Selection of optimal epoch duration in assessment of rodent sleep-wake profiles

Abstract

One of the major inconveniences encountered in sleep studies is the time-consuming labor involved in electroencephalography (EEG) analysis. The choice of epoch duration is of prime importance and is highly related to the outcome of EEG analysis. This study was designed to find the optimal EEG epoch duration that accurately reveals the sleep-wake profiles of animals and relieve researchers of the laborious analysis of rodent sleep-wake parameters. We analyzed mouse and rat EEG signals with commonly used epoch durations (4, 8, 10, 20, and 30 s) and compared the resulting sleep-wake profiles in terms of amounts of sleep and wakefulness, number and duration of episodes, number of sleep-wake stage transitions, and EEG power spectra. There were no statistical differences in the amount and EEG power density of wakefulness, rapid eye movement (REM) and non-REM sleep among the 5 epoch durations we used. However, the shorter the epoch, the more numerous the stage transitions and the larger the episode number. The purpose of an experiment should be a main criterion for determining the optimal epoch length for the experiment. When the amount of each stage and power density of EEG signals are desired, a 30-s epoch duration is appropriate, to save time; and the results obtained are no different from those found by analysis of other epoch durations. If stage transition and duration and number of episodes are to be considered, a 4-s epoch duration is suggested as the best choice for analysis.