Postnatal antibiotic exposure due to maternal group B streptococcus is associated with childhood asthma

Abstract

Background: Previous studies identified early-life antibiotic exposure as a risk factor for childhood asthma. However, this association may be confounded by an indication, as antibiotics are often prescribed for respiratory infections, which themselves promote asthma. To mitigate this bias, we aim to assess the unique contribution of postnatal antibiotic therapy, given to non-infected infants for maternal indication, on childhood asthma risk.

Methods: We screened electronic medical records to identify healthy full-term infants born during 2006-2018 to mothers with a positive group B streptococcus (GBS) vaginal culture. Infants with postnatal respiratory symptoms/pneumonia or positive blood/cerebrospinal fluid cultures were excluded. The primary outcome was an asthma diagnosis by age 6 years. We fitted a multivariable quasi-Poisson regression model to assess the unique contribution of antibiotic treatment to asthma diagnosis. As a validation step, we utilized a propensity model in which infants treated with antibiotics were matched 1:3 with infants not treated.

Results: The cohort included 14,807 infants, of whom 311 received antibiotics. After controlling for potential confounders, postnatal antibiotic exposure was associated with higher asthma risk (adjusted risk ratio [aRR] = 1.3, 95%; confidence interval [CI] 1.04-1.61, p = .017). Higher asthma risk was validated in the propensity model (aRR = 1.49, 95% CI 1.12-1.96 p = .005). Postnatal antibiotic therapy was also associated with secondary outcomes such as the short-acting beta-ag use (aRR = 1.13, 95% CI 0.99-1.28, p = .072) and allergic rhinitis diagnosis (aRR = 3.00, 95% CI 1.43-6.30, p = .003).

Conclusions: Postnatal antibiotic therapy administrated for maternal GBS, not confounded by infants' infections, was associated with higher childhood asthma risk.

Keywords: asthma; group B streptococcus; postnatal antibiotic exposure.

FIGURE 1

Study flowchart describing the cohort selection from the Clalit Health Services database. From 16,481 eligible newborns, 1674 were excluded due to specific conditions listed as exclusion criteria. The final cohort included 14,807 infants, of whom 311 received antibiotics for maternal positive GBS vaginal cultures, and 933 matched newborns who were not treated with antibiotics.

FIGURE 2

Forest plots comparing disease associations across two analyses. (A) presents adjusted risk ratios (aRR) with 95% confidence intervals (CI) and p‐values from multivariable regression in a non‐matched cohort. (B) shows risk ratios (RR) with 95% CI and p‐values from univariable regression in a matched cohort. In both panels, squares represent point estimates (aRR or RR), horizontal lines indicate 95% CI, and vertical dashed lines represent the null value of 1.0. Statistical significance is indicated by p‐values. Asthma diagnosis: At least two documented prescriptions for short‐acting beta‐agonists (SABA) and inhaled corticosteroids (ICS) within two consecutive years at ages 5–6 years, in addition to at least one ICD 9 code for asthma from either a community or hospital record.