The voltage of R wave in lead aVL improves risk stratification in hypertensive patients without ECG left ventricular hypertrophy

C

Original article 1697

The voltage of R wave in lead
aVL improves risk stratificationin hypertensive patients without ECG left ventricularhypertrophyPaolo Verdecchiaa, Fabio Angelia, Claudio Cavallinia, Giovanni Mazzottaa,Salvatore Repacia, Silvia Pedea, Claudia Borgionia, Giorgio Gentileb andGianpaolo Reboldib
Objectives We tested the hypothesis that the voltages of

QRS on ECG improve risk stratification in hypertensive

patients without left ventricular hypertrophy on ECG.

Methods and results We studied 2042 initially untreated

patients with hypertension (mean age 49 years, 46%

women) without left ventricular hypertrophy on ECG and no

history of cardiovascular disease. At entry, all patients

underwent diagnostic tests, including 24-h ambulatory

blood pressure monitoring and echocardiography. Among

the different ECG voltages, the R wave in lead aVL showed

the closest association with left ventricle (LV) mass

(r U 0.31; P < 0.001), followed by the R wave in D1

(r U 0.25) and the S wave in V3 (r U 0.22). Patients were

followed up for a mean of 7.7 years (range 1–22 years),

and treatment was tailored individually. During follow-up,

there were 188 major cardiovascular events. The

relationship between LV voltage and outcome was

assessed using a Cox model with adjustment for

age, sex, diabetes, smoking, total cholesterol, serum

creatinine, LV mass on echocardiography and average

24-h ambulatory blood pressure. A 0.1 mV higher

R wave voltage in lead aVL was associated with a 9%

higher risk of cardiovascular disease (95% confidence

interval U 0.04–0.15%; P < 0.001). Other ECG voltages and

minor repolarization changes were not related to clinical

outcome.

opyright © Lippincott Williams & Wilkins. Unauth

0263-6352 � 2009 Wolters Kluwer Health | Lippincott Williams & Wilkins

Conclusion Our results show for the first time that the

voltage of the R wave in lead aVL improves cardiovascular

risk stratification in hypertensive patients without left

ventricular hypertrophy on ECG. Its prognostic value is

independent of LV mass on echocardiography and 24-h

ambulatory blood pressure. J Hypertens 27:1697–1704

Q 2009 Wolters Kluwer Health | Lippincott Williams & Wilkins.

Journal of Hypertension 2009, 27:1697–1704

Keywords: echocardiography, electrocardiography, epidemiology,hypertension, left ventricular hypertrophy, left ventricular mass, risk factors

Abbreviations: ABP, ambulatory blood pressure; ACE, angiotensin-converting-enzyme; ARB, angiotensin receptor blocker; BP, bloodpressure; CCBs, calcium channel blockers; ECG, electrocardiogram; LV, leftventricle; LVH, left ventricular hypertrophy; LVM, left ventricular mass;PIUMA, Progetto Ipertensione Umbria Monitoraggio Ambulatoriale; ROC,receiver-operated characteristic

aDepartment of Cardiology, Clinical Research Unit ‘Preventive Cardiology’,Hospital S. Maria della Misericordia, Cardiologia and bDepartment of InternalMedicine, University of Perugia, Perugia, Italy

Correspondence to Dr Paolo Verdecchia, MD, FACC, FAHA, Unita di RicercaClinica ‘Cardiologia Preventiva’, Ospedale ‘S. Maria della Misericordia’, PiazzaleG. Menghini, Perugia 06132, ItalyTel: +39 075 5782213; fax: +39 075 5782214; e-mail: [email protected]

Received 3 January 2009 Revised 18 March 2009Accepted 23 March 2009

See editorial commentary on page 1538

IntroductionThe traditional resting ECG remains a simple, non-

invasive and relatively inexpensive diagnostic tool in

patients with high blood pressure (BP). Hypertension

guidelines recommend ECG as first-line diagnostic test

in patients with a clinical diagnosis of hypertension [1,2].

Indeed, there is evidence from longitudinal studies that

left ventricular hypertrophy (LVH) diagnosed by stan-

dard 12-lead ECG is a powerful predictor of cardiovas-

cular disease not only in the general population [3,4], but

also in specific groups of patients with arterial hyperten-

sion [5–7], coronary artery disease [8] and congestive

heart failure [9].

Although the prognostic value of ECG LVH is well

established, surprisingly few data are available on the

prognostic value of ECG findings in hypertensive

patients without evidence of ECG LVH and no overt

cardiovascular disease. In such patients, echocardiogra-

phy would be the obvious diagnostic choice because

ECG is less sensitive than echocardiography for diagnosis

of LVH [10,11]. However, a further exploitation of the

ECG features in this large proportion of patients with

generally mild or moderate hypertension and without

ECG LVH would be relevant for cost and convenience

reasons, if supported by evidence. For example, it was

suggested that a prolonged ventricular repolarization [12]

and minor repolarization abnormalities [13] may improve

prediction of cardiovascular risk even after adjustment for

the confounding effect of LVH.

The present study was conducted to test a hypothesis

generated by some previous studies. In one study, ECG

voltages showed a direct and consistent association with

orized reproduction of this article is prohibited.

DOI:10.1097/HJH.0b013e32832c0031

mailto:[email protected]

http://dx.doi.org/10.1097/HJH.0b013e32832c0031

C

1698 Journal of Hypertension 2009, Vol 27 No 8

left ventricular mass (LVM) [14]. In other studies, a

continuous relationship was found between LVM and

the risk of cardiovascular disease even in individuals with

normal LVM [15,16]. Therefore, we tested the hypoth-

esis of a direct association between ECG voltages and the

risk of cardiovascular disease in hypertensive patients

without LVH on ECG. We also investigated the possible

impact of LVM on such association.

MethodsWe analyzed data from the Progetto Ipertensione Umbria

Monitoraggio Ambulatoriale (PIUMA) study, a prospec-

tive observational registry of morbidity and mortality in

initially untreated patients with essential hypertension.

Details of the study have been provided elsewhere

[7,17,18]. Entry criteria included an office BP of least

140 mmHg systolic or of least 90 mmHg diastolic on at

least three visits and an absence of secondary causes of

hypertension, previous cardiovascular disease and life-

threatening conditions. The PIUMA study was approved

by the local Ethical Committee, and all patients provided

their informed consent to participate in the registry.

ElectrocardiographyStandard 12-lead ECG was recorded in all patients during

a brief end-expiratory apnea. Patients with conditions

potentially precluding a correct ECG assessment of LVH

(complete right bundle branch block, left bundle branch

block, atrial fibrillation, pathological Q waves because of

prior myocardial infarction and Wolf–Parkinson–White

syndrome) were excluded from the study.

LVH was defined by the presence of a typical strain

pattern or a modified Cornell voltage (sum of the S wave

in V3þ the R wave in aVL> 2.0 mV in women and

>2.4 mV in men) [19]. The strain was defined by an at

least 0.5 mm depression of the J point, T-wave inversion

with asymmetric branches and rapid return to baseline.

Minor repolarization abnormalities were defined by a

Minnesota code 4-3, 4-4, 5-3 and 5-4 [20].

EchocardiographyThe M-mode echocardiographic study of the left ven-

tricle (LV) was performed under two-dimensional guide.

Only frames with optimal visualization of interfaces

and showing simultaneous visualization of septum, LV

internal diameter and posterior wall were used for read-

ings. Details about reading procedures and reproducibil-

ity in our laboratory are provided elsewhere [18]. LV mass

was calculated by using a necropsy validated formula

[21] and corrected by height (m) at the power of 2.7

[22]. LVH was defined by an LV mass of more than

51.0 g/height[m]2.7 [22].

Ambulatory blood pressureAmbulatory blood pressure (ABP) was recorded as rou-

tinely performed in our laboratory [17] using an oscillo-

opyright © Lippincott Williams & Wilkins. Unautho

metric device (SpaceLabs 5200, 90202 and 90207; Space-

Labs, Redmond, Washington, USA). The frequency of

measurements was set to one every 15 min throughout

the 24 h.

Follow-upFamily doctors and our hospital staff were responsible for

the follow-up of patients. Treatment was tailored indi-

vidually and based on lifestyle and pharmacological

measures. As the results of diagnostic tests were made

available to family doctors, it is difficult to ascertain to

what extent some of the diagnostic findings, particularly

those related to 24-h ambulatory BP monitoring and

echocardiography, influenced the management of the

single patients. Thiazide diuretics, b-blockers, angioten-

sin-converting enzyme (ACE) inhibitors, angiotensin II

(AT-II) antagonists, calcium channel blockers and a1-

blockers, alone or combined, were the antihypertensive

drugs most frequently used. Periodical contacts with

family doctors and phone interviews and clinical visits

with patients continued in order to ascertain the

vital status and the occurrence of events. The total

duration of follow-up to an event or censoring was

7.7 years (range 1–22).

Assessment of end pointsHospital record forms and other source documents of

patients who died or suffered a vascular event were

reviewed in conference by the authors in the absence

of source data regarding the echocardiographic study or

other investigations such as 24-h ambulatory BP. Details

about the international standard criteria used to diagnose

outcome events in the PIUMA study are provided else-

where [17,18]. Stroke was defined as a new neurological

deficit lasting at least 24 h, in the absence of underlying

potentially important nonvascular causes. Transient

ischemic attack (TIA) was diagnosed by a neurologist

or internist in the presence of a rapid onset of a focal

neurological deficit lasting between 30 s and 24 h, and

presumably due to ischemia. The PIUMA protocol

required the deficit to have lasted until the time of the

qualifying clinical examination in order to be accepted

and coded as terminating event. Myocardial infarction

was diagnosed on the basis of at least two out of three

standard criteria (typical chest pain, ischemic ECG

changes and biochemical markers of myocardial damage).

Myocardial ischemia was diagnosed in the presence of an

exercise test, thallium imaging, Holter ECG monitoring

or coronary angiography (stenosis> 50% in at least one

major coronary artery) positive for ischemia. Arterial

occlusive disease was diagnosed by claudicatio with

evidence of at least more than 50% echographic or

angiographic stenosis in a peripheral artery. Sudden

death was defined as a witnessed death that occurred

within 1 h after the onset of acute symptoms, with no

evidence that violence or accident played any role in the

fatal outcome.

rized reproduction of this article is prohibited.

C

Prognostic value of normal ECG Verdecchia et al. 1699

Fig. 1

3352

Office BP < 140/90 mmHg (n = 495)

Included in the PIUMA registryfrom June 10. 1986 to June 9. 2007

(n = 3844)

3111

ECG unavailable (n = 177)Complete right bundle block (n = 50)Complete left bundle block (n = 7)Wolf parkinson white syndrome (n = 4)

2790

Follow-up data unavailable (n = 321)

2363

Left ventricular hypertrophy at ECG (LV strain or Cornell voltage > 2.0 mV (women) and > 2.4 mV (men)(n = 427)

Unavailable or sub-optimal echocardiographic tracings(n = 321)

2042

Selection procedures of the study population. LV, left ventricle; PIUMA,Progetto Ipertensione Umbria Monitoraggio Ambulatoriale.

Table 1 Main characteristics of the population

Variable Value

Age (years) 48.8 (11)Weight (kg) 75.2 (14)Height (cm) 167.9 (9)Body mass index (kg/m2) 26.6 (4)Known duration of hypertension (years) 3.70Duration of follow-up (years) 7.73 (4.5)Office systolic BP (mmHg) 154 (17)Office diastolic BP (mmHg) 97 (9)Office heart rate (beats/min) 75 (10)Average 24-h systolic BP (mmHg) 135 (13)Average 24-h diastolic BP (mmHg) 86 (9)Average 24-h heart rate (beats/min) 75 (9)Total cholesterol (mmol/l) 5.56 (1.0)Triglycerides (mmol/l) 1.62 (1.1)Glucose (mmol/l) 5.46 (1.03)Uric acid (mmol/l) 278.8 (83)Creatinine (mmol/l) 85.4 (20)Na (mEq/l) 141 (5.7)K (mEq/l) 4.2 (0.4)Ca (mEq/l) 9.43 (4.0)End-diastolic interventricular septum (cm) 1.09 (0.20)End-diastolic left ventricular internal diameter (cm) 4.91 (0.49)End-diastolic posterior wall thickness (cm) 0.99 (0.17)Left ventricular mass (g/height2.7) 46.7 (11)Height of the R wave in lead aVL (mV) 0.54 (0.31)Depth of the S wave in lead V3 (mV) 0.88 (0.41)Cornell voltage (mV)a 1.42 (0.47)Depth of the S wave in lead V1 (mm) 0.97 (0.37)Height of the tallest R wave in leads V5 or V6 (mm) 1.52 (0.57)

Data reported as mean (standard deviations). a Sum of the height of the R wave inlead aVL plus the height of the R wave in lead aVL (mm). BP, blood pressure.

Data analysisWe used the SPSS (SPSS, Inc., Chicago, Illinois, USA)

and the SAS-Stat (SAS Insitute, Cary, North Carolina,

USA) for analysis. Parametric data are reported as

mean�SD. Standard parametric and nonparametric

analysis were performed. For survival analyses, event-

free curves were estimated using Kaplan–Meier product-

limit method and compared by the Mantel (log-rank) test.

For patients who experienced multiple events, survival

analysis was based on the first-occurring event. The

independent effect of several prognostic factors on sur-

vival was tested by a stepwise Cox model. The two

components of the Cornell voltage (height of the R wave

in lead aVL and depth of the S wave in lead V3) were

tested simultaneously in the model. The other tested

variables were age, sex (men, women), BMI (kg/m2),

clinic systolic and diastolic BP (mmHg), average 24-h

systolic and diastolic BP (mmHg), total cholesterol (mg/

dl), serum triglycerides (mmol/l), serum creatinine (mg/

dl), family history of cardiovascular disease at age younger

than 55 years in the father or younger than 65 years in the

mother (no, yes), smoking habits (nonsmokers, current

smokers), LVM on echocardiography (g/height2.7), minor

repolarization changes (Minnesota code 4-3, 4-4, 5-3 and

5-4) [yes, no] and type of drug treatment at the follow-up

visit [diuretic (yes, no), b-blocker (yes, no), ACE inhibitor

(yes, no), angiotensin receptor blocker (ARB) (yes, no),

calcium channel blocker (CCB) (yes, no), other

(yes, no)].

A receiver-operated characteristic (ROC) curve analysis,

focused on the impact of ECG voltages as continuous

variables on the occurrence of cardiovascular events, was

undertaken according to standard methods.

In two-tailed tests, P< 0.05 was considered statistically

significant.

ResultsOut of 3844 consecutive patients who entered the

PIUMA registry from June 1986 to June 2007, 2042 were

selected for this study according to the prespecified

selection criteria reported in Fig. 1. As shown in

Table 1, the mean age of patients was 48.8 years. As

most of the patients (78.6%) were newly diagnosed, the

known duration of hypertension was 3.7 years as mean

and 1.0 years as median. The percentage of women was

45.7% (n¼ 934), that of current smokers was 23.8%

(n¼ 485) and that of patients with type 2 diabetes was

6.3% (n¼ 129). The prevalence of LVH on echocardio-

graphy was 31.5% (n¼ 644).

At the last follow-up visit before censoring or event,

percentages of regular users of diuretics, b-blockers,

ACE inhibitors, CCBs and ARBs were 35%, 18%, 38%,

22% and 18%, respectively. Average clinic BP at entry

was 154/97 mmHg and 24-h BP was 135/86 mmHg (sys-


tolic/diastolic). The distribution of 24-h ambulatory BP in

our population is reported in Fig. 2.

Correlation between ECG voltages and left ventriclemassThe Cornell voltage (sum of the S wave in V3þ the R

wave in aVL) showed the closest association with LV

mass (r¼ 0.36, P< 0.001). The single ECG voltage most


C


Fig. 2

250

200

150

100

50

080 100 120 140 160 180 200 40 60 80 100 120 140

Average 24-hour ambulatory blood pressure, mmHg

Systolic n = 2042Mean 135,1Median 133,0

Diastolic n = 2042Mean 86,2Median 86,0

Num

ber

of p

atie

nts

Distribution of average 24-h systolic and diastolic blood pressure.

Fig. 4

closely associated with LVM was the height of the R

wave in lead aVL (r¼ 0.31, P< 0.001). Progressively

weaker correlations were shown by the R wave in lead

DI (r¼ 0.25; P< 0.001), the S wave in lead DIII (r¼ 0.22;

P< 0.001) and the S wave in lead V3 (r¼ 0.19; P< 0.001).

These data are reported in Fig. 3. The scatter plot of the

R wave in lead aVL vs. LVM is shown in Fig. 4.

Cardiovascular eventsDuring follow-up, there were 188 new cardiovascular

fatal or nonfatal events (Table 2). The overall event rate

(�100 patients per year) was 1.19 [95% confidence inter-

vals (CIs)¼ 1.03–1.37], reflecting the relatively low risk

in this population without ECG LVH. In the three

tertiles of the distribution of the R wave in aVL

(Fig. 5), the event rate was 0.73 (95% CI¼ 0.54–1.01),

1.14 (95% CI¼ 0.89–1.47) and 1.71 (95% CI¼ 1.39–


Fig. 3

−0.1 −0.05 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35

Correlation coefficients with left ventricular mass

0.40

RaVL

RDI

SDIII

SV3

SV1

RV6

RV5

SDI

RDIII

Cornell *

*

*

*

*

*

*

*

*† n = 2042

* P < 0.01† P = n.s.

Sokolow *

Correlation coefficients between left ventricular mass onechocardiography and the electrocardiographic voltages.

2.10), respectively. The Mantel–Cox test for trend of

rates was highly significant (P< 0.0001). In contrast, there

was not any significant increase in the risk of events

(x2¼ 0.31; P¼ 0.319) across the three tertiles of the

distribution of the S wave in V3 [1.02 (95% CI¼ 0.78–

1.33), 1.31 (95% CI¼ 1.04–1.67) and 1.24 (95%

CI¼ 0.97–1.58).

In the three tertiles of the distribution of the Cornell

voltage, which reflects the sum of the R wave in lead aVL

and the S wave in lead V3, the event rate was 0.62 (95%

CI¼ 0.45–0.89), 1.20 (95% CI¼ 0.94–1.53) and 1.75

(95% CI¼ 1.42–2.15). The Mantel–Cox test for trend

of rates was highly significant (x2¼ 27.05; P< 0.0001).


0

0.5

1.0

1.5

2.0

Hei

ght o

f the

R w

ave

in le

ad a

VL

(m

V)

20 40 60 80 100 120

Left ventricular mass (g/height2.7)

r = 0.31P < 0.001

Correlation between left ventricular mass on echocardiography and theheight of the R wave in lead aVL. The 95% confidence ellipse isincluded. Each small circle (dots) is one single patient. The ellipse is the95% confidence ellipse.

C


Table 2 Major cardiovascular events that occurred during thefollow-up period

Event Number of patients

Cardiovascular death 11Non fatal myocardial infarction 48Nonfatal stroke 39Transient ischemic attack 22New-onset myocardial ischemiaa 36New-onset heart failure requiring hospitalization 12New-onset arterial occlusive disease 14Renal failure requiring chronic dialysis 6

Only the first-occurring event is reported for patients who experienced multipleevents. a 15 of these patients underwent aorto-coronary by-pass surgery. SeeMethods for a definition of events.

Fig. 5

C

rud

e ev

ent r

ate

(X 1

00 p

atie

nts

per

yea

r)

0

0.2

0.4

0.6

0.8

1

1.2

1.4

1.6

1.8

Height of the R wave in lead aVL (mV)

Tertiles

Depth of the S wave in lead V3 (mV)

Tertiles

Cornell voltage (mV)

Tertiles

Tertile 1 Tertile 2 Tertile 3

P < 0.0001 P = 0.3194

P < 0.0001

Event rate in the three tertiles of the distribution of the R wave in leadaVL, the S wave in lead V3 and the Cornell voltage.

Fig. 6

2

3

4

5

6

7

0 1 2 3 4 5 6 7 8 9 10

37 g/m2.7 42 g/m2.7 48 g/m2.7 56 g/m2.7

5-ye

ar r

isk-

fact

or

adju

sted

ris

ko

f fir

st C

V e

ven

t

Height of the R wave in lead aVL (mm)

Left ventricular mass (division lines for quintiles)

Five-year risk-factor adjusted risk of cardiovascular events withprogressively greater values of left ventricular mass for each level of theR wave in lead aVL. CV, cardiovascular.

In a Cox model (Table 3), after simultaneous adjustment

for age (P< 0.001), sex (P¼ 0.001), diabetes (P¼ 0.032),

smoking (P< 0.001), total cholesterol (P¼ 0.001), serum

creatinine (P< 0.001), LVM on echocardiography

(P¼ 0.002) and average 24-h ABP (P¼ 0.028), for every

0.1 mV increase in the R wave voltage in lead aVL, there

was a 9% higher risk of cardiovascular disease (95%

CI¼ 4–15%; P< 0.001). Neither of the other ECG vol-

tages nor the minor repolarization changes (Minnesota


Table 3 Results of multivariate analysis

Covariate Comparison

Age 10 yearsSex Men vs. womenDiabetes Yes vs. noCurrent smoking Yes vs. noTotal cholesterol 1 SD (41 mg/dl)Serum creatinine (mg/dl) 1 SD (0.22 mg/dl)Left ventricular mass 1 SD (11 g/height2.7)Average 24-h systolic BP 1 SD (13 mmHg)Height of the R wave in aVL 0.1 mV

BP, blood pressure.

code 4-3, 4-4, 5-3 and 5-4) achieved statistical signifi-

cance.

When the Cornell voltage was included in the model in

place of R in aVL, after adjustment for the above covari-

ates, for each 1 mV higher Cornell voltage, there was a

5.3% higher risk of cardiovascular disease (95% CI¼ 1–

9%; P¼ 0.018).

Figure 6 shows the 5-year risk factor-adjusted risk of a

first cardiovascular event as a function of the height of the

R wave in lead aVL for increasing values of LVM.

The area under ROC curve relating the height of the R

wave in lead aVL to the first-occurring cardiovascular

event was 0.624 (95% CI¼ 0.603–0.645; Fig. 7, left

panel), and the crossing point between sensitivity and

specificity of RaVL was 5.7 mm (right panel). The other

electrocardiographic voltages showed lower perform-

ances. The areas under the curves were 0.540, 0.495,

0.556, 0.521, 0.534, 0.525, 0.549 and 0.551 for RDI, SDI,


Hazard ratio 95% confidence intervals P-value

1.76 1.51–2.05 <0.0011.78 1.25–2.54 0.0011.59 1.04–2.42 0.0321.85 1.32–2.60 <0.0011.28 1.11–1.48 0.0011.17 1.07–1.28 <0.0011.26 1.09–1.46 0.0021.18 1.02–1.37 0.0321.09 1.04–1.15 <0.001

C


Fig. 7

0 20 40 60 80 100

100

80

60

40

20

0

100-specificity

Sen

siti

vity

Area = 0.624 (95% CI 0.603−0.645)

RaVL (mm)0 5 10 15 20

100

80

60

40

20

0

Sensitivity (%)Specificity (%)

Receiver-operated characteristic curve analysis of the R wave in lead aVL in relation to total cardiovascular events. The crossing point betweensensitivity and specificity (5.7 mm) is reported in the right panel. CI, confidence interval.

SDIII, RDIII, SV3, SV1, RV5, RV6, respectively (all

P< 0.001 vs. RaVL).

DiscussionThe main finding of this study is that the voltage of the R

wave in lead aVL is a predictor of cardiovascular disease

risk in a relatively young population of hypertensive

patients without LVH on standard ECG. The relation-

ship between the height of the R wave and the risk of

cardiovascular disease was statistically significant, and it

persisted after adjustment for traditional risk factors, 24-h

ABP and LVM estimated by echocardiography. The

Cornell voltage (sum the of R wave in lead aVLþS wave

in lead V3) was also a predictor of cardiovascular risk and

its association with LV mass was relatively close, but its

prognostic impact was driven by the R wave in aVL

because the depth of the S wave in lead V3 did not show

any significant association with the risk of events.

To the best of our knowledge, our study is the first to

simultaneously test the prognostic value of the ECG

voltages and echocardiographic LVM in the specific

setting of hypertensive patients without LVH on ECG.

Despite the overwhelming evidence that ECG LVH

portends an increased risk of cardiovascular disease [3–

7], ECG is poorly sensitive for diagnosis of LVH [11].

Echocardiography is thus recommended by guidelines

‘when a more sensitive detection of LVH is considered

useful’ [1], which is the case of patients without LVH on

ECG. In a previous study from our group, 21% of low-risk

patients and 32% of medium-risk patients defined by

guidelines actually showed LVH on echocardiography,

thus implying reclassification of these patients into a

high-risk category with indication to drug treatment

[23]. Surprisingly, however, few data exist with respect

to the prognostic value of echocardiography, and possibly


of additional ECG information, in the specific setting of

hypertensive patients without ECG evidence of LVH.

Left ventricular massThe prognostic value of echocardiographic LVM has

been extensively investigated in mixed populations with

and without LVH on ECG [17,15,24]. Because echocar-

diographic LVM is higher in the presence than in the

absence of LV hypertrophy on ECG [14], the prognostic

impact of LVM could have been driven by the subset

with ECG LVH. The present investigation removes such

potential concern by suggesting that the conclusions

regarding the prognostic value of LVM can be fully

applied to a lower risk population without LVH on

ECG. Furthermore, our findings indicate that the signi-

ficant 26% higher risk of cardiovascular disease for

any 11 g/height2.7 higher LV mass is independent of

traditional cardiovascular risk factors and the average

24-h systolic BP. Adjustment for 24-h ABP is valuable

because it is more potent than clinic BP for risk stratifica-

tion of hypertensive patients [17,25–27].

ElectrocardiographyOur findings are consistent with previous studies that

reported a direct relationship between ECG voltages and

risk of major cardiovascular events in hypertensive

patients [6,19,28]. Similar results have been obtained

in a referred sample of veterans [29]. Again, the present

study has the unique feature of being conducted in a

specific population without ECG LVH.

At variance with the results of this and the above studies

[6,19,28,29], LVH defined solely by voltage ECG criteria

was not a significant predictor of cardiovascular risk in the

general population [30,31]. In the Framingham Heart

Study, the adjustment for BP removed the excess risk


C


associated with voltage-only LVH [31], suggesting that

high BP levels may be more important than high ECG

voltages for risk stratification. However, the general

population might include a consistent proportion of you-

ng and healthy individuals. These factors are likely to

limit the prevalence, and hence the sensitivity, of ECG

for LVH. For example, in the Framingham Study, sen-

sitivity of ECG for LVH was as low as 3.4% in patients

below 50 years of age, compared with 13.9% in those over

70 years [32]. The sensitivity of ECG voltages for a

diagnosis of LVH tends also to decrease for the effect

of obesity [32]. Overall, these data suggest that elevated

ECG voltages might be more specific for true LVH in

populations at higher prevalence of LVH, such as hyper-

tensive patients or those with a high-risk phenotype. This

hypothesis is consistent with the results obtained in the

Heart Outcomes Prevention Evaluation Study, in which

ECG LVH defined by voltage criteria was a significant

independent predictor or cardiovascular death, all-cause

death and heart failure [33].

In the present study, the height of the R wave in lead aVL

was the single lead showing the closest association with

echocardiographic LVM, followed by RDI, SDIII and

SV3. The R wave in lead aVL also provided a better

performance, when compared with other ECG voltage, in

terms of sensitivity and specificity for subsequent cardi-

ovascular disease, as suggested by the ROC curve

analysis. From a clinical perspective, a voltage value of

0.57 mV (5.7 mm), which marked the crossing point

between sensitivity and specificity, might be considered

as a diagnostic threshold. The R wave in lead aVL is

believed to reflect most of vectors, directed towards left

and high, originating from activation of the LV. In a study

by Casale et al. [34], the R wave in lead aVL ranked

second, after the S wave in lead V3, for strength of

correlation with echocardiographic LVM. It is possible

that the independent association between the height of

the R wave in aVL and outcome may reflect the direct

relationship of the above ECG voltage with LVM. Such a

relationship was statistically significant, albeit weak

(r¼ 0.31). Again, the exclusion of patients with ECG

LVH could have weakened the degree of association

between the ECG voltages and echocardiographic

LVM. The lack of any relationship of the other ECG

voltages with outcome could also reflect their lesser

degree of association with LVM.

The risk of cardiovascular disease increased with the

voltage of the R wave in lead aVL, over and beyond

LVM. The explanations for this finding are elusive.

Several features of the myocardial tissue may influence

the surface ECG regardless of the amount of LVM. For

example, the proportion of collagen in relation to myo-

cytes [35] and the distorted orientation and coupling of

myocardial fibers [36] can affect the features of surface

ECG.


The minor, nonspecific, abnormalities of LV repolarization

did not yield statistical significance, at variance with a prior

analysis of the PIUMA database [13]. However, in the

present study, we had the opportunity to adjust for echo-

cardiographic LVM, in addition to the voltage of the R

wave in aVL. These factors could include the prognostic

information provided by nonspecific LV repolarization

changes.

Limitations of the studyBecause our population is entirely composed of white

patients, extrapolation of the results to different ethnic

groups is not supported. Black patients have higher ECG

voltages than whites even when matched for LV mass,

and this may increase sensitivity and decrease specificity

of ECG for a diagnosis of LV hypertrophy in blacks [37].

In addition, because only about one-third of our patients

had regular periodical examinations during follow-up, the

prognostic impact of serial changes in the ECG and

echocardiographic markers of LV anatomy and function

could not be determined.

ConclusionThe Cornell voltage and the voltage of R wave in lead

aVL improved risk stratification in hypertensive patients

without LVH on ECG. The prognostic value of the

Cornell voltage was driven by the height of the R wave

in lead aVL. Each 0.1 mV higher value of the R wave

voltage in lead aVL was associated with a 9% higher risk

of cardiovascular disease. Such a relationship was inde-

pendent of traditional and newer risk markers, including

LVM at echocardiography and 24-h ABP, which also

contributed to risk stratification. R wave voltages in lead

aVL in the upper tertile (>6.50 mm) were associated with

an annual event rate of 1.71%, which is a remarkably high

figure in this low-risk population (overall annual event

rate 1.13%). Taken together, these findings reinforce the

clinical value of traditional ECG in the work-up of

hypertensive patients.

AcknowledgementsThis study was supported in part by the no-profit founda-

tion Fondazione Umbra Cuore e Ipertensione, Perugia,

Italy. We thank Mrs Carla jaspers for secretarial assistance.

References1 Mancia G, De Backer G, Dominiczak A, Cifkova R, Fagard R, Germano‘ G,

et al. Guidelines for the Management of Arterial Hypertension: the TaskForce for the Management of Arterial Hypertension of the European Societyof Hypertension (ESH) and of the European Society of Cardiology (ESC).J Hypertens 2007; 25:1105–1187.

2 Chobanian AV, Bakris GL, Black HR, Cushman WC, Green LA, Izzo JL Jr,et al., Joint National Committee on Prevention, Detection, Evaluation, andTreatment of High Blood Pressure. National Heart, Lung, and BloodInstitute; National High Blood Pressure Education Program CoordinatingCommittee. Seventh report of the Joint National Committee on Prevention,Detection, Evaluation, and Treatment of High Blood Pressure.Hypertension 2003; 42:1206–1252.

3 Kannel WB, Gordon T, Offutt D. Left ventricular hypertrophy byelectrocardiogram: prevalence, incidence and mortality in the Framinghamstudy. Ann Intern Med 1969; 71:89–105.


C


4 De Bacquer D, De Backer G, Kornitzer M, Blackburn H. Prognostic value ofECG findings for total, cardiovascular disease, and coronary heart diseasedeath in men and women. Heart 1998; 80:570–577.

5 Sokolow M, Perloff D. The prognosis of essential hypertension treatedconservatively. Circulation 1961; 23:697–713.

6 Fagard RH, Staessen JA, Thijs L, Celis H, Birkenhager WH, Bulpitt CJ,et al., Systolic Hypertension in Europe (Syst-Eur) Trial Investigators.Prognostic significance of electrocardiographic voltages and their serialchanges in elderly with systolic hypertension. Hypertension 2004;44:459–464.

7 Verdecchia P, Schillaci G, Borgioni C, Ciucci A, Gattobigio R, Zampi I,Porcellati C. Prognostic value of a new electrocardiographic method fordiagnosis of left ventricular hypertrophy in essential hypertension. J Am CollCardiol 1998; 31:383–390.

8 Sullivan JM, van der Zwaag RV, el-Zeky F, Ramanathan KB, Mirvis DM. Leftventricular hypertrophy: effect on survival. J Am Coll Cardiol 1993;22:508–513.

9 Hawkins NM, Wang D, McMurray JJ, Pfeffer MA, Swedberg K, Granger CB,et al., CHARM Investigators and Committees. Prevalence and prognosticimplications of electrocardiographic left ventricular hypertrophy in heartfailure: evidence from the CHARM programme. Heart 2007; 93:59–64.

10 Reichek N, Devereux RB. Left ventricular hypertrophy: relationship ofanatomic, echocardiographic and electrocardiographic findings.Circulation 1981; 63:1391–1398.

11 Woythaler JN, Singer SL, Kwan OL, Meltzer RS, Reubner B, Bommer W,DeMaria A. Accuracy of echocardiography versus electrocardiography indetecting left ventricular hypertrophy: comparison with postmortem massmeasurements. J Am Coll Cardiol 1983; 2:305–311.

12 Schillaci G, Pirro M, Ronti T, Gemelli F, Pucci G, Innocente S, et al.Prognostic impact of prolonged ventricular repolarization in hypertension.Arch Intern Med 2006; 166:909–913.

13 Schillaci G, Pirro M, Pasqualini L, Vaudo G, Ronti T, Gemelli F, et al.Prognostic significance of isolated, nonspecific left ventricularrepolarization abnormalities in hypertension. J Hypertens 2004; 22:407–414.

14 Budhwani N, Patel S, Dwyer EM Jr. Electrocardiographic diagnosis of leftventricular hypertrophy: the effect of left ventricular wall thickness, size, andmass on the specific criteria for left ventricular hypertrophy. Am Heart J2005; 149:709–714.

15 Levy D, Garrison RJ, Savage DD, Kannel WB, Castelli WP. Prognosticimplications of echocardiographically determined left ventricular mass inthe Framingham heart study. N Engl J Med 1990; 322:1561–1566.

16 Schillaci G, Verdecchia P, Porcellati C, Cuccurullo O, Cosco C, PerticoneF. Continuous relation between left ventricular mass and cardiovascular riskin essential hypertension. Hypertension 2000; 35:580–586.

17 Verdecchia P, Porcellati C, Schillaci G, Borgioni C, Ciucci A, Battistelli M,et al. Ambulatory blood pressure. An independent predictor of prognosis inessential hypertension. Hypertension 1994; 24:793–801; Erratum in:Hypertension 1995;25:462.

18 Verdecchia P, Schillaci G, Borgioni C, Ciucci A, Gattobigio R, Zampi I, et al.Prognostic significance of serial changes in left ventricular mass inessential hypertension. Circulation 1998; 97:48–54.

19 Verdecchia P, Angeli F, Reboldi GP, Carluccio E, Benemio G, GattobigioR, et al. Improved cardiovascular risk stratification by a simple ECG Index inhypertension. Am J Hypertens 2003; 16:646–652.

20 Prineas RJ, Crow RS, Blackburn H. The Minnesota code manual ofelectrocardiographic findings: standards and procedures for measurementand classification. Littleton, Massachusetts: John Wright-PSG; 1982.

21 Devereux RB, Alonso DR, Lutas EM, Gottlieb GJ, Campo E, Sachs I,Reichek N. Echocardiographic assessment of left ventricularhypertrophy: comparison to necropsy findings. Am J Cardiol 1986;57:450–458.

22 de Simone G, Daniels SR, Devereux RB, Meyer RA, Roman MJ, de DivitiisO, Alderman MH. Left ventricular mass and body size in normotensivechildren and adults: assessment of allometric relations and impact ofoverweight. J Am Coll Cardiol 1992; 20:1251–1260.

23 Schillaci G, De Simone G, Reboldi G, Porcellati C, Devereux RB,Verdecchia P. Change in cardiovascular risk profile by echocardiography inlow- or medium-risk hypertension. J Hypertens 2002; 20:1519–1525.

24 Verdecchia P, Porcellati C, Reboldi G, Gattobigio R, Borgioni C, PearsonTA, et al. Left ventricular hypertrophy as an independent predictor of acutecerebrovascular events in essential hypertension. Circulation 2001;104:2039–2044.

25 Sega R, Facchetti R, Bombelli M, Cesana G, Corrao G, Grassi G, ManciaG. Prognostic value of ambulatory and home blood pressures comparedwith office blood pressure in the general population: follow-up results fromthe Pressioni Arteriose Monitorate e Loro Associazioni (PAMELA) study.Circulation 2005; 111:1777–1783.


26 Hansen TW, Jeppesen J, Rasmussen S, Ibsen H, Torp-Pedersen C.Ambulatory blood pressure and mortality: a population-based study.Hypertension 2005; 45:499–504.

27 Dolan E, Stanton A, Thijs L, Hinedi K, Atkins N, McClory S, et al. Superiorityof ambulatory over clinic blood pressure measurement in predictingmortality: the Dublin outcome study. Hypertension 2005; 46:156–161.

28 Van Hoof R, Staessen J, Fagard R, Thijs L, Amery A. The effect ofantihypertensive treatment on electrocardiogram voltages in the EWPHEtrial. J Cardiovasc Pharmacol 1991; 17 (Suppl 2):S101–S104.

29 Hsieh BP, Pham MX, Froelicher VF. Prognostic value ofelectrocardiographic criteria for left ventricular hypertrophy. Am Heart J2005; 150:161–167.

30 Larsen CT, Dahlin J, Blackburn H, Scharling H, Appleyard M, Sigurd B,Schnohr P. Prevalence and prognosis of electrocardiographic leftventricular hypertrophy, ST segment depression and negative T-wave; theCopenhagen City Heart Study. Eur Heart J 2002; 23:315–324.

31 Kannel WB. Prevalence and natural history of electrocardiographic leftventricular hypertrophy. Am J Med 1983; 75:4–11.

32 Levy D, Labib SB, Anderson KM, Christiansen JC, Kannel WB, Castelli WP.Determinants of sensitivity and specificity of electrocardiographic criteriafor left ventricular hypertrophy. Circulation 1990; 81:815–820.

33 Lonn E, Mathew J, Pogue J, Johnstone D, Danisa K, Bosch J, et al., HeartOutcomes Prevention Evaluation Study Investigators. Relationship ofelectrocardiographic left ventricular hypertrophy to mortality andcardiovascular morbidity in high-risk patients. Eur J Cardiovasc PrevRehabil 2003; 10:420–428.

34 Casale PN, Devereux RB, Kligfield P, Eisenberg RR, Miller DH, ChaudharyBS, Phillips MC. Electrocardiographic detection of left ventricularhypertrophy: development and prospective validation of improved criteria.J Am Coll Cardiol 1985; 6:572–580.

35 Mehta D, McKenna WJ, Ward DE, Davies MJ, Camm AJ. Significance ofsignal-averaged electrocardiography in relation to endomyocardial biopsyand ventricular stimulation studies in patients with ventricular tachycardiawithout clinically apparent heart disease. J Am Coll Cardiol 1989;14:372–379.

36 Punske BB, Taccardi B, Steadman B, Ershler PR, England A, Valencik ML,et al. Effect of fiber orientation on propagation: electrical mapping ofgenetically altered mouse hearts. J Electrocardiol 2005; 38 (4 Suppl):40–44.

37 Lee DK, Marantz PR, Devereux RB, Kligfield P, Alderman MH. Leftventricular hypertrophy in black and white hypertensives. Standardelectrocardiographic criteria overestimate racial differences in prevalence.JAMA 1992; 267:3294–3299; Erratum in: JAMA. 1992;268:3201.


Documents

The voltage of R wave in lead aVL improves risk stratification in hypertensive patients without ECG left ventricular hypertrophy