Item

Abstract

The sleeping brain enters a processing mode in which it continues to monitor the environment, balancing sleep protection with the need to respond to stimuli. However, the relationship between top-down attentional mechanisms and bottom-up stimulus input is unclear. Furthermore, whether pre-sleep light exposure may modulate attentional resources and thereby modify stimulus processing acutely and during subsequent sleep is unknown. To address these questions, 29 healthy participants (18-30 y, 15 women) performed a local-global auditory oddball task (Bekinschtein et al., 2009) during wakefulness and sleep with polysomnography. They came to the laboratory twice with visits differing only in the light exposure (1-h pre-sleep; high- vs. low-melanopic metamers, difference factor 2x in melanopsin activation). We used temporal generalisation analyses based on event-related potentials to evaluate (dis-)similarities in the neural signatures across vigilance stages (i.e., wake, N1-N3, REM) and between the two light exposure conditions. Effects of interest were (i) processing of global deviants, which is modulated by attention, and (ii) processing of omissions (silence where sound was expected) of bottom-up input as a measure of top-down expectations. Global effects were evident during wakefulness, N2 and N3 sleep. Contrasting earlier findings, this indicates that the ability to generate relatively complex task-relevant expectations and compare them to stimulus input is retained during sleep. However, in the absence of bottom-up input, effects were no longer evident during sleep underlining the parsimony and thus sleep-protective nature of this processing mode. Processing was not modulated by pre-sleep light exposure targeting specifically the melanopsin system using metameric light.