or protection for HIV. While we have demonstrated that women with IDU partners are at elevated risk for HIV, they are only at risk if their partner has other concurrent partners. Similarly, risks are elevated in minority populations to the extent to which their sexual networks are broad or limited to same race/ethnicity. Partner selection choices may also reflect limited availability of available sexual partners, as has been seen among inner-city African American women, where a large proportion of the male population is or has been incarcerated, an acknowledged risk for HIV transmission.

Different drug-using populations differ in their epidemiologic risk factors and reflect the dominant culture of their group. MSM who are IDUs are very different from non-IDU drug users, and their HIV prevalence rates clearly show their risk. Nevertheless, MSM differ widely in terms of race/ethnicity, and Black MSM have a vastly increased HIV prevalence compared to their White counterparts. This may represent partner selection, the elevated stigma and discrimination towards MSM in the Black community, or the higher rate of bisexuality in this population.

It is also clear that differentiating risk associated with sexual practices from on-going drug use is especially difficult given the time-varying nature of both behaviors. Only with the use of longitudinal cohort studies can we effectively untangle these risks, but the varying nature of each behavior provides a number of challenges to researchers: how frequently should these behaviors be addressed, using which methods of elicitation, and with what degree of confidence can these sensitive practices be determined. It is this challenge that must be met if we are to determine the temporal ordering and magnitude of the risks imposed by sexual practices and drug use for HIV and STI.

Finally, special attention should be given to vulnerable groups of drug users, including the special issues noted among Black MSM, as well as neglected groups such as WSW, where current data are insufficient to determine whether this population is at elevated risk or not. Approaching populations who are both sexual minorities and race/ethnic minorities in a culturally sensitive and relevant manner is essential to determining the context of HIV risk as well as collaboratively developing approaches to risk reduction. Recent attention has also been given to the use of the terms MSM and WSW with consideration to variations in identity and social interactions of members within each group (56), a limitation of this paper recognized by the authors yet restricted by the specificity of current research. We need to focus on the edges of the risk continuum if we are to effectively respond to the neglected peoples most affected by HIV.