J Occup Health 1997; 39: 154–155

Accuracy of Pulse Rate Variability Parameters
Obtained from Finger Plethysmogram:
A Comparison with Heart Rate Variability
Parameters Obtained from ECG

Takayuki KAGEYAMA¹, Michinori KABUTO¹, Tetsuya KANEKO^{2<,sup>, and Noriko NISHIKIDO3}

¹Urban Environment and Health Project, National Institute for Environmental
Studies, ²Department of Environmental Health, Faculty
of Health Sciences, Kyorin University and ³Health Development Department,
Fujitsu Facom Information Processing Corporation

Key words: Heart rate variability, Pulse rate variability, Finger plethysmogram, Electrocardiogram, Breathing frequency

Two spectral components of heart rate variability, respiratory sinus arrhythmia (RSA) and Mayer wave-related sinus arrhythmia (MWSA), respectively provide noninvasive indices of cardiac vagal activity and systemic sympathetic activity with vagal modification1–3). Their amplitudes are usually obtained from spectral analysis of R-R intervals measured on an electrocardiogram (ECG). In the present study, it was examined whether peak-to-peak intervals measured on a finger plethysmogram (FPG) can be alternative data to the ECG R-R intervals for the same analysis. If we can substitute the FPG for the ECG, this provides an advantage in a routine physical examination at a work site, because the FPG is often easier to be applied to many subjects than the ECG.

Subjects and Methods
Thirty-two healthy volunteers (6 men and 26 women aged 20.2 ± 1.7 (mean ± S.D.)) who gave written informed consent participated in this experiment. Following a rest period of 120 min after a meal, they lay quietly for 5 min in a sound- insulated and electrically shielded room, after which their ECGs (standard lead I), reflective-type FPGs (Model TL612-T, Nihon-Kohden Co., Ltd., Tokyo) on the left index finger, and pneumatogram (PMG: Model TR762-T, Nihon-Kohden Co., Ltd., Tokyo) were simultaneously recorded for 5 min in each of the following positions: supine rest, sitting rest (90° tilt), and standing rest in this order. The time constant for the FPG was set at 0.3 sec in order to minimize spontaneous baseline fluctuation, according to Minami et al.4). The recording started after the heart rate became stable. The subjects were instructed to breathe regularly without deep breaths and to change their posture actively. Data analysis was carried out following the method previously reported1,2,5,6).

Autoregressive spectral analysis and component wave analysis were applied to the 200 successive R-R intervals measured
on the ECG with a tachograph (Model TM55, CERX Co., Ltd., Tokyo). Following the formula proposed by Hayano1), component coefficients of variation and center frequency of RSA (C-CV_RSA, %, and f_RSA, Hz) and those of MWSA (CCV _MWSA, %, and C-CV_MWSA, Hz) were calculated. From the peak-to-peak intervals measured on the FPG for the same time period as the above ECG, similar variables were calculated in the same way. Since the above two C-CV_RSA values were obtained in the same breathing condition, respiratory modification of C-CV_RSA 4,5) was neglected. Mean breathing frequency (BF, Hz) was calculated from the PMG for the same time period.

Results
The pulse rate based on the FPG always agreed with the heart rate based on the ECG. Figure 1 shows a typical power spectrum of HRV based on the ECG and that based on the FPG both of which were simultaneously obtained for the same subjects at supine rest. Similar spectral components, MWSA and RSA, were extracted from both of them by component wave analysis. C-CV_RSA and C-CV_MWSA values obtained from the FPG agreed well with those obtained from the ECG regardless of the posture (Fig. 2). f_RSA values obtained from the FPG agreed well with BF obtained from the PMG (Fig. 3) as well as f_RSA obtained from the ECG. No sex difference was observed in these results.

Read the full study: Accuracy of Pulse Rate Variability Parameters