POWER SPECTRUM ANALYSIS OF HEART-RATE-VARIABILITY TO ASSESS THE CHANGES IN SYMPATHOVAGAL BALANCE DURING GRADED ORTHOSTATIC TILT

Background The powers of the low-frequency (LF) and high-frequency (HF ) oscillations characterizing heart rate variability (HRV) appear to r eflect, in their reciprocal relationship, changes in the state of the sympathovagal balance occurring during numerous physiological and path ophysiological conditions. However, no adequate information is availab le on the quantitative resolution of this methodology. Methods and Res ults We studied 22 healthy volunteers (median age, 46.5 years) who wer e subjected after a rest period to a series of passive head-up tilt st eps randomly chosen from the following angles: 15 degrees, 30 degrees, 45 degrees, 60 degrees, and 90 degrees. From the continuous ECG, afte r appropriate analog-to-digital conversion, a personal computer was us ed to compute, with an autoregressive methodology, time and frequency domain indexes of RR interval variability. Spectral and cross-spectral analysis with the simultaneously recorded respiratory signal excluded its contribution to LF. Age was significantly correlated to variance and to the absolute values in milliseconds squared of very-low-frequen cy (VLF), LF, and HF components. The tilt angle was correlated to both LF and HF (expressed in normalized units [nu]) and to the LF-to-HF ra tio (r=.78, -.72, and .68; respectively). Lower levels of correlation were found with HF (in ms(2)) and RR interval. No correlation was pres ent between tilt angle and variance, VLF, or LF (in ms(2)). Individual analysis confirmed that the use of nu provided the greatest consisten cy of results. Conclusions Spectral analysis of HRV, using nu or LF- t o-HF ratio, appears to be capable of providing a noninvasive quantitat ive evaluation of graded changes in the state of the sympathovagal bal ance.