Revisiting Left Ventricular Ejection Fraction Levels: A Circadian Heart Rate Variability-Based Approach

• Main Authors: Mohanad Alkhodari, Herbert F. Jelinek, Shiza Saleem, Leontios J. Hadjileontiadis, Ahsan H. Khandoker
• Format: Article
• Language: English
• Published: IEEE 2021-01-01
• Series: IEEE Access
• Subjects: Heart failure; coronary artery disease; left ventricular ejection fraction; cardiac circadian rhythm; heart rate variability; Jenks natural breaks
• Online Access: https://ieeexplore.ieee.org/document/9541231/

Analysis of heart failure is important in clinical practice to ensure coronary artery disease (CAD) patients will be provided with appropriate timely treatment. The current gold-standard, echocardiography, although reliable, provides a once-off left ventricular ejection fraction (LVEF) measurement and does not provide information about heart function disturbances during day/night cardiac cycles. Additionally, the discrimination between heart failure with preserved and mid-range ejection fraction remains challenging in echocardiography tests. In this vein, this study was sought to investigate the ability of heart rate variability (HRV) in categorizing CAD patients into multiple LVEF groups throughout the 24-hour circadian cycle and checking its agreement with established gold-standard echocardiography-based guidelines. A total of 92 CAD patients who have suffered from heart failure were included in this study. The newly introduced index, HRV ejection fraction (HRVEF), was based on optimizing indices extracted from HRV data, which are correlated with the sympathetic and parasympathetic nervous systems, to form group membership of the preserved (HFpEF), mid-range (HFmEF), and reduced (HFrEF) LVEF categories. HRVEF groups optimized on hourly basis through Jenks natural breaks algorithm exhibited a consistent pattern with a goodness of variance fit (GVF) of more than 70% accuracy during the late-night to early-morning (01:00-08:00) and evening (17:00-23:00) time periods. At these hours, several HRV indices were found significant (p-value ≤0.05) in differentiating between HRVEF groups using statistical analysis of variance (ANOVA) test. These features include the successive differences between normal heartbeats (RMSSD), low and high frequency (LF, HF) power, standard deviation of normal heartbeats (SD2), short-term scaling exponent (alpha1), and percentage of normal heartbeats in alternation segments (PAS). The findings of this study suggest HRV as a promising supplementary tool to the once-off echocardiography for timely LVEF measurements and heart failure prognosis. It paves the way towards multi-time HRV-based estimations for LVEF according to the association between LVEF and HRV indices to better demonstrate the circadian cardiac function at different LVEF levels in CAD patients.