Could easier access to AIDS treatment increase risky sexual behaviors?


This page in:


By the end of 2009, an estimated 5.2 million people in low- and middle-income countries received antiretroviral therapy (ART). In sub-Saharan Africa, nearly 37% [34%–40%] of people eligible for treatment had access to those life-saving medicines (UNAIDS 2010). This is an extraordinary achievement, considering that as recently as 2003, relatively few people living with HIV/AIDS had access to ART in Africa. The scaling-up of ART in Africa and other regions has saved the lives of countless people and we hope will continue to do so.

 At the same time, access to HIV/AIDS treatment might have transformed the perception of AIDS from a death sentence to a manageable, chronic condition, not necessarily different from any other chronic disease. Such a change in perception could lead to change in sexual behaviors. If AIDS is not perceived as a killer disease anymore, it might induce complacency and increase risky behaviors and the mixing between higher- and lower-risk groups in the population. That’s what has been described as the “disinhibition” hypothesis.

One of the earliest studies of the possibility of disinhibition behaviors looked at change in condom use by sex workers in Nairobi, Kenya (Jha et al., 2001). The findings are summarized in the figure below. The figure offers suggestive evidence that condom use by sex workers decreased when two fake cures of AIDS, Kemron and Pearl Omega, were announced. Such a pattern is consistent with disinhibition behaviors, although the result may not be repeated in the general population, since it is based on a very specific segment of the population.

Figure 1: Percent of condom use in a cohort of sex workers: Nairobi, 1985-1999
Source: Jha et al. (2001)

In a recent working paper, we test the disinhibition hypothesis using data that we collected in Mozambique. We find suggestive evidence that easier access to antiretroviral therapy can lead to more risky sexual behaviors. We reach that conclusion by drawing on Mozambique household panel data, which cover both randomly-selected, HIV-positive individuals and the general population in 2007 and 2008.

Our results suggest that men and women respond differently to increased access to therapy: risky behaviors increase for men who believe, wrongly, that AIDS can be cured; while risky behaviors increase for women who believe, correctly, that antiretroviral therapy can treat AIDS but cannot cure it.

A proposal under active consideration for combating AIDS is the  so-called “Test-and-Treat” policy, that all HIV-infected people in severely affected countries be started immediately on treatment, without waiting 8 to 10 years for symptoms to develop. One of the key assumptions supporting this policy is that better access to antiretroviral treatment will be accompanied by a 40 percent reduction in risky sexual behavior.(Granich et al., Lancet, 2009). The findings of our  study support the doubts raised by Garnett and Baggaley (2009), Baggaley and Fraser (2010) and others regarding this assumption, suggesting instead that better access to antiretroviral therapy may actually increase risky behavior, because people perceive that the risk of AIDS has diminished. These findings have two main policy implications. First, our results suggest that, to the degree that men incorrectly believe that AIDS can be cured (in Mozambique 8% in 2007 and 6% in 2008), prevention programs correcting this misconception may diminish men’s tendency to revert to riskier behavior in reaction to treatment access. Second, correcting misinformation is not enough to prevent disinhibition among the majority of men who already know AIDS cannot be cured or among women. Most of the adult population must instead be persuaded that, even with access to treatment, AIDS is a dangerous disease, warranting caution in sexual contacts.


Damien de Walque

Senior Economist, Development Research Group, World Bank

Mead Over

Senior fellow, Center for Global Development

Harounan Kazianga

Assistant professor, Spears School of Business, Oklahoma State University

Join the Conversation

February 25, 2011

Has the author looked at any studies in the rich countries, where 99% (except in the US maybe) have access to ART treatment? Why does he think that it is different in poor countries once they start to have access, even though still very limited, to similar treatment as we have had for years? Aren't we exposed in that case to the same disinhibition? What is his solution also, to provide less ART treatment in order to 'scare' the people to use more condoms?

Hmmm, seems to me a bit top-down research and analysis...


Damien de Walque
February 28, 2011


Thank you very much for your comment. There are several studies documenting evidence of disinhibition in rich countries. We discuss some of them in the working paper on which our blog is based: on study in France(,one in the Netherlands 9, and three in the US[Kalichman (; Lakdawalla (, Sood and Goldman ( ; Mechoulan (]

Crepaz, Hart and Marks ( have provided a more comprehensive review in 2004, with studies finding evidence of disinhibition and others not.

We are certainly NOT suggesting to reduce access to antiretroviral treatment, but we are stressing that the scaling-up of treatment should go hand-in-hand with reinforced prevention messages.

Best regards,

February 28, 2011

Thank you very much Damien for taking the time to answer my comment. I will look further into the references you mentioned. Sometimes, when reading just a part of a working paper one might not get the whole picture, that's why also it is important to try to be very clear I think when deciding to describe something. I did not in fact read anything about similar studies in rich countries in your test, or how this possible risk has been tackled where we have longer experience in ART. I understood of course that the team did not advocate for less ART treatment, but the recommendations were a bit vague as I saw it.

I want also to apologize for my tone in my comment. Having just came back from Ethiopia, where the situation is rather difficult in terms of access ART (even if it supposedly is provided for free), and some hypocrite comment from people I met there, I suppose I wrongly interpreter your blog. I still think though that giving some information around the behaviors in our countries would make it more balanced.

Kind regards and good luck with further studies.


Edward C Green
May 19, 2011

I was looking for PPT slides that Mead Over presented in South Africa a couple of years ago, and I found this blog. You are onto an extremly important topic. My colleagues and I have found evidence of behavioral disinhibition linked to ARV availability in Uganda, as we say in an article in the current JAIDS. We say much more in a new book AIDS, Behavior, and Culture by Edward C Green and Allison Ruark:
Of course, causality is very hard to determine.

Edward C Green
May 22, 2011

Here are some findings from a recent study we did in Uganda, discussed in our new book AIDS, Behavior and Culture (Left Coast Press 2011).

Our recent Uganda study found ample evidence that AIDS has come to be seen as a treatable, chronic disease not unlike malaria. Because of the availability of ARVs, and because Ugandans no longer see AIDS victims wasting away before their eyes, AIDS is no longer as feared as it once was (use of "fear appeals" was part of a deliberate strategy in earlier years, during which there was unprecedented behavioral change...Green and Witte 2006). Focus group respondents made comments such as, “Ever since people began getting ARVs, it has caused them to have irresponsible sex. They don’t fear each other any more, [or fear] that they might infect each other with the AIDS virus.”
Of course we are not presenting these (only suggestive) qualitative findings to argue that ARVs are not desperately wanted and needed in Africa and everywhere, but only to caution that risk compensation might be a downside to ARV provision that ought to be taken into consideration.