Redundancy among risk predictors derived from heart rate variability and dynamics: ALLSTAR big data analysis

Area of Science:

• Cardiology
• Biomedical Engineering
• Data Science

Background:

• Numerous heart rate variability (HRV) and heart rate dynamics indices are proposed as cardiovascular mortality risk predictors.
• The degree of redundancy between the predictive powers of these indices remains largely unknown.

Purpose of the Study:

• To investigate the redundancy among various heart rate variability (HRV) and heart rate dynamics predictors of cardiovascular mortality.
• To determine if high-risk predictors tend to co-occur in individuals.

Main Methods:

• Utilized 24-hour ambulatory ECG data from the Allostatic State Mapping by Ambulatory ECG Repository (ALLSTAR) project.
• Calculated standard deviation of normal-to-normal R-R interval (SDNN), very-low-frequency power (VLF), scaling exponent α1, deceleration capacity (DC), and non-Gaussianity λ25s.
• Dichotomized calculated values into high-risk and low-risk categories based on established cutoffs for predicting mortality post-acute myocardial infarction.
• Examined the rate of accumulation of multiple high-risk predictors in individuals and compared it to expected rates assuming independence.

Main Results:

• Analysis of 265,291 ECG records revealed varying prevalence of high-risk values for individual predictors (e.g., λ25s at 18.82%, α1 at 15.75%).
• The observed rate of individuals with no high-risk predictors was 66.68%, slightly higher than the expected 60.74% if predictors were independent.
• Significantly higher-than-expected rates of individuals with multiple (two or more) high-risk predictors were observed, particularly for three or more co-occurring risk factors (e.g., 4.26 times expected for three, 47.66 times for four, and 1140.66 times for five).

Conclusions:

• High-risk predictors of HRV and heart rate dynamics demonstrate a tendency to cluster within the same individuals.
• This clustering indicates a substantial degree of redundancy among these proposed cardiovascular mortality risk predictors.