Serial neutrophil-to-lymphocyte ratio changes in neonates evaluated for infection: A developmentally contextualized cross-cohort analysis

Background Serial neutrophil-to-lymphocyte ratio (NLR) changes in neonates may reflect both infection-related processes and rapid postnatal immune adaptation, limiting interpretation based on adult-derived assumptions. Aims To characterize longitudinal NLR changes in neonates evaluated for infection and contextualize relative change against an independent adult infection-related cohort using the same estimand. Study design Secondary exploratory longitudinal analysis of two pre-existing clinical cohorts. Subjects The source cohorts comprised 65 neonates and 100 adult hospital-admission episodes; at least one valid NLR was available for 62 neonates and 99 adult episodes. Outcome measures Cohort-specific random-intercept mixed models estimated relative NLR change across three ordered, non-time-harmonized assessments. Cross-cohort effects were expressed as ratios of geometric mean ratios (RoRs). Results The analysis included 414 NLR observations. From the first to third assessment, geometric mean ratios were 0.814 (95% CI 0.661–1.003) in neonates and 0.580 (0.487–0.691) in adults; the RoR was 1.403 (1.069–1.842). The global mixed-model test gave p = 0.0433. Robust generalized estimating equations yielded a similar RoR of 1.386 (1.029–1.866) but global p = 0.0901. Conclusions Serial NLR declined less in the neonatal cohort than in the adult comparator, but inference on the overall longitudinal pattern was model-sensitive. The findings support developmentally and clinically contextualized interpretation of neonatal serial NLR and do not identify a causal age effect or diagnostic threshold.