We used polytomous logistic regression to estimate relative risk ratios (rRRs) with 95% confidence intervals (CIs)33 for each early-life exposure in association with very early (≤10 years), early (11 years), late (14 years), and very late (≥15 years) menarche relative to typical ages at menarche (12-13 years). We combined women who reported “definitely” and “probably” for prenatal and infancy feeding exposures to estimate their associations with age at menarche compared with women who reported “probably not” or “definitely not”. We initially examined age at menarche associations separately across all birth-order groups and 5-year maternal age categories (data not shown) and then combined categories having similar frequency of early and late menarche for reporting associations. Birth weight was categorized in clinically relevant categories: <2,500 g (low), 2,500-3,999 g (reference), and ≥4,000 g (high). Because there was no difference in age at menarche for high versus reference birth weight (data not shown), we further combined those two categories. All polytomous logistic regression models included race-ethnicity, participant’s birth decade, and childhood family income as confounders because they may influence the exposures and age
