#43 Maternal and Early Life Factors influencing the Human Milk Microbiota in the CHILD Cohort
Shirin Moossavi, University of Manitoba; Ehsan Khafipour, University of Manitoba; Shadi Sepehri, University of Manitoba; Bianca Robertson, University of California San Diego; Lars Bode, University of California San Diego; Allan B Becker, University of Manitoba; Piushkumar J Mandhane, University of Alberta; Padmaja Subbarao, University of Toronto; Stuart E Turvey, University of British Columbia; Diana L Lefebvre, McMaster University; Malcolm R. Sears, McMaster University; Meghan B Azad, University of Manitoba
Previous studies suggested that human milk contains a complex microbiota that likely influences infant health and development. However, the maternal and early life factors determining the composition of milk microbiota is not well understood.
We studied the milk microbiota of 395 lactating mothers in the Canadian Healthy Infant Longitudinal Development study by sequencing the V4 region of the 16S rRNA gene on Illumina MiSeq. Human milk oligosaccharides (HMOs) were analysed by rapid high-throughput HPLC. Maternal factors were self-reported or documented from hospital records; 2010 Healthy Eating Index (HEI) was derived from a modified food frequency questionnaire. Effects of single factors on the β-diversity was estimated with PERMANOVA. Multivariate associations of maternal and early life factors with the milk microbial community was also assessed by redundancy analysis. The overall and by term effects were determined by permutation testing. Samples were clustered into groups based on hierarchical clustering with ward sum of square algorithm and multivariable logistic regression was used to identify factors associated with cluster membership.
While maternal intrapartum antibiotics, secretor status, ethnicity and method of delivery were not associated with milk microbiota β-diversity; parity (p=0.019) and infant sex (p=0.037) significantly affected the community structure. In multivariate analyses, HMO diversity, maternal HEI, birth weight, lactation stage, maternal age, maternal BMI, infant sex, and duration of exclusive breastfeeding collectively accounted for 7% of the variability in microbiota composition. Although this overall model was not significant (p>0.05), analysis by term revealed that duration of exclusive breastfeeding (p=0.019), maternal BMI (p=0.041), and infant sex (p=0.038) were independently associated with the overall structure. We identified three clusters with distinct microbiota profiles; of all factors evaluated, only the duration of exclusive breastfeeding was significantly associated with cluster membership.
Our findings identify several novel maternal, infant, and early life factors that may influence milk microbiota composition.