Accuracy of ultra-short heart rate variability measures

Abstract

Heart rate variability (HRV) measures have been used to assess autonomic cardiac regulation. The standard lengths used in HRV analyses are 5 minutes and 24 hours. In this paper we investigated the accuracy of three HRV measures for ultra-short record length, 10 seconds, which is the length of standard electrocardiograms. The measures chosen were: Standard Deviation Normal-Normal (SDNN), Root Mean Square of Successive Differences (RMSSD), frequency of the peak of the high-frequency (HF) spectra derived using a non-parametric spectrum method. Our analyses indicated that the RMSSD(10)s would be consistent estimates of the 5 minute RMSSD(300)s. The SDNN(10)s were found not to be accurate in our analyses. The HF peak, while promising, would require further studies.