Reply to Putignani et al.: Vagina as a major source of natural inoculum for the newborn

We thank Putignani et al. (1) for their insightful comments on our work describing the initial microbiotas of newborns at the time of delivery (2). Ravel et al. (3) recently reported that vaginal communities not dominated by lactobacilli but by strictly anaerobic bacteria were common in healthy black and Hispanic women. We, therefore, suggest that poor representation of lactobacilli is not necessarily pathologic. The physicians involved in our study reported healthy pregnancies and no symptoms of bacterial vaginosis.

We agree with Putignani et al. (1) that the maternal fecal microbiota may also serve as a source of anaerobic bacteria and could be transferred more readily to vaginal than to caesarean section (C-section)–delivered newborns. Indeed, several culture-based studies have revealed a higher rate of early colonization by Bacteroides in vaginal deliveries (4), and in our study, vaginal and C-section deliveries resulted in 0.54% and 0.004%, respectively, representation by Bacteroides on average. However, Bacteroides can also be found at low abundances in the vagina and on skin, underscoring the difficulty in defining a sole source for these bacteria. We had to restrict our sampling to the approved protocol, which did not include maternal fecal samples. We have reanalyzed our data and incorporated fecal samples from healthy adults from a previous study (5), and we found that the initial microbiotas of vaginally delivered babies were, on average, more similar to vaginal than to gut microbiotas (Fig. 1).

In our study, samples were taken within 15 min of delivery, and the results for C-section–delivered babies were consistent with the reported presence of skin microbes in indoor hospital environments (6). We did not measure the consumption of colostrum by the babies during the first 24 h of life. We agree that future studies with much broader environmental sampling could be informative for tracking the sources of our earliest inocula.

In relation to cephalosporin administration “several hours” before the C-section (2), we did not have any input into the routine procedures of physicians at the study site, and our protocol simply allowed us to record how and what antibiotic was applied.

Finally, we do believe that differences in developmental microbiota may lead to differences in microbiota/disease risks, but this was beyond the scope of our work. We believe that not enough direct evidence has yet been generated showing an association between the establishment of bacterial diversity early in infancy and the individual’s risk for adult diseases, particularly cardiovascular diseases.

We agree with a call for a large cohort study with detailed ongoing sampling of babies, family members, physicians, and environment. Much remains to be discovered about the principles of community assembly of the human microbiome.