Prenatal and postnatal exposure to antibiotics and risk of food allergy in the offspring: A nationwide birth cohort study in South Korea

Background: There are only preliminary studies examining the associations of postnatal antibiotic exposure with food allergy in childhood, and the effect of antibiotic exposure in utero has not been resolved. Thus, we aimed to investigate the effect of prenatal and postnatal antibiotic exposure on the risk of food allergy in childhood. Methods: Using the nationwide birth cohort in South Korea, all 3,163,206 infants (pairing mother; n = 2,322,735) born in South Korea between 2010 and 2017 were included in the analysis. The primary outcome was the diagnosis of food allergy, and the observation period was between January 1, 2009, and December 31, 2020. We implemented four different designs for the study, which consisted of a full unmatched cohort, 1:1 propensity-matched cohort, sibling comparison cohort, and health screening cohort along with multiple subgroup analyses. Results: During the follow-up period (median 6.92 years [IQR, 4.72–9.00]) of the 3,161,858 infants (52.6% male) in the birth cohort, 29,973 (1.9%) were diagnosed with food allergies. After a 1:1 propensity score matching, the use of antibiotics increased the risk of overall food allergy (prenatal [HR, 1.05; 95% CI, 1.04–1.09] and postnatal [HR, 1.05; 95% CI, 1.01–1.10] periods). The association was more significantly accentuated when antibiotic exposure was used in the short term, and the children were born preterm or with low birthweight; however, a trimester-specific effect was not observed. We observed more pronounced risks of food allergy in the health screening cohort (prenatal, 17%; postnatal, 15%), thus addressing the adverse effects of critical factors including maternal BMI, smoking status, and type of infant feeding. Similar trends were observed across all four differnt cohorts. Conclusion: This study reported a moderate association between early-life antibiotic use and subsequent food allergy during childhood throughout four different designs of analyses. This study suggests that clinicians need to consider the risks and benefits of antibiotics when administering antibiotics to individuals in the prenatal and postnatal periods.