Heart rate variability indices for very short-term (30 beat) analysis. Part 2: validation

Area of Science:

• Physiology
• Biomedical Engineering
• Data Science

Background:

• Short-term heart rate variability (HRV) analysis is crucial for monitoring dynamic autonomic nervous system (ANS) activity during non-steady-state conditions.
• Previous research has identified numerous HRV indices potentially suitable for short-term analysis, but validation is needed.

Purpose of the Study:

• To validate 70 previously identified HRV indices for short-term (30-beat windows) physiological state monitoring.
• To assess the ability of these indices to differentiate between various physiological states including rest, activity, exercise, sleep, and meditation.

Main Methods:

• Validation of 70 HRV indices using 10 x 30 beat windows from PhysioNet databases.
• Establishment of baseline 95% confidence intervals via bootstrap resampling (10,000 iterations).
• Assessment of statistical significance using confidence interval overlap.

Main Results:

• Thirty-one HRV indices demonstrated significant differentiation between resting and at least one other physiological state.
• Respiratory sinus arrhythmia and Poincaré plot indices showed strong correlations with time-domain measures (SDNN, RMSSD).
• Spectral indices, particularly using the Lomb-Scargle algorithm, accurately detected power shifts during meditation and exercise; PolVar20 highlighted sympathetic activity during exercise.

Conclusions:

• A subset of HRV indices is validated for short-term monitoring of physiological states.
• Specific indices show promise for detecting dynamic ANS changes during exercise and meditation.
• Further validation using pharmacological blockade is recommended for these promising short-term HRV indices.