Effect of Breastfeeding and Formula Feeding on Transmission of HIV-1
Summary of published clinical trial
Transmission of human immunodeficiency virus type 1 (HIV-1) is known to occur through breastfeeding, but the magnitude of risk has not been precisely defined. Studies have led to estimates of transmission risk during acute maternal infection of 29% (95% CI 16%-42%) and for women with established infection of 14% (95% CI 7%-22%). Whether breast milk HIV-1 transmission risk exceeds the risk of formula feeding-associated diarrhoeal mortality in developing countries is unknown.
A randomised clinical trial1 was conducted in Nairobi to determine the frequency of breast milk transmission of HIV-1 and to compare mortality rates and HIV-1-free survival in breastfed and formula-fed infants. Over five and a half years 425 HIVseropositive women were enrolled in the study. Women were eligible for inclusion if they were Nairobi residents and had access to municipal treated water. Participants did not use antiretroviral therapy. At about 32 weeks of pregnancy women were randomised to breastfeed or formula feed. The formula group were given free formula and shown how to prepare it. They were told to boil water and feed with a cup to minimise bacterial contamination.
Compliance with the assigned feeding method was:
- 96% in the breastfeeding group (defined by any use of breast milk) 83% were exclusively breastfed at 6 weeks, 62% at 3 months and 9% at 6 months.
- 70% in the formula feeding group (defined by complete avoidance of breast milk).
This comparatively low compliance in the formula group results from the fact that breastfeeding is the norm in Kenya and women experienced pressure from family and community and were in some cases concerned about maintaining confidentiality about their HIV- status.
Of the 401 infants in the study analysis2 92 were HIV-1 infected, 61 of these were in the breastfeeding group and 31 in the formula group. At 24 months the cumulative probability of HIV-1 infection was significantly higher for infants in the breastfeeding group than the formula feeding group. The estimated rate of breast milk HIV-1 transmission was 16.2% during the first 2 years of life. Breast milk transmission accounted for 44% in the group exposed to breast milk. The data suggest that substantial transmission occurs early (by 6 months 75% of all breast milk transmission had occurred). Another important finding was that the 2-year mortality rates in the 2 groups were similar. However the HIV-1 free survival at 2 years was significantly lower in the breastfeeding group than in the formula feeding group (58% vs 70% respectively; P = 0.02)
This important study throws some light on risk difference between HIV-1 transmission and formulae associated diarrhoeal mortality. Though the study showed that in this setting the use of formula prevented 44% of infant HIV-1 infections the authors caution about generalisability of results. While HIV-1 transmission via breast milk may be comparable across populations given similar exposure, the risks associated with formula are community specific. In this trial mortality in both groups was similar even though participants had access to clean water and extensive instruction on safe formula use. In developing country situations (especially emergencies involving refugee or displaced populations) where clean water and formula feeding knowledge is limited, the balance of risks and benefits could be shifted. Furthermore, formula is unaffordable to many (about $300 for 6 months in Kenya). Also, an intervention aimed at reducing HIV transmission through promoting use of infant formula amongst HIV sero-positive women requires antenatal HIV-1 testing and a healthcare infrastructure to provide education on formula feeding. The authors conclude that given the high sero-positive prevalence in pregnant women in sub-Saharan Africa the priority is to find ways to make a spectrum of preventative interventions widely accessible.
1Nduati R, et al., Effect of Breastfeeding and Formula Feeding on Transmission of HIV-1: A Randomised Clinical Trial. JAMA, March, 2000-Vol 283, No. 9 p 1167.
2At time of delivery only 408 remained in the study. After exclusion of stillbirths and second born twins the number in the study declined to 401.
Taken from Field Exchange Issue 11, December 2000
http://fex.ennonline.net/11/effect.aspx