2.3 Potential confounders or covariates
We considered the following covariates for potential association with exposure and outcome based on previous studies.2,12Maternal factors included maternal age, urbanization, insurance property, mode of delivery, type of pregnancy, acetaminophen use during pregnancy, maternal atopic disorders, gestational infections, and maternal comorbidities, including hypertension, diabetes mellitus, hyperlipidemia, malignancy, urinary tract infection, maternal asthma, rheumatoid arthritis, systemic lupus erythematosus, Sjögren syndrome, endometriosis, gestational diabetes, and preeclampsia. Infant factors included birth year, child’s sex, length of gestation, birth weight, and APGAR score at birth. In addition, postnatal antibiotic and acetaminophen exposure in the first year of life were also selected as variables based on previous reports.