Early Childhood Antibiotics and Chronic Pediatric Conditions: A Retrospective Cohort Study

Abstract

BACKGROUND: Early childhood antibiotics have been implicated in chronic pediatric conditions, but many studies leave concerns about unmeasured confounding. We evaluated associations between early childhood antibiotics and multiple pediatric conditions. METHODS: We performed a retrospective cohort study using electronic health records data (United Kingdom, 1987-2020). The primary exposure was antibiotic prescriptions between birth and age 2 years. Outcomes were newly diagnosed chronic pediatric conditions (asthma/allergic, autoimmune, neurodevelopmental/psychiatric) or forearm fracture (negative control). Hazard ratios (HRs) with 95% confidence intervals (CIs) were estimated using multivariable Cox regression models adjusted for maternal, child, and area-based socioeconomic covariates. Sibling-matched analyses were conducted using conditional Cox regression. RESULTS: Among 1 091 449 children, antibiotic exposure before age 2 was positively associated with asthma (HR, 1.24 [95% CI, 1.22-1.26]), food allergy (HR, 1.33 [95% CI, 1.26-1.40]), and allergic rhinitis (HR, 1.06 [95% CI, 1.03-1.10]), with stronger associations with multiple antibiotic courses. Sibling-matched analyses yielded similar findings. Early childhood antibiotic exposure was also dose-dependently associated with intellectual disability (≥5 vs 1-2 courses: HR, 1.73 [95% CI, 1.49-2.01]; sibling-matched: HR, 2.79 [95% CI, 1.87-4.18]). Consistent associations were not observed for celiac disease, inflammatory bowel disease, juvenile idiopathic arthritis, psoriasis, type 1 diabetes, attention-deficit/hyperactivity disorder, autism spectrum disorders, or anxiety. Sibling-matched results and a negative control outcome suggested minimal confounding. CONCLUSIONS: Children receiving multiple antibiotic courses before age 2 were more likely to develop asthma, food allergies, allergic rhinitis, and intellectual disability. However, antibiotic-associated risks of most autoimmune, neurodevelopmental, and psychiatric conditions studied were minimal.