Abstract 11057: Measures of Heart Rate Variability as Predictors of Clinical Outcome in Patients with Chronic Heart Failure - Results from the Myovasc Study

Abstract

Introduction: Evidence supports a potential prognostic value of heart rate variability (HRV) in patients with chronic heart failure (HF). Comprehensive data comparing different domains of HRV in relation to clinical outcome in HF are scarce. Methods: Subjects with chronic HF (i.e. AHA stage C or D) from the MyoVasc study (NCT04064450; N=3,289), a prospective cohort study, were analyzed. Individuals with pacemaker or atrial fibrillation were excluded. All participants received a highly standardized medical-technical examination during a five-hour visit in a dedicated study center. Data from 24-hour Holter ECG recordings were analyzed resulting in HRV markers from the time and frequency domain, and non-linear HRV measures. Information on clinical outcome, i.e. worsening of HF (WoHF), was derived from structured follow-up with subsequent validation and independent adjudication of endpoints. Random forest models were used to identify the most predictive marker of each HRV domain, which were further analyzed with multivariable Cox regression. Results: The analysis sample comprised 554 HF subjects (mean age: 66.4±9.9 years, 34.3% female sex) with HRV data. During a median observational period of 4.08 years (interquartile range 2.22; 5.22), 153 participants experienced WoHF. In random forest analyses, the rMSSD (time domain), high frequency power in normalized units (HF nu ; frequency domain), and SD1/SD2 (non-linear measure) were ranked as most predictive HRV markers for WoHF in each HRV domain. Subsequent Cox regression analysis demonstrated that log(SD1/SD2) (hazard ratio (HR) per standard deviation 1.50, 95%CI 1.29;1.75; P<0.0001), HF nu (HR SD 1.47, 95%CI 1.23;1.77; P<0.0001) and log(rMSSD) (HR SD 1.36, 95%CI 1.17;1.57; P<0.0001) were strong predictors of HF outcome independent of age, sex, clinical profile, and medication. In Cox regression analysis with adjustment for NT-proBNP and the established HF risk score MAGGIC, HF nu (HR SD 1.25, 95%CI 1.05;1.48; P=0.012) and log(SD1/SD2) (HR SD 1.24, 95%CI 1.07;1.43; P=0.005) remained independent predictors of WoHF. Conclusions: Markers of HRV were identified as strong predictors for HF outcome independent of established risk markers, suggesting a potential to advance risk stratification.