Defining growth faltering and malnutrition among very preterm infants in a neonatal intensive care unit

OBJECTIVE: Among infants born very preterm (≤32 weeks' gestation), we compared weight-for-age ≤10th percentile (10%ile) at 36 weeks to malnutrition guidelines using z-score changes and evaluated predictors of malnutrition. STUDY DESIGN: Infants (extremely low birthweight (ELBW) n = 699; very low birthweight (VLBW) n = 878) were classified using (1) ≤10%ile and (2) malnutrition categories with z-score changes. Multiple logistic regression identified predictors of malnutrition. RESULTS: 13% of ELBW (n = 91) and 15% of VLBW (n = 131) infants were classified as moderately/severely malnourished by losses of z-scores while weighing >10%ile. Infants with chronic lung disease and/or who had any ventilator use had an increased risk of moderate/severe malnutrition with z-scores. Notably, a loss of ≤0.8 z-scores was not associated with any postnatal clinical risk factors. CONCLUSION: The use of ≤10%ile for weight-for-age as a cut-off to identify growth concerns at 36 weeks postmenstrual age misclassified many preterm infants compared to malnutrition categories using z-scores.