QRS Duration and QT Interval Predict Mortality in Hypertensive Patients With Left Ventricular Hypertrophy
2004; Lippincott Williams & Wilkins; Volume: 43; Issue: 5 Linguagem: Inglês
10.1161/01.hyp.0000125230.46080.c6
ISSN1524-4563
AutoresLasse Oikarinen, Markku S. Nieminen, Matti Viitasalo, Lauri Toivonen, Sverker Jern, Björn Dahlöf, Richard B. Devereux, Peter M. Okin,
Tópico(s)Heart Rate Variability and Autonomic Control
ResumoHomeHypertensionVol. 43, No. 5QRS Duration and QT Interval Predict Mortality in Hypertensive Patients With Left Ventricular Hypertrophy Free AccessResearch ArticlePDF/EPUBAboutView PDFView EPUBSections ToolsAdd to favoritesDownload citationsTrack citationsPermissions ShareShare onFacebookTwitterLinked InMendeleyReddit Jump toFree AccessResearch ArticlePDF/EPUBQRS Duration and QT Interval Predict Mortality in Hypertensive Patients With Left Ventricular HypertrophyThe Losartan Intervention for Endpoint Reduction in Hypertension Study Lasse Oikarinen, Markku S. Nieminen, Matti Viitasalo, Lauri Toivonen, Sverker Jern, Björn Dahlöf, Richard B. Devereux, Peter M. Okin and Lasse OikarinenLasse Oikarinen From the Division of Cardiology (L.O., M.S.N., M.V., L.T.), Department of Medicine, Helsinki University Central Hospital, Helsinki, Finland; Sahlgrenska University Hospital/Östra (S.J., B.D.), Göteborg, Sweden; and Division of Cardiology (R.B.D., P.M.O.), Department of Medicine, Weill Medical College of Cornell University, New York, NY. , Markku S. NieminenMarkku S. Nieminen From the Division of Cardiology (L.O., M.S.N., M.V., L.T.), Department of Medicine, Helsinki University Central Hospital, Helsinki, Finland; Sahlgrenska University Hospital/Östra (S.J., B.D.), Göteborg, Sweden; and Division of Cardiology (R.B.D., P.M.O.), Department of Medicine, Weill Medical College of Cornell University, New York, NY. , Matti ViitasaloMatti Viitasalo From the Division of Cardiology (L.O., M.S.N., M.V., L.T.), Department of Medicine, Helsinki University Central Hospital, Helsinki, Finland; Sahlgrenska University Hospital/Östra (S.J., B.D.), Göteborg, Sweden; and Division of Cardiology (R.B.D., P.M.O.), Department of Medicine, Weill Medical College of Cornell University, New York, NY. , Lauri ToivonenLauri Toivonen From the Division of Cardiology (L.O., M.S.N., M.V., L.T.), Department of Medicine, Helsinki University Central Hospital, Helsinki, Finland; Sahlgrenska University Hospital/Östra (S.J., B.D.), Göteborg, Sweden; and Division of Cardiology (R.B.D., P.M.O.), Department of Medicine, Weill Medical College of Cornell University, New York, NY. , Sverker JernSverker Jern From the Division of Cardiology (L.O., M.S.N., M.V., L.T.), Department of Medicine, Helsinki University Central Hospital, Helsinki, Finland; Sahlgrenska University Hospital/Östra (S.J., B.D.), Göteborg, Sweden; and Division of Cardiology (R.B.D., P.M.O.), Department of Medicine, Weill Medical College of Cornell University, New York, NY. , Björn DahlöfBjörn Dahlöf From the Division of Cardiology (L.O., M.S.N., M.V., L.T.), Department of Medicine, Helsinki University Central Hospital, Helsinki, Finland; Sahlgrenska University Hospital/Östra (S.J., B.D.), Göteborg, Sweden; and Division of Cardiology (R.B.D., P.M.O.), Department of Medicine, Weill Medical College of Cornell University, New York, NY. , Richard B. DevereuxRichard B. Devereux From the Division of Cardiology (L.O., M.S.N., M.V., L.T.), Department of Medicine, Helsinki University Central Hospital, Helsinki, Finland; Sahlgrenska University Hospital/Östra (S.J., B.D.), Göteborg, Sweden; and Division of Cardiology (R.B.D., P.M.O.), Department of Medicine, Weill Medical College of Cornell University, New York, NY. , Peter M. OkinPeter M. Okin From the Division of Cardiology (L.O., M.S.N., M.V., L.T.), Department of Medicine, Helsinki University Central Hospital, Helsinki, Finland; Sahlgrenska University Hospital/Östra (S.J., B.D.), Göteborg, Sweden; and Division of Cardiology (R.B.D., P.M.O.), Department of Medicine, Weill Medical College of Cornell University, New York, NY. and From the Division of Cardiology (L.O., M.S.N., M.V., L.T.), Department of Medicine, Helsinki University Central Hospital, Helsinki, Finland; Sahlgrenska University Hospital/Östra (S.J., B.D.), Göteborg, Sweden; and Division of Cardiology (R.B.D., P.M.O.), Department of Medicine, Weill Medical College of Cornell University, New York, NY. and for the LIFE Study Investigators Originally published22 Mar 2004https://doi.org/10.1161/01.HYP.0000125230.46080.c6Hypertension. 2004;43:1029–1034Other version(s) of this articleYou are viewing the most recent version of this article. Previous versions: March 22, 2004: Previous Version 1 AbstractLeft ventricular hypertrophy is a risk factor for cardiovascular mortality, including sudden cardiac death. Experimentally, left ventricular hypertrophy delays ventricular conduction and prolongs action potential duration. Electrocardiographic QRS duration and QT interval measures reflect these changes, but whether these measures can further stratify risk in patients with electrocardiographic left ventricular hypertrophy is unknown. We measured the QRS duration and QT intervals from the baseline 12-lead electrocardiograms in the Losartan Intervention For Endpoint Reduction in Hypertension (LIFE) study, which included hypertensive patients with electrocardiographic evidence of left ventricular hypertrophy randomized to either losartan-based or atenolol-based treatment to lower blood pressure. In the present study, we related study baseline electrocardiographic measures to cardiovascular and all-cause mortality. There were 5429 patients (male 45.8%; mean age 66±7 years) included in the present analyses. After a mean follow-up of 4.9±0.8 years, there were 417 deaths from all causes, including 214 cardiovascular deaths. In separate univariate Cox regression analyses, QRS duration and several QT measures were significant predictors of cardiovascular mortality and all-cause mortality. However, in multivariate Cox analyses including all electrocardiographic measures and adjusting for other risk factors as well as treatment strategy, only QRS duration and maximum rate-adjusted QTapex interval remained as significant independent predictors of cardiovascular (P=0.022 and P=0.037, respectively) and all-cause mortality (P=0.038 and P=0.002, respectively). In conclusion, in a hypertensive risk population identified by electrocardiographic left ventricular hypertrophy, increased QRS duration and maximum QTapex interval can further stratify mortality risk even in the setting of effective blood pressure-lowering treatment.Left ventricular hypertrophy (LVH) is an important indicator of target organ damage in chronic arterial hypertension. Electrocardiographically and echocardiographically detected LVH independently predict increased morbidity and mortality,1,2 including sudden cardiac death,3,4 and the relative risk of these events increases with increasing LV mass. The impact of LVH on outcome may in part be mediated by adverse changes in LV mechanical function. However, LVH also induces potentially arrhythmogenic changes in LV electrophysiology. Experimental evidence shows that LVH alters ventricular conduction and repolarization.5–14 In the 12-lead ECG, these changes may prolong the QRS and QT interval duration, respectively, and affect T-wave morphology.15–17 In hypertensive patients with ECG LVH, increase in LV mass was associated with prolonged QRS and QT interval duration and measures of QT dispersion.18 However, it is unknown if these ECG measures can further stratify risk in patients with LVH. Therefore, we investigated, in a large longitudinal study of hypertensive patients with ECG evidence of LVH, the value of QRS duration, QT interval duration, and QT dispersion as predictors of cardiovascular and all-cause mortality.MethodsStudy PopulationThe present study population originated from The Losartan Intervention For Endpoint Reduction in Hypertension (LIFE) study. The study design, inclusion and exclusion criteria, baseline characteristics of the LIFE study population, and the main results of the LIFE study have been previously published.19–21 ECG evidence of LVH was an inclusion criterion in the LIFE study.19,20 The study was approved by institutional review committees, and all subjects gave informed consent. After randomization, target blood pressure was <140/90 mm Hg and patients were followed-up for at least 4 years.All 9193 patients included in the LIFE study were considered eligible for the present study, in which 5429 patients were included. The QT substudy center excluded the ECGs from 3764 patients from the present study before any knowledge of the outcome data because of the following reasons: (1) missing data (n=236) or low quality of the received ECG (n=130); (2) ECG recording not suitable for QT dispersion analysis (≤2 simultaneously recorded leads, amplitude calibration other than 10 mm/mV, or only 1 signal-averaged QRS-T complex in the ECG; n=1570); (3) generally accepted exclusion criteria from QT measurements (atrial fibrillation, bundle branch block, heart rate 120 bpm; n=889); and (4) QT dispersion measurement not technically feasible according to predefined criteria (wandering signal baselines, <6 leads with 2 measurable consecutive QRS-T complexes [<3 precordial leads], or drugs possibly affecting QT intervals; n=939).QRS Duration, QT Interval, and QT Dispersion MeasuresA trained technician, unaware of the clinical data or the study design, performed all ECG QRS and QT interval measurements. The measurement procedure and intraobserver variability for measurements have been described in detail previously.18 QRS duration, QTapex (QRS onset to T-wave apex), and QTend (QRS onset to T-wave end) interval measurements were performed from ECGs recorded at the randomization visit. All included ECGs had an ECG paper speed of either 25 or 50 mm/s. For each patient, maximum QRS duration (referred to later as QRS duration) and maximum QTapex and QTend intervals in any lead were determined, and the latter were rate-adjusted with the nomogram (Nc) method (referred to later as QTapexNc and QTendNc, respectively).22 QTapex dispersion was defined as the standard deviation (SD) of all QTapex intervals in the 12 leads (QTapexSD). QTendSD was calculated in a similar fashion. Visible T-wave abnormalities (typical strain pattern and/or Minnesota codes 5:1 or 5:2) were defined as previously described.23,24Outcome MeasuresFor the present study, cardiovascular mortality and all-cause mortality were prespecified as primary endpoints. We also used the LIFE study primary composite endpoint (cardiovascular death, nonfatal myocardial infarction, and nonfatal stroke).19,21Statistical AnalysisThe ECG analyses were prespecified as part of the LIFE ECG substudy protocol. Data were analyzed with SPSS version 10.0.7 software (SPSS Inc, Chicago, Ill). All continuous data are presented as mean±SD. Mean values were compared between groups by use of independent samples t test, except for ECG measures, which were compared by use of 2-way ANOVA adjusting for differences in ECG paper speed, because it has an effect on estimates of QT interval measures.18,25,26 Proportions were compared by χ2 tests. Outcome rates were calculated by the product-limit method and were plotted by the Kaplan-Meier method, with comparisons of outcome rates between dichotomized groups performed with the log-rank test. Dichotomized groups were created for QRS duration and each QT measure by the median value of the respective ECG measure in both ECG paper speed categories, because of the effect of ECG paper speed on QT variables, and because of the fact that outcome rates were slightly higher in centers using an ECG paper speed of 25 mm/s.Outcome analyses for ECG measures were performed by Cox proportional hazard models by studying each ECG variable separately after adjusting for ECG paper speed. To test the independence of each ECG measure as a predictor of outcome, multivariate Cox models were used. Finally, to identify those ECG variables with the strongest predictive power, all ECG variables and covariates were included in the same multivariate Cox model. A 2-tailed P<0.05 was considered statistically significant.ResultsPatient CharacteristicsDuring the follow-up (mean follow-up 4.9±0.8 years), 214 (3.9%) patients died from cardiovascular causes, 417 patients (7.7%) died from all causes, and 581 patients (10.7%) experienced the LIFE study composite endpoint. Patients who experienced cardiovascular death compared with those who did not were older, more often men, and had higher baseline systolic blood pressure, Framingham risk score,27 baseline Sokolow-Lyon voltage, Cornell voltage duration product, and a greater decrease in systolic blood pressure (Table 1). There were no differences between these groups in baseline diastolic blood pressure or change in diastolic blood pressure. The results were similar when comparing those who experienced the LIFE study composite endpoint to those who did not (data not shown). Comparing those who died (all-cause mortality) to survivors, similar differences again were detected between groups, except baseline diastolic blood pressure was lower in those who died (Table 1). TABLE 1. Clinical and Electrocardiographic Variables in Relation to Cardiovascular and All-Cause MortalityVariableCardiovascular MortalityAll-Cause MortalitySurvived (n=5215)Died (n=214)PSurvived (n=5012)Died (n=417)PValues are mean±SD or proportions.Δ indicates change (from baseline); BP, blood pressure; VP, voltage product; SD, standard deviation of the respective QT intervals; Nc, nomogram-corrected for heart rate.Age, y66±771±7<0.000166±770±7<0.0001Gender, % male45.357.90.000345.1%54.4%0.0002Systolic BP, mm Hg174±14177±140.003174±14176±140.008Diastolic BP, mm Hg98±997±100.12898±997±100.005Δ Systolic BP, mm Hg−29.4±18.5−32.4±19.40.034−29.3±18.4−32.4±20.40.007Δ Diastolic BP, mm Hg−16.6±9.7−16.1±11.00.481−16.6±9.7−16.6±10.50.997Framingham risk score, %21.7±9.227.3±10.6<0.000121.6±9.226.1±10.4<0.0001Sokolow-Lyon voltage, mm30.3±10.232.2±10.80.00630.2±10.131.8±11.30.005Cornell VP, mm×ms2645±7662837±8950.0022646±7622741±8810.031ECG paper speed 50 mm/s, %65.951.4<0.000166.353.7<0.0001T-wave abnormality, %23.445.3<0.000123.237.3<0.0001QRS duration, ms105±14109±13<0.0001105±14108±13<0.0001Maximum QTendNc, ms426±25430±240.014426±25431±24<0.0001Maximum QTapexNc, ms357±23363±220.0002357±24362±22<0.0001QTendSD, ms15.4±6.416.5±6.00.00615.3±6.416.2±6.00.008QTapexSD, ms16.4±7.118.2±6.7<0.000116.4±7.217.5±6.80.001QRS Duration, QT Measures, and Cardiovascular MortalityThe relations of QRS duration and QT measures to cardiovascular mortality and all-cause mortality are also shown in Table 1. All such ECG measures were greater in those patients who had cardiovascular death or all-cause death compared with those who did not experience the respective outcome measure. The results were similar comparing those who experienced the LIFE study composite endpoint to those who did not (data not shown).Visible T-wave abnormalities were present in 1315 patients (24.2%) at study baseline (617 patients [11.4%] showed the typical strain pattern). QTapexNc was longer in patients with visible T-wave abnormalities than in those without them (365±23 versus 353±22 ms, respectively; P<0.0001). Visible T-wave abnormalities were associated with cardiovascular and all-cause mortality (Table 1).In Cox regression analyses, after adjusting for ECG paper speed, QRS duration and all QT measures separately predicted cardiovascular mortality (Table 2). After adjusting for QRS duration, QTapexNc (P=0.011) remained a predictor of cardiovascular mortality, whereas QTendNc did not (P=0.434). When stratifying the study population by QRS duration above and below the median values in both ECG paper speed categories, those above the median had cardiovascular death more often (P=0.0006 by log-rank test; Figure). The results were similar for groups similarly dichotomized by median QTapexNc (P=0.0017; Figure), QTapexSD (hazard ratio [HR] 1.67 [95% CI: 1.27 to 2.20], P=0.0002), and QTendSD (HR 1.49 [95% CI: 1.13 to 1.95], P=0.0042), whereas there was no significant difference in QTendNc (P=0.11). In separate multivariate Cox analyses adjusting each ECG measure for covariates (gender, baseline systolic and diastolic blood pressure, treatment arm [losartan or atenolol], baseline Framingham risk score, Sokolow-Lyon voltage and Cornell voltage duration product, ECG paper speed), QRS duration, QTapexNc, and QTapexSD remained significant predictors (Table 2), even after further adjustment for the presence of visible T-wave abnormalities in the ECG. After entering all ECG measures and covariates in the same model, QRS duration (χ2=5.3; P=0.022; HR 1.16 per SD [13 ms] of QRS [95% CI: 1.09 to 1.24]) and QTapexNc (χ2=4.3; P=0.037; HR 1.16 per SD [23 ms] of QTapexNc [95% CI: 1.08 to 1.24]) were independent predictors of cardiovascular mortality. In patients with both QRS duration and QTapexNc ≥ median, cardiovascular mortality rate was significantly higher than in patients who had both measures below the median (90/1514 [5.9%] versus 42/1501 [2.8%] patients, respectively; HR 2.14 [95% CI: 1.49 to 3.09], P<0.0001). TABLE 2. Cox Proportional Hazards Models for Prediction of the Cardiovascular and All-Cause Mortality Examining ECG Measures as Continuous Variables*VariableComparison, ms†Hazard Ratio95% CIχ2Pχ2‡P‡Nc indicates nomogram-corrected for heart rate; SD, standard deviation of the respective QT intervals.*All analyses adjusted for ECG paper speed.†Relative risks calculated for a 1−SD (in the column) increase in the mean.‡After also adjusting for covariates (gender, baseline systolic and diastolic blood pressure, treatment arm [losartan or atenolol], Framingham risk score, baseline Sokolow-Lyon voltage, Cornell voltage duration product).Prediction of cardiovascular mortality QRS duration131.281.21–1.3619.2<0.00019.50.002 QTendNc251.171.09–1.255.10.0240.80.367 QTapexNc231.271.19–1.3613.10.00038.20.004 QTendSD6.31.181.10–1.266.30.0122.30.131 QTapexSD7.11.271.19–1.3613.50.00025.10.024Prediction of all-cause mortality QRS duration131.211.16–1.2720.0<0.000111.70.0006 QTendNc251.221.17–1.2817.3<0.000111.20.0008 QTapexNc231.251.19–1.3120.9<0.000114.80.0001 QTendSD6.31.121.06–1.175.20.0232.20.141 QTapexSD7.11.151.09–1.218.00.0052.60.107Download figureDownload PowerPointKaplan-Meier event-probability curves in patients above and below median values for QRS duration (left panels) and maximum rate-adjusted QTapex interval (QTapexNc; right panels) for cardiovascular mortality (upper panels) and all-cause mortality (lower panels). Median values were 106 ms (ECG paper speed 25 mm/s) and 100 ms (50 mm/s) for QRS duration, and 360 ms and 351 ms for QTapexNc, respectively. See text for the rationale of equal proportions of patients from both paper speed categories in the 2 groups. The probability value is from the log-rank test. Hazard ratios (HR) with 95% confidence intervals (CI) are also shown.QRS Duration, QT Measures, and All-Cause MortalityIn Cox regression analyses, after adjusting for ECG paper speed, QRS duration and all QT measures separately predicted all-cause mortality (Table 2). After adjusting for QRS duration, QTapexNc (P<0.001) and QTendNc (P=0.006) remained as significant predictors of all-cause mortality. When stratifying the study population by QRS duration above and below median values in both ECG paper speed categories, those above the median died more often (P=0.0038 by log-rank test; Figure). The results were similar for groups similarly dichotomized by median QTapexNc (P=0.0002; Figure), QTendNc (HR 1.44 [95% CI: 1.19 to 1.76], P=0.0002), QTapexSD (HR 1.29 [95% CI: 1.06 to 1.56], P=0.010), and QTendSD (HR 1.29 [95% CI: 1.07 to 1.57], P=0.0090). In separate multivariate Cox analyses adjusting each ECG variable for covariates, QRS duration, QTapexNc, and QTendNc remained significant predictors (Table 2), even after further adjustment for the presence of visible T-wave abnormalities in the ECG. After entering all ECG measures and covariates in the same model, QRS duration (χ2=4.3; P=0.038; HR 1.10 per SD [13 ms] of QRS [95% CI: 1.05 to 1.16]) and QTapexNc (χ2=9.6; P=0.002; HR 1.17 per SD [23 ms] of QTapexNc [95% CI: 1.11 to 1.23]) were independent predictors of all-cause mortality. In patients with both QRS duration and QTapexNc ≥ median, all-cause mortality rate was significantly higher than in patients who had both measures below the median (158/1514 [10.4%] versus 91/1501 [6.1%] patients, respectively; HR 1.75 [95% CI: 1.35 to 2.26], P<0.0001).QRS Duration, QT Measures, and LIFE Study Composite EndpointIn separate Cox regression analyses, after adjusting for ECG paper speed, QRS duration (P=0.004), QTapexNc (P=0.0001), QTendNc (P=0.005), QTapexSD (P<0.0001), and QTendSD (P<0.0001) predicted the LIFE study composite endpoint. In separate multivariate Cox analysis adjusting each ECG variable for covariates, QTapexNc (P=0.004), QTapexSD (P=0.019), and QTendSD (P=0.007) remained as significant predictors. After entering all ECG measures and covariates in the same model, QTapexNc (χ2=4.9; P=0.027; HR 1.11 per SD [23 ms] of QTapexNc [95% CI: 1.06 to 1.16]) and QTendSD (χ2=5.1; P=0.024; HR 1.10 per SD [6.3 ms] of QTendSD [95% CI: 1.06 to 1.15]) were independent predictors of the LIFE study composite endpoint.DiscussionMain FindingsThe present results show that in hypertensive patients with ECG evidence of LVH, prolonged QRS duration, and maximum QTapex interval in the baseline ECG were independently associated with increased risk of cardiovascular and all-cause mortality, even in the setting of effective blood pressure-lowering treatment.QRS Duration, QT Intervals, and Mortality in Patients With LVHThe LIFE study included only patients with ECG evidence of LVH. In previous studies it has been shown that patients with ECG LVH, compared with those without ECG LVH, are at higher risk for cardiovascular mortality, including sudden cardiac death.1,3 Within the group of patients with LVH, the relative risk of these events increases as LV mass increases.2,4,28 Previously, we showed in a subgroup from the LIFE study that QRS duration and QT interval measures prolong as LV mass increases,18 suggesting that these ECG measures might identify LVH patients at even higher risk.Our present results show that prolonged QRS duration is, independently of several baseline prognostic variables, associated with increased cardiovascular and all-cause mortality. An increased QRS duration was associated with a 28% higher rate of cardiovascular mortality and a 21% higher rate of all-cause mortality for every SD of the mean increase in value within the range of studied values. Furthermore, QRS duration retained its independent predictive value when all QT interval variables were included in the model. In experimental studies, chronic LVH delays ventricular conduction,5 which seems to be related to an increase in intracellular resistivity caused by increased gap junctional resistance between adjacent cells, and to loss of conduction anisotropy.6,7 In the ECG, QRS duration is a measure of differences in activation times.29 The observed prognostic value of QRS duration in the present study may be related in part to the fact that delayed conduction per se is one of the factors increasing susceptibility to reentrant ventricular arrhythmias. In addition, prolonged QRS duration may be a marker of cellular uncoupling, and even small changes in cellular coupling may have profound effects on repolarization gradients at the myocardial level.17,30We also observed that prolongation of maximum rate-adjusted QT intervals were predictors of all outcome measures. An increased QTapexNc was associated with a 27% higher rate of cardiovascular mortality and a 25% higher rate of all-cause mortality for every SD of the mean increase in value within the range of studied values. QTapexNc was a predictor of all-cause and cardiovascular mortality independently of QRS duration, the other QT measures, and the studied covariates. Even when adjusting for the presence of visible T-wave abnormalities, QTapexNc remained a significant predictor of all-cause and cardiovascular mortality.In experimental studies, LVH has been consistently associated with prolongation of action potential duration (APD),8–13 which may occur in a spatially heterogenous fashion, possibly increasing the arrhythmic vulnerability.10–12,31,32 In LVH, the changes in epicardial versus endocardial APDs have been disparate, with some studies showing prolongation of APD in the epicardium only,13,33 preferentially in the epicardium32 or preferentially in the endocardium.34 The mechanisms of these changes in LVH may be related, in part, to downregulation of the slow component of the delayed rectifier potassium current (IKs),17,35,36 as well as to prolongation of epicardial repolarization time (sum of activation time and APD), because of prolonged transmural activation time and loss of the inverse relation between activation time and APD as a consequence of reduced cell-to-cell coupling induced by fibrosis characteristic of pathological hypertrophy.29,37 In the electrogram, QTapex interval reflects epicardial or endocardial APD depending on T-wave polarity.16,17,29 In the present study, visible T-wave abnormalities were more frequent in those patients who experienced cardiovascular and all-cause mortality, and QTapexNc was longer in patients with these abnormalities. Importantly, however, the predictive value of QTapexNc remained significant even after taking the presence of visible T-wave abnormalities into account, suggesting that QTapexNc is not just another way of characterizing T-wave inversions. In a population at high coronary risk, QTapexNc was a strong predictor of sudden cardiac death, but not of fatal myocardial infarction,38 possibly linking QTapexNc to arrhythmic vulnerability. In a subgroup of the LIFE study population, we observed regression of both ECG indices of LVH and echocardiographic LVH to be associated with a significant shortening of QTapexNc after 1 year of blood pressure-lowering treatment,24 suggesting that effective treatment may partially reverse this adverse repolarization phenotype.Interestingly, in the present study, patients with both prolonged QRS duration and QTapexNc seemed to be at substantially higher risk for mortality. In the normal myocardium, reduction in IKs prolongs APD in the epicardium and endocardium, but not in the M-cell region, prolonging the QTapex interval in the ECG.17 The reduced repolarization reserve caused by decreased IKs activity may become arrhythmogenic in the presence of cellular uncoupling, a condition possibly reflected by prolonged QRS duration.39,40QT Dispersion and Mortality in Patients With LVHIn univariate analysis QTapex dispersion predicted all-cause and cardiovascular mortality. When controlling for several covariates of prognostic importance, QTapexSD remained a predictor of cardiovascular mortality only, but not independently of maximum QTapex interval. QTend dispersion predicted all-cause and cardiovascular mortality only in univariate analysis, whereas it did predict the LIFE study composite endpoint even in multivariate analysis. Thus, QTendSD does not independently further stratify risk of cardiovascular or all-cause mortality in patients with ECG LVH.Study LimitationsThe LIFE study ECG recording procedure was not standardized; therefore, we excluded a significant proportion of ECGs because of technical reasons. However, the exclusion of ECGs was done before any knowledge of the outcome measures, and the present study population seems to be representative of the complete LIFE study population. Because the LIFE study included only patients with ECG-evidence of LVH, it affects the generalizability of the present results; therefore, the prognostic value of all these ECG variables should also be studied in an unselected population of hypertensive patients.PerspectivesThe present results suggest that in hypertensive patients identified as being at high risk by the presence of ECG LVH, baseline QRS duration and QTapexNc interval can further stratify the risk of mortality despite an intervention to effectively lower blood pressure. It needs to be studied if these ECG parameters indeed are related to arrhythmogenesis, either directly or by preconditioning the myocardium in conjunction with, for example, sympathetic stimulation, ischemic episodes, or drugs affecting IKr. Furthermore, the possible differential effects of the various classes of blood pressure-lowering drugs on these ECG markers of risk need to be addressed.This study was supported by a grant from Merck & Co. We thank Paulette A. Lyle for assistance with the preparation of the manuscript.FootnotesCorrespondence to Dr Lasse Oikarinen, Department of Medicine, Division of Cardiology, Helsinki University Central Hospital, Haartmaninkatu 4, 00290 Helsinki, Finland. E-mail [email protected] References 1 Kannel WB, Gordon T, Offutt D. Left ventricular hypertrophy by electrocardiogram. Prevalence, incidence, and mortality in the Framingham Study. Ann Intern Med. 1969; 71: 89–105.CrossrefMedlineGoogle Scholar2 Levy D, Garrison RJ, Savage DD, Kannel WB, Castelli WP. Prognostic implications of echocardiographically determined left ventricular mass in the Framingham Heart Study. N Engl J Med. 1990; 322: 1561–1566.CrossrefMedlineGoogle Scholar3 Kreger BE, Cupples LA, Kannel WB. The electrocardiogram in prediction of sudden death: Framingham Study experience. Am Heart J. 1987; 113: 377–382.CrossrefMedlineGoogle Scholar4 Haider AW, Larson MG, Benjamin EJ, Levy D. Increased left ventricular mass and hypertrophy are associated with increased risk for sudden death. J Am Coll Cardiol. 1998; 32: 1454–1459.CrossrefMedlineGoogle Scholar5 Winterton SJ, Turner MA, O'Gorman DJ, Flores NA, Sheridan DJ. Hypertrophy causes delayed conduction in human and guinea pig myocardium: accentuation during ischaemic perfusion. Cardiovasc Res. 1994; 28: 47–54.CrossrefMedlineGoogle Scholar6 McIntyre H, Fry
Referência(s)