Objective: The aim of this study was to compare the actual nutrient intakes observed in a previously reported study with assumed nutrient intakes based on the customary assumptions about the composition of human milk.
Study design: Fortified human milk is assumed to provide adequate amounts of nutrients for premature infants. This assumption holds if milk has the composition of milk expressed by mothers of premature infants during weeks 2 to 3 of lactation. The assumption does not necessarily hold for milk expressed after 2 to 3 weeks lactation. It also does not hold for donor milk, which is typically provided by mothers of term infants. The size of the disparity between assumed and actual nutrient intakes is not known. Actual nutrient intakes were available for 32 preterm infants participating in the study. Assumed nutrient intakes were calculated for these infants by substituting assumed nutrient concentrations for observed nutrient concentrations. Data were compared separately for each of the 3 study weeks.
Result: Actual protein intakes were significantly and consistently lower than assumed protein intakes during each study week. The differences in mean intakes were large, ranging from 0.5 to 0.8 g kg(-1) per day. Differences in energy intake were small and not consistently significant.
Conclusion: Actual intakes of protein by preterm infants fed fortified human milk are substantially lower than assumed intakes. The discrepancy may in part explain why preterm infants frequently show postnatal growth failure.