Exclusive breastfeeding recommendation unchanged.

Editor – The secondary analysis published in the article by Bahl et al. in the June 2005 issue of the Bulletin (1) confirms the strong protective effect of breastfeeding against death among infants in Ghana, India and Peru, where infant mortality due to infectious disease is high. The strength of this protective effect is somewhat larger than that observed in the pooled analysis of three smaller observational studies (2), thus strengthening the arguments against the use of breast-milk substitutes under these conditions in the first half of infancy, even when the mother is HIVpositive (3). However, we are concerned with one of the study’s findings: “the risks of death or hospitalization associated with being predominantly breastfed were not significantly different from those associated with being exclusively breastfed.” The analysis presented was limited to infants who were older than 6 weeks (in Ghana and India) or 10 weeks (in Peru). The protective effect of exclusive breastfeeding (in comparison with predominant breastfeeding) may be greatest in the first months and may not have been as strong in the older infants in this study. Furthermore, as the authors acknowledge, the number of infants who were exclusively breastfed in Ghana and India restricted their ability to assess the difference in risk between exclusive breastfeeding and predominant breastfeeding. Despite these limitations, the findings section of the abstract and the discussion lead with the statistical similarity of predominant and exclusive breastfeeding. Although this observation is interesting and possibly important, it is not a true “finding” unless the sample sizes are large enough to be confident that the comparison is sufficiently powered to find a difference. Our concern is not purely academic. Some readers may not fully appreciate that this study was restricted to infants older than 6–10 weeks and that failing to find a difference may be due to small sample sizes. Emphasis on the statistical similarity between exclusive and predominant breastfeeding in this study could therefore be misinterpreted as implying that exclusive breastfeeding offers no health benefits over predominant breastfeeding. This might undermine the promotion of exclusive breastfeeding from birth, with negative consequences for infant health and survival. Although Bahl et al. refer to one inconclusive study on the survival benefits of exclusive breastfeeding compared with predominant breastfeeding (4), many other studies do confirm the advantages of exclusive breastfeeding in protection against diarrhoeal diseases and other illnesses (5–8) and provide a solid evidence base for WHO’s recommendation to breastfeed exclusively for the first six months without any additional foods or fluids, even water. O