Download Effect of Age, Sex, and Body Surface Area

Effect of Age, Sex, and Body Surface Area on
Echocardiographic Left Ventricular Wall Mass in
Normal Subjects
JULIUS M. GARDIN, DANIEL D. SAVAGE, JAMES H. WARE, AND WALTER L. HENRY
Downloaded from http://hyper.ahajournals.org/ by guest on June 15, 2017
SUMMARY M-mode echocardiography was used to estimate left ventricular wall mass in 136 older normal
subjects (Group I: 78 men and 58 women, ages 20 to 97 years) and 105 younger normal subjects (Group II: 52 male
and 53 female subjects, ages 1 day to 23 years). Echocardiographic left ventricular mass (in grams) was estimated
from the following formula: left ventricular mass = 1.05 {[left ventricular internal diastolic dimension + ventricular septal thickness (diastole) + posterior wall thickness (diastole)]3 - [left ventricular internal diastolic dimension]3}. In both groups, female subjects had a slightly smaller left ventricular mass than male subjects (mean
difference 7.2% in Group I, p<0.05, and 3.6% in Group II, p = 0.05) for any given age and body surface area.
Left ventricular mass varied linearly with body surface area and increased as a function of age. In group I
subjects, echocardiographic left ventricular mass (in grams) could be estimated by the general formula: left
ventricular mass = 124 (body surface area) + A ± C, where A is the age-dependent intercept; ± C encompasses a
95% prediction interval for normal values, which is assumed to be nearly constant (± 58 g); and body surface area
is expressed in square meters. In the Group II (younger) subjects, with age not considered, left ventricular mass
(in grams) could be estimated from the following formula: left ventricular mass = 115 (body surface
area) - 11 ± C, where ± C = ± 32% and this 95% prediction interval varies as a percentage of the mean. We
conclude that both age and body surface area must be taken into account when evaluating echocardiographic
measurements of left ventricular mass. (Hypertension 9 [Suppl II]: 11-36—11-39, 1987)
KEY WORDS • left ventricular mass • echocardiography • body surface area
N
• aging • sex
Methods
Subject Population
Our study population consisted of two subject groups. The
first group included 136 subjects, 78 men and 58 women, without evidence of heart disease, who ranged in age from 20 to 97
years. None of the subjects had a history suggesting heart disease or hypertension, an abnormal electrocardiogram, or an
abnormal chest x-ray. In addition, none of the subjects was a
trained athlete, gave a history of excessive alcohol intake, or
had a body weight more than 25% above the upper limit for
desirable weight set forth by the Metropolitan Life Insurance
Company.18 The second group of subjects consisted of 105
younger normal individuals between the ages of 1 day and 23
years. None of these younger individuals had evidence of cardiovascular disease on the basis of a history or physical examination. Twelve-lead electrocardiograms were available in 87 of
the younger normal individuals and were normal in each case.
All subjects gave informed consent and were studied in accordance with institutional guidelines.
ON1NVASIVE estimation of left ventricular (LV) wall
mass by echocardiography has become an important
tool in the assessment of prognosis and response to
therapy in patients with systemic hypertension.1"10 In order to
use echocardiographic LV wall mass measurements as a quantitative tool in patients with potential cardiac involvement, it is
important to have reference data for LV mass obtained in normal subjects. Roberts and Perloff" have noted that hearts of
older individuals have smaller ventricular chambers and thicker
walls than those of younger individuals studied at necropsy.
Various investigators have also noted the influence of body
surface area (BSA) on the dimensions and thickness of cardiac
structures.12"17 In order to investigate the effects of age, body
size, and gender on normal LV wall mass as well as on LV
dimensions and wall thickness, we performed a series of Mmode echocardiographic studies in groups of younger and older
normal subjects.13"15
From the Cardiology Branch, National Heart, Lung and Blood Institute,
National Institutes of Health, Bethesda, Maryland, and the Division of
Cardiology, University of California Irvine Medical Center, Orange, California (J.M. Gardin, W.L. Henry). Dr. Savage's present address is National
Center for Health Statistics, Bethesda, Maryland. Dr. Ware's present address is the School of Public Health, Harvard University, Cambridge, Massachusetts.
Address for reprints: Julius M. Gardin, M. D., Director, Cardiology Noninvasive Laboratory, University of California Irvine Medical Center, 101
City Drive South, Bldg 53, Rt 81, Orange, CA 92668.
Echocardiography
M-mode echocardiograms were performed in the left lateral
decubitus position using either a 2.25-MHz, 3.5-MHz or 5.0MHz transducer. Each study was read independently by at least
two investigators. Three consecutive measurements were made
for each echocardiographic parameter and the average taken as
11-36
EFFECT OF AGE, SEX, AND BODY SIZE ON LEFT VENTRICULAR MASS/Gardin et al.
the value for the parameter. In approximately 5% of studies,
interobserver variation was greater than 10% in dimensional
measurements or 2 mm or more in wall thickness measurements. In these instances, the echocardiograms were reviewed
and data included for analysis as agreed by consensus of the
readers.
The LV internal dimensions at end diastole and end systole
were measured immediately below the tips of the mitral valve
leaflets. These measurements corresponded, respectively, to the
maximum and minimum internal dimensions between the ventricular septum and endocardium of the posterior LV free wall
(Figure 1). Thickness of the ventricular septum and posterobasal LV free wall was measured in late diastole, just prior to
thinning of the wall.13' M Measurements were made with the
ultrasound beam passing through the left ventricle at or slightly
below the tips of the mitral leaflets. A switched-gain circuit was
used to simplify identification of the epicardium of the posterobasal free wall." In no patient was the ratio of septal thickness
to posterior wall thickness greater than 1.2:1.
The LV mass was calculated as previously described20"22
using the following equation:
Downloaded from http://hyper.ahajournals.org/ by guest on June 15, 2017
LV mass =1.05 [(LVIDd + VSTd + PWTd)3 - (LVIDd)3]
where LVID represents the LV dimension, VST = ventricular
septal thickness, PWT = posterior LV wall thickness, and
d = diastole.
The BSA was estimated from height and weight by the
Boothby and Standiford23 modification of the formula of DuBois and DuBois.24
Results
Effect of Body Surface Area
Echocardiographic estimates of LV wall mass were analyzed
for the influence of age, sex, and BSA. In the younger normal
11-37
subjects, we found that the relationship of estimated LV mass to
BSA could be described by the regression model:
LVmass= 115(BSA)- 11
where BSA was expressed in square meters and LV mass was
expressed in grams.13
Although ventricular septal thickness and LV free wall thickness varied linearly with the square root of BSA, and LV internal dimensions (systolic and diastolic) varied in linear relation
to the cube root of BSA, estimated LV mass varied linearly with
BSA. Figure 2 demonstrates a plot of our data for estimated LV
mass (in grams) versus a linear function of BSA (in square
meters).
To develop a predictive equation that would contain a new
normal value with 95% confidence, we expanded the above
equation to include a prediction interval term. The expanded
equation could be expressed as follows:
LV mass = B(BSA) - A ± C
where 2C (or ± C) is the width of the 95% prediction interval
(i.e., the interval into which, with 95% confidence, a new
normal observation would fall13"15). In our younger normal individuals, the 95% prediction interval for LV mass varied proportionally with the mean (i.e., became wider as BSA increased)
and could be expressed as ± 32% of the mean (see Figure 2). 13
In order to simplify the prediction equation in the older adult
population (Group I), we assumed that the 95% prediction interval for LV mass had a constant width. Over the range of BSAs
encountered in this older adult population, the constant assumption introduced less than a 5% difference in the prediction intervals compared with the proportionality assumption.14
Effect of Sex
Table 1 summarizes the sex, heart rate, blood pressure, and
BSA data for our older normal (Group I) population. To determine whether sex (independent of age and BSA) had an effect
on LV mass, we analyzed the data for LV mass separately for
men and women after correcting the data for age and BSA in our
Group I subjects. There was a statistically significant difference
between men and women for estimated LV mass (p<0.05),
with women having an average LV mass 7.2% smaller than that
in men for any given age and BSA.14 In our younger normal
subjects (Group II), the average LV mass in women was 3.6%
smaller than that in men for any given BSA (p = 0.05). 13
Because the sex differences in both the older and younger
groups were relatively small, we combined the data for men and
women in order to simplify our calculation of regression equations and prediction intervals.
Effect of Age
As noted above, we analyzed our data for estimated LV mass
in the Group I subjects according to the general regression
equation:
LV mass = B(BSA) + A ± C.
FIGURE 1. Diagram of the methods used for measuring left ventricular dimensions in diastole flvtdj and in systole (Ivtdj, as well
as the diastolic thickness of the ventricular septum (vs) and the
posterior wall (pw). (Reprintedfrom Henry et al.,'3 with permission of the American Heart Association.)
When the older adult subjects were grouped into six age groups
(as noted in Table 1), the slope of the regression relationship for
estimated LV mass versus age was found to be independent of
age (p><3.05), whereas the intercept showed significant variation with age (p< 0.01). Therefore, we assumed that the intercept (A), but not the slope (B), was influenced by age, and that
the width of the 95% prediction interval ( ± C) was constant and
not appreciably influenced by age or BSA. Table 2 summarizes
the value for B (124) and C ( ± 58 g) in our equation, as well as
the values of A derived for each age group.
Figure 3 depicts the age-related changes in estimated LV
mass for our older adult subjects (Group I). For this figure, the
regression equation for LV mass has been analyzed for BSA
D-38
ECHOCARDIOGRAPHY
SUPPL II HYPERTENSION, VOL 9, No
FIGURE 2. Plot of estimated left ventricular mass versus
the linear function of body surface area in our younger
normal subjects (Group II). The dashed lines represent the
95% prediction intervalsfor normal measurements using the
assumption that the prediction intervals varied proportionally with the mean. (Reprinted from Henry et al.,'3 with
permission of the American Heart Association.)
2, FEBRUARY
1987
y
A
300 -
A
s'
A
At
« 250 -' ' *
O
yy
2200
150 _-
ss
ss
LAAA
A
A
100 --
A „-"'
50 -
I
.
.
05
.
1
,
.
10
1 ,
1
i
,
15
BODY SURFACE AREA IN SQUARE METERS
Downloaded from http://hyper.ahajournals.org/ by guest on June 15, 2017
TABLE 1.
Population Characteristics for Group I
BSA (m 2)
Sex
Age (years)
Blood pressure (mm Hg)
DBP
Mean
SBP
Heart
rate
(beats/rrun)
Men
Women
Range
Mean
21-30
15
10
1.48-2 25
1.80
120
71
87
74
31-40
9
15
1.47-2.38
1.80
117
76
90
69
41-50
16
13
1.42-2.20
1.83
120
76
91
73
51-60
19
10
1 51-2.17
1.84
120
71
87
73
61-70
9
g
1.52-2.10
1.75
130
72
92
73
71-97
10
2
1.42-2.13
1.73
131
61
84
70
72
89
72
1.42-2.38
122
1.80
BSA = body surface area, SBP = systolic blood pressure; DBP = diastolic blood pressure.
Total (n=136)
58
78
TABLE 2. Regression Equation for Left Ventricular Mass in Group I Subjects
Slope
(B)
LV mass = 124 (BSA) +
21-30
31-40
41-50
51-60
61-70
71 +
Prediction
interval
(±C)
-9.87
- 1 98
8.56
10.8
26.8
22.2
±58 g
Intercepts (A) by age range (yr)
BSA = body surface area.
values of 1.4, 1.8 and 2.2 m2. The mean value for each age
group is plotted at the mean age in each group. The figure
depicts the mean and the 95% prediction interval for each age
group at each of the three BSAs. Note that for a BSA of 1.8 m2
there was a 15% increase in mean LV mass from 214 g in the
youngest to 246 g in the oldest age group.
Discussion
Echocardiographic studies of an older and younger population without clinically apparent heart disease demonstrated that
estimated LV mass showed a small but progressive increase
with increasing age. This increase in LV mass was related to an
increase in wall thickness (mean increase, 19%) from the 30and-under to the over-70 age groups that was relatively greater
than the decrease in LV diastolic dimension (mean decrease,
6%) over the same age range.14 These findings are consistent
with the previously reported necropsy findings of Roberts and
Perloff." A recent study by Devereux and co-workers17 of 225
normal subjects, age range 18 to 72 years, suggested that age
had no significant effect on measurements of LV mass. However, two other echocardiographic studies of normal individuals
have concluded that both posterior LV wall thickness23 and LV
mass26 increase significantly with advancing age.
In our studies estimated LV mass varied linearly with BSA
but was 7.2% less in adult normal females (/?<0.05) and 3.6%
less in younger normal females (p = 0.05) than in males of the
same age and BSA. These findings suggest a small but significant relationship between sex and echocardiographic measurements of LV mass that is not eliminated by correction for BSA.
In the study by Devereux et al., 17 LV dimensions and mass
were found to be closely related to BSA. Although indexing by
BSA eliminated sex differences in LV wall thickness and internal dimension, a significant sex difference remained in LV mass
index (89 ± 2 1 g/m2 in men vs 69 ± 19 g/m2 in women,
p<0.0001). When LV mass was indexed by lean body mass,
however, no sex difference persisted. The smaller sex difference (7.2%) found in our adult normal population as compared
to that reported by Devereux et al. (20%) may be related to the
smaller number of women in our older age group and perhaps
also to biologic variability in the relationship between BSA and
EFFECT OF AGE, SEX, AND BODY SIZE ON LEFT VENTRICULAR MASS/Gardin et al.
—-<
300 -
•
r"
250 -
L7
——'
i—-<
3.
4.
1
.
t
^-—i
i-—-d
i.i"1-""^
2.2
150-
r—1 1 1-4
, 11
j
'
,
200 -
2.2
1
34
5.
6.
100-
7.
50-
8.
10
20
30
40
SO
60
70
80
90
Age in Years
Downloaded from http://hyper.ahajournals.org/ by guest on June 15, 2017
FIGURE 3. Plot of estimated left ventricular mass versus age. The
mean and95% prediction intervals are depictedfor each age group
at body surface areas of 1.4 m2 (dashed lines), 1.8 m2 (shaded area)
and 2.2 m2 (solid lines). (Reprinted from Gardin et al.14)
9.
10.
11.
12.
13.
lean body mass in the subjects in the two studies. Of interest,
Valdez and co-workers,l6 in their study of a normal population,
found differences between men and women in LV echocardiographic measurements such as wall thickness and internal dimension, but they noted that these differences were eliminated
by indexing for BSA.
These initial echocardiographic studies provide baseline normal data for use in quantitative evaluation of populations with
suspected increased LV mass (e.g., hypertensive subjects) and
for evaluation of serial changes in LV mass in response to
pharmacologic and other therapeutic interventions. Studies of
larger populations of subjects without clinical evidence of cardiac disease should provide additional information regarding the
applicability of our regression equations to large normal and
abnormal populations. In addition, provocative tests such as
exercise electrocardiography or stress echocardiography might
be helpful in further screening of subjects without overt cardiovascular disease in order to eliminate so-called normal subjects
with asymptomatic cardiovascular disease.
14.
15.
16.
17.
18.
19.
20.
21.
22.
Acknowledgments
The authors thank Valerie Williams, Pat Pierce, and Elena Guzman for
their expert typing assistance.
References
Dunn FG, Chandraratna P, deCarvalho JGR, Basta LL, Frohlich ED.
Pathophysiologic assessment of hypertensive heart disease with echocardiography. Am J Cardiol 1977;39:789-795
Schlant RC, Felner JM, Heymsfield SB, et al. Echocardiographic stud-
23.
24.
25.
26.
11-39
ies of left ventricular anatomy and function in essential hypertension.
Cardiovasc Med 1977;2:477-491
Savage DD, Drayer JIM, Henry WL, et al. Echocardiographic assessment of cardiac anatomy and function in hypertensive subjects. Circulation 1979;59:623-632
Hill LS, Monaghan M, Richardson PJ. Regression of left ventricular
hypertrophy during treatment with antihypertensive agents. Br J Clin
Pharmacol 1979;7(suppl 2):255S-259S
Devereux RB, Savage DD, Sachs I, Laragh JH. Effects of blood pressure control on left ventricular hypertrophy and function in hypertension
[Abstract]. Circulation 1980;62:36
Cohen A, Hagan AD, Watkins J, et al. Clinical correlates in hypertensive patients with left ventricular hypertrophy diagnosed with echocardiography. Am J Cardiol 1981;47:335-341
Drayer J, Gardin JM, Weber MA, Aronow WS. Changes in cardiac
anatomy and function during therapy with alpha-methyldopa: an echocardiographic study. Curr Ther Res 1982;32:856-865
Drayer J, Gardin J, Weber M, Aronow WS. Cardiac muscle mass
during vasodilation therapy of hypertension. Clin Pharmacol Ther 1983;
33:727-732
Drayer J, Weber MA, Gardin JM, Lipson JL. Effect of long-term
antihypertensive therapy on cardiac anatomy in patients with essential
hypertension. Am J Med 1983;75:116-120
Drayer JIM, Gardin JM, Weber MA. Anti-hypertensive therapy and
changes in echocardiographic left ventricular mass (LVM) [Abstract].
Clin Pharmacol Ther 1984;35:326
Roberts WC, Perloff JK. Mitral valve disease: a clinico-pathologic
survey of the conditions causing the mitral valve to function abnormally. Ann Inter Med 1972;77:939-975
Krovetz J. Age-related changes in size of the aortic valve annulus in
man. Am Heart J 1975;90:569-574
Henry WL, Ware J, Gardin JM, Hepner SI, McKay J, Weiner M.
Echocardiographic measurements in normal subjects: growth-related
changes that occur between infancy and early adulthood. Circulation
1978;57:278-285
Gardin JM, Henry WL, Savage DD, Ware JH, Burn C, Borer JS.
Echocardiographic measurements in normal subjects: evaluation of an
adult population without clinically apparent heart disease. JCU 1979;
7:439_447
Henry WL, Gardin JM, Ware JH. Echocardiographic measurements in
normal subjects from infancy to old age. Circulation 1980;62:10541061
Valdez RS, Motta JA, London E, Popp R. Evaluation of the echocardiogram as an epidemiologic tool in an asymptomatic population. Circulation 1979;60:921-929
Devereux RB, Lutas EM, Casale TN, et al. Standardization of M-mode
echocardiographic left ventricular anatomic measurements. J Am Coll
Cardiol 1984;4:1222-1230
Robinson CH. Normal and therapeutic nutrition. 14th ed. New York:
McMillan, 1972:703
Griffith JM, Henry WL. Switched gain — a technique for simplifying
ultrasonic measurement of cardiac wall thickness. IEEE Trans Biomed
Eng 1975;22:337-340
Feigenbaum H. Echocardiography. 2nd ed. Philadelphia: Lea & Febiger, 1976:464
Troy BL, Pombo J, Rackley CE. Measurement of left ventricular wall
thickness and mass by echocardiography. Circulation 1972;45:602-611
Rackley CE, Dodge HT, Coble YD, et al. Method of determining left
ventricular mass in man. Circulation 1964;29:666-671
Boothby W, Standiford RB. Nomographic charts for the calculation of
the metabolic rate by the gasometer method. Boston Med Surg J
1921;185:337-354
DuBois D, DuBois EF. Clinical calorimetry: X. A formula to estimate
the approximate surface area if height and weight be known. Arch Intern
Med 1960;17:863-871
Gerstenblith G, Frederiksen J, Yin FCP, Fortuin MJ, Lakatta EG,
Weisfeldt ML. Echocardiographic assessment of a normal adult aging
population. Circulation 1977;56:273-278
Lakatta EG. Alterations in the cardiovascular system that occur in advanced age. Fed Proc 1979;38:163-167
Effect of age, sex, and body surface area on echocardiographic left ventricular wall
mass in normal subjects.
J M Gardin, D D Savage, J H Ware and W L Henry
Downloaded from http://hyper.ahajournals.org/ by guest on June 15, 2017
Hypertension. 1987;9:II36
doi: 10.1161/01.HYP.9.2_Pt_2.II36
Hypertension is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231
Copyright © 1987 American Heart Association, Inc. All rights reserved.
Print ISSN: 0194-911X. Online ISSN: 1524-4563
The online version of this article, along with updated information and services, is located on
the World Wide Web at:
http://hyper.ahajournals.org/content/9/2_Pt_2/II36
Permissions: Requests for permissions to reproduce figures, tables, or portions of articles originally
published in Hypertension can be obtained via RightsLink, a service of the Copyright Clearance Center,
not the Editorial Office. Once the online version of the published article for which permission is being
requested is located, click Request Permissions in the middle column of the Web page under Services.
Further information about this process is available in the Permissions and Rights Question and Answer
document.
Reprints: Information about reprints can be found online at:
http://www.lww.com/reprints
Subscriptions: Information about subscribing to Hypertension is online at:
http://hyper.ahajournals.org//subscriptions/