Human milk oligosaccharides in breast milk and 2-year outcome in preterm infants: An exploratory analysis

Background & aims The health benefit of human milk (HM) for preterm infant development is known but the role of human milk oligosaccharides (HMOs) contained in HM remains underexplored. We explored the relationship between exposure to HMOs contained in mother's milk and growth and neurodevelopment at 2-years corrected age in preterm infants. Methods Exclusively breastfed preterm infants born between 27 and 34 weeks of gestation were enrolled in a monocentric prospective observational study, LACTACOL. Samples of breast milk were collected once a week for 7 weeks after birth. HMOs and sialic acid were measured by liquid chromatography. Age and Stages questionnaire (ASQ) version 2 was used to assess 2-year neurodevelopmental outcome. We analyzed the relationship between HMO content and (i) infant neurodevelopment at 2-years, and (ii) growth outcome at discharge and at 2 years. A secondary analysis was performed among Secretor(+) Lewis(+) mothers. Only associations with a false discovery rate of 10% or less according to the Benjamini-Hochberg procedure were considered significant. Results 137 preterm infants (mean gestational age of 31.3 ± 1.7 weeks, mean birth weight of 1494 g ± 336 g) born to 117 mothers (mean age of 30.8 ± 5.0 years) were enrolled. Total HMOs and most individual HMOs and sialic acid concentrations decreased with advancing postnatal age, except for lacto-N-fucopentaose-III and 3-fucosyllactose, which increased. Total HMOs were positively correlated with neonatal length growth (adjusted p = 0.012). Neither total HMOs nor any individual HMO correlated with ASQ score in the overall cohort. However, lacto-N-fucopentaose-III (LNFP-III) was significantly associated with total ASQ score (adjusted p ≤ 0.015) among the 104 infants born to Secretor(+) Lewis(+) mothers. Conclusions In this exploratory study in very preterm infants, total HMOs and most individual HMOs, except LNFP-III, decreased with advancing postnatal age. Neither the concentration of total HMOs nor that of any individual HMO were associated with ASQ score at 2 years, except for LNFP-III in Secretor(+) Lewis(+) mothers.