Poster Topical Area: Maternal, Perinatal and Pediatric Nutrition

Location: Hall D

Poster Board Number: 258

E04-05 - Diet and growth in hospitalized very preterm infants <=32 weeks' gestation

Sunday, Jun 10
8:00 AM – 6:00 PM

Objectives
Human milk is the recommended diet for very preterm infants during the neonatal intensive care unit (NICU) hospitalization. The extent to which current fortified human milk diets optimally support growth is unknown. Our aims were to: 1) examine associations of diet (human milk, preterm formula) with somatic growth during the NICU hospitalization; and 2) examine differences in growth by type of human milk (donor, maternal).

Methods
We studied 974 very preterm infants from 9 Massachusetts NICUs born 2015-2017 during a statewide quality improvement collaborative to increase maternal milk use. All NICUs routinely fortify human milk. Human milk and formula use were assessed weekly and at NICU discharge/transfer. Our main exposure was the percent of days assessed on which human milk was fed (%HM). For infants with 100% HM (n=569), we determined the percent of days on which any donor milk was fed (%DM). Clinical measures of infant size (weight, length, head circumference) were collected at birth and NICU discharge/transfer. Main outcomes were 1) z-score changes from birth to discharge/transfer which reflects growth relative to a fetal reference (Olsen 2010) and 2) weight gain velocity (exponential model, Patel 2005). We used linear regression to estimate associations of %HM and %DM categories with growth outcomes, adjusting for potential confounders and accounting for clustering by birth hospital and multiple gestation.

Results
Mean %HM was 83%. Of infants with 100% HM, mean %DM was 7.7%. Mean z-score changes were: -0.45 (weight), -0.92 (length), and -0.09 (head). Mean weight gain velocity was 13.3 g/kg/day. Adjusted associations of %HM with growth outcomes are in Table 1. More favorable linear growth was associated with being fed human milk on fewer days. Among infants fed human milk on 100% of days assessed, more favorable growth was seen in infants fed donor milk on fewer days (0-49%DM) but none of the difference was statistically significant (Table 2).

Conclusions
Infants fed human milk on more days had less favorable linear growth, but the extent to which this reduction in linear growth contributes to long-term health or developmental outcomes needs further investigation. Donor milk use was not associated with statistically significant growth differences, but this practice was infrequent in our cohort.



Funding Source: W.K. Kellogg Foundation
Table 1

Table 2

CoAuthors: Patrice Melvin, MPH – Boston Children's Hospital; James Moses, MD, MPH – Boston Medical Center and Boston University School of Medicine; Munish Gupta, MD, MMSc – Beth Israel Deaconess Medical Center and Harvard Medical School; Margaret Parker, MD, MPH – Boston Medical Center and Boston University School of Medicine

Mandy B. Belfort

Physician Investigator
Brigham and Women's Hospital
Boston, Massachusetts