HIV treatment has not led to increasing sexual risk-taking in rural South African communities

Incidence remains unchanged
This article is more than 11 years old. Click here for more recent articles on this topic

A study from a rural area in the province of KwaZulu Natal, South Africa, has found no evidence that providing widespread antiretroviral therapy (ART) for HIV has led to increased sexual risk-taking in the general population.

Rather the reverse: during the period studied, condom use with a regular partner significantly increased and the proportion of people reporting multiple sexual partners decreased.

The researchers comment that: “Evidence that trends in sexual behaviour at the population level are not counter-acting the preventive effects of HIV welcome news”. But they stress that their findings only apply to the initial roll-out period of ART and that continued monitoring will be needed.

Background and study

Randomised controlled trials of ART and prevention methods such as circumcision have found no evidence that providing other ways to control HIV have led to higher rates of sexual risk-taking, but these were in the special context of a highly monitored and supported group. There have been few studies done on the impact of widespread ART provision in a whole population.


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.


A patient’s agreement to take a test or a treatment. In medical ethics, an adult who has mental capacity always has the right to refuse. 


Studies aim to give information that will be applicable to a large group of people (e.g. adults with diagnosed HIV in the UK). Because it is impractical to conduct a study with such a large group, only a sub-group (a sample) takes part in a study. This isn’t a problem as long as the characteristics of the sample are similar to those of the wider group (e.g. in terms of age, gender, CD4 count and years since diagnosis).


The study of the causes of a disease, its distribution within a population, and measures for control and prevention. Epidemiology focuses on groups rather than individuals.


The period of time from conception up to birth.

To remedy this, researchers studied results from the Africa Centre Demographic Surveillance System that began in 2000. It collects twice-yearly demographic data from resident household members (women aged 15 to 49 and men aged 15 to 54). Since 2003, it has also offered to test them for HIV, and since 2005 has collected sexual behaviour data. Because there is a high level of migration and seasonal labour in the community, it also selects a sample of non-resident members who are currently outside their home area.

For the purposes of this study, data was taken from 90,000 people in the Hlabisa area of northern KwaZulu Natal. HIV prevalence here is among the highest in the world in a general population, and was estimated to be 29% in 2011, with one in every 38 adults infected per year (annual incidence 2.63%). In the same area, publicly provided ART started in 2004 and by 2011, 20,598 adults had started treatment. This, the researchers estimate, represents 31% of all people with HIV aged 15 to 49.  

Participation in the sexual behaviour part of the survey declined between 2005 and 2011 from 84.5% of respondents to 57.5%. This was due to a change in the consent procedures so that participants had to consent separately for each part of the survey. It may possibly have had an influence on the results, though supplementary analysis showed that whether or not people participated in the sexual behaviour survey was not associated with specific behaviours.

Results – sexual behaviour

Between 2005 and 2011, people reporting that they had used a condom with their regular partner the last time they had sex increased from 28 to 43% in men and 26 to 52% in women. This is an average increase of 2.6% a year in men and 4.1% a year in women. Condom use with casual partners did not increase but was already high, at about 66%: in any case, very few women (less than 5%) reported casual partners.

The percentage of men reporting more than one partner in the last year fell from 20% in 2007 to 12% in 2011, a decline of 1.2% a year. Few women reported multiple partners but the percentage doing so fell from 3% to 1% between 2005 and 2011. In addition, the percentage of men reporting concurrency, that is, having more than one long-term, regular partner, fell from 12% to 8%. Concurrency was hardly reported at all by women.

Other sexual risk indicators remained unchanged between 2005 and 2011. For instance: 20% of respondents, male and female, reported no sex in the last year; the average annual number of partners stayed at just over one for men and 0.8 for women; and about 15% of men and 4% of women reported casual sex.

One important sexual risk is the age difference between partners; this is because trans-generational sex is much more likely to pass on HIV – principally from an older man to a younger women – than sex between people of the same age. Men were four years older on average than their female partners though this showed a slight but non-significant decrease from 2005 to 2011.

The researchers also looked at whether knowledge of HIV status influenced condom use. Knowledge of HIV status increased dramatically between 2005 and 2011, from 33 to 55% in men and 48 to 83% in women (who are more likely to know their status due to antenatal testing).

In 2005 status knowledge made no difference to men's condom behaviour. By 2011, while condom use among the diminishing number of men who had not tested stayed fixed at about 30%, it increased to 56% in men who had tested – regardless of whether they tested HIV positive or negative. In women, on the other hand, condom use was not affected by whether they had tested, but by their actual HIV status: by 2011 about 35% of HIV-negative women used a condom at last sex and 61% of HIV-positive women.

These changes happened in the absence of any HIV prevention programme activity in the area during the study period, other than HIV testing and treatment. The researchers comment that the counselling received during HIV testing and health monitoring may have changed risk behaviour in itself, but point out that, in women at least, condom use rose in those who had not tested too.

Incidence remains unchanged

One puzzle in this study is that, despite these reported decreases in sexual risk behaviour, HIV incidence remained stubbornly high at 2.6% throughout the study period. The researchers hypothesise that this could be due to inconsistent condom use or because social desirability bias – the tendency of people to report that their behaviour is better than it is to researchers – may have got worse over the research period, possibly due to more awareness of status and of HIV in general. In addition, it is possible that the people who did not consent to the sexual behaviour part of the survey in the last few years were those who were more likely to have taken sexual risks.

The main explanation, however, is probably the fact that ART has started prolonging lifespan in the area and that therefore, the proportion of the population who are living with HIV (the prevalence) has increased. Another paper by the same research team (Zaidi) shows that it increased from 21 to 29% between 2005 and 2011, while in a third paper, a mathematical model predicts little change in HIV incidence over the next few years (Mossong).

Changes in individuals’ sexual risk behaviour do not automatically translate into lower HIV incidence, because it also depends on who people are having risky behaviour with and whether the nature of those partners and networks changes. As the researchers say, “establishing the epidemiological consequences of an increase in condom usage at last sex requires further investigation.”  

Nonetheless, this study at least finds that higher rates of ART and viral suppression in the rural South African context have not led to an increase in sexual risk behaviour.


McGrath N et al. Sexual behaviour in a rural high HIV prevalence South African community: time trends in the antiretroviral treatment area. AIDS 27, e-pub head of print; doi: 10.1097/01.aids.0000432473.69250.19, 2013.

Zaidi J et al. Dramatic increases in HIV prevalence after scale-up of antiretroviral treatment: a longitudinal population-based HIV surveillance study in rural KwaZulu-Natal. AIDS 27, e-pub ahead of print; doi: 10.1097/QAD.0b013e328362e832, 2013.

Mossong J et al. Modelling HIV incidence and survival from age-specific seroprevalence after antiretroviral treatment scale-up in rural South Africa. AIDS 27, e-pub ahead of print; doi:10.1097/01.aids.0000432475.14992.da, 2013.