Full text of DOI:10.1038/sj.jhh.1001129

Journal of Human Hypertension (2001) 15, 63–70
2001 Macmillan Publishers Ltd All rights reserved 0950-9240/01 $15.00
www.nature.com/jhh
ORIGINAL ARTICLE
The spatial QRS-T angle as a marker of
ventricular repolarisation in hypertension
P Dilaveris, E Gialafos, A Pantazis, A Synetos, F Triposkiadis, and J Gialafos
The State Department of Cardiology, Hippokration Hospital, Athens, Greece
Ventricular repolarisation abnormalities are important in
arrhythmia provocation. The dispersion of repolaris-
ation duration is not the only aspect of repolarisation
heterogeneity. Spatial vectorcardiographic descriptors
constitute a novel approach to quantify ventricular
repolarisation. To test the ability of vectorcardiographic
descriptors to discriminate between hypertensives with
high or low blood pressure (BP), 110 treated hyperten-
sives (mean age 63.6 ؎ 12.1 years) were classified in
the high (systolic BP у160 mm Hg or diastolic BP у95
mm Hg) (n ؍ 67), or the low (systolic BP Ͻ 160 mm Hg
and diastolic BP Ͻ95 mm Hg) (n ؍ 43) BP group. The
maximum QT, JT, and T peak–T end intervals and the
QT, JT, and T peak–T end dispersion were calculated
from a digitally recorded 12-lead electrocardiogram
(ECG). X, Y, and Z leads were reconstructed from the
12-lead ECG. The amplitude of the maximum spatial T
vector (spatial T amplitude), the angle between the
maximum spatial QRS and T vectors (spatial QRS-T
angle) and the frontal plane QRS-T angle were calcu-
lated. The spatial QRS-T angle was higher in patients
with high compared to those with low BP (P ؍ 0.025).
All conventional ECG markers of the dispersion of ven-
tricular repolarisation duration failed to demonstrate
significant differences between hypertensives with high
or low BP. In conclusion, the spatial QRS-T angle was
significantly increased in those treated hypertensive
patients who showed repeatedly high BP values. Hence,
we may suggest that the angle between the directions
of ventricular depolarisation and repolarisation is a
sensitive marker of the repolarisation alterations in sys-
temic hypertension.
Journal of Human Hypertension (2001) 15, 63–70
Keywords: spatial QRS-T angle; QT dispersion; repolarisation; blood pressure; electrocardiogram; vectorcardiogram
hence, may reduce its power to assess arrhythmia
Introduction
risk prospectively.3,7
Systemic hypertension affects up to 25% of the
The dispersion of repolarisation duration seems to
adult population and is widely recognised to be a
consist only one aspect of the repolarisation inhom-
potent cardiovascular risk factor.1 Left ventricular
ogeneity. Several studies have already attempted to
hypertrophy in hypertensive patients results in
quantify the spatial aspects of ventricular repolaris-
inhomogeneity of ventricular repolarisation, favour-
ation.8–11 The spatial aspects of the T wave com-
ing the propensity to ventricular tachyarrhythmias.2
plexity have been evaluated by the so-called princi-
The variability in the QT interval duration among
pal component analysis,8 while T axis has been
the different leads of a surface 12-lead electrocardio-
reported to be a strong risk indicator of cardiac
gram (ECG) (QT dispersion) is supposed to reflect
events in the elderly.9 Furthermore, a recent study
local differences in the recovery time of the myocar-
concluded that QT dispersion is largely determined
dium.3 High QT and JT dispersion values have been
by the T-loop morphology, as expressed by T-loop
previously reported in patients with systemic hyper-
amplitude and width,10 whereas an older one has
reported a widened QRS-T angle in patients with
eccentric left ventricular hypertrophy.11
tension.1,2,4,5 Several studies have postulated that
QT dispersion values decrease when adequate blood
pressure (BP) control is achieved.5,6 However, well-
known difficulties in defining the end of the T wave,
the inability to measure the QT interval in all leads,
and the possible presence of U waves may contrib-
ute to poor reproducibility of QT dispersion and,
The objective of the present study was to evaluate
several temporal and spatial descriptors of ventricu-
lar repolarisation in a population of treated hyper-
tensive patients. In particular, our study investi-
gated the ability of repolarisation descriptors to
discriminate between hypertensive patients with
high or low measured BP values.
Correspondence: Dr Polychronis E Dilaveris, 22 Miltiadou Street,
GR-15561 Holargos, Athens, Greece. E-mail: hrodilȰyahoo.com
Received 31 March 2000; revised 24 June 2000; accepted 20
July 2000

QRS-T angle alterations in hypertension
P Dilaveris et al
64
stored ECGs displayed on a high-resolution com-
puter screen. Each lead was separately magnified
(160 mm/s and 60 mm/mV). On screen calipers were
used to define the start and end of all the aforemen-
tioned ECG intervals separately. The QT interval
was measured from the onset of the QRS complex
to the end of the T wave. The point of T-wave offset
was defined as the return to the baseline.3 If a U
wave followed the T wave without an isoelectric
separation, the end of the T wave was taken as the
nadir between the T and U waves. If the end of the
T wave could not be reliably determined or when
the T wave was of very low amplitude (Ͻ50 ␮V), QT
measurements were not made and the lead excluded
from analysis.10 No attempt was made to correct for
missing leads.3 The JT interval (from the J point to
the end of the T wave) was calculated from the equ-
ation JT ϭ QT − QRS, and the T peak–T end interval
was measured from the apex of the T wave to its
end.2 All measurements were performed by two
independent investigators who were blinded to the
clinical data of the patients. The averages of the
measurements of the two observers were used for
comparisons.
Methods
Study population
The study population consisted of 110 consecu-
tively recruited patients (62 women; mean age 63.6
Ϯ 12.1 years) with history of systemic hypertension.
All patients received antihypertensive therapy and
no one had any other risk factors for coronary artery
disease. All patients underwent physical examin-
ation, 12-lead ECG, and serial blood tests. Apart
from systemic hypertension, no other cardiovascular
disease was present in the study population. All
patients were in sinus rhythm. Routine medications
were not withheld during patients evaluation.
Excluded from the study were patients with left
or right bundle branch block, atrioventricular block,
ventricular pre-excitation, history of coronary artery
disease, atrial fibrillation, sick sinus syndrome, prior
pacemaker implantation, clinically overt heart fail-
ure (New York Heart Association (NYHA) classes II-
IV), or pericarditis. Patients receiving digitalis or
any anti-arrhythmic drugs were also excluded.
Repeated BP measurements performed in the out-
patient clinic were used to stratify the study patients
according to their BP levels. The average from the
BP measurements obtained in three different patient
visits within the same week was used for this pur-
pose. Patients were categorised in the high BP group
when the average systolic BP was у160 mm Hg or
diastolic BP у95 mm Hg and in the low BP group
when the average systolic BP was Ͻ160 mm Hg and
diastolic BP Ͻ95 mm Hg. The study was approved
by the local ethics committee and informed consent
was obtained from all participants.
Definition of ECG indices of ventricular
repolarisation
The following indices were derived from each
measurement of each ECG:
(1) The maximum QT, JT, and T peak–T end inter-
vals in any measurable leads (QT maximum, JT
maximum, and maximum
respectively).
T peak–T end,
(2) QT, JT, and T peak–T end dispersion, defined
as the difference between QT maximum and the
minimum QT interval (QT dispersion), JT
maximum and the minimum JT interval (JT
dispersion), maximum T peak–T end and the
minimum T peak–T end interval (T peak–T end
dispersion) in any measurable leads, respect-
ively.
Twelve-lead surface electrocardiogram
In all subjects, a 12-lead digital ECG was recorded in
the supine resting position using a computer-based
electrocardiograph (Cardioperfect, version 1.1, Car-
dioControl NV, Rijswijk, The Netherlands). All 12
leads of each ECG were recorded simultaneously for
20 seconds and sampled at a rate of 1200 Hz. From
each lead, the average complex was calculated by
the MEANS (Modular ECG Analysis) system.12
These average complexes, sampled at 300 Hz, were
used in the analysis. Left ventricular hypertrophy
was electrocardiographically defined when at least
one of the following three criteria was positive: SV3
+ RaVL Ͼ2.4 mV in men or Ͼ2.0 mV in women, a
typical strain pattern, or a Romhilt-Estes score у5.13
The presence of discordant T waves in those
patients who fulfilled the ECG criteria for left ven-
tricular hypertrophy was assessed in all the 12 leads
of each ECG in both study groups.
Twelve-lead vectorcardiogram
To derive vectorcardiographic descriptors of ven-
tricular repolarisation, orthogonal X, Y, and Z leads
were reconstructed from the standard 12 ECG leads
(Figure 1).14 Let QRSx, QRSy, and QRSz be the pro-
jections of the maximum QRS vector on the X, Y,
and Z axes and QRSxy, QRSxz, and QRSyz its pro-
jections on the frontal (xy), horizontal (xz), and right
saggital (yz) planes, respectively. Similarly, let Tx,
Ty, and Tz be the projections of the maximum T vec-
tor on the X, Y, and Z axes and Txy, Txz, and Tyz
its projections on the frontal (xy), horizontal (xz),
and right saggital (yz) planes, respectively. QRSxy,
QRSxz, QRSyz, Txy, Txz, and Tyz were automati-
cally calculated by our analysis system. According
to previously published equations,11 and by use of
QT, JT, and T peak–T end interval measurements
All QT, JT, and T peak–T end interval measure-
ments were performed manually using the digitally
Journal of Human Hypertension

QRS-T angle alterations in hypertension
P Dilaveris et al
65
Figure 1 (A) Twelve-lead electrocardiogram, (B) reconstructed X, Y, and Z leads and P, QRS and T loop projections in the frontal (F),
horizontal (H), and right saggital (R) planes from a 73 years old hypertensive patient.
Journal of Human Hypertension

QRS-T angle alterations in hypertension
P Dilaveris et al
66
the Pythagorean theorem, the amplitude of the
maximum spatial T vector (Spatial T amplitude) was
calculated from the formula:
Spatial T amplitude ϭ [(Txy2 + Txz2 + Tyz2)/2]1/2
ences were noticed between the two groups con-
cerning age, sex, heart rate, the body mass index,
the duration of hypertension, the plasma potassium
levels, and the drugs administered (Table 1).
All ECG repolarisation indices were found to have
similar values between hypertensive patients with
high BP and those with low BP. Only the spatial
QRS-T angle was found to be significantly higher in
hypertensive patients with high BP values compared
to patients with low BP levels. No significant differ-
ences were noticed between the two groups con-
cerning the QRS axis in the frontal plane (Table 2).
From those patients who fulfilled the ECG criteria
for LVH, five (18%) patients with high BP and one
(10%) patient with low BP showed discordant T
waves to the QRS complexes in both precordial and
limb leads (P ϭ 0.131), while seven (25%) patients
with high BP and two (20%) patients with low BP
showed discordant T waves in the precordial leads
(P ϭ 0.456). In those patients who did not fulfill the
ECG criteria for LVH (n ϭ 72), the spatial QRS-T
angle was the only repolarisation parameter to differ
significantly, although only marginally, between
hypertensives with high BP (18.1 Ϯ 12.8 degrees)
and those with low BP (11.8 Ϯ 6.9 degrees), P ϭ
0.05. The differences in the repolarisation para-
meters between hypertensive patients with ECG
documented LVH and those without LVH are shown
in Table 3.
Furthermore, the angle (␪°) between the maximum
spatial QRS and T vectors (Spatial QRS-T angle) was
calculated from the formula:11,15
cos ␪° ϭ (QRSxTx + QRSyTy + QRSzTz)/͉QRS͉ ͉T͉
The angle between the maximum QRS and T vectors
in the frontal plane (Frontal QRS-T angle) was also
computed.
Intraobserver and interobserver variability
Intraobserver and interobserver mean percent error
(absolute difference between two observations div-
ided by the mean and expressed in percent)16 for
maximum QT, JT, and T peak–T end intervals and
for QT, JT, and T peak–T end dispersion measure-
ments were determined in 40 randomly selected
study participants. To define intraobserver errors of
measurements one of the two investigators meas-
ured the QT, JT, and T peak–T end intervals of all
40 ECGs twice.
Statistical analysis
Continuous variables are expressed as mean (Ϯ s.d.).
For comparisons between the different patient
groups, Mann-Whitney test and chi-square test were
used where appropriate. Spearman’s correlation
coefficients were used to assess the association
between different variables. To determine the multi-
variate contribution of other factors to the values of
different repolarisation indices, linear regression
equations were constructed:
Table 1 Clinical characteristics of patients with high or low
blood pressure
Characteristic
Low blood
pressure
group
High blood
pressure
group
P value
X ϭ A0 + A1BP control + A2LVH + A3Age + A4Sex
+ A5Heart rate,
(n = 43)
(n = 67)
Age (years)
Men
63.7 ± 14
51% (22)
27.5 ± 4
63.5 ± 10.9
39% (26)
27.9 ± 6.6
0.876
0.206
0.693
where X was one of the considered repolarisation
indices (QT maximum, QT dispersion, JT maximum,
JT dispersion, maximum T peak–T end, T peak–T
end dispersion, spatial T amplitude, frontal QRS-T
angle, spatial QRS-T angle) and BP control was the
presence of high or low BP during patient evalu-
ation. Left ventricular hypertrophy (LVH) and sex
were treated as categorical variables in the
regression equation. For each repolarisation index,
the statistical significance of the regression coef-
ficients A1, A2, A3, A4, and A5 was evaluated. P
values Ͻ0.05 were considered statistically signifi-
cant.
Body mass index
(kg/m2)
Heart rate (bpm)
Systolic blood
pressure (mm Hg)
Diastolic blood
pressure (mm Hg)
Duration of
hypertension (years)
Left ventricular
hypertrophy
Serum potassium
(mmol/L)
Drug therapy
Diuretics
ACE-inhibitors
Calcium-antagonists
Beta-blockers
Other
68.4 ± 12.4
138.4 ± 11.1
71.4 ± 9.7
169.5 ± 16.4
0.190
Ͻ0.001
80.7 ± 7.5
10.3 ± 4.6
23% (10)
5 ± 0.6
100.8 ± 12.8
10.7 ± 5.9
42% (28)
5.1 ± 0.6
Ͻ0.001
0.715
0.018
0.323
23% (10)
26% (11)
33% (14)
12% (5)
7% (3)
31% (21)
39% (26)
21% (14)
9% (6)
0.276
0.107
0.102
0.513
0.366
Results
Of a total enrolment of 110 hypertensive patients,
43 were categorised in the low BP group and 67 in
the high BP group. ECG documented LVH was more
prevalent in the high BP group. No significant differ-
4% (3)
Values are mean (± s.d.), or % (no.). ACE, angiotensin con-
verting enzyme.
Journal of Human Hypertension

QRS-T angle alterations in hypertension
P Dilaveris et al
67
Table 2 Electrocardiographic data of patients with high or low
angle was significantly associated with the presence
of LVH (P ϭ 0.01), while the spatial QRS-T angle
was significantly associated to sex (P ϭ 0.001) and
the presence of high or low BP in our study popu-
lation (P ϭ 0.008).
The correlation among the different repolarisation
indices is shown in Table 4.
blood pressure
Parameter
High blood
pressure
group
Low blood P value
pressure
group
(n = 43)
(n = 67)
QRS axis
Normal
85% (57)
13% (9)
2% (1)
91% (39)
7% (3)
2% (1)
0.651
0.180
1.0
Left
Right
Accuracy of the measurements
QT maximum (ms)
QT dispersion (ms)
JT maximum (ms)
JT dispersion (ms)
Maximum T peak–T
end (ms)
T peak–T end
dispersion (ms)
Spatial T amplitude
(␮V)
Frontal QRS-T angle
(degrees)
Spatial QRS-T angle
(degrees)
396 ± 26.9
44.4 ± 17.4
303.1 ± 27.6
40.9 ± 14.4
402.3 ± 37.1
39.8 ± 13.9
307.1 ± 34.6
39.7 ± 12.9
0.410
0.149
0.442
0.734
The intraobserver relative errors for QT maximum,
QT dispersion, JT maximum, JT dispersion,
maximum T peak–T end, and T peak–T end disper-
sion were 2.1 Ϯ 1.6%, 24.3 Ϯ 22.2%, 4.3 Ϯ 3.7%,
25.9 Ϯ 21.1%, 7.9 Ϯ 6.2%, and 27 Ϯ 26.1%, respect-
ively. The interobserver relative errors for the same
indices were 4.7 Ϯ 3.9%, 32.8 Ϯ 21.9%, 7.8 Ϯ 9.4%,
34 Ϯ 24.4%, 9.8 Ϯ 8.2%, and 35.2 Ϯ 31.3%, respect-
ively.
95.8 ± 17.7
54.8 ± 23.7
93.7 ± 18.8
52 ± 18.1
0.327
0.769
0.516
0.114
0.025
342.8 ± 158.1 358.4 ± 121.8
23.4 ± 37
16.8 ± 28.4
14 ± 8.1
19.8 ± 12.6
Discussion
The principal finding of this study is that the spatial
QRS-T angle is the only repolarisation marker to be
significantly increased in those treated hypertensive
patients who show repeat office measurements of
high BP. Conventional ECG markers of the disper-
sion of ventricular repolarisation duration failed to
demonstrate significant differences between hyper-
tensives with high or low BP, although they were
useful for the discrimination between hypertensives
with and those without ECG-documented LVH.
Values are mean (± s.d.), or % (no.). See the text for the definition
of the individual indices.
Table 3 Electrocardiographic data of hypertensive patients with
or without left ventricular hypertrophy
Parameter
Patients
with LVH
(n = 38)
Patients
without LVH
(n = 72)
P value
QT maximum (ms)
QT dispersion (ms)
JT maximum (ms)
JT dispersion (ms)
405.3 ± 37.7
47.9 ± 19.4
309.6 ± 37.5
44.1 ± 16.1
395.6 ± 26.9
39.8 ± 13.6
302.1 ± 25.9
38.5 ± 12.1
90.9 ± 14.3
0.159
0.024
0.235
0.044
0.005
Ventricular repolarisation abnormalities in
hypertension
Maximum T peak–T 102.7 ± 22
end (ms)
Hypertension affects approximately 25% of the
adult population, and a large percentage of these
patients develop LVH, which carries a negative
prognosis being a recognised harbinger of ventricu-
lar arrhythmias and mortality.1,2 Although the fac-
tors predisposing to electrical instability and
arrhythmic death are not well established,17 it has
been reported that myocardial hypertrophy alters
the ionic channels that are operative during the
early repolarisation phase.18 Furthermore, LVH is
characterised by an increase in collagen interstitial
matrix that may also lead to alterations in ventricu-
lar repolarisation.19
T peak–T end
62.8 ± 27.8
49 ± 15.8
0.006
0.023
0.029
0.012
dispersion (ms)
Spatial T amplitude 318 ± 186.8
(␮v)
Frontal QRS-T angle 36.2 ± 52.5
(degrees)
Spatial QRS-T angle
(degrees)
365.2 ± 114.7
12.6 ± 11.4
15.2 ± 10.9
21.3 ± 12.6
Values are mean (± s.d.), or % (no.). LVH, left ventricular
hypertrophy. See the text for the definition of the individual indi-
ces.
Associations between repolarisation indices and
clinical variables
Dispersion of repolarisation duration in
hypertension
QT maximum and JT maximum were significantly
associated with sex (P ϭ 0.009 and P Ͻ 0.001,
respectively) and heart rate (P Ͻ 0.001). JT disper-
sion (P ϭ 0.046), maximum T peak–T end (P ϭ
0.006) and T peak–T end dispersion (P ϭ 0.006)
were significantly related to the presence of ECG
documented LVH. The spatial T amplitude was sig-
nificantly associated with sex (P ϭ 0.020) and the
presence of LVH (P ϭ 0.046). The frontal QRS-T
QT dispersion has been suggested to reflect abnor-
mal ventricular repolarisation.3 High QT dispersion
values have been reported in patients with systemic
hypertension and LVH,1,2,4,5 which decrease after
the appropriate treatment of high BP.5,6 However,
the rather weak correlation between QT dispersion
and BP levels,5,6 or LVH1,2,5 and the poor reproduci-
Journal of Human Hypertension

QRS-T angle alterations in hypertension
P Dilaveris et al
68
Table 4 Correlation coefficients between electrocardiographic and vectrocardiographic indices
QT
JT
JT
Maximum
T peak–T end Spatial T
Frontal
Spatial
QRS-T angle
dispersion maximum dispersion T peak–T end
dispersion
amplitude QRS-T angle
QT maximum
QT dispersion
JT maximum
JT dispersion
Maximum
T peak–T end
T peak–T end
dispersion
Spatial T
amplitude
Frontal
QRS-T angle
0.160
0.893‡
0.019
0.131
0.153
0.287†
0.266†
0.249†
0.179
0.150
0.382‡
0.074
0.142
−0.243*
−0.036
−0.281†
−0.155
0.265†
0.087
0.252†
0.198*
0.071
0.293†
0.003
0.174
0.082
0.741‡
−0.137
0.390‡
0.321†
−0.205*
0.262†
0.198*
−0.059
0.300†
−0.233*
*P Ͻ 0.05, †P Ͻ 0.01, ‡P Ͻ 0.001. See the text for the definition of the individual indices.
bility of QT dispersion measurements,3,7 may reduce
its power to assess arrhythmia risk prospectively.4
In this study, to estimate the dispersion of repolar-
isation duration, apart from QT maximum and QT
dispersion, JT maximum and JT dispersion were
used as more accurate measures of the ventricular
repolarisation time and maximum T peak–T end
and T peak–T end dispersion were used as indices
of the transmural dispersion of repolarisation.20
However, no significant differences were noticed
between hypertensives with high and those with
low BP irrespective of the ECG index used (Table
2). The aforementioned results cannot be attributed
either to the extension of hypertension history, nor
to heart rate, nor to the plasma potassium levels, and
nor to the drugs administered (Table 1). The
inability of the dispersion indices to distinguish
hypertensives with high BP from those with low BP
may be due to the questionable accuracy of the man-
ual measurement of those indices. On the other
hand the dispersion indices, and particularly T
peak–T end dispersion, showed significant differ-
ences between hypertensives with and those with-
out ECG-documented LVH. This is in accordance
with previous reports.1,2,4,5 The associations found
between ECG and clinical variables have already
been reported.21,22
T axis has long been known to provide a global mea-
sure of repolarisation abnormality,22 and the spatial
angle between the direction of the repolarisation
and depolarisation waves (spatial QRS-T angle) has
already been used to quantify ventricular repolaris-
ation.11,23,24
In this study, the spatial QRS-T angle was the only
index found to be significantly higher in hyperten-
sives who showed repeatedly high BP values as
compared to those who had low BP levels. Previous
studies have already reported a widened spatial
QRS-T angle in patients with eccentric LVH,11 and
a decrease in the frontal or the horizontal QRS-T
angle in patients following treatment of hyperten-
sion.23 A widened QRS-T angle in patients with
LVH was also reported when Grant’s vectorial
method was used.25 Similar differences between
patients with ECG-documented LVH and those with-
out were also demonstrated in our study concerning
the spatial T amplitude, the frontal QRS-T angle and
the spatial QRS-T angle. Although most of the ECG
and the vectorcardiographic repolarisation para-
meters showed significant differences between
patients with and those without ECG-documented
LVH, only the spatial QRS-T angle was able to dis-
criminate hypertensives according to the adequacy
of the BP control (Table 3). This is in accordance
with a previous report.23 On the other hand, the spa-
tial T amplitude and the frontal QRS-T angle failed
to demonstrate significant differences between
hypertensives with high or low BP. Furthermore,
scalar ECG characteristics like the QRS axis or the
presence of discordant T waves failed to show any
differences between the two study groups.
Spatial aspects of ventricular repolarisation in
hypertension
The dispersion of repolarisation duration is not the
only important aspect of ventricular repolarisation.
The complexity of the shape of the T wave is sup-
posed to be a marker of abnormal repolarisation and
principal component analysis has been used to
quantify T waves with major morphologic abnor-
malities.8 However, subtle changes in the shape of
a monophasic T wave are not detectable by this
method. QT dispersion has recently been reported
to be an attribute of T loop morphology,10 and T axis
has been used as an alternative measure of repolaris-
ation abnormality.9 An abnormal orientation of the
The ability of the spatial QRS-T angle to dis-
tinguish hypertensives with low from those with
high BP is not probably due to the different preva-
lence of ECG-documented LVH between hyperten-
sives with low and those with high BP. This is obvi-
ous from the fact that the spatial QRS-T angle was
the only repolarisation parameter able to distinguish
the patients who did not fulfill the ECG criteria for
LVH according to their BP levels. The significant
Journal of Human Hypertension

QRS-T angle alterations in hypertension
P Dilaveris et al
69
association found between the spatial QRS-T angle
and sex has been reported previously.15
increased in those treated hypertensives who
showed repeatedly high BP values. Conventional
features of the scalar ECG and markers of the disper-
sion of ventricular repolarisation duration failed to
demonstrate significant differences between treated
hypetensives with high or low BP levels.
Hence, the spatial QRS-T angle may serve as a
marker of ventricular repolarisation heterogeneity in
hypertension, even when ECG documentation of
LVH is not obtained. The relationships between the
spatial QRS-T angle and the pathophysiology of spe-
cific repolarisation abnormalities are not estab-
lished. The occurrence of electrophysiological
changes in response to ventricular pressure or vol-
ume overload has been well documented.26 The
ability of the spatial QRS-T angle to detect those
stretch-induced electrophysiological changes, still
undetectable by the standard surface ECG, cannot be
ruled out. However, we do not know if the observed
differences in the magnitude of the spatial QRS-T
angle are due to movement of the QRS or the T vec-
tor. The consistently high BP values obtained in
some (n ϭ 67) of our patients are less likely to be
due to an alerting response (white coat
hypertension). The latter is usually attributed to an
increased sympathetic activation.27 This was not
probably the case in our study because no sympath-
etically-induced differences in heart rate were
noticed. Therefore, the data of the present study may
suggest that the angle between the directions of ven-
tricular depolarisation and repolarisation is a sensi-
tive and early marker of the repolarisation alter-
ations in systemic hypertension. Moreover, it can be
measured easily, is not affected by observation
biases and is likely to be less susceptible to noise
and problems of definition than conventional ECG
measures of the dispersion of the repolarisation dur-
ation. Future studies should clarify the ability of the
spatial T loop morphology analysis and particularly
of the spatial QRS-T angle to offer a more precise
and reproducible measure of ventricular repolaris-
ation. Furthermore, the ability of those vectorcardio-
graphic descriptors of ventricular repolarisation to
identify high-risk patients for ventricular arrhyth-
mias should be prospectively assessed.
References
1 Clarkson PBM et al. QT dispersion in essential hyper-
tension. Q J Med 1995; 88: 327–332.
2 Perkiomaki JS et al. Dispersion of the QT interval and
¨
¨
autonomic modulation of heart rate in hypertensive
men with and without left ventricular hypertrophy.
Hypertension 1996; 28: 16–21.
3 Statters DJ, Malik M, Ward DE, Camm AJ. QT disper-
sion: problems of methodology and clinical signifi-
cance. J Cardiovasc Electrophysiol 1994; 5: 672–685.
4 Davey PP et al. QT interval dispersion in chronic heart
failure and left ventricular hypertrophy: relation to
autonomic nervous system and Holter tape abnormali-
ties. Br Heart J 1994; 71: 268–273.
5 Mayet J et al. Left ventricular hypertrophy and QT dis-
persion in hypertension. Hypertension 1996; 28:
791–796.
6 Lim PO et al. Irbesartan reduces QT dispersion in
hypertensive individuals. Hypertension 1999; 33:
713–718.
7 Savelieva I et al. Comparative reproducibility of QT,
QT peak, and T peak-T end intervals and dispersion
in normal subjects, patients with myocardial infarc-
tion, and patients with hypertrophic cardiomyopathy.
Pacing Clin Electrophysiol 1998; 21: 2376–2381.
8 Priori SG et al. Evaluation of the spatial aspects of T-
wave complexity in the Long-QT syndrome. Circu-
lation 1997; 96: 3006–3012.
9 Kors JA et al. T axis as an indicator of risk of cardiac
events in elderly people. Lancet 1998; 352: 601–605.
10 Kors JA, Van Herpen G, Van Bemmel JH. QT disper-
sion as an attribute of T-loop morphology. Circulation
1999; 99: 1458–1463.
11 Ishizawa K et al. High reliability rates of spatial pat-
tern analysis by vectorcardiogram in assessing the
severity of eccentric left ventricular hypertrophy. Am
Heart J 1976; 91: 50–57.
12 Van Bemmel JH, Kors JA, Van Herpen G. Methodology
of the modular ECG analysis system MEANS. Methods
Inf Med 1990; 29: 346–353.
13 Fisch C. Electrocardiography. In: Braunwald EG (ed).
Heart Disease: A Textbook of Cardiovascular Medi-
cine. WB Saunders: 1995, pp. 116–117.
14 Edenbrandt L, Pahlm O. Vectorcardiogram synthesized
from a 12-lead ECG: superiority of the inverse Dower
matrix. J Electrocardiol 1988; 21: 361–367.
15 Liebman J, Lee MH, Rao PS, Mackay W. Quantitation
of the normal Frank and McFee-Parungao orthogonal
electrocardiogram in the adolescent. Circulation 1973;
48: 735–752.
Limitations
Only ECG documentation of LVH was obtained in
this study. Additional studies using echocardio-
graphically determined LVH may prove to be more
informative. Neither coronary angiography, nor
other imaging techniques were applied to our study
patients to rule out the presence of significant but
not clinically apparent coronary artery disease.
Finally, our patients were not prospectively evalu-
ated to assess the ability of the vectorcardiographic
markers of ventricular repolarisation to predict the
occurrence of life-threatening ventricular arrhyth-
mias.
16 Dilaveris PE et al. Simple electrocardiographic mark-
ers for the prediction of paroxysmal idiopathic atrial
fibrillation. Am Heart J 1998; 135: 733–738.
17 Levy D et al. Prognostic implications of echocardio-
graphically determined left ventricular mass in the
Framingham Heart Study. N Engl J Med 1990; 322:
1561–1566.
Conclusion
In conclusion, the spatial QRS-T angle was the only
repolarisation marker found to be significantly
18 Kleiman RB, Houser SR. Outward currents in normal
Journal of Human Hypertension

QRS-T angle alterations in hypertension
P Dilaveris et al
70
and hypertrophied feline ventricular myocytes. Am J
Physiol 1989; 256: H1450–H1461.
19 Pye MP, Cobbe SM. Mechanisms of ventricular
arrhythmias in cardiac failure and hypertrophy. Car-
diovasc Res 1992; 26: 740–750.
20 Yan GX, Antzelevitch C. Cellular basis for the normal
T wave and the electrocardiographic manifestations of
the Long – QT syndrome. Circulation 1998; 98:
1928–1936.
23 Dern PL, Pryor R, Walker SH, Searls DT. Serial electro-
cardiographic changes in treated hypertensive patients
with reference to voltage criteria, mean QRS vectors,
and the QRS-T angle. Circulation 1967; 36: 823–829.
24 Talwar KK et al. Spatial quantitative vectorcardiogra-
phy in aortic stenosis: correlation with hemodynamic
findings. Int J Cardiol 1988; 18: 151–161.
25 Hurst JW. Abnormalities of the S-T segment. Part I.
Clin Cardiol. 1997; 20: 511–520.
21 Larsen JA, Kadish AH. Effects of gender on cardiac
arrhythmias. J Cardiovasc Electrophysiol 1998; 9:
655–664.
22 Cooksey JD, Dunn M, Massie E (eds). Clinical Vector-
cardiography and Electrocardiography. Year Book
Medical Publishers: Chicago, 1977.
26 Hansen DE. Mechanoelectrical feedback effects of alt-
ering preload, afterload, and ventricular shortening.
Am J Physiol 1993; 264: H423–H432.
27 Lantelme P et al. White coat effect and reactivity to
stress: cardiovascular and autonomic nervous system
responses. Hypertension 1998; 31: 1021–1029.
Journal of Human Hypertension