A study claiming good news on Africa's fight against AIDS has reignited conflicts over the effectiveness of HIV prevention campaigns, particularly the merits of focusing on abstinence and monogamy.

Greater condom use and monogamy are being linked to a fall in HIV in Zimbabwe. Credit: R. SMITH/CORBIS

Researchers reported last week that HIV infection rates in Zimbabwe fell between 1998 and 2003 from 23% to 20.5% (S. Gregson et al. Science 311, 664–666; 2006). They claimed that the gains came thanks to behavioural changes by Zimbabweans, who were having sex at a later age and with fewer partners. Women also reported more frequent condom use.

Some scientists and government officials hailed the finding as showing that campaigns persuading people to change their behaviour can succeed. “This supports the idea that behavioural change has resulted in a reduction in new infections in this study population,” says Richard Hayes, an epidemiologist at the London School of Hygiene and Tropical Medicine.

More controversially, US officials claim that the study validates their foreign policy. A third of US money spent on AIDS prevention overseas goes on programmes promoting abstinence and being faithful to one partner — the ‘A’ and ‘B’ in the ‘ABC model’, where ‘C’ stands for condoms. Critics say that focus is based on morality, not science, and can be unrealistic, especially for women who may have no choice about with whom they or their partners sleep.

Deputy US global AIDS coordinator Mark Dybul says the Zimbabwe study and earlier gains in Uganda help to put such criticisms to rest. “The greatest behaviour change was in abstinence and fidelity,” he points out.

But Simon Gregson of Imperial College London, who led the Zimbabwe study, insists that “we need to be promoting all the different prevention possibilities”.

And some warn that other factors have been overlooked. Maria Wawer of Columbia University in New York, co-leader of a long-running study in Uganda, thinks deaths from AIDS play a role. Although prevention programmes deserve some credit, she says, “there's no reason to believe mortality is not contributing to this”. After a decline in the late 1990s, HIV prevalence in Uganda now seems to be levelling off or even climbing back up.

The drop in Zimbabwe could also have been affected by violence and unrest, which have escalated there since 2000. Epidemiologist Christopher Beyrer of Johns Hopkins University in Baltimore, Maryland, says forced relocations and the departure of millions of young people seeking work may have skewed the study's results.