Early breastfeeding cessation among infants and risk of serious gastroenteritis: findings from a perinatal prevention trial in Kampala, Uganda

Polly Clayden, HIV I-Base

Carolyne Onyango and coworkers from Uganda (where PMTCT guidelines recommend early weaning at 3 to 6 months for breastfeeding HIV-positive women) evaluated rates of gastroenteritis and mortality pre- and post-breastfeeding cessation in infants enrolled in an ongoing immunoglobulin/antiretroviral trial for PMTCT (HIVIGLOB). [6]

In this trial, mothers exclusively breastfeed for 3 to 6 months and to then stopped all breastfeeding by 6 months. Data on infant feeding practices, illnesses, hospitalisations, and deaths were collected. Hazard rates were calculated for first serious gastroenteritis episode (hospitalisations and dysentery) by month of age and compared to mortality among HIV-negative, exposed infants.

The investigators followed 579 HIV-negative infants; 65 (11.2%) had serious gastroenteritis. The median age of breastfeeding cessation was at 3 months (IQR 2 to 6 months); and median age of serious gastroenteritis was 6 months. Rates of serious gastroenteritis seen within 3 months pre- and 3 months post-weaning were 1.7% (95% CI  0.9 to 3.0) and 3.9% (95% CI 2.6 to 5.8), respectively.

Serious gastroenteritis rates among HIV-negative infants rose from 3 to 5 months and similarly to other reports peaked between 6 and 7 months of age (see the table). There were 15 infant deaths, 5 of which were due to gastroenteritis among the HIV-negative children post breastfeeding cessation, while no infant deaths occurred in the pre-weaning period.

The investigators in the HIVIGLOB trial found that breastfeeding cessation was associated with increased risk of serious gastroenteritis among HIV-negative infants. The rates doubled from the breastfeeding period until the 3-month period post-breastfeeding cessation; and infant deaths rose sharply within 3 months after breastfeeding cessation. They wrote: “Current WHO and Ministry of Health guidelines for early weaning of HIV-exposed infants need careful reassessment as to risks and benefits including risk of infant mortality.”

Table 1: Early breastfeeding cessation among infants and risk of serious gastroenteritis in Kampala, Uganda

Age in months 0-1 1-2 2-3 3-4 4-5 5-6 6-7 7-8 8-9 9-10 10-11 11-12
Rates of gastroenteritis per 1000 infants 2 20 36 30 33 35 47 47 26 30 36 14

Comment (on all breastfeeding and MTCT reports)

What is a poorly resourced mother to do – breastfeed and risk infection of her baby with HIV or wean early and risk death from enteritis? Do these five studies from Africa help?

One third of HIV infection in children is due to breast-feeding and to date we have only one randomised controlled trial of infant feeding. However, several of the studies from which these data have been taken are exploring interventions that may allow a mother to breastfeed more safely, but the observational data presented at CROI suggest that questions relating to duration of therapy will remain. In particular, it is interesting to note that these interventions seem designed on the assumption that weaning at 6 months is a good idea.

Although evidence based on observational data with historical comparison is not the best, some interesting trends and observations are presented:

  1. Rates of diarrhoea peak in the second six months of life and this was independent of the time of weaning (Uganda PPT data).
  2. It seems likely that the greater mobility from this age combines with the oral inquisitive nature of infants to compound the weaning risk.
  3. Circumstances vary and data from one setting cannot necessarily be extrapolated to another – in Malawi 41% of all children get diarrhoea whereas in Kenya only 5.7% in the KiBS and 8.7% in the historical control had diarrhoea in the first year of life.
  4. Circumstances can change – weaning during the rainy season seems to be bad news, especially in Botswana.
  5. Exclusive breast-feeding following by abrupt weaning at 3, 4 or 6 months is associated with problems and resources for maternal-infant services to ensure correct nutrition, monitoring and support are needed only with antiretroviral therapy.
  6. Not all mothers have the same risk of transmission and more information on the predictive ability of CD4 counts (ZEBS) may lead to risk based management.
  7. Babies already infected at birth should be breastfed.


  1. Onyango C, Mmiro F, D Bagenda D et al. Early breastfeeding cessation among HIV-exposed negative infants and risk of serious gastroenteritis: findings from a perinatal prevention trial in Kampala, Uganda. 14th CROI, 25-28 February 2007, Los Angeles. Abstract 775.

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