Should heart rate variability be “corrected” for heart rate? Biological,quantitative, and interpretive considerations

Metrics of heart period variability are widely used in the behavioral and biomedical sciences, although somewhat confusingly labeled as heart rate variability (HRV). Despite their wide use, HRV metrics are usually analyzed and interpreted without reference to prevailing levels of cardiac chronotropic state (i.e., mean heart rate or mean heart period). This isolated treatment of HRV metrics is nontrivial. All HRV metrics routinely used in the literature exhibit a known and positive relationship with the mean duration of the interval between two beats (heart period): as the heart period increases, so does its variability. This raises the question of whether HRV metrics should be “corrected” for the mean heart period (or its inverse, the heart rate). Here, we outline biological, quantitative, and interpretive issues engendered by this question. We provide arguments that HRV is neither uniformly nor simply a surrogate for heart period. We also identify knowledge gaps that remain to be satisfactorily addressed with respect to assumptions underlying existing HRV correction approaches. In doing so, we aim to stimulate further progress toward the rigorous use and disciplined interpretation of HRV. We close with provisional guidance on HRV reporting that acknowledges the complex interplay between the mean and variability of the heart period.     

Estimation of maximal heart rate using the relationship between heart rate variability and exercise intensity in 40–67 years old men

Large interindividual variation in maximal heart rate (HR_max) may not be accounted for by age alone. In order to evaluate a novel method in the prediction of HR_max, this study examined the profile of HR variability (HRV) during exercise in 40–67 years old men (n = 74). R–R intervals were recorded during supine rest and during a graded exercise test by cycle ergometry until exhaustion. A third-degree polynomial function was fitted to the HRV data recorded during exercise to represent the HRV profile of each subject. The instantaneous beat-to-beat R–R interval variability (SD1), high (HF) and low frequency power decreased between all consecutive exercise intensities (P < 0.033). The relationship between HR_max and a variable illustrating the declining rate of HF (HR_HF50%) was stronger (r = 0.50, P < 0.001) than between HR_max and age (r = −0.36, P < 0.01). The regression analysis showed that a more accurate estimation of HR_max was attained when HRV was used in the equation in addition to age: HR_max = 160.633–0.603(age) + 0.441(HR_HF50%) (SEE = 9.8 bpm vs. 11.6 bpm in the equation based on age alone). The decline of HRV during incremental exercise seems to be useful for accurate estimation of maximal heart rate.     

Does motor activity during psychophysiological paradigms confound the quantification and interpretation of heart rate and heart rate variability measures in young children?

Respiratory sinus arrhythmia (RSA), heart period, and motor activity were monitored in preschoolers during a variety of tasks varying in required movement. The data analyses indicate: (1) that when activity increases during tasks, there are synchronous decreases in heart period and RSA; (2) that correlations between changes in RSA and heart period are related to activity only during exercise when there is a major demand for increased metabolic resources; and (3) that the covariation among the variables within each condition is low except during exercise. These findings suggest that the slight increases in motor activity (i.e., hand movements) often required in attention demanding psychophysiological protocols are not related to RSA and heart period responses. However, when tasks necessitate large increases in motor activity (e.g., exercise), the decreases in heart period and RSA are related to the change in motor activity.     

Does preprocessing change nonlinear measures of heart rate variability?

This work investigated if methods used to produce a uniformly sampled heart rate variability (HRV) time series significantly change the deterministic signature underlying the dynamics of such signals and some nonlinear measures of HRV. Two methods of preprocessing were used: the convolution of inverse interval function values with a rectangular window and the cubic polynomial interpolation. The HRV time series were obtained from 33 Wistar rats submitted to autonomic blockade protocols and from 17 healthy adults. The analysis of determinism was carried out by the method of surrogate data sets and nonlinear autoregressive moving average modelling and prediction. The scaling exponents , ₁ and ₂ derived from the detrended fluctuation analysis were calculated from raw HRV time series and respective preprocessed signals. It was shown that the technique of cubic interpolation of HRV time series did not significantly change any nonlinear characteristic studied in this work, while the method of convolution only affected the ₁ index. The results suggested that preprocessed time series may be used to study HRV in the field of nonlinear dynamics.     

Could heart rate play a role in pericardial inflammation?

Purpose and medical hypothesis: Rest is usually recommended in acute pericarditis, as it could help to lower heart rate (HR) and contribute to limit "mechanical inflammation". Whether HR on admission could be correlated and perhaps participate to inflammation has not been reported. Methods: Between March 2007 and February 2010, we conducted a retrospective study on all patients admitted to our center for acute pericarditis. Diagnosis criteria included two of the following ones: typical chest pain, friction rub, pericardial effusion on cardiac echography, or typical electrocardiogram (ECG) findings. Primary endpoint was biology: CRP on admission, on days 1, 2, 3, and especially peak. Results: We included 73 patients. Median age was 38years (interquartiles 28-51) and median hospitalization duration was 2.0days (1.5-3.0). Median heart rate was 88.0beats per minute (bpm) on admission (interquartiles 76.0-100.0) and 72.0 on discharge (65.0-80.0). Heart rate on admission was significantly correlated with CRP peak (p<0.001), independently of temperature on admission, hospitalization duration and age. Recurrences occurred within 1month in 32% of patients. Heart rate on hospital discharge was correlated with recurrence, independently of age. Conclusion: In acute pericarditis, heart rate on admission is independently correlated with CRP levels and heart rate on discharge seems to be independently correlated to recurrence. This could suggest a link between heart rate and pericardial inflammation.     

Effect of β-adrenoblockers on heart rate variability in awake and narcotized rats

In narcotized rats, β-adrenoceptor blockers changed all parameters of heart rate variability indicating up-regulation of parasympathetic activity and down-regulation of the sympathetic one. In immobilized awake rats, the temporal and geometrical parameters varied similarly, while spectral analysis indicated possible activation of other neurohumoral mechanisms against the background of β-blocker treatment. __________ Translated from Byulleten’ Eksperimental’noi Biologii i Meditsiny, Vol. 144, No. 10, pp. 372–264, October, 2007     

Automated detection of perinatal hypoxia using time–frequency-based heart rate variability features

Perinatal hypoxia is a cause of cerebral injury in foetuses and neonates. Detection of foetal hypoxia during labour based on the pattern recognition of heart rate signals suffers from high observer variability and low specificity. We describe a new automated hypoxia detection method using time–frequency analysis of heart rate variability (HRV) signals. This approach uses features extracted from the instantaneous frequency and instantaneous amplitude of HRV signal components as well as features based on matrix decomposition of the signals’ time–frequency distributions using singular value decomposition and non-negative matrix factorization. The classification between hypoxia and non-hypoxia data is performed using a support vector machine classifier. The proposed method is tested on a dataset obtained from a newborn piglet model with a controlled hypoxic insult. The chosen HRV features show strong performance compared to conventional spectral features and other existing methods of hypoxia detection with a sensitivity 93.3 %, specificity 98.3 % and accuracy 95.8 %. The high predictive value of this approach to detecting hypoxia is a substantial step towards developing a more accurate and reliable hypoxia detection method for use in human foetal monitoring.     

Chronic administration of interferon-α decreases blood pressure and heart rate in rats

We studied the effects of interferon-α on rat cardiovascular system. Intravenous administration of intron-A in a dose of 100,000 IU/kg for 3 days led to a permanent and statistically significant decrease in blood pressure (on days 2 and 3) and reduction in heart rate. These effects were not associated with changes in baroreflex regulation of the cardiovascular system. Translated fromByulleten' Eksperimental'noi Biologii i Meditsiny, Vol. 129, No. 5, pp. 491–495, May, 2000     

Quantitative Poincaré plot analysis of heart rate variability: effect of endurance training

The aim of the study was to evaluate the effectiveness of the Poincaré plot analysis of heart rate variability (HRV) in observing endurance training-induced changes. Four 10-min manoeuvres were performed (supine lying, standing, steady state exercising and subsequent recovery) by eight control subjects before and after a short-term endurance training and by eight subjects trained for at least 3 years. HRV was assessed by traditional time- and frequency-domain indexes, in parallel with the Poincaré plot analysis. In the latter each R-R interval is plotted as a function of the previous one, and the standard deviations of the instantaneous and long-term R-R interval variability are calculated. In our subjects, the Poincaré scatter grams became gradually narrower from supine to exercising, with progressive parasympathetic withdrawal. Short- and long-term endurance training led to higher aerobic power (p<0.05) and ventilatory threshold shifted towards higher power output (p<0.05). All HRV evaluation methods showed that HRV values were higher after training both during supine lying and standing (p<0.05). The Poincaré scatter grams were wider in the trained state. Standard deviations of the Poincaré plot were significantly correlated with the main parameters of the time- and frequency-domain analyses, especially concerning the parasympathetic indicators. These results suggested that Poincaré plot parameters as well as the "width" of the scatter gram could be considered as surrogates of time- and frequency-domain analysis to assess training-induced changes in HRV.     

Estimation of %V˙O2 reserve from heart rate during arm exercise and running

The purpose of the present investigation was to examine the relationship between the percent heart rate reserve (%HRR) in arm exercise and the corresponding percent oxygen uptake (V˙O₂) reserve, and to compare this relationship to that occurring in running. Fourteen male physical education students took part in the study. Each subject performed a maximal running exercise test and a maximal arm cycling test. The subjects also performed three submaximal exercise bouts (in both exercise modes) at 30%, 60% and 80% of their HRR. The subjects were monitored for their heart rate (HR) at rest, maximal HR (HR_max), HR at submaximal work loads, maximal V˙O₂ (V˙O_2max), V˙O₂ at rest and V˙O₂ at submaximal loads. For each subject, load and exercise mode, %HRR and %V˙O₂ reserve were calculated (from HR_max and V˙O_2max as measured during running and arm cycling) and the relationship between the two was evaluated. The main finding of the present investigation is that the prediction of %V˙O₂ reserve in arm cycling from %HRR is grossly overestimated when calculated from HR_max and V˙O_2max measured during running. The prediction is better but still overestimated when calculated from HR_max and V˙O_2max measured during arm cycling. The findings indicate a better prediction of %V˙O₂ reserve from %HRR for running than for arm exercise. These findings should be taken into consideration when prescribing the target HR for arm training. Accepted: 24 July 2000     

Sub-maximal and maximal Yo�CYo intermittent endurance test level 2: heart rate response, reproducibility and application to elite soccer

The aims of this study were to (1) determine the reproducibility of sub-maximal and maximal versions of the Yo-Yo intermittent endurance test level 2 (Yo-Yo IE2 test), (2) assess the relationship between the Yo-Yo IE2 test and match performance and (3) quantify the sensitivity of the Yo-Yo IE2 test to detect test-retest changes and discriminate between performance for different playing standards and positions in elite soccer. Elite (n = 148) and sub-elite male (n = 14) soccer players carried out the Yo-Yo IE2 test on several occasions over consecutive seasons. Test-retest coefficient of variation (CV) in Yo-Yo IE2 test performance and heart rate after 6 min were 3.9% (n = 37) and 1.4% (n = 32), respectively. Elite male senior and youth U19 players Yo-Yo IE2 performances were better (P < 0.01) than elite youth U16s and sub-elite players (2,603 ± 451 and 2,534 ± 549 vs. 1,855 ± 535 vs. 1,749 ± 382 m). The intra- and inter-season CV for Yo-Yo IE2 test performance were 4.2 and 5.6%, respectively. A correlation was observed (P < 0.05) between Yo-Yo IE2 test performance and the total (r = 0.74) and high-intensity (r = 0.58) running distance covered in a match. A correlation was also evident (P < 0.01) between Yo-Yo IE2 test heart rate after 6 min expressed in percentage of maximal heart rate and the peak values for high-intensity running performed by midfielders in 5-min (r = -0.71), 15-min (r = -0.75) and 45-min periods (r = -0.77). The present data demonstrate that the Yo-Yo IE2 test is reproducible and can be used to determine the capacity of elite soccer players to perform intense intermittent exercise. Furthermore, the Yo-Yo IE2 test was shown to be a sensitive tool that not only relates to match performance but can also differentiate between intermittent exercise performance of players in various standards, stages of the season and playing positions.     

Seroprotection rate, mean fold increase, seroconversion rate: which parameter adequately expresses seroresponse to influenza vaccination?

Serological parameters intend to describe antibody response to influenza vaccine in a population. However, there is uncertainty about the mathematical appropriateness and the biological or clinical meaning of conventionally used parameters. Theoretical considerations and exploration of a data-set of 16 studies with an inactivated (subunit) influenza vaccine involving 1176 adult subjects suggest the following conclusions. In a population seronegative before vaccination, the post-vaccination geometric mean titre (post-GMT) is a meaningful immunological parameter adequately expressing antibody response after vaccination. The related protection rate (PR) is a good surrogate parameter for protection provided by a given vaccine, thus relevant to public health. However, in a population partially seropositive before vaccination (due to previous exposition to influenza antigens), the same parameters may, under certain conditions, seriously overestimate the antibody response, as they do not account for the pre-vaccination state. Conventional attempts to address pre-vaccination antibody are associated with either loss of information (exclusion of seropositive subjects) or incomplete control of pre-vaccination state (mean fold increase (MFI), response rate (RR)). Although not devoid of theoretical limitations (heteroscedasticity), correction of post-GMT and PR by linear regression appears to provide better estimates of antibody response and vaccine immunogenicity.     

Correlation among Poincare´plot indexes and time and frequency domain measures of heart rate variability

Primary objective: In order to corroborate the reported performance advantage of Poincar´eplot indexes as autonomic activity markers, the correlation among these indexes and those computed from the time and frequency domains were obtained. Methods and procedures: Starting from the RR series derived from the ECG of 21 healthy volunteers during five manoeuvres, longitudinal (L), transverse (T) axis, and autocorrelation (r) from Poincar´eplots, rMSSD and standard deviation (SD) in the temporal domain, and frequency domain indexes were computed. Main outcomes and results: Poincar´e plot indexes were correlated in a better way with the time indexes, rather than the spectral measures. A strong correlation (0.997) between L and SD was observed, while an underlying mathematical relationship was established for T vs rMSSD. Conclusions: Poincaré plot indexes may be considered as equivalent or surrogates of the temporal ones, and they do not have a better performance as autonomic markers.     

Recovery of heart rate variability after treadmill exercise analyzed by lagged Poincaré plot and spectral characteristics

The aim of this study was to analyze the recovery of heart rate variability (HRV) after treadmill exercise and to investigate the autonomic nervous system response after exercise. Frequency domain indices, i.e., LF(ms²), HF(ms²), LF(n.u.), HF(n.u.) and LF/HF, and lagged Poincaré plot width (SD1 _m ) and length (SD2 _m ) were introduced for comparison between the baseline period (Pre-E) before treadmill running and two periods after treadmill running (Post-E1 and Post-E2). The correlations between lagged Poincaré plot indices and frequency domain indices were applied to reveal the long-range correlation between linear and nonlinear indices during the recovery of HRV. The results suggested entirely attenuated autonomic nervous activity to the heart following the treadmill exercise. After the treadmill running, the sympathetic nerves achieved dominance and the parasympathetic activity was suppressed, which lasted for more than 4 min. The correlation coefficients between lagged Poincaré plot indices and spectral power indices could separate not only Pre-E and two sessions after the treadmill running, but also the two sessions in recovery periods, i.e., Post-E1 and Post-E2. Lagged Poincaré plot as an innovative nonlinear method showed a better performance over linear frequency domain analysis and conventional nonlinear Poincaré plot.