Session S73.1

Time Varying Heart Rate Variability Analysis of Active Orthostatic and Cold Face Tests Applied Both Independently and Simultaneously

AR Mejía-Rodríguez, MJ Gaitán-González*, S Carrasco-Sosa, A Guillén-Mandujano

Universidad Autonoma Metropolitana
Mexico City, Mexico

Active orthostatic test (AOT) causes a sympathetic response, while cold face test (CFT) produces primarily a vagal response. Although the common HRV analysis of these responses assumes stationarity, major transitions are present when the stimulus is applied or removed, and it may be difficult to achieve stationarity in both AOT and CFT. To delve into the dynamics of any variable, the appropriate approach for analysis must be time-varying. The aim of this paper is to analyze from a non-stationary perspective these two maneuvers, carried out either independently or simultaneously.
RR interval from ECG records of 20 subjects were obtained on three different conditions: AOT, CFT and during the simultaneous application of both stimuli (SS). Each maneuver consisted on three one minute stages: control, stimulus and recovery. Root mean square of successive differences of the cardiac intervals (rMSSD), low frequency component (LF), high frequency component (HF), as well as total power (TP) and LF/HF ratio were estimated beat-by-beat through a time-varying autoregressive model based on adaptive filtering. Each record was divided into six intervals: control; start, course and end of the maneuver; early and late recovery (at 65, 90, 115, 130 and 170 s, respectively). Difference with respect to control of 5 s average of each HRV index around intervals of interest was used for the statistical comparison of maneuvers and intervals by Friedman's test with multiple comparison using a significant level of 0.05.
In general, vagal indexes (rMSSD, HF) showed that CFT and SS behaved similarly (p<0.05) although SS presented values between CFT and AOT, while sympathetic indexes (LF/HF, mean RR) indicated an AOT sympathetic predominance over SS (p>0.05) except for late recovery. An important change was observed in several indexes during early recovery for all conditions. This overshoot should be further explored in future research using time-varying analysis.
The behavior of short duration maneuvers with substantial transients such as the ones studied could be explored using time-varying variability indicators showing the dynamics of autonomic modulation in the course of the maneuvers.

(Abstract Control Number: 144)