TOP 16 - Cumulative Enteral Dose of Maternal Human Milk, Donor Milk, and Formula: Differential Impact on Neurodevelopmental Outcomes in Very Low Birth Weight Neonates.

Publication Number: 2768.TOP 16

Mikael Ann Worsham, MUSC Children's Hospital, Charleston, SC, United States; Lakshmi D. Katikaneni, Medical University Of south carolina, Charleston, SC, United States; David A. Rueff, Ochsner Hospital for Children, Metairie, LA, United States; Tammy N. Tran, MUSC Children's Hospital, Charleston, SC, United States; Amy Ruddy, Medical University of South Carolina College of Medicine, Charleston, SC, United States; Katherine E. Chetta, Medical University of South Carolina College of Medicine, Charleston, SC, United States

Background: Maternal milk, donor human milk, and preterm formula each offer unique nutritional profiles that can influence an infant’s growth and development. While extensive data compares human milk and formula, as well as maternal and donor milk, there is a paucity of quantitative, comparative data that includes all three nutritional sources that reflect modern neonatal care practices.

Objective: This study aims to investigate the association between maternal milk, donor human milk, and preterm formula intake and neurodevelopmental outcomes at 9-12 months corrected age in very low birthweight (VLBW) and very preterm (PT) infants born under 32 weeks gestation. We hypothesized that VLBW/PT infants that receive greater quantities of maternal breast milk during their NICU stay score higher on neurodevelopmental testing than do neonates that receive more donor human milk and/or formula.

Design/Methods: A retrospective cohort of 300 VLBW/PT infants admitted to our NICU were followed from birth to their follow up at an outpatient NICU graduate clinic from 2017 to 2020. The total enteral intakes and feeding type during the hospital stay from birth until discharge was obtained from the electronic medical record. If any direct breast feeding occurred, it was not included in the compositional data as volume was unable to be ascertained. Neurodevelopmental outcomes were measured using Capute Scales, CAT/CLAMS, and Peabody gross motor scores obtained during their NICU graduate follow up visits and were reported as a developmental quotient, standardized by age. Scoring from visits between 9-12 months corrected age was prioritized. If unavailable, the closest visit was selected. We will conduct a rigorous compositional data analysis (CoDA) to examine the associations between the relative proportions of different milk types (maternal milk, formula, and donor human milk) consumed by infants, modeled as a compositional independent variable. Neurodevelopmental testing scores will be continuous dependent variables. We will apply appropriate log-ratio transformations, such as centered log-ratio (CLR) or isometric log-ratio (ILR) coordinates. To account for confounding, the models will adjust for gestational age, birth weight, sex, race/ethnicity, insurance type, and advanced maternal age. Model diagnostics will include checks for multicollinearity, residual normality and heteroscedasticity using R (vs. 4.4) software. This will be completed by the end of 2025.