| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Department of Cardiological Sciences, St. George's Hospital Medical School, London SW17 0RE, United Kingdom
 |
ABSTRACT |
|---|
 TOP
 ABSTRACT
 INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
|
|---|
Recently, it was demonstrated that the
QT-RR relationship pattern varies significantly among healthy
individuals. We compared the intra- and interindividual variations of
the QT-RR relationship. Twenty-four-hour 12-lead digital
electrocardiograms (ECGs; SEER MC, GE Marquette; 10-s ECG recorded
every 30 s) were obtained at baseline and after 24 h, 1 wk,
and 1 mo in 75 healthy subjects (42 women, 33 men, age 27.9 ± 9.6 vs. 26.8 ± 7.5 yr, P = not significant). QT
interval was measured automatically in each ECG by six different algorithms, and the mean of the six measurements was analyzed. In each
recording of each individual, QT-RR relationship was assessed by 10 different regression models including linear (QT = 
+ 
× RR) and parabolic (QT = × RR
)
models. Standard deviations (SDs) of regression parameters and of consecutive recordings of each individual were compared with SD of
the individual means. Intrasubject stability and interindividual variability were further tested by ANOVA. With all models,
intraindividual SDs of the regression parameters were highly
significantly smaller than SD of individual means (P < 10
5-109). The intrasubject stability
was further confirmed by ANOVA (P < 1019-1030). The QT-RR relationship
exhibits substantial intersubject variability as well as a high
intrasubject stability. This has practical implications for a precise
estimation of the heart rate-corrected QT interval in which optimized
subject-specific rate correction formulas should be used.
corrected QT interval; repolarization
| |
INTRODUCTION |
|---|
|
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
|
|---|
THE RELATIONSHIP between duration of the QT interval and heart rate has long been debated. Studies of this relationship are particularly important for improved calculations of heart rate-corrected QT interval, which is a recognized marker of arrhythmic risk (7, 26, 33, 34) including the risk of drug-induced proarrhythmia (12). Although many mathematical models have been proposed to describe the QT-RR relation, none of the models (and corresponding heart rate correction formulas) has been accepted as universally applicable (1, 10, 13, 19, 29). As repeatedly pointed out, any formula describing the relationship between the QT interval and the preceding RR interval is limited by physiological factors, such as the extra-heart rate influences on myocardial repolarization (4, 5, 30) and the "lag hysteresis" of the QT-RR relationship (8, 17).
Less attention has so far been paid to individual differences in the QT-RR relationship, which also limit the success of any "universal" mathematical model and contribute to the errors of any fixed heart rate correction formula. Substantial interindividual differences of the QT-RR relationship have been reported (6, 25) but, until recently, not studied systematically.
In a recent study (22), the 24-h QT-R relationship of 50 healthy subjects was investigated with different generic regression models. With all these models, the regression parameters varied very
significantly between different individuals. For example, with the
parabolic model (QT = × RR), the
individually optimized coefficient varied from 0.23 to 0.49, that
is, the interindividual differences were greater than the difference
between the Bazett ( = 0.5) and Fridericia ( = 0.33) corrections.
These results strongly suggest that the traditional approach for rate correction of the QT interval with the Bazett correction or any other fixed formula is at best problematic. A formula that appropriately corrects the QT intervals (i.e., renders them heart rate independent) in one individual may significantly over- or undercorrect in another individual.
There are several possible reasons for the high interindividual variability. These include environmental influences and interindividual differences in heart rate variability as well as inherent genetically determined differences in the QT-RR pattern. To differentiate between these possibilities, the intraindividual stability and the interindividual variability must be compared.
With this in mind, this study was designed to investigate systematically the intraindividual and interindividual variations of the QT-RR relationship. For this purpose, a series of repeated 24-h 12-lead electrocardiographic (ECG) Holter recordings was obtained in a group of healthy subjects.
| |
METHODS |
|---|
|
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
|
|---|
Study population. The study investigated 75 subjects (42 women, mean age 27.9 ± 9.6 yr, range 18-59 yr; 33 men, mean age 26.8 ± 7.5 yr, range 18-43 yr) with no history of cardiac disease, a normal physical examination, and a normal baseline 12-lead ECG. No drugs were taken by any participant for at least 2 wk before the first recording and/or during the subsequent phases of the study. The study was approved by the local ethics committee, and all participants gave written consent.
Data acquisition. In each subject, four consecutive 24-h 12-lead ECG Holter recordings (250 Hz, 12-bit analog-to-digital resolution; SEER MC digital recorder; GE Marquette, Milwaukee, WI) were obtained. After a baseline recording, subsequent recordings were obtained after 24 h, 1 wk, and 1 mo. The participants were either employees of St. George's Hospital and the Medical School or undergraduate medical students. The recordings were obtained during standard working days, and although the participants were asked to refrain from substantial physical excesses of their usual daily routine, no other attempts were made to standardize environmental influences during the recording phases. Generally, the environment and occupational routines under which the recordings were obtained were fairly similar for all participants.
One 10-s ECG was recorded every 30 s for the whole nominal 24-h Holter recording. To avoid confusion, we use the term "ECG recording" to describe the 24-h 12-lead ECG Holter recording and the term "ECG sample" to describe the individual 10-s 12-lead ECG snapshot.QT interval measurement. It was technically impossible to measure the QT interval in all ECG samples manually, because the study involved ~840,000 individual ECG samples. Therefore, all ECG samples were processed automatically with the following strategy developed to improve the accuracy of algorithmic QT interval measurement. A lead-specific median ECG beat was constructed from the sinus beats in each 10-s ECG (37). The QT interval was measured in each lead of each median ECG beat by six different automatic methods of the QT Guard package (GE Marquette; Ref. 37). For methods 1 and 2, the intersect between the baseline and the downslope tangent of the T wave was computed from the interpolation of 6 (method 1) and 12 (method 2) samples around the inflex point. For methods 3-6, the final drop of the ECG signal (methods 3 and 4) and of its first derivative (methods 5 and 6) below 5% (methods 3 and 5) and 15% (methods 4 and 6) of the maximum value within the T wave were used.
For each of these methods, the median QT interval among all measurable leads was obtained for each ECG sample. The QT measurement of an ECG sample was accepted if at least six leads were measurable (as judged by the QT Guard package; Ref. 37) and if the median QT intervals provided by the six methods differed by
40 ms. The threshold of six measurable leads and 40-ms difference between different methods was based on previous experience with the QT Guard
package (2). If the measurement was accepted, the average of the six median QT intervals was taken as the valid QT interval.
Compared with the maximum QT interval of all measurable leads, the
automatically measured median QT interval provides superior data
stability (21). In healthy hearts, one is also unlikely to
observe substantial T wave abnormalities in only one lead
("regional" QT prolongation; Ref. 21), and hence, the
difference between median and maximum QT interval is less expressed
than in patients with repolarization abnormalities and/or nonspecific T
wave changes.
Heart rate measurement. Within each ECG sample, individual cardiac cycles were identified by pattern matching between the median beat and the native ECG signal (23). From these, the average RR interval was obtained for each ECG sample.
To avoid mismatch between QT and RR intervals due to the hysteresis of QT-RR adaptation, only ECG samples with relatively stable heart rate were considered. For this purpose, linear regressions were calculated between the duration and the order number of individual RR intervals. ECG samples were excluded if, through the whole 10-s recording, RR intervals were significantly (judged by 95% confidence interval) rising or falling by
5 ms/interval. Although this permitted sinus
arrhythmia in accepted ECG samples, the samples from episodes of
systematic heart rate acceleration or deceleration were excluded.
Data analysis. Only ECG samples with accepted QT interval measurements and accepted stable RR interval measurements were considered. ECG recordings with at least 900 acceptable samples were used in the analysis. The limit of 900 samples was chosen arbitrarily to exclude ECG recordings with a poor definition of the QT-RR pattern. Finally, only data of subjects for whom all four repeated ECG recordings were acceptable were considered.
The QT-RR data of each 24-h recording were investigated with 10 different regression models. These models included a variety of mathematically defined functions the graphs of which resembled the possible physiological shapes of the QT-RR relationship. Thus the set of all the models was constructed to cover the physiological patterns as comprehensively as possible. By investigating all different models, the bias was avoided that might have otherwise been introduced by making inappropriate assumptions about the QT-RR relationship. The following regression models were used: linear model (QT = + × RR); hyperbolic model (QT = + /RR); parabolic log/log model (QT = × RR);
logarithmic model [QT = + × ln(RR)]; shifted
logarithmic model [QT = ln( + × RR)];
exponential model (QT = + × eRR); arcus tangent model [QT = + × arctan(RR)]; hyperbolic tangent model [QT = + × tgh(RR)]; arcus hyperbolic sine model [QT = + × arcsinh(RR)]; and arcus hyperbolic cosine
model [QT = + × arccosh(RR + 1)].
In all formulas the QT and RR intervals were expressed in seconds. All
models were designed to have two regression parameters to make the
closeness of regression fit easily comparable.
Statistical analysis.
With each regression model, four values of the parameters and were obtained from the repeated ECG recordings of each subject with
custom-written software. These were averaged for each individual, and
the individually mean parameters in men and women were compared with a
nonparametric two-sample Wilcoxon test.
and in each participant
were compared with the standard deviation of the subject-specific mean
values of and with a nonparametric one-sample Wilcoxon test. To
eliminate the influence of sex differences on the intersubject variability, the comparison of intra- and intersubject standard deviations was repeated separately for the groups of men and women in
the study. The intrasubject reproducibility of individual regression parameters from the repeated recordings was also compared with repeated-measures ANOVA.
To compare the closeness of fit of the different regression models, the
averaged residuum was obtained for each ECG recording and each model.
For each subject, the regression model was identified that led to the
lowest residua averaged over the four ECG recordings. The residua of
different models were compared with a nonparametric one-sample Wilcoxon
test. Statistica version 5.1 (StatSoft) was used for analysis.
Data are presented as means ± SD unless otherwise stated. A
P value <0.05 was considered statistically significant.
| |
RESULTS |
|---|
|
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
|
|---|
Figure 1 shows examples of QT-RR
patterns of the repeated recordings in two women and two men. The
patterns clearly suggest both substantial intersubject variability and
high intrasubject stability of the QT-RR relationship. Figure 1 also
confirms that in addition to heart rate, there are other important
determinants of QT interval, because at the same RR interval, QT
interval differed substantially in different ECG samples.
|
Of all subjects, 46 had all four ECG recordings accepted for analysis (22 men and 24 women, 26.8 ± 7.3 and 27.3 ± 8.1 yr, respectively; P = not significant). The 184 accepted ECG recordings contained on average 1,504 ± 283 ECG samples/recording. The intervals between the first and subsequent ECG recordings were 1.0 ± 0.0, 7.7 ± 2.0, and 31.2 ± 3.9 days, respectively.
Table 1 shows the regression parameters
and for all regression models in men and women. Except for the
parameter of the hyperbolic model, the regression parameters of all
models differed significantly between the sexes. In general, women had steeper and more curved QT-RR patterns than men.
|
Table 2 shows the distribution of the
optimum regression models as judged by the regression residua. In 52%
of all subjects the linear or the arcus hyperbolic sine model led to
the best fit to the QT-RR data, whereas the parabolic model was optimal in only 4% of the subjects and the hyperbolic model in none.
|
The intrasubject and intersubject variability of the regression
parameters in all subjects, in women, and in men are compared in
Table 3. With all models, the
within-individual standard deviations of both parameters were highly
significantly smaller than the standard deviations of individual means.
The ANOVA analysis confirmed a highly significant intrasubject
repeatability of regression parameters compared with intersubject
variability.
|
Figure 2 presents the distribution of
individual values of the regression parameters for the linear model and
arcus hyperbolic sine model, that is, the most frequent optimum models.
It is clearly visible that the spread of the individual means of the
regression parameters was substantially larger than the
intrasubject variability.
|
Table 4 shows the comparison of the
intrasubject stability of the regression parameters (expressed by
standard deviations of the regression parameters derived from the 4 repeated ECG recordings in each subject) between women and men. With
several regression models (including the linear and arcus hyperbolic
sine model), the intrasubject stability was significantly or nearly
significantly greater in men than in women. However (see Table 3), the
differences in the individual stability between sexes were minute
compared with the intersubject variability in the whole group.
|
| |
DISCUSSION |
|---|
|
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
|
|---|
Findings of study. With the use of a different population and more detailed analysis, this study confirms the previous findings (22) that the QT-RR relationship differs substantially between different healthy individuals. At the same time, the study also shows that the individual QT-RR pattern assessed from 24-h 12-lead Holter recordings is stable over a period of (at least) 1 mo. The observation that the intraindividual variability is lower than the interindividual variability is not in principle surprising. However, it is the degree of the difference that is unexpected. Despite the broad similarity in the recording conditions, there were very substantial differences between different individuals. However, despite this large variability within the group, the individual patterns were surprisingly stable. Thus the results suggest that the inherent, probably genetically determined, differences in cardiac repolarization are substantially greater than the differences due to environmental influences and autonomic regulation. On the basis of these observations, it does not seem unreasonable to speculate that the individual QT-RR relationship has, in addition to autonomic and environmental influences, unique "fingerprint-like" properties.
Relation to previous studies. Although the sex differences in the QT-RR relationship pattern were already noted by Bazett (3), the individual variations of the QT-RR pattern have attracted much less attention. Because the QT-RR relationship in the general population is likely to vary within certain limits, it is possible that analyses of pooled data from very large populations can produce very similar values of interpopulation parameters purely due to regression to the mean [e.g., the Framingham (32) and Rotterdam (7) studies]. However, smaller studies, such as those from which most heart rate correction formulas have been derived, have very often produced substantially different values of parameters in proposed formulas for heart rate correction of the QT interval (see discussion in Ref. 22). Traditionally, these differences have been attributed to differences in heart rate, age, physiological conditions, etc., but not to inherently different individual QT-RR patterns.
The interindividual differences of the QT-RR regression parameters observed in this study are comparable to those previously reported between different clinically defined groups. For example, the slope of the linear QT-RR model in this study ranged from 0.137 to 0.199 in men and from 0.145 to 0.243 in women. These ranges are practically the same as the statistically significant differences reported between groups of patients with exercise-induced ventricular tachycardia with and without structural heart disease (11) or between patients with idiopathic ventricular fibrillation and healthy subjects (36). With multiple QT-RR data from standard Holter recordings in 21 healthy subjects, Molnar et al. (25) compared five correction formulas with individually optimized formulas derived from the data of each patient. They found that the individual optimizations were superior to conventional heart rate corrections. Davey (6) estimated the slope of the individual QT-heart rate linear regression lines during exercise in healthy subjects and patients with hypertrophic cardiomyopathy and heart failure. He obtained corrected QT intervals (QTc) by extrapolating the QT-heart rate regression line to a heart rate of 60 beats/min in each participant. These "individual" QTc separated the three study groups better than the Bazett-corrected QTc and, unlike the latter, were not correlated to the heart rate. No previous author, to our knowledge, has investigated or even formulated the problem of the stability of the individual QT-RR pattern. Lecocq et al. (18) studied the QT-RR relationship of 11 subjects at rest, during exercise, and after intravenous isoproterenol and of 12 subjects at rest and after oral propranolol. An exponential model (QT = A
Bek×RR, where A,
B and k are the regression parameters) was fitted
to the QT-RR data. Four subjects participated in both studies
conducted 10 mo apart. When comparing individually the equations from
both studies, the authors mentioned that they "... could not find
any significant change between the 2 periods ...," which was not
commented on further but is in full agreement with the results of the
present study.
Very recently, Malik (19) tested the practical
appropriateness of an "individualized" rate correction approach in
a study of drug-induced QT interval prolongation. That study suggested that the individual QT-RR relationship could be reliably estimated from
a much smaller number of QT-RR data pairs.
Not only the regression parameters but also the general form of the
regression model that best fits the QT-RR data varied significantly
among different individuals. The linear model (QT = + × RR; Refs. 7, 13, 32)
was optimum in 31 out of a total of 96 participants in this study and
in the previous study (22). On the other hand, the
parabolic model (QT = + RR), from which
the Bazett formula (3) and other popular heart rate
correction formulas (9, 14) have been derived, was optimum in only seven subjects, whereas the hyperbolic model (QT = + /RR; 15) was not an optimum in any subject. Different
models reported in the literature have been tested in very different databases [pooled data of all subjects, multiple recordings in each
subject, or even a mixture of both, as in the original study of Bazett
(3)] under different physiological conditions (rest, exercise test, or 24-h recordings), various heart rate ranges, and
numbers of QT-RR samples. To our knowledge, no previous study has
addressed systematically the distribution of different QT-RR regression
shapes in a healthy population.
These results [as well as those of our previous study
(22)] should not be interpreted as a suggestion of the
superiority of any regression model for studies of QT-RR relationship
in the general population. On the contrary, it seems that no
"superior" or "physiological" QT-RR regression model exists. It
is likely that similar intersubject difference and intrasubject
stability also exist in the shapes of QT-RR relationships and that
different individuals would be best studied with different regression
curvatures. Analyses of different groups of healthy subjects and/or
cardiac patients would most likely render significantly different (and probably nonreproducible) distributions of the "optimum" regression models.
It can be speculated that the intersubject variability of the QT-RR
pattern may reflect interindividual variations in the expression and/or
autonomic control of different ionic repolarizing currents. It is known
that different gene mutations affecting different repolarizing currents
lead not only to different TU morphologies but also to distinctly
different patterns of adaptation of QT interval duration to exercise
and other autonomic stimuli (27, 35). Furthermore, it has
been demonstrated that even phenotypically normal long-QT syndrome
gene carriers may exhibit characteristic changes in QT dynamicity
(16). Recent studies have suggested that the frequency of
asymptomatic long-QT syndrome gene carriers in the general population
is greater than previously considered (28). Whether
"normal" genetic variations or various silent gene mutations are
the cause of the diversity of QT-RR relationship patterns in the
population is not known. If the latter is the case, it remains to be
determined whether particular QT-RR patterns signify slight variations
in "repolarization reserve" (28) and differences in
the propensity to repolarization-related arrhythmias, such as those
induced by repolarization active drugs. The genetic background of the
stable QT-RR pattern is also more likely than other, e.g.,
environment-related determinants, because the participants of our study
were recorded during normal daily routines with variable levels of
mental and physical stress that are very unlikely to remain stable over
a period of 1 mo.
The slightly lower intrasubject stability in women compared with men
might have been contributed by menstrual cycle-related variability or
by the positional variability of ambulatory precordial electrodes in women.
Limitations of study. Our study has several important technical and physiological limitations. The major technical limitation is the reliance on automatic measurement of QT intervals without the possibility of systematic visual verification and manual editing of many thousands of ECG samples. Not only the type of automatic measurement algorithm (24) but even differences in the parameter setting within one algorithm can significantly influence the measurement results (2). Although we used an advanced strategy for improving the quality of automatic QT measurement, the principal problem of automatic measurement remains. The exclusion of 10-s ECG samples with significant variations of the heart rate most likely has significantly diminished, but not completely removed, the effect of QT hysteresis, because the changes in QT interval duration after abrupt changes in heart rate are completed only in ~3 min (17). However, these potential technical inaccuracies could only have caused our assessment to underestimate the degree of intrasubject stability of QT-RR patterns.
Instead of the maximum QT interval among all measurable leads, we used the median QT interval duration. In healthy individuals, a range of QT intervals measured in different ECG leads has been repeatedly reported around only 20-30 ms and attributed predominantly to measurement inaccuracies and differences in T wave loop projection (20). Moreover, the maximum QT interval is more likely influenced by measurement inaccuracies that are bound to be present with fully automatic measurement. Thus with the automatic QT interval measurement, the median QT interval is both relevant and a more stable representation of repolarization duration in healthy subjects. However, to ensure that no bias had been introduced, we also recomputed the whole study with the maximum QT interval of all measured leads and with the limit of 60 ms as a permissible threshold for the discrepancy between the different measurement algorithms. Although the QT-RR regression residua increased (there were obvious artificial outliers in the QT interval data), the principal results of the study remained the same (P < 104-109 for
the comparison of individual SD vs. the SD of individual means of
regression coefficients).
Although we observed that different QT-RR regression models optimally
fit the data of different subjects, our statistical comparisons were
based on fitting the data of all subjects with the same model. This is
because the numerical differences between coefficients of different
regression models are difficult to compare and the spread of the
individual QT-RR patterns would be difficult to quantify if the
different patterns were described with different models. This approach
might have led to underestimation of the differences between
intrasubject stability and intersubject variability because forcing
individual QT-RR data into a nonoptimum regression equation may
increase intrasubject variability. Instead of the battery of different
biparametric regression models, it might have been more appropriate to
develop a more flexible regression model with more parameters that
would have described not only the slope and intercept but also the
different properties of the optimum curvature fitting the QT-RR data.
However, simple attempts of this kind (e.g., QT = 
+ × RR, where is a regression parameter) are
easily influenced by atypical outliers in the data and lead to
physiologically unrealistic equations (e.g., > 1).
Our observations made in healthy young subjects may not necessarily be
applicable to the general population and/or cardiac patients. It
is not clear how much the individual QT-RR relationship might be
influenced by pathological (e.g., subclinical evolution of the
underlying heart disease) or pharmacological factors.
Finally, the protocol of our study did not include the investigation of
cellular electrophysiological mechanisms and/or collection of genetic
material that would permit us to link the differences in QT-RR patterns
to the subclinical differences in cellular electrophysiology and/or in
genes responsible for different repolarization channels. We also have
not collected data on the menstrual cycle in women and are therefore
unable to comment on the influence of menstrual cycle on the (small)
intrasubject variability of QT-RR patterns.
Implications of study. These results provide the physiological basis for a new approach for heart rate correction of the QT interval based on estimation of the individual QT-RR pattern. In studies requiring an increased precision, current strategies for rate correction of the QT interval should be reevaluated. The use of a fixed rate correction formula should probably be restricted only to cases when approximate estimation of the corrected QT interval is sufficient and within a narrow band of physiological heart rates. In investigations requiring precise heart rate correction of the QT interval, e.g., in studies assessing the effect of drugs on QT interval duration, the use of an individual rate correction formula derived in each patient from multiple QT-RR data in a baseline, drug-free state should be adopted as a standard approach. To derive this formula, the optimum QT-RR regression model should first be identified in each individual. If the design of the study permits, the investigation of QT interval changes (e.g., drug related) should include not only the assessment of the changes of heart rate-corrected QT interval but also the changes of QT-RR pattern.
The present study as well as our previous study (22) clearly showed that any standard "ad hoc" accepted rate correction formula can lead to substantial under- or overcorrection of the QT interval in a large number of individuals, even with relatively small changes of the heart rate (e.g., 70-90 beats/min).| |
ACKNOWLEDGEMENTS |
|---|
This work was supported in part by the Wellcome Trust, London, UK, and the British Heart Foundation, London, UK.
| |
FOOTNOTES |
|---|
Address for reprint requests and other correspondence: V. Batchvarov, Dept. of Cardiological Sciences, St. George's Hospital Medical School, Cranmer Terrace, London SW17 0RE, UK (E-mail: vbatchva{at}sghms.ac.uk).
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
First published February 21, 2002;10.1152/ajpheart.00860.2001
Received 3 October 2001; accepted in final form 19 February 2002.
| |
REFERENCES |
|---|
|
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
|
|---|
1.
Aytemir, K,
Maarouf N,
Gallagher MM,
Yap YG,
Waktare JEP,
and
Malik M.
Comparison of formulae for heart rate correction of QT interval in exercise electrocardiograms.
Pacing Clin Electrophysiol
22:
1397-1401,
1999[Medline].
2.
Batchvarov, V,
Yi G,
Guo XH,
Savelieva I,
Camm AJ,
and
Malik M.
QT interval and QT dispersion measured with the threshold method depend on threshold level.
Pacing Clin Electrophysiol
21:
2372-2375,
1998[Medline].
3.
Bazett, HC.
An analysis of the time-relations of electrocardiograms.
Heart
7:
353-370,
1920.
4.
Browne, KF,
Prystowsky E,
Heger JJ,
and
Zipes D.
Modulation of the Q-T interval by the autonomic nervous system.
Pacing Clin Electrophysiol
6:
1050-1056,
1983[Medline].
5.
Cappato, R,
Alboni P,
Pedroni P,
Gilli G,
and
Antonioli GE.
Sympathetic and vagal influences on rate-dependent changes of QT interval in healthy subjects.
Am J Cardiol
68:
1188-1193,
1991[ISI][Medline].
6.
Davey, P.
A new physiological method for heart rate correction of the QT interval.
Heart
82:
183-186,
1999
7.
De Bruyne, MC,
Hoes AW,
Kors JA,
Hofman A,
van Bemmel JH,
and
Grobbee DE.
Prolonged QT interval predicts cardiac and all-cause mortality in the elderly. The Rotterdam Study.
Eur Heart J
20:
278-284,
1999
8.
Franz, MR,
Swerdlow CD,
Liem LB,
and
Schaefer J.
Cycle length dependence of human action potential duration in vivo: effects of single extrastimuli, sudden sustained rate acceleration and deceleration, and different steady-state frequencies.
J Clin Invest
82:
972-979,
1988[ISI][Medline].
9.
Fridericia, LS.
Die Systolendauer in Elektrokardiogram bei normalen Menschen und bei Herzkranken.
Acta Med Scand
53:
469-486,
1920.
10.
Funck-Brentano, C,
and
Jaillon P.
Rate-corrected QT interval: techniques and limitations.
Am J Cardiol
72:
17B-22B,
1993[Medline].
11.
Gill, JS,
Baszko A,
Xia R,
Ward DE,
and
Camm AJ.
Dynamics of the QT interval in patients with exercise-induced ventricular tachycardia in normal and abnormal hearts.
Am Heart J
126:
1357-1363,
1993[ISI][Medline].
12.
Haverkamp, W,
Brethardt G,
Camm AJ,
Janse M,
Rosen MR,
Antzelevitch C,
Escande D,
Franz M,
Malik M,
Moss A,
and
Shah R.
The potential for QT prolongation and proarrhythmia by non-antiarrhythmic drugs: clinical and regulatory implications. Report on a Policy Conference of the European Society of Cardiology.
Eur Heart J
21:
1216-1231,
2000
13.
Hodges, M.
Rate correction of the QT interval.
Card Electrophysiol Rev
3:
360-363,
1997.
14.
Kawataki, M,
Kashima T,
Toda H,
and
Tanaka H.
Relation between QT interval and heart rate. applications and limitations of Bazett's formula.
J Electrocardiol
17:
371-375,
1984[ISI][Medline].
15.
Kova
, SJ.
The duration of the QT interval as a function of heart rate: a derivation based on physical principles and comparison to measured values.
Am Heart J
110:
872-878,
1985[ISI][Medline].
16.
Lande, G,
Kyndt F,
Baró I,
Chabannes D,
Boisseau P,
Pony J-C,
Escande D,
and
Le Marec H.
Dynamic analysis of the QT interval in long QT1 syndrome patients with a normal phenotype.
Eur Heart J
22:
410-422,
2001
17.
Lau, CP,
Freedman AR,
Fleming S,
Malik M,
Camm AJ,
and
Ward DE.
Hysteresis of the ventricular paced QT interval in response to abrupt changes in pacing rate.
Cardiovasc Res
22:
67-72,
1988[ISI][Medline].
18.
Lecocq, B,
Lecocq V,
and
Jaillon P.
Physiologic relation between cardiac cycle and QT duration in healthy volunteers.
Am J Cardiol
63:
481-486,
1989.
19.
Malik, M.
Problems of heart rate correction in assessment of drug-induced QT interval prolongation.
J Cardiovasc Electrophysiol
12:
411-420,
2001[ISI][Medline].
20.
Malik, M,
and
Batchvarov V.
Measurement, interpretation and clinical potential of QT dispersion.
J Am Coll Cardiol
36:
1749-1766,
2000
21.
Malik, M,
and
Camm AJ.
Evaluation of drug-induced QT interval prolongation.
Drug Saf
24:
323-351,
2001[ISI][Medline].
22.
Malik, M,
Färbom P,
Batchvarov V,
Hnatkova K,
and
Camm AJ.
The relationship between QT and RR intervals is highly individual among healthy subjects
Implications for heart rate correction of the QT interval.
Heart
87:
220-228,
2002
23.
Malik, M,
Kautzner J,
Hnatkova,
and
Camm AJ.
Identification of electrocardiographic patterns.
Pacing Clin Electrophysiol
19:
245-251,
1996[Medline].
24.
McLaughlin, NB,
Campbell RWF,
and
Murray A.
Comparison of automatic QT measurement techniques in the normal 12 lead electrocardiogram.
Br Heart J
74:
84-89,
1995
25.
Molnar, J,
Weiss J,
Zhang F,
and
Rosenthal JE.
Evaluation of five QT correction formulas using a software-assisted method of continuous QT measurement from 24-hour Holter recordings.
Am J Cardiol
78:
929-926,
1996.
26.
Okin, PM,
Devereux RB,
Howard BV,
Fabsitz RR,
Lee ET,
and
Welty TK.
Assessment of QT interval and QT dispersion for prediction of all-cause and cardiovascular mortality in American Indians. The Strong Heart Study.
Circulation
101:
61-66,
2000.
27.
Paavonen, KJ,
Swan H,
Piippo K,
Hokkanen L,
Laitinen P,
Viitasalo M,
Toivonen L,
and
Kontula K.
Response of the QT interval to mental and physical stress in type LQT1 and LQT2 of the long QT syndrome.
Heart
86:
39-44,
2001
28.
Priori, SG,
Napolitano C,
and
Schwartz PJ.
Low penetrance in the long-QT syndrome. Clinical impact.
Circulation
99:
529-533,
1999.
29.
Rautaharju, PM,
Warren JW,
and
Calhoun HP.
Estimation of QT prolongation. A persistent, avoidable error in computer electrocardiography.
J Electrocardiol
23, Suppl:
111-117,
1990.
30.
Rickards, AF,
and
Norman J.
Relation between QT interval and heart rate. New design of physiologically adaptive cardiac pacemaker.
Br Heart J
45:
56-61,
1981
31.
Roden, D.
Taking the "idio" out of "idiosyncratic": predicting torsades de pointes.
Pacing Clin Electrophysiol
21:
1029-1034,
1998[Medline].
32.
Sagie, A,
Larson MG,
Goldberg RJ,
Bengtson JR,
and
Levy D.
An improved method for adjusting the QT interval for heart rate (the Framingham Heart Study).
Am J Cardiol
70:
797-801,
1992[ISI][Medline].
33.
Schouten, EG,
Dekker JM,
Meppelink P,
Kok FJ,
Vandenbroucke JP,
and
Pool J.
QT interval prolongation predicts cardiovascular mortality in an apparently healthy population.
Circulation
84:
1516-1523,
1991.
34.
Schwartz, PJ,
and
Wolf S.
QT interval prolongation as predictor of sudden death in patients with myocardial infarction.
Circulation
57:
1074-1077,
1978.
35.
Swan, H,
Viitasalo M,
Piippo K,
Laitinen P,
Kontula K,
and
Toivonen L.
Sinus node function and ventricular repolarization during exercise stress test in long QT syndrome patients with KvLQT1 and HERG potassium channel defects.
J Am Coll Cardiol
34:
823-829,
1999
36.
Tavernier, R,
Jordaens L,
Haerynck F,
Derycke E,
and
Clement DL.
Changes in the QT interval and its adaptation to rate, assessed with continuous electrocardiographic recordings in patients with ventricular fibrillation, as compared to normal individuals without arrhythmias.
Eur Heart J
18:
994-999,
1997
37.
Xue, Q,
and
Reddy S.
New algorithms for QT dispersion analysis.
Comput Cardiol
23:
293-296,
1996.
This article has been cited by other articles:
|
|
 |
|
  K. E. Odening, O. Hyder, L. Chaves, L. Schofield, M. Brunner, M. Kirk, M. Zehender, X. Peng, and G. Koren Pharmacogenomics of anesthetic drugs in transgenic LQT1 and LQT2 rabbits reveal genotype-specific differential effects on cardiac repolarization Am J Physiol Heart Circ Physiol, December 1, 2008; 295(6): H2264 - H2272. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Malik, K. Hnatkova, T. Novotny, and G. Schmidt Subject-specific profiles of QT/RR hysteresis Am J Physiol Heart Circ Physiol, December 1, 2008; 295(6): H2356 - H2363. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Malik, K. Hnatkova, J. Ford, and D. Madge Near-Thorough QT Study as Part of a First-In-Man Study J. Clin. Pharmacol., October 1, 2008; 48(10): 1146 - 1157. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. J. Goldberger, M. E. Cain, S. H. Hohnloser, A. H. Kadish, B. P. Knight, M. S. Lauer, B. J. Maron, R. L. Page, R. S. Passman, D. Siscovick, et al. American Heart Association/American College of Cardiology Foundation/Heart Rhythm Society Scientific Statement on Noninvasive Risk Stratification Techniques for Identifying Patients at Risk for Sudden Cardiac Death: A Scientific Statement From the American Heart Association Council on Clinical Cardiology Committee on Electrocardiography and Arrhythmias and Council on Epidemiology and Prevention J. Am. Coll. Cardiol., September 30, 2008; 52(14): 1179 - 1199. [Full Text] [PDF]
|
|
|
|
 |
|
 J. J. Goldberger, M. E. Cain, S. H. Hohnloser, A. H. Kadish, B. P. Knight, M. S. Lauer, B. J. Maron, R. L. Page, R. S. Passman, D. Siscovick, et al. American Heart Association/American College of Cardiology Foundation/Heart Rhythm Society Scientific Statement on Noninvasive Risk Stratification Techniques for Identifying Patients at Risk for Sudden Cardiac Death: A Scientific Statement From the American Heart Association Council on Clinical Cardiology Committee on Electrocardiography and Arrhythmias and Council on Epidemiology and Prevention Circulation, September 30, 2008; 118(14): 1497 - 1518. [Full Text] [PDF]
|
|
|
|
|
|
S. Sundaram, M. Carnethon, K. Polito, A. H. Kadish, and J. J. Goldberger Autonomic effects on QT-RR interval dynamics after exercise Am J Physiol Heart Circ Physiol, January 1, 2008; 294(1): H490 - H497. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Chan, G.K. Isbister, C.M.J. Kirkpatrick, and S.B. Dufful Drug-induced QT prolongation and torsades de pointes: evaluation of a QT nomogram QJM, October 1, 2007; 100(10): 609 - 615. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. R. Kowey and M. Malik The QT interval as it relates to the safety of non-cardiac drugs Eur. Heart J. Suppl., September 1, 2007; 9(suppl_G): G3 - G8. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Robert, M. Salva, R. Segarra, M. Pavesi, R. Esbri, D. Roberts, and G. Golor The Prokinetic Cinitapride Has No Clinically Relevant Pharmacokinetic Interaction and Effect on QT during Coadministration with Ketoconazole Drug Metab. Dispos., July 1, 2007; 35(7): 1149 - 1156. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. Ballo, D. Bernabo, and S. A. Faraguti Abnormal QT responses to adenosine in subjects with long-QT syndrome Eur. Heart J., July 2, 2006; 27(14): 1759 - 1760. [Full Text] [PDF]
|
|
|
|
 |
|
 P. Smetana, V. Batchvarov, K. Hnatkova, A John Camm, and M. Malik Sex differences in the rate dependence of the T wave descending limb Cardiovasc Res, June 1, 2003; 58(3): 549 - 554. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Visit Other APS Journals Online |
) and 2 men
(bottom 2 rows; 
). Each row presents 4 recordings of 1 subject, and the 4 columns present the 1st and
subsequent recordings of all subjects, respectively. Note the stability
of the QT-RR pattern in each subject and the variability of the
patterns in different subjects. Note also that at the same RR interval,
the QT interval differed substantially in different
electrocardiographic samples. However, in this aspect, the figure is
potentially misleading because it shows only the range of all QT
intervals measured at the same heart rate rather than their
distribution. A detailed analysis (not presented in this article)
showed that on average, the 90th percentile spread of QT intervals
recorded at the same heart rate varied <20 ms in both women and men.
