Page 25 - Read Online

Horwell et al. Microbiome Res Rep 2025;4:1 https://dx.doi.org/10.20517/mrr.2024.32 Page 3 of 20

knowledge base and where the future of this new discipline is going.

MICROBIOME VARIATION WITH AGE
The first months and years of life are a very dynamic period for the microbiome, with wide-reaching
consequences for future health. Several large international longitudinal studies have identified three distinct
periods of rapid and dynamic microbiomic progression [14-17] . These can be summarised as a developmental
phase (the first 12 months of life) that is dominated by the effects of breastfeeding, a transitional phase (the
second year of life, or the cessation of breast milk feeding) where the microbiome is highly plastic and
adjusts to solid food, and culminating in a stable phase at approximately 36-60 months that closely
resembles the microbiome of an adult [14,18] .

Starting with a blank slate?
There has been a long-held belief that the human foetus develops in a sterile uterine environment . The
[19]
typical 48 h delay in collecting meconium that might allow ex-utero colonisation to occur
origin of this goes back to research that was unable to detect microbial life in the meconium, amniotic fluid,
or placenta using traditional culture-based methods and microscopy . Indeed, chorioamnionitis, an intra-
[20]
amniotic infection typically secondary to E. coli or group B streptococcus, is associated with miscarriage,
preterm birth, and increased neonatal mortality, ergo the presence of bacteria was deemed to be an
abnormal phenomenon.

New techniques, however, such as next-generation sequencing (NGS), have transformed the detection and
identification of microbiota. A growing body of evidence has raised questions over the “sterile uterus
hypothesis”. Aagaard et al. were the first group to systematically describe a microbiome of placental samples
from both healthy pregnancies and preterm births. They found a low-abundance but metabolically rich
microbiome, which interestingly reflected the mother’s dental microbiome, raising the possibility of
haematogenous spread from the oral cavity to the placenta . Several other studies have since published
[21]
evidence for distinct microbiomes relating to the placenta, amniotic fluid, and meconium in healthy
neonates, resembling the faecal, oral, and vaginal microbiome of the mother [22-24] . Furthermore, animal
models have shown a similar effect. A Spanish group inoculated pregnant BALB/c mice with an oral
solution of a tagged mutant strain of Enterococcus faecium. This strain was then detected from samples of
amniotic fluid taken at the time of caesarean section (CS), an effect not seen in the control group . In our
[25]
view, this is perhaps the most compelling evidence against the sterile uterus hypothesis.

This antithesis to the canon has caused controversy, however. Several papers have subsequently highlighted
methodological errors pertaining to contamination from both the collection method (e.g., epidermal
contamination from the puncture site for amniocentesis, or contamination during veterinary CS), and the
[26-28] .
Accordingly, an international group repeated these experiments, with more rigorous controls for
contamination, and failed to find evidence of intra-uterine colonization of the foetal GIT .
[29]

An international consensus statement has made clear that the principal viewpoint of the academic
community is still that of the sterile womb, citing, amongst several reasons, the high rates of false-positive
[28]
results that exquisitely sensitive NGS can produce . It should be noted that even in the event of true-
positive results, bacterial DNA does not equate to living and metabolically active bacterial species that have
colonised the foetal microbiome. Furthermore, there have been over seven decades of germ-free animal
experimentation , where mammals are born via CS at term. A valid assumption follows that if in-utero
[30]
colonisation had occurred, this would have been detected in the numerous control groups during this
time .
[27]