Secondary Logo

Journal Logo

Original Articles: Gastroenterology

Mode of Delivery Determines Neonatal Pharyngeal Bacterial Composition and Early Intestinal Colonization

Brumbaugh, David E.; Arruda, Jaime; Robbins, Kristen; Ir, Diana; Santorico, Stephanie A.§; Robertson, Charles E.‡,||; Frank, Daniel N.‡,||

Author Information
Journal of Pediatric Gastroenterology and Nutrition: September 2016 - Volume 63 - Issue 3 - p 320-328
doi: 10.1097/MPG.0000000000001124


What Is Known

  • Bacterial colonization of the infant intestine is a critical event in early immune system development.
  • Cesarean delivery is associated with an increased incidence of allergic diseases in childhood.
  • Cesarean delivery is one of the several important early-life events that shape bacterial colonization of the infant intestine.

What Is New

  • Mode of delivery determines the bacterial communities seen in the initial oropharyngeal inoculum in newborns.
  • Cesarean-born infants display a persistent failure of intestinal Bacteroidetes colonization.
  • Oropharyngeal inoculation with physiologically appropriate bacterial communities may represent a potential method for restoration of normal intestinal colonization in infants born by cesarean delivery.

See Commentary on “More Than Just the Delivery” by Cabana on page 314.

Initial bacterial colonization of the human intestinal tract is a critical early-life process that shapes development of immune function (1–3). Mounting evidence from gnotobiotic models indicates that variance in intestinal bacterial colonization can lead to a proallergic and proinflammatory phenotype (4,5). Studies using molecular techniques for identification and classification of bacteria have shown vast differences in colonization patterns between vaginally born and cesarean section (c-section) born infants (6–8). Epidemiologic studies suggest a 20% increased risk for development of childhood asthma and other atopic diseases following c-section (9). Via mechanisms that are still unclear, c-section may also increase offspring risk of obesity (10–12). These findings have public health relevance as the frequency of c-section in the United States has climbed to 30% of all births (13).

From a cohort of 9 Venezuelan mother/infant pairs (4 vaginal deliveries, 5 c-sections), Dominguez-Bello et al (6) reported that vaginally born neonates carried skin, oral, and rectal microbiomes that were similar to maternal vaginal microbiomes, whereas those neonates delivered by c-section carried microbiomes dominated by skin microbiota. As the adult microbiota varies widely by geography and diet (14), and because clinical obstetric practice regarding cesarean delivery also varies widely by country (15), we sought to understand the relation between delivery mode and offspring intestinal colonization in the context of a large academic medical center in the United States. In a prospective, observational study of mother/infant pairs, we hypothesized that oral bacterial inoculation would differ by delivery mode as reflected in the bacterial taxa present in the oropharynx immediately after birth. In addition to characterizing infant oropharyngeal and fecal microbiomes immediately following birth, we profiled infant fecal microbiomes at 2 and 6 weeks of age to analyze longer-term consequences of delivery mode. In recognition that functional roles for commensal bacteria may be conserved across lower-level taxa (16), we used shotgun metagenomic sequencing to identify potential functional differences in gut microbiota between infants by delivery mode.


Human Subjects

Healthy pregnant women of 18 to 35 years were recruited during the third trimester from an academic obstetrics clinic. The objective was to recruit equal numbers of women planning vaginal delivery and c-section, indications for which included repeat c-section and breech positioning. All women planned to exclusively breast-feed for at least 2 months. Exclusion criteria included preterm delivery, maternal use of probiotics, maternal antibiotic use within 30 days of delivery (except perinatally administered antibiotics), conversion of vaginal delivery to c-section, and scheduled c-section with predelivery rupture of membranes. Informed consent was obtained from subjects before participation in the study. Authorization was obtained from the Colorado Multiple Institution Review Board.

Sample Collection

Using BBL Culture Swabs (Copan Diagnostics, Murrieta, CA), maternal vaginal and rectal specimens were obtained immediately before delivery. After all deliveries, the infant oropharynx was immediately suctioned using a sterile bulb syringe (Amsino International, Pomona, CA). The bulb syringe was rinsed with 5 mL of PBS to provide the oropharyngeal aspirate specimen. Fecal samples were collected from infants before nursery discharge, at 2 weeks, and 6 weeks of life. Samples were collected in 2 mL centrifuge tubes and frozen at −80°C until DNA extraction performed. At 2 and 6 weeks of life, mothers completed a questionnaire regarding infant feeding practices and antibiotic exposures. Primarily breast-fed was defined as an infant receiving the majority of feeds from human milk.

16S Amplicon Library Construction

DNA was extracted from samples using the Powerfecal DNA Isolation Kit (Mo Bio, Carlsbad, CA). Bacterial profiles were determined by broad-range amplification and sequence analysis of 16S rRNA genes following our previously described methods (17,18). In brief, amplicons were generated using primers that target approximately 300 b.p. of the V1V2 variable region of the 16S rRNA gene (primers 27FYM (19) and (338R) (20), modified by adding dual indexes and Illumina adapter sequences). Polymerase chain reaction (PCR) products were normalized using a SequalPrep kit (Invitrogen, Carlsbad, CA), pooled, lyophilized, purified and concentrated using a DNA Clean and Concentrator Kit (Zymo, Irvine, CA). Pooled amplicons were quantified using Qubit Fluorometer 2.0 (Invitrogen, Carlsbad, CA). The pool was diluted to 4 nM and denatured with 0.2 N NaOH at room temperature. The denatured DNA was diluted to 15 pM and spiked with 25% of the Illumina PhiX control DNA before loading the sequencer. Illumina paired-end sequencing was performed on the Miseq platform with version v2.3.0.8 of the Miseq Control Software and version v2.3.32 of MiSeq Reporter, using a 600-cycle version 3 reagent kit.

As previously described (18), Illumina Miseq paired-end sequences were sorted by sample via barcodes in the paired reads with a python script. Sorted paired reads were merged using phrap assembly software (21,22). Pairs that did not assemble were discarded. Assembled sequence ends were trimmed over a moving window of 5 nucleotides until average quality met or exceeded 20. Trimmed sequences with >1 ambiguity or <200 nt were discarded. Potential chimeras identified with Uchime (usearch6.0.203_i86linux32) (23) using the Schloss (24) Silva reference sequences were removed from subsequent analyses. A total of 9,159,483 high-quality sequences were generated for 89 samples (average sequence length, 314 nt; average sample size, 102,916 sequences/sample; minimum sample size, 7530; maximum samples size, 343,499, exclusive of negative controls that were near zero). Assembled sequences were aligned and classified with SINA (1.2.11) (25) using the 418,497 bacterial sequences in Silva 115NR99 (26) as reference configured to yield the Silva taxonomy. Operational taxonomic units (OTUs) were produced by clustering sequences with identical taxonomic assignments. Relative abundances of OTUs were calculated for each subject by dividing the sequence counts observed for each OTU by the total number of sequences generated. Higher-level OTUs (eg, phyla) were constructed by first summing the OTU counts for their constituent taxa, then normalizing by total sequence counts per subject. The median Goods coverage score was at least 99.97% at the rarefaction point of 7530 (Supplemental Table 1,

Statistical Analysis

Subject characteristics were compared using the Fisher exact test for categorical variables and Wilcoxon rank sum test for continuous variables. Comparison of percent relative abundances of OTUs between groups was performed using a Wilcoxon Rank Sum Test with median relative abundance reported as a summary measure. Standard ecological alpha diversity indices (eg, SChao1, Shannon diversity), inferred by rarefaction and bootstrap analysis with 1000 replicate re-samplings, were assessed by analysis of variance (27,28). Dissimilarity between microbiomes of pairs of patients (ie, beta-diversity) was measured using the abundance-based Morisita-Horn index (using the “vegdist” R command). Hierarchical clustering of subjects by microbiome was conducted by applying the “hclust” R function with average linkage to a pairwise dissimilarity table calculated with the Morisita-Horn index. The R (v3.0.3, (29) and Explicet (v2.9.4, (27) software packages were used for data display, analysis, and figure generation.

Metagenomic Analysis

Bulk fecal DNA samples prepared for 16S PCR were subjected to multiplexed shotgun sequencing using the Nextera XT kit (Illumina Inc, San Diego, CA) and the 600-cycle MiSeq Reagent Kit v3 (Illumina Inc, San Diego, CA). Raw, paired-end reads were trimmed of poor-quality bases at 5′ and 3′ ends (by excising bases in a 10 nucleotide sliding window with mean phred Q < 15) (30), assembled using FLASH (31), and 200,000 high-quality merged reads per specimen uploaded to the metagenomic RAST server (MG-RAST;; accessed Oct 2014) for automated sequence classification and analysis (32). Human sequences were identified and culled by comparison to the Homo sapiens HG19 reference genome; remaining sequences were annotated using the MD5 nonredundant database (33). Results are presented for annotations against the KEGG ortholog (KO) hierarchy (34). Sequence annotations were downloaded and analyzed using the R package “matR” v1.0.0 (35). For analyses at each level of the KO hierarchy, sequence counts were standardized per individual to a relative proportion. Comparison of KO annotation classes between groups was performed using a Wilcoxon Rank Sum Test with the exact distribution as implemented in the R package coin (36). Because of the pilot nature of this study, P values were not adjusted to control for multiple testing. Heatmaps were imported into Adobe Illustrator to add annotations. Lower level KO categories were mapped onto higher-order categories after downloading the KO ontology using the MG-RAST API (, accessed August 2014). Principal coordinates analysis (PCoA) (37–39) of metagenomic sequence data was conducted by first constructing a pairwise dissimilarity matrix, based on the Bray-Curtis index, using the “vegdist” command of the R vegan package (40). This matrix was then analyzed by the “wcmdscale” function. In this analysis, principle coordinates (PC) axes 1, 2, and 3 accounted for 43.2%, 13.5%, and 9.9% of overall variance, respectively. The Spearman rank correlation test was used to assess correlations between PC scores and OTU abundances.

Sequence Repository

Paired end 16S rRNA reads and merged metagenomic sequence data were deposited in the NCBI Short Read Archive under project number PRJNA278085.


Study Design and Implementation

Twenty-three subjects participated in the study (Table 1). Twelve mothers delivered by c-section and 11 vaginally. Women in the vaginal and c-section groups did not differ in age or body mass index. As expected, a difference in perinatal antibiotic exposure was noted because all mothers undergoing c-section received a single dose of perioperative intravenous cefazolin. Three mothers delivering vaginally tested positive for Group B Streptococcus during prenatal screening; of these, 2 mothers received doses of IV penicillin while in labor. Infants did not differ in gestational age by mode of delivery. No infant received antibiotics postnatally during study follow-up. Infants were primarily breast-fed during follow-up in both vaginal and c-section groups (91% vs 92% at 2 weeks, 91% vs 75% at 6 weeks, respectively). The earliest infant stool was collected later after c-section compared to vaginal birth (40 hours vs 25 hours, P = 0.04; Table 1), but there was no difference in timing of stool collection at 2 weeks and 6 weeks of life.

Characteristics of study population

Bacterial Abundance by Mode of Delivery

All samples contained detectable bacterial 16S DNA, from which an average of 102,916 high-quality 16S rRNA sequences per sample were generated (Good's coverage was >99% for all samples; Supplemental Fig. 1,

The phylum- and genus-level microbiotas of maternal rectal and vaginal specimens did not differ appreciably between those mothers delivering vaginally or by c-section (Fig. 1a and b). The principal bacterial phylum represented in maternal vaginal samples was Firmicutes, almost entirely lactobacilli. Maternal rectal samples were dominated by the phyla Firmicutes and Bacteroidetes. Only the Proteobacteria detected in rectal samples differed significantly in relative abundance (RA) by delivery mode (2.7% RA in c-section mothers vs 0.5% RA in mothers delivering vaginally, P = 0.03). No differences in ecological measures of alpha-diversity (eg, richness, evenness, complexity) in maternal samples were seen by delivery mode (Supplemental Fig. 1,

Distributions of abundant bacterial taxa in maternal and infant specimens. Panel A: heatmaps showing individual variation in phylum-level microbiomes of mother-infant dyads. Each column represents a single subject. Values to the right of each phylum name represent median percent relative abundances of taxa in CSec and Vag birth groups. Uncorrected P values from Wilcoxon rank-sum tests are indicated for values <0.1. Panel B: genus-level distributions in specimens. Values represent average percent relative abundances for taxa identified in specimens, grouped by sample type and mode of birth. P values from PERMANOVA tests are displayed under the figure: # P < 0.01, P < 0.05. For clarity of presentation, only phyla or genera with mean relative abundances ≥2% are displayed in Panels A and B. CSec = c-section; Vag = vaginal.

In contrast to maternal specimens, significant differences in phylum-level (Fig. 1a) and genus-level (Fig. 1b) bacterial abundance were observed by delivery mode for all infant sample types. Oropharyngeal aspirates from infants born by c-section compared to vaginal delivery had increased relative abundances of Actinobacteria (20.1% vs 3.8% RA, P = 0.05) and Proteobacteria (17.8% vs 1.6% RA, P = 0.02), whereas aspirates from vaginally delivered infants were enriched in Firmicutes (62.6% vs 30.1% RA, P = 0.01). Within the phylum Actinobacteria, there was increased abundance of the genus Propionibacterium, members of which include common skin commensal species (eg, Propionibacterium acnes), in oropharyngeal aspirates from c-section infants (18.3% vs 3.1% RA, P = 0.009). The genus Lactobacillus trended toward increased abundance in oropharyngeal aspirates of vaginally born infants compared with c-section infants (32.3% vs 2.8% RA, P = .09). Although not reaching significance, oropharyngeal aspirates from vaginally delivered infants also were characterized by increased OTU counts (47.4 vs 36.0 genus-level OTUs, P = 0.08) and decreased evenness (50.5% vs 65.2%, P = 0.06), compared with c-section delivery.

In comparing the earliest stool sample between delivery groups, the RA of Bacteroidetes was greater for infants born vaginally (8.5% vaginal vs 0.7% c-section, P = 0.01). The disparity in Bacteroidetes RA by delivery mode persisted in both the 2-week (31.4% vaginal vs 0.3% c-section, P < 0.001) and 6-week (22.7% vaginal vs 0.1% c-section, P = .007) fecal samples (Fig. 1a). No differences in infant fecal abundances of Lactobacillus spp were observed by delivery mode. Indeed, the abundance of lactobacilli in infant fecal samples was low through 6 weeks (Fig. 1b). No differences between groups in alpha-biodiversity (Supplementary Fig. 1, were evident in the initial and 2-week fecal samples. At week 6, vaginally delivered infants exhibited lower OTU richness (24.9 vs 36.6 OTUs, P = 0.0025) and elevated complexity (ie, Shannon diversity 2.46 vs 2.72; P = 0.09).

Relation Between Maternal and Infant Microbiomes

We measured the similarities in bacterial communities between different anatomical sites and at different time points for all subjects and samples. Hierarchical clustering of maternal rectal, maternal vaginal, and infant oropharygneal aspirate specimens based on Morisita-Horn similarity scores indicated that with c-section birth there was distinct, nonoverlapping clustering of all rectal, vaginal, and oropharyngeal aspirate samples (Fig. 2a). In contrast, 6 of 11 oropharyngeal aspirate samples from vaginal deliveries were clustered with either the maternal rectal or vaginal samples (Fig. 2a). An additional pair of maternal vaginal and infant oropharyngeal aspirate samples from subject S03 clustered together, but apart from all other vaginal specimens.

Mode of birth influences similarity of infant pharyngeal microbiome to maternal rectal and vaginal microbiomes. Panel A: dendograms comparing microbiomes of maternal and infant specimens. Panel B: Morisita-Horn microbiome similarity scores. REC: maternal rectal swab; VAG: maternal vaginal swab; ASP: infant pharyngeal swab; Fec0: age 0 weeks (∼36 hours) fecal sample; Fec2: age 2 weeks fecal sample; and Fec6: age 6 weeks fecal sample. Birth mode: C (c-section) or V (vaginal). Significant differences in similarity scores between vaginal and c-section groups are indicated as follows: P < 0.05, ∗∗ P < 0.01, ∗∗∗ P < 0.001.

Oropharyngeal aspirates from infants born vaginally were more similar (ie, had higher Morisita-Horn similarity scores) to both maternal rectal and vaginal samples than were oropharyngeal aspirates from c-section infants (Fig. 2b); these scores were calculated for each mother-infant dyad, rather than between unrelated mothers and infants. Surprisingly, infant stool samples taken at any time point from either delivery group were dissimilar to maternal vaginal and rectal swabs, as well as to the initial oropharyngeal samples. More important, greater similarity between successive stool samples was evident in infants born vaginally compared with stools from c-section infants, suggesting more continuity in bacterial succession in infants born vaginally.

Metagenomic Analyses of Fecal Microbiota

To explore functional consequences of intestinal microbiota differences seen by delivery mode, we performed shotgun metagenomic sequencing of week 6 stools and analyzed relative abundances of genes annotated using KEGG (34) categories (assigned through the MG-RAST server (32)). Exploratory analysis of KEGG-annotated metagenomic datasets was performed using PCoA to visualize similarities in fecal microbiome functional capacities among infants at 6 weeks of age. PCoA revealed that separation of infants along principal coordinates axis PC2 was significantly influenced by delivery mode (P = 0.0004 for t test of PC2 scores by delivery mode, Fig. 3a). Following adjustment for delivery mode, no clinical/demographic variables listed in Table 1 were associated with PC scores along axes 1, 2, or 3 (data not shown). We next examined correlations between PC scores and abundances of bacterial phyla determined by 16S sequencing (Fig. 3a, right panel). PC1 scores were positively correlated with the abundance of Proteobacteria (Spearman rho = 0.78, P < 0.0001) suggesting that the primary source of variation in metagenomes resulted from differences in fecal proteobacterial taxa, such as Escherichia spp and other Enterobacteria. In contrast, PC2 and PC3 scores were associated with variability in other predominant fecal phyla, including Actinobacteria, Bacteroidetes, and Firmicutes. Although mode of delivery was significantly correlated with PC2 scores, this relation was lost when adjusted for Bacteroidetes abundance, suggesting that colonization by Bacteroidetes mediated the relation between delivery mode and metagenomic differences in the stool.

Fecal microbiomes at 6 weeks age differ in genomic coding capacity. Panel A: PCoA of metagenomic data. PC1, PC2, and PC3, accounted for 43.2%, 13.5%, and 9.9% of the overall variance, respectively. Panel B: heatmap of KEGG categories with significantly different abundances in vaginal vs c-section deliveries (P < 0.0025; cutoff chosen to display a representative selection of genes). Z scores were calculated across each KEGG ortholog (KO) using normalized read abundances. Each row of the heatmap represents a single KEGG ortholog classification (KO), as indicated in the column labeled “KO” to the right of the heatmap. The higher-order KEGG categories are indicated in the remaining columns. Blue font color indicates genes belonging to the broad “metabolism” KEGG category. Each column represents an individual infant, with mode of birth indicated below the heatmap. C = c-section; PCoA = principal coordinates analysis; V = vaginal birth.

Fecal bacterial genes that were differentially abundant by delivery mode (at a cutoff P < 0.0025) are displayed in Figure 3b (P values of higher-order KEGG categories are presented in Supplemental Table 2, The microbiomes of vaginally delivered infants were enriched in genes involved in amino acid, carbohydrate, and vitamin metabolism. In contrast, genes encoding putative membrane transport proteins were more abundant following c-section. For example, KEGG orthologs K11707, K11708, and K11710 are predicted to be ATP-binding cassette transporter proteins that function in transport of metals such as iron, manganese, and zinc, which are essential for microbial growth.


The incidence of c-section in the United States is 32.8%, reflecting a 60% rise between 1996 and 2009 (41). Worldwide, the incidence of c-section varies widely both between and within countries (15,42,43). Advances in molecular identification of bacteria have accelerated our understanding of newborn intestinal colonization and observational evidence suggests that delivery mode is a major determinant of offspring intestinal microbiota composition in early life (6–8).

We hypothesized that delivery mode would be a significant factor influencing types of microbes present in the infant oropharynx immediately after birth. Specifically, we hypothesized that in vaginal birth, maternal vaginal and intestinal bacterial communities would transfer via oral inoculation to newborn infants. Indeed, we found that the oropharyngeal aspirates of vaginally delivered infants were dominated by typical rectovaginal organisms of the genera Lactobacillus, Ureaplasma, Bacteroides, Prevotella, and unclassified Lachnospiraceae, and Ruminococccaceae. In contrast, the oropharyngeal inocula of infants born by c-section were enriched in typical skin commensals, belonging to the genera Propionibacterium and Staphylococcus(16,44–46). Interestingly, gut and vaginal taxa, such as Lactobacillus, Bacteroides, and Lachnospiraceae, were also detected in oropharyngeal aspirates from c-section infants, perhaps indicating ectopic transfer of rectovaginal microbes during delivery. Alternatively, we cannot rule out the possibility of contamination during specimen collection or downstream processing (though negative extraction controls were devoid of 16S rRNA amplicons). Our study supports the notion of transfer of maternal bacterial communities to offspring, as reported by Dominguez-Bello et al (6), and clearly demonstrates that this important physiologic connection between mother and infant is severed in delivery by c-section.

We further hypothesized that stool samples from vaginally delivered infants, in comparison to c-section infants, would reflect bacterial communities derived from maternal fecal and vaginal microbiomes. Surprisingly, this was not the case because fecal samples collected approximately 25 hours following vaginal birth did not resemble maternal rectovaginal bacterial communities (Fig. 2A). These results indicate that infant intestinal bacterial communities are not simply inherited en masse from the mother. This is consistent with research demonstrating that environment is as important as heritability in determining the development of the intestinal microbiota (14,47). Nevertheless, mode of delivery was associated with distinct bacterial profiles in fecal samples collected at 0, 2, and 6 weeks age. This was most evident in the persistently higher abundances of Bacteroidetes observed in vaginally delivered infants, as reported previously (8). Conversely, there was a trend toward lower Firmicutes in the stool of vaginally born infants. No difference in abundance of Bacteroidetes was observed between oropharyngeal aspirates taken from vaginal and c-section infants. Thus, failure of Bacteroidetes intestinal colonization in infants born by c-section could not be explained simply by lack of exposure.

Failure or delay in colonization of commensal bacteria such as Bacteroidetes may have important consequences for the developing immune system and later inflammatory response. Bacteroides fragilis has been shown to suppress interleukin-17 production in the gut via polysaccharide A production (5). Exposure to intestinal commensal bacterial antigens and bacterial products such as short-chain fatty acids stimulates differentiation of tolerogenic T-regulatory cells (48,49). The tolerogenic response to intestinal microbiota exposure appears to be developmentally regulated, strengthening the importance of early exposure to normal commensal bacteria such as the Bacteroidetes (4).

Failure to colonize the intestine with Bacteroidetes may influence host metabolism toward weight gain. A higher ratio of intestinal Firmicutes to Bacteroidetes has been associated with obesity in adults and children (50,51). In our study, 6-week-old c-section infants had a median Firmicutes/Bacteroidetes ratio of 381.7 compared with 1.1 in vaginally born infants. Studies in mice and humans have linked a higher Firmicutes/Bacteroidetes ratio to obesity (52–55). These observations may begin to explain the recognized association between c-section and overweight in childhood (10–12).

What is the physiologic benefit of vaginal delivery to offspring intestinal bacterial colonization? We speculate that “pioneer” bacteria in the initial oropharyngeal inoculum create intestinal conditions favorable to physiologic bacterial succession with commensals such as Bacteroidetes. Lactobacilli likely are a key early determinant of early colonization dynamics, owing to their dominance in the vaginal canal of expectant mothers and in pharyngeal aspirates of vaginally delivered infants. Yet Lactobacillus was in low abundance in infant stool through 6 weeks of life, consistent with previous molecular studies that report low or fleeting abundance of Lactobacillus species after the first week of life (56). The physiologic role of lactobacilli in the neonatal intestine may exceed its measurable abundance in stool. As facultative anaerobes, lactobacilli may change the oxygen tension in the neonatal intestine, making conditions more favorable for anaerobic bacteria such as Bacteroides spp and clostridia. In addition, fermentation of carbohydrate to lactate by lactobacilli could impact the pH of the intestinal microenvironment. Finally, lactobacilli influence epithelial mucin production via epidermal growth factor receptor mediated stimulation of goblet cells (57), and Bacteroides spp bind to mucin and metabolize mucin-derived glycans during intestinal colonization (58,59). Thus, stimulation of mucin production by lactobacilli and other pioneer bacteria may enhance opportunities for establishment of Bacteroides and other members of the climax community of the adult gastrointestinal tract.

Strengths of our study include prospective collection of both maternal and infant specimens, equal breast-feeding intensity between delivery groups, and absence of infant postnatal antibiotic exposure. Our technique for isolating the oropharyngeal inoculum in infants was novel and involved coordination with delivering obstetricians.

Real and theoretical limitations of the present study deserve discussion. Our study was limited to 6 weeks follow-up and we cannot speculate on the impact of mode of delivery on intestinal microbial communities beyond early infancy. Fecal samples may not adequately characterize physiologically relevant adherent bacteria in the intestine or anatomic sites more proximal to the colon. Initial fecal samples from c-section infants were obtained 15 hours later on average than from vaginally delivered infants, but there was no difference in timing of the 2 and 6 fecal samples. Maternal exposure to antimicrobial therapy could potentially confound the relation between mode of delivery and offspring Bacteroidetes colonization because all mothers delivering by c-section received a single dose of a parenteral first-generation cephalosporin antibiotic at the time of surgery. Several lines of evidence, however, suggest that confounding is unlikely. Two mothers in our vaginal delivery group received multiple doses of parenteral penicillin before delivery, yet their infants were able to establish Bacteroidetes colonization after birth. In addition, the genera Bacteroides is generally resistant to both penicillin and first generation cephalosporins because of production of a beta-lactamase. Three of our infant aspirate samples from vaginal deliveries did not resemble maternal vaginal or rectal communities—this could represent normal physiologic variation or may represent contaminated specimens.

In summary, this study sheds light on the dynamics of infant intestinal microbiota colonization and succession. Mode of delivery impacts the microbiota of the initial neonatal oropharyngeal inoculum, and c-section leads to a more chaotic succession pattern in the infant intestine and a persistent defect in colonization of Bacteroidetes. The principal importance of the oropharyngeal inoculum at vaginal delivery is likely its effect on preparing the intestinal microenvironment for physiologic bacterial succession. A disturbance in newborn intestinal colonization and succession may influence long-term risk of disease. As opportunities for mitigating these risks in offspring are considered, further research into the physiology of infant bacterial colonization is required.


1. Peterson DA, Frank DN, Pace NR, et al. Metagenomic approaches for defining the pathogenesis of inflammatory bowel diseases. Cell Host Microbe 2008; 3:417–427.
2. Round JL, Mazmanian SK. The gut microbiota shapes intestinal immune responses during health and disease. Nat Rev Immunol 2009; 9:313–323.
3. Kau AL, Ahern PP, Griffin NW, et al. Human nutrition, the gut microbiome and the immune system. Nature 2011; 474:327–336.
4. Olszak T, An D, Zeissig S, et al. Microbial exposure during early life has persistent effects on natural killer T cell function. Science 2012; 336:489–493.
5. Mazmanian SK, Round JL, Kasper DL. A microbial symbiosis factor prevents intestinal inflammatory disease. Nature 2008; 453:620–625.
6. Dominguez-Bello MG, Costello EK, Contreras M, et al. Delivery mode shapes the acquisition and structure of the initial microbiota across multiple body habitats in newborns. Proc Natl Acad Sci U S A 2010; 107:11971–11975.
7. Azad MB, Konya T, Maughan H, et al. Gut microbiota of healthy Canadian infants: profiles by mode of delivery and infant diet at 4 months. CMAJ 2013; 185:385–394.
8. Jakobsson HE, Abrahamsson TR, Jenmalm MC, et al. Decreased gut microbiota diversity, delayed Bacteroidetes colonisation and reduced Th1 responses in infants delivered by caesarean section. Gut 2014; 63:559–566.
9. Thavagnanam S, Fleming J, Bromley A, et al. A meta-analysis of the association between Caesarean section and childhood asthma. Clin Exp Allergy 2008; 38:629–633.
10. Blustein J, Attina T, Liu M, et al. Association of caesarean delivery with child adiposity from age 6 weeks to 15 years. Int J Obes (Lond) 2013; 37:900–906.
11. Mueller NT, Whyatt R, Hoepner L, et al. Prenatal exposure to antibiotics, cesarean section and risk of childhood obesity. Int J Obes (Lond) 2014; 39:665–670.
12. Huh SY, Rifas-Shiman SL, Zera CA, et al. Delivery by caesarean section and risk of obesity in preschool age children: a prospective cohort study. Arch Dis Child 2012; 97:610–616.
13. MacDorman MF, Menacker F, Declercq E. Cesarean birth in the United States: epidemiology, trends, and outcomes. Clin Perinatol 2008; 35:293–307.
14. Yatsunenko T, Rey FE, Manary MJ, et al. Human gut microbiome viewed across age and geography. Nature 2012; 486:222–227.
15. Lumbiganon P, Laopaiboon M, Gulmezoglu AM, et al. Method of delivery and pregnancy outcomes in Asia: the WHO global survey on maternal and perinatal health 2007-08. Lancet 2010; 375:490–499.
16. Human Microbiome Project Consortium. Structure, function and diversity of the healthy human microbiome. Nature 2012; 486:207–214.
17. Hara N, Alkanani AK, Ir D, et al. Prevention of virus-induced type 1 diabetes with antibiotic therapy. J Immunol 2012; 189:3805–3814.
18. Markle JG, Frank DN, Mortin-Toth S, et al. Sex differences in the gut microbiome drive hormone-dependent regulation of autoimmunity. Science 2013; 339:1084–1088.
19. Frank JA, Reich CI, Sharma S, et al. Critical evaluation of two primers commonly used for amplification of bacterial 16S rRNA genes. Appl Environ Microbiol 2008; 74:2461–2470.
20. Lane DJ. Stackebrandt E, Goodfellow M. 16S/23S rRNA sequencing. Nucleic Acid Techniques in Bacterial Systematics. New York: Wiley; 1991. 115–175.
21. Ewing B, Green P. Base-calling of automated sequencer traces using phred. II. Error probabilities. Genome Res 1998; 8:186–194.
22. Ewing B, Hillier L, Wendl MC, et al. Base-calling of automated sequencer traces using phred. I. Accuracy assessment. Genome Res 1998; 8:175–185.
23. Edgar RC, Haas BJ, Clemente JC, et al. UCHIME improves sensitivity and speed of chimera detection. Bioinformatics 2011; 27:2194–2200.
24. Schloss PD, Westcott SL. Assessing and improving methods used in operational taxonomic unit-based approaches for 16S rRNA gene sequence analysis. Appl Environ Microbiol 2011; 77:3219–3226.
25. Pruesse E, Peplies J, Glockner FO. SINA: accurate high throughput multiple sequence alignment of ribosomal RNA genes. Bioinformatics 2012; 28:1823–1829.
26. Quast C, Pruesse E, Yilmaz P, et al. The SILVA ribosomal RNA gene database project: improved data processing and web-based tools. Nucleic Acids Res 2013; 41:D590–D596.
27. Robertson CE, Harris JK, Wagner BD, et al. Explicet: graphical user interface software for metadata-driven management, analysis and visualization of microbiome data. Bioinformatics 2013; 29:3100–3101.
28. Magurran AE. Measuring Biological Diversity. MA: Malden; 2004.
29. R: A Language and environment for statistical computing. R Foundation for Statistical Computing, 2014. Accessed June 1, 2015.
30. Frank DN, Robertson CE. de Bruijn FJ. The Phyloware Project: a software framework for phylogenomic virtue. John Wiley & Sons, Inc, Handbook of Molecular Microbial Ecology I: Metagenomics and Complementary Approaches. Hoboken, NJ: 2011.
31. Magoc T, Salzberg SL. FLASH: fast length adjustment of short reads to improve genome assemblies. Bioinformatics 2011; 27:2957–2963.
32. Meyer F, Paarmann D, D'Souza M, et al. The metagenomics RAST server—a public resource for the automatic phylogenetic and functional analysis of metagenomes. BMC Bioinformatics 2008; 9:386.
33. Wilke A, Harrison T, Wilkening J, et al. The M5nr: a novel non-redundant database containing protein sequences and annotations from multiple sources and associated tools. BMC Bioinformatics 2012; 13:141.
34. Kanehisa M, Goto S, Sato Y, et al. KEGG for integration and interpretation of large-scale molecular data sets. Nucleic Acids Res 2012; 40:D109–D114.
35. Braithwaite DT, Keegan KP. matR: Metagenomics Analysis Tools for R [software]. ed 2013. Accessed July 1, 2015.
36. Torsten H, Kurt H, van de Wiel MA, et al. Implementing a Class of Permutation Tests: The Coin Package. J Stat Softw 2008; 28:1–23.
37. Frank DN, Bales ES, Monks J, et al. Perilipin-2 modulates lipid absorption and microbiome responses in the mouse intestine. PLoS One 2015; 10:e0131944.
38. Everitt B, Hothorn T. An introduction to applied multivariate analysis. New York, NY: Springer; 2011.
39. Quackenbush J. Computational analysis of microarray data. Nat Rev Genet 2001; 2:418–427.
40. Oksanen J, Kindt R., Legendre P., O’Hara B., Simpson G.L., Solymos P., Stevens H.H., Vegan Wagner S.. Community Ecology Package. R package version 1.15–1. 2008.
41. Martin JA, Hamilton BE, Ventura SJ, et al. Births: final data for 2011. Natl Vital Stat Report 2011; 62:1–69.72.
42. Freitas PF, Drachler Mde L, Leite JC, et al. Inequalities in cesarean delivery rates by ethnicity and hospital accessibility in Brazil. Int J Gynaecol Obstet 2009; 107:198–201.
43. Kozhimannil KB, Arcaya MC, Subramanian SV. Maternal clinical diagnoses and hospital variation in the risk of cesarean delivery: analyses of a National US Hospital Discharge Database. PLoS Med 2014; 11:e1001745.
44. Grice EA, Kong HH, Conlan S, et al. Topographical and temporal diversity of the human skin microbiome. Science 2009; 324:1190–1192.
45. Grice EA, Kong HH, Renaud G, et al. A diversity profile of the human skin microbiota. Genome Res 2008; 18:1043–1050.
46. Frank DN, Feazel LM, Bessesen MT, et al. The human nasal microbiota and Staphylococcus aureus carriage. PLoS One 2010; 5:e10598.
47. Palmer C, Bik EM, DiGiulio DB, et al. Development of the human infant intestinal microbiota. PLoS Biol 2007; 5:e177.
48. Lathrop SK, Bloom SM, Rao SM, et al. Peripheral education of the immune system by colonic commensal microbiota. Nature 2011; 478:250–254.
49. Furusawa Y, Obata Y, Fukuda S, et al. Commensal microbe-derived butyrate induces the differentiation of colonic regulatory T cells. Nature 2013; 504:446–450.
50. Ley RE, Backhed F, Turnbaugh P, et al. Obesity alters gut microbial ecology. Proc Natl Acad Sci U S A 2005; 102:11070–11075.
51. Xu P, Li M, Zhang J, et al. Correlation of intestinal microbiota with overweight and obesity in Kazakh school children. BMC Microbiol 2012; 12:283.
52. Kelly CJ, Colgan SP, Frank DN. Of microbes and meals: the health consequences of dietary endotoxemia. Nutr Clin Pract 2012; 27:215–225.
53. DiBaise JK, Frank DN, Mathur R. Impact of the gut microbiota on the development of obesity: current concepts. Amer J Gastro 2012; 1:22–27.
54. Delzenne NM, Cani PD. Interaction between obesity and the gut microbiota: relevance in nutrition. Annu Rev Nutr 2011; 31:15–31.
55. Ley RE. Obesity and the human microbiome. Curr Opin Gastroenterol 2010; 26:5–11.
56. Jost T, Lacroix C, Braegger CP, et al. New insights in gut microbiota establishment in healthy breast fed neonates. PLoS One 2012; 7:e44595.
57. Wang L, Cao H, Liu L, et al. Activation of epidermal growth factor receptor mediates mucin production stimulated by p40, a Lactobacillus rhamnosus GG-derived protein. J Biol Chem 2014; 289:20234–20244.
58. Ulmer JE, Vilen EM, Namburi RB, et al. Characterization of glycosaminoglycan (GAG) sulfatases from the human gut symbiont bacteroides thetaiotaomicron reveals the first GAG-specific bacterial endosulfatase. J Biol Chem 2014; 289:24289–24303.
59. Huang JY, Lee SM, Mazmanian SK. The human commensal Bacteroides fragilis binds intestinal mucin. Anaerobe 2011; 17:137–141.

cesarean; infant; microbiome; microbiota

Supplemental Digital Content

Copyright 2016 by ESPGHAN and NASPGHAN. Unauthorized reproduction of this article is prohibited.