Beschreibung:
<jats:sec><jats:label /><jats:p>An increasing interest has been focused on the assessment of autonomic regulation during sleep by means of heart rate variability (HRV) analysis. The common tools to analyze HRV require signal stationarity, but no studies addressed this problem during sleep. Aim of our study was to evaluate the prevalence of non‐stationarities on HRV series during different sleep stages.</jats:p><jats:p>We recorded ECG in rats for 8 hours with episodes of wake (W), slow wave sleep (SWS) and REM sleep and evaluated the stationarity of HRV series of lengths ranging from 100 to 300 beats (with steps of 50). For each series, 8 sub‐sequences of 50 samples were randomly selected and checked if mean and variance remained constant over all the patterns by means of analysis of variance (F‐test) and Bartlett test respectively.</jats:p><jats:p>We found increasing percentages of non‐stationarity (from 60 to 100%) with series length increasing from 100 to 200 beats, while all the series longer than 200 were totally non‐stationary. The percentage of non‐stationarity were similar between W, SWS and REM. In conclusion, our data suggest to use time series shorter than 200 beats to reduce non stationarities when assessing HRV during sleep in rats. Moreover, the sources on non‐stationarity may be different in different sleep‐wake stages, being prevalently related to motor activity during W and SWS and more dependent upon sympathetic‐parasympathetic shifts during REM sleep.</jats:p></jats:sec>