Sleep defined by arousal threshold reveals decreases in corticocortical functional correlations independently from the conventional sleep stages

Abstract

Reports of sleep-specific brain activity patterns have been constrained by assessing brain function as it related to the conventional polysomnographic sleep stages. This limits the variety of sleep states and underlying activity patterns that one can discover. The current study used all-night functional MRI sleep data and defined sleep behaviorally with auditory arousal threshold (AAT) to characterize sleep depth better by searching for novel neural markers of sleep depth that are neuroanatomically localized and temporally unrelated to the conventional stages. Functional correlation values calculated in a four-min time window immediately before the determination of AAT were entered into a linear mixed effects model, allowing multiple arousals across the night per subject into the analysis, and compared to models with sleep stage to determine the unique relationships with AAT. These unique relationships were for thalamocerebellar correlations, the relationship between the right language network and the right "default-mode network dorsal medial prefrontal cortex subsystem," and the relationship between thalamus and ventral attention network. These novel neural markers of sleep depth would have remained undiscovered if the data were merely analyzed with the conventional sleep stages. Significance StatementThe original classification of sleep into distinct stages used behavioral characteristics. With the proliferation of new techniques, the first experiments might have been to perform correlations with arousal threshold. These experiments have never been performed, either with functional MRI or with any other modern technique. The amount of communication between brain regions as measured by all-night functional magnetic resonance imaging was correlated with arousal threshold. This revealed novel neural markers of sleep depth that would have remained undiscovered if the data were merely analyzed with the conventional stages. This expands our understanding of sleep and its functions beyond the constraints imposed by the conventional stages.