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

Anne-Louise Smith, Harry Owen, Karen Reynolds

    Research output: Contribution to journalArticlepeer-review

    41 Citations (Scopus)


    Heart rate variability (HRV) analysis over shorter periods may be useful for monitoring dynamic changes in autonomic nervous system activity where steady-state conditions are not maintained (e.g. during drug administration, or the start or end of exercise). This study undertakes a validation of 70 HRV indices that have previously been identified as possible for short-term use. The indices were validated over 10 × 30 beat windows using PhysioNet databases with physiological states of rest, active, exercising, sleeping, and meditating (N from 12 to 20). Baseline 95 % confidence intervals of the median were established with bootstrap resampling (10,000x). Statistical significance was assessed using the overlap of 95 % confidence intervals. Thirty-one indices could differentiate between resting and at least one physiological state using 30 beat windows. All respiratory sinus arrhythmia indices and Poincaré plot indices were strongly correlated to time domain measures (SDNN or RMSSD). Spectral indices using the Lomb-Scargle algorithm were able to correctly identify paradoxical shifts in power with meditation and reduced power in exercise. Some less-known indices gave interesting results: PolVar20 identified the higher sympathetic activity of exercise with the largest positive magnitude. These indices should now be considered for rigorous gold standard tests with pharmacological blockade.

    Original languageEnglish
    Pages (from-to)577-585
    Number of pages9
    JournalJournal of Clinical Monitoring and Computing
    Issue number5
    Publication statusPublished - Oct 2013


    • Autonomic nervous system (ANS) modulation
    • Biomedical signal processing
    • Heart rate variability (HRV)
    • Software algorithms


    Dive into the research topics of 'Heart rate variability indices for very short-term (30 beat) analysis. Part 2: validation'. Together they form a unique fingerprint.

    Cite this