Associations of species and amplicon sequence variants (ASVs) differential abundance with risk factors from the year 1 or birth surveys, after FDR correction. Infant diet features were dichotomized based on frequency (daily or 3–5 times weekly versus seldom or 1–2 times weekly). Months of breastfeeding (total/exclusive) were dichotomized as 1–4 months compared to 5 or more. Psychosocial vulnerability was dichotomized for high/low risk based on the total index. Risk factors included gastroenteritis; infection requiring antibiotics; otitis in the first year; increased psychosocial vulnerability; maternal smoking during pregnancy; fewer months of total/exclusive breastfeeding; more frequent chocolate, fries, and chips; and birth by cesarean section, while protective factors represented the inverse. Log2FC values were determined in DESeq2. See also Table S3 for comprehensive statistics across ND models.
