Coming Into Resistance: Clinically Important Antibiotic Resistance Genes in NICU vs. Healthy Infants

Importance: Antibiotic-resistant infections in neonates often arise from the infant's own colonizing bacteria harboring antibiotic resistance genes (ARG). This study evaluates ARG that impact antibiotic selection and drive resistance-related mortality, termed clinically important antibiotic resistance genes (CI-ARG). Objective: To determine differences in gut CI-ARG abundance in children from birth through 3 years old, stratified by neonatal intensive care unit (NICU) admission and antibiotic (abx) exposure in the first two months Design: Cohort study (sequencing, analysis completed 2025-2026) using shotgun metagenomic sequencing of stool samples from infants in two studies: Neonatal Intestinal Microbiome (+NICU groups, 2015-2022), The First 1000 Days of Life and Beyond (-NICU groups, 2018-2021). Infants grouped into four cohorts: +NICU/+abx, +NICU/-abx, -NICU/-abx, and -NICU/+abx. Prespecified CI-ARG abundance compared longitudinally among groups. Setting: NICU, well-baby nursery, Inova Fairfax Hospital, Virginia Participants: +NICU criteria: admission to the NICU, anticipated length of stay (LOS) of at least five days, maternal age at least 18 years. -NICU: parents with good general health and age at least 18 years enrolled prior to infant birth. Stool samples: adequate stool processing volume, at least one stool sample from before one month and after one year of age, 7-day minimum between serial samples. Exposures: NICU versus well-baby nursery; antibiotics versus none in first two months of corrected age Main Outcomes and Measures: Predefined primary outcome- longitudinal CI-ARG abundance by group, compared using linear mixed effects models with adjustment for multiple comparisons (Benjamini-Hochberg procedure). Predefined secondary outcomes- differences in overall microbiota and resistome composition. Results: 1000 serial stool samples, 179 infants (birth-39 months). Longitudinal abundance significantly differed among cohorts for 11 CI-ARG (p < 0.05); often higher in the +NICU groups in early life, with or without antibiotic exposure. Abundance of ARG conferring resistance to aminoglycosides, macrolides, and fluoroquinolones later decreased below -NICU levels. mecA, which often drives methicillin resistance in Staphylococcus, was detected only in +NICU infants during premature period. Conclusions: This longitudinal study demonstrates significant longitudinal CI-ARG differences in children grouped by early-life NICU status and/or antibiotic exposure. Future studies should explore the mechanisms driving these CI-ARG changes and evaluate their impact on infection risk and community spread.