Creamatocrit, carbon content, and energy value of pooled banked human milk: implications for feeding preterm infants.

It is convenient to calculate milk energy from the creamatocrit but in order to assess the accuracy with which energy content can be predicted in this way, the creamatocrit, the percentage of carbon, and the calorie value of milk were determined in 11 samples of pooled pasteurised human milk. The calculated milk energy was compared with direct measurements of milk energy by static bomb calorimetry. The errors in calculation ranged from -5.6 kcal/100 ml to + 19.5 kcal/100 ml in milk samples whose measured energy was 34.5 to 63.1 kcal/100 ml. In 9 milk samples energy values were over-estimated by calculation and in the remaining 2 milk samples energy values were underestimated. The correlation between creamatocrit and measured energy value in pooled pasteurised milk was weaker than in previous studies using fresh milk. The percentage carbon was determined in our milk samples as a measure of their total organic constituents, and this appeared to be a more accurate predictor of milk energy than the predictive value of the creamatocrit which is only a measure of milk fat. In pooled pasteurised milk the relatively weak predictive value of the creamatocrit may be due to variations in the other constituents of milk apart from fat which provide energy, namely protein and lactose.