Mother to child transmission during exclusive breastfeeding

Polly Clayden, HIV i-Base

Breastfeeding remains an important route of mother to child transmission.

A study by Coovadia and co-workers, published in the 31 March 2007 issue of the Lancet, assessed the HIV transmission risk and survival associated with exclusive breastfeeding, and mixed breastfeeding and formula feeding among a cohort of women and infants attending antenatal clinics in Kwazulu Natal, South Africa.

This was a non-randomised intervention cohort study in which 2722 HIV-positive and HIV-negative women were enrolled. Infant feeding data were obtained weekly from mothers and samples taken monthly from infants for HIV testing. Of the infants born to HIV-positive mothers, complete feeding data were available for 1276.

The median duration of exclusive breastfeeding for women who initiated breastfeeding and whose infants had HIV diagnosis results available (n=1034) was 159 days (IQR, 122-174 days). 847 women (82%) reported exclusively breastfeeding for at least six weeks, 688 (67%) for at least 3 months and 415 (40%) for 6 months.

The investigators reported, of the exclusively breastfed infants, 175 had been diagnosed with HIV before 6 months of age. Kaplan-Meier survival analysis, conditional on exclusive breastfeeding, found cumulative infection rates were 14.1% (12.0-16.4) at 6 weeks of age, 18.1% (15.8-20.8) by 4 months, 18.6% (16.2-21.4) by 5 months and 19.5% (17.0- 22.4) by 6 months.

Of 723 exclusively breastfed infants who were uninfected at or after 6 weeks the estimated risk of infection was 1.1% (0.28-1.84) after 1 month, 2.2% (1.05-3.34) after 2 months, 2.7% (1.44-4.02) after 3 months, 3.3% (1.88-4.77) after 4 months and 4.0% (2.29-5.76) after 5 months (ie at about 6 months of age).

For infants who were HIV-negative at or after 6 weeks the overall transmission rate per 100 child-days, including infants who were replacement fed and those with missing data excluded was 0.032 (0.0222-0.0455). This rate varied from 0.0290 (0.0195-0.442) for 100 days of exclusive breastfeeding and 0.0436 (0.0208-0.0915) for breastmilk plus other foods or fluids. The investigators noted that this result equates to an estimated risk of 10.72 per 100 child-years of exposure to exclusive breastfeeding (or 0.89% per child month). With exclusive breastfeeding as reference they found the hazard ratio for breastmilk plus other fluids was, HR 1.56 (0.66-3.69, p=0.308. Infants who were breastfed but also received solid foods at any time were nearly 11 times more likely to become HIV infected than exclusively breast fed children, HR,10.87, 1.51-78.00, p=0.018. Numbers in non-exclusive breastfeeding categories however were small. The proportions of infants with HIV test results available who were fed a mixture of breastmilk and other fluids at 6 weeks, 3 months and 6 months were 3.8%, 5.75 and 15.4% respectively.

Transmission risk was strongly associated with maternal CD4 count <200 cells/mm3, HR 3.79 (2.35-6.12), p=<0.001 (and less strongly with maternal age, birthweight below 2500g, vaginal delivery and long duration of ruptured membranes). Estimated transmission at 6 months in exclusively breastfed women born to women with <200 cells/mm3 and >200 cells/mm3 were 34% and 17% respectively. Data for single dose nevirapine were inconsistent and not included in the analysis.

In multivariate analysis only infant HIV status was significantly associated with greater infant mortality risk. HIV positive infants were 15 times more likely to die than HIV negative children (HR, 15.28, 9.20-25.40, p<0.0001). In multivariate analysis, in a model that included infant HIV status, the only significant factor was maternal CD4 count. Infants born to mothers with CD4 count 200-500 cells/mm3 were almost twice as likely to have died, HR 1.89, 1.16-3.08, p=0.011, than those born to mothers with CD4 count >500 cells/mm3. Those born to mothers with a CD4 counts <200 cells/mm3 were more than three times likely to have died, HR, 3.19, 1.73-5.88, p=0.0001.

Infants born to mothers with CD4 counts <200 cells/mm3 were nearly four times more likely to be HIV positive or die than those born to mothers with CD4 cell counts >500 cells/mm3. Those born to mothers with CD4 counts between 200 and 500 cells/mm3 were 2.2 times more likely to acquire HIV or die.


Important lessons from this study are that 1) high rates of exclusive breast feeding can be achieved with adequate support, 2) mixed feeding with early introduction of solids greatly increases the risk of HIV transmission (HR 10.87) and 3) HIV-infected babies were 15 times more likely to die than uninfected infants.

Additionally, the study found increased early mortality associated with exclusive replacement feeding at 3 months in a small group (15% vs 6%) but that the gap narrowed by 6 months if HIV-free survival is the goal.

An important limitation of the study is the lack of data beyond 6 months, when, as reported in HTB (May 2007), a number of studies have shown high morbidity rates after weaning which seem also to be associated with the increased mobility of the infant. Additionally the effect of HIV-related infant mortality and morbidity will be clearer. Hopefully, further data from this cohort will cover this period. The study supports both the use of exclusive formula feeding (replacement feeding) where resources allow this to be conducted safely as well as exclusive breast feeding rather than mixed breast-feeding in other settings and identifies the need to target resources at the most vulnerable mother-infant pairs ie those with maternal CD4 counts less than 200.


Coovadia HM, Rollins NC, Bland RM et al. Mother- to-child-transmission of HIV-1 infection during exclusive breastfeeding in the first 6 months of life; an intervention cohort study. The Lancet, Vol 369; 31 March 2007.

Links to other websites are current at date of posting but not maintained.