EEG microstates reveal distinct network dynamics in lucid and non-lucid REM sleep
Xinlin Wang, Emma Peters, Johannes Strelen, Nathan Lockhart, Michael S. Franklin, Stephen LaBerge, Thomas Koenig, Martin Dresler, Daniel Erlacher
Consciousness and Cognition April 13, 2026 DOI: 10.1016/j.concog.2026.104047 (opens in new tab) via OpenAlex
Summary
AI-generated from the abstractDuring lucid dreaming, the dreamer is aware they are dreaming. To identify brain network activity linked to this conscious experience, 39 sleep recordings (lucid vs. non-lucid REM sleep) from 8 lucid dreamers were analyzed using EEG microstate analysis. Microstates A and G dominated during lucid REM, while microstates B, C, and D dominated during non-lucid REM. These microstates were associated with emotional processing (A), visual processing (B), salience network activity (C), executive functions (D), and the default mode network (G). Results suggest lucidity involves increased self-visualization, metacognition, and executive processing, with decreased emotional processing and default mode network activity, indicating distinct network dynamics in lucid REM sleep.
Study at a glance
| Characteristics | Observational study Peer reviewed |
|---|---|
| Sample size | 8 |
| Population | Lucid dreamers |
| Topics | Lucid dreaming |
| Keywords | Electroencephalography Sleep system call Psychology Dynamics music |
| Key finding | Lucid REM sleep is characterized by increased microstates A and G, associated with emotional processing and default mode network, and decreased microstates B, C, and D, linked to visual processing, salience, and executive functions. |
Abstract
Abstract During lucid dreaming, the dreamer is aware that they are dreaming. In past years, numerous studies have explored the characteristics of lucid dreams, which link lucid dreams to vivid visual perception, positive emotions, and wake-like metacognition and executive functions. To determine brain network activity associated with the conscious experience during lucid dreaming, we analyzed 39 sleep recordings (non-lucid REM vs. lucid REM) with 32-channel polysomnography from 8 lucid dreamers, using EEG microstate analysis. We found that microstates A and G dominated during lucid REM sleep compared to non-lucid REM sleep, and microstates B, C, and D dominated during non-lucid REM sleep compared to lucid REM sleep. We explored the correlation of our microstate maps with previous findings based on topographical similarities of the microstate maps in our study. This suggests that in our study, microstate A might be associated with emotional processing, microstate B with visual processing, microstate C with salience network activity, microstate D with executive functions, and microstate G with the default mode network. Our results suggest that lucidity during REM sleep is associated with increased self-visualization, metacognition, and executive processing, along with decreased emotional processing and reduced default mode network activity. Additionally, we found the inverse relationship between the presence of microstates/networks associated with regions that serve specific functions and evidence for the function being used. This might indicate the inhibitory function of the EEG microstates during sleep. Our study provides novel insight into the distinct network dynamics in lucid and non-Lucid REM sleep.