Update of cancer surveillance of veterans in Massachusetts, USA.

Ahmed recently reviewed the evidence for an association between prolonged breastfeeding and malnutrition in developing countries. 1 They stated that most prior studies had demonstrated a negative association (increased malnutrition or lower mean nutritional status, both in terms of height-forage and weight-for-height, among prolonged breast-fed children), although an absence of a relationship, or even positive relationships have been described. We agree with the authors that confounding alone cannot explain the negative association between breast-feeding status and nutritional status, and also that 're-verse causality' is more likely to explain it. The authors state that according to their knowledge, 'there is little more than anecdotal support from the scientific liter-ature' to support the hypothesis of reverse causality. However, at least three papers based on longitudinal studies provide results that suggest that malnourished children are weaned later than other children. In rural Guatemala, children weaned before the age of 36 months were taller than the children weaned later, and since the height differences existed from infancy, they were not a result of earlier weaning. 2 In rural Bangladesh, mean weight-forage of children aged 12–36 months, who were to be weaned during the following month was slightly, but significantly higher than the mean of children who were to remain breast-fed, after adjustment for age (70.4% versus 69.6% of the NCHS median, P Ͻ 0.05). 3 In a periurban area of Guinea-Bissau in West Africa, children with a very low weight-forage (Ͻ–2.5 z-scores) at 10–17 months were weaned significantly later than other children (24.1 versus 22.3 months, P Ͻ 0.05). This difference remained significant after adjustment for ethnic group and mother's age and education. 4 Conversely, the study cited by the authors, which suggests that the child's length might influence weaning , 5 does not consider weaning as cessation of breast-feeding, but rather as the introduction of complementary food in addition to breastmilk during infancy; thus, it is not relevant in this context. We agree that further longitudinal studies are needed to assess whether age at weaning is related to nutritional status at weaning in other settings, and how nutritional status is affected by weaning in different environments. Nevertheless, the evidence that malnutrition is not a consequence of prolonged breastfeeding is already strong enough to promote continued breast-feeding in case of malnutrition, rather than to promote weaning, as has been suggested by some authors. 6 REFERENCES 1 Caulfield L E, Bentley M E, Ahmed S. …