Serum transferrin levels in the longitudinal assessment of protein-energy status in preterm infants.

Abstract

We prospectively measured serum transferrin levels weekly from birth until discharge in 33 preterm newborn infants hospitalized on the newborn intensive care unit (n = 130 weeks) to study whether transferrin levels accurately reflect recent nutritional intakes and predict subsequent changes in anthropometric measurements and serum protein levels. Mean daily protein and caloric intakes were no greater during weeks when transferrin levels increased than when levels decreased. There were weak but statistically significant linear relationships between protein intake and transferrin levels (r = 0.24, p less than 0.01), caloric intake and transferrin levels (r = 0.27, p less than 0.01), and magnitudes of weekly changes in protein intake and transferrin levels (r = 0.31, p less than 0.001), and magnitudes of weekly changes in caloric intake and transferrin levels (r = 0.27, p less than 0.01). Transferrin levels did not reflect same-week weight or midarm circumference (MAC) gains, nor did they predict the following week's gains. Mean anthropometric measurement gains were similar following weeks when transferrin levels increased or decreased. There were no positive linear relationships between the magnitudes of changes in transferrin levels and same-week weight gain (r = -0.35), same-week MAC gain (r = -0.27), or following-week MAC gain (r = 0.01). Weak correlations were found with following-week albumin levels (r = 0.32, p less than 0.001) and with same-week transthyretin levels (r = 0.44, p less than 0.001). Weekly serum transferrin levels are not useful for longitudinal surveillance of protein-energy status in preterm infants.