Cesarean Section, Formula Feeding, and Infant Antibiotic Exposure: Separate and Combined Impacts on Gut Microbial Changes in Later Infancy

Farzana Yasmin, Hein Min Tun, Theodore Brian Konya, David S. Guttman, Radha S. Chari, Catherine J. Field, Allan B. Becker, Piush J. Mandhane, Stuart E. Turvey, Padmaja Subbarao, Malcolm R. Sears, CHILD Study Investigators, James A. Scott, Irina Dinu, Anita L. Kozyrskyj, S. S. Anand, M. B. Azad, A. B. Becker, A. D. Befus, M. Brauer, J. R. Brook, E. Chen, M. M. Cyr, D. Daley, S. D. Dell, J. A. Denburg, Q. L. Duan, T. Eiwegger, H. Grasemann, K. HayGlass, R. G. Hegele, D. L. Holness, P. Hystad, M. Kobor, T. R. Kollmann, A. L. Kozyrskyj, C. Laprise, W. Y. W. Lou, J. Macri, P. J. Mandhane, G. Miller, T. J. Moraes, P. Paré, C. Ramsey, F. Ratjen, A. Sandford, J. Scott, J. A. Scott, M. R. Sears, F. Silverman, E. Simons, P. Subbarao, T. Takaro, S. J. Tebbutt, T. To, S. E. Turvey

Established during infancy, our complex gut microbial community is shaped by medical interventions and societal preferences, such as cesarean section, formula feeding, and antibiotic use. We undertook this study to apply the significance analysis of microarrays (SAM) method to quantify changes in gut microbial composition during later infancy following the most common birth and postnatal exposures affecting infant gut microbial composition. Gut microbiota of 166 full-term infants in the Canadian Healthy Infant Longitudinal Development birth cohort were profiled using 16S high-throughput gene sequencing. Infants were placed into groups according to mutually exclusive combinations of birth mode (vaginal/cesarean birth), breastfeeding status (yes/no), and antibiotic use (yes/no) by 3 months of age. Based on repeated permutations of data and adjustment for the false discovery rate, the SAM statistic identified statistically significant changes in gut microbial abundance between 3 months and 1 year of age within each infant group. We observed well-known patterns of microbial phyla succession in later infancy (declining Proteobacteria; increasing Firmicutes and Bacteroidetes) following vaginal birth, breastfeeding, and no antibiotic exposure. Genus Lactobacillus, Roseburia, and Faecalibacterium species appeared in the top 10 increases to microbial abundance in these infants. Deviations from this pattern were evident among infants with other perinatal co-exposures; notably, the largest number of microbial species with unchanged abundance was seen in gut microbiota following early cessation of breastfeeding in infants. With and without antibiotic exposure, the absence of a breast milk diet by 3 months of age following vaginal birth yielded a higher proportion of unchanged abundance of Bacteroidaceae and Enterobacteriaceae in later infancy, and a higher ratio of unchanged Enterobacteriaceae to Alcaligenaceae microbiota. Gut microbiota of infants born vaginally and exclusively formula fed became less enriched with family Veillonellaceae and Clostridiaceae, showed unchanging levels of Ruminococcaceae, and exhibited a greater decline in the Rikenellaceae/Bacteroidaceae ratio compared to their breastfed, vaginally delivered counterparts. These changes were also evident in cesarean-delivered infants to a lesser extent. The clinical relevance of these trajectories of microbial change is that they culminate in taxon-specific abundances in the gut microbiota of later infancy, which we and others have observed to be associated with food sensitization.