Colostrum has important anti-infective properties. It may also somehow
promote the development of the child's immunological system. Discardi
ng colostrum, as practised in some cultures, could thus have adverse h
ealth consequences beyond the neonatal period. To test this hypothesis
, the age at breast feeding start of 734 healthy newborns in urban Gui
nea-Bissau was ascertained. The children were then prospectively follo
wed up to 3 years of age. Eighty nine deaths occurred during the study
. The probability of death in the age interval 28 days to 3 years was
about 20%. The child's age at breast feeding start had no statistical
impact on postneonatal growth or survival. As a single measure, early
breast feeding start is not likely to make much difference for the lon
g term growth or survival of children living under material poverty co
nditions.