Abstract
Fresh human and bovine milk are bacteriostatic in vitro for only some (milk-sensitive) strains of E. coli. The addition of bicarbonate to the test system potentiates the bacteriostasis so that otherwise milk-resistant strains are inhibited. By titration of the bicarbonate in the milk, it is possible to determine the minimum concentration that will activate milk aganinst a milk resistant strain but be ineffective in boiled milk, i.e. it potentiates a heat-labile system in milk and does not merely exert a direct toxic effect. This concentration is lower for human milk than for cow's milk and can be reduced even further by the addition of more iron-binding protein. Lactoferrin and bicarbonate may be present in the gut of the newborn. In an attempt to imitate conditions in the infant gut, we therefore reinvestigated, in vitro and in the presence of added bicarbonate and transferrin, the bacteriostatic activity against E. coli of fresh breast-milk, commercial bottle-milk, and mixtures of these as fed to infants in the study. The results, and information about events in vivo deduced from the ratio of milk-sensitive to milk-resistant strains of E. coli isolated from babies' stool suggest that neonatal intestinal secretions may contribute to the bacteriostatic activity of their feeds so that (1) in fully breastfed babies all strains of E. coli are inhibited to the same extent; there is no selection on the basis of milk sensitivity and equal numbers of strains resistant and sensitive to milk are found in the stools; (2) in fully bottle-fed babies E. coli is not inhibited since the milk is non-bacteriostatic and again there is no selection; (3) in babies fed at the breast but bottle-milk supplemented, only milk-sensitive strains are inhibited; milk-resistant strains are not, and preferentially colonize the large intestines.