RESUMEN
BACKGROUND: The paediatric early versus late parenteral nutrition in critical illness (PEPaNIC) multicentre, randomised, controlled trial showed that, compared with early parenteral nutrition, withholding supplemental parenteral nutrition for 1 week in the paediatric intensive care unit (PICU; late parenteral nutrition) reduced infections and accelerated recovery from critical illness in children. We aimed to investigate the long-term impact on physical and neurocognitive development of early versus late parenteral nutrition. METHODS: In this preplanned 2-year follow-up study, all patients included in the PEPaNIC trial (which was done in University Hospitals Leuven, Belgium; Erasmus Medical Centre-Sophia Children's Hospital, Rotterdam, Netherlands; and Stollery Children's Hospital, Edmonton, AB, Canada) were approached for possible assessment of physical and neurocognitive development compared with healthy children who were matched for age and sex, and who had never been admitted to a neonatal ICU or a PICU. Assessed outcomes comprised anthropometric data; health status; parent-reported or caregiver-reported executive functions and emotional and behavioural problems; and tests for intelligence, visual-motor integration, alertness, motor coordination, inhibitory control, cognitive flexibility, and memory. To address partial responses among the children tested, we did multiple data imputation by chained equations before univariable and multivariable linear and logistic regression analyses adjusted for risk factors. This trial is registered with ClinicalTrials.gov, number NCT01536275. FINDINGS: At the 2-year follow-up, 60 (8%) of 717 children who received late parenteral nutrition and 63 (9%) of 723 children who received early parenteral nutrition had died (p=0·81). 68 (9%) of 717 children who received late and 91 (13%) of 723 children who received early parenteral nutrition were too disabled for neurocognitive assessment (p=0·059), and 786 patients (395 assigned to late and 391 assigned to early parenteral nutrition) consented for testing. 786 patients and 405 healthy control children underwent long-term outcome testing between Aug 4, 2014, and Jan 19, 2018, and were included in the imputation model for subsequent multivariable analyses. Late parenteral nutrition did not adversely affect anthropometric data, health status, or neurological functioning, and improved parent-reported or caregiver-reported executive functioning (late vs early parenteral nutrition ß estimate -2·258, 95% CI -4·012 to -0·504; p=0·011), more specifically inhibition (-3·422, -5·171 to -1·673; p=0·0001), working memory (-2·016, -3·761 to -0·270; p=0·023), and meta-cognition (-1·957, -3·694 to -0·220; p=0·027). Externalising behavioural problems (ß estimate -1·715, 95% CI -3·325 to -0·106; p=0·036) and visual-motor integration (0·468, 0·087 to 0·850; p=0·016) were also improved in the late parenteral nutrition group compared with the early parenteral nutrition group. After Bonferroni correction for multiple comparisons, the effect on inhibitory control remained significant (p=0·0001). INTERPRETATION: Withholding early parenteral nutrition for 1 week in the PICU did not negatively affect survival, anthropometrics, health status, and neurocognitive development, and improved inhibitory control 2 years after PICU admission. FUNDING: European Research Council Advanced Grant, Methusalem programme provided by the Flemish Government, Flemish Agency for Innovation by Science and Technology (IWT), Research Foundation Flanders (FWO), Sophia Children's Hospital Foundation (SSWO), Stichting Agis Zorginnovatie, Erasmus Trustfonds, and European Society for Parenteral and Enteral Nutrition (ESPEN) research grant.