Epidemiologic evidence for time variation in HIV infectivity.

Data from three epidemiologic studies of heterosexual transmission of HIV among monogamous couples are used to assess evidence for time variation in HIV infectivity, possibly related to varying levels of infectiousness following infection in the primary infected partner. Analyses are based on statistical techniques that account for the inherent incompleteness of exposure information from such studies, and that allow direct assessment of the hypotheses that infectivity varies with time since infection and across partnerships. Data include findings from 302 couples from the California Partners' Study and 51 and 31 couples, respectively, from two U.S. Center for Disease Control and Prevention (CDC)-sponsored studies of infection in partners of transfusion recipients. Results indicate weak evidence for higher infectivity following infection of the primary partner, decreasing to relatively lower levels from 2 to 10 years after. Although these findings are consistent with biologic observations of time variation in viral levels, other explanations of the observed pattern (e.g., heterogeneity of infectivity) are equally plausible, pointing out some inherent limitations of data from such studies.