Scoring transitions to REM sleep in rats based on the EEG phenomena of pre-REM sleep: An improved analysis of sleep structure

Abstract

Algorithms for scoring sleep/waking states and transitions to REM sleep (NRTs) in rats are presented and validated. Both algorithms are based on electroencephalographic (EEG) power in delta (0.5-4.0 Hz), theta (6-9 Hz) and sigma (10-14 Hz) frequency bands, and electromyogram (EMG) intensity. Waking is scored when EMG intensity is high or (sigma power)-(theta power) is low. Nonrapid eye movement (NREM) sleep is scored in nonwaking epochs having high (delta power)/(theta power). Rapid eye movement (REM) sleep is scored in nonwaking epochs having low (delta power)/(theta power). NRTs are identified by the EEG phenomena of the pre-REM sleep phase of NREM sleep. Algorithms are validated by comparison with records scored independently by two investigators based on visual examination of EEGs and EMGs. The sleep/waking- state scoring algorithm produces greater than 90% agreement with visual scoring. The NRT-scoring algorithm produces 88-92% agreement with visual scoring. Scoring NRTs based on the phenomena of the pre-REM sleep phase of NREM sleep, instead of relying solely on REM sleep expression for identification of REM sleep onset, reveals a significant population of brief REM sleep episodes that are ignored by most sleep cycle analyses and allows independent quantification of REM sleep timing and maintenance.