Jce_40580.tex

Effects of Psychologic Stress on Repolarization and Relationship
to Autonomic and Hemodynamic Factors
RACHEL LAMPERT, M.D.,∗ VLADIMIR SHUSTERMAN, M.D., PH.D.,† MATTHEW M. BURG, PH.D.,∗‡ FORRESTER A. LEE, M.D.,∗ CHRISTINE EARLEY, M.S.,∗ ANNA GOLDBERG, B.S.,† CRAIG A. MCPHERSON, M.D.,∗ WILLIAM P. BATSFORD, M.D.,∗ ∗Yale University School of Medicine, New Haven, Connecticut; †PinMed, Inc., and University of Pittsburgh, Cardiovascular Institute, Pittsburg, Pennsylvania; ‡VA Connecticut Healthcare System, New Haven, Connecticut, USA Psychological Stress and Repolarization. Introduction: Psychological stress can precipitate ventric-
ular arrhythmias in patients with ICDs, as well as sudden death. However, the physiologic pathways remain
unknown. We sought to determine whether psychological stress induced in the laboratory setting alters
indices of repolarization associated with arrhythmogenesis.
Methods and Results: Patients with ICDs and a history of ventricular arrhythmia underwent ambula-
tory ECG monitoring during a laboratory mental stress protocol (anger recall and mental arithmetic).
Continuous changes in repolarization indices which have correlated with temporal and spatial myocardial
heterogeneity of repolarization, including T-wave alternans (TWA), T-wave amplitude (Tamp), and T-wave
area (Tarea) were analyzed in the time domain. In the 33 patients (85% male, 88% with coronary artery
disease, mean ejection fraction 30%), norepinephrine, epinephrine, BP, and HR increased during mental
stress. TWA increased from 22 (interquartile range 16–27) at baseline to 29 (21–38) uV during mental stress
(P < 0.001). Changes in TWA correlated with changes in HR, systolic BP, and catecholamines. Tamp and
Tarea also increased with mental stress (P < 0.01) but did not correlate with changes in other variables.
Conclusion: Psychological stress increased TWA, Tamp, and Tarea. Autonomically mediated repolar-
ization changes may be a pathophysiologic link between emotion and arrhythmia in susceptible patients.
(J Cardiovasc Electrophysiol, Vol. 16, pp. 1-6, April 2005)
stress, catecholamines, repolarization, ventricular arrhythmia, T-wave alternans Introduction
circadian variation, with highest values in the morning,8 whencatecholamine levels also peak.10 Further, beta-blockade can Psychological stress increases sudden cardiac death in decrease TWA.11 Eliciting an anger-like state in dogs in- populations during emotionally devastating disasters such creases TWA.12 One previous study has demonstrated that as earthquake or war,1,2 alters induced arrhythmias,3,4 and psychological stressors can increase TWA,13 but evaluated precipitates spontaneous ventricular arrhythmias in patients only limited potentially mediating factors.
with implantable cardioverter defibrillators (ICDs).5 How- To further investigate whether and how psychological ever, the physiologic pathways through which stress can trig- stressors may alter indices which reflect temporal and spatial ger arrhythmia remain poorly understood. Stress may alter heterogeneity of repolarization, we compared TWA, as well electrophysiological properties of the myocardium, through as T-wave amplitude (Tamp) and T-wave area (Tarea)14-17 the actions of stress hormones or via efferents descending at rest and during laboratory-induced psychological stress from the CNS. One electrophysiological property necessary in a group of subjects likely to manifest repolarization het- for ventricular arrhythmogenesis is non-uniform recovery of erogeneity, patients with ICDs and a history of ventricular ventricular excitability.6 Indices reflecting temporal and spa- arrhythmia, and evaluated these changes in relationship to tial heterogeneity of repolarization, such as T-wave alternans changes in autonomic and hemodynamic factors.
(TWA) predict vulnerability to sudden death and ventricu-lar arrhythmia.7-9 Indirect evidence suggests that autonomic fluctuations can alter TWA. For example, TWA demonstrates Patient Population
Dr. Lampert received grant from American Heart Association, Scientist Patients were included with ischemic or dilated cardiomy- Development Grant no. 0030190N; Dr. Shusterman from AHA, SDG opathy and a history of spontaneous or induced ventricular 0030248N and NIH, 1R43HL077116-01; Dr. Soufer from NIH, RO1 no.
arrhythmias. Patients with atrial fibrillation, diagnosed psy- HL59619-01 and HL071116-01, catecholamine analysis: Yale General Clin-ical Research Center, NIH no. M01 RR00125.
chiatric disorders, or severe comorbidities were excluded.
Thirty-three patients receiving ICDs between 12/00 and Address for correspondence: Rachel Lampert, M.D., Yale School of 12/02 agreed to participate and provided written informed Medicine, Section of Cardiology, 333 Cedar Street, FMP 3; New consent. The study was approved by the Yale Human Inves- Haven, CT 06520. Tel.: 203-737-4068; Fax: 203-737-2437; E-mail:rachel.lampert@yale.edu Manuscript received 16 August 2004; Revised manuscript received 4 Study Design
October 2004; Accepted for publication 21 October 2004.
Patients underwent laboratory mental stress testing in the morning immediately prior to scheduled non-invasive ICD Journal of Cardiovascular Electrophysiology
testing, 3 months after implantation (standard of care at our Analysis of TWA was methodologically similar to the institution), and prior to administration of sedatives or anal- modified moving average approach introduced by Nearing gesics. An intravenous cannula was placed in an antecubital and Verrier26 but computationally simpler for examining vein for blood sampling. Pacing was programmed to the VVI gross alternations in the T-wave, at the expense of a lower sen- mode at 40 bpm, which allowed emergence of sinus rhythm sitivity to small-amplitude TWA.24 This automated computer with native AV conduction in all patients. Continuous 12-lead algorithm consisted of (1) detection of the fiducial points (T- ECG was monitored throughout the protocol (GE CardioLab onset, T-peak, and T-end), (2) calculation of the mean am- plitude of the corresponding segment of the T-wave (fromT-onset to the T-peak (Tonset-peak); from the T-peak to the Mental Stress Protocol
T-end (Tpeak-end); and from the T-onset to the T-end for “to-tal” (TWA); (3) calculation of the time series of the differ- In accordance with our standard mental stress protocol,4 ences between the corresponding amplitudes under consecu- the room lights were dimmed and quiet maintained. In the tive even and the odd beats; (4) averaging of these time series baseline period, starting at least 30 minutes after IV place- over 5-minute intervals; and (5) calculation of the results ment, patients were encouraged to think about past relax- for each test as a percent change compared to the individual ing situations. This was followed by the two mental stress baseline values. This algorithm was validated on simulated tasks, first arithmetic and then anger recall, interspersed with signals with various levels of TWA in the presence of random a second baseline period. Each phase lasted 5–7 minutes. For noise, spurious artifacts, and phase-shifts using the method- the mental arithmetic task, subjects were asked to subtract ology described by Nearing and Verrier.26 The algorithm ac- 7 serially from a 3-digit number rapidly and accurately. For curately detected changes in the level of TWA both in clean anger recall, patients described a recent event eliciting irri- signals and in signals contaminated by noise and artifacts,27 tation, annoyance, or frank anger, with the insertion of fre- with performance similar to that of modified moving average quent irritating questions by the interviewer. Two stressors were utilized because this approach provides greater gener-alizability to naturalistic settings.18 Consistent with previous Catecholamine Collection and Analysis
investigations, the maximum response to the two stressorswas analyzed.4,19 Blood for catecholamine assay was withdrawn continu- ously by exfusion pump (Dakmed, Buffalo, NY) at a rate Repolarization Analysis
of 1 ml/min. Samples were immediately placed on iceand brought to the Yale General Clinical Research Cen- Ambulatory ECGs (Holters) with modified V1 and V5 ter within 30 minutes where they were spun and stored leads were recorded on GE Medical (Milwaukee, WI) Mar- at −70◦C. Levels of epinephrine (EPI) and norepinephrine quette Series 8500 direct (amplitude-modulated) recorders.
(NE) were determined by high-performance liquid chro- These recorders have a flat frequency response and a lin- matography (ESA, Inc., Chelmsford, MA) using electro- ear phase between 0.67 and 50 Hz (±3 dB).20,21 This range chemical detection (Coulochem II) after alumina extraction.
is somewhat narrower than 0.3–50 Hz recommended by Samples from each patient were run in the same batch in Nearing et al. for TWA monitoring.22 However, since the low- duplicate. Four patients were not included in the analysis of est frequency components of the cardiac complexes in more catecholamines due to the inability to gain or maintain ade- than 99% of adults, 99% of the time are greater than 0.67 Hz, most ambulatory recorders (including Marquette 8500)have a 0.67 Hz (3 dB) cutoff frequency of the high-pass filter Heart Rate Variability Analysis
for baseline correction.23 The frequency of the cardiac com-plexes can be lower than 0.67 Hz at very slow heart rates (<40 After editing as above, an annotated list of R-R intervals bpm), which were not observed in our study. The recordings was analyzed using customized software. The R-R interval were digitized at 400 Hz sampling frequency and effective data were edited to remove ectopic beats and noise, and gaps were filled in by interpolated linear splines.28 Holter record- Holter recordings were digitized at 400 Hz using a com- ings with >20% interpolated R-R intervals were excluded mercial scanning system (Burdick, Inc., Syracuse, NY). A from further analysis (N = 2). The R-R interval time series single lead (that with larger magnitude T-wave) was analyzed was sampled using a boxcar window29 to obtain 1024 sam- for each repolarization index, the same lead for all stages.
ples per 5 minutes (3.41333 Hz). The power spectrum was A previously validated program for adaptive baseline cor- computed using a fast Fourier transform with a Parzen win- rection24 was then applied to assure accurate detection of the dow on 4-minute segments with a 1-minute sliding window, isoelectrical line with a minimal distortion of the repolariza- corrected for attenuation due to windowing and sampling30 tion waveforms. The QRS complexes were classified using and integrated over five standard frequency bands.31 High- custom software and verified by an experienced technician.
frequency power (HF, 0.15–0.40 Hz), a marker of parasym- After exclusion of ectopy, series of consecutive sinus beats pathetic activity,32,33 was compared between experimental were processed to identify fiducial points, including the on- set of the Q-wave, the end of the S-wave, and the beginning, Statistical Analysis
peak, and end of the T-wave as described elsewhere.24 Be-cause the changes in repolarization are complex and highly Changes in TWA, Tamp, and Tarea and in catecholamines, variable among individuals, and because no single parameter from baseline to mental stress showed a highly skewed distri- on the surface ECG can reliably represent the entire spectrum bution (Shapiro–Wilk W < 0.05) and hence were analyzed by of repolarization changes, we used a set of several descriptors, the paired non-parametric Wilcoxon signed-rank test. Other including Tamp, Tarea, and TWA, as previously described.25 variables were compared using paired t-test. Relationships Psychological Stress and Repolarization
Autonomic and Hemodynamic Changes with Mental Stress Baseline
Mental Stress
during anger by 78% (50–140%, P < 0.001). Responses to arithmetic and anger did not differ significantly. About half of subjects showed greater response to anger (53–59% for Tamp, Tarea, and TWA) and half to arithmetic.
Median increase in Tamp was 9% (1–26%, P = 0.001), and ∗With VT induced at electrophysiology study.
in Tarea, 11% (0–24%, P < 0.001). QT interval decreased.
†CAD, EF < 35%, inducible VT ‡including sotalol.
Relationship of Repolarization Changes to Autonomic and
Hemodynamic Changes
between mental-stress induced changes in repolarization in- The rise in TWA during mental stress was significantly dices, autonomic factors, and heart rate were evaluated using correlated with increases in EPI, systolic BP, and HR Pearson product-moment correlation coefficients. Relation- (Table 4). In a multivariable model including EPI, BP, and ship between TWA change and factors with significant corre- HR, only EPI independently predicted increase in TWA lations was further evaluated with standard least-squares re- (P <0.01). The relationship between changes in cate- gression. In addition, because NE and EPI response to stress cholamines and in repolarization indices was further evalu- was highly skewed, patients were dichotomized into high and ated by comparing TWA changes between subjects with high low NE and EPI responders at the top quartile (top 7 of 29 pa- versus low-normal NE and EPI responses. While TWA in- tients). Mental-stress induced changes in repolarization were creased with stress in patients with both high and low-normal catecholamine response, those with greater rise in NE and inEPI showed significantly greater increases in TWA (Fig. 2).
Changes in Tamp and Tarea did not correlate with heart ratechange or with other variables.
Patient Population
Participants’ demographic and clinical characteristics are Discussion
shown in Table 1. Typical of most ICD populations, the ma- In patients with heart disease and a history of ventricu- jority of subjects were male and had coronary artery disease.
lar arrhythmia, laboratory-induced psychological stress in- The mean ejection fraction was 30%. Most were on beta- creased temporal and spatial heterogeneity of repolarization blockers and few on anti-arrhythmic medications.
as measured by TWA, Tamp, and Tarea in association with Autonomic, Hemodynamic, and Electrocardiographic
changes in catecholamines as well as in hemodynamic pa- Effects of Mental Stress
rameters. Thus, stress-induced sympathetic activation maybe one pathway through which stress increases heterogene- NE and EPI rose significantly from baseline to men- tal stress (Table 2), implying sympathetic activation, and TWA not only predicts both induced7,9 and sponta- HF power decreased, indicating parasympathetic withdrawal.
neous ventricular arrhythmia,7,8 but also immediately pre- The increase in heart rate was consistent, although of small cedes development of ventricular fibrillation in animal mod- magnitude. Systolic and diastolic BP rose during stress. No els,14,26,34,35 suggesting that TWA may be mechanistically ischemic changes in ST segment or T wave were seen on related to arrhythmia. In 10% of our patients, the stress- induced increase in TWA was similar to that previously asso-ciated with vulnerability to ventricular fibrillation.26 Tamp, Effects of Mental Stress on Repolarization
which also increased in response to stress, increases prior to All heterogeneity-related T-wave measurements, TWA, spontaneous sustained monomorphic VT,36 suggesting that Tamp, and Tarea, increased significantly from baseline to an increase in Tamp may also reflect an arrhythmogenic pro- mental stress (Fig. 1 and Table 3). Median increase in TWA cess. Thus, changes in repolarization may be one mechanism was 78% (interquartile range 43–107%, P < 0.001), and in through which psychological stress may trigger ventricular 10% of patients, TWA at least doubled with stress. In patients with CAD, median increase in TWA with stress was 79%, and Physiologic Correlations of Repolarization Measurements
in those without CAD, 66% (P > 0.5). TWA increased duringboth Tonset-peak and Tpeak-end (Table 3 and Fig. 1). TWA in- The importance of non-uniform recovery of ex- creased during arithmetic by 67% (42–146%, P < 0.001) and citability to developing ventricular fibrillation is well Journal of Cardiovascular Electrophysiology
Tonset-peak
Tpeak-end
Figure 1. Repolarization changes with mental stress. Box plots represent median and interquartile range. Tamp = T-wave amplitude; TWA = T-wave
alternans; Tarea = T-wave area.
recognized.6,14,26,34,35 However, the specific electrophysio- Potential Pathways Linking Psychological Stress and
logic perturbations responsible for repolarization heterogene- Repolarization Changes
ity are not well understood. In experimental models, hetero-geneity of repolarization associated with the development of The pathways through which stress alters repolarization arrhythmia has been demonstrated at multiple levels: first, be- are unknown. TWA increases with heart rate,7,14,41 as also tween myocardial regions epicardially,17 next, transmurally seen here. However, sympathetic activation may increase across the myocardium,15,16,37 and finally, between neighbor- TWA beyond the effects of heart rate. Experimentally, stellec- ing cells.14 Each of these mechanisms may be important in tomy abolishes, while stellate ganglion stimulation increases, TWA.34 In clinical studies, intravenous beta-blockade11 de- Each of the indices used here—Tarea, Tamp, and TWA— creased the magnitude of TWA, and TWA induced with ex- has been correlated with heterogeneity. For example,17 in ercise is greater than that with atrial pacing at the same heart multisite epicardial recordings, regional dispersion of recov- rate.41 In this study, HR increase was minimal, and EPI inde- ery times correlates with the width of the root-mean-square pendently predicted TWA changes, further supporting a role T-wave (similar to Tarea). Whether Tamp also correlates with of catecholamines in increasing heterogeneity of repolariza- epicardial dispersion is controversial.38-40 However, T-wave height does reflect transmural voltage differences.16 While Whether the parasympathetic nervous system influences TWA measures temporal changes in action potential duration repolarization is unclear. With atrial pacing to control heart at the level of single cells, cells differ in the timing and/or rate, while vagal stimulation reduced TWA induced by extent of these changes, creating spatial heterogeneity.14,15 coronary occlusion in one study,42 in another, atropine Transmural heterogeneity of repolarization, with epicardial did not alter TWA.11 In this study, while HF power de- cells displaying the shortest, and subendocardial M-cells the creased overall with mental stress, as expected,43 there was longest, action potential duration, correlates with TWA.15 no correlation between vagal withdrawal and changes in TWA also correlates with cell-to-cell heterogeneity.14 Effects of Mental Stress on Repolarization Baseline
Mental Stress
∗TWA = T-wave alternans; †Data expressed as median (interquartile range) (distribution highly skewed); ‡Tamp = T-wave amplitude; §Tarea = T-wave area.
Psychological Stress and Repolarization
Correlations of Changes in T-wave Alternans with Changes in Autonomic and Hemodynamic Factors in Response to Mental Stress Correlation Coefficient
TWA changes also correlated strongly with increases in systolic BP. It is possible that mental stress affected TWAin part through mechanoelectrical feedback, as changes inafterload can alter ventricular refractoriness.44 While ischemia invokes TWA in animals,45 as well as hu- mans during coronary angioplasty,45,46 in this study no ECG Figure 2. Increase in TWA with stress based on catecholamine response to
changes were seen, and a recent study13 showed no correla- stress. High NE, subjects whose NE (norepinephrine) response to stress fell tion between stress-induced TWA changes and ischemia on in the top quartile; High EPI (epinephrine) similarly. SPECT. This suggests that mental stress-induced TWA is notischemically mediated. Determining mechanisms throughwhich stress increases TWA remains an important avenue Tamp, and Tarea to a greater degree than demonstrated in this Tamp and Tarea also increased with stress, but changes To evaluate potential electrophysiological changes link- did not correlate with measured potential mediators. These ing stress to arrhythmia, patients having a history of arrhyth- indices may measure different electrophysiologic phenomena mia were studied, who were most likely to manifest repo- than does TWA. In normals, Tamp decreases with stress.47 larization abnormalities. Without a control group, this study Similarly, with sympathetic activation by upright tilt, Tamp cannot state definitively whether the relationship between increases in patients with heart disease but decreases in nor- stress, catecholamines, and TWA is abnormal. However, as development of TWA with exercise or pacing is clearlypathologic, stress-induced TWA most likely represents a Methodology
pathologic process also. Whether vulnerability to mental- While time-domain methods to quantify TWA are less fre- stress-induced changes in repolarization in the laboratory is quently used than spectral, the clinical predictive value does associated with vulnerability to ventricular arrhythmia is an not differ,8,9 and time domain analysis may provide signif- important avenue of future research.
icant advantages.26 In clinical ECG recordings, TWA esti-mation has been limited by signal non-stationarity and the Conclusions
presence of noise and artifacts,26 rendering spectral meth-ods inaccurate. However, the differences between consecu- Psychological stress induced in a laboratory setting in- tive even and odd beats calculated in the time domain provide creased indices associated with temporal instability and spa- a more accurate dynamic estimate of the TWA changes over tial heterogeneity of repolarization including TWA, Tamp, time.26 Methodology used here further enhances the speci- and Tarea in individuals with heart disease and a history ficity of TWA in settings of noise.27 First, serial analysis of of arrhythmia. Stress-induced changes in TWA were asso- changes in the averaged amplitude of the T-wave35 (or its seg- ciated with sympathetic activation. These findings suggest ment) provides an averaged (over the corresponding segment) that autonomically mediated repolarization changes may be estimate of the changes in the T-wave energy.24 While this one pathophysiologic link between emotion and arrhythmia.
method is not as sensitive to subtle TWA as spectral methodsor modified moving average analysis,26 the use of averaging References
renders this method less sensitive to spurious spikes or arti-facts, which often contaminate clinical ECG recordings.26 To 1. Leor J, Poole WK, Kloner RA: Sudden cardiac death triggered by an earthquake. N Engl J Med 1996;334:413-419.
further decrease noise, patients were supine, and recordings 2. Meisel SR, Kutz I, Dayan KI, Pauzner H, Chetboun I, Arbel Y, Daviv D: were carefully corrected for artifacts and baseline wander.24 Effect of Iraqi missile war on incidence of acute myocardial infarction Finally, each individual served as his own control, and a per- and sudden death in Israeli civilians. Lancet 1991;338:660-661.
cent change in TWA calculated. The use of normalized units 3. Kirby DA, Pinto JMB, Hottinger S, Johnson DA, Lown B: Behavioral arousal enhances inducibility and rate of ventricular tachycardia. Am J (percent change) is relatively insensitive to inter-individual variations in spatial electrophysiologic properties.35 4. Lampert R, Jain D, Burg MM, Batsford WP, McPherson CA: Destabi- lizing effects of mental stress on ventricular arrhythmias in patients with Limitations
implantable cardioverter-defibrillators. Circulation 2000;101:158-164.
5. Lampert R, Joska T, Burg M, Batsford W, McPherson C, Jain D: Emo- Most patients were taking beta-blocking medications, tional and physical precipitants of ventricular arrhythmia. Circulation2002;106:1800-1805.
which decreased the magnitude of TWA in pharmacologi- 6. Han J, Moe GK: Nonuniform recovery of excitability in ventricular cal studies.11 It is thus likely that mental stress alters TWA, Journal of Cardiovascular Electrophysiology
7. Rosenbaum DS, Jackson LE, Smith JM, Garan H, Ruskin JN, Cohen 27. Shusterman V, Goldberg A: Tracking repolarization dynamics in real- RJ: Electrical alternans and vunerability to ventricular arrhythmias. N life data. J Electrocardiol 2004;37:180-186.
28. Albrecht P, Cohen RJ: Estimation of heart rate power spectrum bands 8. Verrier RL, Nearing BD, La Rovere M, Pinna GD, Mittleman MA, from real-world data: Dealing with ectopic beats and noisy data. Comput Bigger JT, Schwartz PJ: Ambulatory electrocardiogram-based track- ing of T wave alternans in postmyocardial infarction patients to assess 29. Berger RD, Akselrod S, Gordon D, Cohen RJ: An efficient algorithm risk of cardiac arrest or arrythmic death. J Cardiovasc Electrophysiol for spectral analysis of heart rate variability. IEEE T Bio-Med Eng 9. Gold MR, Bloomfield DM, Anderson KP, El-Sherif NE, Wilber DJ, 30. Hamming R: Numerical Methods for Scientists and Engineers. 2nd ed.
Groh WJ, Estes NA III, Kaufman ES, Greenberg M, Rosenbaum D: A comparision of T-wave alternans, signal averaged electrocardiography 31. Bigger JT, Fleiss JL, Steinman RC, Rolnitsky LM, Kleiger RE, Rottman and programmed ventricular stimulation for arrhythmia risk stratifica- JN: Frequency domain measures of heart period variability and mortality tion. J Am Coll Cardiol 2000;36:2247-2253.
after myocardial infarction. Circulation 1992;85:164-171.
10. Turton MB, Deegan T: Circadian variations of plasma catecholamine, 32. Akselrod S, Gordon D, Ubel FA, Shannon DC, Barger AC, Cohen cortisol, and immunoreactive insulin concentrations in supine subjects.
RJ: Power spectrum analysis of heart rate fluctuation: A quantita- tive probe of beat-to-beat cardiovascular control. Science 1981;213: 11. Rashba EJ, Cooklin M, MacMurdy K, Kavesh N, Kirk M, Sarang S, Peters RW, Shorofsky SR, Gold M: Effects of selective autonomic 33. Pagani M, Lombardi F, Guzzetti S, Rimoldi O, Furlan R, Pizzinelli P, blockade on T-wave alternans in humans. Circulation 2002;105:837- Sandrone G, Malfatto G, Dell’Orto S, Piccalluga E, Turiel M, Basellli G, Cerutti S, Malliani A: Power spectral analysis of heart rate and arterial 12. Kovach JA, Nearing BD, Verrier RL: Angerlike behavioral state poten- pressure variability as a marker of sympatho-vagal interaction in man tiates myocardial ischemia-induced T-wave alternans in canines. J Am and conscious dog. Circ Res 1986;59:178-193.
34. Nearing BD, Huang AH, Verrier RL: Dynamic tracking of car- 13. Kop W, Krantz D, Nearing B, Gottdiener J, Quigley J, O’Callahan M, diac vulnerability by complex demodulation of the T wave. Science DelNegro A, Friehling T, Karasik P, Suchday S, Levine J, Verrier R: Ef- fects of acute mental stress and exercise on T-wave alternans in patients 35. Adam DR, Smith JM, Akselrod S, Nyberg S, Powell AO, Cohen RJ: with implantable cardioverter defibrillators and controls. Circulation Fluctuations in T-wave morphology and susceptibility to ventricular fibrillation. J Electrocardiol 1984;17:209-218.
14. Pastore JM, Girouard SD, Laurita KR, Akar FG, Rosenbaum DS: Mech- 36. Anderson KP, Shusterman V, Beigel A, Aysin B, Brode S, Gottipaty anism linking T-wave alternans to the genesis of cardiac fibrillation.
VK, Weiss R, for the ESVEM Investigators: Changes in ventricular repolarization preceding the onset of spontaneous sustained ventricular 15. Shimizu W, Antzelevitch C: Cellular and ionic basis for t-wave alternans tachycardia (Abstract). PACE 1999;22:837.
under long-QT conditions. Circulation 1999;99:1499-1507.
37. Akar FG, Rosenbaum DS: Transmural electrophysiological hetero- 16. Yan GX, Antzelevitch C: Cellular basis for the normal T wave and the geneities underlying arrhythmogenesis in heart failure. Circ Res electrocardiographic manifestations of the long-qt syndrome. Circula- 38. di Bernardo D, Murray A: Medical physics: Explaining the T-wave shape 17. Fuller MS, Sandor G, Punske B, Taccardi B, MacLeod RS, Ershler PE, Green LS, Lux RL: Estimates of repolarization dispersion from 39. Russell D, Dart A: T wave amplitude as a quantitative index of re- electrocardiographic measurements. Circulation 2000;102:685-691.
gional myocardial sympathetic responsiveness. J Cardiovasc Pharm 18. Kamarck TW, Debski TT, Manuck SB: Enhancing the laboratory-to-life generalizability of cardiovascular reactivit;y using multiple occasions 40. Nearing B, Verrier R: Tracking cardiac electrical instability by com- of measurement. Psychophysiology 2000;37:533-542.
puting interlead heterogeneity of T-wave morphology. J Appl Physiol 19. Krantz DS, Santiago HT, Kop WJ, Merz CNB, Rozanski A, Gottdiener JS: Prognostic value of mental stress testing in coronary artery disease.
41. Hohnloser SH, Klingenheben T, Zabel M, Li Yg, Albrecht P, Cohen RJ: T wave alternans during exercise and atrial pacing in humans. J 20. Schluter P: Magnetic tape recording and playback for ST-segment anal- Cardiovasc Electrophysiol 1997;8:987-993.
ysis. J Electrocardiol 1988;21(Supp I):S20-S26.
42. Verrier RL, Nearing BD: Electrophysiologic basis for T wave alternans 21. Shook TL, Balke CW, Kotilainen PW, Hubelbank M, Selwyn AP, as an index of vulnerability to ventricular fibrillation. J Cardiovasc Elec- Stone PH: Comparison of amplitude-modulated (direct) and frequency modulated ambulatory techniques for recording ischemic electrocardio- 43. Pagani M, Mazzuero G, Ferrare A, Liberati D, Cerutti S, Vaitl D, Tavazzi graphic changes. Am J Cardiol 1987;60:895-900.
L, Malliani A: Sympathovagal interaction during mental stress: A study 22. Nearing BD, Stone PH, Verrier RL: Frequency response characteris- using spectral analysis of heart rate variability in healthy control subjects tics required for detection of T-wave alternans during ambulatory ECG and patients with a prior myocardial infarction. Circulation 1991;83:II- 23. Bailey JJ, Berson AS, Garson A Jr, Horan LG, Macfarlane PW, Mortara 44. Coulshed DS, Cowan JC, Drinkhill MJ, Hainsworth R: The ef- DW, Zywietz C: Recommendations for standardization and specifica- fects of ventricular end-diastolic and systolic pressures on action po- tions in automated electrocardiography: Bandwidth and digital signal tential and duration in anaesthetized dogs. J Physiol 1992;457:75- processing. Circulation 1990;81:730-739.
24. Shusterman V, Shah SI, Beigel A, Anderson KP: Enhancing the pre- 45. Nearing B, Oesterle S, Verrier R: Quantification of ischaemia induced cision of ECG baseline correction: Selective filtering and removal of vulnerability by precordial T wave alternans analysis in dogs and hu- residual error. Comp Biomed Res 2000;33:144-160.
mans. Cardiovasc Res 1994;28:1440-1449.
25. Shusterman V, Beigel A, Shah SI, Aysin B, Weiss R, Gottiparty VK, 46. Mart´ınez J, Olmos S, Laguna P: T wave alternans and acute is- Schwartzman D, Anderson KP: Changes in autonomic activity and ven- chemia in patients undergoing angioplasty. Comp Cardiol 2002; tricular repolarization. J Electrocardiol 1999;32:185-192.
26. Nearing BD, Verrier RL: Modified moving average analysis of T-wave 47. Furedy JJ, Szabo A, Peronnet F: Effects of psychological and physi- alternans to predict venricular fibrillation with high accuracy. J Appl ological challenges on heart rate, T-wave amplitude, and pulse-transit time. Int J Psychophysiol 1996;22:173-183.

Source: http://www.pinmed.net/solutions/JCE4_2005.pdf