Lead. The rigid separation of sleep into discrete REM and NREM stages has been a cornerstone of sleep scoring since Rechtschaffen and Kales. But a new paper in Sleep Medicine challenges that tidy picture. Qi Li, Zhijun Zhong, Qian Ouyang, Xueliang Zou, Fang Yi, Chunmei Gan, and Dongyuan Yao at Jiangxi Provincial People’s Hospital and Nanchang Medical College report that characteristic EEG waveforms from both REM and NREM sleep can appear simultaneously during the transition between sleep stages. The finding, published online ahead of print, raises important questions about how sleep stage boundaries are defined and whether the conventional categorical framework captures the true neurophysiology of sleep state change.
What it argues. The study, titled “Coexistence of REM and NREM sleep characteristic EEG waves during sleep transition” (DOI: 10.1016/j.sleep.2026.109150), provides electrophysiological evidence that the border between sleep stages is not a sharp switch but a zone of mixed activity. During the transition period, the brain may generate sleep spindles or K-complexes (hallmarks of NREM sleep) at the same time as rapid eye movements or sawtooth waves characteristic of REM sleep. Rather than a clean stage boundary, the EEG shows a hybrid state where the brain has not fully committed to one sleep type. This challenges the assumption underlying standard polysomnography that a given 30-second epoch belongs exclusively to one stage. If hybrid EEG patterns occur regularly, then conventional scoring rules may systematically misclassify transitional epochs, flattening the richness of sleep architecture into a single label that does not reflect the underlying neural state.
Why it matters. Sleep stage classification drives clinical diagnosis and research. Misclassification of transitional sleep could affect metrics such as REM latency, NREM-REM cycle duration, and stage shift counts (parameters used to assess disorders from narcolepsy to depression). More broadly, the paper points toward a graded or continuous model of sleep state regulation, where the brain does not jump discretely between stages but passes through intermediate states with mixed electrophysiological signatures. This aligns with recent work using intracranial recordings and high-density EEG that suggests sleep stage transitions involve gradual changes in multiple oscillatory networks rather than simultaneous, all-or-nothing shifts. If hybrid EEG states are a normal part of sleep microarchitecture, then scoring guidelines may need updating to account for them, and automated sleep staging algorithms may need retraining to recognize transitional patterns rather than forcing them into discrete bins. The findings also open a window into the neural mechanisms that coordinate the transition (specifically, how thalamocortical circuits driving spindles interact with brainstem and basal forebrain systems that generate REM at the same moment).
Source. Li Q, Zhong Z, Ouyang Q, Zou X, Yi F, Gan C, Yao D. Coexistence of REM and NREM sleep characteristic EEG waves during sleep transition. Sleep Medicine. 2026. DOI: 10.1016/j.sleep.2026.109150. Published online ahead of print; abstract not yet indexed in PubMed. Source URL: https://pubmed.ncbi.nlm.nih.gov/42470900/