Risk of sudden cardiac death associated with QRS, QTc,
and JTc intervals in the general populationJani T. Tikkanen, MD,* Tuomas Kentta, PhD,* Kimmo Porthan, MD,†
Olli Anttonen, MD,*‡ Antti Eranti, MD,* Aapo L. Aro, MD,x Tuomas Kerola, MD,*‡
Harri A. Rissanen, MSc,k Paul Knekt, MD,k Markku Heli€ovaara, MD,k Arttu Holkeri, MD,*
Anette Haukilahti, MD,* Teemu Niiranen, MD,{ Jussi Hernesniemi, MD,#
Antti Jula, MD,k Markku S. Nieminen, MD,x Robert J. Myerburg, MD,**
Christine M. Albert, MD,††‡‡ Veikko Salomaa, MD,k Heikki V. Huikuri, MD,*
M. Juhani Junttila, MD*From the *Research Unit of Internal Medicine, Medical Research Center Oulu, University of Oulu and Oulu
University Hospital, Oulu, Finland, †Department of Medicine, University of Helsinki and Minerva
Foundation Institute for Medical Research, Helsinki, Finland, ‡P€aij€at-H€ame Central Hospital, Lahti,
Finland, xDivision of Cardiology, Heart and Lung Center, Helsinki University Hospital, Helsinki,
Finland, kTHL-Finnish Institute for Health and Welfare, Helsinki, Finland, {Department of Medicine,
Turku University Hospital and University of Turku, Turku, Finland, #Tays Heart Hospital, Tampere
University Hospital, Tampere, Finland, **Division of Cardiology, Miller School of Medicine,
University of Miami, Miami, Florida, ††Division of Preventive Medicine, Department of Medicine,
Brigham and Women’s Hospital and Harvard Medical School, Boston, Massachusetts, and ‡‡Division
of Cardiovascular Medicine, Department of Medicine, Brigham and Women’s Hospital and Harvard
Medical School, Boston, Massachusetts.BACKGROUND QRS duration and corrected QT (QTc) interval have
been associated with sudden cardiac death (SCD), but no data are
available on the significance of repolarization component (JTc in-
terval) of the QTc interval as an independent risk marker in the gen-
eral population.
OBJECTIVE In this study, we sought to quantify the risk of SCD
associated with QRS, QTc, and JTc intervals.
METHODS This study was conducted using data from 3 population
cohorts from different eras, comprising a total of 20,058 individuals.
The follow-up period was limited to 10 years and age at baseline to
30–61 years. QRS duration and QT interval (Bazett’s) were measured
from standard 12-lead electrocardiograms at baseline. JTc interval
was defined as QTc interval – QRS duration. Cox proportional hazards
models that controlled for confounding clinical factors identified at
baseline were used to estimate the relative risk of SCD.
RESULTS During a mean period of 9.7 years, 207 SCDs occurred
(1.1 per 1000 person-years). QRS duration was associated with aFunding Sources: This work was supported by grants from the Finnish Foundat
tila), Sigrid Juselius Foundation (Drs Tikkanen, Aro, and Junttila), Paavo Nurmi F
(Dr Tikkanen), FinnishMedical Foundation (Drs Tikkanen, Aro, and Niiranen), Aca
and Porthan). Disclosures: Dr Salomaa has received honoraria from Novo Nordisk
ongoing research collaboration with Bayer Ltd (all unrelated to the present study). T
and correspondence: Dr Juhani Junttila, Research Unit of Internal Medicine, Med
P.O. Box 5000, Oulu 90014, Finland. E-mail address: juhani.junttila@oulu.fi.
1547-5271/© 2022 Heart Rhythm Society. This is an open access article under the
(http://creativecommons.org/licenses/by/4.0/).significantly increased risk of SCD in each cohort (pooled hazard ra-
tio [HR] 1.030 per 1-ms increase; 95% confidence interval [CI]
1.017–1.043). The QTc interval had borderline to significant associ-
ations with SCD and varied among cohorts (pooled HR 1.007; 95% CI
1.001–1.012). JTc interval as a continuous variable was not associ-
ated with SCD (pooled HR 1.001; 95% CI 0.996–1.007).
CONCLUSION Prolonged QRS durations and QTc intervals are asso-
ciated with an increased risk of SCD. However, when the QTc interval
is deconstructed into QRS and JTc intervals, the repolarization
component (JTc) appears to have no independent prognostic value.KEYWORDS Electrocardiography; Epidemiology; Sudden cardiac
death; Depolarization; Repolarization(Heart Rhythm 2022;-:1–7) © 2022 Heart Rhythm Society. This is
an open access article under the CC BY license (http://
creativecommons.org/licenses/by/4.0/).ion for Cardiovascular Research (Drs Tikkanen, Porthan, Salomaa, and Junt-
oundation (Drs Tikkanen, Eranti, and Niiranen), Instrumentation Foundation
demy of Finland (Dr Junttila), and Orion Research Foundation (Drs Tikkanen
and Sanofi for consulting and travel support from Novo Nordisk. He also has
he rest of the authors report no conflicts of interest.Address reprint requests
ical Research Center Oulu, University of Oulu and Oulu University Hospital,
CC BY license https://doi.org/10.1016/j.hrthm.2022.04.016
2 Heart Rhythm, Vol-, No-,- 2022Introduction
Sudden cardiac death (SCD) remains a frequent cause of pre-
mature death in Western countries, with an annual incidence
estimated in the range of 50–100 SCDs per 100,000 per-
sons.1,2 Despite the improvements in heart disease prevention
and treatment, the prediction and prevention of SCD remains
a clinical and public health challenge. The majority of SCDs
occur in the general population, and up to 50% are the first
manifestation of previously unrecognized heart disease.3
Thus, improved risk stratification is urgently required and
as SCD is primarily a result of electrical disturbance of the
normal cardiac rhythm, the standard 12-lead electrocardio-
gram (ECG) remains an attractive, inexpensive, and easily
applied noninvasive tool beyond clinical risk factors for po-
tential risk prediction.
Abnormalities in depolarization and repolarization of the
myocardium are reflected in the ECG. Changes could be
related to structural abnormalities in the heart affecting electri-
cal activity or functional electrophysiological abnormalities.
The duration of the QRS complex reflect impulse propagation
andmyocardial activation properties and its possible patholog-
ical processes, whereas the QT interval is affected by multiple
factors but predominantly used as a measure of repolarization.
Both the QRS duration and the QT interval have previously
been associated with an increased risk of SCD, but the data
are somewhat limited or conflicting.4–10 This study was
designed to estimate the risk of SCD associated with these
easily obtained ECG measures in the population by analysis
of data from 3 large cohorts from different eras.Methods
Study populations and data collection
This study included data from 3 population samples gathered
in Finland: the FinnishMobile Clinic study (FMC-study), the
Mini-Finland Health Survey, and the Health 2000 Survey. Of
these, the Mini-Finland Health Survey and the Health 2000
Survey are representative samples of the adult Finnish popu-
lation. The FMC-study and the Mini-Finland Health Survey
were conducted by the Social Insurance Institution of
Finland, and the Health 2000 Survey was conducted by the
National Public Health Institute. All the data of these cohorts
are now controlled by the National Institute for Health and
Welfare. The methodology and procedures used at the base-
line examinations for each cohort have been described in
detail elsewhere.11–13 Briefly, in addition to undergoing
standard 12-lead ECG, all subjects underwent a thorough
clinical examination at baseline, including blood pressure,
body mass index (BMI), and serum cholesterol. Known dis-
eases, health behavior, and medications were also recorded,
and all participants were followed in national registries.
The FMC-study included 10,962 individuals (52% men)
aged 30–61 years drawn from 12 community populations,
and participants were enrolled between 1966 and 1972. The
Mini-Finland Health Survey included 7217 individuals
(51% men) aged over 30 years, with enrollment done be-
tween 1978 and 1980 in 40 study areas around the country.Similarly, the Health 2000 Survey included 6354 individuals
(51% men) aged over 30 years and enrollment was done be-
tween 2000 and 2001 in 80 regions in Finland.
In order to standardize the analyses and minimize the in-
fluence of changing risks of death, we limited the follow-
up period to 10 years. Individuals aged over 61 years were
excluded from the Mini-Finland Health Survey and the
Health 2000 cohorts, as the inclusion age for the FMC-
study had been 30–61 years. After other exclusions for
missing or uninterpretable data, left or right bundle branch
block, and atrial fibrillation on the ECG, the FMC-study
cohort included 10,770 subjects, the Mini-Finland Health
Survey 5030 subjects, and the Health 2000 cohort 4258 indi-
viduals. Thus, the pooled study cohort consisted of 20,058 in-
dividuals (49.9% men; mean age 44 6 12 years). Our study
was approved by the institutional review board.Outcome assessment
Participants in each of the study cohorts were continuously
monitored for major events through extensive national regis-
ters. Monitored events included myocardial infarction and all
other hospitalizations and death. Causes of death were deter-
mined by examining detailed death certificates from Statistics
Finland. Every death in the country is recorded, and the qual-
ity as well as the reliability of these registers has been vali-
dated previously.14,15
To identify the cases of sudden death due to presumed
arrhythmia, all deaths due to cardiac causes have been eval-
uated and adjudicated by an independent committee of expe-
rienced clinicians. The primary end point of this study was
sudden death due to arrhythmia, with definite and probable
deaths due to arrhythmia included in the SCD end point. In
FMC-study and Mini-Finland Health Survey, SCDs were
classified according to the definitions described in the Car-
diac Arrhythmia Pilot Study.16 In this definition, a sponta-
neous cessation of respiration and circulation with loss of
consciousness is considered as death from arrhythmia in
one of the following situations: witnessed and instantaneous,
without new or accelerating symptoms; witnessed and pre-
ceded or accompanied by symptoms attributable to myocar-
dial ischemia in the absence of heart failure; witnessed and
preceded by symptoms attributable to cardiac arrhythmia
(syncope); and unwitnessed but without evidence of another
cause. Cardiac arrests during hospitalizations for severe
congestive heart failure were not classified as SCDs. In
Health 2000 Survey, SCDs were classified similarly, as
described elsewhere in more detail.17ECG measurements
At baseline, standard resting 12-lead ECGs were recorded at
a paper speed of 50 mm/s and stored for possible further an-
alyses. In the FMC-study, the ECG intervals were measured
from paper ink tracings; in the Mini-Finland Health Survey,
the intervals were measured digitally (paper tracings were
transformed into digital format).18 In the Health 2000 Survey,
the ECG intervals were measured digitally from native digital
Tikkanen et al Risk of SCD Associated With QRS and JTc Intervals in the ECG 3tracings and a custom-made software was used to measure
QT intervals.19 The measurements were reviewed on-
screen in a blinded fashion. All ECGs were analyzed blinded
to both the clinical and the outcome data.
QRS duration was defined as the maximum duration in
leads II and V5 in the FMC-study and the Mini-Finland
Health Survey and as the mean of all leads in the Health
2000 Survey. The QT interval was measured by tangent in
leads II and V5, and the maximum value was used for ana-
lyses and corrected for heart rate using the Bazett formula
(corrected QT [QTc] 5 QT/OR-R). JTc interval was defined
as QTc interval – QRS duration.Statistical analyses
Continuous variables are presented as mean6 SD, and cate-
gorical variables are presented as percentage in each group.
Linear regression was used to compare the mean values for
continuous variables and the prevalence of categorical vari-
ables between the no SCD and SCD groups in the 3 different
cohorts. The reported P values are 2-sided. The hazard ratios
(HRs) and 95% confidence intervals (CIs) for death were
calculated with Cox proportional hazards models, and multi-
variable models were adjusted for confounding factors,
including age, gender, BMI, smoking status, history of hy-
pertension, history of myocardial infarction, history of dia-
betes mellitus, serum total cholesterol, heart rate, and left
ventricular hypertrophy by Sokolow-Lyon ECG criteria.
These confounding factors were selected a priori as the
most common and significant clinical risk factors for SCD.
Model discrimination was further assessed with C-statistics
and integrated discrimination increment (IDI).
A 2-step meta-analysis was performed. Analyses were
performed separately in each individual study. Heterogeneity
across study-specific HRs was tested using the Q statistic.
The cohort-specific logs of HRs were weighted by the inverse
of their variance, and an overall pooled estimate of the R-R
interval was computed using a random effects model.20
The P value for the test of trend was based on a Wald test
of the pooled estimates. In addition, Cox models with penal-
ized splines were used to assess and plot the relationships of
continuous QRS, QTc, and JTc intervals to SCD risk. Statis-
tical analyses were performed using SAS version 9.3 (SAS
Institute, Cary, NC) and R Statistics version 3.4.3 (The R
Foundation for Statistical Computing, Vienna, Austria). All
reported P values are 2-sided, with a P value of less than
.05 considered to indicate statistical significance. In addition,
we applied the Fine and Gray competing risk model for the
assessment of adjusted and unadjusted HRs and their 95%
CIs in the competing risk regression model, where all-
cause mortality was used as the competing event.
Adjustments in the multivariate model included the same
confounding factors as the Cox model.Results
In the pooled cohort of 20,058 participants, a total of 1757
deaths, including 207 SCDs, occurred during a meanfollow-up period of 9.7 6 1.4 years (113 SCDs in the
FMC-study, 39 in the Mini-Finland Health Survey, and 55
in the Health 2000 Survey, respectively), yielding an inci-
dence of approximately 1.1 SCDs per 1000 person-years.
In all cohorts, when comparing those who suffered SCD
with the rest of the cohort, they were more often male and
were older. Cohort-specific differences among groups are
presented in Table 1.
QRS duration and the risk of SCD
The mean QRS duration was significantly longer in SCD
cases than in survivors in each cohort (P , .001) (Table 1).
When analyzed as a continuous variable, QRS duration
was significantly related to the risk of SCD (Table 2 and
Figure 1A). The elevation in the relative risk of SCD associ-
ated with every millisecond increase in QRS duration varied
from 1.2% to 3.6% between the cohorts, averaging to 3.0%
(HR 1.030; 95% CI 1.017–1.043) when all cohorts were
pooled. When examined according to quintiles in the pooled
cohort, participants in the highest vs lowest quintile of QRS
duration had a nearly 2-fold adjusted increased risk of SCD
(Table 3).
QTc interval and the risk of SCD
The mean QTc interval was significantly longer in SCD cases
than others in all cohorts (Table 1). An increase in QTc inter-
val was associated with a significantly higher adjusted risk of
SCD in the FMC-study and the Mini-Finland Health Survey
cohorts, but not in the Health 2000 Survey cohort (Table 2).
When all data were pooled, each millisecond increase in QTc
interval predicted a 0.7% higher adjusted risk of SCD (HR
1.007; 95% CI 1.001–1.012) (Table 2 and Figure 1B). The
highest vs lowest QTc quintile had a nearly 2-fold adjusted
risk of SCD (Table 3).
JTc interval and the risk of SCD
The mean JTc interval did not statistically differ between
those with SCD and the rest of the cohorts in the FMC-
study and Mini-Finland Health Survey, but was statistically
longer in SCD cases than others in the Health 2000 cohort
(Table 1). In the pooled cohort, JTc interval did not predict
SCD (Tables 2 and 3 and Figure 1C).
Risk prediction value and competing risk
To examine the incremental risk prediction value of QRS,
QTc, and JTc intervals, we analyzed C-index and IDI for
the pooled cohort. The IDI analysis for QRS duration re-
vealed a minor improvement in risk prediction for SCD
(IDI 0.0035; P 5 .015) (Online Supplemental Table 1),
whereas QTc and JTc intervals had no improvement
compared to the original model (IDI 20.0004; P 5 .89 and
IDI 20.0001; P 5 .10, respectively) (Online Supplemental
Table 1). Additionally, none of the examined ECG intervals
showed a significant improvement in C-index for SCD
compared to the original model (Online Supplemental
Table 2).
Table 1 Baseline characteristics of the study cohorts
Characteristic
Finnish Mobile Clinic study Mini-Finland Health Survey Health 2000 Survey
No SCD SCD P No SCD SCD P No SCD SCD P
No. of subjects 10,657 113 4991 39 4203 55
Age (y) 44 6 8 52 6 7 ,.001 44 6 9 51 6 7 ,.001 45 6 8 52 6 6 ,.001
Male gender 51.9 85.9 ,.001 48.5 82.1 ,.001 47.0 89.1 ,.001
BMI (kg/m2) 26 6 4 26 6 4 .59 26 6 4 27 6 5 .07 26 6 5 28 6 5 .005
Current smoker 33.7 62.9 ,.001 27.1 53.9 ,.001 26.1 65.5 ,.001
Hypertension 4.2 14.2 ,.001 9.4 28.3 ,.001 35.3 65.5 ,.001
History of MI 1.1 2.7 .11 1.8 25.7 ,.001 0.6 5.5 ,.001
Diabetes mellitus 1.8 5.4 .004 2.7 23.1 ,.001 3.6 9.1 .03
Total cholesterol (mmol/l) 6.5 6 1.3 6.9 6 1.4 ,.001 6.8 6 1.3 7.2 6 1.2 .10 5.8 6 1.1 6.5 6 1.0 ,.001
Heart rate (beats/min) 76 6 15 76 6 15 .75 67 6 13 72 6 16 .02 63 6 11 69 6 13 ,.001
QRS duration (ms) 85 6 11 90 6 14 ,.001 88 6 11 96 6 15 ,.001 93 6 9 97 6 11 .005
QTc interval (ms) 408 6 27 414 6 33 .027 412 6 28 422 6 37 .03 400 6 23 410 6 23 .001
JTc interval (ms) 323 6 29 323 6 34 .95 324 6 30 327 6 39 .54 306 6 25 313 6 21 .04
LVH by SL 31.3 49.6 ,.001 11.6 30.8 ,.001 8.0 14.6 .08
Values are presented as mean 6 SD or percentage.
QRS MAX II/V5 5 longest QRS duration in lead II or V5. QTc MAX II/V5, tangent, Bazett’s 5 longest Bazett equation corrected QT interval measured with
tangent method in lead II or V5.
BMI 5 body mass index; JTc 5 repolarization component of the corrected QT interval (JTc 5 QTc 2 QRS); LVH by SL 5 left ventricular hypertrophy by
Sokolow-Lyon ECG criteria; MI 5 myocardial infarction; QTc 5 corrected QT; SCD 5 sudden cardiac death.
4 Heart Rhythm, Vol-, No-,- 2022In univariate models accounting for the competing risk of
non-SCD mortality, QRS and QTc intervals were associated
with SCD. In competing risk models further adjusted for clin-
ical risk factors, only QRS duration retained a significant
association with SCD (Table 4).Discussion
In these 3 population cohort studies performed at separate
time points over 30 years, QRS duration was found to be
consistently and independently associated with SCD in a
linear fashion. The QTc interval demonstrated a more modest
statistical association with SCD, but when the QTc interval
was deconstructed into QRS and JTc intervals, the repolariza-
tion component (JTc) was no longer associated with SCD
risk. This finding is novel and merits further discussion.
To date, this is the largest study to quantify the association
between QTc, QRS, and JTc intervals and SCD risk. We
investigated 3 cohorts collected in 3 different eras: late
1960s, late 1970s, and early 2000. Although the diagnostics
and treatment of cardiovascular disease have improved overTable 2 HRs of SCD for continuous variables
Model
Continuous, per ms HR (95% CI)
FMC Mini-Finland HECG measure
Univariable QRS duration 1.046 (1.02921.063) 1.056 (1.029–
QTc interval 1.008 (1.00121.014) 1.010 (1.002–
JTc interval 1.000 (0.994–1.007) 1.003 (0.993–
Multivariable QRS duration 1.034 (1.01721.051) 1.036 (1.008–
QTc interval 1.006 (0.99821.014) 1.006 (0.996–
JTc interval 0.999 (0.991–1.007) 1.001 (0.990–
Covariates in the multivariate analyses: age, gender, body mass index, current
prior acute myocardial infarction (yes/no), heart rate, and left ventricular hypertr
CI5 confidence interval; ECG5 electrocardiographic; FMC5 Finnish Mobile Cl
interval (JTc 5 QTc 2 QRS); QTc 5 corrected QT; SCD 5 sudden cardiac death.the decades, the clinical risk profiles for SCD have generally
remained unchanged.
In our analysis, QTc interval was significantly associated
with SCD, although some discrepancy was observed between
cohorts after adjustments for all clinical factors. Further, in
our analysis the QTc interval did not improve the original
risk prediction model in C-statistics. In older populations,
QTc prolongation has been associated with a 3-fold risk of
SCD,6,7 whereas in patients with coronary heart disease,
those with a prolonged QTc interval but without diabetes
or QT-prolonging drugs had up to a 5-fold risk of SCD.8
QTc prolongation has also been associated with SCD risk
when observed during Holter monitoring.9 However, no as-
sociation between QTc interval and SCD was observed in
the Framingham study.10
The QT interval incorporates both depolarization (QRS
complex) and repolarization (JT interval) components.
When analyzed separately, the distribution of durations of
the QRS complexes generated a linear increase in SCD
risk. The independent SCD predictive value of JTc interval,
assessed first time in the population study, was borderlineP for
heterogeneityealth Health 2000 Pooled
1.082) 1.040 (1.013–1.069) 1.047 (1.034–1.059) .75
1.019) 1.018 (1.007–1.028) 1.011 (1.005–1.016) .30
1.013) 1.010 (1.000–1.021) 1.003 (0.998–1.009) .27
1.066) 1.012 (0.984–1.040) 1.030 (1.017–1.043) .37
1.016) 1.009 (0.998–1.021) 1.007 (1.001–1.012) .89
1.011) 1.008 (0.995–1.020) 1.001 (0.996–1.007) .52
smoker (yes/no), arterial hypertension (yes/no), diabetes mellitus (yes/no),
ophy by Sokolow-Lyon ECG criteria (yes/no).
inic; HR5 hazard ratio; JTc5 repolarization component of the corrected QT
Figure 1 Adjusted hazard ratios for sudden cardiac death for QRS, corrected QT (QTc), and repolarization component of the QTc interval (JTc) intervals.
SCD 5 sudden cardiac death.
Tikkanen et al Risk of SCD Associated With QRS and JTc Intervals in the ECG 5or nonsignificant across the studied cohorts. We have earlier
reported that prolongation of the QRS complex without
bundle branch blocks associates with an increased risk of
SCD in one of the cohorts (the FMC-study) of the present
study. Additionally, a similar association has been presented
in a separate population of middle-aged men.4,5 When we as-
sessed the incremental risk prediction value of QRS and JTc
intervals in the present study, only QRS interval improved
the original model with robust clinical risk factors (age,Table 3 HRs of SCD for quintiles in the pooled cohort
Model
HR (95% CI)
Q1 Q2 Q3
QRS interval
range (ms)
60–78 80 82–88
QTc interval
range (ms)
304–384.8 384.9–399.6 399.7–4
JTc interval
range (ms)
220–295.1 295.2–312.1 312.2–3
Univariable QRS interval 1 1.132 (0.595–2.154) 0.974 (
QTc interval 1 0.967 (0.575–1.626) 1.577 (
JTc interval 1 1.564 (0.575–4.254) 1.073 (
Multivariable QRS interval 1 1.021 (0.535–1.951) 0.862 (
QTc interval 1 0.926 (0.546–1.570) 1.575 (
JTc interval 1 1.396 (0.557–3.501) 1.038 (
Covariates in the multivariate analyses: age, gender, body mass index, current
prior acute myocardial infarction (yes/no), heart rate, and left ventricular hypertr
CI5 confidence interval; HR5 hazard ratio; JTc5 repolarization component of
2/3/4/5; QTc 5 corrected QT; SCD 5 sudden cardiac death.gender, BMI, smoking status, history of hypertension, his-
tory of myocardial infarction, history of diabetes mellitus,
serum total cholesterol, heart rate, and left ventricular hyper-
trophy by Sokolow-Lyon ECG criteria). We also now show
that those with longer QRS duration were at increased risk
of SCD without meaningful elevations of competing deaths.
On the basis of the present results, one can argue that at
least in the general population with a low absolute event
rate of SCD, prolongation of depolarization is moreP for
trendQ4 Q5
90–98 100–144
12.9 413–428.7 428.8–703
26.5 326.6–343.8 343.9–631
0.358–2.649) 1.763 (1.006–3.089) 2.910 (1.664–5.087) ,.001
0.996–2.497) 1.525 (0.959–2.426) 2.033 (1.299–3.183) ,.001
0.536–2.147) 1.477 (0.791–2.755) 1.129 (0.707–1.802) .23
0.314–2.364) 1.281 (0.723–2.269) 1.793 (1.011–3.181) ,.001
0.977–2.539) 1.508 (0.917–2.478) 1.728 (1.034–2.887) .02
0.539–2.000) 1.420 (0.794–2.539) 1.091 (0.636–1.872) .65
smoker (yes/no), arterial hypertension (yes/no), diabetes mellitus (yes/no),
ophy by Sokolow-Lyon ECG criteria (yes/no).
the corrected QT interval (JTc5 QTc2 QRS); Q1/Q2/Q3/Q4/Q55 quintile 1/
Table 4 Competing risk regression model (pooled cohort)
Variable
Univariate Multivariate
HR (95% CI) P HR (95% CI) P
QRS
duration
1.03 (1.02–1.04) ,.001 1.015 (1.006–1.025) .0013
QTc interval 1.01 (1.01–1.02) ,.001 1.003 (0.999–1.008) .16
JTc interval 1.01 (1.0–1.01) .0079 1.00 (0.995–1.005) .92
Covariates in the multivariate analyses: age, gender, body mass index,
current smoker (yes/no), arterial hypertension (yes/no), diabetes mellitus
(yes/no), prior acute myocardial infarction (yes/no), heart rate, and left
ventricular hypertrophy by Sokolow-Lyon ECG criteria (yes/no).
CI 5 confidence interval; HR 5 hazard ratio; JTc 5 repolarization
component of the corrected QT interval (JTc 5 QTc 2 QRS); QTc 5 cor-
rected QT.
6 Heart Rhythm, Vol-, No-,- 2022significant than moderate prolongation of the repolarization
component. Thus, although completely speculative, it is
somewhat reasonable to speculate that in subjects with a
slightly prolonged QT interval due to prolonged QRS dura-
tion, strict restrictions on QT-affecting medications might
not be required. The QTc interval is not only a measure of de-
polarization and repolarization but also reflective of the de-
gree of underlying structural heart disease.21 Whether the
risk of SCD associated with QTc prolongation is actually
due to an abnormality of cardiac depolarization merits further
investigation.
Although this study has strengths in volume and method-
ology, several limitations require attention. First, there was
modest variance in the ECG collection as they were ob-
tained in different eras and errors in measurements due to
technical issues cannot be ruled out. In Health 2000 popula-
tion, ECGs were digitally recorded, which could cause dif-
ferences in analysis and may explain some of the differences
between Health 2000 cohort and the other cohorts in this
study. Second, the measurements were performed only at
baseline and both QRS and QTc intervals could have
changed if recordings were repeated during follow-up.
Thus, we limited the follow-up period to 10 years, since it
is reasonable to hypothesize that the risk conferred by a sin-
gle parameter that is subject to change over time may also
change. This limitation in follow-up and limited number
of SCD events could have negatively affected the statistical
power of the study and its results. Third, our study included
only middle-aged individuals of Caucasian origin and thus
the results cannot be extrapolated to other patient groups.
Nevertheless, middle-aged populations might yield the
greatest benefit in SCD prevention as compared with older
populations because of life expectancy and compared with
younger populations because of low event rates. Fourth,
the cohorts were obtained in 3 different eras and the diag-
nostics and treatment of cardiovascular disease have
improved significantly over time and pooling the data might
have yielded some errors owing to variability in the cohort
data. It should also be acknowledged that our analyses
included only the most commonly used formula in clinical
settings (Bazett’s) for the QT interval, while the presentguidelines prefer other formulas for heart rate adjust-
ments.22 Also, the formal test between different ECG
indices studied could not be statistically performed because
of interactions, although the major conclusion of the study
was that QRS interval should be identified as a significant
part of the risk related to prolonged QT interval, and in
our view this conclusion can be drawn with the analyses
chosen for the study. Additionally, no sufficient medication
data were available for all cohorts and thus there is a possi-
bility of confounding effects of drugs on the QT interval.
Finally, we could not adjust for left ventricular systolic
function as no data were available, which might have
affected the results.Conclusion
Our data suggest that depolarization abnormality detected as
a prolongation of QRS duration is an important and easily ob-
tained independent predictor of SCD in the general popula-
tion, but prolonged repolarization measured as JTc interval
has no independent prognostic value.Appendix
Supplementary data
Supplementary data associated with this article can be found
in the online version at 10.1016/j.hrthm.2022.04.016.References
1. Chugh SS, Reinier K, Teodorescu C, et al. Epidemiology of sudden cardiac death:
clinical and research implications. Prog Cardiovasc Dis 2008;51:213–228.
2. Hayashi M, Shimizu W, Albert CM. The spectrum of epidemiology underlying
sudden cardiac death. Circ Res 2015;116:1887–1906.
3. Myerburg RJ, Goldberger JJ. Sudden cardiac arrest risk prediction: population
science and the individual risk mandate. JAMA Cardiol 2017;2:689–694.
4. Aro AL, Anttonen O, Tikkanen JT, et al. Intraventricular conduction delay in a
standard 12-lead electrocardiogram as a predictor of mortality in the general pop-
ulation. Circ Arrhythm Electrophysiol 2011;5:704–710.
5. Kurl S, M€akikallio TH, Rautaharju P, et al. Duration of QRS complex in resting
electrocardiogram is a predictor of sudden cardiac death in men. Circulation 2012;
125:2588–2594.
6. Dekker JM, Schouten EG, Klootwijk P, et al. Association between QT interval
and coronary heart disease in middle-aged and elderly men: the Zutphen Study.
Circulation 1994;90:779–785.
7. Straus SM, Kors JA, De Bruin ML, et al. Prolonged QTc interval and risk of sud-
den cardiac death in a population of older adults. J Am Coll Cardiol 2006;
47:362–367.
8. Chugh SS, Reinier K, Singh T, et al. Determinants of prolonged QT interval and
their contribution to sudden death risk in coronary artery disease: the Oregon Sud-
den Unexpected Death Study. Circulation 2009;119:663–670.
9. Algra A, Tijssen JG, Roelandt JR, et al. QTc prolongation measured by standard
12-lead electrocardiography is an independent risk factor for sudden death due to
cardiac arrest. Circulation 1991;83:1888–1894.
10. Goldberg RJ, Bengtson J, Chen ZY, et al. Duration of the QT interval and total
and cardiovascular mortality in healthy persons (the Framingham Heart Study
experience). Am J Cardiol 1991;67:55–58.
11. Heistaro S, ed. Methodology Report: Health 2000 Survey. Helsinki: Publications
of National Public Health Institute B26/2008; 2008. http://urn.fi/URN:NBN:fi-fe2
01204193320. Accessed September 20, 2021.
12. Knekt P, Rissanen H, J€arvinen R, Heli€ovaara M. Cohort profile: the Finnish Mo-
bile Clinic Health surveys FMC, FMCF and MFS. Int J Epidemiol 2017;
46:1760–1761.
13. Aromaa A, Heli€ovaara M, Impivaara O, Knekt P, Maatela J, eds. The execution of
the Mini-Finland Health Survey. Helsinki and Turku, 88. Publications of the So-
cial Insurance Institution ML; 1989 (in Finnish, with English summary). http://
hdl.handle.net/10138/162731. Accessed September 20, 2021.
Tikkanen et al Risk of SCD Associated With QRS and JTc Intervals in the ECG 714. Pajunen P, Koukunen H, Ketonen M, et al. The validity of the Finnish Hospital
Discharge Register and Causes of Death Register data on coronary heart disease.
Eur J Cardiovasc Prev Rehabil 2005;12:132–137.
15. Rapola JM, Virtamo J, Korhonen P, et al. Validity of diagnoses of major coronary
events in national registers of hospital diagnoses and deaths in Finland. Eur J Epi-
demiol 1997;13:133–138.
16. Greene HL, Richardson DW, Barker AH, et al. Classification of deaths after
myocardial infarction as arrhythmic or nonarrhythmic (the Cardiac Arrhythmia
Pilot Study). Am J Cardiol 1989;63:1–6.
17. Lahtinen AM, Noseworthy PA, Havulinna AS, et al. Common genetic variants
associated with sudden cardiac death: the FinSCDgen study. PLoS One 2012;
7:e41675.
18. Holkeri A, Eranti A, Kentt€a TV, et al. Experiences in digitizing and digitally
measuring a paper-based ECG archive. J Electrocardiol 2018;51:74–81.19. Porthan K, Marjamaa A, Viitasalo M, et al. Relationship of common candidate
gene variants to electrocardiographic T-wave peak to T-wave end interval and
T-wave morphology parameters. Heart Rhythm 2010;7:898–903.
20. DerSimonian R, Laird N. Meta-analysis in clinical trials. Control Clin Trials
1986;7:177–188.
21. Moss AJ. QTc prolongation and sudden cardiac death: the association is in the
detail. J Am Coll Cardiol 2006;47:368–369.
22. Rautaharju P, Surawicz B, Gettes L, et al. AHA/ACCF/HRS recommendations
for the standardization and interpretation of the electrocardiogram: Part IV: the
ST segment, T and U waves, and the QT interval: a scientific statement from
the American Heart Association Electrocardiography and Arrhythmias Commit-
tee, Council on Clinical Cardiology; the American College of Cardiology Foun-
dation; and the Heart Rhythm Society. Endorsed by the International Society for
Computerized Electrocardiology. J Am Coll Cardiol 2009;53:982–991.