Abstract
Background The PEPaNIC RCT, which investigated critically ill children admitted to the pediatric intensive care unit, showed that early administration of parenteral nutrition (early PN) as compared with withholding PN for 1 week (late PN) negatively affected 6 neurocognitive functions assessed 2 years later (1). However, it is theoretically possible that age at time of exposure determines whether early PN has negative or positive impact on long-term physical and cognitive development, possibly resulting in a neutral outcome for the total study population. In this secondary analysis of the PEPaNIC RCT, we investigated whether age at exposure to early PN determined its 2 years developmental impact. Methods The 786 children who were evaluated 2 years after inclusion in the PEPaNIC RCT for health status, anthropometrics, executive functioning, emotional and behavioral problems, intelligence and visual motor integration were categorized for age at randomization (0–17 years). We defined 4 ±similarly sized age groups based on previously reported timing of cerebral maturation spurts: neonates ≤28 days old (n=121), infants 29 days to <11 months old (n=239), toddlers 11 months to <5 years old (n=223), children 5 years or older (n=203). For each outcome, interaction between the randomized intervention and age at randomization was assessed with a multivariable linear regression analysis adjusted for baseline risk factors. For the outcomes with an interaction p≤0.15, we subsequently compared the adjusted effect of early PN versus late PN within each age group. Results Interaction between randomization and age group was identified for weight, development of inhibitory control, flexibility, working memory, planning and organization, metacognition, total executive functioning and internalizing and total behavioral problems. In particular among infants 29 days to <11 months old, harm by early PN was observed for several neurocognitive functions [inhibitory control (p=0.008), flexibility (p=0.02), working memory (p=0.009), planning and organization (p=0.004), metacognition (p=0.008), total executive functioning (p=0.004), internalizing (p=0.005) and total behavioral problems (p=0.01)]. Among toddlers 11 months to <5 years old, neurocognitive harm by early PN was only observed for inhibitory control (p=0.003) and total executive functioning (p=0.02). In neonates ≤28 days old, early PN did not affect neurocognitive development whereas it increased weight (p=0.03) but not height. Among children 5 years or older, early PN only appeared to affect development of planning and organization in a positive manner (p=0.03). Conclusion Critically ill children who were exposed to early PN at an age between 29 days and 11 months were found to be most vulnerable for neurocognitive developmental harm evoked by early administration of PN, as assessed 2 years later. 1 Verstraete et al. Lancet Respir Med 2018