Study of spectral components of ventricular repolarization variability by multiple correspondence analysis

In this study we propose a novel approach, based on multiple correspondences analysis (MCA), to identify the relationships between the spectral variables obtained from HRV signals and those from ventricular repolarization variability (VRV). Parameters were extracted from RR and RT time series from 47 ECG records from aerobic and anaerobic sportsmen and sedentary subject. These numerical variables were transformed in categorical variables using a 7 classes fuzzy coding and several MCA were carried out. Results show that: i) aerobic disciplines present more variability than anaerobic ones, ii) mean RT interval are longer in sportsmen than in control subjects, iii) only the low frequencies of the HRV and VRV series are correlated, evidencing the heart rate modulation on ventricular repolarization, iv) the high frequencies of the RT space seem to be inadequate markers for the characterization of the ANS modulation on ventricular repolarization