Sexual behavior and illicit drug use

Sexual behavior and illicit drug use

Ross, Michael W

Sexual behavior in drug users varies in association with the drug used, the drug subculture and setting, and the need to maintain the drug addiction. We review the literature on sexual behavior in drug-using subcultures, most of which centers on HIV risk reduction, and the drug-associated and economic pressures that influence sexual behavior. Most data available are on opioid (predominantly heroin) users and crack cocaine users. The recent phenomenon of the circuit party has led to investigation of the context in which drug use and sex have become the focus of large gay-oriented parties over long weekends. Sexual behavior in association with drugs in Western societies is important for its role in the spread of sexually transmissible pathogens, as well as injection-related pathogens.

Key Words: circuit party, crack, drugs, heroin, HIV, sex, STDs.

The relationships between illicit (illegal) drug use and sexual behavior are complex, not always direct, and present significant methodological difficulties (Leigh & Stall, 1993; Stall, McKusick, Wiley, Coates, & Ostrow, 1986). However, there has been increased attention paid to the interactions between drug use and sexual behavior, given the rise of the HIV pandemic in the past 2 decades. Five hypotheses may explain associations between drug use and sexual behavior. First, there may be a causal relationship between drug use and sexual behavior, with drugs acting as an aphrodisiac. Second both substance abuse and sexual behaviors may be significantly associated with a third variable, sensation seeking. Third, drugs may disinhibit previously controlled sexual behavior. Fourth, the social context in which drug use occurs may also encourage sex, and fifth, there may be multifactorial interrelationships (Ostrow, 2000). In this paper, we discuss the findings and interactions with regard to sexual behavior in illicit drug users, particularly with a view toward possible interventions to reduce the risk of transmission of HIV and other sexually transmissible pathogens. The physiologic effects of illicit drugs on behavior, however, are a major subject in their own right and will be dealt with only tangentially here when relevant to the data.

Relationships Between Sexual Identity and Drug Use

Sometimes, sexual activity can be seen as a side effect of the larger goal, drug acquisition. The question of such relationships between drug use and sexual behavior, particularly concerning variables relating to context and identity (e.g., as a drug user, as a means of survival, or as having a particular sexual orientation), can often best be answered by qualitative research determining the meanings of behaviors. In a multisite study of HIV risks in drug-using men who have sex with men (MSMs), Rhodes et al. (2000) described the findings of focus groups using the same methodology in five sites across the United States. Consistent findings emerged regarding the importance of the participant’s attachment to, and primary identification with, drugs rather than sex. In fact, the drug scene appeared to be the primary means of social orientation, rather than the “gay scene.” Some respondents expressed the view that in many cases, the drugs have replaced having intercourse, companionship, or a partner. In these cases, the primary drug was crack (smokeable freebase cocaine) or heroin: Male crack smokers may not be able to achieve an erection when they smoke crack, and heroin is known to depress sexual appetite. The fact that drug use was the primary organizational theme among these men led to some prevarication about sexual activities and sexual identity among a number of men who had sex with men. Many defined themselves as heterosexual and minimized or denied their sexual activities with other men. Labeling of self sexual identity was eschewed and most of the men had little contact with the “gay scene,” most of them meeting sexual partners in the course of their daily activities. For the heterosexually identified MSMs, it was clear that some sexual activities were not defined as “sex,” oral sex being one, for example. For some, sex was not defined as sex unless both partners reached orgasm.

Drugs for Sex and Sex for Drugs

The pervasiveness of exchange of sex for drugs or drugs for sex in Rhodes et al.’s (2000) study should not be underestimated. They concluded that almost all of the gay, transgender, homosexual, bisexual, and heterosexual participants had participated at some time in exchanges of male-male sex for drugs or money to buy drugs. Access to drugs, including exchanges involving drugs, sex, and money, was a dominant theme across the focus groups, whether participants noted pleasure, sexual stimulation, withdrawal, or craving as motivation. Drugs were central to drug-using MSM encounters, and such encounters rarely happened without drugs. The drugs may either spark a sexual excitement or spark the interest to seek a partner. Finding a sex partner might continue the process, with sex engaged in to access the drug, and money might justify the sex that occurred while using the drugs from the sexual encounter (Rhodes et al., 2000).

Baseman, Ross, and Williams (1999) studied the exchange of drugs for sex and sex for drugs in two predominantly African American neighborhoods in Houston, Texas. Based on street-intercept interviews, they found that 58% of men and 20% of women reported that they had bought sex for drugs or money, and 22% of men and half of the women reported they had sold sex for drugs or money. They concluded that in an economically disadvantaged milieu, drugs and sex form a currency of an underground economy: Drugs are a market currency and sex is a market commodity. Both drugs and money are well established as a means of exchange for sex. Baseman et al. emphasized that where a large underground economy exists, a sex market coexists with a crack market and the economic context of drug use (its status as a medium of exchange), as well as individual factors, must be taken into account in explaining the interactions between sex and drugs.

Impact of Specific Drugs on Behavior The links of drugs to specific impacts on sexual behavior are not always invariant: In Rhodes et al.’s (2000) study, individuals tended to prefer one drug in sexual contexts, but for the same drug, reports could vary from individual to individual in terms of its sexually related effects. For example, crack aroused some, yet left others uninterested in sex; it was commonly reported that crack caused sexual craving but eventually interfered with their sexual performance, and for others, heroin produced the same dichotomy of arousal in some cases and disinterest in others. Methamphetamine, however, was more generally reported to cause craving but not to interfere with performance. In the Californian section of Rhodes et al.’s study, methamphetamine and crack were reported to increase sexual desire, increase sexual stamina, and eliminate sexual inhibitions. It is useful to consider the variables of drug, mental set, and setting (Zinberg, 1984) as all contributing to the effect of a drug, but also to consider individual variation as a significant contributor as well. For this reason, it is difficult to unequivocally attribute a particular sexual impact to a specific drug.

Nor is it possible to assume, even in a single country, that drug use patterns within a circumscribed subgroup are similar. Sullivan, Nakashima, Purcell, and Ward (1998), in a study of HIV seropositive MSMs in the United States found that, comparing 12 different states and metropolitan areas in nearly 10,000 HIV seropositive MSMs, 33% reported alcohol use, 51% marijuana use, 31% noninjected cocaine use, 16% crack cocaine use, 13% injected cocaine use, and 8% each injected stimulants and used heroin, in the past 5 years. White MSMs were more likely than Hispanic MSMs to report use of hallucinogens, marijuana, nitrites, noninjected amphetamines, and diazepam, whereas African American MSMs were more likely than Hispanic MSMs to report the use of noninjected crack. Use of injected stimulants was more likely in White MSMs and those residing in the west versus the east.

Changes Across Time

Sexual behavior changes as a result of treatment or other interventions to reduce HIV-related risk in injecting drug users (IDUs) have been measured by several studies in several countries. Over an 18month period, Calsyn, Saxon, Wells, and Greenberg (1992) measured the sexual behaviors of 220 IDUs engaged in an AIDS prevention project and HIV testing and counseling in Seattle. There was a 70% followup rate, with those available for follow-up being older, longer in treatment, more likely to be female, and to have been recruited form methadone maintenance programs. The sample was primarily heterosexual. Significantly fewer men than women had multiple sex partners. The frequency of condom use significantly increased for men, but not women, from a baseline of 12% to a follow-up of 28%. The percentage of men using condoms for >90% of episodes of vaginal intercourse increased from 4% to 22%. Continuing unsafe sexual behavior was associated with exchanges of sex for drugs or money and with use of drugs to meet sexual needs for both men and women. For men, alcohol use to intoxication and intranasal cocaine were associated with unsafe sex; for women, all drugs except sedative hypnotics were associated with unsafe sex. Half the sample remained in treatment over the follow-up period, but only in women was treatment associated with reduced risk. These data indicate both that drug users are responsive to interventions, and that sex for drugs/money and drugs/money for sex are risk factors.

In a similar study in male IDUs in New York, Dolezal et al. (1999) compared 144 HIV seropositive and seronegative men over 4 years. They noted that reviewers have generally found that IDUs are more likely to report changes toward reduced risk in injecting behavior than sexual behavior in every study in which the two are compared (Booth & Watters, 1994; Des Jarlais, Friedman, Choopanya, Vanichseni, & Ward, 1992). Dolezal et al.’s sample was three quarters African American, and 40% of the sample were not available by the last visit (which were scheduled every 6 months). Thus, it may be that those least likely to change were those who were more likely lost to follow-up, biasing these data toward a higher rate of success. Substance abuse was common in this cohort: At baseline, 80% reported alcohol use with sex, and 93% reported drugs in conjunction with sex (most commonly methadone, heroin, and cocaine). The proportion using any drugs at the last visit declined to 80% (similar proportions were found for those lost to attrition). The continual sexual risk reduction shown over the 4 years for both HIV seropositive and seronegative men was attributable to increases in abstinence and monogamy and decreases in the frequency of unprotected sex. These changes took place in the absence of any systematic intervention (1988-1993) and were maintained over time. Nevertheless, there was a substantial amount of risk behavior, with over a third of the HIV seropositive and over half the HIV seronegative samples reporting unprotected sex at any given point of the study. For those completing the study and infected with HIV, as HIV disease progressed, abstinence and monogamy increased. However, it should not be assumed that risk behavior remains constant over time: Posner, Collins, Longshore, and Anglin (1996) found that Californian IDU respondents moved in all directions between high, medium, and low risk stages over a period of a year, and that there is apparently no orderly progression through stages to a low-risk end stage.

Beneficial change has not been reported in every study, however. In a Californian longitudinal study of young MSMs, Katz et al. (1998) found no changes in either HIV-associated risk behaviors or in injecting drug use behaviors in men aged 15-22. Nor can it be assumed that there is any relationship between reducing drug-associated HIV risk and sexually-associated HIV risk. Wodak, Stowe, Ross, Gold, and Miller (1995) studied over 1,200 injecting drug users (almost all heroin users) in Australia and found no significant relationship between sexual risk and injecting risk for HIV. Of their sample, 17% were at low risk for both, and 33% were unsafe both sexually and through injecting. The remaining half were unsafe on one or the other behavior. For sexual behavior, 33% were safe, and, for injecting behavior, 52% were safe. The lack of any significant relationship between risk in injecting and risk in sexual behavior suggests that, in this sample, there is not a general sensationseeking or risk-taking construct that accounts for risk behavior in both areas of potential risk.

Change in the sexual milieu may also influence sexual behavior. For example, where interventions have targeted people according to sexual orientation, sexual orientation interacts with drug use and HIV risk behavior. In a study of a large sample of Australian IDUs, Ross, Wodak, Gold, and Miller (1992) found that sexual orientation was the best predictor of condom use and safer sex. It was highest in the gay men and lowest in the heterosexually identified men, with bisexual men in an intermediate position; however, for numbers of sexual partners, HIV seroprevalence, and anal sex, the patterns were reversed. There were no differences across sexual orientation for either sex on injecting risk behaviors (consistent condom use for anal or vaginal sex). These data also suggest that sexual risk behavior is likely to be determined by the prevailing intervention program educational and informational context with regard to sexual risk, and that there is not necessarily any connection between sexual risk and drug-taking risk.

High rates of sexual risk are not necessarily characteristic of drug users entering treatment. In a sample of 165 opioid users entering methadone treatment in Baltimore, King et al. (1994) found a median of one sex partner, and only 16% reported any involvement in drugs for sex/money or sex/money for drugs. Patients involved in commercial sex had significantly higher condom use rates (42% vs. 15%). Reported rates of monogamy or sexual abstinence in King et al.’s study in the year prior to admission were 80% of males and 82% of females. These high rates may reflect the impact of opioids on sexual desire and functioning.

Sexual Partners of Drug Users

Sexual partners of drug users have rarely been studied and yet are an important part of the description of sexual behavior. Kotranski et al. (1998) reported on people who reported having IDUs as sexual partners and noted that over half had used cocaine in some form other than by injection in the previous 6 months and that three quarters had been in drug treatment in their lifetimes. For these partners, condom use was uncommon: Among females and males with one IDU partner, 30% and 38% respectively, and among those with five or more partners, 72% of females and 47% of males reported sometimes or always using condoms. Although almost all of the sample reported vaginal intercourse in the past 6 months, 22% of men reported anal sex with a female partner. In a study of the last and second-last sexual episode in over 1,200 predominantly heroin-injecting drug users in Sydney, Australia, Ross, Wodak, Miller, and Gold (1993) found that 75% of the partners of females were also IDUs, compared with a lower 59% for men. Homosexual contact for the last sexual contact was 13% for men and 12% for women. There was high agreement between partner characteristics (sexual orientation, gender) for the last and second-to-last sexual encounter. These data are similar to Italian data, where Sasse, Salmaso, Conti, and Rezza (1991) also found that women were 2.5 times more likely to have had only IDU sexual partners than men. This may relate to issues of stigma, sexual availability, and power, with IDU women being more heavily stigmatized and only able to attract other drug-using partners.

Using the same Sydney data set, Ross, Kelaher, Wodak, and Gold (1994) reported that both males and females, regardless of sexual orientation, reported more than half of their sexual contacts occurred under the influence of drugs. Drugs on which respondents were most likely to be intoxicated while having sex were heroin 61%, Cannabis 60%, alcohol 55%, amphetamines 28%, cocaine 24%, benzodiazepines 23%, methadone 17%, other opioids 10%, LSD 9%, and ecstasy (methylenedioxymethamphetamine) 9%. Although these data may not be widely generalizable, they do indicate that sex under the influence of drugs is likely to be high, at least in a predominantly opioid-injecting sample. Predictors of intoxicated sex also included being intoxicated while injecting, more widespread acceptance of injecting equipment from a broad range of others, sharing injection equipment more recently, having more sex partners, and being more likely to be paid for sex. Klee, Faugier, Hayes, Boulton, and Morris (1990), in a British sample, noted that 81% of their sample of IDUs had had intercourse in the past 6 months. This is important to note because both heroin and alcohol have been suggested to significantly depress libido and sexual activity as a result of long-term use, partly as a consequence of the pharmacological actions of the drug group. The available data do not support this, and in fact raise the possibility that stimulants may be used to counter the depressive actions of opioids. Self-medication may also occur to counter the effects of other drugs: Mansergh et al. (2001), interviewing MSMs who used illicit drugs in circuit parties, reported the use of Viagra at circuit parties to counter the effects of illicit drugs on sexual behavior. Although beyond the scope of this paper, we need to note that interactions between illicit and licit drugs are likely to be common: Fenaughty and Fisher (1998) noted that there is evidence of associations between alcohol use and risky sex among samples of IDUs and crack smokers.

Contexts of Drug Use

The “circuit party” or “dance party” phenomenon has been associated with the confluence of disinhibited drug use and sex in popular discourse, but the few studies of the phenomenon are equivocal with regard to risk. Lewis and Ross (1995) carried out a major qualitative analysis of gay circuit parties in Australia. They examined through extensive interviews the history, social context, drug use, and sexual behavior associated with the two biggest events in the gay circuit party scene, Mardi Gras and Sleaze Ball (described as the “Christmas and Easter for gay men”) in Sydney, Australia. Mardi Gras usually had over 17,000 gay men dancing in a sports pavilion and originally arose in the early 1980s from a combination of the political legitimization of the gay subculture and as a consequence of the HIV pandemic. Lewis and Ross found that the circuit parties were consistently described as a rite of social transformation in an atmosphere in which the twin taboos of death and sex could be dealt with. Parties were described by patrons as a healing and regeneration ritual in a community that constantly faces, outside the parties, stigmatization and the specter of AIDS and death. They noted that historically plagues and disasters were associated with such carpe them (“seize the day”) themes and that choreomania, dancing in the face of plague, was described in the 17th century and more recently. In fact, many participants described the parties as a “sacred time and space” in which reality was transformed and in which the music, light shows, and atmosphere were designed to enhance the effect of drugs-in this case, ecstasy and crystal methamphetamine, ketamine, and alcohol. Lewis and Ross’ data also suggest that under such circumstances, magical thinking regarding risks combined with drug use could lead to unsafe sexual behaviors. The circuit party played a major function in reinforcing the gay men’s identity through its role as a gay-controlled venue (participants were usually in their 20s or early 30s), provided a physically and psychologically safe atmosphere for expressing sexuality and for anonymity, and a place to forget the HIV pandemic while being able to express themselves. Thus, the broader psychological and historical contexts which may underlie the physical settings of drug use may be as important as the immediate environment in their promotion of safety of drug use and sexual behavior.

The sexual aspect of the parties was emphasized through the party themes, for example, the Bacchanalia, based upon the orgiastic ancient Greek rite named for Bacchus, god of ecstasy. Lewis and Ross (1995) described at length the polydrug combinations used by party patrons, the sexual behaviors associated with varying drug combinations, particularly the unsafe sex associated with being “out of it” (usually by combining too many drugs with unpredictable consequences). The data showed that drugs were taken to peak at 2 a.m. and following this time, sexual behavior in dark areas and toilets where the lights had been knocked out was commonplace. The circuit party was a place to reinforce identity. Stereotyped, eroticized fashion and the emphasis on fit, muscular bodies characterized these circuit parties and continues to do so. In particular, Lewis and Ross noted the importance of Zinberg’s (1984) concept of the interaction of drug, set, and setting in circuit parties and the need for interventions to be sensitive to, and integrated into, the party ethos and psychological function. Interventions that may reduce unsafe sex and promote safer drug use include attention to state-dependent learning in the drug intoxicated state, use of luminous and laser-lit messages, use of appropriate role models, such as disk jockeys, and timing of messages in relation to drug peaks.

Lewis and Ross (1995) also noted the introduction of a drug, crystal methamphetamine (“Ice”), into the circuit party scene during their study. The action of this drug as reported by party patrons was strongly aphrodisiac, and its effect in blocking memory of what the respondent had been doing for periods of hours. Impaired judgment, loss of control and heightened sexual arousal in the context of a circuit party were described by respondents as an extremely dangerous combination which had a strong possibility of leading to unsafe behavior.

More recent data on circuit parties are also equivocal regarding drugs use and sexual risk. Mattison, Ross, Wolfson, Franklin, and HNRC Group (2001) investigated over 1,000 gay circuit party patrons and found that drug use was high: More than half the patrons reported using alcohol, ecstacy, and ketamine at circuit parties in the past 12 months. Over a third reported using cocaine, crystal methamphetamine, nitrites, and marijuana in the same period, and over a quarter gamma hydroxybutyrate. Frequent use of ecstacy, ketamine, and nitrites was associated with unsafe sex, and the modal number of drugs used was four. Several reasons for attending circuit parties were significantly associated with reporting unsafe sex in the past 12 months: to have sex, to be uninhibited and wild, and to look and to feel good. These data suggest that the intention to have sex, possibly unsafe sex, may precede the event, and that drug use may be an accompaniment but not necessarily a cause of unsafe sex. Generally, unsafe sexual behavior reported was low: 15% in the past 12 months (including at parties). Unpredictable combinations of drugs may be partially associated with lack of safety: Mattison et al. found that only 10% of those using only one drug, compared with 26% using seven or eight drugs, were unsafe.

In a similar study, Mansergh et al. (2001) interviewed nearly 300 men who attended circuit parties in California and found that a quarter reported a drug “overuse” incident in the past year. Ecstasy, ketamine, crystal methamphetamine, and gamma hydroxybutyrate were the drugs reported in order of preference, and 28% of their sample reported having unprotected anal sex during the circuit-party weekend. As the number of drugs used increased, so did the probability of unprotected anal sex. These data are strikingly consistent across studies and suggest that intentions to get “out of it” and unpredictable combinations of drugs are both associated with unsafe sexual behavior. Sampling biases, however, make it difficult to assess the prevalence of risk behavior at circuit parties. Mattison et al. (2001) found that their sample were disproportionately White, well-educated, and affluent.

Sexually Transmissible Infections in Drug Users

Discussion of sexual risk in illicit drug users would not be complete without reference to STDs in addition to HIV. In their Sydney sample of predominantly opioid users, Ross, Wodak, Gold, and Miller (1991) found a high reported lifetime prevalence of STDs with both men (over a third) and women (over a half) reporting at least one STD. Using biological markers from blood samples collected from over 400 crack-house patrons sampled actually at the crack houses in Houston, Texas, Ross, Hwang, Leonard, Teng, and Duncan (1999) found evidence of widespread infections, with evidence of infection with syphilis in 13%, genital herpes in 61%, HIV in 12%, hepatitis B in 52%, and hepatitis C in 41%. They noted that the association of crack and sexual behavior has been extensively documented, with increased sexual activity hypothesized as due to heightened sexual arousal or decreased inhibitions, as well as the addictive nature of the drug, and the exchange of sex for money or drugs associated with this.

In the crackhouse data of Ross et al. (1999), sexual activity was reported significantly more frequently for women than for men, although this is consistent with a smaller number of women than men being involved in the crack house milieu. The most common sexual acts were vaginal sex, followed by oral sex (with almost all acts involving vaginal sex, and half of the acts oral sex for men, and almost all involving vaginal sex and two thirds oral sex for women: The anal sex rate was negligible). Condom use was low, with half of both men and women reporting never using condoms despite the multiple partners, and only a quarter reporting always using condoms. With such high rates of STD and viral hepatitis biological markers, and relatively low consistent condom use, the crack house population has all the characteristics of a core STD transmission (and reception) population.

The data suggest that in illicit drug using populations, STD treatment should be routine. Hwang et al. (2000), in a cross-sectional study of drug users in treatment, found prevalence of syphilis over 3%, of chlamydia nearly 4%, and of gonorrhea nearly 2%, with over 44% having markers for genital herpes infection. In drug-abusing populations, Hwang et al. suggested that rather than considering “dual diagnosis” (coexisting mental health and drug pathology), we should be considering “triple diagnosis” (drug pathology, mental health issues, and sexually transmissible infection pathology). The data on crack users, however, suggest that there is a close association between crack use and HIV infection, and thus the sexual behavior of those infected with HIV is as important to understand as the sexual behavior of those who are uninfected.

Crack and Sex

Determining how crack and sex interact relies heavily on qualitative research. Quantitatively, the association of crack and sex has been extensively documented (DeHovitz et al., 1994; Falck, Wang, Carlson, & Siegal, 1997; Minkoff et al., 1990). Williams (1992), in an ethnographic study of a crack house in New York city, noted that the exchange of sex for drugs among women is common. However, his research shows that “Most women act on the basis of individual choice, depending on their own desires and the opportunities that emerge in any given situation” (p. 112). Williams noted that men come to crack houses for sex, bringing drugs as the medium of exchange, but that for some of the people in the crack house, “sex is as much of a drug as the freebase they inhale. … The men say the drug stimulates the female; the women say the drug excites the male. It is clear from the frequency of sexual encounters that there is some truth to both assessments” (p. 115). He suggested that although oral sex is preferred, being less complicated and more manageable given the need to perform in any accessible place, it is difficult to use condoms given the power of the man (who usually has the money or the drugs) in the situation, and the fact that there is little caution displayed once intoxicated (a finding echoed by Falck et al., 1997). Williams (1992) observed that the practice of safe sex was rare in the crackhouses he studied.

Sexual Behavior in HIV Seropositive Drug Users

We know very little about the sexual behaviors of HIV positive (HIV+) crack smokers. The few data that do exist have been collected as part of epidemiological and prevention studies broadly targeting drug users or other at-risk populations (Booth, Kwiatkowski & Weissman, 1999; Novotna et al., 1999; Wilson et al., 1999). Despite the lack of research focusing on HIV+ crack smokers, developing effective programs to decrease the sexual risk behaviors of HIV+ crack smokers must be a research priority (Wilson et al., 1999). In many localities in the United States, crack smokers have higher rates of HIV infection than do drug injectors.

Despite limited research on the sexual risk behaviors of HIV+ crack smokers, we can extrapolate from studies of injection drug users or mixed samples of drug users. In a study of drug injectors, Friedman et al. (1999) found higher rates of consistent condom use among drug injectors who were HIV+ compared to those who were HIV negative. Even so, less than half of HIV+ drug injectors, 46%, consistently used condoms. Sociodemographic predictors of inconsistent condom use were the sexual partners’ age, closeness of the relationship between sexual partners, and normative support for unprotected vaginal sex (Higginbotham et al., 2000; Lansky, Nakashima & Jones, 2000; Myers et al., 1995; Singh et al., 1993). Kwiatkowski and Booth (1998) compared the sexual behaviors of HIV+ crack smokers and HIV+ drug injectors. Data were collected in five cities in the United States. Kwiatkowski and Booth found that almost half of the HIV+ drug users had engaged in unprotected sex in the 6 months before they were interviewed. Comparing the behaviors of HIV+ crack smokers and HIV+ injectors, the authors found that crack smokers were more likely to have had unprotected sex. To the authors this suggested that HIV+ crack smokers should be targeted for prevention activities. Metsch, McCoy, Lai, and Miles (1998), using data collected from a sample of injectors and crack smokers in Miami, Florida, reported similar findings.

Using a sample of 541 drug users residing in New York, Deren, Beardsley, Tortu, and Goldstein (1998) found that HIV+ drug users, including crack smokers, showed no significant change in unprotected sex after completing a behavior change intervention. Deren et al. concluded that HIV+ drug users, including crack smokers, decrease the level of risky sexual behavior after learning of their positive HIV status rather than after participation in a risk reduction program. Although encouraging in a fashion, the findings are inconclusive as to the reason for the apparent difference in risk levels. Even so, simple knowledge of HIV infection does not eliminate nor decrease risk sufficiently. In a recent study of HIV+ cocaine and opiate users entering a methadone maintenance clinic, Avants, Warburton, Hawkins, and Margolin (2000) found that a large percentage of those entering treatment had engaged in unprotected vaginal sex since learning that they were HIV+. The sample was mostly male and African American. Data were that 33 of 50 program entrants reported having had unprotected sex or reported that they had engaged in risky drug use behavior since learning of their HIV diagnosis. Approximately half reported having had unprotected sex with more than one partner since being diagnosed. Of the 35 who reported having had unprotected sex, 40% reported that they did not know the HIV status of their most recent sexual partner (or that their most recent sexual partner was negative). Although the majority surveyed said that they planned to use condoms in the next 6 months, only 21 of 35 had acquired condoms in the previous 6 months, and 9 had a condom with them at the time they were interviewed. Extrapolating from these studies, we can conclude that knowledge of an HIV diagnosis is insufficient to result in consistent or sustained risk reduction behavior. As these studies demonstrate, HIV+ drug users continue to engage in sexual behaviors that put themselves at risk for re-infection with resistant strains of HIV and other STDs, and to put others at risk for HIV infection.

Condom Use in Drug Users

In lieu of an effective vaccine or complete abstinence from sex, consistent condom use remains the most effective means for preventing the sexual transmission of HIV. Results of a meta-analysis examining the efficacy of condom use in reducing the probability of HIV transmission were that consistent condom use by HIV+ persons reduces the risks of transmission by as much as 69% (Johnson, 1994; Weller, 1993). As Heitman and Ross (1999) noted, the successful treatment of HIV+ persons is not just a matter of appropriate medication and compliance with dosing schedules. Successful treatment is dependent on the adherence of the patient to safer sex or safe drug use practices, because the question is not just the development of drug-resistant strains of the virus, but their transmission to others.

Achieving consistent condom use with HIV+ crack cocaine smokers may have to take place within a reality of continued crack use. Crack is a powerfully self-rewarding and addictive substance. Smoking the drug produces a strong sense of euphoria, heightened feelings of mental or physical agility, or other feelings of mood elevation (McCoy & Inciardi, 1995). Crack smokers rarely partake of a single hit or dose of the drug. Rather, users smoke crack in hinging cycles, smoking as much as resources allow for up to 3 or 4 days, then crashing after the binge (Inciardi, Lockwood, & Pottieger, 1993; Williams, 1992). With regard to HIV infection, crack smoking has pharmacological and sociological effects on sexuality. Most notably, crack is reputed to cause hypersexuality. Users often experience reduced sexual inhibition from smoking crack, exceeding that which can be achieved using Valium, opium derivatives, or alcohol. For some new users, the disinhibiting effect of the drug results in more extreme sexual orgasms than would otherwise be experienced, reinforcing other positive effects. However, long-term use of crack leads to sexual dysfunction in both men and women, particularly decreased sexual desire. For some men, smoking results in intense sexual desire accompanied by impotence (Inciardi, Lockwood, & Pottieger, 1993). For individuals who have become dependent on crack, deprivation of the drug results in feelings of depression and extreme craving. These smokers are often compulsively motivated to obtain and use the drug, sometimes going to extreme measures to do so. Among the extreme measures that are relevant to the transmission of HIV is the exchange of sex for crack.

Crack use may also have an effect on the psychological functioning (including cognitive impairment) of some users. Although it is unclear whether crack use causes psychological impairment or whether people who are impaired are more likely to use crack, it is apparent that crack use interacts with impaired psychological functioning to increase risky sexual behaviors (Compton, Cooler, Spitznagel, Ben-Abdallah, & Gallagher, 1998). At the very least, crack use may affect a smoker’s sexual expectations, increasing the likelihood of having sex without a condom (McKirnan, Vanable, Ostrow, & Hope, 2001; Ostrow, 2000; Seage et al., 1998). If, for example, an individual expects that smoking crack will be sexually disinhibiting, then disinhibited behavior will likely follow. Disinhibited sexual behavior may be more likely and extreme for crack smokers experiencing feelings of depression, hostility, or anxiety.

Entry into drug treatment and cessation of crack smoking would significantly reduce the risk of HIV transmission by infected users. However, crack smokers are less likely to benefit from a single episode of drug treatment than are users of other drugs. Hser, Joshi, Anglin, and Fletcher (1999) found that pretreatment crack use is strongly related to negative treatment outcomes. After treatment, most crack smokers return to using the drug and engaging in sexual behaviors in high risk sexual settings (Ross et al., 1999). Hser et al. also found that crack smokers who repeatedly entered and stayed in treatment were likely to eventually achieve abstinence. Therefore, the authors recommended that multiple treatment episodes be considered a normal part of treatment for crack addiction. The importance of this conclusion for HIV prevention is that drug treatment, although an important part of a prevention regimen, is not a panacea that will significantly lower the risk of HIV infection among crack smokers. Given a cyclic process of recovery, crack smokers will remain at risk through several treatment episodes. Therefore, it is critically important that HIV risk reduction be a part of the between treatment periods.

Possible Biological Interactions

There may be a biological explanation for the close association between the crack and HIV epidemics (Petry, 1999). Bagasra and Pomerantz (1993) observed that HIV replication in vitro significantly increases in peripheral blood mononuclear cells in the presence of cocaine metabolites. Although the exact effect of cocaine on HIV replication in vivo has not been established, Bagasra and Pomerantz concluded that in vitro, epidemiological, and animal studies strongly suggest that cocaine is a cofactor in HIV infection. Cocaine and other substances cause immunomodulations that support an environment in which replication of HIV is more efficient. Thus, cocaine’s immunomodulatory effects significantly increase the risk of infection after exposure to the virus and possibly increase viral shedding in those who are HIV+. Furthermore, Bagasra and Pomerantz suggested that the presence of cocaine and alcohol may have an interactive effect on HIV expression. Although not widely investigated, most crack smokers abuse alcohol (Booth et al., 1999; Fenaughty & Fisher, 1998; Petry, 1999) largely in an effort to moderate cravings. As Jewell and Shiboski (1993) warned, any factor or combination of factors (including alcohol) that influence susceptibility to an infectious agent will have a profound effect at the individual level and on the distribution of a pathogen in society as a whole.

Nonuse of condoms by not-in-treatment drug users is the norm (Calsyn et al., 1992; Desenclos, Papaevangelou, & Ancelle-Park, 1993). Initiation and maintenance of consistent condom use has proven to be extremely resistant to change (Cottler et al., 1998), even among HIV+ drug users (Deren et al., 1998). Published results of trials examining interventions to reduce the sexual risk of HIV transmission among drug users are discouraging, yet provide directions for the development of risk reduction programs.

Using a sample of 684 not-in-treatment drug users, Kotranski et al. (1998) found that users participating in a risk reduction intervention decreased the number of sexual partners from baseline to follow-up and that 37% engaging in risky sexual practices at baseline initiated or increased condom use during vaginal sex. However, analyses showed no significant differences between drug users participating in the NIDA Standard Risk Reduction Protocol (Coyle, 1993) or a standard plus intervention. Using a sample of 715 crack cocaine smokers, Cottler et al. (1998) replicated Kotranski et al.’s findings with a sample of crack smokers. Cottler et al. examined the number of sexual partners and rates of condom use after participating in a standard or a standard plus risk-reduction intervention. Analysis showed that 75% of the sample reduced the number of sexual partners. Cottler et al. also found that the proportion of the sample reducing their sexual risk, defined as having no sex or consistently using condoms during vaginal sex, increased from approximately 15% at baseline to 20% 6 months after intervention. However, there were no differences between those assigned to the standard or the standard plus interventions. Cottler et al. also reported that those in the standard plus did not improve condom use as much as did those in standard group. Given these results, Cottler et al. concluded that drug users are willing to reduce the number of sexual partners, but unwilling to initiate consistent condom use.

Kwiatkoski, Stober, Booth, and Zhang (1999) examined the sexual risk behaviors, defined as sex without a condom, of 3,357 heterosexual, sexually active drug users participating in a multisite risk reduction study. The authors found that there was a marginal, but significant increase in condom use between intake and follow-up. Condom use increased from 15% of sexual encounters at baseline to 22% 6 months later. Slightly over a quarter of the sample reported more condom use after participating in an intervention, but 71% reported no change or less condom use. An investigation of factors predicting increased condom use showed that HIV+ status, being single with multiple partners, or engaging in money for drugs or sex exchanges predicted increased condom use. Even so, the authors noted that only half of drug users who were HIV+ increased condom use. The authors concluded that increasing the amount (time) of the intervention may be what is needed to increase condom use.

Robles et al. (1998) supported this conclusion. Robles et al. found that an eight-session risk reduction intervention significantly increased condom use compared to a standard intervention. Using a sample of 1,004 drug users residing in areas around San Juan, Puerto Rico, of whom 20% were crack smokers, Robles et al. found that the use of condoms during vaginal sex increased from 26% prior to intervention to 37% after intervention. In a multivariate analysis of increased condom use, the authors found that being HIV+, having a diagnosis for an STD, and participation in the standard plus intervention predicted increased condom use, while having a steady sex partner significantly decreased the odds of condom use. The odds of condom use during vaginal sex by drug users who had an STD diagnosis were 27 times greater than the odds of condom use by those who did not have an STD diagnosis. The odds of condom use by HIV+ drug users were five times greater than the odds of HIV negative drug users. The odds of condom use by drug users participating in the standard plus group were two times greater than the odds of condom use by those who were in the standard protocol. Thus, awareness of HIV/STD status may be an important adjunct to behavioral interventions in promoting condom use.

Although measuring condom use as increases in condom use by drug users after intervention produced a statistically significant result in three of the four studies, condom use was not measured as consistent use in any of the studies. If consistent condom use had been the evaluation criteria, pre-/posttest results would have been significantly less positive. If it is assumed that HIV transmission is a no-error event, which is to say that any instance of unprotected vaginal or anal sex can result in HIV infection, then the appropriate standard for evaluating condom use is 100% or consistent use. For studies examining condom use among MSMs, consistent use has been the standard (Ross & Kelly, 2000; Stall & Purcell, 2000), although modification of this high standard has sometimes been recommended for condom use with primary partners.

Describing the rates of condom use or the characteristics associated with condom use does not explain the behavior, nor does it provide the basis for developing interventions to increase adoption or maintenance of sexual risk reduction. Borrowing from theories developed to explain smoking cessation, most investigators now examine the adoption of risk reduction behavior as a change process, rather than a discrete change event (Bowen & Trotter, 1995). For example, Bowen (1996) examined the effect of an intervention based on concepts from several psychosocial models of rational behavior change. Using a sample of 78 injection drug users and crack smokers, Bowen assessed the effect of a face-to-face risk reduction intervention on condom use. Bowen found that intention to use condoms, as measured by a stage of change measure, is directly influenced by higher levels of condom use assertiveness, a self-efficacy measure. Condom use assertiveness, in turn, is directly affected by an intention to begin using condoms in the next 6 months, positive attitudes toward condom use, and having multiple sexual partners. Bowen concluded that her findings were consistent with the models proposed by Social Cognitive Theory and the Theory of Planned Behavior.

Even though there appears to be a consensus as to what constitutes a successful condom use intervention, the relative importance of specific psychosocial precursors is still being assessed. Especially in dispute is the importance of situational determinants of condom use. For example, most psychosocial models of condom use have affective attitudinal components. Researchers have confirmed that, with certain partners and in certain sexual situations, condom use is almost wholly the result of a situationally determined affective cognitive process (Ross & Ferreira-Pinto, 2000). For example, Aron, Paris, and Aron (1995) found that individuals in love experience reduced levels of self-monitoring. Reduced self-monitoring is the opposite of what is desired if condom use is to be increased. Using a sample of 155 mostly African American MSM, Marks, Bingman, and Duval (1998) found that affective states, mediated by alcohol or drug use, are significantly associated with unprotected anal sex. Along a slightly different line, Kelly and Kalichman (1998) found that only the affective reinforcement value of unsafe sex and substance use accounted for levels of condom use in a sample of 297 HIV negative MSM. These findings strongly suggest that affective aspects of sexuality must be considered in developing sexual risk reduction interventions (Ostrow, 2000; Ross & Ferreira-Pinto, 2000).

It is clear that the sexual behavior of drug users has become an important consideration in the spread of not only HIV but also of other sexually transmissible pathogens. As is apparent, there are frequently significant interactions between drugs and the contexts (including physical, economic, and subcultural) within which drugs are taken, and sexual behaviors. Thus, sexual behavior in drug users must be considered above and beyond the physiological impact of the drugs used. The results of this review reinforce the importance of both qualitative and quantitative data to understand the influence of drug, individual, subculture, economy, and setting on the sexual behavior of drug users.

References

Aron, A., Paris, M., & Aron, E. (1995). Falling in love: Prospective studies of self-concept change. Journal of Personality and Social Psychology, 69, 1102-1112.

Avants, S., Warburton, L., Hawkins, K, & Margolin, A. (2000). Continuation of high-risk behavior by HIV-positive drug users: Treatment implications. Journal of Substance Abuse Treatment, 19, 15-22.

Bagasra, O., & Pomerantz, R. (1993). Human immunodeficiency virus type 1 replication in peripheral blood mononuclear cells in the presence of cocaine. Journal of Infectious Diseases, 168,1157-1164.

Baseman, J., Ross, M., & Williams, M. (1999). Sale of sex for drugs and drugs for sex: An economic context of sexual risk behavior for STDs. Sexually Transmitted Diseases, 26, 444-449.

Booth, R., Kwiatkowski, C., & Weissman, G. (1999). Health-related utilization and HIV risk behaviors among HIV infected injection drug users and crack smokers. Drug Alcohol Dependence, 55, 69-78.

Booth, R., & Watters, J. (1994). How effective are risk-reduction interventions targeting injecting drug users? AIDS, 8, 1515-1524.

Bowen, A. (1996). Predicting increased condom use with main partners: Potential approaches to intervention. Drugs and Society, 9, 57-74.

Bowen, A., & Trotter, R. (1995). HIV risk in IV drug users and crack smokers: Predicting stage of change for condom use. Journal of Consulting and Clinical Psychology, 63, 238-248.

Calsyn, D., Saxon, A., Wells, E., & Greenberg, D. (1992). Longitudinal sexual behavior changes in injecting drug users. AIDS, 6, 1207-1211.

Compton, W., Cottler, L., Spitznagel, E., Ben-Abdallah, A., & Gallagher, T. (1998). Cocaine users with antisocial personality improve HIV risk behaviors as much as those without antisocial personality. Drug & Alcohol Dependence, 49, 239-247.

Cottler, L., Compton, W, Ben-Abdallah, A., Cunningham-Williams, R., Abram, F., Fichtenbaum, C., et al. (1998). Peer-delivered interventions reduce HIV risk behaviors among out-of-treatment drug abusers. Public Health Reports, 113, 31-41.

Coyle, S. (1993). The NIDA HIV Counseling and Education Intervention Model (NIH Publication No. 93-3580). Rockville, MD: U.S. Department of Health and Human Services, National Institute on Drug Abuse.

DeHovitz, J., Kelly, P., Feldman, J., Sierra, M., Clarke, L., Bromberg, J., et al. (1994). Sexually transmitted diseases, sexual behavior, and cocaine use in inner city women. American Journal of Epidemiology, 140, 1125-1134.

Deren, S., Beardsley, M., Tortu, S., & Goldstein, M. (1998). HIV serostatus and changes in

risk behaviors among drug injectors and crack users. AIDS & Behavior, 2, 171-176. Desenclos, J., Papaevangelou, G., & Ancelle-Park, R. (1993). Knowledge of HIV serostatus

and prevention behaviour among European injecting drug users. AIDS, 7, 1371-1377. Des Jarlais, D., Friedman, S., Choopanya, K., Vanichseni, S., & Ward, T. (1992). International epidemiology of HIV and AIDS among injecting drug users. AIDS, 6, 1053-106.

Dolezal, C., Meyer-Bahlburg, H., Liu, X., Ehrhardt, A., Exner, T., Rabkin, J., et al. (1999). Longitudinal changes in sexual risk behavior among HIV+ and HIV- male injecting drug users. American Journal of Drug and Alcohol Abuse, 25, 281-303.

Falck, R., Wang, J., Carlson, R., & Siegal, H. (1997). Factors influencing condom use among heterosexual users of injection drugs and crack cocaine. Sexually Transmitted Diseases, 24, 204-210.

Fenaughty, A., & Fisher, D. (1998). High risk sexual behavior among drug users: The utility of a typology of alcohol variables. Sexually Transmitted Diseases, 25, 38-43.

Friedman, S., Chapman, T., Perlis, T., Sotheran, J., Rockwell, R., Paone, D., et al. (1999). Modulators of “Activated Motivation”: Event-specific condom use by drug injectors who have used condoms to prevent HIV/AIDS. AIDS & Behavior, 3, 85-98.

Heitman, E., & Ross, M. (1999). Ethical issues in the use of new treatments for HIV. In D. Ostrow & S. Kalichman (Eds.), Mental health and behavioral issues in new HIV therapies (pp. 113-135). New York: Plenum.

Higginbotham, S., Holmes, R., Stone, H., Beil, J., Datu, G., Costa, S., et al. (2000). Adoption of protective behaviors among persons with recent HIV infection and diagnosis-Alabama, New Jersey, and Tennessee, 1997-1998. Morbidity Mortality Weekly Report, 49, 512-515.

Hser Y., Joshi, V., Anglin, D., & Fletcher, B. (1999). Predicting posttreatment cocaine abstinence for first-time admissions and treatment repeaters. American Journal of Public Health, 89, 666-667.

Hwang, L., Ross, M., Zack, C., Bull, L., Rickman, C., & Holleman, M., et al. (2000). Prevalence of sexually transmitted infections and associated risk factors among populations of drug abusers. Clinical Infectious Diseases, 31, 920-926.

Inciardi, J., Lockwood, D., & Pottieger, A. (1993). Women and crack-cocaine. New York: Macmillan.

Jewell, N., & Shiboski, S. (1993). The design and analysis of partner studies of HIV transmission. In D. Ostrow & R. Kessler (Eds.), Methodological issues in AIDS behavioral research (pp. 291-342). New York: Plenum.

Johnson, A. (1994). Condoms and HIV transmission. New England Journal of Medicine, 331, 391-392.

Katz, M., McFarland, W, Guillin, V., Fenstersheib, M., Shaw, M., Kellogg, T., et al. (1998). Continuing high prevalence of HIV and risk behaviors among young men who have sex with men: The young men’s survey in the San Francisco bay area in 1992 to 1993 and in 1994 to 1995. Journal of Acquired Immune Deficiency Syndromes and Human Retrovirology, 19, 178-181.

Kelly, J., & Kalichman, S. (1998). Reinforcement value of unsafe sex as a predictor of condom use and continued HIV/AIDS risk behavior among gay and bisexual men. Health Psychology, 17, 328-335.

King, V, Brooner, R., Bigelow, G., Schmidt, C., Felch, L., & Gazaway, P. (1994). Condom use rates for specific sexual behaviors among opioid abusers entering treatment. Drug and Alcohol Dependence, 35, 231-238.

Klee, H., Faugier, J., Hayes, C., Boulton, T., & Morris, J. (1990). Factors associated with risk behavior among injecting drug users. AIDS Care, 2, 133-145.

Kotranski, L., Semaan, S., Collier, K., Lauby, J., Halbert, J., & Feighan, K. (1998). Effectiveness of an HIV risk reduction counseling intervention for out-of-treatment drug users. AIDS Education and Prevention, 10, 19-33.

Kwiatkowski, C., & Booth, R. (1998). HIV-seropositive drug users and unprotected sex. AIDS & Behavior, 2, 151-159.

Kwiatkoski, C., Stober, D., Booth, R., & Zhang, Y (1999). Predictors of increased condom use following HIV intervention with heterosexually active drug users. Drug & Alcohol Dependence, 54, 57-62.

Lansky, A., Nakashima, A., & Jones, J. (2000). Risk behaviors related to heterosexual transmission from HIV-infected persons. Sexually Transmitted Diseases, 27, 483-489. Leigh, B., & Stall, R. (1993). Substance abuse and risky sexual behavior for exposure to

HIV. American Psychologist, 48, 1035-1045.

Lewis, L., & Ross, M. (1995). A select body: The gay dance party subculture and the HIV/AIDS pandemic. London: Cassell.

McCoy, C., & Inciardi, J. (1995). Sex, drugs, and continuing spread of AIDS. Los Angeles: Roxbury.

McKirnan, D., Vanable, P., Ostrow, D., & Hope, B. (2001). Expectancies of sexual “escape” and sexual risk among drug and alcohol-involved gay and bisexual men. Journal of Substance Abuse, 13, 137-154.

Mansergh, G., Colfax, G., Marks, G., Rader, M., Guzman, R., & Buchbinder, S. (2001). The circuit party men’s health survey: Findings and implications for gay and bisexual men. American Journal of Public Health, 91, 953-958.

Marks, G., Bingman, C., & Duval, T (1998). Negative affect and unsafe sex in HIV-positive men. AIDS & Behavior, 2, 89-99.

Mattison, A., Ross, M., Wolfson, T., Franklin, D., & HNRC Group (2001). Circuit party attendance, club drug use and unsafe sex in gay men. Journal of Substance Abuse, 13, 1-8.

Metsch, L., McCoy, C., Lai, S., & Miles, C. (1998). Continuing risk behaviors among HIV seropositive chronic drug users in Miami, Florida. AIDS Behavior 2, 161-170.

Minkoff, H., McCalla, S., Delke, I., Stevens, R., Salween, M., Feldmanet, J., et al. (1990). The relationship of cocaine use to syphilis and human immunodeficiency virus infections among inner city parturient women. American Journal of Obstetrics and Gynecology, 163, 521-526.

Myers, T., Millson, M., Rigby, J., Kurtz, R., Jackson, E., Orr, K, et al. (1995). A comparison of the determinants of safe injecting and condom use among injecting drug user. Addict, 90, 217-226.

Novotna, L., Wilson, T., Minkoff, H., McNutt, L., DeHovitz, J., Ehrlich, L, et al. (1999). Predictor of risk-taking consequences of drug use among HIV-infected women. Journal Acquired Immune Deficiency Syndromes and Human Retrovirology, 20, 502-507.

Ostrow, D. (2000). The role of drugs in the sexual lives of men who have sex with men: Continuing barriers to researching this question. AIDS & Behavior 4, 205-219.

Petry, N. (1999). Alcohol use in HIV patients: What we don’t know may hurt us. International Journal of STD & AIDS, 10, 561-570.

Posner, S., Collins, L., Longshore, D., & Anglin, M. (1996). The acquisition and maintenance of safer sexual behaviors among injection drug users. Substance Use and Misuse, 31, 1995-2015.

Rhodes, F., Deren, S., Cooler, L., Siegal, H., Stark, M., & Reich, W. (2000). A multisite study of HIV risks in drug-using men who have sex with men: Focus group summaries. Long Beach, CA: California State University Press.

Robles, R., Marrero, C., Matos, T., Colon, H., Finlinson, H., Reyes, J., et al. (1998). Factors associated with changes in sex behaviour among drug users in Puerto Rico. AIDS Care, 10,329-338.

Ross, M., & Ferreira-Pinto, J. (2000). Toward a public health of situations: The re-contextualization of risk. Cadernos de Saade Publica, 16, 59-71.

Ross, M., Hwang, L., Leonard, L., Teng, M., & Duncan, L. (1999). Sexual behavior, STDs and drug use in a crack house population. International Journal of STD and AIDS, 10, 224-230.

Ross, M., Kelaher, M., Wodak, A., & Gold, J. (1994). Predictors of intoxicated sex in injecting drug users. Journal of Addictive Diseases, 13, 69-80.

Ross, M., & Kelly, J. (2000). Interventions to reduce HIV transmission in homosexual men. In J. Peterson & R. DiClemente (Eds.), Handbook on HIV prevention (pp. 201216). New York: Kluwer/Plenum.

Ross, M., Wodak, A., Gold, J., & Miller, M. (1991). Sexually transmissible diseases in injecting drug users. Genitourinary Medicine, 67, 32-36.

Ross, M., Wodak, A., Gold, J., & Miller, M. (1992). Differences across sexual orientation on HIV risk behaviors in injecting drug users. AIDS Care, 4, 139-148.

Ross, M., Wodak, A., Miller, M., & Gold, J. (1993). Sexual partner choice in injecting drug users from a “critical incident” measure: Its implications for assessing HIV spread. Sexological Review, 1, 77-92.

Sasse, H., Salmaso, S., Conti, S., & Rezza, G. (1991). British Journal of Addiction, 86, 1083-1089.

Seage, G., Mayer, K, Wold, C., Lenderking, W., Goldstein, R., Cai, B., et al. (1998). The social context of drinking, drug use, and unsafe sex in the Boston Young Men Study. Journal of Acquired Immune Deficiency Syndromes & Human Retrovirology, 17, 368-375.

Singh, B., Koman, J., Catan, V., Souply, K, Birkel, R., & Golaszewski, T. (1993). Sexual risk behavior among injection drug-using human immunodeficiency virus positive clients. International Journal of the Addictions, 28, 735-747.

Stall, R., McKusick, L., Wiley, J., Coates, T., & Ostrow, D. (1986). Alcohol and drug use during sexual activity and compliance with safe sex guidelines for AIDS: The AIDS behavioral research project. Health Education Quarterly, 13, 359-371.

Stall, R., & Purcell, D. (2000). Intertwining epidemics: A review of research on substance use among men who have sex with men and its connection to the AIDS epidemic. AIDS & Behavior 4, 181-192.

Sullivan, P, Nakashima, A., Purcell, D., & Ward, J. (1998). Geographic differences in noninjection and injection substance use among HIV-seropositive men who have sex with men: Western United States versus other regions. Journal of Acquired Immune Deficiency Syndromes and Human Retrovirology, 19, 266-273.

Weller, S. (1993). A meta-analysis of condom effectiveness in reducing sexually transmitted HIV. Social Science Medicine, 36, 341-346.

Williams, T (1992). Crackhouse: Notes from the end of the line. New York: Penguin. Wilson, T., Massad, L., Riester, K., Barkan, S., Richardson, J., Young, M., et al. (1999). Sexual, contraceptive, and drug use behaviors of women with HIV and those at high risk for infection: Results from the Women’s Interagency HIV Study. AIDS, 13, 591-598.

Wodak, A., Stowe, A., Ross, M., Gold, J., & Miller, M. (1995). HIV risk exposure of injecting drug users in Sydney. Drug and Alcohol Review, 14, 213-222.

Zinberg, N. (1984). Drug, set and setting: The basis for controlled intoxicant use. New York: Yale University Press.

Michael W. Ross and Mark L. Williams

WHO Center for Health Promotion and Prevention Research, School of Public Health, University of Texas

Parts of this research were supported by a grant from the Centers for Disease Control to the first author and a grant from the National Institute on Drug Abuse to the second author. The opinions expressed herein are solely those of the authors. Correspondence concerning this article should be addressed to Michael W. Ross, WHO Center for Health Promotion and Prevention Research, School of Public Health, University of Texas, PO. Box 20036, Houston TX 77225, USA. (MRoss@sph.uth.tmc.edu)

Copyright Society for the Scientific Study of Sex 2001

Provided by ProQuest Information and Learning Company. All rights Reserved