`doi:10.1093/qjmed/hci018
`
`Review
`
`Downloaded from https://academic.oup.com/qjmed/article/98/2/87/1570548 by guest on 19 April 2021
`
`Heart rate variability measurements and the prediction of
`ventricular arrhythmias
`
`M.J. REED, C.E. ROBERTSON and P.S. ADDISON1
`
`From the Department of Emergency Medicine, Royal Infirmary of Edinburgh, Edinburgh, and
`1Cardiodigital Ltd, Elvingston Science Centre, Gladsmuir, UK
`
`Introduction
`Heart rate variability (HRV) is the temporal variation
`between sequences of consecutive heartbeats. On a
`standard electrocardiogram (ECG),
`the maximum
`upwards deflection of a normal QRS complex is at
`the peak of the R wave (Figure 1), and the duration
`between two adjacent R wave peaks is termed the
`R-R interval. The ECG signal requires editing before
`HRV analysis can be performed, a process requiring
`the removal of all non-sinus-node-originating beats.
`The resulting period between adjacent QRS com-
`plexes resulting from sinus node depolarizations is
`interval.1 HRV
`termed the N-N (normal-normal)
`is the measurement of the variability of the N-N
`intervals.
`Although counter-intuitive, it is possible that HRV
`confers a survival advantage. Any system exhibiting
`intrinsic variability is primed to respond rapidly
`and appropriately to demands placed upon it.
`HRV is a measure of the balance between sympa-
`thetic mediators of heart rate (HR) (i.e. the effect
`of epinephrine and norepinephrine, released from
`sympathetic nerve fibres, acting on the sino-atrial
`and atrio-ventricular nodes), which increase the rate
`of cardiac contraction and facilitate conduction
`at the atrio-ventricular node, and parasympathetic
`mediators of HR (i.e. the influence of acetylcholine,
`released by the parasympathetic nerve fibres, acting
`on the sino-atrial and atrio-ventricular nodes), lead-
`ing to a decrease in the HR and a slowing of con-
`duction at the atrio-ventricular node. Sympathetic
`
`mediators appear to exert their influence over longer
`time periods and are reflected in the low frequency
`power (LFP) of the HRV spectrum (between 0.04 Hz
`and 0.15 Hz.2,3 Vagal mediators exert their influ-
`ence more quickly on the heart, and principally
`affect the high frequency power (HFP) of the HRV
`spectrum (between 0.15 Hz and 0.4 Hz).4 Thus,
`at any point in time, the LFP:HFP ratio is a proxy
`for the sympatho-vagal balance.
`Physiological and pathological process may
`influence N-N interval variability. Under normal
`conditions, the balance between sympathetic and
`parasympathetic activity favours the latter. Physio-
`logical
`influences may modulate central and
`peripheral receptor (i.e. carotid sinus) activity. This
`is demonstrated in the slowing of HR with expira-
`tion, and a quickening with inspiration (respiratory
`sinus arrhythmia). These effects are apparent on
`the HFP spectrum. Circadian alterations in HRV are
`present in normal subjects, with higher LFP in the
`daytime and higher HFP at night.5,6 Exercise,
`standing and stress in human subjects, and hypoten-
`sion, and coronary or cerebral ischaemia in dogs,
`increases sympathetic drive and LFP. Conversely,
`cold stimulation of the face increases parasympa-
`thetic drive and increases HFP.5–7
`In normal subjects, a variable heart rate is the
`normal physiological state. It has been suggested
`that the healthy heart has a long range ‘memory’
`which prevents it from developing extremes of pace,
`
`Address correspondence to Dr M.J. Reed, Emergency Department, Royal Infirmary of Edinburgh, 51 Little
`France Crescent, Edinburgh EH16 4SA. e-mail: mattreed1@hotmail.com
`QJM vol. 98 no. 2 ! Association of Physicians 2005; all rights reserved.
`
`APPLE 1017
`
`1
`
`
`
`88
`
`M.J. Reed et al.
`
`R-R interval
`
`R wave peak 1
`
`R wave peak 2
`
`R wave
`
`Downloaded from https://academic.oup.com/qjmed/article/98/2/87/1570548 by guest on 19 April 2021
`
`P wave
`
`Q wave
`
`S wave
`
`T wave
`
`Figure 1. The normal electrocardiogram with component waves labelled.
`
`this facility erodes as age or disease
`and that
`develops.8 A loss of variability is associated with
`an increased mortality in patients post myocardial
`infarction.9 In animal studies, an increase in sym-
`pathetic activity can provoke ventricular
`tachy-
`arrythmias (VTAs)10,11 and lower the ventricular
`fibrillation threshold.12 This effect
`is exacerbated
`ischaemia.13,14 Conver-
`by coexistent myocardial
`sely, vagal activity seems to provide a protective
`effect against the development of VTAs.15,16 With
`this exerting the greatest effect on the heart rate
`in normal conditions and predominantly effecting
`HFP, a heart
`rate with much variability is the
`optimal state most likely to prevent the development
`of fatal VTAs.
`Drug therapy may alter HRV; beta-blocker
`therapy has been shown to have a favourable
`effect on HRV17,18 in patients with heart
`failure.
`However, changes in HR dynamics observed before
`VTAs in patients taking anti-arrhythmic drugs were
`independent of the drug regimen.19
`
`The history of HRV
`Heart rate variability (HRV) was first used clinically
`in 1965 when Hon and Lee20 noted that
`fetal
`distress was accompanied by changes in beat-
`to-beat variation of
`the fetal heart, even before
`
`In the
`there was detectable change in the HR.
`1970s, Ewing et al. used short-term HRV measure-
`ments as a marker of diabetic autonomic neuro-
`pathy.21 In 1977, Wolf et al.9 showed that patients
`with reduced HRV after a myocardial infarction had
`an increased mortality, and this was confirmed by
`studies showing that HRV is an accurate predictor
`infarction (MI).22–24
`of mortality post myocardial
`HRV falls within 2 to 3 days after MI, begins to
`recover within a few weeks, and is maximally but
`not fully recovered by 6 to 12 months.25 Patients
`with persisting low HRV have mortality almost three
`times greater than those with a normal HRV.23
`Over the last decade, alterations in HRV have
`been found in patients with many cardiovascular
`conditions. Patients with hypertension exhibit
`increased LFP and reduced circadian patterns,26
`congestive heart failure is associated with reduced
`vagal but preserved sympathetic activity,27 and
`patients with denervated transplanted hearts show
`a 90% reduced HRV.28 HR and ULFP may be good
`prognostic indicators for mortality, progression to
`surgery and the development of atrial
`fibrillation
`in patients with mitral regurgitation,29 and patients
`with mitral valve prolapse show reduced HFP.30
`Radio frequency
`ablation of
`supraventricular
`arrhythmia pathways leads to an increase in HR,
`reduced HRV and vagal tone measurements,31 and
`patients with cardiomyopathies exhibit
`reduced
`
`2
`
`
`
`Downloaded from https://academic.oup.com/qjmed/article/98/2/87/1570548 by guest on 19 April 2021
`
`Heart rate variability
`
`89
`
`tone.32 HRV has also been extensively
`vagal
`investigated as a tool to predict the risk of sudden
`cardiac death. Low HRV is an independent risk
`factor for the development of later cardiac arrest
`in survivors of cardiac arrest.33 Both reduced HF
`power and reduced LF power are independent
`predictors of later sudden death following survival
`from cardiac arrest. Reduction in HF power appears
`superior at risk-stratifying patients.34 To date, most
`studies have concentrated on identifying HRV
`characteristics to predict
`the longer-term risk of
`developing
`fatal
`ventricular
`tachyarrhythmias
`(VTAs). Much less research has focussed on the
`changes that occur in HRV in the period immedi-
`ately prior to the development of VTAs.
`
`Problems with measuring HRV
`To detect HRV changes over a period of hours or
`days requires a large volume of ECG data to be
`
`collected and analysed. This has traditionally been
`done with Holter devices that record the ECG in out-
`patients over periods from 24 h up to several weeks.
`Data can also be collected from patients who are
`monitored in hospital (Figure 2). Data capture on
`dynamic changes in HRV in the period prior to
`arrhythmias or ischaemic events is harder to attain,
`due to the relative infrequency of such events. In the
`laboratory environment, studies of patients with
`exercise or electrically-induced VT are possible,
`and implanted cardio-defibrillator devices (ICDs),
`are able to store information prior to an episode
`of ventricular fibrillation (VF) or ventricular tachy-
`cardia (VT).
`Signal quality and elimination of background
`‘noise’ is important when analysing HRV. Interpre-
`tation of HRV is extremely difficult in patients who
`are not
`in sinus rhythm (e.g. atrial
`fibrillation),
`or those with an extremely irregular HR (Figure 3)
`or multiple ectopic (VE) beats. Most HRV studies
`
`Figure 2. A section of an ECG waveform and oxygen saturation waveform obtained from an Emergency Department
`monitor. The ECG shows the onset of VF, with the accompanying loss of cardiac output demonstrated by the loss of the
`oxygen saturation waveform.
`
`3
`
`
`
`90
`
`M.J. Reed et al.
`
`Downloaded from https://academic.oup.com/qjmed/article/98/2/87/1570548 by guest on 19 April 2021
`
`Figure 3. A section of an ECG signal obtained from a Coronary Care Unit, showing an irregular rhythm that makes HRV
`interpretation extremely difficult. The ECG demonstrates multiple premature atrial ectopic beats. The locations of the R wave
`peaks as detected by our R wave peak detection algorithm are represented by the vertical lines.
`
`in sinus rhythm,
`exclude patients who are not
`but there is controversy over the issue of multiple
`VEs. Some authors advocate excluding signals that
`contain more than 10 ectopic beats per hour.35
`Others accept patients where ectopic beats com-
`prise up to 5%,19 8%,36 10%37,38 or even 15%39 of
`all R-R intervals.
`If a signal with ectopic beats is to be analysed,
`most authors advise removing the ectopic beats and
`correcting for them by adjusting the position of the R
`wave peak and placing a beat midway between the
`two adjacent beats. Editing algorithms are available
`that remove all R-R intervals that differ by a certain
`percentage from the preceding normal one. The
`level at which this editorial exclusion is performed
`(usually between 20–30% of
`the preceding R-R
`interval40) obviously reflects a balance between
`editing genuine data and missing ectopic beats. The
`frequency of ectopic beats is also of interest, since
`their frequency can increase prior to an arrhythmic
`event.41 Since the vast majority of patients at risk
`of sudden cardiac death have underlying cardiac
`abnormalities and are more likely to have irregular
`rhythms such as atrial fibrillation and multiple VEs,
`
`these signals is problematic. The
`the analysis of
`process of signal editing prior to analysis is complex,
`and is poorly performed by conventional algorithms.
`Some authors suggest that manual filtering, although
`time consuming, is more accurate.
`A final problem when measuring HRV is of
`accurately locating the successive R-wave peaks
`on the ECG. This requires a robust R-wave detector
`algorithm. The more accurate the R-wave detector,
`the less error in the analysed HRV spectrum. A
`completely missed R wave will cause greater error
`than a slightly miscorrected R wave, and this error is
`reflected more in the HFP than in the LFP of the
`HRV spectrum. This is due to the greater influence
`of a single N-N interval on short-term variability
`measures (HFP), compared to longer-term variability
`measures (LFP) where a single N-N interval effect
`becomes smoothed out.
`
`Measuring HRV
`frequency
`HRV can be measured in time or
`domains. Time domain methods are the simplest
`
`4
`
`
`
`Heart rate variability
`
`91
`
`Downloaded from https://academic.oup.com/qjmed/article/98/2/87/1570548 by guest on 19 April 2021
`
`LF HR changes can also be quantified (i.e. para-
`sympathetic and sympathetic influences) and the
`interactions between them, preceding an event
`such as a VTA, quantified.
`Spectral methods have been used to analyse HRV
`for 40 years.42 These measure how the variance
`(or power) of the ECG signal changes as a function
`of frequency. Non-parametric methods of spectral
`analysis employing the Fast Fourier Transform (FFT)
`algorithm are commonly used.1 This technique
`involves splitting the ECG waveform into small
`subunits (usually from 2 to 5 min long for
`the
`measurement of HFP, LFP and VLFP, but can be
`up to 24 h when analysing ULF components).
`These signal segments are then ‘transformed’ from
`a temporal signal
`into a spectral
`representation
`whereby the ECG signal is reinterpreted as the sum
`of multiple simpler (sinusoidal) waves of a given
`amplitude and frequency (Figure 4). The amplitudes
`of the component waves are then plotted to give a
`power spectrum by plotting power (the square of
`amplitude in volts) versus frequency.
`VLFP, LFP and HFP components of the HRV can
`be calculated for recordings of 5 min or greater.
`For longer recordings of 24 h, ULFP can also be
`calculated, and reflects influences that occur on the
`heart rate over periods of days.
`More recently, new time-frequency signal analy-
`sis methods have been used in the analysis of HRV.
`As their name suggests,
`these offer simultaneous
`interpretation of the signal in both time and fre-
`quency, which allows local, transient or intermittent
`components to be elucidated.
`(These are often
`obscured due to the averaging inherent within
`spectral-only methods, i.e. the FFT.) Several time-
`frequency methods
`are
`currently
`available,
`including the short time Fourier transform (STFT),
`Wigner-Ville transform (WVT), Choi-Williams dis-
`tribution (CWD) and the continuous wavelet trans-
`form (CWT). Of these, the CWT has become the
`most favoured tool by researchers, as it does not
`contain the cross-terms inherent in the WVT and
`CWD methods, and provides frequency-dependent
`windowing, which allows for arbitrarily high resolu-
`tion of
`the high frequency signal components
`(unlike the STFT)43 (Figure 5). Accordingly, high
`frequency components (the ‘fine detail’ of the ECG
`signal) are not lost to analysis. CWT has recently
`shown an increase in LFP:HFP ratio prior to the
`onset of non-sustained VTA.44 Other recent studies
`involving the wavelet analysis of HRV have
`allowed the detection of patterns directly associated
`with changes in myocardial perfusion,45 and the
`association between autonomic tone and sponta-
`neous coronary spasm in patients with variant
`angina.46
`
`power
`
`
`
`frequency
`
`Figure 4. Any complex wave can be broken down into
`sine waves that when added together give the original
`complex wave. The figure shows the first four of the sine
`waves (middle) that when combined will make up the
`approximation to the square wave (top). The amplitudes of
`each sine wave are then converted to power and plotted
`against the frequency of the sine wave to give the power
`spectrum (bottom). To form the square wave, an infinite
`number of sine waves of decreasing amplitude are
`required. While this is a simplified example, it demon-
`strates the process of Fast Fourier Transform.
`
`to perform. Each N (or R) point is determined in
`the ECG trace and variables such as mean HR and
`longest and shortest N-N intervals calculated.
`More complex calculations such as SDNN (standard
`deviation of
`the N-N intervals, representing the
`overall HRV) and NN50 (the number of adjacent
`N-N intervals that differ by more than 50 ms) can be
`performed using this data. Variables can also be
`derived that estimate the short- and long-term
`components of HRV (i.e. RMSDD, the square root
`of the mean squared differences between adjacent
`N-N intervals gives an estimate of short-term HRV,
`and SDANN, the standard deviation of the average
`N-N interval over periods of about 5 min, gives an
`estimate of long-term HRV). The calculation of all
`these variables enables the temporal variability of
`the HR to be quantified. The contribution of the
`various factors that manifest themselves in HF and
`
`5
`
`
`
`92
`
`M.J. Reed et al.
`
`Downloaded from https://academic.oup.com/qjmed/article/98/2/87/1570548 by guest on 19 April 2021
`
`0.003 Hz
`
`0.04 Hz
`
`0.15 Hz
`
`0.4 Hz
`
`Figure 5. Consecutive sinus beat intervals (the HRV signal) from a healthy subject (top) together with its associated wavelet
`transform plot (below). The boundaries of the HF (0.15–0.4 Hz), LF (0.04–0.15 Hz) and VLF (0.003–0.04 Hz) regions are
`plotted across the transform surface.
`
`HRV and the onset of VTAs
`VTAs have a circadian rhythm,47 with increased
`frequency during the early morning (7–11 am) and
`early evening (6–7 pm). This is consistent with
`Peckova’s findings of a circadian distribution of
`cardiac arrests, with a low incidence at night, and
`peaks between 0800–1100 h and 1600–1900 h.48
`The evening peak may be attributable to VF,48 while
`the morning peak may be attributable to patients
`in non-VF and non-VT rhythms. These findings have
`been replicated in other studies.49,50
`Diurnal variation is also found in HRV. Higher
`LFP occurs in the daytime, and higher HFP at
`night.5,6 Markers of vagal activity (RMSSD and HFP)
`display circadian variability and are maximal during
`sleep. An inverse circadian rhythm is seen in
`patients with a morning VTA peak.51,52
`The characteristics of HRV immediately prior to
`the onset of VTAs are, however, unclear. Some
`studies report significant changes in HRV in the period
`immediately preceding a VTA, and HR increasing
`prior to an episode of VTA is common.36,41,53–56
`Vybrial et al., however, found no consistent changes
`in HRV indices (or HR) in 24 patients wearing Holter
`devices who developed VF.57 Some studies have
`reported an increase in HR, but also no change in
`HRV spectra characteristics prior to VTAs.53
`
`found a significant reduction in
`Huikuri et al.
`HR, SDANN, HFP, LFP and VLFP in post-MI patients
`who developed VT or cardiac arrest compared to
`normal controls and post-MI patients not suffering
`arrhythmias.58 These changes occurred in the 1-h
`period prior to the onset of the VTA, and were more
`pronounced in patients developing sustained VT
`than those with non-sustained VT. These findings
`were confirmed by Shusterman et al., who noted
`a rise in HR and a fall in LFP, and LFP:HFP ratio
`before the onset of VT.59 Pruvot et al. found an
`increase in HR and a significant reduction in HRV
`prior to the onset of a VTA in post-MI patients.19
`Other studies have shown a rise in VLF power and
`a decline in HF power,60 a decline in HF power
`but no VLF power changes,19 and a rise in LF:HF
`ratio.36,44
`These results strongly suggest an alteration in the
`interaction between the sympathetic and parasym-
`pathetic nervous system prior to the onset of VTAs.
`The effect on HR and HRV variables is likely to be
`heterogeneous and affected by individual patient
`characteristics, which may explain the conflicting
`evidence in the literature. This is compounded by
`the small numbers of patients studied, their differing
`drug treatment, underlying cardiac pathology,
`pre-existing medical problems and methods of
`recording and analysing the ECG.
`
`6
`
`
`
`Downloaded from https://academic.oup.com/qjmed/article/98/2/87/1570548 by guest on 19 April 2021
`
`Heart rate variability
`
`93
`
`Two recent studies have addressed these diffi-
`culties. Meyerfeldt et al. found more reduced HRV
`periods before VT occurred in patients with ICDs.37
`They divided the VT events into those with a cycle
`length 5270 ms (‘fast VT’) and those with a cycle
`length 4270 ms (‘slow VT’). Episodes of slow VT
`occurred during periods of increased HR, whereas
`fast VT occurred during periods of decreased HR.
`Shusterman et al. have suggested that baseline
`levels of HRV may determine HRV changes in
`response to short-term autonomic perturbations
`prior to VTAs.54 Groups with higher initial
`total
`power, LFP and LFP: HFP ratio exhibited a fall in
`LFP in the 2 h prior to the VT episode. This group
`had a fall in HFP in the 15 min prior to VT onset. A
`group with lower
`initial
`total power, LFP and
`LFP:HFP HRV parameters, showed an increase or
`no change in LFP in the 2 h prior to VT, and no
`change in HFP immediately preceding VT. This
`suggests that merely a change in HRV, rather than
`the magnitude or nature of the change, facilitates the
`development of VTAs.
`
`Conclusions and further work
`Initially, the focus of HRV investigation was its use
`in the prediction of long-term survival in patients
`who had suffered myocardial
`infarction, or had
`valvular or congestive heart disease. More recently,
`work has concentrated on attempts to predict the
`timing of onset of fatal VTAs. Given the poor results
`currently achieved in out-of-hospital cardiac arrest,
`efforts directed towards the identification of VTAs
`are crucial.61 The area of HRV behaviour before the
`onset of life-threatening VTAs offers exciting possi-
`bilities. Newer
`improved analytical
`techniques
`such as wavelet analysis, together with improved
`processing power, have simplified and speeded up
`signal acquisition and analysis.
`While the prognostic value of HRV post-myocar-
`dial infarction is well established, evidence of value
`in VTAs and sudden death is less clear. Although
`it is not yet possible to predict the onset of ventricular
`arrhythmias using HRV techniques, there is now a
`better understanding of HRV behaviour before these
`events. Prediction of the exact time of onset of VTA
`occurrence remains a distant and probably unreal-
`istic goal using HRV analysis alone. It is, however,
`likely that patients with a high risk of developing
`VTAs may be identified with some accuracy,
`enabling prophylactic approaches to be made.
`trials such as AVID,62 MUSTT,63 and
`Results of
`MADIT I and II64 have shown that ICDs improve
`survival in patients surviving an episode of cardiac
`arrest, and in those with coronary heart disease and
`
`advanced left ventricular dysfunction. HRV inves-
`tigation,
`together with other ECG markers (e.g.
`T wave alternans), may further risk-stratify these
`patients,65,66 enabling the targeting of groups who
`may benefit from such devices.
`Recent awareness of the heterogeneity of HRV
`behaviour has emphasized the importance of
`acquiring large ECG databases from patients at risk
`of developing life-threatening VTAs.67 The increas-
`ing ease with which data can be captured from
`monitor-defibibrillator devices used in and out of
`hospital should facilitate this process and enable
`future algorithm development.
`
`References
`1. Heart rate variability. Standards of measurement, physiolog-
`ical interpretation, and clinical use. Task for of the European
`Society of Cardiology and the North American Society of
`Pacing and Electrophysiology. Eur Heart J 1996; 17:354–81.
`
`2. Akselrod S, Gordon D, Ubel FA, et al. Power spectrum
`analysis of heart rate fluctuation: a quantitive probe of beat-
`to-beat cardiovascular control. Science 1981; 213:220–2.
`
`3. Pomeranz B, Macauley RJ, Caudill MA, et al. Assessment of
`autonomic function in humans by heart rate spectral analysis.
`Am J Physiol 1985; 248:H151–3.
`
`4. Task force of the European Society of Cardiology and the
`North American Society of Pacing and Electrophysiology.
`Heart rate variability: standards of measurement, physiologi-
`cal
`interpretation, and clinical use. Circulation 1996;
`93:1043–65.
`
`5. Malliani A, Pagani M, Lombardi F, Cerruti S. Cardiovascular
`neural
`regulation explored in the frequency domain.
`Circulation 1991; 84:1482–92.
`
`6. Furlan R, Guzetti S, Crivellaro W, et al. Continuous 24-hour
`assessment of the neural regulation of systemic arterial pres-
`sure and RR variabilities in ambulant subjects. Circulation
`1990; 81:537–47.
`
`7. Kamath MV, Fallen EL. Power spectral analysis of heart rate
`variability: a non-invasive signature of cardiac autonomic
`function. Crit Revs Biomed Eng 1993; 21:245–311.
`
`8. Buchanan M. Fascinating rhythm. New Scientist 1998; 3
`January:20–25.
`
`9. Wolf MM, Varigos GA, Hunt D, et al. Sinus arrhythmia in
`acute myocardial infarction. Med J Australia 1978; 2:52–3.
`
`10. Armour JA, Hageman GR, Randall WC. Arrhythmias induced
`by local cardiac nerve stimulation. Am J Physiol 1972;
`223:1068–75.
`
`11. Han J, Garcia de Jalon P, Moe GK. Adrenergic effects on
`ventricular vulnerability. Circ Res 1964; 14:516–24.
`
`12. Hockman CH, Mauck HP Jr, Hoff EC. ECG changes resulting
`from cerebral
`stimulation.
`II. Spectrum of ventricular
`arrhythmias of
`sympathetic origin. Am Heart
`J 1966;
`71:695–700.
`
`13. Maling HM, Moran NC. Ventricular arrhythmias induced by
`sympathomimetic amines in unanaesthetized dogs following
`coronary occlusion. Circ Res 1957; 5:409–13.
`
`7
`
`
`
`Downloaded from https://academic.oup.com/qjmed/article/98/2/87/1570548 by guest on 19 April 2021
`
`94
`
`M.J. Reed et al.
`
`14. Kliks BR, Burgess MJ, Abildskov JA. Influence of sympathetic
`tone on ventricular fibrillation threshold during experimental
`coronary artery occlusion. Am J Cardiol 1975; 36:45–9.
`
`15. Kent KM, Smith ER, Redwood DR, et al. Electrical stability
`of acutely ischemic myocardium. Influences of heart rate and
`vagal stimulation. Circulation 1973; 47:291–8.
`
`16. Myers RW, Pearlman AS, Hyman RM, et al. Beneficial effects
`of vagal stimulation and bradycardia during experimental
`acute myocardial infraction. Circulation 1974; 49:943.
`
`17. Lin JL, Chan HL, Du CC, et al. Long-term beta-blocker
`therapy improves autonomic nervous regulation in advanced
`congestive heart failure: a longitudinal heart rate variability
`study. Am J Cardiol 1999; 137:658–65.
`
`18. Pousset F, Copie X, Lechat P, et al. Effects of bisoprolol on
`heart rate variability in heart failure. Am J Cardiol 1996;
`77:612–17.
`
`19. Pruvot E, Thonet G, Vesin JM, et al. Heart rate dynamics at
`the onset of ventricular tachyarrhythmias as retrieved from
`implantable
`cardioverter-defibrillators
`in patients with
`coronary artery disease. Circulation 2000; 101:2398–404.
`
`20. Hon EH, Lee ST. Electronic evaluations of the fetal heart rate
`patterns preceding fetal death, further observations. Am J
`Obstet Gynae 1965; 87:814–26.
`
`21. Ewing DJ, Martyn CN, Young RJ, Clarke BF. The value of
`cardiovascular autonomic function tests: 10 years experience
`in diabetes. Diabetes Care 1985; 8:491–8.
`
`22. Bigger JT Jr, Fleiss JL, Steinman RC, Rolnitzky LM. Kleiger RE,
`Rottman JN. Frequency domain measures of heart period
`variability and mortality after myocardial infarction. Circula-
`tion 1993; 85:164–71.
`
`23. Kleiger RE, Miller JP, Bigger JT Jr, Moss AJ. Decreased heart
`rate variability and its association with increased mortality
`after acute myocardial
`infarction. Am J Cardiol 1987;
`59:256–62.
`
`24. Malik M, Farrell T, Cripps T, Camm AJ. Heart rate variability
`in relation to prognosis after myocardial infarction: selection
`of optimal processing techniques. Eur Heart
`J 1989; 10:
`1060–74.
`
`25. Bigger JT, Fleiss JL, Rolnitzky LM, et al. Time course of
`recovery of heart period variability after myocardial infarc-
`tion. J Am Coll Cardiol 1987; 59:256–62.
`
`26. Guzzetti S, Dassi S, Pecis M, et al. Altered pattern of
`circadian neural control of heart period in mild hypertension.
`J Hypertension 1991; 9:831–8.
`
`27. Casolo G, Balli E, Taddei T, et al. Decreased spontaneous
`heart rate variability on congestive cardiac failure. Am J
`Cardiol 1989; 64:1162–7.
`
`28. Sands KE, Appel ML, Lelly LS, et al. Power spectrum analysis
`of heart
`rate variability in human cardiac transplant
`recipients. Circulation 1989; 79:76–82.
`
`29. Stein KM, Bores JS, Hoxchreites C, et al. Prognostic value
`and physiological correlates of heart
`rate variability in
`chronic severe mitral
`regurgitation. Circulation 1993;
`88:127–35.
`
`30. Marangoni S, Scalvini S, Mai R, et al. Heart rate variability
`assessment in patients with mitral valve prolapse syndrome.
`Am J Noninvas Cardiol 1993; 7:210–14.
`
`31. Kocovic DZ, Harada T, Shea JB, et al. Alterations of heart rate
`and of heart rate variability after radiofrequency catheter
`ablation of supraventricular tachycardia. Circulation 1993;
`88:1671–81.
`
`32. Counihan PJ, Fei L, Bashir Y, et al. Assessment of heart
`rate variability in hypertrophic cardiomyopathy. Association
`with clinical and prognostic features. Circulation 1993;
`88:1682–90.
`
`33. Dougherty CM, Burr RL. Comparison of heart rate variability
`in survivors and nonsurvivors of sudden cardiac arrest.
`Am J Cardiol 1992; 70:441–8.
`
`34. Algra A, Tijssen JG, Roelandt JR, et al. Heart rate variability
`from 24-hour electrocardiography and the 2-year risk for
`sudden death. Circulation 1993; 88:180–5.
`
`35. Kleiger RE, Stein PK, Bosner MS, et al. Time-domain
`measurements of heart rate variability. In: Malik M, Camm
`AJ, eds. Heart rate variability. Armonk NY, Futura, 1995;
`33–46.
`
`Implantable
`36. Lombardi F, Porta A, Marzegalli M, et al.
`Cardioverter Defibrillator-Heart Rate Variability Italian
`Study Group. Heart rate variability patterns before ventricular
`tachycardia onset in patients with an implantable cardiover-
`ter defibrillator. Participating Investigators of ICD-HRV Italian
`Study Group. Am J Cardiol 2000; 86:959–63.
`
`37. Meyerfeldt U, Wessel N, Schutt H, et al. Heart rate variabil-
`ity before the onset of ventricular tachycardia: differences
`between slow and fast arrhythmias.
`Int
`J Cardiol 2002;
`84:141–51.
`
`38. Tsuji H, Venditti FJ, Manders ES, et al. Determinants of heart
`rate variability. JACC 1996: 28:1539–46.
`
`39. Ozdemir O, Soylu M, Demir AD, et al. Sympathetic
`overactivity as cause of exercise-induced ventricular tachy-
`cardia in patients with normal coronary arteries. Texas Heart
`Inst J 2003; 30:100–4.
`
`40. Valkama JO, Huikuri HV, Koistinen J, et al. Relation between
`heart rate variability and spontaneous and induced ventri-
`cular arrhythmias in patients with coronary artery disease.
`JACC 1995; 25:437–43.
`
`41. Nemec J, Hammill SC, Shen WK, et al. Increase in heart rate
`precedes episodes of ventricular tachycardia and ventricular
`fibrillation in patients with implantable cardioverter defibril-
`lators. Pacing Clin Electrophysiol 1999; 22:1729–38.
`
`42. Kay SM, Marple SL. Spectrum analysis: a modern perspec-
`tive. Proc IEEE 1981; 69:1380–419.
`
`transform handbook:
`43. Addison PS. The illustrated wavelet
`introductory theory and applications in science, engineering,
`medicine and finance. Bristol, Institute of Physics Publishing,
`1992.
`
`44. Chen S-W. A wavelet-based heart-rate variability analysis
`for the study of nonsustained ventricular tachycardia. IEEE
`Transact Biomed Eng 2002; 49:736–42.
`
`45. Toledo E, Gurevitz O, Hod H, et al. Wavelet analysis of
`instantaneous heart rate: a study of autonomic control during
`thrombosis. Am J Physiol Regul Integr Comp Physiol 2003;
`284:R1079–91.
`
`46. Tan B-H, Shimizu H, Hiromoto K, et al. Wavelet transform
`analysis of heart rate variability to assess the autonomic
`changes associated with spontaneous coronary spasm of
`variant angina. J Electrocardiol 2003; 36:117–24.
`
`47. Kozak M, Krivan L, Semrad B. Circadian variations in the
`occurrence of ventricular tachyarrhythmias in patients with
`implantable cardioverter defibrillators. Pacing Clin Electro-
`physiol 2003; 26:731–5.
`
`48. Peckova M, Fahrenbruch CE, Cobb LA, et al. Circadian
`variations in the occurrence of cardiac arrests: initial and
`repeat episodes. Circulation 1998; 98:31–9.
`
`8
`
`
`
`Downloaded from https://academic.oup.com/qjmed/article/98/2/87/1570548 by guest on 19 April 2021
`
`Heart rate variability
`
`95
`
`49. Levine RL, Pepe PE, Fromm RE, et al. Prospective evidence of
`a circadian rhythm for out-of-hospital cardiac arrests. JAMA
`1992; 267:2935–7.
`
`50. Soo LH, Gray D, Young T, Hampton JR. Circadian variation
`in witnessed out of hospital cardiac arrest. Heart 2000;
`84:370–6.
`
`51. Fries R, Hein S, Konig J. Reversed circadian rhythms of
`heart
`rate variability and morning peak occurrence of
`sustained ventricular
`tachyarrhythmias
`in patients with
`implanted cardioverter defibrillator. Med Sci Mon 2002;
`8:CR751–6.
`
`52. Sezgin A, Demircan S, Yalcin F. Reversed circadian rhythms
`of heart rate variability and morning peak occurrence of
`sustained ventricular
`tachyarrhythmias
`in patients with
`implanted cardioverter defibrillator. Med Sci Mon 2003;
`9:LE1, author reply LE2.
`
`53. Huikuri HV, Seppanen T, Koistinen MJ, Airaksinen J,
`Ikaheimo MJ, Castellanos A, Myerburg RJ. Abnormalities in
`beat-to-beat dynamics of heart rate before the spontaneous
`onset of
`life-threatening ventricular
`tachyarrhythmias in
`patients with prior myocardial infarction. Circulation 1996;
`93:1836–44.
`
`54. Shusterman V, Aysin B, Weiss R, et al. Dynamics of low-
`frequency R-R interval oscillations preceding spontaneous
`ventricular tachycardia. Am Heart J 2000; 139:126–33.
`
`55. Pruvot E, Thonet G, Vesin JM, et al. Heart rate dynamics
`at
`the onset of ventricular tachyarrhythmias as retrieved
`from implantable cardioverter-defibrillators in patients with
`coronary artery disease. Circulation 2000; 101:2398–404.
`
`rate
`56. Lombardi F, Porta A, Marzegalli M, et al. Heart
`variability patients before ventricular
`tachycardia onset
`in patients with an implantable cardioverter defibrillator.
`Am J Cardiol 2000; 86:959–63.
`
`57. Vybiral T, Glaeser DH, Goldberger AL, et al. Conventional
`heart
`rate variability analysis of ambulatory electrocar-
`diographic recordings fails to predict imminent ventricular
`fibrillation