Growth of HIV Exposed and Unexposed Infants

Includes bibliographical references.BACKGROUND AND OBJECTIVES Malnutrition, which is widespread in many parts of South Africa, is a problem that affects child growth and predisposes children to early death. Another driver of child mortality in South Africa is the HIV/AIDS epidemic. An understanding of the interactions between malnutrition and HIV is therefore important, especially for vulnerable groups such as infants. This study is aimed to report on growth, in the first 36 weeks of life, of infants in three cohorts of motherinfant pairs: those infected by their HIV-positive mothers (infected), HIV-negative infants born to HIV-positive mothers (uninfected), and HIV-negative infants born to HIV-negative mothers (unexposed). Infant growth was also compared between the three different settings in South Africa. METHODS A prospective cohort study, called the Good Start Study, was conducted in three different settings in South Africa. Mothers were recruited at 28-36 weeks of pregnancy and followed up until the 36 weeks post delivery. Infant growth measurements were taken at 3, 24 and 36 weeks during scheduled home visits. The work presented in this document was a secondary analysis of data collected during the Good Start Study. Mean z-scores were calculated for length-for-age (LAZ), weight-for-age (WAZ) and weight-for-length (WLZ), and if they were below minus two, the infant was considered as moderately stunted, underweight and wasted, respectively. Mean z-scores, stratified by infant HIV exposure and infection status at three weeks, were plotted against infant age to assess growth over time. RESULTS The final sample included 98 infected, 386 uninfected and 193 unexposed infants. Although these infants differed significantly with regard to some demographic characteristics, these differences were minor. Infected infants had significantly lower mean WAZ (-1.11) compared to uninfected (-0.55) and unexposed (-0.55) infants at the three week (P0.05) in all mean z-scores was observed between uninfected infants and unexposed infants. Results from the multivariate analysis showed a significant (P=0.01) effect of time on the difference in mean WAZ between uninfected infants and unexposed infants between the 3 and 24 week visit times. Uninfected infants had a steeper growth trajectory compared to unexposed infants. Infants living in Rietvlei were significantly (P