Abstract
This study examined and compared the characteristics of the power spectral indices of heart rate variability (HRV) for the different stages of sleep in order to determine which indices are useful for sleep assessment. Sixteen sets of polysomnogram (PSG) and electrocardiogram data were collected synchronously from 8 healthy women (age range, 20-44 years). PSG data were analyzed in 1-min segments, while HRV data were analyzed in 1- and 5-min segments using the MemCalc method and calculated at 3 frequency band powers: higher area in very low frequency band (VLF-hi, 0.016-0.04 Hz), low frequency band (LF, 0.04-0.15 Hz), and high frequency band (HF, 0.15-0.4 Hz). Significant differences were observed in LF/HF, HF/ (LF+HF), HF/(VLF-hi+LF+HF), and VLF-hi between all sleep stages. Compared to the HRV results obtained from 5-min segments, those from the 1-min segments showed a stronger relationship with sleep stage. Although HRV analysis during sleep typically involves the use of 5-min segments, the present results imply that improvements in technology enable the detection of sleep stages at shorter 1-min segments which correspond to PSG analysis segments.
