For the past decade, the much lauded reduction of HIV/AIDS in Uganda has been the subject of intense debate. Was the decline real? Was it due to conscious changes in sexual behaviour? If so, why did Ugandans change their behaviour when so few people in other African countries have done so, including countries like Zambia and Zimbabwe, where the epidemic is almost as old as it is in Uganda?

Recently, researchers from Columbia University who have been studying the epidemic in Rakai district for more than a decade have claimed that the HIV decline in Uganda was due to a large number of deaths from AIDS and to a lesser extent, to an increase in the use of condoms. (1)

In a rebuttal, a number of evangelical Christians and African traditionalists have argued that the HIV decline in Uganda was due to increased sexual abstinence and faithfulness in marriage, arising from a remarkable commitment to Christian values, and a return to African traditions that cherished virginity.

However, a sober review of AIDS research over the past 20 years suggests that both parties in this debate may be off the mark. Research findings point to two types of behaviour change as the main causes of the HIV decline in Uganda. (2) The first was a 60% reduction in the proportion of men and women with casual sexual partners (meaning relationships lasting less than a year) in the late 1980s and early 1990s. These changes were promoted by campaigns urging people to "love carefully' and to 'zero-graze' and were reinforced by the threatening sounds of drums on radio and TV, and an effectively executed public health program that delivered the message throughout the country.

This change in behaviour probably began some 5 years before the Rakai researchers began collecting behavioural data in 1994. (3) During the late 1980s, focus groups suggested than many people distrusted condoms and preferred to protect themselves by avoiding short term affairs, and sticking to long-term, sometimes multiple partners.

The second type of behaviour change began in the early 1990s, when condom use increased sharply. However, by the time condom use became widespread, the HIV decline was well underway. After 1995, when condom social marketing reached its peak, the proportion of men with "non-regular" and "short-term" partners rose again. Yet HIV rates remained stable, almost certainly because of consistent condom use in casual relationships. Recent increases in HIV prevalence in some districts may be due to a worrying combination of a relaxation of both Zero Grazing and consistent condom use.

Contrary to what the evangelical Christians and African traditionalists maintain, abstinence-until-marriage seems to have made little contribution to the decline of HIV infection rates in Uganda. Some Ugandan officials have maintained that the age of sexual debut increased from 13 to 16 years during the 1990s, but they have provided no substantive evidence for this claim. The claim is further contradicted by the fact that Uganda's teenage pregnancy rates-among the highest in the world-- remained high during the first half of the 1990s, even when HIV rates were falling in the same pregnant teenage girls. The more recent reduction of teenage pregnancy, for example, from 43.3% in 1995 to 23.2% in 2000 among 17 year olds, occurred after HIV prevalence had significantly declined, and was accompanied by a sharp increase in the use of modern contraceptives, and thus cannot be attributed to abstinence alone.

Interestingly, research does show that boys and young men of the current generation are less polygamous and more cautious with sex. But they are not necessarily more religious or traditionally "African" than previous generations of men. Compared to girls, the boys' age of sexual debut does appear to have increased significantly. This is supported by the fact that girls 15 -19 are 6 times more likely to be infected with HIV than boys in the same age group.

While it certainly makes sense to advise young people to avoid sexual contact for as long as possible, and to remain faithful when they do have sex, a moralistic abstinence-until-marriage message could be disastrous. The experience of Swaziland, the nation with the highest HIV infection rate in the world is sobering. Four years ago, King Mswati III urged all young women to abstain from sex for five years. Many did as they were told, and since then, there has been a decline in both teen pregnancy and teen HIV infection rates. However, there has also been a shocking increase in the infection rate in those 20 and older. This suggests that the abstinence decree served only to postpone the age at which young people became infected, but did nothing to eliminate their risk. (4)

Finally, the conclusion that deaths from AIDS played a key role in HIV reduction in Uganda is simplistic because prevalence declines of similar magnitude have not occurred in other countries-such as Zimbabwe and Zambia-where the epidemic is almost as old as Uganda's and where a great many people have died.

The highly partisan debate between advocates of condoms versus abstinence has obscured the vital lessons that Uganda's fight against HIV holds for other countries. It seems likely that HIV reduction in Uganda was caused by a) a significant reduction of casual sex-partners, largely out of fear of death, reinforced by an energetic AIDS campaign, not out of the fear of God; b) faithfulness in stable relationships, sometimes multiple or polygamous (i.e. faithfulness not necessarily based on Christian values); c) increasing consistent use of condoms, especially in casual relationships; and d) a pragmatically more sexually careful but not necessarily religious or traditional new generation of young-men.

Uganda's experience is not unique. Wherever HIV infection rates have fallen-in Uganda, in Thailand and in the homosexual communities of the West--partner reduction, especially on the part of men, combined with vigilant condom use in casual relationships, seem to have been the key behaviour changes that led to the decline. (5) The public health community must urgently consider how best to implement programs to encourage similar behavioural changes in other countries severely affected by AIDS.