Characteristics of QT Interval and QT Dispersion in Exercise Electrocardiogram: Healthy Persons versus Stable Angina Patients

Transcription

1 ORIGINAL ARTICLE Korean Circ J 2007;37: Print ISSN / On-line ISSN Copyright c 2007 The Korean Society of Cardiology Characteristics of QT Interval and QT Dispersion in Exercise Electrocardiogram: Healthy Persons versus Stable Angina Patients Jin Hong Park, MD, Dae Young Kim, MD, Bong Soo Kim, MD, Hyun Jik Lee, MD, Jae Hoon Kim, MD, Hee Sang Jang, MD, Seung Min Shin, MD, Hyun Jae Kang, MD, Bong Ryeol Lee, MD and Byung Chun Jung, MD Department of Cardiology, Fatima General Hospital, Daegu, Korea ABSTRACT Background and Objectives: The QT interval (QTi) and QT dispersion (QTd), which represent the myocardial electrical heterogeneity of repolarization, were studied to recognize the differences between normal controls (n=32) and stable angina patients (n=78). Subjects and Methods: During the treadmill exercise test, standard 12 lead Electrocardiogram (ECG) was obtained at every stage, with the QTi and QTd measured. The corrected QT interval (cqti) and QT dispersion (cqtd) were calculated using Bazett s formula, with the Delta QT interval (ΔQTi) measured on leads V5 and avf. Results: During exercise, the QTi had a reverse relationship with the heart rate in both groups, but was decreased by a lesser extent in the patient group. The QTd also had a tendency to decrease according to increasing heart rate in both groups and was significantly greater in the patient group. The corrected QTi increased during exercise in both groups, and reached maximum during the pre-peak stage, but was minimized during a 1 minute recovery stage in the patient group. The corrected QTd reached a maximum during the peak exercise stage in both groups, but the values between the two groups were significantly different. Both the QTi and cqti had tendencies to increase according to the number of vessels with stenosis. The ΔQTi tended to reflect a regional ischemia in a single vessel disease. Conclusion: The QTi, QTd, cqti and cqtd were increased in the stable angina patients compared with the normal controls, and augmented during the exercise test. (Korean Circ J 2007;37: ) KEY WORD: Exercise test. Introduction It is well known that the ventricular repolarization process is preferentially affected by ischemia, which is represented by ST segment changes or T wave inversions. The underlying cellular mechanism is known to be related with the shortening of action potential duration in ischemic myocardium. 1) QT dispersion (QTd), defined as the difference between the longest and shortest QT intervals (QTi s) measured on standard 12 lead Electrocardiogram (ECG), has been proposed as an index of the spatial dispersion during myocardial repolarization, and the prolongation of the QTd is believed to be linked to increased Received: May 14, 2007 Revision Received: July 5, 2007 Accepted: July 18, 2007 Correspondence: Byung Chun Jung, MD, Department of Cardiology, Fatima General Hospital, Sinam-dong, Dong-gu, Daegu , Korea Tel: , 7459, Fax: Augustjbc@yahoo.co.kr heterogeneity of ventricular repolarization in ischemic heart disease. The QTd is clinically considered as a predictor of ventricular arrhythmias, sudden cardiac deaths and other adverse events in various cardiac diseases. 2)3) Many previous studies have also reported that an increased QTd in ischemic heart disease was generally ameliorated after reperfusion using thrombolytic therapy or percutaneous coronary intervention; 4)5) thus, the change in the QTd is believed to reflect the degree of myocardial ischemia. The principle purposes of the present study were to assess the differences in the time course changes of the QTi and QTd during an exercise test by comparing normal individuals and stable angina patients, and evaluate whether the QTi and QTd show any differences according to the location and severity of coronary artery disease. Additionally, the corrected QTi and QTd values by heart rate were estimated in terms of the relationship between the QTi and QTd. The ability of the delta QTi to reflect the location of regional ischemia was also tested. 543

2 544 QT Dispersion in Stable Angina Subjects and Methods Subjects One hundred and ten subjects, who visited our hospital and underwent exercise myocardial SPECT between May 2003 and July 2006, were enrolled in the study. The control group consisted of 32 subjects, who showed negative results on both exercise testing and myocardial SPECT. The patient group consisted of 78 subjects, who had no anginal symptoms at rest and showed positive exercise test results. The patient group also showed reversible, not fixed perfusion abnormalities on myocardial SPECT and significant coronary artery stenoses on coronary angiography. Significant coronary artery stenosis was defined as more than 70% stenosis of at least one of any of the three coronary vessels. Subjects with atrial fibrillation or bundle branch block on surface ECG were excluded from the study group. Also, the patients who had past histories of coronary angioplasty or CABG were excluded. The patient group was divided into three subgroups according to the number of vessels with stenosis, and the single vessel disease group was further classified into three groups according to the location of stenosis. Methods An exercise stress test was carried out according to modified Bruce s protocol, using a CASE 8000 Marquette treadmill. The standard 12 lead ECG was recorded at a paper speed of 25 mm/s and with a sensitivity of 10 mm/mv. Resting ECG, blood pressure and heart rate were recorded in the supine and standing positions prior to exercise. During exercise, ECG recordings were taken during the pre-peak and peak exercise stages. Also, the peak blood pressure and heart rate were recorded. During the recovery stage, ECGs were recorded at 1 and 3 minutes after completing the exercise. The QTi was measured in as many of the 12 leads as possible on the ECG recorded at every stage, and was measured manually from the first deflection of the QRS complex to the end of the T wave, where T wave returns to the TP baseline. When the T wave was fused with the U wave, a straight line was drawn tangential to the down stroke of the T wave, and the point where the extrapolated line intersected the isoelectric line was taken as the end point of the T wave. If the end of the T wave could not be reliably determined, had very low amplitude or the QTi could not be measured due to artifacts, these leads were excluded from analysis. Also, ECGs with fewer than 6 leads available for analysis were excluded from the study. The QTd was defined as the difference between the maximal and minimal QTi (max QTi and min QTi, respectively). The QTi was corrected for the heart rate using Bazett s formula, 6) with the corrected QT dispersion (cqtd) calculated in the same way as the QTd. The delta QT interval (ΔQTi) was defined as the difference between the peak exercise and the resting QTi. The ΔQTi of lead V5 of the precordial leads and lead avf of the limb leads were measured and compared according to the location of stenosis in the single vessel disease group. Statistical analysis Data are expressed as the mean values±sd (standard deviation). All measurements in the control and patient groups were compared using an unpaired, two tailed Student s t-test. Data between subgroups in the patient group were compared with each other using a one-way ANOVA. A p of less than 0.05 was considered statistically significant. Results Characteristics and exercise test results The baseline characteristics and treadmill exercise test data of the control and patient groups (including sub- Table 1. Baseline characteristics and exercise test results One vessel disease Control Patient 1VD 2VD 3VD Parameters (n=32) (n=78) LAD LCX RCA (n=51) (n=17) (n=10) (n=31) (n=14) (n=6) Age (year) 055.1± ± ± ± ± ± ± ±11.7 EF (%) 063.4± ± ± ± ± ± ± ±10.2 BMI 023.8± ± ± ± ± ± ± ±3.3 METs 011.3± ± ± ± ± ± ± ±1.5 rhr (bpm) 069.5± ± ± ± ± ± ± ±13.1 rsbp (mmhg) 124.1± ± ± ± ± ± ± ±20.2 rdbp (mmhg) 073.8± ± ± ± ± ± ± ±9.0 phr (bpm) 154.0± ± ± ± ± ± ± ±13.1 psbp (mmhg) 189.4± ± ± ± ± ± ± ±32.4 pdbp (mmhg) 103.9± ± ± ± ± ± ± ±12.3 Statistically significant difference compared with the control group (p<0.05). EF: ejection fraction, BMI: body mass index, METs: metabolic equivalents, r: rest, p: peak, HR: heart rate, SBP: systolic blood pressure, DBP: diastolic blood pressure, VD: vessel disease, LAD: left anterior descending artery, LCX: left circumflex artery, RCA: right coronary artery

3 Jin Hong Park, et al. 545 groups according to the number of diseased vessels and the location of stenosis) are shown in Table 1. There were no significant differences in the age and BMI (body mass index) between the control and patient groups (55.1±13.8 years vs. 58.7±9.9 years and 23.8±3.2 kg/m 2 vs. 24.6±2.7 kg/m 2 ). Echocardiography revealed nearly normal ventricular systolic function in both groups (EF>50%). The patient group was divided into three subgroups; single vessel disease (1VD, n=51), two vessel disease (2VD, n=17) and three vessel disease (3VD, n=10). Also, the single vessel disease subgroup was further divided into three groups according to the location of stenosis, LAD (n=31), LCX (n=14) and RCA (n=6). There was no significant difference in the total work load achieved during the exercise testing between the control and patient groups (11.3±2.6 METs vs. 9.6±4.7 METs), but the 2VD and 3VD subgroups showed significantly lower values than the control group (8.9±3.0 METs and 7.4±1.5 METs). No significant differences were observed between the control and patient groups with regard to the resting blood pressure, resting heart rate and peak systolic blood pressure. However, the peak heart rate and peak diastolic blood pressure were significantly greater in the control than the patient group. QT interval and QT dispersion (Table 2) In both the control and patient groups, the QTi significantly decreased as the heart rate was raised, and gradually increased again during the recovery stage after exercise (Fig. 1). The QTi during the resting state showed no significant difference between the two groups, but during the pre-peak, peak exercise and recovery stages after exercise was significantly greater in the patient group. The QTd also slightly decreased during exercise in both groups, but there was no statistical significance, as with the QTi. The QTd was significantly greater in the patient than the control group at rest, and the differences between the QTd values became greater with exercise stress in both groups (Fig. 2). Among the 1VD, 2VD and 3VD subgroups, the QTi showed no significant differences on the resting ECG. However, at the pre-peak, during the peak and 3 minutes after the exercise, the QTi of the three subgroups showed significant differences, which tended to increase according to the number of vessels involved. The QTi also showed statistically significant differences between the three subgroups according to the location of stenosis in a single vessel disease. Conversely, the QTd showed no significant differences Table 2. QT interval and QT dispersion during the exercise test Supine Stand Pre-Peak Peak 1 minute 3 minute QTi (msec) ANOVA Control 396.3± ± ± ± ± ±35.7 <0.01 Patients 406.4± ± ± ± ± ±34.1 <0.01 p NS NS <0.01 <0.01 <0.01 < vessel disease 407.1± ± ± ± ± ±28.7 < vessel disease 405.9± ± ± ± ± ±24.6 < vessel disease 404.0± ± ± ± ± ±47.0 <0.01 ANOVA NS NS <0.01 <0.01 NS <0.01 LAD 398.7± ± ± ± ± ±25.4 <0.01 LCX 415.7± ± ± ± ± ±21.4 <0.01 RCA 430.0± ± ± ± ± ±36.7 <0.01 ANOVA NS <0.01 <0.02 <0.02 NS <0.01 QTd (msec) Control 035.0± ± ± ± ± ±9.90 NS Patients 041.5± ± ± ± ± ±13.3 NS p 0.02 <0.01 <0.01 <0.01 <0.01 < vessel disease 040.0± ± ± ± ± ±14.7 NS 2 vessel disease 044.7± ± ± ± ± ±11.2 NS 3 vessel disease 044.0± ± ± ± ± ±8.40 NS LAD 040.0± ± ± ± ± ±14.3 NS LCX 038.5± ± ± ± ± ±18.3 NS RCA 043.3± ± ± ± ± ±8.20 NS NS: no statistical significance, QTi: QT interval, QTd: QT dispersion, LAD: left anterior descending artery, LCX: left circumflex artery, RCA: right coronary artery, ANOVA: analysis of variance

4 546 QT Dispersion in Stable Angina QTi (msec) Supine Stand Pre-peak Peak 1 min 3 min RRi (IHD) RRi (control) QTi max (IHD) according to the number of diseased vessels or the location of stenosis. However, during peak exercise, the QTd tended to increase according to the number of diseased vessels, and also showed a tendency to increase in the RCA subgroup compared to the LAD and LCX subgroups with a single vessel disease. Exercise stage QTi max (control) QTi min (IHD) QTi min (control) Fig. 1. Maximal QT interval (Max QTi) and minimal QT interval (Min QTi) of control and patient groups during exercise test. QT interval (QTi) shortens according to increasing heart rate in both groups. Max QTi of patient group decreases lesser than that of controls, showing significant differences at exercise and recovery stages. Min QTi of patient group falls more rapidly than min QTi of controls, and at peak exercise min QTi s of the two groups show no significant difference ( ). Significant difference between two groups. IHD: ischemic heart disease, RRi: RR interval. QTd (msec) Supine Stand Pre-peak Peak 1 min 3 min cqtd (IHD) cqtd (control) Exercise stage QTd (IHD) QTd (control) Fig. 2. QT dispersion (QTd) and corrected QT dispersion (cqtd) of the control and patient groups during the exercise test. Comparing the control and patient groups, the QTd and cqtd show significant differences throughout the test. However, there are no significant changes in the QTd values throughout the exercise stages in either group; whereas, the cqtd was significantly increased with exercise in both groups. Significant difference between two groups. IHD: ischemic heart disease RRi (sec) cqtd (msec) Corrected QT interval and corrected QT dispersion (Table 3) The Corrected QTi increased as heart rate rose in both the control and patient groups. In the control group, the corrected QTi was greatest during peak exercise and in the patient group during the pre-peak stage. The corrected QTi tended to be greater in the patient than the control group, but showed statistical significance only during the pre-peak stage and 1 minute after exercise. The corrected QTd also increased according to the heart rate, with both groups showing the greatest corrected QTd values during the peak exercise stage. The corrected QTd was significantly greater in the patient group throughout the entire test (Fig. 2). In the 1VD group, the corrected QTi showed significant differences according to the location of stenosis during the peak-exercise, 1 and 3 minutes after exercise, which was greater in the RCA subgroup compared to the LAD and LCX subgroups. However, the cqtd showed no significant differences according to the number of diseased vessels and the location of stenosis. Delta QT interval in V5 and avf (Table 4) In the 1VD group, the ΔQTi showed no significant difference between the subgroups according to the location of stenosis. However, the ΔQTi of the V5 lead was greater in the subgroup with LAD lesions (120.7±40.5 msec) than the LCX and RCA subgroups (119.3±37.3 msec and 103.3±26.6 msec). In the RCA lesion subgroup, the ΔQTi of the avf lead tended to be higher compared to the other subgroups (RCA: 133.3±16.3 msec vs. LAD: 120.7±39.2 msec and LCX: 118.6± 40.4 msec). Discussion The shortening of the QTi with increasing heart rate is a well recognized consequence of the interval and duration relationship. 7) In our study, the QTi had a reverse relationship with the heart rate during exercise in both normal individuals and stable angina patients, but gradually decreased as the heart rate rose, and reached minimum values during peak exercise, then progressively increased throughout the recovery stages. It has been noted in several studies that the QTi in the resting state is prolonged in patients suffering acute or remote myocardial infarction and during balloon inflation in percutaneous coronary intervention, but discrepancies still remain with stable angina. 8) In our study with stable angina patients, the QTi in the resting state showed no significant difference between the control and patient groups. However, during the treadmill exercise, the QTi of the patient group progressively exhibited a tendency to decrease to a lesser degree than that of the controls, showing significantly greater values compared with those

5 Jin Hong Park, et al. 547 Table 3. Corrected QT interval and corrected QT dispersion during the exercise test cqti (msec) Supine Stand Pre-Peak Peak 1 minute 3 minute ANOVA Control 421.2± ± ± ± ± ±31.8 <0.01 Patients 431.7± ± ± ± ± ±31.1 <0.01 p NS NS <0.01 NS <0.05 NS 1 vessel disease 428.6± ± ± ± ± ±30.6 < vessel disease 432.5± ± ± ± ± ±25.0 < vessel disease 446.2± ± ± ± ± ±35.4 NS ANOVA NS NS NS <0.01 NS <0.03 LAD 424.4± ± ± ± ± ±20.0 <0.01 LCX 427.7± ± ± ± ± ±38.1 <0.01 RCA 452.2± ± ± ± ± ±33.6 NS ANOVA NS 0.03 NS <0.01 <0.04 <0.01 cqtd (msec) Control 37.0± ± ± ± ± ±12.3 <0.01 Patients 44.2± ± ± ± ± ±16.1 <0.01 p <0.01 <0.01 <0.01 < vessel disease 42.3± ± ± ± ± ±17.8 < vessel disease 47.3± ± ± ± ± ±12.2 < vessel disease 49.1± ± ± ± ± ±13.5 NS LAD 42.8± ± ± ± ± ±18.1 <0.01 LCX 39.5± ± ± ± ± ±20.0 <0.04 RCA 46.1± ± ± ± ± ±9.50 <0.01 NS: no statistical significance, cqti: corrected QT interval, cqtd: corected QT dispersion, LAD: left anterior descending artery, LCX: left circumflex artery, RCA: right coronary artery, ANOVA: analysis of variance of the control group during the pre-peak and peak exercises (Table 2) (Fig. 1). Previous reports, which showed a significant difference of the QTi in the resting state, have mainly been based on study groups with myocardial infarction patients. However; in our study, the patient group was restricted to stable angina patients, who presented no anginal symptoms at rest. Therefore, the difference in the study subjects could be the reason for the QTi in the resting state showing no significant difference in our study. During exercise, stress induced ischemia caused abnormal changes in the ventricular repolarization process, resulting in a significant difference between the QTi of the normal controls and stable angina patients. A dynamic change of the QTi during exercise is also affected by impaired responses to circulating catecholamines and autonomic control in ischemic myocardium, which have been observed in several previous studies. 9-12) This is regarded as contributing to the blunt decrease of the QTi in stable angina patients. With respect to our findings, the QTi was sensitive enough to present the occurrence of myocardial ischemia, even in stable coronary artery disease, which was considered concordant with previous studies. The QTd has been calculated by deduction of the min QTi from the max QTi. 13) In our study, the QTds of the control and patient groups were revealed relatively constant values throughout the exercise test, even those that exhibited a slight decrease during the exercise stage compared to the resting state. The QTd showed a significant difference between the two groups at rest, and this difference increased during exercise. The Max and min QTi of both groups all decreased as the heart rate rose, but the extent of the min QTi shortening in the patient group was much greater than that of the control group. As a result, the min QTi of the patient group fell to the point of the min QTi of the control group during peak exercise. The min QTis of both groups during peak exercise showed no significant difference, while the max QTis of both groups were still significantly different (Fig. 1). Therefore, it can be considered that the difference in the QTds between the control and patient groups during exercise testing is more likely due to the differential shortening of the min QTi rather than that of the max QTi. In other words, the increase in the QTd was mainly attributed to the preferential shortening of the min QTi in ischemic myocardium. This is also thought to be related to the alteration of potassium channels caused by ischemia. 14)15) The QTd showed sig-

6 548 QT Dispersion in Stable Angina Table 4. Delta QT interval in leads V5 and avf of single coronary artery disease LAD LCX RCA Rest Peak Rest Peak Rest Peak ANOVA V5 QTi 389.0± ± ± ± ± ±32.0 Δ QTi 120.7± ± ± avf QTi 386.5± ± ± ± ± ±35.2 Δ QTi 120.7± ± ± LAD: left anterior descending artery, LCX: left circumflex artery, RCA: right coronary artery, ΔQTi: delta QT interval, difference between QTi at peak exercise and QTi at rest, ANOVA: analysis of variance, VF: ventricular fibrillation nificant differences between the two groups, even at rest. Therefore, the QTd is considered sensitive enough to detect myocardial ischemia in patients that present with no angina symptoms at rest. The corrected QTi, which was calculated using Bazett s formula with our data, presented maximally prolonged values during peak exercise in the control group and during the pre-peak exercise stage in the patient group, and minimum values during the 1 minute recovery stage in both groups. These changes in the behavior patterns seemed to be more complicated and non-linear compared to the above results for the uncorrected QTi. Several previous studies have already reported that Bazett s formula, commonly used to correct the effect of heart rate on QTi, causes under- and over-correction around the peak exercise period and during the early period of the recovery stage, respectively. 16) The resting corrected QTd of the patient group was noted to be significantly prolonged compared with that of the controls, and were increased throughout both the exercise test and maximized during peak exercise, which showed quite a different pattern from the uncorrected QTds, but maintained relatively constant values throughout the exercise test in both groups. Therefore, these discordant aspects between before and after correction suggest that any intentions to correct the effect of heart rate on the QTi will inevitably result in some distortion of the original data and its patterns. Therefore, this demands very prudent attention when applying corrected QTi and QTd values for assessing the dynamic changes of these values during exercise testing or on a 24 hour ambulatory ECG. Whether the QTd and cqtd can represent the severity of myocardial ischemia, in terms of the number of diseased vessels, is still controversial. Pak and colleagues 17) reported that the QTd and cqtd showed significant differences according to the severity of the disease. Conversely, Dilaveris and colleagues reported that the QTd and cqtd were not affected by the severity of CAD. 18) In our study, the QTd and cqtd showed no significant relationship with the number of diseased vessels. Several factors affecting the dynamic changes of the QTd during exercise testing have been reported. For example, circulating catecholamines and the autonomic nervous system are regarded to be more actively involved in the change of the QTd during the exercise test. A recent study has also suggested the lactate extraction rate had excellent correlation with the QTd, which indicates that the facilitated condition of lactate production, such as during exercise testing, is supposed to augment the QTd. 19) However, the spatial distribution of myocardial ischemia is considered an important factor in manifesting the degree of the QTd. 20) In other words, the QTd is reduced when ischemia is global, as the regional difference of the QTi is diminished. Therefore, the severity of ischemia is not merely determined by the number of diseased vessels, but also by the percentage stenosis of an individual diseased vessel, collateral circulation, the amount of electrically responsible viable myocardium and the distance from the surface ECG leads. It is considered that further studies are needed on the relationship of the QTi and QTd with the severity of coronary artery disease during exercise testing, which should encompass all of the above factors. Several recent studies 21)22) have presented discordant statements as to whether the QTi and QTd can show any difference according to the location of the diseased vessel in a single vessel disease; in other words, whether the QTi and QTd could reflect the anatomic location of myocardial ischemia. Tikiz and colleagues 23) reported that the QTi and QTd showed no relation with the anatomic location of ischemia, but only the relevance of the severity of ischemia in single vessel diseased patients. However, in most studies, including the present study, the QTi was generally measured at the lead presenting the longest QT in standard 12 lead ECGs, regardless of the anatomic location of ischemia, which means the QTi usually comes from different leads during each stage of exercise testing, and not constantly from the same lead. To solve this issue, the ΔQTi was used, which is the difference between the QTi at peak exercise and that during the rest stage in individually concerned leads. In our study, the concerned leads were the V5 and avf leads, which were assumed to represent the local electrical activities of the left anterior descending and right coronary artery territory, respectively. Although our study failed to demonstrate any statistical significance, the ΔQTis showed tendencies to increase in the V5 lead in patients with LAD disease, and in the avf lead in pa-

7 Jin Hong Park, et al. 549 tients with RCA disease (Table 4). Therefore, the ΔQTi was presumed to be a reasonable approach to address the location of ischemia. Conclusively, in stable angina patients, the QTi and QTd were significantly prolonged compared with those in normal individuals, and augmented by exercise stress, regardless of the extent of the disease, but rarely reflected the location or severity of coronary artery disease. Only the QTi and cqti had tendencies to increase according to the number of vessels with stenosis during peak exercise. In terms of the relationships between the QTi and QTd, cqti and cqtd, no mirror relationship with QTi and QTd was observed; thus, careful interpretation will be required to concurrently deal with both these values; even the cqti and cqtd were prolonged in stable angina patients. Additionally, to possibly address the location of regional ischemia in a single vessel disease, the ΔQTi could be considered, but further study will be required for confirmation. REFERENCES 1) Shu J, Zhu T, Yang L, Cui C, Yan GX. ST-segment elevation in the early repolarization syndrome, idiopathic ventricular fibrillation, and the Brugada syndrome: cellular and clinical linkage. J Electrocardiol 2005;38: ) Zareba W, Moss AJ, le Cessie S. Dispersion of ventricular repolarization and arrhythmic cardiac death in coronary artery disease. Am J Cardiol 1994;74: ) Day CP, Mc Comb JM, Campbell RW. QT dispersion: an indication of arrhythmia risk in patients with long QT intervals. Br Heart J 1990;63: ) Kim BS, Kang JH, Lee SW, et al. Effect of coronary angioplasty on QT and JT dispersion. Korean Circ J 1998;28: ) Choi KJ, Lee IS, Lee SG, et al. Change of QT dispersion following PTCA in angina patients. Korean Circ J 1998;28: ) Ahnve S. Correction of QT interval for heart rate: review of different formulas and the use of Bazett s formula in myocardial infarction. Am Heart J 1985;109: ) Kligfield P, Lax KG, Okin PM. QT interval-heart rate relation during exercise in normal men and women: definition by linear regression analysis. J Am Coll Cardiol 1996;28: ) Elming H, Sonne J, Lublin HK. The importance of the QT interval: a review of the literature. Acta Psychiatr Scand 2003;107: ) Ahnve S, Vallin H. Influence of heart rate and inhibition of autonomic tone on the QT interval. Circulation 1982;65: ) Boudoulas H, Ruff PD, Fulkerson PK, Lewis RP, Dervenagas S. Effect of propranolol on post exercise left ventricular ejection time index. Am J Cardiol 1981;48: ) Boudoulas H, Geleris P, Lewis RP, Leier CV. Effect of increased adrenergic activity on the relationship between electrical and mechanical systole. Circulation 1981;64: ) Davey P, Bateman J. Heart rate and catecholamine contribution to QT interval shortening on exercise. Clin Cardiol 1999;22: ) Malik M, Batchvarov VN. Measurement, interpretation and clinical potential of QT dispersion. J Am Coll Cardiol 2000;36: ) Roukema G, Singh JP, Meijs M, Carvalho C, Hart G. Effect of exercise induced ischemia on QT interval dispersion. Am Heart J 1998;135: ) Lukas A, Antzelevitch C. Differences in the electrophysiological response of canine ventricular epicardium and endocardium to ischemia: role of the transient outward current. Circulation 1993; 88: ) Benatar A, Decraene T. Comparison of formulae for heart rate correction of QT interval in exercise ECGs from healthy children. Heart 2001;86: ) Pak HN, Kim YH, Park SW, et al. Diagnostic value of QT and JT dispersion in exercise ECG. Korean Circ J 1995;25: ) Dilaveris P, Andrikopuolos G, Metaxas G, et al. Effects of ischemia on QT dispersion during spontaneous anginal episodes. J Electrocardiol 1999;32: ) Stierle U, Giannitsis E, Sheikhzadeh A, et al. Relation between QT dispersion and the extent of myocardial ischemia in patients with three-vessel coronary artery disease. Am J Cardiol 1998;81: ) Roukema G, Singh JP, Meijs M, Carvalho C, Hart G. Effect of exercise-induced ischemia on QT interval dispersion. Am Heart J 1998;135: ) Yunus A, Gillis AM, Traboulsi M, et al. Effect of coronary angioplasty on precordial QT dispersion. Am J Cardiol 1997;79: ) Moreno FL, Vilaneuva T, Karagounis LA, Anderson JL. Reduction in QT interval dispersion by successful thrombolytic therapy in acute myocardial infarction. Circulation 1994;90: ) Tikiz H, Terzi T, Balbay Y, et al. QT dispersion in single coronary artery disease: is there a relation between QT dispersion and diseased coronary artery or lesion localization? Angiology 2001; 52:43-51.