| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Cardiovascular Research Group, University of Calgary, Calgary, Alberta, Canada T2N 4N1
 |
ABSTRACT |
|---|
Top
Abstract
Introduction
Methods
Results
Discussion
Appendix
References
|
|---|
The purpose of this work was to determine the temporal origins of the standard deviation of successive 5-min mean heart period sequences (SDANN) and the power of the ultralow-frequency (ULF) spectral band (<0.0033 Hz). We hypothesized that SDANN and ULF might have their origins in changes in human activity rather than slow oscillatory rhythms. Heart period sequences were obtained from 24-h Holter electrocardiograms of 10 healthy ambulatory subjects. There was no evidence of any persistent oscillation within the ULF band. Using moving 4-h windows in short-time Fourier transforms, we showed that the amplitude of ULF fluctuated markedly, particularly during times bordering sleep. The local ULF amplitude correlated (r = 0.59 ± 0.09) with large-scale changes in heart period quantified with 2- and 4-h wavelet transforms. Local SDANN also fluctuated, mainly around times of sleep. Although the 24-h SDANN and ULF values correlated highly, there was little correlation between their temporal distributions (r = 0.10 ± 0.25). The temporal distributions of measures of long-range heart period variability suggest that they reflect changes in human activity levels.
electrocardiogram; power spectral analysis; wavelets
| |
INTRODUCTION |
|---|
|
Top
Abstract
Introduction
Methods
Results
Discussion
Appendix
References
|
|---|
MEASURES OF LONG-RANGE CHANGES in heart rate variability, or more correctly, heart period variability, on 24-h Holter electrocardiograms have been reported to be prognostic factors of outcome in patients. Reductions in either the power in the ultralow-frequency (ULF) spectral band (<0.0033 Hz) or the standard deviation of successive 5-min means of heart periods (SDANN) predict poor survival for patients with chronic severe mitral regurgitation (22), patients with acute (24) or recent myocardial infarction (4, 5), and a variety of outpatients (2). SDANN is an estimate of the changes in heart period due to fluctuations lasting longer than 5 min, and ULF is derived from spectral components having periods longer than 5 min. SDANN correlates well with the square root of the power of the ULF band (5). The physiological basis of slow fluctuations in heart period is unclear (23). Older work suggests that the ULF band might reflect fluctuations in heart period due to peripheral vasomotor (8), thermoregulatory (8), or renin-angiotensin (1) systems. However, enalapril does not change heart rate variability in healthy subjects (13), and the evidence for true oscillatory behavior with periods of several minutes to hours is scant.
If long-period oscillations do not exist, then what might be the origin of the ULF power and SDANN? The mathematical basis of spectral analysis is frequency decomposition, which computes the theoretical contribution of different frequencies to the theoretical function that describes the heart period sequence. Abrupt changes of large magnitude require a number of different frequencies, with powers heavily weighted toward the low frequencies. Thus the times of largest physiological transitions and changes in heart period, i.e., between sleeping and wakefulness, might be an important source of ULF, even though each transition is one-way rather than oscillatory. Bernardi et al. (3) showed that powers in the very low frequency band (<0.03 Hz) were higher in exercising than in resting subjects. These transitions may also be important for SDANN. Because SDANN is a measure of the deviation of 5-min heart period means about the 24-h heart period mean, transitions that shift the local mean heart period away from the 2-h mean should result in increased SDANN.
We addressed the hypothesis that ULF and SDANN simply are due to large-scale transitions in heart period. First, we used time-frequency analysis to determine whether the contributions to ULF power were evenly distributed throughout the recording or were maximal at times of maximal change in heart period. Novel mathematical tools called wavelets were used to quantify these local changes. Second, we measured local contributions to SDANN to assess whether SDANN was evenly distributed throughout the recording or was derived from local deviations in heart period. Finally, we constructed a simple model day based on our findings and showed that it faithfully generated realistic ULF and SDANN characteristics.
| |
METHODS |
|---|
|
Top
Abstract
Introduction
Methods
Results
Discussion
Appendix
References
|
|---|
Study subjects. There were four male and six female healthy subjects, all without a history of structural heart disease, ventricular tachycardia, diabetes, autonomic neuropathy, syncope, or hypertension. None were taking any medications. The mean age was 38 ± 17 yr (range 17-69 yr). The wide age range was selected to reflect the general population. All subjects were reviewed by R. Sheldon, and all gave informed consent.
Ambulatory electrocardiography. The subjects underwent 24-h ambulatory electrocardiography using a system that incorporates a synchronization pulse to reduce variability due to tape speed fluctuations. The recordings were analyzed using the Marquette 8000 Scanner with version 5.7 of the Marquette Arrhythmia Analysis Program to identify and label each QRS. Entire recordings were analyzed by an operator to eliminate cycles in which ventricular beats had abnormal morphological characteristics or were without normal P waves. These beats were replaced by linear interpolation between adjacent normal beats. Unclassified beats were corrected manually and verified. No recordings had ectopy more severe than isolated complexes or atrial tachycardia >3 beats long. The proportion of ectopic complexes ranged from 0 to 0.17%, and no patients had >7 ectopic beats/h. The corrected heart period sequences of ~105 beats were then transferred into MATLAB for further analysis.
Frequency-domain analysis. Three different frequency-domain analyses were performed. The first method computes the power spectra for the 24-h recordings. For these analyses we computed a 24-h in toto Fourier transformation (21). With the use of cubic spline interpolation, the 24-h heart period sequence was uniformly sampled 218 times using a sample interval of 0.329 s. The interpolated series had their means removed, were Hanning windowed, and were then transformed to the frequency domain using a fast Fourier transform. Thus the direct current components were excluded. ULF power is the power <0.0033 Hz in the band, excluding the direct current component, and ULF amplitude is the amplitude of the ULF components. Note that ULF amplitude is the square root of ULF power and is dimensionally equivalent to SDANN; both are expressed in units of milliseconds.
The second method was time-frequency analysis of the ULF band (1.39 × 10
4 to 3.3 × 103 Hz) throughout the 24-h
recording period. We first constructed a 24-h uniformly sampled time
series from the raw heart period sequences. To sample at uniform
intervals and to detect only frequencies <0.0033 Hz, we constructed a
sequence composed of the mean heart periods within successive 150-s
epochs. The 150-s sampling interval allowed for a Nyquist frequency of
0.0033 Hz. We then constructed subsequences, each starting at the
beginning of each 150-s epoch and each lasting 4 h. Fast Fourier
transforms were performed on the Hanning-windowed 4-h subsequences as
they were moved incrementally along in 150-s steps; results were
displayed in conventional time-frequency analysis plots (see Fig. 2).
The 4-h window size was chosen as a compromise between the need for
temporal localization and the need to include as much of the ULF band
as possible. With this procedure we were able to localize temporal
changes from frequencies of 1.39 × 104 to 3.3 × 103 Hz.
The third method involved summing the ULF power within each of these
4-h subsequences to produce an estimate of temporally localized ULF
power. This local ULF power is an estimate of the contribution that the
local 4-h structures make toward the 24-h ULF power.
Time-domain analysis. Total SDANN for an entire 24-h recording is the standard deviation of a sequence of the mean heart periods within successive 300-s epochs. The contribution that the mean heart period from each 300-s epoch makes toward the total SDANN value was then calculated, and this was denoted as the local SDANN. We replaced each 300-s heart period mean value with the 24-h heart period mean and recalculated the standard deviation of this altered 24-h sequence. The difference between the standard deviation of this new 24-h sequence and total SDANN is the local SDANN, i.e., the contribution made by that particular 300-s epoch toward total SDANN. Local SDANN values were normalized such that their sum equaled the total SDANN (see APPENDIX).
Wavelet analysis. To focus on the contributions of temporally localized changes in heart period to overall measures of heart period variability, we used an independent method of detecting and quantifying these localized changes. Wavelets are mathematical tools particularly suited for this purpose. They are localized mathematical functions of specified scale (duration), size (amplitude), and shape that, when passed through a sequence with the use of convolution, identify events on the basis of their own characteristics (14, 17). The wavelet we used was the first derivative of the Gaussian function in the interval between the mean ± 3 SD. This wavelet has a mathematical shape resembling the tilde (~) key on a keyboard. Convolution is the stepwise passage of the wavelet through the heart period sequence, summing at each step the multiplication between the wavelet and the small part of the sequence sharing the same duration as the wavelet. This detects and quantifies the changes in the heart period sequence for each incremental step. Wavelet analysis emphasizes the local properties of the heart period sequences. In this study we wanted to detect and quantify those changes in heart period that occur mainly over a duration of 4 h. We convolved the 24-h sequences of successive 300-s mean heart periods with the 4-h-duration wavelet. The result was a sequence of values representing the 4-h heart period derivative (i.e., the Gaussian center-weighted change in heart period with respect to time, computed at the 4-h timescale) as measured at times throughout the 24-h recording period. The changes detected by the 4-h wavelet transforms are expressed in units of milliseconds per hour.
Statistical analysis. Results are expressed as means ± SD. The correlation of variables between groups was determined by linear regression analysis.
| |
RESULTS |
|---|
|
Top
Abstract
Introduction
Methods
Results
Discussion
Appendix
References
|
|---|
Table 1 lists the 24-h SDANN and ULF power values for the 10 subjects. The 24-h mean heart periods range from 687 to 903 ms, and the total SDANN values range from 81 to 207 ms. The mean SDANN is 134 ± 34 ms, closely resembling a reported normative mean value of 127 ± 35 ms (5, 23). The mean ULF is 13,469 ± 9,299 ms2/Hz, close to a reported normative mean value of 18,420 ± 10,639 ms2/Hz (6). Because total SDANN and ULF amplitude correlate highly in patients with structural heart disease, we tested this correlation in the study population. Figure 1 shows that total SDANN correlated with ULF amplitude (r = 0.90, P < 0.01). These values closely resemble those of other studies (5, 6, 12).
|
|
Local contributions to ULF. We performed conventional time-frequency analysis of 24-h recordings to determine whether the power in the ULF band was due to a collection of ULF oscillatory processes or to the computed contributions of ULF related to large changes in heart period. Figure 2 displays the time-frequency analyses of three recordings, along with their corresponding local mean heart period sequences. There is no distinct spectral peak within the ULF band. The large fluctuations in the local mean heart period sequences are associated with a broad range of local ULF components, with the power increasing markedly toward the low end of the ULF band.
|
|
|
Local contributions to SDANN. The close correlation between total SDANN and total ULF amplitude suggested that SDANN might also reflect temporally localized heart period changes (see APPENDIX). Figure 4 displays 24-h plots from three representative subjects of local mean heart period (Fig. 4A), the absolute difference between the local mean heart period and the 24-h mean (Fig. 4B), and local SDANN (Fig. 4C). The absolute differences and the local SDANN values are elevated during nighttime hours. Also, numerous local contributions to SDANN are scattered throughout the recording period, particularly at times of maximal local deviation of heart period from the global mean. Finally, we correlated the local SDANN for all subjects with the squares of the differences between the local mean heart period and the 24-h mean heart period. The correlation coefficients for all subjects were 1.00. This level of correlation and the reason for studying the square of the differences are explained in the APPENDIX. Thus local SDANN simply reflects the local deviation of heart period from the 24-h mean.
|
Comparison of local SDANN and local ULF. One explanation for the significant correlation between total SDANN and total ULF amplitude is that they both are derived from the same large-scale changes in the heart period sequences but measure different aspects of these changes. Figures 3 and 4 show that local SDANN and local ULF amplitude do not occur at the same times, with a temporal correlation coefficient of 0.10 ± 0.25. This suggests that these measures reflect different aspects of the same structures.
| |
DISCUSSION |
|---|
|
Top
Abstract
Introduction
Methods
Results
Discussion
Appendix
References
|
|---|
The major new finding of this study is that the contributions to SDANN and ULF vary markedly throughout a 24-h recording in a fashion that has implications for our understanding of the nature of the large-scale structures of heart period variability.
Local structure of ULF. Our data show that no characteristic frequency dominates the ULF band, that the power in the ULF band is distributed discontinuously throughout the recording period, and that the ULF band is maximal at times of maximal local change in heart period. Furthermore, these changes often are unidirectional, in contrast to the bidirectional changes of true oscillatory behavior. This suggests that ULF is not derived from persistent, slow oscillations but from large, local, unidirectional changes in heart period. The explanation for this observation lies in the nature of the Fourier transform. This tool decomposes any function into its theoretical frequency components, which when summed form the original function. Spectral analysis is not proof of the presence of harmonic oscillators but is merely a method of describing a sequence. The largest changes in heart period are described by a broad band of frequencies with heavy weighting toward the lowest frequencies. Thus the power in the ULF band simply reflects large, infrequent events such as going to sleep and awakening.
Local structure of SDANN. Like ULF, SDANN is distributed discontinuously. Local SDANN simply reflects the square of the difference between local mean heart period and the 24-h mean heart period. Furthermore, the SDANN sequences are dominated by irregularities. There are many local structures scattered throughout the recordings, which precludes the existence of a persistent oscillation as a dominant source of SDANN. Given that these abrupt structures are due to local deviations from the 24-h mean heart period, we suggest that local SDANN is attributable to changes in the physical and physiological behavior of the subjects. These changes occur on scales ranging from a few minutes to several hours. The SDANN contribution from sleep is recognizable by its prolonged nature.
Model of large-scale heart period variability. We propose that heart period changes that are quantifiable by these two methods reflect changes in human behavior. Because SDANN expresses the local heart period relative to the global mean heart period, it reflects the current behavior relative to the long-term average behavior. In contrast, ULF reflects the transitions between behavioral states. That is, local SDANN reflects a temporally localized state, and local ULF reflects transitions to and from that state. Because states are bounded by transitions, total ULF magnitude and total SDANN will correlate for each subject, whereas their local ULF amplitudes and local SDANN values will not. The most prominent large-scale structure is the sleep-wake cycle, with ULF peaking at the times of going to sleep and awakening and SDANN peaking during sleep.
A simple illustration of this concept is depicted in Figure 5. Figure 5A displays a model 24-h day with a baseline daytime heart period of 650 ms and a mean nighttime heart period of 1,000 ms. During wakeful hours we added a 0.1-Hz oscillator with an amplitude of 100 ms, and during sleep hours we added a 0.25-Hz oscillator with an amplitude of 100 ms. Figure 5B shows the 24-h power spectrum for this simple day. The total ULF power is 27,069 ms2/Hz, and the total SDANN is 165 ms. Figure 5, C and D, shows the local contributions to ULF amplitude and SDANN, calculated as was done for the clinical data. Local SDANN is maximal during the night, when local heart period is most different from the global mean heart period, and local ULF is maximal at the transitions between sleep and wakefulness. A similarly simplified conclusion regarding the power-law structure of the QRS complex shape was reached by Lewis and Guevara (16) and Chialvo and Jalife (7).
|
Events-based approach to heart period variability. These findings suggest the benefits of an events-based approach to the analysis of heart period variability. This events-based approach focuses on the individual integrated responses of the sinus node to transient physiological stresses. This approach does not assume the existence of any global properties (10, 11, 20); rather, it focuses on temporally localized structures within the recordings. Furthermore, it overcomes the problem of stationarity, which confounds frequency-domain analyses. An events-based approach focuses on events as they occur and does not rely on long-term statistical constancy in the overall sequence.
Clinical implications. This report may both simplify further uses of large-scale heart period variability and open avenues of investigation of its physiological basis. Given the close correlation between total SDANN and total ULF amplitude and the understanding that they simply reflect different aspects of the same large-scale increases and decreases in heart period, it seems reasonable to recommend that only SDANN be used (23). SDANN is computationally simpler than ULF, is more tolerant of data imperfections, and involves none of the assumptions that underlie spectral analysis. It is tempting to speculate that other, simpler measures such as the difference between mean supine heart rate and mean standing heart rate over a brief period might be a simpler and equally adequate measure. It is noteworthy that chronotropic incompetence during exercise was a prognostic factor of poor outcome in the Framingham study (15).
Limitations. The subjects were healthy volunteers who were not representative of most patients with structural heart disease. This healthy population was chosen to obviate anticipated problems with medication use, autonomic changes due to aging or heart failure, possible physical limitations, and computational problems due to low SDANN or ULF in some patients with significant left ventricular disease. The physiological basis for diminished SDANN and ULF in patients with structural heart disease is not addressed by this study but does become amenable to analysis. The possible causes include reduced intrinsic chronotropic responsiveness (15), the use of negatively chronotropic medications, and the inanition and lethargy that often accompany chronic severe disease.
The limitations of power spectral analysis include the nonstationarity of the sequence and the problem with singular-type structures within the sequences. Many of these problems are inescapable in Fourier analyses of Holter electrocardiograms and are reasons for favoring the use of time-domain analyses such as SDANN. Finally, the activities of the subjects during the recording period were not controlled.| |
APPENDIX |
|---|
|
Top
Abstract
Introduction
Methods
Results
Discussion
Appendix
References
|
|---|
The following section demonstrates that local SDANN is proportional to (and thus correlates highly with) the square of the difference between the local mean heart period (HP300 s) and the global mean heart period (HP24 h). That is
![local SDANN(<IT>k</IT>) ∝ [HP<SUB>300s</SUB>(<IT>k</IT>) − HP<SUB>24h</SUB>]<SUP>2</SUP>](/content/vol274/issue5/fulltext/H1465/img001.gif)
|
To demonstrate this relationship, let
xk = HP300 s(k)
and 
= HP24 h.
|
|
. Consider the Taylor series
expansion of at
a = 0
|
|
|
|
|
![local SDANN(<IT>k</IT>) ≈ <FR><NU>[HP<SUB>300s</SUB>(<IT>k</IT>) − HP<SUB>24h</SUB>]<SUP>2</SUP></NU><DE>574 <RAD><RCD><IT>q</IT><SUB><IT>k</IT></SUB></RCD></RAD></DE></FR>](/content/vol274/issue5/fulltext/H1465/img012.gif)
|
constant
![local SDANN(<IT>k</IT>) ∝ [HP<SUB>300s</SUB>(<IT>k</IT>) − HP<SUB>24h</SUB>]<SUP>2</SUP>](/content/vol274/issue5/fulltext/H1465/img014.gif)
|
| |
ACKNOWLEDGEMENTS |
|---|
This work was supported by Grant PG11188 from the Medical Research Council of Canada, Ottawa, ON, Canada, and the Calgary General Hospital Research and Development Committee, Calgary, AB, Canada.
| |
FOOTNOTES |
|---|
D. Roach was a Postdoctoral Fellow of the Heart and Stroke Foundation of Canada, Ottawa, ON, Canada.
Address for reprint requests: R. Sheldon, Faculty of Medicine, Univ. of Calgary, 3330 Hospital Dr., N.W., Calgary, AB, Canada T2N 4N1.
Received 9 June 1997; accepted in final form 5 January 1998.
| |
REFERENCES |
|---|
|
Top
Abstract
Introduction
Methods
Results
Discussion
Appendix
References
|
|---|
1.
Akselrod, S.,
D. Gordon,
F. A. Ubel,
D. C. Shannon,
A. C. Barger,
and
R. J. Cohen.
Power spectrum analysis of heart rate fluctuation: a quantitative probe of beat-to-beat cardiovascular control.
Science
213:
220-222,
1981
2.
Algra, A.,
J. G. Tijssen,
J. R. Roelandt,
J. Pool,
and
J. Lubsen.
Heart rate variability from 24-hour electrocardiography and the 2-year risk for sudden death.
Circulation
88:
180-185,
1993
3.
Bernardi, L.,
F. Valle,
M. Coco,
A. Calciati,
and
P. Sleight.
Physical activity influences heart rate variability and very low frequency components in Holter electrocardiograms.
Cardiovasc. Res.
32:
234-237,
1996
4.
Bigger, J. T.,
J. L. Fleiss,
R. C. Steinman,
L. M. Rolnitzky,
R. E. Kleiger,
and
J. N. Rottman.
Frequency domain measures of heart period variability and mortality after myocardial infarction.
Circulation
85:
164-171,
1992
5.
Bigger, J. T.,
J. L. Fleiss,
R. C. Steinman,
L. M. Rolnitzky,
W. J. Schneider,
and
P. K. Stein.
RR variability in healthy, middle-aged persons compared with patients with chronic coronary heart disease or recent acute myocardial infarction.
Circulation
91:
1936-1943,
1995
6.
Bigger, J. T.,
R. C. Steinman,
L. M. Rolnitzky,
J. L. Fleiss,
P. Albrecht,
and
R. J. Cohen.
Power law behavior of RR-interval variability in healthy middle-aged persons, patients with acute myocardial infarction, and patients with heart transplants.
Circulation
93:
2142-2151,
1996
7.
Chialvo, D. R.,
and
J. Jalife.
1/f
power spectral density of the cardiac QRS complex is not associated with a fractal Purkinje system.
Biophys. J.
60:
1303-1305,
1991.
8.
Fallen, E. L.,
M. V. Kamath,
and
D. N. Ghista.
Power spectrum of heart rate variability: a non-invasive test of integrated neurocardiac function.
Clin. Invest. Med.
11:
331-340,
1988[Medline].
9.
Hrushesky, W. J.,
D. Fader,
O. Schmitt,
and
V. Gilbertsen.
The respiratory sinus arrhythmia: a measure of cardiac age.
Science
224:
1001-1004,
1984
10.
Kanters, J. K.,
M. V. Hojgaard,
E. Agner,
and
N. H. Holstein-Rathlou.
Short- and long-term variations in non-linear dynamics in heart rate variability.
Cardiovasc. Res.
31:
400-409,
1996[Medline].
11.
Kanters, J. K.,
N. H. Holstein-Rathlou,
and
E. Agner.
Lack of evidence for low-dimensional chaos in heart rate variability.
J. Cardiovasc. Electrophysiol.
5:
591-601,
1994[Medline].
12.
Katona, P. G.,
and
J. I. H. Felix.
Respiratory sinus arrhythmia: noninvasive measure of parasympathetic cardiac control.
J. Appl. Physiol.
39:
801-805,
1975
13.
Kaufman, E. S.,
M. Bosner,
J. T. Bigger,
P. K. Stein,
R. E. Kleiger,
L. M. Rolnitzky,
R. C. Steinman,
and
J. L. Fleiss.
Effects of digoxin and enalapril on heart period variability and response to head-up tilt in normal subjects.
Am. J. Cardiol.
72:
95-99,
1993[Medline].
14.
Lasserre, M.,
N. R. Eldridge,
and
D. Roach.
A continuous wavelet transform analysis of the singularity behaviour of synthetic aperture radar data.
Soc. Photo-Opt. Instrumentation Eng. Proc.
2491:
682-696,
1995.
15.
Lauer, M. S.,
P. M. Okin,
M. G. Larson,
J. C. Evans,
and
D. Levy.
Impaired heart rate response to graded exercise. Prognostic implications of chronotropic incompetence in the Framingham Heart Study.
Circulation
93:
1520-1526,
1996
16.
Lewis, T. J.,
and
M. R. Guevara.
1/f power spectrum of the QRS complex does not imply fractal activation of the ventricles.
Biophys. J.
60:
1297-1300,
1991[Medline].
17.
Mallat, S.,
and
W. L. Hwang.
Singularity detection and processing with wavelets.
IEEE Trans. Info. Theor.
38:
617-643,
1992.
18.
Malliani, A.,
M. Pagani,
F. Lombardi,
and
S. Cerutti.
Cardiovascular neural regulation explored in the frequency domain.
Circulation
84:
482-492,
1991
19.
Pagani, M.,
F. Lombardi,
S. Suzzetti,
O. Rimoldi,
R. Furlan,
P. Pizzinelli,
G. Sandrone,
G. Malfatto,
S. Dell'Orto,
E. Piccaluga,
M. Turiel,
G. Baselli,
S. Cerutti,
and
A. Malliani.
Power spectral analysis of heart rate for arterial pressure variabilities as a marker of sympatho-vagal interaction in man and conscious dog.
Circ. Res.
59:
178-193,
1986
20.
Roach, D.,
and
R. Sheldon.
Short-term predictability in heart period variability (Abstract).
Can. J. Cardiol.
11:
129E,
1995.
21.
Rottman, J. N.,
R. C. Steinman,
P. Albrecht,
J. T. Bigger, Jr.,
L. M. Rolnitzky,
and
J. L. Fleiss.
Efficient estimation of the heart period power spectrum suitable for physiologic or pharmacologic studies.
Am. J. Cardiol.
66:
1522-1524,
1990[Medline].
22.
Stein, K. M.,
J. S. Borer,
C. Hochreiter,
P. M. Okin,
E. M. Herrold,
R. B. Devereux,
and
P. Kligfield.
Prognostic value and physiologic correlates of heart rate variability in chronic severe mitral regurgitation.
Circulation
88:
127-135,
1993
23.
Task Force of the European Society of Cardiology and the North American Society of Pacing and Electrophysiology.
Heart rate variability. Standards of measurement, physiological interpretation, and clinical use.
Circulation
93:
1043-1065,
1996
24.
Vaishnav, S.,
R. Stevenson,
B. Marchant,
K. Lagi,
K. Ranjadayalan,
and
A. Timmis.
Relation between heart rate variability early after acute myocardial infarction and long-term mortality.
Am. J. Cardiol.
73:
653-657,
1994[Medline].
This article has been cited by other articles:
|
|
 |
|
  N. Aoyagi, K. Ohashi, and Y. Yamamoto Frequency characteristics of long-term heart rate variability during constant-routine protocol Am J Physiol Regulatory Integrative Comp Physiol, July 1, 2003; 285(1): R171 - R176. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. E. Burgess, D. C. Randall, R. O. Speakman, and D. R. Brown Coupling of sympathetic nerve traffic and BP at very low frequencies is mediated by large-amplitude events Am J Physiol Regulatory Integrative Comp Physiol, March 1, 2003; 284(3): R802 - R810. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. Aoyagi, K. Ohashi, S. Tomono, and Y. Yamamoto Temporal contribution of body movement to very long-term heart rate variability in humans Am J Physiol Heart Circ Physiol, April 1, 2000; 278(4): H1035 - H1041. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. Roach, R. Haennel, M. L. Koshman, and R. Sheldon Origins of heart rate variability: relationship of heart rate burst morphology to work duration and load Am J Physiol Heart Circ Physiol, October 1, 1999; 277(4): H1491 - H1497. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. Roach, E. Thakore, and R. S. Sheldon Large-magnitude, transient, bradycardic events in rabbits Am J Physiol Regulatory Integrative Comp Physiol, July 1, 1999; 277(1): R243 - R249. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J.-O. Fortrat, C. Formet, J. Frutoso, and C. Gharib Even slight movements disturb analysis of cardiovascular dynamics Am J Physiol Heart Circ Physiol, July 1, 1999; 277(1): H261 - H267. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. Roach, P. Malik, M. L. Koshman, and R. Sheldon Origins of Heart Rate Variability : Inducibility and Prevalence of a Discrete, Tachycardic Event Circulation, June 29, 1999; 99(25): 3279 - 3285. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Visit Other APS Journals Online |
