Infant-feeding practices of mothers of known HIV status in Lusaka, Zambia.

Background: Between 25 and 44% of mother-to-child transmission (MTCT) of the human immunodeficiency virus (HIV) occurs through breastfeeding. As a result, feeding guidelines for infants of HIV-infected mothers are being formulated in many resource-poor countries. The impact of introducing these guidelines on mothers' actual feeding practices has not previously been examined. Infant-feeding practices of mothers of known HIV status who should have received advice during pre- and post-test HIV counselling were assessed and compared with those of uninfected mothers.

Methods: Mothers of infants aged 2-12 months, 55 HIV-infected and 85 HIV-uninfected, were recruited from the HIV Family Support Unit in Lusaka, Zambia. HIV status was known to 121 of these mothers, who had all received pre- and post-test HIV counselling. Feeding practices were determined by verbal questionnaire.

Results: All mothers breastfed but only 35% of infants below 4 months were exclusively breastfed (received breast milk only). HIV-infected mothers introduced fluids and weaned their infants significantly earlier than HIV-uninfected mothers (p = 0.03 and p = 0.002, respectively). Infants of HIV-infected mothers had significantly lower weight for age Z (WAZ) scores indicating poorer nutritional or health status (p = 0.004). Commercial formula milk and cow's milk were used by 36 mothers as breast milk substitutes, and were introduced at a median age of 2.5 months. Thirteen mothers gave cow's milk, and no mother added water to cow's milk (as recommended), with two adding sugar and four adding salt.

Conclusion: Infant-feeding practices of HIV-infected mothers differed significantly from HIV-uninfected mothers, and this may contribute to their poorer growth. Paradoxically these mothers feeding practice could be putting these infants at greater risk of both non-HIV-related morbidity and HIV transmission, as early introduction of foods other than breast milk may increase MTCT.