Could HIV treatment mean safer sex?

'Risk compensation' does not happen, study finds
Caitlin Kennedy, from Johns Hopkins University, speaking at AIDS 2014. Photo: International AIDS Society/Steve Forrest.
This article is more than 10 years old. Click here for more recent articles on this topic

A meta-analysis of every study that has looked at the sexual behaviour of people after starting HIV treatment has found not a single instance of so-called ‘risk compensation’ – the idea that if people start taking HIV treatment they will fear transmitting HIV less and so start taking more risks. 

On the contrary: in 15 studies that met inclusion criteria for the analysis, starting HIV treatment was consistently associated with a fall in sexual risk behaviour, as measured by increased condom use. This association held true for both genders; for committed relationships and casual sex; and for partners both known, and not known, to be HIV negative.

The results were presented at the 20th International AIDS Conference (AIDS 2014) in Melbourne, Australia.


middle income countries

The World Bank classifies countries according to their income: low, lower-middle, upper-middle and high. There are around 50 lower-middle income countries (mostly in Africa and Asia) and around 60 upper-middle income countries (in Africa, Eastern Europe, Asia, Latin America and the Caribbean).

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.


When the statistical data from all studies which relate to a particular research question and conform to a pre-determined selection criteria are pooled and analysed together.

inclusion criteria

The conditions which a person must meet to join a research study.


low income countries

The World Bank classifies countries according to their income: low, lower-middle, upper-middle and high. While the majority of the approximately 30 countries that are ranked as low income are in sub-Saharan Africa, many African countries including Kenya, Nigeria, South Africa and Zambia are in the middle-income brackets. 

The limitations of this meta-analysis are that most of the studies were conducted in sub-Saharan Africa and among heterosexual people, so the results might not hold for other areas and for other populations such as gay men. But the meta-analysis does appear to demolish the idea that riskier behaviour is an inevitable, or even a normal, reaction to taking HIV treatment.

The Evidence Project is a programme sponsored by the US National Institute of Mental Health examining the strength of evidence for different behavioural interventions in people living with HIV. Researchers from Johns Hopkins and South Carolina Universities in the US combed through the literature to find studies published between 1990 and May 2012 that were conducted in low- or middle-income countries, provided treatment and, when it became available, combination antiretroviral therapy (ART) to people living with HIV; and which compared sexual risk behaviour either in people on ART with people not on ART, or in the same people before and after starting ART.

They found 37 studies that met these criteria (and another twelve only conducted in high-income countries). Thirty of the 37 came from sub-Saharan Africa, four from Thailand, two from Brazil and one from India.

Sexual risk was defined in a number of different ways including condom use, number of partners, proportion having casual sex, proportion abstinent, time of sexual debut, and incidence of sexually transmitted infections (STIs). For the sake of consistency, therefore, the current analysis only looked at 15 studies that rated the most commonly measured risk behaviour, condom use.

Overall, for both genders and any type of partner, the studies found that people taking ART used condoms 80% more often than people not on ART. In women on ART, rates of condom use doubled compared with women not on ART, and in men on ART, condom use was 50% higher. The association was even stronger when restricted to particular kinds of partner: the four studies that looked specifically at sex with partners of opposite or unknown HIV status found that condom use in people on ART rose by 160%, and it also rose by 160% (albeit from a much lower level) with spouses or regular partners.

These figures were for consistent condom use. The study found that the figures also held when condom use at last sex was considered. Overall, condom use at last sex was 130% higher in people on ART. It was also 110% higher in women specifically, 40% higher in men and 120% higher in people of negative or unknown HIV status, but in these three cases, the association was not statistically significant.

As well as holding true across groups, the association also held true over time: people on HIV treatment were more likely to use condoms from the earliest trial of HIV drugs in 1990 to the latest trial of ART in 2012. The majority of studies in fact came from the last decade, and the number of studies relating ART to sexual risk behaviour has increased exponentially recently: a new search has found 50 studies in the last two years that meet the inclusion criteria. This should enable new analyses to be made of other risk factors such as number of partners and of non-self-reported outcomes such as STI incidence.

The meta-analysis was restricted to heterosexual people and to low- and middle-income countries. But the consistency of its findings is striking.

“This is encouraging news for the continued expansion of ART in low- and middle-income countries”, said presenter Caitlin Kennedy. She suggested that instead of ‘treatment optimism’ causing complacency, the provision of ART could lead to decreased HIV risk behaviour partly due to regular medical contact and counselling, and partly due to an increased hope for the future and sense of agency.

“They suggest that the phrase ‘treatment is prevention’ may be true in more ways than one,” she added.


Kennedy C et al. Is use of antiretroviral treatment (ART) associated with decreased condom use? A meta-analysis of studies from low- and middle-income countries (LMICs). 20th International AIDS Conference, Melbourne, abstract WEAC0104, 2014.

View this abstract on the conference website.

See conference presentation here.