The dose-to-mother method to measure milk intake in infants by deuterium dilution: a validation study.

A validation study of the dose-to-mother deuterium dilution method to measure breast-milk intake has been carried out on ten infants from a Nutrition Recovery Centre in Santiago, Chile. Formula milk labelled with exponentially decreasing doses of deuterium oxide represented milk from a 'pseudo mother' of these exclusively bottle-fed infants. Unlabelled formula represented milk from other sources. Deuterium levels in the infants' body water were measured on saliva samples by mass spectroscopy. The data were fitted in a two-compartment steady-state model of the mother-child system to estimate the flow of labelled water from the 'mother' to the infant and the 'mother' and infant's water elimination constants. A dose-to-child experiment was also carried out on each infant to determine the deuterium dilution space and total daily water intake. Total and labelled water flows from deuterium dilution were used to calculate total, labelled and unlabelled formula milk intakes which were compared to the same quantities measured by bottle weighing. Water elimination constants calculated from the dose-to-mother experiments underestimated by 17 per cent the initially set mother constant and overestimated by 21 per cent the infant constants calculated from the dose-to-child experiments. Rate constants, however, have little effect on the estimated water flows so that there was good agreement between values from deuterium dilution and from direct weighing. Mean (SD) total formula, labelled formula and unlabelled formula intakes were 865 (129) g/d, 417 (74) g/d and 448 (106) g/d, respectively, when measured by deuterium dilution, compared with 856 (116) g/d, 414 (71) g/d and 441 (68) g/d when measured by direct weighing. The mean per cent differences were 1.1 per cent, 0.7 per cent and 1.6 per cent and the mean absolute differences for individuals, 4.3 (1.9) per cent, 4.8 (3.1) per cent and 9.7 (6.6) per cent, respectively.