Fetal and Early Life Exposure to Antibiotics and the Risk of Neurodevelopmental Disorders: A Nationwide Population-Based Cohort Study with Sibling Analysis

Background: While antibiotics are commonly used during pregnancy and infancy to treat infections, growing evidence indicates that early life microbiome disruption is associated with childhood neurodevelopmental disorders. Given the strong influence of antibiotics on microbiota composition, a comprehensive investigation of this topic is warranted.

Objectives: To evaluate the association between antibiotics use during pregnancy and infancy and the risk of neurodevelopmental disorders in children.

Methods: We used Korea’s mother-child linked healthcare database to identify all pregnancies resulting in live births between 2009 and 2020. Exposure was defined as ≥1 systemic antibiotic prescription during 1) pregnancy and 2) first six months of life in infants. The reference group was those unexposed to antibiotics in the respective time window. Study outcomes, or neurodevelopmental disorders, were autism spectrum disorder (ASD), intellectual disorder (ID), language disorder (LD), and epilepsy. Infants were followed up from birth (pregnancy analysis) or six months after birth (infancy analysis) until the earliest of an outcome diagnosis, death, or Dec 31, 2021. After 1:1 propensity score matching based on various maternal and infant characteristics (e.g. indication), hazard ratio (HR) with 95% CI for each outcome was estimated using Cox hazard model. A sibling analysis was performed to additionally account for unmeasured familial factors by using a stratified Cox model.

Results: After matching, we identified 980,872 mother-child pairs for pregnancy analysis and 804,887 infants for infancy analysis; all covariates were well-balanced between groups in both analyses. While antibiotic exposure during pregnancy was associated with increased risks of all four study outcomes in the overall cohort, these estimates were all attenuated towards the null in the sibling analyses (ASD [HR 1.06, 95% CI 1.00-1.12], ID [1.00, 0.93-1.07], LD [1.05, 1.02-1.09], and epilepsy [1.03, 0.98-1.08]). Likewise, no association was observed between antibiotics exposure during infancy and ASD (1.00, 0.96-1.03), ID (1.07, 0.98-1.15), and LD (1.04, 1.00-1.08) in the sibling analyses; however, a small increased risk of epilepsy was observed (1.13, 1.09-1.18).

Conclusions: The findings of this study suggest that after accounting for various confounding factors, antibiotic use during pregnancy or infancy was not associated with an increased risk of ASD, ID, and LD in children. However, antibiotic use during infancy was modestly associated with epilepsy, even after controlling for indications and familial factors.