Neonatal brain activity across sleep states: Evidence from resting EEG and auditory event-related potentials

The collection of electroencephalography (EEG) data in neonates typically occurs during sleep. EEG activity is highly sleep-state dependent, therefore differentiating between states during data processing can provide important insights into neurodevelopment. Despite this, there have been a paucity of studies directly comparing how infant EEG data, especially event-related potentials, differ between these sleep states. Here, we adapted the Maryland Analysis of Developmental EEG pipeline (MADE) to integrate sleep-state coding into its automated preprocessing pipeline. We recorded EEG in 102 sleeping one month old infants and evaluated their responses during a resting state and during a three-stimuli auditory oddball paradigm. Examination of resting-state power revealed significant differences between two sleep states, namely active (AS) and quiet (QS) sleep across all frequency bands in both absolute and relative power. For the auditory oddball paradigm, we computed responses to both standard and deviant tones and then created a difference score reflecting the Mismatch Response (MMR). For the novel tones we examined the evoked response (P300). Results revealed for the MMR, a significant electrode cluster by sleep-state interaction (F = 5.36, p = .01), indicating that the MMR was present at all three electrode clusters during AS (p-values <.05), but only at the frontal cluster during QS (t = 2.05, p = .04). There were no differences in the amplitude of the P300 to the novel sound as a function of sleep state.