One hundred and eighty seven (187) infants, including 10 twins, were born alive to the HIV-positive mothers. After excluding the second-born twins, 182 infants were included in this analysis. Seventy five (41%) of mothers chose to formula-feed and 107 (59%) mothers chose to breast-feed.
shows baseline maternal and infant characteristics. Mothers who practiced formula-feeding were significantly older, had higher parity, and were more likely to have received ART or pMTCT drugs (92.0%) compared to breast feeding mothers (73.0%, P-value<0.01). Mothers who used formula also had lower CD4+ T cell counts (412 cells/mm3 than breast-feeding mothers (606 cells/mm3,P-value <0.01). There were no differences between the feeding groups with respect to education or access to protected water supplies. There were no differentials between the infants in both groups with respect to gender, birth weight and prematurity.
Baseline Maternal and Infant Characteristics.
is a consort diagram showing the follow-up of infants. Cumulative follow-up (including deaths) rates were similar among formula-fed and breast-fed infants at all visits (100% vs. 96 %, P-value
0.09 at one month, 97% vs. 93%, P-value
0.23 at six months and 93% vs. 83%, P-value
0.06 at twelve months).
Consort diagram showing follow-up of infants, losses to follow-up, deaths and HIV-infection.
None of mothers who chose to use formula reported any breast-feeding at all subsequent visits. Among the breast-fed infants, all infants seen alive at one month were still breast-feeding. The proportions breast-fed among surviving infants were 85% (82/97) at six months and 70% (61/87) at twelve months. However, exclusive breast-feeding was practiced by only 25% (25/101) of lactating women seen at one month and by 18% (17/97) at six months. The major infant feeds in addition to breast milk used by non-exclusive breast-feeding mothers at one month consisted of cow's milk (25%), mushroom soup (65%), and water (25%). At six months, the additional infant feeds were cow's milk (83%), mushroom soup (58%), water (40%), maize porridge (56%), millet porridge (41%), soya porridge (5%) and solid feeds (56%). The major feeds offered to infants whose mothers weaned them by six months included cow's milk (88%), mushroom soup (59%), millet porridge (88%), maize porridge (59%) and solid feeds (53%).
Infant mortality, HIV-infection and infant HIV-free survival
The proportion of HIV infected infants at one month were 13.0% (12/92) among the breast-fed compared to 4.4% (3/69) among the formula-fed infants (P-value
shows the Kaplan-Meier cumulative probabilities of survival from death. The 12-month cumulative probability of death was 18% (95% CI
11%–29%) among the formula-fed compared to 3% (95% CI
1%–9%) among the breast-fed (unadj. HR
6.1(95% CI, 1.7–21.4, P-value<0.01). We only present unadjusted analysis for this mortality endpoint since infant HIV-testing started at one month and we could not differentiate between in utero/peripartum transmission and early breast milk transmission between birth and one month, so adjusting for infant HIV-infection when analyzing the mortality endpoint would lead to censoring of the seven failure events that occurred between birth and one month prior to HIV testing. Since some infants in the breast-feeding group were weaned before twelve months we explored whether their exclusion would affect the results. The hazard ratio of mortality comparing formula-fed to breast-fed after exclusion of infants weaned before twelve months was 6.3 (95% CI
Kaplan-Meier cumulative probabilities of survival from death by feeding group.
All the three infants infected with HIV by one month of age in the formula-feeding group died by twelve months, whereas no deaths occurred among the twelve infants infected with HIV at one month in the breast-feeding group. The hazard ratio after exclusion of the HIV-infected formula fed infants was 4.9 (95% CI
shows the Kaplan-Meier cumulative probabilities of HIV-free survival. The 12-month cumulative probability of infant HIV infection or death was 14% (95% CI 8%–25%) in the formula-fed infants and 8% (95% CI
4%.–16%) among the breast-fed infants (unadjusted HR
3.9 [95%CI 1.1–14.0, P-value
0.04), and adjusted HR
2.8 [95%CI 0.67–11.7,] after adjustment for maternal age and maternal ART use. The corresponding infant HIV-free survival at twelve months was 86% in the formula group compared to 96%.among the breast-fed (P-value
Kaplan-Meier cumulative probabilities of HIV-free survival by feeding group.
After a review by the Study investigators confirmed the excess mortality in non-breastfed infants, further provision of formula in the pMTCT service program was stopped on 6th August 2007 for newly delivered babies.
The pMTCT program then decided to encourage exclusive breast-feeding for the first six months of life followed by gradual introduction of complementary feeds (not provided by the program) with continuation of breast-feeding for a duration of the mothers' informed choice. At this time however, ongoing formula-feeding was still being practiced by 29 mothers. Since breast-feeding could not be re-established, these infants continued to formula-feed. Mothers of the 29 infants were re-visited in their homes to evaluate their hygiene practices and to support safe continued use of formula. During the home visits, MWCs observed preparation and storage of formula and completed a household hygiene survey to assess hygienic practices. Monthly visits were planned to support the mothers for the rest of the formula-feeding time.
In the household hygiene survey, all mothers reported that they used soap and water to wash feeding utensils, but 17 (59%) mothers admitted to re-use of feeding utensils without washing. Infant feeding bottles, which were contraindicated by the program, were used by 25 (86.7%) of mothers. Storage of leftover feeds for later use was practiced by 5 (17.2%) of mothers. Overall, 9 (31.1%) of mothers reported difficulty maintaining clean utensils and 19 (65.5%) reported difficulty measuring the correct amount of formula powder. Nine mothers (31.0%) reported difficulty keeping utensils clean for the night feeds. Seventeen mothers (59%) reported having to delegate formula feeding to caretakers. Eight out of 17 (47.1%) of mothers delegated feeding to older children, while the remainder delegated feeding to the father or another adult relative. Twenty-three homes (79.3%) had a toilet facility, but only 11 of the 23 (47.8%) of homes with a toilet had hand washing facilities.