No increase in risk behaviour in people taking antiretroviral therapy, meta-analysis shows

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

One of the anxieties in relation to ‘treatment as prevention’ is that it may discourage people with HIV from using condoms and other prevention methods. However, a meta-analysis has found no increase in risk taking in people who are taking antiretroviral treatment, compared to other people with HIV. In fact, people on treatment had less unprotected sex and fewer sexually transmitted infections.

The study by Joseph Doyle and colleagues was presented to the 7th International AIDS Society Conference on HIV Pathogenesis, Treatment and Prevention in Kuala Lumpur this week.

One important caveat is that the researchers examined studies from all parts of the world and at any time since the introduction of antiretroviral therapy, with only some of them gathering data in recent years. However, understandings of the meanings and benefits of taking HIV treatment are changing. No studies related to people specifically prescribed antiretrovirals for prevention purposes.


odds ratio (OR)

Comparing one group with another, expresses differences in the odds of something happening. An odds ratio above 1 means something is more likely to happen in the group of interest; an odds ratio below 1 means it is less likely to happen. Similar to ‘relative risk’. 

statistical significance

Statistical tests are used to judge whether the results of a study could be due to chance and would not be confirmed if the study was repeated. If result is probably not due to chance, the results are ‘statistically significant’. 

heterogeneous or heterogeneity

Diverse in character or content. For example, the ‘heterogeneity’ of clinical trials means that they, and their results, are so diverse that comparisons or firm conclusions are difficult.


The origin and step-by-step development of disease.


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.

The researchers identified all previous studies which met a predetermined criteria and pooled their results. Relevant studies were of HIV-positive adults, with some taking antiretroviral therapy and some not, providing information on unprotected sex, diagnoses of sexually transmitted infections or unsafe injecting. The methodology of the Cochrane Collaboration was followed.

Fifty-five different studies provided data on self-reports of unprotected sex or inconsistent condom use, with over 30,000 participants. The pooled data showed that people on treatment were 28% less likely to report unprotected sex than people not taking treatment (odds ratio 0.72, 95% confidence interval 0.63 – 0.81). However, there was considerable heterogeneity in these results – the findings varied considerably from study to study.

Moreover, people taking treatment were less likely to report unprotected sex with a person of unknown or negative HIV status (odds ratio 0.57, 95% confidence interval 0.45 – 0.71).

Eleven studies with around 16,000 participants had information on diagnoses on sexually transmitted infections. Again, the results suggested that people on treatment were less likely to acquire an infection (odds ratio 0.58, 95% confidence interval 0.33 – 1.01), although this was not quite statistically significant.

These results were again quite heterogeneous, but essentially because of a single study with very different results to the others. The researchers felt justified in excluding this one from a further analysis. With this study removed, the results became statistically significant (odds ratio 0.48, 95% confidence interval 0.38 – 0.61).

Only four studies, with 1600 participants, provided information on unsafe injecting practices (lending, borrowing or re-use). This analysis showed that people on treatment were neither more nor less likely to have unsafe practices (odds ratio 0.90, 95% confidence interval 0.60 – 1.35).

The researchers say that the lower levels of risk behaviour reported by people on treatment should be interpreted cautiously, as this data cannot tell us anything about the causal relationships.

While it is possible that the counselling and support associated with engagement with healthcare help individuals to limit their risk taking, it is equally possible that the causation could run the other way – people who have more stability in their lives or who are more risk averse may be more likely to go on treatment. 

But the data remains reassuring. “Strategies to scale up use of treatment in the context of care are unlikely to adversely affect risk-taking behaviour,” the authors conclude.


Doyle JS et al. Meta-analysis of effects of antiretroviral therapy use on sexual and injecting risk-taking behaviour. 7th IAS Conference on HIV Pathogenesis, Treatment and Prevention, Kuala Lumpur, abstract WEPDB0105, 2013. View the abstract on the IAS conference website.

This article is also available in Portuguese.