Neonatal sleep and early executive function in preterm children

A prospective Leuven cohort recorded overnight multichannel EEG in preterm infants at term-equivalent age and reassessed the children at 2 years corrected age with standardized tests of early executive function; the team used automated sleep classification and general linear models to examine associations between specific neonatal sleep features and later executive functioning.

Researchers followed a single-center prospective cohort (the Resilience study) of preterm infants born between August 2016 and July 2018. Infants born before 34 weeks gestation and/or with birth weight under 1500 g had overnight multichannel polysomnography at term-equivalent age before discharge. At 2 years corrected age a blinded assessment of motor, cognitive, and early executive function was performed using established tasks. The study used validated automated pipelines to detect artefacts and classify neonatal sleep stages and planned statistical models to relate neonatal sleep features to later executive function. Study details: - Cohort and inclusion: Infants born at University Hospitals Leuven (Aug 2016–Jul 2018) born <34 weeks gestational age and/or birth weight <1500 g; several exclusion criteria applied and parents provided written consent. - PSG methods: Overnight multichannel EEG (9 electrodes) recorded before discharge; no sedative drugs were given within one month of the recording; recordings were preprocessed and visually displayed according to standard settings. - Automated analysis: Two validated pipelines were applied—automated artefact detection and a convolutional neural network sleep classifier—with preprocessing, hidden Markov smoothing, and postprocessing to flag and handle unreliable segments. - Primary sleep features: Preselected explanatory variables included percentage of total sleep time (%TST), transitional sleep, AS1, QS-HVS, QS-TA, and the number of unexpected transitions per hour; survival analysis and Cox regression were used to evaluate state durations and PMA effects. - Outcome assessment: At 2 years corrected age, executive function was measured with three tasks (Multisearch Multilocation for working memory, Reversed Categorization for cognitive flexibility, Snack Delay for inhibitory control) and a composite EF z-score was computed. - Statistical approach: Associations between neonatal sleep features and 2-year EF were planned to be examined in separate general linear models adjusted for seven covariates (PMA, sex, birth weight, days of mechanical ventilation, neonatal skin breaking procedures, amount of kangaroo care, and maternal education) with bootstrapping and checks for collinearity. Summary: The article describes the cohort, recording and automated analysis methods, the preselected neonatal sleep features, the age-appropriate executive function tests at 2 years, and the planned adjusted statistical models. Undetermined at this time.