EEG Microstate Dynamics Associated with Dream-Like Experiences During the Transition to Sleep

Abstract

Consciousness always requires some representational content; that is, one can only be conscious about something. However, the presence of conscious experience (awareness) alone does not determine whether its content is in line with the external and physical world. Dreams, apart from certain forms of hallucinations, typically consist of non-veridical percepts, which are not recognized as false, but rather considered real. This type of experiences have been described as a state of dissociation between phenomenal and reflective awareness. Interestingly, during the transition to sleep, reflective awareness seems to break down before phenomenal awareness as conscious experience does not immediately fade with reduced wakefulness but is rather characterized by the occurrence of uncontrolled thinking and perceptual images, together with a reduced ability to recognize the internal origin of the experience. Relative deactivation of the frontoparietal and preserved activity in parieto-occipital networks has been suggested to account for dream-like experiences during the transition to sleep. We tested this hypothesis by investigating subjective reports of conscious experience and large-scale brain networks using EEG microstates in 45 healthy young subjects during the transition to sleep. We observed an inverse relationship between cognitive effects and physiological activation; dream-like experiences were associated with an increased presence of a microstate with sources in the superior and middle frontal gyrus and precuneus. Additionally, the presence of a microstate associated with higher-order visual areas was decreased. The observed inverse relationship might therefore indicate a disengagement of cognitive control systems that is mediated by specific, inhibitory EEG microstates.

Keywords: Dream-like experiences; EEG microstates; Hypnagogic state; Phenomenal consciousness; Reflective consciousness; Transition to sleep.

Fig. 1

Microstate topographies averaged across subjects and wakefulness levels as well as mean values of the spatiotemporal parameters for each microstate class. Contribution: percentage of time covered by each microstate class. Occurrence: mean number of times that each microstate occurred per second. Duration: mean duration in milliseconds per occurrence of each microstate class. Mean GFP: mean global signal strength of each microstate class

Fig. 2

Predicted mean values of contribution (a), occurrence (b), duration (c) and global field power (d) as a function of rating score (centered mean scores) for each microstate class. Percentage of time covered by microstate 4 significantly increased with increasing mean rating scores and percentage of time covered by microstate 2 significantly decreased with increasing mean rating scores in a-c. Mean rating score: positive values indicate more dream-like experiences being indicative for decreased reflective awareness. Negative values correspond to more ordinary ratings. Shaded areas indicate 95% confidence intervals. The F-values show the fixed effects of microstate x rating score

Fig. 3

Source localization of EEG microstate class 2 (a) and 4 (b). Yellow areas depict estimated average source maxima