HEART RATE VARIABILITY INDICES EXTRACTED FROM NIGHT-TIME PERIOD ARE SUPERIOR TO THOSE EXTRACTED FROM DAY-TIME PERIOD OF SIGNAL ANALYSIS FOR VENTRICULAR ARRHYTHMIA PREDICTION
Authors: Polychronis Dilaveris, Petros Arsenos, Konstantinos Gatzoulis, George Manis, Dimitros Mytas, Anna Kefala, Dimitrios Tsiachris, Apostolos-Ilias Vouliotis, Stefanos Archontakis, Christodoulos Stefanadis
Purpose: To examine circadian properties of heart rate dynamics in ventricular arrhythmia prediction among Heart Failure (HF) patients (pts).
Methods: We screened 232 HF pts (age:64±13years, male:84%, LVEF: 32±10, NYHA:2.3±0.5, CAD:82%,
DCMP:18%) under optimum treatment with ECG, SAECG, ECHO and 24 hours HOLTER. After 21 months of follow up the study population was divided to the HIGH risk(54 pts) and the LOW risk(178pts) groups according to three arrhythmia events/surrogates:1.clinical VT/VF 2. ICD’s appropriate activation 3. confirmed Sudden Cardiac Death. All indices were calculated and statistically analyzed for the two groups. Holter’s signals were analyzed for periods: Entire 24 hours:08.00-0.80, Day-time: 06.00-23.00, Night time:23.00-06.00.
Results: A standard model adjusted for LVEF, fQRS, NSVT 1/24hour and VPBs 240/24 hour was used for multiple logistic regression analysis. Three separate analysis were performed with this model including SDNN24h /SDNNday /SDNNnight .The only independent and important arrhythmia predictor was proved SDNNnight with OR: 0.981 (p=0.006, 95%CI: 0.969-0.994). (Table 22.1.)
Conclusions: In this HF population, HRV indices extracted from signal’s night-time period were superior from those extracted from the entire or the day-time period signals. Impaired night-time heart rate adaptation and loss of circadian properties of HR dynamics proved of prognostic importance for arrhythmia prediction.