Columbia study

Validation study report (short version)

Research Holter Laboratory Columbia University


  1. Statement of Study Purpose

Heart Rhythm Instruments, Inc. has developed algorithms for power spectral analysis of RR intervals (the Nerve Express algorithms). Heart Rhythm Instruments, Inc. has packaged the Nerve Express algorithms in an instrument suitable for stand-alone office use. The purpose of this evaluation was to compare the Nerve Express algorithms with the “gold standard” CHRONOS algorithms to test the reliability of the Nerve Express algorithms for power spectral analysis of RR intervals. The CHRONOS algorithms have been shown to predict death in coronary heart disease and to quantify physical fitness, but they are accessed via Holter technology and are not packaged as a stand-alone office instrument.
The Nerve Express algorithms are conveniently packaged for office use. If they are equivalent to CHRONOS, they would constitute a reliable office system used for many purposes: assessment of risk in cardiovascular disorders; assessment of physical fitness; documentation of benefit for cardiac, chiropractic, or orthopedic rehabilitation; and quantification of drug effects on the autonomic nervous system.

B. Background and Rationale

Power spectral measures of normal RR intervals recorded for a short (5 minute) or long (24-hour) period of time predict cardiac death in patients with coronary heart disease or valvular heart disease and in random samples of middle-aged persons (1-10).
Power spectral measures of normal RR intervals also provide an excellent measure of physical fitness (11-14). Short-term recordings of the ECG show cyclic fluctuations that are well characterized by frequency domain methods, either non-parametric fast Fourier transforms or parametric autoregression (15-17). These methods are capable of quantifying the energy of cyclic fluctuations in RR intervals. The Joint Task Force on Heart Rate Variability of the European Society of Cardiology and the North American Society for Electrophysiology and Pacing recommended that, for assessment of autonomic modulation of RR intervals, frequency domain methods be applied to 5-minute ECG recordings (17). Frequency domain analysis reveals two peaks of energy (power) in the power spectra obtained with these methods: a low-frequency peak (LF) that represents modulation of RR intervals by baroreflex activity and a high-frequency peak (HF) that represents modulation of RR intervals by breathing. The vagus nerve carries efferent impulses in both cases. High values for measures of RR interval variability indicate greater physical fitness and a prognosis.
The CHRONOS algorithms developed jointly by The Research Holter Laboratory at Columbia University and Dr. Jeffrey N. Rottman are the best-documented algorithms for predicting cardiac death in patients with coronary heart disease (18). However, data for the landmark studies with the CHRONOS algorithms were acquired using Holter recorders, a cumbersome way to acquire short-term data in the office. Also, highly trained research
technicians have extensively edited ECG recordings before the CHRONOS algorithms are applied to them.
Nerve-Express is a PC-based system for assessing the state of the autonomic nervous system using proprietary power spectral technology developed over a 15-year period. The algorithms are packaged for convenient office use, but the ability of the Nerve Express algorithms to predict cardiac death or physical fitness has not been validated. The purpose of the current research is to test the reliability of the Nerve Express algorithms for power spectral analysis of RR intervals by comparing them with the CHRONOS algorithms.
If the Nerve Express algorithms prove reliable for predicting death and/or predicting physical fitness, this system will be the first convenient, inexpensive tool for the office assessment of cardiovascular risk and physical fitness will finally be available.


Research Holler Laboratory Columbia University

The Research Holter Laboratory at Columbia University has conducted many studies to document the ability of the CHRONOS algorithms to predict cardiac death and/or physical fitness. Heart Rhythm Instruments, Inc. has developed Nerve Express algorithms to analysis fluctuations in RR intervals. Among these are routines to process artifact and ectopic beats in the RR data stream. These routines have the potential to make Nerve Express more fully automated which will not only broaden the venues in which the instrument can be used but also decrease the cost of assessing the autonomic nervous
system by reducing expensive technician time.


The results of our comparison of the Nerve Express algorithms with the CHRONOS algorithms indicate that the results obtained are similar and suggest that the Nerve Express algorithms should predict death in coronary heart disease and level of physical fitness.
P.S. Heart Rhythm Instruments Inc, and trade name “Nerve-Express”  are previous names of the company and Intellewave device.
Copy of full study report available upon request.