Breastfeeding patterns are associated with human milk microbiome composition: The Mother-Infant Microbiomes,...

The human milk microbiome (HMM) is hypothesized to be seeded by multiple factors, including the infant oral microbiome during breastfeeding. However, it is not known whether breastfeeding patterns (e.g., frequency or total time) impact the composition of the HMM. As part of the Mother-Infant Microbiomes, Behavior, and Ecology Study (MIMBES), we analyzed data from naturalistic observations of 46 mother-infant dyads living in the US Pacific Northwest and analyzed milk produced by the mothers for its bacterial diversity and composition. DNA was extracted from milk and the V1-V3 region of the 16S rRNA gene was amplified and sequenced. We hypothesized that number of breastfeeding bouts (breastfeeding sessions separated by >30 seconds) and total time breastfeeding would be associated with HMM α-diversity (richness, diversity, or evenness) and differential abundance of HMM bacterial genera. Multiple linear regression was used to examine associations between HMM α-diversity and the number of breastfeeding bouts or total time breastfeeding and selected covariates (infant age, maternal work outside the home, frequency of allomother physical contact with the infant, non-household caregiving network). HMM richness was inversely associated with number of breastfeeding bouts and frequency of allomother physical contact, but not total time breastfeeding. Infants' non-household caregiving network was positively associated with HMM evenness. In two ANCOM-BC analyses, abundances of 5 of the 35 most abundant genera were differentially associated with frequency of breastfeeding bouts (Bifidobacterium, Micrococcus, Pedobacter, Acidocella, Achromobacter); 5 genera (Bifidobacterium, Agreia, Pedobacter, Rugamonas, Stenotrophomonas) were associated with total time breastfeeding. These results indicate that breastfeeding patterns and infant caregiving ecology may play a role in influencing HMM composition. Future research is needed to identify whether these relationships are consistent in other populations and if they are associated with variation in the infant's gastrointestinal (including oral) microbiome.