A new international growth reference is being prepared based on childr
en who are fed according to World Health Organization recommendations,
which entail exclusive breastfeeding for the first 4-6 mo of life. Ho
wever, the number of exclusively breast-fed infants in most societies
is small, so that selection biases may result from using such a popula
tion. We examined the variability in linear and soft tissue growth of
infants according to their feeding patterns, emphasizing differences b
etween exclusively and predominantly (breast milk plus fluids) breast-
fed infants. About 650 infants from a relatively developed urban area
in southern Brazil were examined at the ages of 1,3, 6, and 12 mo, and
an additional 800 infants at 6 and 12 mo. At each visit, infants were
weighed and measured and a 24-h dietary recall was completed with the
infants' caretakers. In the analyses of growth, care was taken to add
ress the biases of reverse causality, regression to the mean, and conf
ounding. There was little association between feeding pattern and grow
th in the first month. From 1 to 3 mo, partially breast-fed infants te
nded to gain more weight, followed by those who were completely weaned
. From 3 to 6 mo, fully weaned infants grew fastest in weight and leng
th and exclusively breastfed infants grew slowest. After 6 mo of age,
the growth in length of partially breast-fed and fully weaned infants
was similar, but the latter gained more weight. There were virtually n
o differences between exclusively and predominantly breast-fed infants
in the first 6 mo of life. There results suggest that both of these g
roups may be pooled for the purpose of constructing growth references.