High HIV-1 incidence in young women masked by stable overall seroprevalence among childbearing women in Kinshasa, Zaïre: estimating incidence from serial seroprevalence data.

In Zaire health workers screened 5937 and 4623 pregnant women delivering at the central maternity hospital in Kinshasa for HIV-1 antibodies. HIV/AIDS researchers used a birth-year cohort analysis and adjusted for mortality and fertility changes related to HIV infection to estimate HIV-1 incidence in women of reproductive age. The birth cohorts were 23-year-old women in 1986 and 26-year-old women in 1989 (i.e. birth-year cohort of 1963). The overall HIV-1 seroprevalence increased only slightly from 5.8 to 6.5%. Between 1986 and 1989 HIV-1 seroprevalence increased significantly among 35 years old. The HIV-1 seroprevalence curve had shifted from 27-year-old women in 1986 to 24-year-old women in 1989 (p = .03) (peaks of 7.7% and 7.5% respectively). Between 1986 and 1989 none of the HIV-1 seronegative women died while 13% of the HIV-1 infected women died. Fertility rates were consistently lower in HIV-1 infected women than in HIV-1 seronegative women within any birth-year cohort. When the researchers adjusted for higher mortality and lower fertility they determined the 3-year cumulative HIV-1 incidence rate to be 2.8/100 uninfected women. The highest HIV-1 incidence rate was among 20-24 year old women and the lowest was among > 40 year old women (5.7% vs. -0/5%). Further estimations showed that around 40% of the women identified as HIV-1 seropositive in 1989 had become infected between 1986 and 1989 and 60% of these women were < 25 years old in 1989 (i.e. younger than 22 in 1986). These findings provide a clearer understanding of the dynamics of the HIV-1 epidemic in Zaire and show the importance of age-specific HIV-1 seroprevalence data as means to interpret data from serial cross-sectional seroprevalence surveys. 1989