Metabolomic markers of exclusive breastfeeding in infants

Breast milk is the gold standard for infant nutrition because in addition to nutrients, it also provides bioactive components that benefit the health of the developing infant. Infants under six months are recommended to be exclusively breastfed (EBF). Infant feeding practices determined by parental recall methods have been reported to have discrepancies when compared with results measured by the deuterium dose-to-mother (DTM) technique. However, the DTM method is limited by its high workload of sample collection and its validity outside the sample collection window. Finding markers of infant feeding practice using a readily accessible biological sample (such as urine or feces) is a promising novel approach to determine infant feeding practices.Breast milk is dynamic and its composition is impacted by many factors among which maternal secretor status and Lewis blood type impacts the levels of several human milk oligosaccharides (HMOs). Previous research has demonstrated the influence of maternal HMO phenotype on the infant gut microbiome. However, these studies used recall methods to determine infant feeding practice and the sample size used were relatively small. Therefore, it is important to illustrate this impact in a larger infant population whose feeding practice were determined objectively. This dissertation investigates the impact of maternal HMO phenotype on the milk metabolome, infant metabolism and gut microbiome, and seeks to develop metabolic biomarkers of feeding practice from samples of infant feces and urine. Chapter 1 reviews the compositional differences between human milk and infant formula, as well as the recent efforts to narrow the gap between the two. Chapter 2 demonstrates the impact of maternal secretor status and Lewis blood type on breast milk metabolome. Fold change analysis showed that the non-secretor (Se-) Lewis negative (Le-) milk had major differences in free fatty acids, free amino acids, and metabolites related to energy metabolism. Chapter 3 shows the impact of maternal secretor status on the infant fecal metabolome and gut microbiome. Maternal secretor status did not alter the within-community (alpha) diversity, between-community (beta) diversity, or the relative abundance of bacterial taxa at the genus level. However, differences in infant gut microbial fermentation products were observed with succinate, amino acids and their derivatives, and 1,2-propanediol. Chapters 4 and 5 develop and validate metabolic biomarkers for the discrimination between EBF and nonexclusively breastfed (non-EBF) infants using fecal and urine samples, respectively. Metabolites with excellent prediction performance (area under the receiver operating curve (AUC) above 0.8) were identified as biomarkers, and these included HMOs, short-chain fatty acids, amino acids (AAs) and organic acids in infant feces, and AA derivatives, nicotinamide-adenine dinucleotide degradation products, and metabolites from dietary sources and/or host-microbial co-metabolism in infant urine. This work provides new insights into the impact of maternal secretor status and Lewis blood type on the breast milk metabolome, as well as a further knowledge on how maternal secretor status influences the EBF infant gut metabolome and microbiome. The results in this study pave the road to a deeper understanding of how the choice of feeding practices impacts infant metabolism and the gut microbiome, and demonstrates the power of utilizing metabolic biomarkers as a novel approach to determine infant feeding practice