This study, which took place between 1994 and 1998 in Rakai, Uganda, tracked the HIV-1 transmission rate in 415 heterosexual couples with discordant HIV-1 status. Of the couples surveyed, 22% of the HIV negative partners seroconverted during the course of the study. While age and circumcision status were significant factors in the risk of seroconversion, the study found that the infected partner's viral load was the factor most strongly predictive of the risk of transmission.
While most of the factors surveyed--such as history of sexually transmitted diseases, time outside the area, or number of sexual partners--were not predictive of risk, whether or not a male HIV negative partner was circumcised was a significant factor in transmission. In HIV-1 negative males, there were no infections among 50 circumcised men. Amongst uncircumcised males, however, 40 out of 137 seroconverted in the course of the study.
HIV transmission between partners also increased if the HIV-positive partner had a history of genital discharge, painful urination, or AIDS defining symptoms. Age was also associated with infection risk in discordant partners, and risk decreased with age.
Viral load was, overall, the most significant transmission risk factor. Among couples where one partner was seronegative and later converted, their partners had, on average, a significantly higher viral load level than those who remained seronegative. Similarly, there were no transmissions in couples where the seropositive partner had an undetectable viral load. The study did not, however, isolate the extent to which individuals with low viral loads were also on highly active antiretroviral therapy or other antiretroviral treatments. Other studies, such as the Swiss "How Reliable is the Undetectable Viral Load," also have shown that people who have an undetectable viral load and take HIV antiretroviral consistently are effectively noninfectious.