David A. Brodie

Heart rate variability (HRV) is a known risk factor for mortality in both healthy and patient populations. There are currently no normative data for short-term measures of HRV. A thorough review of short-term HRV data published since 1996 was therefore performed. Data from studies published after the 1996 Task Force report (i.e., between January 1997 and September 2008) and reporting short-term measures of HRV obtained in normally healthy individuals were collated and factors underlying discrepant values were identified. Forty-four studies met the pre-set inclusion criteria involving 21,438 participants. Values for short-term HRV measures from the literature were lower than Task Force norms. A degree of homogeneity for common measures of HRV in healthy adults was shown across studies. A number of studies demonstrate large interindividual variations (up to 260,000%), particularly for spectral measures. A number of methodological discrepancies underlined disparate values. These include a systematic failure within the literature (a) to recognize the importance of RR data recognition/editing procedures and (b) to question disparate HRV values observed in normally healthy individuals. A need for large-scale population studies and a review of the Task Force recommendations for short-term HRV that covers the full-age spectrum were identified. Data presented should be used to quantify reference ranges for short-term measures of HRV in healthy adult populations but should be undertaken with reference to methodological factors underlying disparate values. Recommendations for the measurement of HRV require updating to include current technologies.

INTRODUCTION: Chronic exercise training produces a resting bradycardia that is thought to be due partly to enhanced vagal modulation. PURPOSE: The aim of the present study was to determine the effects of exercise training on heart rate and measures of heart rate variability associated with vagal cardiac modulation and to quantify the relationship between changes in these measures. METHODS: A random effects model of effect size (d) for change in high frequency (HF) power and RR interval was calculated. Within-group heterogeneity was assessed using the Q statistic. Where heterogenous effects were found, subgroup analyses were performed using the between-group Q statistic. RESULTS: A meta-analysis of 13 studies measuring HF (N=322 cases) produced an overall effect size of d=0.48 (C.I. 0.26-0.70, P=0.00003). Twelve studies (298 cases) reported a change in RR interval with an overall effect size of d=0.75 (C.I. 0.51-0.96, P<0.00001). Effect sizes for RR interval data were significantly heterogenous. Subgroup analysis revealed significantly smaller responses of RR interval to training in older subjects (P<0.1). Effect sizes for change in HF were homogenous, although a trend toward an attenuated response to training was exhibited in older subjects (P>0.10). Linear, quadratic, and cubic fits all revealed weak (P>0.05) relationships between effect sizes for change in HF and RR interval. DISCUSSION: Exercise training results in significant increases in RR interval and HF power. These changes are influenced by study population age. The smaller effect size for HF and weak relationship between HF and RR interval suggest factors additional to increased vagal modulation are responsible for training bradycardia.

PURPOSE: : To assess the validity and the reliability of short-term resting heart-rate variability (HRV) measures obtained using the Polar S810 heart-rate monitor and accompanying software. METHODS: : Measures of HRV were obtained from 5-min R to R wave (RR) interval data for 19 males and 14 females during 10 min of quiet rest on three separate occasions at 1-wk intervals using the Polar S810. Criterion measures of HRV were obtained simultaneously using the CardioPerfect (CP; Medical Graphics Corporation, St Paul, MN) 12-lead ECG module. Measures of validity of the Polar S810 were estimated by regression analysis, and measures of reliability of both devices were estimated by analysis of change scores. Measures of the SD of normal-to-normal intervals (SDNN), the root mean square of successive differences (RMSSD), and the low-frequency (LF) and the high-frequency (HF) spectral power and their ratio (LF/HF) were analyzed after log transformation, whereas mean RR and LF and HF in normalized units were analyzed without transformation. RESULTS: : There were marginal differences between the Polar and the CP mean measures of HRV, and the uncertainty in the differences was small. The Polar S810 demonstrated high correlations (0.85-0.99) with CP for all measures of HRV indicating good to near-perfect validity. Except for the low- and the high-frequency normalized units, Polar S810 did not add any substantial technical error to the within-subject variability in the repeated measurements of HRV. CONCLUSION: : HRV measures obtained with the Polar S810 and accompanying software have no appreciable bias or additional random error in comparison with criterion measures, but the measures are inherently unreliable over a 1-wk interval. Reliability of HRV from longer (e.g., 10 min) and/or consecutive 5-min RR recordings needs to be investigated with the Polar and criterion instruments.