Reliability of Heart Rate Variability during Polysomnographic Sleep

Abstract

Spectral analysis of heart rate variability (HRV) is often used as a noninvasive index of cardiac sympathetic and/or parasympathetic activity during sleep. During controlled wake recordings, the low frequency (LF) component of HRV can be unreliable within experimental sessions and against other direct sympathetic measurements. In contrast, the high frequency (HF) component of HRV has higher reported reproducibility within controlled wake sessions. It remains unknown if the frequency component of HRV is reliable across sleep cycles within individuals during stable sleep. The purpose of the present study was to determine HRV reliability during stable stage II (N2) sleep, slow wave sleep (SWS) and rapid eye movement (REM) sleep. We hypothesized that both LF and HF components of HRV would be less reliable during REM sleep than during SWS and N2 sleep. Twenty‐seven participants (11 male, 16 female, 26±1 years, 27±1 kg/m2) were enrolled in the present study. Overnight polysomnography (PSG; NATUS, Middleton, WI) and continuous two‐lead electrocardiography (ECG) were recorded during an 8‐hour sleep opportunity. The nocturnal ECG recordings were imported into a custom software (WinCPRS, Absolute Aliens, Finland) for analysis of HRV via fast‐Fourier transformation function. Autonomic cardiac activity was quantified as LF (0.04‐0.15 Hz) and HF (0.15‐0.4 Hz) HRV. Two separate stable periods of sleep (range, 5‐10 min) absent of cortical arousal were selected from each sleep stage for comparison. Raw LF and HF HRV underwent logarithmic transformation due to non‐normal distribution. Statistical analysis included bivariate correlation of HRV across differing sleep stage cycles (α = 0.05). LF HRV was significantly correlated across two independent sleep cycles in N2 (r=0.694, p<0.001, n=26), SWS (r=0.765, p<0.001, n=25), and REM sleep (r=0.699, p<0.001, n=20). Similarly, HF HRV was reliable across two independent sleep cycles in N2 sleep (r=0.839, p<0.001, n=26), SWS (r=0.765, p<0.001, n=25), and REM sleep (r=0.881 p<0.001, n=21). Taken together, both LF and HF HRV measures were reliable across stable N2, SWS, and REM sleep. Our results support the reliability of frequency domain analysis of HRV during key PSG sleep stages within a single night, but further research across nights is warranted.