Antibiotic use and infant skin microbiome
There has been some evidence from meta-analyses for the link between antibiotic use in pregnancy (n=18 studies)47 and early life (n=22 studies)48 and infant AD. These studies varied considerably in design, ranging from cohort studies to case-control studies and retrospective medical record reviews, yielding high heterogeneity. Wan et al47 in particular showed that significant links were demonstrated only for retrospective studies but not for prospective or cross-sectional studies, suggesting that a degree of ascertainment bias may exist. Confounding by indication may also be present in studies where the indication for antibiotic prescriptions may be due to early skin disease or infections. The meta-analyses were also not able to delineate the specific at-risk window in pregnancy or early life where antibiotic administration might conclusively increase the risk of infant AD.
It has also been shown that the skin microbiome in healthy adults receiving antibiotic treatment manifested greater changes, as demonstrated by increased Bray-Curtis dissimilarities after treatment compared to baseline.49 Longer antibiotic courses induced more profound and persistent changes, suggesting that there was also a dose/duration-dependent effect.
Antenatal, perinatal, and postnatal antibiotic exposure may theoretically likewise cause perturbations in the host microbiome and this has been demonstrated in gut microbiome studies.50 There is, however, limited data describing the impact of maternal antibiotic exposures on their infants’ skin microbiome and its potential impact on infant AD. Hence, definitive conclusions on whether the skin microbiome plays any role in the weak link between maternal and early life antibiotic exposures and infant AD cannot be made.