Covariates
For each exposure-outcome association, we identified a minimally sufficient set of confounders using directed acyclic graphs (Figure S1).30,31 These included maternal age (continuous), parity (0 vs. ≥1 previous delivery), pre-pregnancy Body Mass Index (BMI; continuous), maternal asthma (yes vs. no), maternal depression (yes vs. no), and smoking (yes vs. no), alcohol consumption (yes vs. no), and use of calcium-containing supplements (yes vs. no) during pregnancy. All confounder data were obtained from the prenatal questionnaires.