VENTRICULA

R MECHANICS

From the Un

Medicine, Kit

University of

and Kawasak

Dr. Langhas re

Reprint reque

cine, Universi

1.1. Iseigaoka

med.uoeh-u.a

0894-7317/$3

Copyright 201

doi:10.1016/j.

Age- and Gender-Dependency of Left VentricularGeometry Assessed with Real-Time

Three-Dimensional Transthoracic Echocardiography

Kyoko Kaku, MD, Masaaki Takeuchi, MD, Kyoko Otani, MD, Lissa Sugeng, MD, Hiromi Nakai, BS,Nobuhiko Haruki, MD, Hidetoshi Yoshitani, MD, Nozomi Watanabe, MD, Kiyoshi Yoshida, MD,

Yutaka Otsuji, MD, Victor Mor-Avi, PhD, and Roberto M. Lang, MD, Kitakyushu and Kurashiki, Japan;Chicago, Illinois

Background: Aging and gender may affect left ventricular (LV) mechanics. The aim of this study was to deter-mine the age and gender dependency of LV mechanical indices obtained from real-time three-dimensionalechocardiography (RT3DE).

Methods: RT3DE was performed in 280 healthy subjects (age range, 1–88 years; 137 men). From full-volumedata sets, LV endocardial and epicardial borders were semiautomatically traced using quantitative software.LV volumes and corresponding long-axis diameter were measured throughout the cardiac cycle. Sphericityindex was defined as the ratio of LV volume and spherical volume, calculated as 4/3 � p � (long-axis diam-eter/2)3. LV mass was calculated as (LV epicardial volume � LV endocardial volume) � 1.05. The ratio of LVmass to LV volume was also calculated.

Results: The mean value of LV ejection fraction did not change with age. However, LV volumes, mass, sphe-ricity index, and LV mass/volume ratio were altered by age: (1) sphericity index was highest in the first decadeof age and then declined until the fifth decade, (2) LV mass/volume ratio significantly increased in older age,and (3) LV mass/volume ratio was significantly higher in aged women compared with age-matched men.

Conclusions: Age has heterogeneous effects on LV shape and LV mass/volume ratio, potentially due to thegrowing process of myocardial fibers and the surrounding architecture in the younger population, as wellas the aging process, with an increase in vascular stiffness and a loss of myocytes in older populations. HigherLVmass/volume ratios in older womenmight be a contributor to the preferential development of diastolic heartfailure in this population. (J Am Soc Echocardiogr 2011;24:541-7.)

Keywords: Aging, Echocardiography, Sex, Ventricles

Because of rapid increase in the aged population, the assessment ofage-related and gender-related changes in left ventricular (LV) geom-etry and function are of paramount importance in the understandingof LV mechanics. Previous studies have demonstrated that heart fail-ure with normal ejection fraction (EF) is frequently more common inaged women compared with aged men.1,2 The age and genderdependency of LV remodeling has been also reported in LVpressure overload hypertrophy.3,4 Specifically, women had greater

iversity of Occupational and Environmental Health, School of

akyushu, Japan (K.K., M.T., K.O., H.N., N.H., H.Y., Y.O.); the

Chicago Medical Center, Chicago, Illinois (L.S., V.M.-A., R.M.L.);

i Medical School, Kurashiki, Japan (N.W., K.Y.).

ceived researchgrants fromPhilipsMedical Systems (Andover,MA).

sts: Masaaki Takeuchi, MD, Second Department of Internal Medi-

ty of Occupational and Environmental Health, School of Medicine,

, Yahatanshi-ku, Kitakyushu 807-8555, Japan (E-mail: takeuchi@

c.jp).

6.00

1 by the American Society of Echocardiography.

echo.2011.01.011

fractional shortening and achieved smaller end-systolic chamber sizesand smaller LV mass than men in severe senile aortic stenosis.Moreover, it is demonstrated that age-related increase in arterial stiff-ening is associated with elevated systolic ventricular stiffening, evenwithout LV hypertrophy.5 Inmicroscopic animal studies, LVmyocytesincreased in length and became irregular in shape with aging.However, the width of myocytes remained constant with advancedaging.6 Age-related decrease in the number of myocytes and increasein collagen content have been associated with elevated LV end-diastolic pressure and decreased dP/dt.7 Therefore, aging and gendermay have different impacts on LV volumes, mass, and shape.However, the effects of age and gender on LV geometry in normalsubjects have not been extensively studied.

Real-time three-dimensional echocardiography (RT3DE) hasproven useful for assessing LVmechanics because of its high feasibilityrate of data acquisition, its relatively low observer and test-retestvariability, and the widely available and convenient semiautomatedoffline analysis.8,9 Previous studies have reported thatdetermination of LV volume and mass on RT3DE is accuratecompared with cardiac magnetic resonance.10-14 RT3DE has beenalso useful for the assessment of LV shape.15

541

Abbreviations

EF = Ejection fraction

LV = Left ventricular

RT3DE = Real-time three-

dimensionalechocardiography

542 Kaku et al Journal of the American Society of EchocardiographyMay 2011

We hypothesized that age andgender may have different ef-fects on LV geometry, whichcan be reliably assessed byRT3DE. Accordingly, we soughtto determine the effects of ageand gender on LV volumes,mass, and shape using transtho-

racic RT3DE in a relatively large number of healthy subjects of bothgenders over a wide range of ages.

METHODS

Study Subjects

A total of 322 healthy subjects over a wide range of ages (1–88 years;150 men) were enrolled. Eligibility criteria included (1) normal bloodpressure without a history of hypertension, (2) absence of diabetesand/or cardiovascular disease, and (3) no cardiac medication use.Subjects were recruited from three university hospitals from theUnited States and Japan and were predominantly hospital employeesor their relatives and/or volunteers recruited through advertisements.All subjects underwent physical examinations and two-dimensionalechocardiography to exclude valvular disease and the presence of re-gional wall motion abnormalities. The study protocol was approvedby the ethics committee of each hospital, and informed consent ob-tained from all subjects.

RT3DE

A Sonos 7500 or iE33 scanner (Philips Medical Systems, Andover,MA) equipped with a fully sampled matrix-array transducer (X4 orX3-1, respectively) was used. Studies were acquired by experiencedcardiac sonographers with subjects in the left lateral decubitus posi-tion. Full-volume data sets were acquired from the apical transducerposition during held end-expiration. To ensure the inclusion of the en-tire LV volume within the pyramidal scan volume, data sets were ac-quired using the wide-angle mode, whereby four wedge-shapedsubvolumes (93� � 21�) were acquired during a single 5-second to7-second breath hold.

Doppler Echocardiography

Pulsed-wave Doppler examination of mitral inflow obtained in theapical four-chamber view was performed in each subject. Doppler tis-sue imaging, used to measure septal mitral annular velocity, was ob-tained by placing the sample volume in the septal corner of themitral annulus in the apical four-chamber view.

Image Analysis

LV Volume, EF, and Shape Indices. Data sets were analyzed off-line using commercially available software (3DQ ADV, QLAB ver-sion 7.0; Phillips Medical Systems), as described previously.11-13

Briefly, five anatomic landmarks (four points for the mitral annulusand one for the LV apex) were manually initialized on the end-diastolic frame using nonforeshortened apical four-chamber andtwo-chamber views selected from the pyramidal data set. Thethree-dimensional endocardial surface was automatically recon-structed using a deformable shell model. Subsequently, the end-systolic frame was manually selected by identifying the frame withthe smallest LV cavity cross-sectional area in both apical views.After a similar initialization, LV surface detection was repeated on

this frame to calculate end-systolic volumes. Finally, the computer al-gorithm automatically tracked the endocardial border throughout allframes of the cardiac cycle. Dynamic ‘‘casts’’ of the LV endocardiumand LV volume-versus-time curves were displayed, from which LVend-diastolic and end-systolic volumes, stroke volume, and LV EFwere then computed. In addition, the corresponding long-axis diam-eter, measured from the mid mitral annulus to the LV apex, was cal-culated.

For the determination of LV shape, we measured the sphericity in-dex,15 which was defined as the ratio between the measured LV vol-ume divided by the spherical volume of the left ventricle, calculated as4/3 � p � (long-axis diameter/2)3 at either end-diastole or end-systole. For the assessment of arterial stiffness, we calculated total sys-temic arterial compliance, which was defined as stroke volume indexdivided by pulse pressure.

LV Mass and LV Mass/Volume Ratio. End-diastolic epicardialcontours were manually traced to calculate LV epicardial volume.LV mass was calculated as (LV epicardial volume � LV endocardialvolume) multiplied by the specific mass of myocardial tissue (1.05g/mL). LV mass/volume ratio was also calculated as LV mass dividedby LV end-diastolic volume.

Doppler Echocardiographic Indices. From mitral inflow veloc-ities, the E-wave and A-wave velocities, E-wave deceleration time,and E/A velocity ratio were measured. Peak diastolic annular velocityduring early diastolic rapid filling (E0) was also measured to calculatethe E/E0 ratio in all subjects. All Doppler measurements were aver-aged from three consecutive beats.

Intraobserver and Interobserver Variability

Intraobserver variability was determined by having one observer re-peat the three-dimensional measurements of LVend-diastolic volumeand LVmass in 30 randomly selected subjects 1 month after complet-ing the initial measurements. Interobserver variability was determinedby having a second observer repeat these measurements in thesesame subjects. Intraobserver and interobserver variability valueswere calculated as the absolute difference between the correspond-ing two measurements in terms of percentage of their mean.

Statistical Analysis

Data are expressed as mean6 SD or as median (interquartile range).Frequencies are expressed as percentages. All statistical analysis wascarried out using commercially available statistical software (JMP ver-sion 7.0 or StatView version 5.0; SAS Institute Inc., Cary, NC).Differences in continuous variables among groups were calculated us-ing one-way analysis of variance with post hoc Bonferroni correction.Categorical variables were compared using Fisher’s exact tests or c2

tests. Linear or polynomial regression analysis was used to study therelation between two parameters. Multivariate linear regression anal-ysis was used to test for independent associations between LV mass/volume ratio and clinical and known echocardiographic parametersassociated with diastolic function, including age, sex, total systemic ar-terial compliance, E/A velocity ratio, E-wave deceleration time, andE0. P values < .05 were considered significant.

RESULTS

Of the 322 subjects screened, 42 (13%) were excluded from analysisbecause of elevated systolic blood pressure (>140 mmHg; n = 23) atthe time of physical examination or poor image quality (n= 19). Thus,

Table 1 Baseline characteristics and results of quantitative analysis on age dependency

Variable

Age group (y)

P

1–9

(n = 53)

10–19

(n = 34)

20–29

(n = 29)

30–39

(n = 32)

40–49

(n = 25)

50–59

(n = 32)

60–69

(n = 51)

>70

(n = 24)

Men 27 17 13 17 12 20 18 13

BSA (m2) 0.79 6 0.22 1.42 6 0.23 1.70 6 0.18 1.86 6 0.22 1.74 6 0.20 1.70 6 0.24 1.57 6 0.16 1.56 6 0.20 <.0001

HR (beats/min) 91 6 18 71 6 15 66 6 11 68 6 9 66 6 12 63 6 9 64 6 11 65 6 11 <.0001

SBP (mm Hg) 103 6 10 116 6 10 120 6 11 120 6 13 121 6 12 127 6 11 126 6 10 130 6 8 <.0001

LVEDV (mL) 40.8 6 14.0 80.3 6 21.7 103.3 6 22.0 106.9 6 18.7 91.3 6 15.8 89.8 6 20.1 73.6 6 15.0 73.3 6 15.9 <.0001

LVEDVI (mL/m2) 51.3 6 7.6 56.8 6 12.3 60.5 6 9.8 57.4 6 8.0 52.5 6 6.4 52.2 6 7.4 46.9 6 8.1 46.7 6 7.7 <.0001

LVESV (mL) 13.8 6 5.0 28.3 6 10.1 36.0 6 10.8 37.8 6 10.6 29.8 6 9.2 30.4 6 11.1 24.1 6 7.2 26.1 6 7.7 <.0001

LVESVI (mL/m2) 17.4 6 3.7 19.8 6 5.6 20.9 6 5.1 20.1 6 4.9 17.0 6 4.4 17.3 6 5.3 15.4 6 4.3 16.5 6 4.0 <.0001

SV (mL) 27.0 6 9.7 52.1 6 13.1 67.3 6 13.0 69.1 6 9.9 61.5 6 9.2 59.4 6 10.7 49.5 6 8.8 47.2 6 9.1 <.0001

SVI (mL/m2) 33.8 6 5.6 36.9 6 8.1 39.6 6 6.2 37.3 6 4.6 35.5 6 3.8 34.9 6 3.3 31.6 6 4.7 30.2 6 4.8 <.0001

LV EF (%) 66.0 6 5.2 65.2 6 5.2 65.6 6 5.0 65.2 6 4.9 68.2 6 5.7 67.1 6 5.9 67.7 6 4.7 64.9 6 4.9 n.s.

LVLAD (cm) 6.14 6 0.85 8.09 6 0.85 8.93 6 0.65 9.10 6 0.57 8.67 6 0.60 8.49 6 0.85 7.81 6 0.59 7.81 6 0.69 <.0001

LVSAD (cm) 3.27 6 0.44 4.05 6 0.45 4.34 6 0.37 4.41 6 0.51 4.15 6 0.31 4.25 6 0.52 3.86 6 0.41 3.82 6 0.49 <.0001

SI (ED) 0.33 6 0.06 0.29 6 0.06 0.28 6 0.04 0.27 6 0.03 0.27 6 0.03 0.28 6 0.05 0.29 6 0.04 0.29 6 0.04 <.0001

SI (ES) 0.22 6 0.05 0.19 6 0.05 0.17 6 0.04 0.17 6 0.03 0.16 6 0.04 0.17 6 0.04 0.17 6 0.05 0.17 6 0.04 <.0001

LVM (g) 43.5 6 16.5 88.9 6 20.3 110.5 6 21.1 118.1 6 22.9 101.2 6 19.1 106.4 6 23.3 96.7 6 16.2 103.8 6 19.3 <.0001

LVMI (g/m2) 54.1 6 8.5 63.1 6 11.6 64.7 6 9.0 63.5 6 8.6 58.1 6 7.4 62.0 6 9.5 61.6 6 7.4 65.7 6 10.2 <.0001

LV mass/volume

ratio (g/mL)

1.06 6 0.14 1.12 6 0.14 1.08 6 0.08 1.10 6 0.09 1.11 6 0.07 1.19 6 0.11 1.34 6 0.19 1.45 6 0.29 <.0001

Total systemic arterial

compliance (mL/mm Hg/m2)

0.82 6 0.22 0.72 6 0.19 0.81 6 0.22 0.86 6 0.24 0.79 6 0.10 0.72 6 0.15 0.63 6 0.15 0.58 6 0.15 <.0001

E/A ratio 2.0 6 0.6 2.3 6 0.6 2.3 6 0.8 1.9 6 0.5 1.5 6 0.4 1.3 6 0.4 1.0 6 0.3 0.8 6 0.3 <.0001

DT (msec) 147 6 31 167 6 27 175 6 30 187 6 25 195 6 37 192 6 33 218 6 49 227 6 49 <.0001

E0 12.1 6 1.6 12.4 6 2.1 12.7 6 1.6 11.6 6 1.6 10.0 6 2.1 8.2 6 1.7 7.0 6 1.5 6.0 6 1.4 <.0001

E/E0 ratio 9.3 6 1.5 8.6 6 1.8 7.4 6 1.8 7.7 6 1.4 8.3 6 1.6 10.2 6 2.6 10.3 6 2.2 12.4 6 5.0 <.0001

Data are expressed as mean 6 SD.

BSA, Body surface area; DT, deceleration time; ED, end-diastole; ES, end-systole; HR, heart rate; LVEDV, LV end-diastolic volume; LVEDVI, LVend-diastolic volume index; LVESV, LV end-systolic volume; LVESVI, LV end-systolic volume index; LVLAD, LV long-axis diameter; LVSAD, LV

short-axis diameter; LVM, LVmass; LVMI, LVmass index;SBP, systolic blood pressure;SI, sphericity index;SV, stroke volume;SVI, stroke volume

index.

Journal of the American Society of EchocardiographyVolume 24 Number 5

Kaku et al 543

the final study group consisted of 280 subjects (mean age, 38 6 24years; age range, 1–88 years; 137 men).

LV Volume, EF, and Shape Indices

LV end-diastolic volumes, end-systolic volumes, and stroke volumeswere age dependent. LV volumes and stroke volumes reached theirpeak value during the third or fourth decade, followed by a mildbut significant reduction during the remaining decades (Table 1).An identical tendency was noted when these values were correctedfor body surface area. In contrast, the mean value of LV EF remainedunchanged throughout all decades. As expected, gender differenceswere noted in LV volumes and stroke volumes, depicting larger valuesin men (Table 2). However, LV EF was slightly but significantly higherin women compared with men.

In keeping with LV volumes, the LV sphericity index was signifi-cantly larger at end-diastole compared with end-systole. The LV sphe-ricity index at both end-diastole and end-systole varied with age(Figure 1). The sphericity index was largest in the earlier decades, in-dicating a more spherical LV shape, which decreased subsequently,reaching its minimal value during the fourth and fifth decades oflife, in contrast to LV volume, reaching its maximal value during thethird and fourth decades of life. Figure 1 also depicts the long-axisand short-axis diameter changes according to age. In the earlierdecades, the absolute increase was larger in the long-axis comparedwith the short-axis diameter.

LV Mass and LV Mass/Volume Ratio

Aging significantly affected LV mass (Figure 2). LV mass increasedfrom the earlier decades, reaching its peak value during the fourth de-cade of life, and then slightly but significantly decreased until the old-est decade. When indexed to body surface area, the age dependencyof LVmass was still observed between the second and eighth decadesof life. Significant gender differences were noted in LVmass (Table 2).LVmass/volume ratio remained constant from the first to the sixth de-cade of life. Subsequently, this value increased throughout the re-maining decades. A negative correlation was noted between LVmass/volume ratio and stroke volume index beyond 30 years ofage (r =�0.61, P < .001). Figure 3 depicts age and gender differencesin LV mass index, LV end-diastolic volume index, and LV mass/vol-ume ratio. LV mass/volume ratio was significantly higher in womencompared with men aged >60 years, when grouped by gender.Total systemic arterial compliance also had a tendency to be lowerin women than men aged >60 years. A weak but significant negativecorrelation was noted between LV mass/volume ratio and total sys-temic arterial compliance (r=�0.41, P < .001). LVmass/volume ratiowas also significantly correlated with E/A velocity ratio (r = �0.46,P < .001), E-wave deceleration time (r = 0.41, P < .001), E0

(r = �0.59, P < .001), and E/E0 (r = 0.33, P < .01), respectively.Multivariate regression analysis revealed that total systemic arterialcompliance (t ratio = �3.21, P = .0016) and E0 (t ratio = �2.29,P = .0236) were independent predictors of LV mass/volume ratio.

Table 2 Gender difference of LV mechanical index

Variable

1–19 y 20–39 y 40–59 y >60 y

Male

(n = 44)

Female

(n = 43) P

Male

(n = 30)

Female

(n = 31) P

Male

(n = 32)

Female

(n = 25) P

Male

(n = 31)

Female

(n = 44) P

BSA (m2) 1.02 6 0.39 1.05 6 0.38 .66 1.86 6 0.18 1.70 6 0.22 <.01 1.83 6 0.21 1.57 6 0.13 <.0001 1.70 6 0.16 1.47 6 0.12 <.0001LVEDV (mL) 58.6 6 28.6 53.9 6 23.1 .40 113.1 6 19.9 97.5 6 17.7 <.01 100.7 6 15.2 77.4 6 12.5 <.0001 84.2 6 12.4 66.0 6 12.2 <.0001

LVEDVI (mL/m2) 56.2 6 11.5 50.5 6 7.3 <.01 60.9 6 10.1 56.8 6 7.4 .07 55.0 6 6.5 49.1 6 6.0 <.01 49.7 6 7.3 44.9 6 7.9 <.05LVESV (mL) 20.7 6 11.6 18.2 6 8.5 .24 42.4 6 10.3 31.6 6 8.1 <.0001 35.3 6 9.4 23.5 6 7.0 <.0001 29.7 6 6.2 21.3 6 6.1 <.0001

LVESVI (mL/m2) 19.6 6 5.1 17.1 6 3.8 <.05 22.8 6 5.2 18.2 6 3.5 <.001 19.1 6 4.8 14.9 6 4.0 .001 17.5 6 3.7 14.5 6 4.1 <.01SV (mL) 37.9 6 17.7 35.7 6 15.4 .55 70.8 6 11.1 65.8 6 11.4 .09 65.4 6 8.8 53.9 6 7.5 <.0001 54.5 6 7.8 44.7 6 7.4 <.0001

SVI (mL/m2) 36.6 6 7.8 33.4 6 5.2 <.05 38.2 6 5.9 38.6 6 5.3 .77 36.0 6 3.3 34.2 6 3.5 .07 32.1 6 4.4 30.5 6 4.8 .14

LV EF (%) 65.3 6 5.3 66.2 6 5.1 .43 62.9 6 4.1 67.8 6 4.4 <.0001 65.6 6 5.6 70.0 6 5.1 <.01 64.8 6 3.8 68.2 6 5.2 <.01

SI (ED) 0.32 6 0.06 0.31 6 0.06 .31 0.26 6 0.03 0.28 6 0.04 <.05 0.27 6 0.04 0.28 6 0.04 .28 0.29 6 0.04 0.30 6 0.04 .11

SI (ES) 0.21 6 0.05 0.20 6 0.06 .28 0.17 6 0.03 0.17 6 0.04 .75 0.16 6 0.04 0.17 6 0.04 .92 0.17 6 0.03 0.17 6 0.05 .91

LVM (g) 62.8 6 30.2 59.7 6 27.2 .62 124.2 6 21.2 105.1 6 19.1 <.001 116.5 6 18.5 88.3 6 13.1 <.0001 108.1 6 17.1 92.6 6 14.7 <.0001

LVMI (g/m2) 59.9 6 11.6 55.2 6 9.2 <.05 66.8 6 9.5 61.4 6 7.2 <.05 63.8 6 8.9 56.1 6 6.6 <.001 62.8 6 7.3 63.0 6 9.3 .91

LV mass/volume ratio (g/mL) 1.07 6 0.13 1.10 6 0.15 .41 1.10 6 0.08 1.08 6 0.09 .32 1.16 6 0.10 1.15 6 0.11 .63 1.30 6 0.21 1.43 6 0.23 <.05

Total systemic arterial

compliance (mL/mm Hg/m2)

0.78 6 0.19 0.78 6 0.23 .94 0.87 6 0.27 0.78 6 0.16 .34 0.76 6 0.15 0.73 6 0.13 .55 0.65 6 0.18 0.59 6 0.12 .08

E/A ratio 2.1 6 0.6 2.1 6 0.6 .90 2.0 6 0.7 2.2 6 0.7 .60 1.3 6 0.4 1.4 6 0.4 .42 1.0 6 0.3 1.0 6 0.3 .84

DT (msec) 157 6 32 155 6 30 .87 182 6 28 178 6 29 .75 202 6 38 186 6 29 .15 218 6 50 222 6 49 .75

E0 12.4 6 2.1 12.0 6 1.4 .34 12.4 6 1.7 12.1 6 1.7 .56 8.2 6 1.5 9.4 6 2.2 .09 7.1 6 1.6 6.6 6 1.5 .23

E/E0 9.1 6 1.7 8.9 6 1.6 .58 7.0 6 1.4 8.0 6 1.7 .05 9.6 6 2.3 9.5 6 2.6 .93 9.9 6 2.2 11.3 6 3.6 .11

Data are expressed as mean 6 SD.

Abbreviations as in Table 1.

544

Kaku

etal

Journalo

ftheAmerican

Society

ofEchocard

iograp

hy

May

2011

Figure 2 Box plots depicting LV mass, LV end-diastolic volume (LVEDV), and LV mass/volume ratio according to age. (Bottom left)Polynomial correlation between LV mass and LVEDV.

Figure 1 Box plots showing sphericity indices at end-diastole (ED) and end-systole (ES) and LV long-axis and short-axis diametersaccording to age. In each box plot, upper and lower bars represent the 90th and 10th percentiles, respectively. The top of the boxrepresents the 75th percentile, the line represents the median value, and bottom of the box represents the 25th percentile. Horizontalbars represent P < .05 between two groups.

Journal of the American Society of EchocardiographyVolume 24 Number 5

Kaku et al 545

Intraobserver and Interobserver Variability

Intraobserver variability for LVend-diastolic volume and LV mass was3.2% and 6.6%, respectively, and interobserver variability was 5.7%and 7.8%, respectively.

DISCUSSION

The main findings of this study were as follows. First, LV volumes,stroke volume, and LV mass were age dependent, in agreement

Figure 3 Box plots showing LV mass index, LV end-diastolic volume index (LVEDVI), and LV mass/volume ratio among different ageand gender groups.

546 Kaku et al Journal of the American Society of EchocardiographyMay 2011

with previous studies using cardiac magnetic resonance imaging.16-19

Specifically, LV volumes and mass increased until the fourth decadeof life and decreased thereafter, and the LV mass/volume ratio wasage dependent, suggesting that changes in LV volume and masswere not uniform and varied with age, particularly in the olderdecades of life. Second, LV EF remained constant between the firstand eighth decades. Third, LV shape was affected by age: the hearthad a more spherical shape in the earlier decades of life and thengradually changed to an elliptical shape in the middle and olderdecades. Fourth, LV end-diastolic volume index and LV mass indexwere significantly lower in women compared with men from the firstto the eighth decade of life. However, the LV mass/volume ratio inolder subjects was significantly higher in women compared withmen. Fifth, multivariate regression analysis revealed that total systemicarterial compliance and E0 were independent predictors of LV mass/volume ratio.

LV Volume and Mass

We used RT3DE to evaluate several LV indices because this technol-ogy has high feasibility and low interobserver and intraobserver vari-ability.11-14 As expected, LV volume and mass progressively increasedfrom the earliest to the fourth decade. After reaching their peakvalues, these values slightly decreased until the seventh decade oflife, reflecting the aging process of the myocardium. On the otherhand, LV EF remained unchanged with age. The majority ofpreviously published studies have also reported decreases in LVvolumes with age using echocardiography or cardiac magneticresonance imaging.16,17,19

LV mass/volume ratio using RT3DE is a robust index to evaluateLV hypertrophy.20 However, whether aging affects the LV mass/vol-ume ratio is unknown. Interestingly, we noted that the LV mass/vol-ume ratio remained constant until the sixth decade of life, andthereafter it increased in later decades. Our finding agrees with those

of previous studies reporting that in children, cardiac growth is char-acterized by parallel increases in cardiac mass and volume, with con-stant LV mass/volume ratios until adolescence.17 Increases in LVmass/volume ratios in the older subjects were also reported in a pre-vious magnetic resonance study.16 The increase in LV mass/volumeratio in an older population probably reflects a finding noted on a cel-lular level wherein the number of cardiomyocytes decreases becauseof necrotic or apoptotic death, and the remaining cardiomyocytes be-come more elongated and hypertrophied,21 with increased fibrousstructure. This might also be a physiologic cardiac response to in-creased vascular afterload caused by arterial stiffening,22 which wasobserved in our elderly subjects. Palmieri et al.23 reported that esti-mated arterial stiffness using systemic arterial compliance was associ-ated with concentric LV geometry and diastolic LV dysfunction inolder and more severely hypertensive patients. Thus, the increase inLV mass/volume ratio subsequently results in decreased LV compli-ance, leading to age-related change in diastolic function observedby Doppler echocardiography, which were also observed in thisstudy. We also observed gender differences in the LVmass/volume ra-tio in the older volunteers. Higher LV mass/volume ratios in agedwomen reflects a more advanced stage of LV concentric remodeling,which might contribute to the preferential development of diastolicheart failure, if they have comorbidities predisposing to LV hypertro-phy.24,25

LV Shape

The assessment of LV shape provides important functional and prog-nostic information in various cardiovascular diseases. RT3DE is a use-ful methodology for the analysis of LV shape, because full-volumedata sets derived from RT3DE incorporate the entire left ventricle,and the analysis does not depend on geometric assumptions, whichis a major limitation of previous two-dimensional echocardiographicstudies focused on LV shape. The three-dimensional sphericity index

Journal of the American Society of EchocardiographyVolume 24 Number 5

Kaku et al 547

has been shown to be predictive in LVremodeling after acute myocar-dial infarction.15 In this study, we found that the sphericity index waslargest in the youngest subjects and subsequently decreased until thefifth decade of life, indicating that the left ventricle is more spherical inthe earlier decades. The age-related decrease in sphericity index is re-lated to the fact that the increase in long-axis diameter was larger com-pared with the short-axis diameter during the early decades. Whetherthis preferential increase in the longitudinal direction might be relatedto myocardial fiber reorientation (from more horizontal to more lon-gitudinal) and physiologic alternation of cellular morphology needs tobe clarified in future studies.

Study Limitations

The numbers of male and female subjects were not equal, particularlyin the middle decades of life, raising the possibility that some genderdifferences might not be accurately detected. Because the number ofsubjects aged >70 years was relatively small, we could not divide pa-tients into the eighth and ninth decades of life to examine age andgender differences in these decades. However, the enrollment of trulyhealthy subjects at this age is difficult. Although the number of youn-ger subjects was relatively large, it was not still enough to divide morefine groups to determine the effect of growth on LV mechanics.Finally, all subjects had no chest pain and dyspnea on exertion, andwe could not exclude the possibility of coronary artery disease, espe-cially in older subjects, which might affect LV mechanics.

CONCLUSIONS

Aging and gender significantly affect LV morphologic changes, evenin normal subjects. Thus, the classification of age-specific andgender-specific normal values should be required when assessingLV geometry, especially in patients with pressure-overload hypertro-phy, such as those with hypertension and aortic stenosis. RT3DE isa noninvasive and feasible method to evaluate age-related andgender-related changes in LV geometry.

REFERENCES

1. Regitz-Zagrosek V, Brokat S, Tschope C. Role of gender in heart failurewith normal left ventricular ejection fraction. Prog Cardiovasc Dis 2007;49:241-51.

2. ZabalgoitiaM, Rahman SN, HaleyWE,Mercado R, Yunis C, Lucas C, et al.Comparison in systemic hypertension of left ventricular mass and geom-etry with systolic and diastolic function in patients. Am J Cardiol 1998;82:604-8.

3. Villari B, Vassalli G, Schneider J, Chiariello M, Hess OM. Age dependencyof left ventricular diastolic function in pressure overload hypertrophy. JAm Coll Cardiol 1997;29:181-6.

4. Weinberg EO, Thienelt CD, Katz SE, Bartunek J, Tajima M, Rohrbach S,et al. Gender differences in molecular remodeling in pressure overload hy-pertrophy. J Am Coll Cardiol 1999;34:264-73.

5. Chen CH, Nakayama M, Nevo E, Fetics BJ, Maughan WL, Kass DA. Cou-pled systolic-ventricular and vascular stiffening with age: implications forpressure regulation and cardiac reserve in the elderly. J Am Coll Cardiol1998;32:1221-7.

6. Capasso JM, Fitzpatrick D, Anversa P. Cellular mechanisms of ventricularfailure: myocyte kinetics and geometry with age. Am J Physiol 1992;262:H1770-81.

7. Anversa P, Palackal T, Sonnenblick EH, Olivetti G, Meggs LG, Capasso JM.Myocyte cell loss and myocyte cellular hyperplasia in the hypertrophiedaging rat heart. Circ Res 1990;67:871-85.

8. Jenkins C, Bricknell K, Hanekom L, Marwick TH. Reproducibility and ac-curacy of echocardiographic measurements of left ventricular parametersusing real-time three-dimensional echocardiography. J Am Coll Cardiol2004;44:878-86.

9. Lang RM, Mor-Avi V, Sugeng L, Nieman PS, Sahn DJ. Three-dimensionalechocardiography: the benefits of the additional dimension. J Am CollCardiol 2006;48:2053-69.

10. Corsi C, Lang RM, Veronesi F, Weinert L, Caiani EG, MacEneaney P, et al.Volumetric quantification of global and regional left ventricular functionfrom real-time three-dimensional echocardiographic images. Circulation2005;112:1161-70.

11. Jacobs LD, Salgo IS, Goonewardena S, Weinert L, Coon P, Bardo D, et al.Rapid online quantification of left ventricular volume from real-time three-dimensional echocardiographic data. Eur Heart J 2006;27:460-8.

12. Mor-Avi V, Sugeng L, Weinert L, MacEneaney P, Caiani EG, Koch R, et al.Fast measurement of left ventricular mass with real-time three-dimen-sional echocardiography: comparison with magnetic resonance imaging.Circulation 2004;110:1814-8.

13. Sugeng L, Mor-Avi V, Weinert L, Niel J, Ebner C, Steringer-Mascherbauer R, et al. Quantitative assessment of left ventricular sizeand function: side-by-side comparison of real-time three-dimensionalechocardiography and computed tomography with magnetic resonancereference. Circulation 2006;114:654-61.

14. Takeuchi M, Nishikage T, Mor-Avi V, Sugeng L, Weinert L, Nakai H, et al.Measurement of left ventricular mass by real-time three-dimensionalechocardiography: validation against magnetic resonance and comparisonwith two-dimensional and m-mode measurements. J Am Soc Echocar-diogr 2008;21:1001-5.

15. Mannaerts HF, van der Heide JA, Kamp O, Stoel MG, Twisk J, Visser CA.Early identification of left ventricular remodelling after myocardial infarc-tion, assessed by transthoracic 3D echocardiography. Eur Heart J 2004;25:680-7.

16. Germans T, G€otte MJ, Nijveldt R, Spreeuwenberg MD, Beek AM,Bronzwaer JG, et al. Effects of aging on left atrioventricular coupling andleft ventricular filling assessed using cardiac magnetic resonance imagingin healthy subjects. Am J Cardiol 2007;100:122-7.

17. Poutanen T, Jokinen E. Left ventricular mass in 169 healthy children andyoung adults assessed by three-dimensional echocardiography. PediatrCardiol 2007;28:201-7.

18. Poutanen T, Jokinen E, Sairanen H, Tikanoja T. Left atrial and left ventric-ular function in healthy children and young adults assessed by three di-mensional echocardiography. Heart 2003;89:544-9.

19. Sandstede J, LipkeC, BeerM, Hofmann S, Pabst T, KennW, et al. Age- andgender-specific differences in left and right ventricular cardiac function andmass determined by cine magnetic resonance imaging. Eur Radiol 2000;10:438-42.

20. De Castro S, Caselli S, Maron M, Pelliccia A, Cavarretta E, Maddukuri P,et al. Left ventricular remodelling index (LVRI) in various pathophysiolog-ical conditions: a real-time three-dimensional echocardiographic study.Heart 2007;93:205-9.

21. Lakatta EG, Levy D. Arterial and cardiac aging: major shareholders in car-diovascular disease enterprises: part II: the aging heart in health: links toheart disease. Circulation 2003;107:346-54.

22. Avolio AP, Deng FQ, Li WQ, Luo YF, Huang ZD, Xing LF, et al. Effects ofaging on arterial distensibility in populations with high and low prevalenceof hypertension: comparison between urban and rural communities inChina. Circulation 1985;71(2):202-10.

23. Palmieri V, Bella JN, Roman MJ, Gerdts E, Papademetriou V, Wachtell K,et al. Pulse pressure/stroke index and left ventricular geometry and func-tion: the LIFE study. J Hypertens 2003;21:781-7.

24. Kitzman DW, Little WC, Brubaker PH, Anderson RT, Hundley WG,Marburger CT, et al. Pathophysiological characterization of isolated dia-stolic heart failure in comparison to systolic heart failure. JAMA 2002;288:2144-50.

25. Sanderson JE, FraserAG. Systolic dysfunction in heart failurewith a normalejection fraction: echo-Doppler measurements. Prog Cardiovasc Dis2006;49:196-206.