Making prevention work: abstinence and partner reduction reviewed at World AIDS conference

This article is more than 17 years old. Click here for more recent articles on this topic

What of the debate about ‘ABC’ and whether programmes to encourage abstinence work? In terms of the ‘A’ we had two analyses of randomised and quasi-randomised trials of abstinence programmes, one mainly from Africa16 and one from the United States17.

Kevin O’Reilly from the World Health Organization (WHO) reviewed eleven rigorously-structured studies among adolescents in five African countries, India and the Philippines. Importantly, he was reviewing studies that encouraged both abstinence and condom use – only one was an abstinence-only intervention. He found that three studies produced a weak effect on increasing condom use (a 23% increase, which was not significant) and five produced a significant though not huge increase in sexual abstinence (40%): however this figure was distorted by one study from Uganda which produced a nearly 400% increase in the number of teenagers staying abstinent: none of the others produced more than a 40% increase.

O’Reilly called the results ‘significant but minimal’ and said the trials had worked better in changing gender attitudes than in reducing risk behaviour. (In a similar review, which unlike the abstinence study was not chosen for presentation, O’Reilly found that four out of five African trials of condom social marketing – none of them recent - produced some degree of condom use increase, two of them significant).



In a bacteria culture test, a sample of urine, blood, sputum or another substance is taken from the patient. The cells are put in a specific environment in a laboratory to encourage cell growth and to allow the specific type of bacteria to be identified. Culture can be used to identify the TB bacteria, but is a more complex, slow and expensive method than others.

mathematical models

A range of complex mathematical techniques which aim to simulate a sequence of likely future events, in order to estimate the impact of a health intervention or the spread of an infection.


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 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.

not significant

Usually means ‘not statistically significant’, meaning that the observed difference between two or more figures could have arisen by chance. 

Kristen Underhill from the Centre for Evidence-based Prevention reviewed eight trials of abstinence-only programmes for young people in the United States (one included adults) between 1993 and 2001. She found that six programmes produced no significant behaviour change: one produced a significant decrease in the frequency of recent heterosexual sex, but not in the proportion of it that was unprotected; and one, while producing higher rates of condom use in adults (presumably not the intended outcome), produced increases in sexually transmitted infections (STIs) in adults, and sexual frequency and pregnancy in teenagers. She concluded that the programmes ‘do not appear’ to affect HIV risk or abstinent behaviour, though because they had such different structures she was unable to generate a meta-analysis of their overall effect.

What of the ‘B’ in ABC? In two separate sessions (18,19), the second a PEPFAR-sponsored satellite, Daniel Halperin of the USAID Swaziland Programme and David Miller of the World Bank presented the same very convincing case that multiple partners, and in particular the high prevalence of long-term concurrent relationships, is the primary driver of the still-uncontrolled HIV pandemic in southern Africa.

Most of the evidence is epidemiological, showing that partner reduction tracks falls in HIV incidence and prevalence much better than condom use does. But they also presented an interesting mathematical model showing that small increases in the average number of concurrent partners could produce a huge increase in the number of people connected in sexual ‘networks’ and therefore able to transmit HIV. In a situation where the mean number of concurrent partners was 1.68, the largest proportion of a local population linked into a single network was 2%. When the mean number of partners increased to 1.86, no less than 64% of the sexually-active population became linked in a single network.

The question is, how to promote partner reduction and discourage concurrent relationships? In the PEPFAR satellite Derek von Wissell of the Swazi National Emergency Response Council on HIV/AIDS presented the decidedly ambiguous results of the ‘Makwapheni’ or ‘Secret Lover’ campaign. This campaign, with the tagline ‘your secret lover will kill you’, simulated text messages arranging assignations with second partners. It caused huge controversy in Swaziland, including one of the largest demonstrations the small country has ever seen, called by the Swaziland National Network of People Living with HIV/AIDS, which described the campaign as ‘insulting’ to the one-third of the population that already has HIV.

Von Wissell said that the campaign, which due to the protests only lasted ten days, reached 100% recognition in Swaziland, but admitted that “we should have done more pre-testing.” The previous phase of the campaign, featuring poster images of men and women saying ‘I will not share my partner’, achieved a reduction in the number of people who had had two or more partners in the last four weeks (from 16% to 9%). But this was accompanied by a decline in condom use (from 25% to 16%).

During questions, Bernard Gardiner of the International Federation of the Red Cross commented that the tradition of multiple concurrent partners must have arisen in the area because it had cultural and economic value – or, in other words, it had a context. Moderator Suzanne Leclerc-Madlala, an anthropologist at the University of KwaZulul Natal, answered that concurrent relationships were based in patterns that assured the continuity of productivity in agricultural cultures, such as wife inheritance. But society had changed, and now many women did not like the tradition.

One fascinating examination of the cultural and economic drivers behind multiple sexual partners in Africa, and which supported the viewpoints of both Gardiner and Leclerc-Madlala, was given the intriguing title of “A cow dies with grass in its mouth” 20. This was an examination of sex and relationship patterns among the Luo fisherfolk of Lake Victoria near Kisumu in Kenya. The remark was a retort made by a fisherman about the ‘zero grazing’ monogamy policy of President Museveni in neighbouring Uganda.

He said: “Initially people took AIDS seriously because people were dying and getting infected. But nowadays we say ‘A cow dies with grass in its mouth. If you think of it, AIDS can be considered an accident like the boat capsizing.”

Researcher Anjali Sharma found that this apparently fatalistic attitude was driven both by tradition and by economic necessity in a group of people whose livelihoods were threatened by international canneries and who were poor even by local standards. On the one hand, sex had a ritual place in Luo culture: “Traditionally among us Luo most of our customs end with sex – whether it is the planting season or when you want to harvest, everything ends with sex,” said one man.

But sex was also part of the local economy. Abstinence would impact on the earnings of female fishmongers, who were often widows. Much of the sex was transactional: women would sell sex to fishermen and get paid in fish that they then sold to earn a living. In turn, other men would form sexual relationships with fish-sellers for both comfort and food security – because the women that sold most sex had most fish. Not surprisingly, multiple concurrent relationships were the norm and the local HIV prevalence was 29%.

The Luo fisherfolk themselves, however, were not really fatalistic about HIV. When asked what their own solutions would be to the HIV/AIDS problem they came back with one biomedical solution (local antiretroviral dispensaries), but all their other solutions were economic: diversification of livelihoods, a cash float that could be rotated between women’s groups in times of need, fishermen’s co-operatives, and local microcredit and savings and loan programmes.

The social scientists are right: sex is all about status and making a living within your culture, and we if we ignore these drivers we will always be scratching our heads as to why people continue to put themselves at risk of HIV.


16. O’Reilly K. Systematic review of the impact of abstinence-only programmes on risk behavior in developing countries (1990-2005). Sixteenth International AIDS Conference, Toronto, abstract THAX0301, 2006.

17. Underhill K et al. Systematic review of abstinence-only and abstinence-plus programs for the prevention of HIV infection in high-income countries. Sixteenth International AIDS Conference, Toronto, abstract THPE0382, 2006.

18. Halperin D. Assessing a national HIV behavior change campaign focusing on multiple concurrent partnerships in Swaziland. Sixteenth International AIDS Conference, Toronto, abstract MOAC0201, 2006.

19. Miller D. Refining the prevention paradigm: exploring the evidence and programmatic models for behavior change. Sixteenth International AIDS Conference, Toronto, Satellite session TUSA04, 2006.

20. Sharma A. “A cow dies with grass in its mouth” – Fishermen’s response to “zero grazing” in Kisumu, Kenya. Sixteenth International AIDS Conference, Toronto, abstract THAX0302, 2006.