Content uploaded by Francesca Mantovani
Author content
All content in this area was uploaded by Francesca Mantovani
Content may be subject to copyright.
.....................................................................................................................................................................................
.....................................................................................................................................................................................
Left ve ntricular hypertrophy r eclassification
and death: applica tion of the Recommenda tion of
the American Society of Echocardiogr aphy/
Eur opean Association of Echocardiogr aphy
†
Andrea Barbieri
1
, Francesca Bursi
1
*
, Francesca Mantovani
1
, Chiara Valenti
1
,
Michele Quaglia
1
, Elena Berti
2
, Massimiliano Marino
2
, and Maria Grazia Modena
1
1
Department of Cardiology, Policlinico Hospital, Modena and Reggio Emilia University, Via del Pozzo 71, 41100, Modena, Italy; and
2
Healthcare and Social Agency,
Regione Emilia Romagna, Bologna, Italy
Received 2 June 2011; accepted after revision 29 August 2011; online publish-ahead-of-print 5 October 2011
Aims Despite the American Society of Echocardiography (ASE)/European Association of Echocardiography (EAE) rec-
ommended the use of left ventricular (LV) mass to diagnose left ventricular hypertrophy (LVH), several laboratories
continue to use only the septal thickness by M-mode because it appears easier to measure. Aim of the study was to
investigate the discrepancy between the categorization of LVH severity based on measurement of septal thickness
and indexed LV mass and the relative prognostic utility of these two methods.
Methods
and results
Observational cohort study. Unselected adults (.18 years) referred to the echocardiography laboratory for any indi-
cation had septal thickness and LV mass measured by the ASE/EAE formula using LV linear dimensions indexed to
body surface area. LVH was categorized as absent, mild, moderate, and severe according to the ASE/EAE guideline
sex-specific categorization cut-offs for septal thickness and LV mass. Follow-up for death was obtained from the
national death index. A total of 2545 subjects (mean age 61.9 + 15.8, 53% women, mean diastolic septal thickness
10.3 + 2.2 mm, and mean indexed LV mass 107.5 + 37.3 g/m
2
) were enrolled. Agreement between the two
methods in classifying LVH degree across the four categories was 52.6% (Kappa ¼ 0.29, 95% confidence interval
(CI): 0.26–0.32, P , 0.001). Of the 2513 subjects without severely thickened septum, 472 (18.9%) had severely
abnormal indexed LV mass. Vice versa, of the 2045 individuals without severely abnormal indexed LV mass, only
4 (0.1%) were classified as severe LVH by septal thickness. After a mean follow-up of 2.5 + 1.2 years 121 (4.7%)
deaths occurred. Using indexed LV mass partition values there was a graded association between LVH degree and
survival. Compared with patients with normal indexed LV mass, the adjusted hazard ratio (HR) for death from all
causes was 2.17 for mild (95% CI: 1.23–3.81, P ¼ 0.007), 3.04 for moderate (95% CI: 1.76– 5.24, P , 0.001), and
3.81 for severe (95% CI: 2.43–5.97, P , 0.001) LVH by indexed LV mass. The area under the receiver-operator
characteristic (ROC) curve for the four degrees of LVH by indexed LV mass was superior [area under the curve
(AUC) ¼ 0.66] to that of the septal thickness partition values (AUC ¼ 0.58, P ¼ 0.0004).
Conclusion In a large cohort study of unselected adult outpatients referred to the echocardiography laboratory, the measure-
ments of indexed LV mass applying the ASE/EAE recommended cut-offs yielded remarkable discrepancy in the diag-
nosis of LVH severity and offered prognostic information beyond that provided by septal thickness only criteria.
-----------------------------------------------------------------------------------------------------------------------------------------------------------
Keywords Left ventricular hypertrophy † Septal wall thickness † Indexed left ventricular mass † Prognosis
†
Application of the Recommendation of the American Society of Echocardiography/European Association of Echocardiography.
* Corresponding author. Tel: +39 059 422 3145; fax: +39 059 422 2843, Email: bursi@libero.it
Published on behalf of the European Society of Cardiology. All rights reserved.
& The Author 2011. For permissions please email: journals.permissions@oup.com
European Heart Journal – Cardiovascular Imaging (2012) 13, 109–117
doi:10.1093/ehjci/jer176
at Mayo Clinic, Division of Biostatistics on May 9, 2013http://ehjcimaging.oxfordjournals.org/Downloaded from
Introduction
Previous studies reported that marked discrepancies between the
echocardiographic measurement of left ventricular (LV) wall thick-
ness and LV mass leading to uncertainty regarding the actual echo-
cardiographic diagnosis of left ventricular hypertrophy (LVH).
1
The
American Society of Echocardiography (ASE)/
European Association of Echocardiography (EAE) guidelines for
chamber quantification
2
proposed sex-specific cut-offs to categor-
ize the degree of LVH based on the distribution of LV mass and
wall thickness in relation to reference limits in a ethnically hetero-
geneous North-American population.
3
These ASE/EAE guidelines
concluded that LV mass from the ASE-recommended formula
using LV linear dimensions indexed to body surface area (BSA) is
preferred in the diagnosis of LVH over linear measurements such
as septal or posterior wall thickness since it provides a more accu-
rate and reproducible estimation of LV mass, when compared with
reference standards such as autoptic studies.
4,5
Nonetheless,
several laboratories continue to use only the septal thickness by
M-mode to evaluate the LVH because it appears easier to measure.
To the best of our knowledge, there are no studies concurrently
assessing the traditionally measured LV septal thickness by
M-mode and the LV mass as an estimate of LVH applying the
recently published cut-offs. For these reasons, we sought to deter-
mine, in a large group of unselected outpatients referred to a ter-
tiary care echocardiography laboratory, whether sex-specific
categorization of LV mass based on the ASE-recommended
cut-offs resulted in diagnostic and prognostic reclassification then
if only the septal wall thickness by M-mode were used.
Methods
The study population comprised unselected elective adult outpatients
who underwent standard Doppler echocardiography for any indication
in the period from January 2005 to March 2009 at the echocardiogra-
phy laboratory of Modena University Hospital. Criteria for enrolment
were: (i) age ≥18 years; (ii) complete resting two-dimensional echo-
cardiogram including prospective real-time measurement of LV mass;
(iii) residency in Modena province, Italy. For patients undergoing
more than one echocardiographic exam during the aforementioned
time frame, we considered only the first access to the echocardiogra-
phy laboratory.
Echocardiographic data
All exams were performed using Acuson Sequoia ultrasound system,
Siemens Medical Solutions, Mountain View, California and were per-
formed and/or supervised by cardiologists fully trained in echocardio-
graphy with long-standing experience with the technique and intense
hands-on training period with interpretation of .750 studies.
6
The LV diameters were measured from 2D-guided M-mode method
in the parasternal short-axis view. LV end-diastolic dimensions were
measured at the onset of the QRS complex. The LV volumes were
derived according to the modified biplane Simpson’s method in the
apical four- and two-chamber views. The LV ejection fraction was cal-
culated in the standard fashion from LV end-diastolic and end-systolic
volume.
7
Regional LV function was assessed with a standard 16-segment
model.
2
Segmental scores were assigned as follows: normal or
hyperkinesis ¼ 1; hypokinesis ¼ 2; akinesis ¼ 3; dyskinesis ¼ 4; and
aneurismal ¼ 5. The LV wall motion score index was derived as the
sum of all scores divided by the number of segments visualized.
Left atrial volume was assessed by the modified biplane Simpson’s
method from apical four- and two-chamber views an indexed to
BSA. Measurements were obtained in end-systole from the frame pre-
ceding mitral valve opening.
8
Individual echocardiographic Doppler
parameters (mitral inflow pattern, tissue Doppler, and Valsalva man-
oeuver where necessary) were integrated to grade diastolic function
in four stages: normal diastolic function; impaired relaxation with
normal or near-normal filling pressures (grade I/IV); impaired relax-
ation with moderate elevation of filling pressures, pseudonormal
filling (grade II/IV); and impaired relaxation with marked elevation of
filling pressures, restrictive filling (grades III – IV/IV) as previously
described.
9
Each value represented the average of three consecutive beats. Sig-
nificant left side valve disease severity was defined as the presence of
aortic or mitral prosthesis or the presence of greater than moderate
native mitral or aortic valve stenosis or insufficiency, similarly to pre-
vious reports.
10
Valve disease severity was defined according to the
American Heart Association/American College of Cardiology guide-
lines for the management of valvular heart disease.
11
By applying
these guidelines, the cardiologist performing the exam graded valve
disease as absent, mild, moderate, and severe and this information
was embedded in the echo report. The methods used included
pulsed wave and continuous wave Doppler velocities and gradients,
direct measurement of valve area planimetry, continuity equation,
colour Doppler to assess the jet width, or proximal isovelocity
surface area for quantitative evaluation. The method/s used to classify
valve disease severity was at the discretion of the physician performing
the exam and often the final judgment was based on the combination
of more than one method. Valve disease severity was defined accord-
ing to the American Heart Association/American College of Cardiol-
ogy guidelines for the management of valvular heart disease. Patients
with greater than moderate valvular heart disease were considered
having significant valvular disease.
All measurements were performed online and immediately entered
in an electronic database at the time of the echocardiogram; no modi-
fication from the original database was applied and no measurement
was made off line. Hence, the study consisted in a retrospective analy-
sis of data prospectively entered in the electronic echocardiographic
database.
Echocardiographically determined left
ventricular mass
The ASE-recommended formula for estimation of LV mass from LV
linear dimensions based on modelling the LV as a prolate ellipse of
revolution. LV mass was calculated thus: LV mass ¼ 0.8[1.04
(LVIDD + IVST + PWT)
3
2 (LVIDD)
3
] + 0.6, where LVIDD rep-
resents LV end-diastolic internal dimension and IVST and PWT indicate
the end-diastolic thicknesses of the interventricular septum and LV
posterior wall, respectively.
5
LV mass were indexed to BSA.
Cut-off limits for left ventricular hypertrophy
The ASE/EAE guidelines suggest the following cut-offs for LVH: LV
septal wall thickness . 0.9 cm for women and .1.0 cm for men, LV
mass/BSA .95 g/m
2
for women and LV mass/BSA .115 g/m
2
for
men. Furthermore, the ASE/EAE guidelines subdivided on an ordinal
scale the values exceeding the reference limits to identify patients
with mild LVH (LV septal thickness 1.0–1.2 cm, LV mass/BSA
96–108 g/m
2
for women and LV septal thickness 1.1–1.3 cm, LV
mass/BSA 116–131 g/m
2
for men), moderate LVH (LV septal thickness
A. Barbieri et al.110
at Mayo Clinic, Division of Biostatistics on May 9, 2013http://ehjcimaging.oxfordjournals.org/Downloaded from
1.3–1.5 cm, LV mass/BSA 109 – 121 g/m
2
for women and LV septal
thickness 1.4–1.5 cm, LV mass/BSA 132 – 148 g/m
2
for men) and
severe LVH (LV septal thickness 1.6 cm, LV mass/BSA ≥122 g/m
2
for
women and LV septal thickness ≥1.7 cm, LV mass/BSA ≥149 g/m
2
for men).
2
Left ventricular geometry
The relative wall thickness (RWT) were measured by the formula
(2×PWTd)/LVIDd which permitted the categorization of an increase
in LV mass as either concentric (RWT ≥0.42) or eccentric (RWT
,0.42) hypertrophy
1
and allowed the identification of concentric
remodelling (normal LV mass with increased RWT).
Clinical data
Age, sex, height, weight, BSA, body mass index, and cardiac rhythm
were recorded at the time of the echocardiogram and entered in
the electronic echocardiography report. Prior history of cardiovascular
diseases (prior acute coronary syndromes—including ST elevation
myocardial infarction and unstable angina/non-ST elevation myocardial
infarction; history of chronic coronary artery disease; prior acute or
chronic heart failure; and prior stroke) was obtained using the hospital
discharge codes of public hospitals of Modena province. Risk factors
were obtained retrospectively by manual review of electronic clinical
notes of the public hospitals of the province. These electronic
records allow assessment of outpatient visits as well as hospital dis-
charge notes.
To collect information on medications, we combined two methods:
(i) we retrospectively reviewed through the electronic databases of
Modena Province public hospital discharge letters and ambulatory car-
diology visits; (ii) we used the Emilia Romagna Region pharmacy cen-
tralized electronic database. This electronic database contains all
prescriptions that are filled in all pharmacies of the region by all resi-
dents regardless of who is the physician prescribing the medication.
We examined all prescriptions filled in 60 days before and 30 days
after the date of the exam.
Follow-up
All cause death was the endpoint of the study. Follow-up information
for death was obtained from the national death index, where the status
of all citizens is steadily constantly updated and is 100% complete.
Indeed in Italy it is mandatory by law that all deceased patients are
immediately registered in this national data bank.
Data are presented as frequency or mean + standard deviation. To
measure the strength of the relation between the septal thickness and
indexed LV mass Pearson’s correlation coefficients (r) was calculated.
The Kappa and Weighted Kappa statistics were used to calculate the
strength of the accord in categorizing the presence and the degree
of LVH of the two methods.
12
The per cent of agreement was calcu-
lated as the ratio between agreed-on measures and the total.
Statistical analysis
Kaplan–Meier curves were constructed to show survival according to
LVH degrees and groups were compared with the log rank test con-
sidering a linear trend across the levels. Univariate and multivariable
Cox proportional hazards models were used to assess the association
between LVH degrees and the risk of death, the risk was presented as
hazard ratio (HR) and 95% confidence interval (CI).
Receiver-operator characteristic (ROC) curves were generated to
assess the overall performance for death prediction of septal thickness
and indexed LV mass both as continuous variables as well as using
guideline proposed partition values. The area under the curve
(AUC) of the ROC curves were compared using the DeLong method.
All tests were two-tailed. A P , 0.05 indicated statistical significance.
All analyses were performed using SPSS (Chicago, IL, USA) version
15.0 for Windows and MedCalc 11.5.0.
Results
During the study period, 2545 subjects (mean age 61.9 + 15.8,
53% women) met the inclusion criteria, and were considered for
the analysis. Characteristics of the patients included in the study
are summarized in Table 1.
................................................................................
Table 1 Baseline characteristics of the patients
included in the study
Total,
n 5 2545
Baseline clinical characteristics (%)
Age, years 61.9 + 15.8
Women 1434 (53)
BSA, m
2
1.82 + 0.22
BMI, m/kg
2
26.2 + 4.6
Hypertension, n
a
1325 (71.6)
Diabetes mellitus, n
a
228 (12.3)
Hyperlipidaemia, n
a
444 (24.0)
Smoking status, n
a
187 (10.1)
Prior acute coronary syndrome 230 (9.0)
History of chronic/stable coronary artery disease 309 (12.1)
History of heart failure 208 (8.2)
Prior stroke 42 (1.7)
Significant valvular heart disease 274 (10.8)
Atrial fibrillation 196 (7.7)
Echocardiographic characteristics
LVEF, % 64.6 + 12.2
WMSI 1.07 + 0.24
LAVi, mL/m
2
, n
b
34.7 + 17.0
Mass indexed, g/m
2
107.5 + 37.3
Relative wall thickness 0.40 + 0.09
Relative wall thickness .0.42 (%) 1007 (39.6)
Medications (%) 440 (17.3)
Antiplatelets 469 (18.4)
Beta-blockers 333 (13.1)
Diuretics 262 (10.3)
Calcium channel blockers 741 (29.1)
ACE inhibitors/angiotensin receptor blockers 327 (12.8)
Statins 2 (0.1)
Anti-arrhythmics (class IA, IB) 26 (1.0)
Anti-arrhythmics (class IC) 18 (0.7)
Amiodarone 117 (4.6)
Insulin/oral hypoglycaemic drugs 174 (6.8)
Anticoagulants 440 (17.3)
BSA, body surface area; BMI, body mass index; LVEF, left ventricular ejectio n
fraction; WMSI, wall motion score index; LAVi, left atrium volume indexed.
a
n ¼ 1851.
b
n ¼ 1624.
Left ventricular hypertrophy reclassification and death 111
at Mayo Clinic, Division of Biostatistics on May 9, 2013http://ehjcimaging.oxfordjournals.org/Downloaded from
Among all subjects enrolled mean LV mass was 195.6 + 72.2 g
and mean indexed LV massed to BSA was 107.5 + 37.3 g/m
2
.
Mean septal diastolic thickness was 10.3 + 2.2 mm. There was a
significant positive correlation between septal thickness and LV
indexed mass (r ¼ 0.75, r
2
¼ 0.57, P , 0.001) but a large scatter
was observed (Figure 1).
Reclassification of left ventricular
hypertrophy
According to the ASE/EAE guideline cut-offs, 1335 (52.5%) sub-
jects were classified as having LVH when the septal thickness
was used and 1198 (47.1%) by indexed LV mass. The observed
proportion of overall agreement between the two methods in clas-
sifying LVH was 79.5%, with a similar proportion of positive and
negative agreement (Kappa ¼ 0.592, 95%CI: 0.560–0.623, P ,
0.001), indicating moderate agreement. Accordingly, in about one
of five subjects there was a discordance between the two defi-
nitions in classifying LVH. Particularly, of 1210 subjects with
normal septal thickness, 192 (15.9%, 7.5% of total) were reclassi-
fied as LVH by indexed LV mass. Vice versa, of the 1347 with
normal indexed LV mass, 329/1347 (24.4%, 12.9% of total) had
LVH by the septal thickness only criterion (Table 2). Patients in
whom LVH was identified by only one criterion were not signifi-
cantly different from those who classified having LVH by both cri-
teria, except for a higher prevalence of diabetes mellitus (Table 3).
Reclassification of left ventricular
hypertrophy severity
There was a different distribution of LVH severity by using septal
thickness and indexed LV mass partition values (Figure 2). Using
septal thickness partition values, LVH was classified as mild in
1069 (42%), moderate in 234 (9.2%), and severe in 32 (1.2%) sub-
jects while using indexed LV mass it was categorized as mild in 391
(15.4%), moderate in 307 (12.1%), and severe in 500 (19.6%)
patients (P , 0.001).
Agreement in classifying LVH degree using the two methods was
met in 1340 subjects and the proportion of overall agreement
across the four categories was 52.6%. Kappa was 0.29 (95% CI:
0.265–0.323, P , 0.001) indicating fair agreement (Table 4). Con-
sidering the close matches, the weighted Kappa was 0.41, which
indicates a moderate agreement. The interclass correlation was
0.52 (95% CI: 0.42– 0.60, P , 0.001).
Of the 2513 subjects without severely thickened septum, 472
(18.9%) had severely abnormal indexed LV mass. Conversely, of
the 2045 individuals without severely abnormal indexed LV mass
only 4 (0.1%) were classified as severely hypertrophic by septal
thickness.
Mortality and left ventricular
hypertrophy reclassification
After a mean follow-up of 2.5 + 1.2 years 121 (4.7%) deaths
occurred. Survival was worse with greater LVH. Using septal thick-
ness partition values 3-year survival was 96.6 + 0.6% among
patients without LVH, 93.4 + 1.0% among those with mild,
92.0 + 2.4% among those with moderate, and 93.8 + 4.3 among
those with severe LVH (P for trend , 0.001, x
2
of the model
13.2, Figure 3). Compared with patients with normal septal thick-
ness those with mildly abnormal septal thickness had a 1.7 (95%
CI: 1.14–2.54, P ¼ 0.009) increased risk of death, those with mod-
erately abnormal septal thickness had a 2.33 (95% CI: 1.34 – 4.06,
P ¼ 0.003) increased risk and those with severely abnormal
septal thickness had a 3.00 (95% CI: 0.92–9.71, P ¼ 0.06)
increased risk.
Using indexed LV mass partition values, there was a graded
association between LVH degree and survival, at 3 years survival
was 97.4 + 0.5% among subjects with normal indexed LV mass,
94.3% + 1.5 among those with mild LVH, 91.5 + 2.0% among
those with moderate, 90.9 + 1.5% among those with severe LVH
(P for trend ,0.001, x
2
40.0 of the model, Figure 4).
Compared with patients with normal indexed LV mass those
with mildly abnormal LV mass had a 2.17 (95% CI: 1.23–3.81,
P ¼ 0.007) increased risk of death, those with moderately abnor-
mal indexed LV mass had a 3.04 (95% CI: 1.76–5.24, P , 0.001)
increased risk, and those with severely abnormal indexed LV
................................................................................
Table 2 Agreement between septal thickness and
indexed left ventricular mass in classifying left
ventricular hypertrophy
Normal LV by
indexed LV mass,
n 5 1347 (52.9%)
Hypertrophy by
indexed LV mass,
n 5 1198 (47.1%)
Normal by septal
thickness, n ¼ 1210
(47.5%)
1018 (40.0%) 192 (7.5%)
Hypertrophy by septal
thickness, n ¼ 1335
(52.5%)
329 (12.9%) 1006 (39.5%)
Figure 1 Correlation between left ventricular septal diastolic
thickness and indexed left ventricular mass. Regression line is
shown as solid line.
A. Barbieri et al.112
at Mayo Clinic, Division of Biostatistics on May 9, 2013http://ehjcimaging.oxfordjournals.org/Downloaded from
mass had a 3.81 (95% CI: 2.43–5.97, P , 0.001) increased risk.
Considering death as dependent variable, the area under the
ROC curve for the four degrees of LVH by indexed LV mass
was superior (AUC ¼ 0.66, 95% CI: 0.64–0.68) to that based on
the septal thickness partition values (AUC ¼ 0.58, 95% CI: 0.56–
0.60), P ¼ 0.0004. Similarly, the AUC was 0.71 (95% CI: 0.69–
0.73) for indexed LV mass and 0.61 (95% CI: 0.60–0.63)
for septal thickness both considered as continuous variables,
P , 0.0001.
At multivariable analyses, after adjusting for age, gender, LV ejec-
tion fraction, atrial fibrillation, presence of significant valvular heart
disease, and LV wall motion score index, indexed LV mass partition
values retained a significant association with death, whereas LV
septal thickness cut-offs were no longer significant (Table 5).
When both LVH definitions were entered in the multivariable
model only indexed LV mass-based partition values retained stat-
istically significant association with death during the follow-up
(compared with normals HR 1.80, 95% CI: 0.97–3.31, P ¼ 0.060
for mildly abnormal indexed LV mass, HR: 2.3, 95% CI: 1.25 –
4.22, P ¼ 0.007 for moderately abnormal indexed LV mass and
HR: 2.38, 95% CI: 1.27–4.44, P ¼ 0.007 for severely abnormal
indexed LV mass).
...............................................................................................................................................................................
Table 3 Baseline characteristics according to the discrepancy in left ventricular hypertrophy categorization.
Abbreviations as in Table 1
LVH by LV mass and septal
thickness, n 5 1006
LVH by septal thickness
only, n 5 329
LVH by LV mass only,
n 5 192
P-value
Baseline characteristics
Age 62.2 + 5.6 62.3 + 15.9 62.3 + 14.5 0.989
Women (%) 572 (56.9) 183 (55.6) 97 (50.5) 0.268
BSA, m
2
1.82 + 0.22 1.82 + 0.21 1.84 + 0.23 0.756
BMI, kg/m
2
26.3 + 4.7 26.2 + 4.7 26.1 + 4.5 0.840
Hypertension
a
(%) 538 (72.9) 167 (68.4) 104 (75.9) 0.242
Diabetes mellitus
a
(%) 82 (11.1) 44 (18.0) 26 (19.0) 0.003
Hyperlipidaemia
a
(%) 183 (24.8) 60 (24.6) 38 (27.7) 0.750
Smoking status
a
(%) 72 (9.8) 27 (11.0) 19 (13.9) 0.521
Prior acute coronary syndrome (%) 99 (9.8) 32 (9.7) 32 (12.0) 0.646
History of chronic/stable coronary
artery disease (%)
131 (13.0) 38 (11.6) 25 (13.0) 0.777
History of heart failure (%) 85 (8.4) 28 (8.5) 16 (8.3) 0.998
Prior stroke (%) 16 (1.6) 10 (3.0) 1 (0.5) 0.084
Significant valvular heart disease (%) 108 (10.7) 32 (9.7) 21 (10.9) 0.859
Atrial fibrillation (%) 81 (8.1) 20 (6.1) 18 (9.4) 0.349
Echocardiographic characteristics
LVEF, % 64.9 + 12.1 62.2 + 12.8 64.8 + 13.3 0.706
WMSI 1.07 + 0.23 1.07 + 0.24 1.09 + 0.29 0.486
LAVi, mL/m
2b
34.2 + 17.1 34.7 + 16.1 35.4 + 16.4 0.772
LV mass, g/m
2
107.5 + 36.6 109.7 + 41.7 111.2 + 39.7 0.735
Relative wall thickness 0.40 + 0.09 0.41 + 0.09 0.40 + 0.09 0.127
Relative wall thickness .0.42 (%) 404 (40.2) 149 (45.3) 72 (37.5) 0.152
BSA, body surface area; BMI, body mass index; LVEF, left ventricular ejection fraction; WMSI, wall motion score index; LAVi, left atrium volume indexed.
a
n ¼ 1851.
b
n ¼ 1624.
Figure 2 Distribution of left ventricular hypertrophy severity
by using septal thickness and indexed mass partition values.
Left ventricular hypertrophy reclassification and death 113
at Mayo Clinic, Division of Biostatistics on May 9, 2013http://ehjcimaging.oxfordjournals.org/Downloaded from
After further adjustment for prior history of acute coronary syn-
drome, history of chronic coronary artery disease, history of heart
failure, prior stroke, and medications use (antiplatelets, anticoagu-
lants, beta-blockers, calcium channel blockers, ACE inhibitors/
angiotensin receptor blockers, diuretics, statins, amiodarone,
other class antiahrrythmics, and insulin/oral hypoglycaemic drugs)
the association between indexed LV mass partition values and
death remained significant (compared with subjects without
LVH, HR: 1.65, 95% CI: 0.93–2.93, P ¼ 0.085 for mildly abnormal
indexed LV mass, HR: 2.12, 95% CI: 1.21 – 3.70, P ¼ 0.008 for mod-
erately abnormal indexed LV mass and 2.17, 95% CI: 1.31 – 3.58,
P ¼ 0.003 for severely abnormal indexed LV mass).
Furthermore, after adjustment for cardiovascular risk factors
(hypertension, diabetes mellitus, hyperlipidaemia, and smoking)
the results were similar. Interaction for gender and cardiac
rhythm were not significant.
Discussion
The main finding of the present study was that estimation of
indexed LV massed by sex and BSA while applying the ASE/EAE
linear recommended formula and cut-offs allowed identification
of a significant number of patients with echocardiographic diagno-
sis of moderate or severe LVH that would have been missed if
...............................................................................................................................................................................
Table 4 Agreement between septal thickness and indexed left ventricular mass in classifying left ventricular
hypertrophy severity
No hypertrophy by
mass, n 5 1347 (52.9%)
Mild hypertrophy by
mass, n 5 391 (15.4%)
Moderate hypertrophy by
mass, n 5 307 (12.1%)
Severe hypertrophy by
mass, n 5 500 (19.6%)
No hypertrophy by septal
thickness, n ¼ 1210
(47.5%)
1018 (40%) 103 (4.0%) 45 (1.7%) 44 (1.7%)
Mild hypertrophy by septal
thickness, n ¼ 1069
(42.0%)
317 (12.4%) 267 (10.5%) 232 (9.1%) 253 (9.9%)
Moderate hypertrophy by
septal thickness, n ¼ 234
(9.2%)
12 (0.4%) 20 (0.7%) 27 (1.0%) 175 (6.8%)
Severe hypertrophy by septal
thickness, n ¼ 32 (1.2%)
0 1 3 (0.1%) 28 (1.1%)
Figure 3 Survival according to left ventricular septal thickness
partition values proposed by the American Society of Echocar-
diography/European Association of Echocardiography classifi-
cation scheme. Solid black line indicates patients without
hypertrophy; dotted line, mild hypertrophy; dashed line, moder-
ate hypertrophy; dash-dotted line, severe hypertrophy.
Figure 4 Survival according to left ventricular mass partition
values proposed by the American Society of Echocardiography/
European Association of Echocardiography classification
scheme. Solid black line indicates patients without LV hypertro-
phy; dotted line, dashed hypertrophy; dashed line, moderate
hypertrophy; dash-dotted line, severe hypertrophy.
A. Barbieri et al.114
at Mayo Clinic, Division of Biostatistics on May 9, 2013http://ehjcimaging.oxfordjournals.org/Downloaded from
classified only by septal thickness cut-offs. The fact that indexed LV
mass partition values represented a more sensitive risk marker of
death than septal thickness in our population reinforces these
findings.
Reclassification of echocardiographic
diagnosis of left ventricular hypertrophy
Prevalence estimates of echocardiographic LVH are closely depen-
dent on the criteria used. In a large study population of patients
with moderate to severe hypertension, the echocardiographic
diagnosis of LVH was present in 42– 77% depending on the
method of indexation and the partition value used.
13
In the present study, we have demonstrated that septal thickness
is often an unreliable approximation of the indexed LV mass,
despite a statistically significant linear correlation between these
two parameters. Of note, the proportion of subjects with LVH
was similar using indexed LV mass and wall thickness criteria, but
the agreement between the two methods was only modest
(kappa ¼ 0.59). Furthermore, indexed LV mass varied greatly
among patients with a similar degree of LV septal thickness.
Indeed, when the septal wall thickness measurements were used,
only about 80% of the cases were classified correctly, whereas
the remainders were misclassified.
Our result confirms and extends previous findings. Devereux
et al.
1
demonstrated in a heterogeneous population with moderate
or severe hypertension, mitral regurgitation, or dilated cardiomyo-
pathy that indexed LV mass was abnormal in 73% while posterior
wall thickness, septal thickness, and relative wall thickness in only
11–32% (P , 0.001 vs. indexed LV mass). The Framingham
Study showed a substantial overlap in septal thickness among sub-
jects with and without echocardiographic LVH defined on the basis
of indexed LV mass.
14
In the present study, the septal thickness criteria tended to clas-
sify LVH severity of lesser degree than the indexed LV mass
method (i.e. 19.6% of subjects had severe LVH by indexed LV
mass compared with only 1.2% by septal thickness). Equally, the
agreement between the two methods was much lower in
classifying the severity of LVH (kappa ¼ 0.29). Therefore, a large
proportion of patients with moderate or severe echocardiographic
LVH by indexed LV mass was missed if classified only by septal
thickness. In contrast, we found that an increased indexed LV
mass was not invariably present in patients with increased wall
thickening, consistent with previous reports demonstrating that
various stressors may have more influence on LV wall thickness
than chamber size.
15,16
The mechanisms responsible for this considerable interindividual
variability in LV mass increase are beyond the scope of this study. It
seems likely that the presence of unmeasured risk factors, including
genetic risk factors and multiple variants, may be involved in mod-
ulating a complex trait such as the development and pattern of
LVH.
17
Accordingly, in a large cohort study, multivariable models
accounting for known risk factors explained only 50% of the varia-
bility of LV mass as assessed by echocardiography.
18
Discordance between left ventricular
wall thickness and mass
We examined the clinical and echocardiographic characteristics
associated to the discrepancy in LVH categorization. One
notable result was that those patients in whom LVH was identified
by only one echocardiographic criterion in general had no signifi-
cant difference in risk factors, medical history, or echocardio-
graphic characteristics when compared with the concordant
groups with LVH present by both criteria, apart from for the
higher prevalence of diabetes. These findings are not surprising
considering that cardiac hypertrophy in vivo is a multifactorial
disease and that the natural history of the challenged heart is a
‘continuum’ in which the overall thickness or mass modifications
of the LV can all take place at various stages of cardiac disease.
19
Notwithstanding, in our population a history of diabetes was
associated to the presence of discordance in the diagnosis of
LVH. This finding parallels the result of studies
16,20
showing that,
in the clinical context, the presence of metabolic syndrome and
diabetes have a distinctive influence on the evolution of the LV
remodelling despite an indistinguishable initial cardiac insult.
...............................................................................................................................................................................
Table 5 Multivariable regression analysis for death using partition values of left ventricular hypertrophy based on septal
thickness and indexed left ventricular mass
LV mass HR (95% CI) P of trend Septal thickness HR (95% CI) P-value
No LVH 1 No LVH 1
Mild LVH 1.82 (1.03 –3.20) 0.040 Mild LVH 1.52 (1.01–2.81) 0.44
Moderate LVH 2.31 (1.33 –4.01) 0.003 Moderate LVH 1.44 (0.81 –2.55) 0.217
Severe LVH 2.30 (1.39 –3.79) ,0.001 Severe LVH 1.91 (0.58– 6.28) 0.287
Age 1.02 (1.00–1.03) 0.007 Age 1.02 (1.01– 1.03) 0.004
Gender 0.63 (0.43 –0.92) 0.016 Gender 0.68 (0.47–0.99) 0.043
Atrial fibrillation 1.06 (0.63 –1.81) 0.821 Atrial fibrillation 1.07 (0.63–1.83) 0.80
EF 0.99 (0.98 –1.01) 0.250 EF 0.99 (0.97–1.00) 0.140
Significant valvular disease 2.78 (1.85–4.19) ,0.001 Significant valvular disease 3.21 (2.13– 4.84) ,0.001
WMSI 1.52 (0.89 –2.58) 0.121 WMSI 1.74 (1.05–2.90) 0.03
BSA, body surface area; BMI, body mass index; LVEF, left ventricular ejection fraction; WMSI, wall motion score index; LAVi, left atrium volume indexed.
Left ventricular hypertrophy reclassification and death 115
at Mayo Clinic, Division of Biostatistics on May 9, 2013http://ehjcimaging.oxfordjournals.org/Downloaded from
Previous data reported that diabetes accelerates the development
of LVH independent of arterial pressure.
21
In a cross-sectional
population-based study, increase in wall thickening without
increase in LV mass characterized the main modality by which
metabolic syndrome abnormalities influenced cardiac structure.
The Authors hypothesized that these findings were consistent
with a possible influence of underlying factors such as insulin resist-
ance and a haemodynamic or vascular process on myocardial
thickening.
16
In clinical practice, it is important to distinguish this unique phe-
notype since the abnormal relative wall thickness/normal mass LV
remodelling pattern represents a stronger cardiovascular risk
factor in diabetic than non-diabetic patients.
20
Reclassification of left ventricular
hypertrophy and prognosis
Ideally, to provide normative data about the degree to which echo-
cardiographic LVH measurements deviate from normal, a practical
approach that predict outcomes or prognosis would be preferred.
2
However, the use of linear dimensions for LVH determination pre-
sents potential limitations. Indeed, M-mode echocardiography does
not allow one to identify the distribution of hypertrophy in the
ventricular septum precisely. This is the case when those portions
of the septum which are thickened are not accessible to the
M-mode ultrasonic beam, typically in hypertrophic cardiomyopathy
patients where the heterogeneous distribution of hypertrophy
results in a distortion of internal LV shape allowing algorithms, gen-
erally used to measure LV mass, not applicable in this disease.
22
Limitation notwithstanding, our data demonstrated that both the
indexed LV mass and LV thickness partition values proposed by
the ASE/EAE were strongly related to death and were able to
entail a risk gradient. Notably, the prognostic accuracy for the
model including the four degrees of indexed LV mass was superior
to the model based on the septal thickness partition values. In
addition, at multivariable analyses, after adjusting for age, gender,
LV ejection fraction, atrial fibrillation, presence of greater than
moderate valvular heart disease, and LV wall motion score index,
indexed LV mass partition values retained a significant association
with death, whereas LV septal thickness values were no longer sig-
nificant. Furthermore, when both definitions were considered con-
temporary, only the indexed LV mass classification of LVH severity
maintained a statistical significant and graded association with
death.
Our data reinforce previous findings demonstrating that indexed
LV mass represented a strong independent marker of cardiovascu-
lar risk both in the general population
23 – 26
and in high-risk
groups
27 – 31
and suggest that, given its prognostic utility, should
be preferred to qualify LVH severity.
Limitations and strengths
The present study population was referral based, and the extent to
which the data can be generalized to other populations remains
unknown. Particularly, because the vast majority of subjects
enrolled was Caucasian results may not be applicable to other
ethnicities.
Information on risk factors was retrievable only for 73% of the
patients. For this reason multivariable analysis including risk factors
as covariates was performed separately as ancillary analysis.
Strengths of the study include the real-time contemporary
measurement of LV mass and septal thickness in a large number
of unselected outpatients and completeness of follow-up.
Conclusions
The application of the ASE/EAE recommended formula and
cut-offs of indexed LV mass and septal wall thickness produced a
significantly different echocardiographic classifications of LVH
degree. Furthermore, the categorization based on measurement
of indexed LV mass provided greater prognostic information
than septal thickness. Therefore, this parameter should be pre-
ferred when evaluating subjects for the presence and severity of
echocardiographic determined LVH.
Conflict of interest: none declared.
References
1. Devereux RB, Casale PN, Kligfield P, Eisenberg RR, Miller D, Campo E et al. Per-
formance of primary and derived M-mode echocardiographic measurements for
detection of left ventricular hypertrophy in necropsied subjects and in patients
with systemic hypertension, mitral regurgitation and dilated cardiomyopathy.
Am J Cardiol 1986;57:1388 –93.
2. Lang RM, Bierig M, Devereux RB, Flachskampf FA, Foster E, Pellikka PA et al. Rec-
ommendations for chamber quantification: a report from the American Society of
Echocardiography’s Guidelines and Standards Committee and the Chamber
Quantification Writing Group, developed in conjunction with the European
Association of Echocardiography, a branch of the European Society of Cardiology.
J Am Soc Echocardiogr 2005;18:1440 –63.
3. Ilercil A, O’Grady MJ, Roman MJ, Paranicas M, Lee ET, Welty TK et al. Reference
values for echocardiographic measurements in urban and rural populations of dif-
fering ethnicity: the Strong Heart Study. J Am Soc Echocardiogr 2001;14:601– 11.
4. Devereux RB, Reichek N. Echocardiographic determination of left ventricular
mass in man. Anatomic validation of the method. Circulation 1977;55:613 –8.
5. Devereux RB, Alonso DR, Lutas EM, Gottlieb GJ, Campo E, Sachs I et al. Echo-
cardiographic assessment of left ventricular hypertrophy: comparison to necropsy
findings. Am J Cardiol 1986;57:450 –8.
6. Beller GA, Bonow RO, Fuster V. ACCF 2008 Recommendations for Training in
Adult Cardiovascular Medicine Core Cardiology Training (COCATS 3) (revision
of the 2002 COCATS Training Statement). J Am Coll Cardiol 2008;51:335– 8.
7. Schiller NB, Shah PM, Crawford M, DeMaria A, Devereux R, Feigenbaum H et al.
Recommendations for quantitation of the left ventricle by two-dimensional echo-
cardiography. American Society of Echocardiography Committee on Standards,
Subcommittee on Quantitation of Two-Dimensional Echocardiograms. J Am Soc
Echocardiogr 1989;2:358– 67.
8. Ujino K, Barnes ME, Cha SS, Langins AP, Bailey KR, Seward JB et al. Two-
dimensional echocardiographic methods for assessment of left atrial volume.
Am J Cardiol 2006;98:1185 –8.
9. Redfield MM, Jacobsen SJ, Burnett JC Jr, Mahoney DW, Bailey KR, Rodeheffer RJ.
Burden of systolic and diastolic ventricular dysfunction in the community: appre-
ciating the scope of the heart failure epidemic. JAMA 2003;289:194– 202.
10. Ommen SR, Nishimura RA, Appleton CP, Miller FA, Oh JK, Redfield MM et al.
Clinical utility of Doppler echocardiography and tissue Doppler imaging in the
estimation of left ventricular filling pressures: a comparative simultaneous
Doppler-catheterization study. Circulation 2000;102:1788 –94.
11. Bonow RO, Carabello BA, Chatterjee K, de Leon AC Jr, Faxon DP, Freed MD
et al. ACC/AHA 2006 guidelines for the management of patients with valvular
heart disease: a report of the American College of Cardiology/American Heart
Association Task Force on Practice Guidelines (writing Committee to Revise
the 1998 guidelines for the manageme nt of patients with valvular heart disease)
developed in collaboration with the Society of Cardiovascular Anesthesiologists
endorsed by the Society for Cardiovascular Angiography and Interventions and
the Society of Thoracic Surgeons. J Am Coll Cardiol 2006;48:e1 –148.
12. Landis JR, Koch GG. The measurement of observer agreement for categorical
data. Biometr ics 1977;33:159 –74.
A. Barbieri et al.116
at Mayo Clinic, Division of Biostatistics on May 9, 2013http://ehjcimaging.oxfordjournals.org/Downloaded from
13. Wachtell K, Bella JN, Liebson PR, Gerdts E, Dahlof B, Aalto T et al. Impact of different
partition values on prevalences of left ventricular hypertrophyand concentric geome-
try in a large hypertensive population: the LIFE study. Hypertension 2000;35:6–12.
14. Savage DD, Garrison RJ, Kannel WB, Levy D, Anderson SJ, Stokes J III et al. The
spectrum of left ventricular hypertrophy in a general population sample: the Fra-
mingham Study. Circulation 1987;75:I26– 33.
15. Lauer MS, Anderson KM, Levy D. Influence of contemporary versus 30-year
blood pressure levels on left ventricular mass and geometry: the Framingham
Heart Study. J Am Coll Cardiol 1991;18:1287 –94.
16. Burchfiel CM, Skelton TN, Andrew ME, Garrison RJ, Arnett DK, Jones DW et al.
Metabolic syndrome and echocardiographic left ventricular mass in blacks: the
Atherosclerosis Risk in Communities (ARIC) Study. Circulation 2005;112:819 –27.
17. Drazner MH. The progression of hypertensive heart disease. Circulation 2011;123:
327– 34.
18. Devereux RB, Roman MJ, Paranicas M, O’Grady MJ, Wood EA, Howard BV et al.
Relations of Doppler stroke volume and its components to left ventricular stroke
volume in normotensive and hypertensive American Indians: the Strong Heart
Study. Am J Hypertens 1997;10:619– 28.
19. Knoll R, Iaccarino G, Tarone G, Hilfiker-Kleiner D, Bauersachs J, Leite-Moreira AF
et al. Towards a re-definition of ‘cardiac hyp ertrophy’ through a rational charac-
terization of left ventricular phenotypes: a position paper of the Working Group
‘Myocardial Function’ of the ESC. Eur J Heart Fail 2011;13:811 –9.
20. Eguchi K, Ishikawa J, Hoshide S, Ishikawa S, Pickering TG, Schwartz JE et al. Differential
impact of left ventricular mass and relative wall thickness on cardiovascular prognosis
in diabetic and nondiabetic hypertensive subjects. Am Heart J 2007;154:79 e9– 15.
21. Grossman E, Shemesh J, Shamiss A, Thaler M, Carroll J, Rosenthal T. Left ventri-
cular mass in diabetes-hypertension. Arch Intern Med 1992;152:1001– 4.
22. Olivotto I, Maron MS, Autore C, Lesser JR, Rega L, Ca solo G et al. Assessment
and significance of left ventricular mass by cardiovascular magnetic resonance
in hypertrophic cardiomyopathy. J Am Coll Cardiol 2008;52:559 – 66.
23. Levy D, Garrison RJ, Savage DD, Kannel WB, Castelli WP. Prognostic implications
of echocardiographically determined left ventricular mass in the Framingham
Heart Study. N Engl J Med 1990;322:1561 –6.
24. Vasan RS, Larson MG, Levy D, Evans JC, Benjamin EJ. Distribution and categoriz-
ation of echocardiographic measurements in relation to reference limits: the
Framingham Heart Study: formulation of a height- and sex-specific classification
and its prospective validation. Circulation 1997;96:1863 –73.
25. Tsang TS, Barnes ME, Gersh BJ, Takemoto Y, Rosales AG, Bailey KR et al. Predic-
tion of risk for first age-related cardiovascular events in an elderly population:
the incremental value of echocardiography. J Am Coll Cardiol 2003;42:
1199– 205.
26. Gar din JM, McClell and R, Kitzman D, Lima JA, Bom mer W, Klopfenstein HS
et
al. M-mode echocardiographic predictor s of six- to seven-ye ar incidence
of coronary heart disease, stroke, con gestive heart failure, and mortality in
an elderly cohort (the Cardiovascular Hea lth Study). Am J Cardiol 2001;87:
1051– 7.
27. Bolognese L, Dellavesa P, Rossi L, Sarasso G, Bongo AS, Scianaro MC. Prognostic
value of left ventricular mass in uncomplicated acute myocardial infarction and
one-vessel coronary artery disease. Am J Cardiol 1994;73:1–5.
28. Cooper RS, Simmons BE, Castaner A, Santhanam V, Ghali J, Mar M. Left ventri-
cular hypertrophy is associated with worse survival independent of ventricular
function and number of coronary arteries severely narrowed. Am J Cardiol
1990;65:441 –5.
29. Stevens SM, Farzaneh-Far R, Na B, Whooley MA, Schiller NB. Development of an
echocardiographic risk-stratification index to predict heart failure in patients with
stable coronary artery disea se: the Heart and Soul study. JACC Cardiovasc Imaging
2009;2:11–20.
30. Verma A, Meris A, Skali H, Ghali JK, Arnold JM, Bourgoun M et al. Prognostic
implications of left ventricular mass and geometry following myocardial infarction:
the VALIANT (VALsartan In Acute myocardial iNfarcTion) Echocardiographic
Study. JACC Cardiovasc Imaging 2008;1:582 –91.
31. Quinones MA, Greenberg BH, Kopelen HA, Koilpillai C, Limacher MC,
Shindler DM et al. Echocardiographic predictors of clinical outcome in patients
with left ventricular dysfunction enrolled in the SOLVD registry and trials: signifi-
cance of left ventricular hypertrophy. Studies of Left Ventricular Dysfunction. JAm
Coll Cardiol 2000;35:1237 –44.
Left ventricular hypertrophy reclassification and death 117
at Mayo Clinic, Division of Biostatistics on May 9, 2013http://ehjcimaging.oxfordjournals.org/Downloaded from