AIDS in the world, 2007

A Kenyan grandmother has been left to fend for orphans after her own children died of AIDS. (Antony Njuguna/Reuters)

LAST WEEK, the lead United Nations agency in the fight against AIDS announced that for several years it had been overestimating the number of people infected with HIV, the virus that causes AIDS. The number of new infections, officials at UNAIDS declared, likely peaked around 1999. This news is welcome, but it should not lead to any weakening of financial support or social commitment to the global fight against AIDS.

Even with the lowered estimate, UNAIDS still believes that 33.2 million people, down from 39.5 million, are infected. Only a small minority in developing countries are getting the best drug treatments.

Meanwhile, recent news on the vaccine front has been discouraging. Recently, Merck announced that a vaccine it had been testing had failed. Indeed, it actually raised a person's risk of infection. The difficulty of developing a vaccine against HIV puts a premium on the best available means of preventing infection. The "ABC" strategy - Abstinence, Being faithful to one partner, and Condoms - remains the best hope of keeping the number of people with HIV from rising.

Past UNAIDS figures overstated that number because the agency's data were based largely on the testing of pregnant women in urban clinics. Researchers now believe that extrapolating from those results was a mistake, because, they say, patients in urban settings are likely to have more sex partners than rural women and thus are more likely to be infected. To get more accurate figures, researchers financed by the US Agency for International Development have chosen households at random in urban and rural areas and deployed health workers to collect medical and lifestyle histories and blood samples.

The information gleaned from such surveys can be invaluable, not just in gauging the state of the epidemic but also in figuring out how to deal with it. In parts of sub-Saharan Africa where a high percentage of people are HIV-positive, one important factor in lower rates of new infections appears to be more fidelity in sexual relationships, according to Paul De Lay, the director of evidence, monitoring, and policy for UNAIDS.

An emerging hypothesis for the high prevalence of AIDS in sub-Saharan Africa is that, while people there have roughly the same number of sexual partners over a lifetime as people elsewhere, it is more common there for a man to have a sexual relationship with two or more women at the same time. A person who is newly infected with HIV has a high viral load in the first month or so and is especially likely to transmit the disease to partners. For this reason, according to Helen Epstein, a public health specialist and former AIDS vaccine researcher, "overlapping relationships" are riskier than consecutive ones.

Someone who has as few as one long-term sexual partner, Epstein has noted, is at risk of infection if that partner has another long-term partner who is on, as she puts it, the "HIV superhighway." In a teleconference this week, Epstein said it is important to get this information to people, so that they will "have an understanding of where the real risk comes from."

This hypothesis also points to the importance of raising the status of women so that they do not have to enter such overlapping relationships for economic reasons, and, if they are in such relationships, can insist on the consistent use of condoms. A study in 2004 and 2005 by Physicians for Human Rights found that lack of control by women over sexual decision-making was a major factor increasing their vulnerability to the disease. In sub-Saharan Africa, 61 percent of those with AIDS are women.

De Lay, the UNAIDS official, cited other factors besides increased fidelity to one partner that have likely played a role in reducing infections, such as more frequent use of condoms by prostitutes and other high-risk groups, and more attention to other sexually transmitted diseases, which if untreated can increase a person's vulnerability to HIV.

The improved surveys on HIV infection rates by health workers have another benefit: They can also alert officials to any tendency among individuals who have access to effective antiretroviral drugs to engage in risky sexual behaviors or to be more willing to use intravenous drugs with contaminated needles. De Lay said reasearchers are finding that, in both the United States and Uganda, the transformation of AIDS into a chronic, largely treatable disease is leading more people to return to unsafe practices.

Officials and grass-roots public health workers alike must fight this attitude. While promising evidence suggests that new infections have declined in recent years, the trend will only reverse if broader access to HIV drugs makes the infection appear to be a risk worth taking. Defeating AIDS requires the best drugs and the best prevention strategies, buttressed by the best research on what is working and what is not. Paying for this will continue to require generous support for the Global Fund to Fight AIDS, Malaria, and Tuberculosis, and this country's own international program of AIDS relief. The new calculations must prove to be not just a statistical blip, but a long term trend to greater survival.