Declining HIV prevalence and incidence in perinatal women in Harare, Zimbabwe.

Background: In several recent papers it has been suggested that HIV prevalence and incidence are declining in Zimbabwe as a result of changing sexual behavior. We provide further support for these suggestions based on an analysis of more extensive age-stratified HIV prevalence data from 1990 to 2009 for perinatal women in Harare as well as data on incidence and mortality. Methodology/principal findings: Pooled prevalence incidence and mortality were fitted using a simple susceptible-infected (SI) model of HIV transmission; age-stratified prevalence data were fitted using double-logistic functions. We estimate that incidence peaked at 5.5% per year in 1991 declining to 1% per year in 2010. Prevalence peaked in 1998/9 [35.9% (CI95: 31.3-40.7)] and decreased by 67% to 11.9% (CI95: 10.1-13.8) in 2009. For women < 20. y 20-24. y 25-29. y 30-34. y and = 35. y prevalence peaked at 25.4% 34.2% 47.1% 44.0% and 33.5% in 1993 1996 1997 1998 and 1999 respectively declining thereafter in every age group. Among women < 25 y prevalence peaked in 1994 at 28.8% declining thereafter by 69% to 8.9% (CI95: 6.8-11.5) in 2009. Conclusion/significance: HIV prevalence declined substantially among perinatal women in Harare after 1998 consequent upon a decline in incidence starting in the early 1990s. Our model suggests that this was primarily a result of changes in behavior which we attribute to a general increase in awareness of the dangers of AIDS and the ever more apparent increases in mortality.