International AIDS Conference hears how drugs, sex and HIV interlock

This article is more than 12 years old. Click here for more recent articles on this topic

Two sessions at the 19th International AIDS Conference (AIDS 2012) in Washington looked at the way injecting drug use, recreational drug use and unsafe sex interlock to multiply HIV risk, and how new outbreaks of injecting drug use continue to spawn localised HIV epidemics.

Harm reduction in Tanzania

One study looked at the work done by the NGO Médecins du Monde (MdM) in Dar es Salaam, Tanzania. Injecting drug use in Africa seems like a relatively new phenomenon, but has in fact been around since white soluble heroin arrived in 1998. There are an estimated 50,000 people who inject drugs (PIDs) in Tanzania and a large community of them in Temeke, a poor urban area with a highly visible drug-use problem.

In 2011, MdM followed a qualitative and quantitative research study with a needle and syringe exchange project, involving the training of seven peer educators. Tanzania is virtually ‘virgin territory’ for harm-reduction initiatives, MdM’s Mark Stoové told the conference, but the harm-reduction model has been embraced by local and national government and the ‘Temeke model’ rolled out in other regions.

The quantitative research looked at 267 PIDs, of whom 36 (13%) were women. They had been injecting heroin for a median of five years and using it for ten: before injecting they had been smoking it. Worryingly, though, in people under 24, the time from smoking to injecting was only two years.


harm reduction

Harm reduction is a set of practical strategies and ideas aimed at reducing negative consequences associated with drug use (including safer use, managed use and abstinence). It is also a movement for social justice built on a belief in, and respect for, the rights of people who use drugs.

risky behaviour

In HIV, refers to any behaviour or action that increases an individual’s probability of acquiring or transmitting HIV, such as having unprotected sex, having multiple partners or sharing drug injection equipment.

multivariate analysis

An extension of multivariable analysis that is used to model two or more outcomes at the same time.


Quantitative research involves precise measurement and quantification of data, using methods like clinical trials, case-control studies, longitudinal cohorts, surveys and cost-effectiveness analyses.


Affecting a specific body site, organ or system.

HIV prevalence was 30% in men and 67% in women (24 out of 36) and, very unusually for a PID population, hepatitis C prevalence was lower – it was 28% in both sexes, though while only 52% of hepatitis C-positive men were co-infected with HIV, all 10 women with hepatitis C also had HIV.

Few of the PIDs reported needle-sharing, though 17% (23% of those who were HIV positive) shared other equipment, but sharing was much more common in a subgroup who injected in groups, usually in a specific geographical location like a building site; these open-air injecting places are known locally as maskani (camps). In maskani users, 15% used unsterilised needles at last injection, compared with 4% in other locations.

Over half of PIDs reported recent unprotected sex, leading to concerns that without scale-up of harm-reduction services, prevalence in Tanzania, currently running at about 6%, could increase again.

Contrasting epidemics in Russia and India

Russia is often cited as a country with a PID-driven epidemic, but in fact features a multitude of local mini-epidemics at all stages of growth. Presenter Kseniya Eritsyan told the conference that HIV prevalence in drug users ranges from 2.6% to 65% in different cities. She contrasted the situation in two cities, Ivanovo near Moscow and Novosibirsk in Siberia, amongst both drug users (PIDs) and their partners (PPIDs). These partners were both regular and casual partners, recruited by the PIDs themselves using contact slips.

Drug users tended to be male and their partners female, though by no means exclusively: about a quarter of drug users in Ivanovo and a third in Novosibirsk were women. In some ways the users and partners, and the groups in the two cities, were similar: their median age was in all groups about 25. In other ways, however, they were very different and showed that, while Ivanovo has a fairly settled drug-using population, in Novosibirsk there is a newer and rapidly expanding epidemic of HIV that comprises both PIDs and their partners.

HIV prevalence in Ivanovo was 34% in users (though only 10% of PIDs had ever tested) and 38% in their contacts. The majority of partnerships (72%) were regular ones and 76% of the partners knew their contact was a PID. In Novosibirsk, HIV prevalence in PIDs was only 3.8% (16% had ever tested) but it was actually higher – 22% – in their largely female contacts. The majority of sexual partnerships were casual and only 29% of partners knew their contact was a PID. This would make sense if many of the partners of drug users were female sex workers or otherwise engaged in sex with a number of different drug users.

The researchers calculated what potential these concentrated epidemics in PIDs had to spread into the general population. Given factors like condom use (about 22% among PIDs and 38% [Ivanovo] or 47% [Novosibirsk] among their contacts) and the number of partners people had, Eritsyan told the conference that her team had calculated that, for every 100 drug users who contracted HIV in Ivanovo, 70 people would be vulnerable to contracting HIV from them and, via them, 14 people in the general population (the partners of partners of people who inject drugs or PPPIDs). In contrast, in Novosibirsk, for every 100 PIDs with HIV, 118 people would be vulnerable to contracting HIV from them and via them 95 in the general population, or close to the figure needed to sustain a general-level epidemic.     

A poster study from India found similar heterogeneity in the risk behaviour of PIDs in different locations there. It compared risk behaviour in PIDs in Imphal, the capital of Manipur state, which has a long-standing heroin and HIV problem; the town of Sonipat about 15 miles north of Delhi, and the capital Delhi itself. Despite their geographical separation, risk behaviours in Sonipat and Imphal were relatively low, with high levels of condom use (in Imphal almost universal even with regular partners, in Sonipat 50% with regular partners, 70% with casual and 87% with commercial partners) and low levels of syringe reuse (6% in the last month). In Delhi in contrast there were high levels of needle and syringe re-use (53% in the last month) and lower levels of condom use: in fact, condom use was lower in casual (33%) and commercial (29%) than it was in regular partners.

Presenter Kaushik Biswas explained that, in the other two locations, PIDs in general had accommodation, but in Delhi we are talking about a group of homeless people who live on the city streets and who, if they have sex at all, often have it as hurried and often transactional sex with other homeless people.

Drug-injecting partnerships in California

One interesting way of looking at the intersection of HIV risk via injecting and via sex came from researcher Meghan Morris from San Francisco, who looked at partnerships based, not on sex, but on injecting. As part of a study that looks at the spread not of HIV, but at the faster-spreading hepatitis C in PIDs, she identified 53 ‘injecting partnerships’ of people of differing hepatitis C status who usually injected drugs together, lived together and often pooled money for drugs – a common phenomenon in PIDs, as drug injecting is often a highly social activity.

Thirty-three (62%) of these partnerships were male/female, of whom 11 (21% of all partners) were also sexual partners; 18 (34%) were male/male partnerships, of whom only two were sexual partners; and there were two female/female injecting partners. Individuals had been injecting drugs for a median of six years but had only been in an ‘injecting partnership’ for four months. Male/female partners were 1.8 times more likely to share syringes, but if they also had sex, then that risk multiplied to 2.8 times more likely and, in multivariate analysis, partners who had sex were four times more likely to share syringes than partners who did not have sex. The risk of HIV and hepatitis C was therefore far higher for women, especially, who combined sharing syringes with sex with the same partner.

Methamphetamines in Cape Town and the US midwest

An interesting study from South Africa looked at non-injected methamphetamine (meth) use in the townships of Cape Town. There has been a dramatic increase in smoked methamphetamine in the last decade, and fears that it will lead to increases in the already high HIV prevalence in the area.

In a sample of 3328 users of shebeens and bars, researchers found  212 meth users (6%), 117 men and 95 women (meth use is in fact highly stigmatised in the community and as a result users are often hard to find). Meth use was much more common amongst the 'coloured' (historically mixed-race) community than in the black community (10.5% versus 3.5%) and was associated with high levels of transactional sex, sexual assault, multiple sex partners, and male/male sex. There was evidence that some meth users were targeted for sex as they were seen as easily available.  

A study by Jichuan Wang and Brian Kelly looked at patterns of drug use and sexual behaviour in rural areas of the US midwest. They identified three different types of drug user: ones whose main stimulant drug of choice was powder cocaine; ones who almost exclusively used crack cocaine; and ones who mainly used methamphetamine, though combined it with other drugs. Although the patterns of drug use varied between these groups, the prevalence of unprotected sex was, contrary to their expectations, actually lower in the polydrug/methamphetamine users (59%, versus 67% for crack users). In this rural and mainly heterosexual population, methamphetamine may be used in its own right and not as a stimulant for sex.

Recreational sex and drugs in US gay men: San Francisco and Chicago

Finally, two studies looked at the intersection of drug use and HIV risk in US gay men. Milo Santos of the San Francisco Department of Public Health looked at the association between recreational drug use and unprotected sex both in terms of how it was related to different drugs, and how it was related to frequency of use. The three drugs most predictive of unsafe sex were methamphetamine, cocaine and poppers (alkyl nitrites). Unprotected sex was more common in all drug users compared to non-users, and in regular users it was more common than episodic users: episodic cocaine users were twice as likely to report unprotected sex as non-users, for instance, and regular users four times as likely.

Methamphetamine users, however, were very much more likely (about seven times more likely) to have unprotected sex than non-drug users, regardless of whether their use was regular or episodic. This finding set off an interesting discussion on how to work with episodic ‘meth’ users. These are not people who will probably regard their drug use as problematic, Santos commented, and far from their drug use setting off unsafe sex that they otherwise would not have had, they are probably using the meth in order to disinhibit themselves enough to have the unsafe sex they want.

The fact some gay men may use drugs in order to have unsafe sex, rather than drugs causing it to happen, was also suggested by the last study, an analysis of drug used during sex in the long-standing MACS cohort of gay men in Chicago. Researcher David Ostrow looked at drug use patterns in 1551 gay men over a 15-year period. He found that the largest group (39%) never used drugs; that the next-largest group often used one drug of choice in sex, and their use was stable over time (29%); but that there were smaller groups of men with continuing high polydrug levels (10%); a group whose use declined to zero as they got older (11%); and a group whose use actually increased as they got older (12%).

He looked at psychological and social factors associated with these different patterns and found that the factor most strongly associated with any pattern of drug use apart from the increasing pattern was sexual sensation-seeking: in other words these were men who were sexually adventurous and were using drugs to enhance sex. Polydrug use was also associated with high, rather than low income and also with lots of social connection to the gay scene as was moderate drug use. Decreasing use was associated with high educational levels. The only group that expressed concern about their drug use was the moderate-but-steady users, who might be more likely to see their drug as a habit they would like to do without. Low income was associated with drug use that increased in middle age.  

What these studies all indicate is that the drivers and motivators of unprotected sex and of drug use, injecting or otherwise, are very complex; whereas in some communities unsafe sex may be a consequence of drug use, either because of disinhibition or because of transactional sex, in other cases recreational sex, or relationship and intimacy, may be what then drives drug use. Finding people’s primary motivator in their risk behaviour may be important to help people reduce it though, as David Ostrow said, “What these studies show is that there are many places to intervene, and if you can find somewhere that works, that’s better than nothing working.”


All references are to the 

Stoové M et al. A critical need to scale-up of HIV prevention and harm reduction services for people who inject drugs in Tanzania: results from a HIV and hepatitis C prevalence study in Dar es Salaam, 2011. 19th International AIDS Conference, abstract MOAC0402, Washington DC, 2012..

Eritsyan K et al. Estimation of HIV sexual transmission potential from IDU to general population in two Russian cities. 19th International AIDS Conference, abstract MOAC0403, Washington DC, 2012.

Biswas K et al. Getting high, getting laid: injecting practices and sexual behaviour of people who inject drugs (PWID) in three Indian states (findings from the Hridaya baseline study). 19th International AIDS Conference, abstract TUPDD0204, Washington DC, 2012.

Morris M et al. Risk profiles of injecting partnerships: correlates of receptive syringe and cooker sharing among a cohort of young injecting drug users in San Francisco, California. 19th International AIDS Conference, abstract MOAC0404, Washington DC, 2012.

Meade CS et al. Methamphetamine use is associated with sexual abuse and HIV sexual risk behaviours among patrons of alcohol serving venues in Cape Town, South Africa. 19th International AIDS Conference, abstract MOAC0405, Washington DC, 2012. 

Wang J, Kelly B. Typology of polydrug use and unsafe sex practices among rural stimulant users. 19th International AIDS Conference, abstract TUPDD0201, Washington DC, 2012. 

Santos G-M et al. Substance use and high-risk sexual behavior: dose-response associations in episodic and high-frequency substance-using men who have sex with men (SUMSM). 19th International AIDS Conference, abstract TUPDD0205, Washington DC, 2012. 

Ostrow DG et al. Predictors of long-term trajectories (2003-2010) of sex-drug and heavy alcohol (SDA) use among MSM. 19th International AIDS Conference, abstract TUPDD0202, Washington DC, 2012.

View a complete listing of the session presentations, with links to abstracts and webcasts, on the conference website