Research released Monday and published in the African Journal of AIDS Research suggests the AIDS epidemic in South Africa, the nation with the highest HIV/AIDS caseload, may be leveling off. Study authors Thomas Rehle, an independent US researcher, and Olive Shisana, of South Africa's Human Sciences Research Council (HSRC), developed a model based on South African antenatal clinic seroprevalence data. They said the nation's AIDS epidemic likely peaked in 2002 - when 4.69 million of South Africa's 45 million people were living with the disease - and would level off as fewer infections are recorded.

The epidemiological model uses data from South Africa's first national HIV prevalence study, released in late 2002 by HSRC and the Nelson Mandela Foundation.

The study projects that South Africa's annual AIDS-related deaths will peak in 2008 at 487,320, before declining to about 470,000 in 2010. New HIV infections among people ages 15-49 - a leading indicator of how the epidemic will progress - have already dropped from 4.2 percent in 1997 to 1.7 percent in 2003. Overall HIV prevalence among people ages 15-49 is also projected to drop from 17.3 percent in 2001 to an estimated 15.2 percent in 2010.

Earlier studies placed the incidence, prevalence and projected death numbers much higher. A US Census Bureau report estimated that as much as 37.9 percent of South Africa's sexually active adult population could be HIV-positive by 2010, with more than 900,000 deaths annually.

Among factors the new study credits with the slowdown are HIV/AIDS education and prevention programs that have led to changes in sexual behavior, and the fact that death is quickly thinning the ranks of the HIV-positive.

There is "still much uncertainty" about the course of the epidemic, and the authors noted that South Africa's decision to make antiretroviral treatment more widely available could have a profound change on future numbers. They also cautioned against interpreting the new, lower estimate as a sign the country's problems might have been exaggerated.

www.aegis.org