bioRxiv2018
Rapid Eye Movement (REM) sleep is a peculiar neural state showing a combination of muscle atonia and intense cortical activity. REM sleep is usually considered as a unitary state in neuroscientific research; however, it is composed of two different microstates: phasic and tonic REM. These differ in awakening thresholds, sensory processing, and cortical oscillations. Nevertheless, studies examining cortical oscillations during REM microstates are scarce, and used low spatial sampling. Here, we analyzed the data of 18 healthy individuals assessed by high-density sleep EEG recordings. We systematically contrasted phasic and tonic REM periods in terms of topographical distribution, source localization, as well as local, global and long-range synchronization of frequency-specific cortical activity. Tonic periods showed relatively increased high alpha and beta power over frontocentral derivations. In addition, higher frequency components of beta power exhibited increased global synchronization during tonic compared to phasic states. In contrast, in phasic periods we found increased power and synchronization of low frequency oscillations coexisting with increased and synchronized gamma activity. Source localization revealed several multimodal, higher-order associative, as well as sensorimotor areas as potential sources of increased high alpha/beta power during tonic compared to phasic REM. Increased gamma power in phasic REM was attributed to medial prefrontal and right lateralized temporal areas associated with emotional processing. Our findings emphasize the heterogeneous nature of REM sleep, expressed in two microstates with remarkably different neural activity. Considering the microarchitecture of REM sleep may provide new insights into the mechanisms of REM sleep in health and disease.\n\nSignificance StatementRapid Eye movement (REM) is composed of the alternation of two markedly different microstates: phasic and tonic REM. Here we present the first high-resolution topographical study that systematically contrasted spectral power, neural synchronization and brain topography of spontaneous cortical oscillations across phasic and tonic microstates. Phasic REM was characterized by the combination of synchronized low frequency oscillations resembling deeper sleep stages and high frequency oscillations reflecting intense cortical activity. In contrast, tonic REM sleep showed increased oscillatory activity in the alpha and beta ranges that may reflect a shift towards increased environmental alertness resembling resting wakefulness. Exploring the neurophysiological aspects of REM microstates may shed new light on the mechanisms of REM sleep in healthy and pathological conditions.