Global and local complexity of intracranial EEG decreases during NREM sleep.
Level 4 - case-series / case-control
Case series of clinical patients evaluated under an observational within-subject protocol.
PubMed 30042832 · doi:10.1093/nc/niw022
What was done
The authors analyzed spontaneous intracranial depth electrode (iEEG) recordings from 10 patients with epilepsy during wakeful rest and various sleep stages: early-night non-rapid eye movement (NREM), late-night NREM, and rapid eye movement (REM) sleep. They calculated three measures of dynamical complexity globally across distributed channels and locally within single regions: Lempel-Ziv complexity, amplitude coalition entropy, and synchrony coalition entropy.
What was found
The abstract reports no numerical values. In all 10 participants, all three complexity measures decreased substantially across distributed brain regions during early-night NREM sleep relative to wakeful rest. This decrease was partially reversed during late-night NREM sleep, while REM sleep complexity values were similar to wakeful rest. This pattern was mirrored locally in almost all cases. Frontal lobe channels exhibited elevated signal complexity compared to other regions, and differences were not solely attributable to spectral power changes.
Why it matters
This study provides direct human intracranial evidence that neural dynamical complexity tracks conscious state transitions across sleep cycles, supporting theoretical models that link consciousness to dynamical complexity.
Limits
The study is limited by a small sample size (n = 10) consisting entirely of patients with epilepsy undergoing invasive monitoring, which may not generalize to healthy brains. The abstract reports no numerical estimates, effect sizes, confidence intervals, or precise channel numbers.
Cited by
- supports States of reduced consciousness such as deep sleep, coma, or general anesthesia are characterized by low brain entropy and highly predictable ongoing brain activity.