Echocardiographic Assessment of Cardiac Anatomy

and Function in Hypertensive Subjects

DANIEL D. SAVAGE, M.D., PH.D., JAN I.M. DRAYER, M.D., WALTER L. HENRY, M.D.,
EMMETT C. MATHEWS, Jr., M.D., JAMES H. WARE, PH.D., JULIUS M. GARDIN, M.D.,
ESTELLE R. COHEN, B.S., STEPHEN E. EPSTEIN, M.D., AND JOHN H. LARAGH, M.D.

SUMMARY Cardiovascular complications are a major source of morbidity and mortality in hypertensive
patients. To assess the prevalence of anatomic and functional abnormalities of the heart in such patients, we
studied 234 asymptomatic subjects with mild-to-moderate systemic hypertension by echocardiography. After
adjusting the echocardiographic values for age and body surface area, we found abnormally increased ventric-
ular septal and/or posterobasal free-wall thickness in 61% of the hypertensive subjects. We found increased
left atrial, aortic root, and left ventricular internal dimension (at end-diastole) in 5-7%, and decreased mitral
valve closing velocity (E-F slope) and left ventricular ejection fraction were noted in six and 15% of the sub-
jects, respectively. Four percent of the patients had disproportionate septal thickening (i.e., ventricular septal-
to-left ventricular free-wall thickness ratio > 1.3). In contrast to the high prevalence of cardiac abnormalities
detected by echocardiography, less than 10% of the hypertensive subjects had abnormal 12-lead ECGs or ab-
normal chest x-rays. These findings demonstrate a high prevalence of cardiac abnormalities in a population of
Downloaded from http://circ.ahajournals.org/ by guest on February 5, 2018

asymptomatic hypertensive subjects. These abnormalities can be detected by echocardiography before they are
otherwise apparent.

ABNORMALITIES OF CARDIAC STRUCTURE compared the relative sensitivity of the echocardio-

and function occur in response to sustained hyperten-
sion. Before echocardiography, however, it was
difficult to detect early cardiac complications in
patients with hypertension. As a result, the first in-
dication of cardiac involvement in many patients was
the development of overt congestive heart failure.
Echocardiography, which provides a direct means
of assessing both functional and anatomic abnor-
malities of the heart, has recently been applied to the
study of hypertensive subjects. The results of these
studies have suggested that increases in left atrial
dimension' and left ventricular wall thickness'"3 and
decreases in closing velocity (E-F slope) of the
anterior leaflet of the mitral valve3 occur frequently
even in the absence of other signs of cardiac involve-
ment. However, similar changes have been reported to
occur in normal subjects as a result of aging.4 I
To determine whether some or all of the abnor-
malities reported in hypertensive subjects could be
related to age rather than blood pressure elevation, we
evaluated a population of subjects with mild-to-
moderate hypertension and compared the results with
normal data using regression equations that accounted
for both body surface area and age. We also examined
the relation of the various echocardiographic
measurements to severity of hypertension. Finally, we
From the Cardiology Branch, National Heart, Lung, and Blood
Institute, NIH, Bethesda, Maryland, and the Cardiovascular
Center, The New York Hospital-Cornell Medical Center, New
York, New York.
Presented in part at the 49th Annual Scientific Sessions,
American Heart Association, November 1976, Miami Beach,
Florida.
Address for reprints: Dr. Daniel D. Savage, Cardiology Branch,
National Heart, Lung, and Blood Institute, National Institutes of
Health, Building 10, Room 7B-15, Bethesda, Maryland 20014.
Received March 21, 1978; revision accepted November 7, 1978.
Circulation 59, No. 4, 1979.
623
gram, the standard 12-lead ECG, and the routine
chest x-ray for detecting cardiac abnormalities in
hypertensive subjects.
Materials and Methods
Subjects
Two hundred sixty hypertensive subjects and 124
normotensive subjects gave informed consent and
were studied by echocardiography. The hypertensive
subjects were selected from the outpatient hyperten-
sion clinic at the New York Hospital - Cornell
Medical Center. The following patients were ex-
cluded: 1) patients with malignant hypertension; 2)
patients on antihypertensive medications other than
diuretics, propranolol, clonidine, a-methyldopa or
guanethidine; and 3) patients with a history of angina
pectoris, myocardial infarction, atrial fibrillation or
congestive heart failure (i.e., orthopnea or paroxysmal
nocturnal dyspnea). Twenty-six of the 260 subjects
(10%) were excluded because their echocardiograms
were technically unsatisfactory for measurement of
the anatomic variables included in the study. The re-
maining 234 hypertensive subjects included 134 on
antihypertensive therapy at the time of the echocar-
diographic study and 100 who had had therapy discon-
tinued for at least 3 weeks before the study. The selec-
tion and evaluation of the normotensive subjects are
extensively described in a previous report.4
Clinical characteristics of the hypertensive and nor-
motensive subjects are shown in table 1. Blood
pressures were measured in the supine position with
standard sphygmomanometric methods. Mean
arterial pressure was estimated from the sum of the
diastolic pressure and one-third of the pulse pressure.
Seventy-three of the 134 patients (54%) on antihyper-
tensive therapy and 53 of the 100 patients (53%) not on
 624
TABLE 1. Clinical Characteristics of Subjects
Hypertensive subjects
Therapy
discontinued
Number of subjects 100
Sex-males:females 66:34
Age (years)
Mean - SD (range) 47 13 (19-69)
Race-whites:blacks 89:11
Systolic blood pressure
(mm Hg)
Mean SD (range) 150 20 (112-190)
Diastolic blood pressure
(mm Hg)
Mean - SD (range) 95 10 (80-150)
Number of subjects with
hypertensive retinopathy
Keith-Wagener grade > 2* 16 (16%)
Number of subjects with
Downloaded from http://circ.ahajournals.org/ by guest on February 5, 2018
serum creatinine
> 1.7 mgWc (range) 77 = 7 _ (1 .7-2.7 mg%) 16 = 7 % (1.7-2.2 mg %)
*Only one hypertensive subject who was continued on therapy had grade 3 hypertensive retinopathy. All
other subjects had grades < 2.
such therapy had diastolic blood pressures lower than
95 mm Hg at the time of echocardiographic study.
However, all hypertensive subjects had had diastolic
blood pressure . 95 mm Hg on two or more
measurements within 1 year before study. Only two
hypertensive subjects had diastolic blood pressures
greater than 130 mm Hg at the time of echocardio-
graphic study. None of the subjects had evidence of a
known secondary cause of hypertension by history,
physical examination or routine laboratory tests.
One hundred twenty-eight of the hypertensive sub-
jects had systemic arterial blood pressure recorded
(sitting) on two or three clinic visits during the 6
months before echocardiographic study (with no
change in treatment status during these visits). The
average of these blood pressures was used for assess-
ment of the relation of echocardiographic measure-
ments to severity of hypertension.
Echocardiographic Measurements
M-mode echocardiograms were performed with the
patient on his or her left side. An Aerotech transducer
ultrasound receiver, a Honeywell 1856 Line Scan
Recorder, a Hewlett-Packard X-Y display, and a
custom-built video amplifier were used. Studies were
performed with the transducer in the fourth intercos-
tal space near the left sternal edge. If necessary, the
transducer was moved laterally and/or to a different
interspace so that both mitral leaflets could be
visualized with the transducer perpendicular to the
chest wall. To obtain consistent and reproducible
recordings, the T-scan method was used.6
Figure 1 shows illustrative echocardiograms which
indicate where measurements were taken. The
thickness of the ventricular septum and posterobasal
left ventricular wall were measured at or slightly
CIRCULATION VOL 59, No 4, APRIL 1979
Therapy Normotensive
continued subjects
134 124
81:53 69:55
48 9 (24-67) 44 = 14 (19-70)
123:11 122:2
144 - 23 (106-218) 121 13 (94-166)
94 12 (68-140) 73 = 10 (50-90)
35 (26%)
below the tips of the mitral valve leaflets. Both
thicknesses were measured in the portion of the car-
diac cycle that occurs after rapid ventricular filling but
before atrial systole (fig. lA).6 Posterobasal left ven-
tricular wall was measured near the damped portion
of the record (fig. lA). Left ventricular transverse
dimensions at end-diastole and at end-systole were
measured in the same portion of the record and taken
as the maximal and minimal distances between sep-
tum and posterior left ventricular wall, respectively
(fig. lA).7 Left atrial dimension was measured as the
maximal distance between the posterior aortic root
wall and the posterior left atrial wall. Measurements
were taken in the damped portion of the record when
the ultrasonic beam passed through the aortic valve
leaflets.7-9 Aortic root dimension was measured in the
same portion of the recording. This measurement was
made from the midpoint of the line denoting the
anterior aortic root wall to the midpoint of the line
denoting the posterior aortic root wall at end-diastole.
The mitral valve E-F slope, corresponding to the rate
of early diastolic closure of the anterior leaflet of the
mitral valve, was measured at a point in which the ex-
cursion of the anterior mitral leaflet was maximal and
both leaflets were visualized (fig. lB).10
Values derived from the measurements included es-
timated left ventricular mass," percent fractional
shortening of the left ventricular transverse dimen-
sion," and left ventricular ejection fraction, which was
calculated using the cubed assumption to estimate left
ventricular volume.'3
Electrocardiographic Measurements
Two hundred seventeen of the hypertensive subjects
had standard 12-lead ECGs within 1 day to 12 weeks
of the echocardiographic study (median 1 day). No
 ECHOCARDIOGRAPHY IN HYPERTENSIVE SUBJECTS/Savage et at.
Downloaded from http://circ.ahajournals.org/ by guest on February 5, 2018
FIGURE 1. M-mode echocardiograms from hypertensive
subjects demonstrating where dimensions were measured.
VS= ventricular septum; PW left ventricular postero-
basal free wall; L VTDD = left ventricular transverse dimen-
sion at end-diastole; L VTD/, left ventricular transverse
dimension at end-systole. E-F slope (and not E-F0) was
measured. ICM indicates I cm.
subject had a change in treatment status during the
time between electrocardiographic and echocardio-
graphic study. The Romhilt-Estes scoring system'4
was used to assess electrocardiographic evidence of
left ventricular hypertrophy. A point score of four in-
dicated' probable left ventricular hypertrophy and five
or more points indicated definite left ventricular
hypertrophy.
Chest Roentgenographic Measurements
One hundred sixty-eight of the hypertensive subjects
had routine posteroanterior chest x-rays within 1 day
to 12 weeks of the echocardiographic study (median 1
day). As with the ECG, no patient had a change
in treatment status during the time between the chest
x-ray and the echocardiogram. A cardiothoracic ratio
> 0.5 was considered evidence of left ventricular
enlargement."
625
TS 0'71
.JF,*, \' >
Statistical Analyses
Statistical analyses used included standard regres-
sion analysis and analysis of covariance. Where ap-
propriate, the t test or the chi square test was used to
assess statistical significance.
Results
Effects of Body Surface Area and Age
on Echocardiographic Measurements
of Hypertensive Subjects
To assess the effect of body surface area on echocar-
diographic measurements, each hypertensive subject
was placed into 1 of 5 decades by age. For each age
decade the echocardiographic measurements were
plotted vs the previously derived appropriate root
function'6 of body surface area. The slopes from these
plots were compared with those of similar plots
 626 CIRCULATION VOL 59, No 4, APRIL 1979

18 r TABLE 2. Blood Pressure Data of Hypertensive Subjects in

Each Age Group

16 F (24) Age group Number of Blood pressure (mm Hg)

(years) subjects Systolic* Diastolic*
E
E (64) '(88) 21-30 29 137 12 91 9§
z
en
14 k 31-40 29 138 20 94 16
UJ
z 41-50 64 143 14 96 9
C-) 51-60 88 150 19t 96 9
I- 12 F
I
61-70 24 167 19: 95 8
M: (29) (24 *Mean f SD.
U) 10 F
0-
J

z
w
8
;I ,:,(!,1I7I tMean systolic blood pressure was significantly higher than
that of each of the three younger age groups (p <0.01).
tIean systolic blood pressure was significantly higher than
that of eaah of the four younger age groups (p <0.001).
§Mean diastolic blood pressture was significantly lower than
those of the subjects aged 41 to 60 (p <0.02) but was similar
to suibjects in the other age groups (p >0.05).
6
Downloaded from http://circ.ahajournals.org/ by guest on February 5, 2018

* Mean For Hypertensive Subjects

o Mean For Normotensive Subjects
I_ 20 30
I
40 50
I I
60
I
70
subjects which were independent of body surface area.
To assess the effect of age on the echocardio-
AGE IN YEARS graphic measurements of hypertensive subjects, the
FIGURE 2. Effect of age on ventricular sep tal thickness In measurements were each adjusted to a body surface
area of 1.8 m2 using regression equations derived in
hypertensive and norrnotensive subjects. Each echocardio-
graphic value was adjiusted by regression analysis to a body our laboratory. These adjusted values were sub-
surface area of 1.8 n 2*. Numbers
. . . divided into 5 decades as before, and the values were
in parentheses indicate. averaged. The adjusted values for ventricular septal
numbers of subjects. JIn this and all subsequent figures the
symbol I = mean ± SL thickness are plotted vs age in figure 2. As in the study
of normal subjects,4 ventricular septal thickness, free-
wall thickness, left atrial dimension and aortic root
dimension increased, while mitral valve E-F slope
derived from norm(otensive subjects.4 No significant decreased with increasing age. Left ventricular
difference was found between these slopes from hyper- transverse dimension (at end-systole and at end-
tensive and normo)tensive subjects for any of the diastole), ejection fraction and percent fractional
echocardiographic nneasurements, including left ven- shortening showed no significant changes with age.
tricular transverse dimension at end-diastole and at The magnitude of change in each echocardiographic
end-systole, ventrictilar septal thickness, left ventric- measurement with age is summarized in figure 3.
ular free-wall thickmness, left ventricular mass, left Blood pressure data for each of the five age groups is
atrial and aortic ro)ot dimensions, and mitral valve given in table 2. The changes in echocardiographic
E-F slope. As in nor*motensive subjects,4 ejection frac- measurements with age remained evident after adjust-
tion and percent fraictional shortening were the only ment (using covariate analysis) for differences in blood
echocardiographic nneasurements in the hypertensive pressures among the five age groups.

SEPTAL THICKNESS
z 20 LEFT ATRIAL DIMENSION
_FREE=-WREE-WALL
_ THICKNESS FIGURE 3. Effect of age on echocar-
<C 10 __-
-1
AORTIC ROOT DIMENSION
00O 0- : 7--7_ -._. diographic measurements in hypertensive
FRACTION subjects. Mean echocardiographic values for
ZHF
-Z -10 each age decade is compared with that of
wF
z
Z w
-20 hypertensive subjects in the 21-30 age group
ID -20) after adjustment of echocardiographic
w
values to a body surface area of 1.8 m2,
-40 *For mitral valve E-F slope n = 227.
-50

AGE IN YEARS
 E
E
z

w
z
I
-j
>
-J
30 F

20

10 r-
A
ECHOCARDIOGRAPHY IN HYPERTENSIVE SUBJECTS/Savage et al.

SEPTAL THICKNESS

- .40
I-IDooe"

---410690"

3,-,'--Ab9*AAA
1
FREE-WALL THICKNESS

D.)O"

...

.Q&l 1.8486.;.16.6665=2612122t=
.'

,
-1

, 4040416SAAA
700

600 F

500

400

300
k
LV MASS (grams)

'A
A

4!!.A
A

,)1-1, *A

A
-.o

.-
*AAA
_''-1041."
.AA

..* 1000"

.400AAAAA
40**A&"
.-4MDA&"
175

150

125

100_

75
'(206)

,
MITRAL VALVE
E-F SLOPE* (mm/sec)

.t:

ii'

X-4-
4.X
±4
..
627
Downloaded from http://circ.ahajournals.org/ by guest on February 5, 2018

-j OAA&A"

Ljl .I
A
00" e
200 O"
A
4MAAA
50 IL
9O"
,1R
I ;.j 35
O 1 o1 I4
FIGURE 4. Distribution of absolute echocardiographic measurements of ventricular septum, left ventric-
ular free wall, mitral valve E-F slope, and derived left ventricular mass in 234 hypertensive subjects. On this
and all subsequent figures: open circles = patients on no medication, closed circles = patients on antihyper-
tensive medication other than propranolol, and closed triangles = patients on propranolol with or without
other antihypertensive medication. LV = left ventricular. *For mitral valve E-F slope n = 227.

LEFT ATRIAL AORTIC ROOT

LVTDD LVTDS DIMENSION DIMENSION
I

60 OA

A.
I.: A
OSAA
0
00"

E 50 .,. OAAAAAA
E 41408SAAAA "A

A
z OOAAAAAA
40 S"
SO
.
*-0
-:10 -:4

z
w 4a"

0 .
-:-,. -,( .-

-..
..-

I
I

.-.

z 301- O"
. OA
-4"WAA
1-
.- .- , 's &A"

w 400
-

A&
0 -A
00

20 F O"

10i

n I
u
FIGURE 5. Distribution of absolute echocardiographic measurements of left ventricular dimensions at end-
diastole and end-systole, and left atrial and aortic root dimensions in 234 hypertensive subjects.
L VTDD = left ventricular transverse dimension at end-diastole; L VTD, = left ventricular transverse dimen-
sion at end-systole.
 628 CIRCULATION
LV EJECTION LV FRACTIONAL
FRACTION SHORTENING
z
Downloaded from http://circ.ahajournals.org/ by guest on February 5, 2018
0 50
0u
0 1
FIGURE 6. Distribution of ejection fraction and percent
fractional shortening measured by echocardiography in 234
hypertensive subjects. Shaded areas represent the 95%
prediction interval derived from normal data. LV left
ventricular.
Prevalence of Anatomic and Functional Echocardiographic
Abnormalities in Hypertensive Subjects
Figures 4, 5 and 6 show the distribution of the ab-
solute echocardiographic measurements of the hyper-
tensive subjects. Since the frequencies of echocardio-
graphic abnormalities in subjects who remained on
therapy, including the subgroup on propranolol, was
similar to that of subjects who had therapy discon-
tinued, we considered their data together. The effect of
hypertension on the various echocardiographic
measurements was assessed by calculating the ratio of
the actual measurement to the predicted value for
each hypertensive subject. The predicted value was
calculated using our previously derived regression
equations4 which account for the effects of both age
and body surface area. Figures 7 and 8 show the dis-
VOL 59, No 4, APRIL 1979
TABLE 3. Prevalence of Echocardiographic Abnormalities in
234 Hypertensive Subjects
Percent of
Echocardiographic measurement patients*
Ventricular septal thickness 50
Left ventricular free-wall thickness 61
Disproportionate septal thickeningt 4
Left ventricular mass 51
Left ventricular transverse dimension
at end-diastole 5
Left ventricular transverse dimension
at end-systole 12
Left atrial dimension 5
Aortic root dimension 7
Ejection fraction 15
Percent fractional shortening 13
MIitral valve E-F slope: 6
*Echocardiographic values were considered abnormal if
they were above (or below for ejection fraction, percent
fractional shorteniing and mitral valve E-F slope) the 95%0
prediction interval derived from normotensive subjects.
tDisporportionate septal thickening = ratio of ventricular
septal thickness to free-wall thickness > 1.3.
tThe number of subjects who had mitral valve E-F slope
measured was 227.
tribution of these ratios for each of the echocardio-
graphic measurements of the hypertensive subjects.
Table 3 summarizes the percentage of echocar-
diographic abnormalities that fall above (or below for
ejection fraction, percent fractional shortening, and
mitral valve E-F slope) the 95% prediction interval
derived from the normotensive subjects.
Mean ventricular septal thickness, left ventricular
free-wall thickness, and left ventricular mass for the
hypertensive subjects were significantly greater than
those of normal subjects (p < 0.001) (fig. 7). Mean
values of left ventricular transverse dimensions (at
end-diastole and at end-systole), left atrial and aortic
root dimensions, left ventricular ejection fraction, and
percent fractional shortening for the hypertensive sub-
jects were not significantly different from values in
normal subjects (figs. 6 and 8). There was no signifi-
cant correlation (positive or negative) between ventric-
ular septal or left ventricular free-wall thickness and
ejection fraction. Although only a small number of
hypertensive subjects had a mitral valve E-F slope
below the 95% prediction interval, the mean E-F slope
of hypertensive subjects was significantly lower than
that of normal subjects (p < 0.001) (fig. 7).
Nine hypertensive subjects had ventricular septal
thickening that was disproportionate to the left ven-
tricular free-wall thickness (i.e., septal-free wall ratio
> 1.3). Their measured septal thicknesses ranged
from 15-27 mm, with septal-free wall ratios from
1.3-1.9. None of the nine had systolic anterior motion
of the anterior leaflet of the mitral valve. However,
one subject who had concentric left ventricular wall
thickening did have systolic anterior motion of the
anterior mitral leaflet.
 Lu
225

200

175

1 125 t
F- 125
O'-~ ~ ~ -. .:- . -. .
SEPTAL
THICKNESS
*(244)
ECHOCARDIOGRAPHY IN HYPERTENSIVE SUBJECTS/Savage et al.

. E W
S
.
FREE-WALL
THICKNESS

T Wt
W ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~..........
LV MASS
-,(299)
(230)

E
MITRAL VALVE
E-F SLOPE*

...
FIGURE 7. Distribution

slope.
of
diographic ventricular septal thickness, left
r 0 ;|; : ventricular free-wall thickness, left ventric-
............ular mass, and mitral valve F-F slope in 234
629

echocar-

1 - -0 --p°0 20 0 0 0 0 0 040:0 0 0 0 00000000000:......hypertensive subjects. Each value is plotted

Downloaded from http://circ.ahajournals.org/ by guest on February 5, 2018

O~~~~~~~~~~~~~~~~~~~~~~~..... ...

25 .

D ,'.-~~~~~~... __ . ....---

]<7Xr
LEFT ATRIAL AORTIC ROOT
LVTDD LVTD5 DIMENSION DIMENSION

a-~ ~ ~~~~~~.... ie rmnrmldt.Sae ra

17

< 150

Z1252. -J yPL
.....0.

< 50___
Z
z
w.25___

o_ _ _ I I I
FIGURE 8. Distribution of echocardiographic left ventricular transverse dimension at end-diastole
(L VTDD) and at end-systole (L VTD3), and left atrial and aortic root dimensions in 234 hypertensive sub-
jects. Each value is plotted as a percentage of the predicted value determined from normal data. Shaded
areas represent the 95% prediction interval derived from normal data.
 630 CIRCULATION VOL 59, No 4, APRIL 1979

TABLLE 4. Correlation Coefficients of Blood Pressure vs Echo- Left ventricular enlargement was detected by chest
cardiographic .Ieasurcertents in 128 Hypertensive Subjects x-ray in nine of 168 hypertensive subjects (5%) who
Systolic Diastolic Mean had recent chest x-rays. Six of the nine subjects (67%)
blood blood arterial had increased left ventricular wall thickness and in-
pressure pressure pressure creased left ventricular mass by echocardiogram. Left
Ventricular septal ventricular enlargement was detected by chest x-ray in
thickness 0.209* 0.346t 0.314$ only six of 81 hypertensive subjects (7%) who had left
Left ventricular ventricular enlargement (at end-diastole) or increased
free-wall thickness 0.291$ 0.4681 0.430$ left ventricular mass by echocardiogram, or both.
Left ventricular mass 0.231t 0.319$ 0.311t
Left ventricular transverse Discussion
dimension at end-diastole 0.022 -0.026 -0.003
Left ventricular transverse Previous studies of normotensive subjects without
dimension at end-systole - 0.083 0.010 -0.040 clinically apparent heart disease have shown changes
Left atrial dimension 0.204* 0.166 0.208*
in echocardiographic measurements with increasing
age and body surface area.4' I The changes associated
Aortic root dimension -0.082 0.010 -0.040 with increasing age include increased aortic root and
Ejection fraction 0.120 -0.029 0.050 left atrial size, increased left ventricular wall
Percent fractional thickness, and decreased mitral valve E-F slope. In the
Downloaded from http://circ.ahajournals.org/ by guest on February 5, 2018

shortening 0.136 -0.032 0.058 present study echocardiographic measurements of

Mitral valve E-F slope -0.001 -0.064 -0.037
*p <0.05.
tp <0.01.
$p <0.001.
Relation of Echocardiographic Measurements
to Severity of Hypertension
One hundred twenty-eight of the hypertensive sub-
jects had systemic arterial blood pressure recorded on
two or three clinic visits during the 6 months before
echocardiographic study (with no change in treat-
ment status during these clinic visits). There was only
a modest (but statistically significant) correlation
between the averaged mean arterial pressures
recorded during these visits and the ratios of actual to
predicted values for ventricular septal thickness, left
ventricular free-wall thickness, left atrial dimension
and left ventricular mass (table 4). No significant cor-
relations were found between mean arterial blood
pressure and aortic root dimension, left ventricular
transverse dimension (at end-systole or end-diastole),
left ventricular ejection fraction or mitral valve E-F
slope.
Comparison of the Echocardiogram with the 12-Lead
Electrocardiogram and the Chest X-ray
The echocardiogram was compared with both the
standard 12-lead ECG and the routine chest x-ray for
detecting cardiac abnormalities in hypertensive sub-
jects. Probable or definite left ventricular hyper-
trophy was detected by ECG in seven of the 217 hyper-
tensive subjects (3%) who had recent ECGs. Six of
these seven subjects (86%) had increased left ventric-
ular mass by echocardiogram and the seventh had an
echocardiographic left ventricular mass at the upper
limit of normal (as well as increased left ventricular
free-wall thickness). Probable or definite left ventric-
ular hypertrophy by ECG was detected in only six of
108 hypertensive subjects (6%) who had increased left
ventricular mass by echocardiogram.
hypertensive subjects showed changes with increasing
age and body surface area similar in direction and
degree to those seen in normotensive subjects. The
magnitude of these changes suggests that the effects of
both age and body surface area should be accounted
for in any attempt to assess the effects of hyperten-
sion on echocardiographic measurements. This was
accomplished in the present study by using the ratio of
actual to predicted values. These predicted values were
based on previously derived regression equations from
normal subjects.4
Using this approach, structural abnormalities of the
heart were detected in more than 60% of our popula-
tion of subjects with mild-to-moderate hypertension.
The most frequent abnormalities were ventricular sep-
tal and left ventricular free-wall thickening and in-
creased left ventricular mass. Similarly, Schlant et al.2
found increased left ventricular mass by echocardiog-
raphy in 36 of 73 (49%) of their hypertensive subjects.
The mean values and scatter of the data of left atrial,
aortic root, end-diastolic and end-systolic left ventric-
ular dimensions for our hypertensive population were
similar to those found for the normotensive popula-
tion. If published criteria7 for left atrial enlargement
(even those adjusted for body surface area) were used
for our hypertensive population, the prevalence of left
atrial enlargement would have been three times as
high as that found after accounting for both age and
body surface area (i.e., 15% rather than 5%). Thus, left
atrial enlargement was uncommon in our hyperten-
sive subjects, whether or not they had left ventricular
hypertrophy, and did not appear to be an early in-
dicator of cardiac involvement in hypertensive disease,
as we previously thought.'7 Dunn et al.' found larger
echocardiographic left atrial dimensions in hyperten-
sive subjects than in normal subjects. However, the
hypertensive subjects in their study were older than the
normal subjects. Thus, the larger left atrial dimen-
sions might at least partially be explained by changes
which normally occur with age, rather than by the
effects of hypertension.
In our study, modest but statistically significant
 ECHOCARDIOGRAPHY IN HYPERTENSIVE SUBJECTS/Savage et al.
correlations were found between systemic arterial
blood pressure and left ventricular wall thicknesses,
left ventricular mass and left atrial dimension. Schlant
et al.2 found poorer correlations between left ventric-
ular wall thicknesses and diastolic pressure than those
found in the present study. However, their series was
smaller than ours. In addition, they correlated left
ventricular wall thicknesses with a single diastolic
pressure measurement, while two or more measure-
ments from separate clinic visits were used in the pres-
ent study.
Disproportionate septal hypertrophy has been re-
ported in 47%,18 30%,19 10%,' and less than 1%2 of sub-
jects with hypertension. We found this abnormality in
4% of our hypertensive subjects (nine of 234). Exclu-
sion of subjects with angina pectoris, evidence of
myocardial infarction and severe hypertension from
our study population may have lowered the prevalence
of disproportionate septal thickening. However, it
Downloaded from http://circ.ahajournals.org/ by guest on February 5, 2018
would appear that this abnormality is uncommon in
subjects with mild-to-moderate hypertension.
We did not study the relatives of patients who had
disproportionate septal thickening and, therefore, do
not know whether this abnormality represents
genetically transmitted hypertrophic cardiomyopathy
(i.e., asymmetric septal hypertrophy), is an atypical
form of left ventricular wall thickening secondary to
hypertension, or neither. Of note, a recent necropsy
study of hypertensive subjects from our laboratory20
showed that the disproportionate septal thickening
present in two patients with hypertension did not have
the features characteristic of genetic hypertrophic car-
diomyopathy.
Left ventricular systolic function (as assessed by
ejection fraction) was reduced in 35 of the 234 hyper-
tensive subjects (15%) in the present study, and in two
subjects ejection fraction was considerably reduced
(23% and 35%). This was found despite our exclusion
of subjects with severe hypertension. Two other
echocardiographic studies, which included subjects
with more severe hypertension, showed that mean left
ventricular ejection fraction diminished as the severity
of hypertension increased." 2 Left ventricular diastolic
function (as measured by mitral valve E-F slope) was
below the lower limit of normal in only 14 of the 227
hypertensive subjects (6%), but overall was shifted
toward low-normal values so that the mean E-F slope
of the hypertensive population was significantly below
that of the normotensive population.
The echocardiogram appears to be superior to the
routine chest x-ray and the standard 12-lead ECG for
detecting cardiac abnormalities in hypertensive sub-
jects. Thus, while the echocardiogram identified an in-
crease in estimated left ventricular mass in over 50%
of the patients, the chest x-ray and the ECG identified
left ventricular enlargement or hypertrophy in less
than 10%. Echocardiograms of sufficient quality to
make all measurements could not be obtained in some
subjects. More important, ECGs and chest x-rays
may convey different information than echocardio-
grams, and therefore, echocardiograms should be used
in conjunction with them, rather than instead of them.
631
These findings are consistent with those of Schlant et
al.2 and Pisarczyk and Ross.2'
In summary, echocardiography identifies anatomic
and functional cardiac abnormalities in a large per-
centage of asymptomatic hypertensive subjects before
abnormalities are detected by ECG or chest x-ray.
This conclusion is strengthened by correcting the
echocardiographic values for effects of body surface
area and age. Whether echocardiography should be
used for routine screening of such subjects will depend
partly on the prognostic significance of these abnor-
malities.
Acknowledgments
The authors gratefully acknowledge the excellent technical
assistance of Joyce McKay and Cora Burn in obtaining the echocar-
diograms. The assistance of Pamela Peters and Rose Aceto in the
logistics of evaluating the large number of patients is greatly ap-
preciated. We also acknowledge Erica Brittain for help with the
statistical analyses.
References
1. Dunn FG, Chandraratna P, deCarvallo JGR, Basta LL,
Frohlich EF: Pathophysiologic assessment of hypertensive
heart disease with echocardiography. Am J Cardiol 39: 789,
1977
2. Schlant RC, Felner JM, Heynsfield SB, Gilbert CA, Shulman
NB, Tuttle EB, Blumenstein BA: Echocardiographic studies of
left ventricular anatomy and function in essential hypertension.
Cardiovasc Med 2: 477, 1977
3. Mashiro I, Kinoshita M, Tomonaga G, Hoshino T, Shimono
Y, Kusukawa R: Echocardiographic observations in hyperten-
sion. Jpn Circ J 39: 1097, 1975
4. 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. J Clin Ultrasound. In press
5. Gerstenblith G, Frederiksin J, Yin FCP, Fortuin NJ, Lakatta
EF, Weisfeldt ML: Echocardiographic assessment of a normal
adult aging population. Circulation 56: 273, 1977
6. Henry WL, Clark CE, Epstein SE: Asymmetric septal hyper-
trophy (ASH): echocardiographic identification of the pathog-
nomonic anatomic abnormality of IHSS. Circulation 47: 225,
1973
7. Feigenbaum H: Echocardiography, 2nd ed. Philadelphia, Lea
and Febiger, 1976, pp 236, 311, 464
8. Brown OR, Harrison DC, Popp RL: An improved method for
echocardiographic detection of left atrial enlargement. Circula-
tion 50: 58, 1974
9. Henry WL, Morganroth J, Pearlman AS, Clark CE, Redwood
DR, Itscoitz SB, Epstein SE: Relation between echocardio-
graphically determined left atrial size and atrial fibrillation.
Circulation 53: 273, 1976
10. Quinones MA, Gaasch WH, Waisser E, Alexander JK: Reduc-
tion in the rate of diastolic descent of the mitral valve echo-
gram in patients with altered left ventricular diastolic pressure-
volume relations. Circulation 49: 246, 1974
ll. Troy BL, Pombo J, Rackley CE: Measurements of left ventric-
ular wall thickness and mass by echocardiography. Circulation
45: 602, 1972
12. Fortuin NJ, Hood WP Jr, Sherman ME: Determination of left
ventricular volume by ultrasound. Circulation 44: 575, 1971
13. Popp RL, Harrison DC: Ultrasonic cardiac echography for
determining stroke volume and valvular regurgitation. Circula-
tion 41: 493, 1969
14. Romhilt DW, Estes EH Jr: A point-score system for the ECG
diagnosis of left ventricular hypertrophy. Am Heart J 75: 752,
1968
15. Paul LW, Juhl JH: The Essentials of Roentgen Interpretation,
2nd ed. New York and London, Harper and Row, 1965, p 784
 632 CIRCULATION VOL 59, No 4, APRIL 1979

16. Henry WL, Ware JH, Gardin JM, Hepner SI, McKay J,
Weiner M: Echocardiographic measurements in normal sub-
jects: growth-related changes that occur between infancy and
early adulthood. Circulation 57: 278, 1978
17. Drayer JIM, Savage DD, Henry WL. Mathews EC Jr., Laragh
JH, Epstein SE: Incidence of echocardiographic left ventric-
ular hypertrophy and left atrial enlargement in essential hyper-
tension. (abstr) Circulation 54 (suppl II): 11-233, 1976
18. Criley JM, Blaufuss AH, Abbasi AS: Nonobstructive IHSS.
Circulation 52: 963, 1975
19. Toshima H, Koga Y, Yoshioka H, Ayiyoshi T, Kimura N:
Echocardiographic classification of hypertensive heart disease:
a correlative study with clinical features. Jpn Heart J 16: 377.
1975
20. Maron BJ, Edwards JE, Epstein SE: Occurrence of dispropor-
tionate ventricular septal thickening in patients with systemic
hypertension. Chest 73: 466, 1978
21. Pisarczyk MJ, Ross AM: Cardiac measurements in hyperten-
sion: echocardiogram, electrocardiogram and x-ray com-
parison. (abstr) Am J Cardiol 37: 162, 1976

Carotid Baroreflex Function in Young Men

with Borderline Blood Pressure Elevation
DWAIN L. ECKBERG, M.D.
Downloaded from http://circ.ahajournals.org/ by guest on February 5, 2018

SUMMARY Carotid baroreflex function was assessed in 10 normotensive young men and 20 age-matched
subjects with borderline hypertension (successive blood pressures above and below 140/90 mm Hg) by measur-
ing sinus node responses to brief neck suction. Subjects with borderline hypertension were divided into two
equal groups according to their average systolic arterial pressures. Baroreflex responses were reset to function
at higher pressure levels than normal in subjects with mild borderline hypertension, but reflex sensitivity was
normal. Responses were also reset in subjects with more severe borderline hypertension, but reflex sensitivity
was subnormal. The results suggest that a gradation of baroreflex responsiveness exists among patients
classified as having borderline hypertension: Subnormal responsiveness was found in those subjects whose
resting average systolic arterial pressure was 140 mm Hg.

SOON AFTER THE CAROTID ARTERIAL
BAROREFLEX was discovered, Koch and Mies' and
Volhard' suggested that defective baroreflex buffering
of blood pressure might cause essential hypertension.
Despite numerous subsequent attempts to delineate
hypertensive mechanisms, the role of the arterial baro-
reflex in the pathogenesis of hypertension remains an
enigma. The validity of the theory of Koch and Mies
and Volhard has been questioned,3 but the theory has
not been discredited altogether. The study of Bristow
and co-workers4 supports their postulate: In their
study arterial baroreceptor-cardiac reflex responses
were found to be strikingly depressed in patients with
moderate (average mean arterial pressure 123 mm
Hg), sustained hypertension. In these patients, baro-
receptor reflex malfunction might have contributed to
the development of hypertension, or it might have
been a consequence of hypertension.
If a defective baroreceptor reflex mechanism causes
hypertension, subnormal baroreflex responses should
From the Cardiovascular Center, the Cardiovascular Division,
the Department of Internal Medicine, and the Veterans Administra-
tion and University Hospitals, the University of Iowa College of
Medicine, Iowa City, Iowa.
Supported by grants HL 14388 and HL 18083 from the Veterans
Administration and the National Institutes of Health.
Dr. Eckberg is a Clinical Investigator for the Veterans Ad-
ministration.
Address for reprints: Dwain L. Eckberg, M.D., Cardiovascular
Research, McGuire VA Hospital, Richmond, Virginia 23249.
Received August 7, 1978; revision accepted November 8, 1978.
Circulation 59, No. 4, 1979.
be found in patients with mild degrees of blood
pressure elevation. Data on this issue are conflicting.
Takeshita and associates5 found subnormal baroreflex
responses in young men whose average blood pressure
was 160/82 mm Hg, but Julius6 found normal
responses in borderline hypertensive patients whose
blood pressures were reported to be lower than those
studied by Takeshita and co-workers. I have
attempted to clarify this issue by using new tech-
niques to measure baroreflex responses of asymp-
tomatic young men whose blood pressures oscillate
above and below 140/90 mm Hg.
Methods
Seven intensities of neck suction were delivered
briefly to stretch carotid baroreceptors of young men
with normal blood pressures and young men with
borderline hypertension (defined as blood pressures
above and below 140/90 mm Hg on successive ex-
aminations). Sinus node responses were measured.
Hypertensive and Normal Volunteers
Volunteers, ages 19-25 years, were recruited from
900 university students whose blood pressures were
measured during registration. Blood pressures of each
volunteer were measured two to four (average 2.85)
times with subjects in the sitting position, after 10
minutes of rest. Weight, height and skin fold
thickness, measured in the midline midway between
the chin and the upper margin of the thyroid cartilage
were also measured.
 Echocardiographic assessment of cardiac anatomy and function in hypertensive subjects.
D D Savage, J I Drayer, W L Henry, E C Mathews, Jr, J H Ware, J M Gardin, E R Cohen, S E
Epstein and J H Laragh

Circulation. 1979;59:623-632
Downloaded from http://circ.ahajournals.org/ by guest on February 5, 2018

doi: 10.1161/01.CIR.59.4.623
Circulation is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231
Copyright © 1979 American Heart Association, Inc. All rights reserved.
Print ISSN: 0009-7322. Online ISSN: 1524-4539

The online version of this article, along with updated information and services, is located on
the World Wide Web at:
http://circ.ahajournals.org/content/59/4/623

Permissions: Requests for permissions to reproduce figures, tables, or portions of articles originally
published in Circulation 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 Circulation is online at:

http://circ.ahajournals.org//subscriptions/