Objectives: To examine the implications of variation in maternal infectivity on the timing of mother-to-child HIV transmission through breastfeeding.
Design and methods: A mathematical model of mother-to-child HIV transmission was developed that incorporates two main features: (i) the fetus/child potentially experiences a series of exposures (in utero, intrapartum, and via breastmilk) to HIV; and (ii) variation in maternal infectivity. The model was estimated from different sources of epidemiological data: a retrospective cohort study of children born to HIV-1-infected women in Sao Paulo State, Brazil, the International Registry of HIV-Exposed Twins, and the AIDS Clinical Trials Group 076 trial, which assessed the effectiveness of zidovudine in preventing mother-to-child HIV transmission.
Results: Variation in maternal infectivity results in higher average risk of breastfeeding-related transmission in the early stages of breastfeeding than in the late stages, even in the absence of a direct relationship between transmission risk and the age of the child. However, the available data were unable to resolve the quantitative importance of this mechanism.
Conclusions: Our model has helped identify a previously unrecognized determinant of the timing of breastfeeding-related HIV transmission, which may have adverse implications for the effectiveness of certain interventions to reduce mother-to-child HIV transmission such as maternal antiretroviral therapy in breastfeeding populations and the early cessation of breastfeeding.
PIP: By 2000, an estimated 5 million children will have been infected with HIV, the majority of them in sub-Saharan Africa. 30-50% of such infections could be the result of mother-to-child viral transmission through breast-feeding. Findings are presented from a study conducted to examine the implications of variation in maternal infectivity upon the timing of mother-to-child HIV transmission through breast-feeding. A mathematical model of mother-to-child HIV transmission was developed which incorporates the possibility of the fetus/child being exposed to HIV in utero, during the intrapartum period, and through breast milk; and variation in maternal infectivity. The model was estimated from epidemiological data drawn from a retrospective cohort study of children born to HIV-1-infected women in Sao Paulo State, Brazil, the International Registry of HIV-Exposed Twins, and the AIDS Clinical Trials Group 076 trial, which assessed the effectiveness of zidovudine in preventing mother-to-child HIV transmission. The effect of duration of breast-feeding upon the overall probability of mother-to-child HIV transmission, and therefore the age-specific risk of breast-feeding-related transmission, is highly sensitive to the degree of variation in infectivity. When substantial, the average risk of breast-feeding-related transmission declines rapidly with age and most infections occur in the early stages of breast-feeding. When the variation is less, infections attributable to breast-feeding are more evenly spread across the period of exposure to breast milk, although an imbalance towards early transmission remains.