Dynamic Analysis of Heart Sounds in Right

and Left Bundle-Branch Blocks

By JAN ORAVETZ, M.D., STANLEY WISSNER, M.D., BEATRICE ARGANO, M.D.,
AND ALDO A. LUISADA, M.D.

SUMMARY
A clinical, electrocardiographic, phonocardiographic, and arteriographic study was
made of 20 normal subjects, 20 patients with right bundle-branch block (RBBB) and
20 patients with left bundle-branch block (LBBB). Ventricular asynchronism was
documented in nearly all cases of bundle-branch block by the behavior of the arterial
pulse and the heart sounds. The first heart sound was of normal duration in all cases and
usually had the normal number and position of its three components. The first sound
had a normal relationship with the QRS component of the electrocardiogram in RBBB,
but its onset was markedly delayed in LBBB. This fact confirms the exclusive left-sided
origin of the first sound. The third (or ejection) component of the first sound had a
greater delay in LBBB, due to intraventricular block, and this delay was closely related
to that of the carotid pulse rise. Reversed splitting of the second sound in LBBB was
common.

Additional Indexing Words:

Auscultation Phonocardiogral Electrocardiography
Downloaded from http://ahajournals.org by on November 26, 2019

N the basis of theoretical considerations,
splitting of the first heart sound has
been considered typical of bundle-branch
block (BBB) for a long time in spite of in-
adequate evidence supporting this view.
Studies based on clinical auscultation in-
clude those of King,' King and McEachern,2
Cossio and associates,3 Laubry and Pezzi,4
Lewis,5 Lian and associates6 (quoted by
Calo7 ), and Segall.8 Various percentages of
splitting of the first sound were reported,
from an occasional finding4' 8 to a 56% in-
cidence.2
Graphic studies were made by Wolferth
and Margolies,9 Battro and associates,'0 Con-
tro and Luisada," Kelly,'2 Haber and Lea-
From the Divisions of Cardiovascular Research and
Cardiology (Medicine), The Chicago Medical School,
and the Division of Cardiology (Medicine), Mount
Sinai Hospital Medical Center, Chicago, Illinois.
This study was supported by Training Grants HE-
5002 and HE-5182, and Research Grant HE-09350
from the National Heart Institute, U.S. Public Health
Service.
Circulation, Volume XXXVI, August 1967
tham,13 and Brachetti and associates.'4 Three
components were described within the first
sound in BBB by Battro and associates,'0
and Contro and Luisadall (at that time, the
division of the normal first sound into three
components had not been recognized as yet).
Haber and Leatham'3 described frequent
splitting of the first sound in right bundle-
branch block (RBBB) in contrast with left
bundle-branch block (LBBB) where there
was no splitting.
Studies based on auscultation or on phono-
cardiography gave a clear-cut description of
the common abnormalities of the second
sound. These were (1) wide, fixed splitting
of the second sound with delay of the pul-
monary component in RBBB, and (2) single
second sound or reversed* splitting with de-
lay of the aortic component in LBBB.
*"Reversed splitting" is a new term, which
seems more appropriate than the older term "para-
doxical splitting."
275
 276 ORAVETZ ET AL.
X
While the nature of the changes of the "c

10 10
second sound has not been challenged, new Cd
0
0) 01
cO~ COI
concepts on the mechanism of production of 0) v

the first heart sound suggested a new study a

of the problem. O > 4 0

cl
C]
=o
H.
Methods
Our study was conducted on 60 subjects: 20 s:
u0

normal control subjects ranging in age between 01

Cv 10
r- COI
10
17 and 60 years; 20 patients with RBBB rang- -e
o
H. 0b
ing in age from 17 to 83 years; and 20 patients C
C
U.
U0
with LBBB ranging in age from 26 to 74 years.
Patients with rheumatic valvular disease, sub- -5Q)
aortic muscular stenosis, obvious myocarditis, a;
C.
00

or uncorrected congenital heart disease were ex- 0

cluded in order to avoid limited conduction C)
C.)

block or alterations of ventricular dynamics. n C=

Patients with systemic hypertension orpulmonary CZ
emphysema were excluded for the same reason. Co

The patients were selected following a clear- >

0CZ
cut electrocardiographic diagnosis of BBB made
according to Wilson's criteria. O a 0
Following a clinical and electrocardiographic ."
Co

r--
Co
V
study. phonocardiograms were recorded using 02-4
0
X..
a Sanborn multichannel apparatus that included
two 1700-B amplifiers with high-pass filters and C)
two dynamic microphones. Four tracings were Co q *
o~ 1-1
C
Cl
a
simultaneously recorded: an electrocardiogram, .) n C- CO
CO
1Z
Downloaded from http://ahajournals.org by on November 26, 2019

+1 +(
two phonocardiograms in the same frequency 0
C Ca

band, and a carotid tracing, recorded as low as C)

Co
Ca
.

possible in the neck. The sound tracings were 4-i

recorded, first, at the apex and over the second
left interspace; then, over the third left and the 1 1-

+I Cl1
second or third right interspaces. In addition
to a nonfiltered tracing, records were taken with 2..
*

nominal frequencies of 50, 100, 200, 400, and 600 +1 r24 +1

cps, respectively. CO
0
The data reported in table 1 summarize the ca
+1
10
10
tl) tlC C)
0z
averages of all cases. cn
Oa
The following measurements were made: (1) tZC)Ca
%'C4
Q-Ia interval; (2) Q-Ib interval; (3) Q-Ic inter- ca
t .<

val; (4) overall duration of first heart sound; m Ck co

*

a ~
CI

(5) interval between the onsets of the aortic 9: 0

O0

(A) and pulmonary (P) components of the sec-

ond heart sound or vice versa; (6) Q-carotid
pulse rise interval; and (7) Q-carotid pulse in- Ca
0IS -.,

cisura interval.
The measurements of heart sounds were 8 00
made mostly in the tracings with 0 or 50 filtra-
tion for the first sound and with 100 or 200 filtra-
tion for the second sound because, in BBB, high
filtration often gives a poorly defined first sound. .4.2 4.2 +4.
In normal subjects, the degree of splitting was -4 C)

CZ zm-P
'--
t) C)
measured only at the end of inspiration; in EBB,
the interval was measured in apnea. z = I

Circulation, Volume XXXVI, August 1967

 w_r-~ -oi_:+t
HEART SOUNDS
2
.PCG
2L
50/ P6C
5L 1 0: -
50/
t1,> ,iE,lVl-
7v
ECG =i
F 4-} =U7Z4r4.
Figure 1
Case of right bundle-brcanch block. QRS 1.12 sec The (a) coimponent of the first heart soulndl
starts 0.05 sec after the B rise. Phonocardiograms at second left (5 L) intercostal space (above)
Downloaded from http://ahajournals.org by on November 26, 2019
and apical (4 L; below) areas, both with 50 cps filter. Three components within the first sound
at apex, which has a normal amplitude.
Results
Auscultation
Right Bundle-Branch Block
Two of the 20 patients had muffled first
sounds while the other 18 had normal first
sounds; two patients out of 20 seemed to
have splitting of the first sound on ausculta-
tion. Splitting of the second sound was heard
in seven of the 20, and the splitting was
considered wider than normal in four pa-
tients. A third sound was heard in one pa-
tient, and an ejection souind at the base was
heard in one of the 20 patients.
Left Bunidle-Branch Block
Seven of the 20 patienits had a dull or
muffled first sound; two had a split first sounid
on auscultation. The second sound was con-
sidered as having a reversed splitting in six
of 20 patients; a third sound was heard in
one; and a fourth sound was heard in six.
Circulation, Volume XXXVI, August 1967
277
j'Ie
rz- ; =~~~~~~~~~~~~~~~~- = ...-..F
~z-_E.P
w.. -..-,-.
,.-+w(,Z
,_S
".----,,
An ejection sound was heard in one of 20
patients.
Phonocardiography
Normal Subjects
First Sound. The average duration of the
first heart sound was 126 msec. Three com-
ponents were recorded in all. The Q-Ia in-
terval averaged 59 msec; the Q-Ib, 86 msec;
and the Q-Lc, 125 msec.
Second Sound. Normal type of splitting
was found in 18 subjects after the end of
inspiration while two had a single second
sound (ages, 33 and 60 years, respectively).
The interval between- the A and P compo-
nents of the second souind averaged 42 msec.
Right Bundle-Branch Block (Figs. 1 and 2)
First Sound. The amplitude was normal
in all 20 patients. Three components were
recorded in all of them. The overall duration
 278 ORAVETZ ET AL.

I II1
m r-

... ... ..

..... ~~ ~~~~~~~~~~~
.. .. .... .....

PCG
4R3 u~~~~
f

PCG
4L3
0

Car.
Downloaded from http://ahajournals.org by on November 26, 2019

ECG
Figure 2
Case of right bundle-branch block. QRS 0.14 sec. The (a) component of the first heart souuLd
starts0.055 sec after the R rise. Phonocardiograms fourth right (above) and fourth left
at

(below) areas. Components (a) and (b) are visible at the left, cornponenlts (a) and (c) at the right
of the sternum (unfiltered tracings). The first sound has a normal amplitude. There is normcal
type of splitting of the second sournd, buit the interval between the A and the P components
is wide (0.07 sec). There is a third sounrd. The carotid rise starts 0.10 sec after the R rise.

of the sound varied from 80 to 160 msec interval between the A and P components
(average, 116 msec). The Q-Il interval av- was 57 msec.
eraged 55 msec with four of the 20 patients Third Sound. This sound was present in
having 70 msec while the others had lower one of the 20 patients.
figures. The Q-Ib interval averaged 97 msec Fourth Sound. This sound was present in
(average Ia-Ib interval, 42 msec). The Q-Ic five of the 20 patients.
interval averaged 133 msec (average la-Ic
interval, 78 msec). The Ic component (ejec- Left Butndle-Brcanch Block (Figs. 3 and 4)
tion sound) was abnormally large in three First Sound. The amplitude was decreased
of the 20 patients. in 10 of 20 patients, especially with filters
Second Sound. Normal type of splitting 200 higher. In all of them, three compo-
or
was found in 19 of the 20 patients, while a nents were recorded. The overall duration
single sound was present in one. The average of the first sound had an average of 110 msec.
Circulation, Volume XXXVI, Augwsa 1967
 HEART SOUNDS 279

I 21
.*..
.A I

I
PCG
2R3 Nr4
0
l!Pyrtfrnwwwa\ Yxv*lwa
vJAsk%

I
I 11
PCG
3L4 WVNJ .,
I
I
1.
0 LI I .L
lj,.
i

E C G
ECG~~~ ~~~~~~~~~~~~~~~~~~~~~~~~~~ .. ----|rf T

Figure 3
Case of left bundle-branch block. QRS 0 12 Phon-ocardiogicrans at second right (above)
sec.

and third left initercostal spaces (unfiltered). The first sound at the left of the sternutm starts
Downloaded from http://ahajournals.org by on November 26, 2019

0.08 sec after the rise. Three compotnents (a, b, and c) are visible at the right of the sternum
te-hile only components (a and c) are visible at the left. The first sotiund has a normal amplitude.
There is a close btut reversed type of splitting of the second sonntid (interval P-A = 0.04 sec).
There is loat frequtency fouirth souind in both sotud tracings.
a

The average Q-Ia interval was 90 msec. The Carotid Pulse

average Q-lb interval was 134 msec (average
Ia-lb interval, 44 msec). The average Q-Ic
interval was 171 msec (average Ia-Ic interval,
81 msec). The Ic component (ejection sound)
was abnormally large in two of the 20 pa-
tients.
Second Sound. The second sound had a
normal type of splitting in one patient and
was single in two of the 20 patients (in one
there was reversed splitting only in expiration) .
There was a reversed type of splitting in 17.
The average interval between the P and A
components of these patients was 53 msec.
Third Sound. This sound was present in
four of the 20 patients and was low pitched.
Fourth Sound. This sound was present in
13 of 20 patients. In two, it was prolonged
causing a short presystolic rumble.
Circulation, Volume XXXVI, Auguwt 1967
The interval Q-rise of the carotid pulse
averaged 120 msec in normal subjects, 114
msec in RBBB, and 167 msec in LBBB. The
interval Q-incisura of the carotid pulse av-
eraged 400 msec in normal subjects, 395 msec
in RBBB, and 435 msec in LBBB.
Auscultatory-Graphic Correlations
Splitting of the first heart sound was noted
on auscultation in two of 20 patients with
RBBB and in two of 20 patients with LBBB.
In all these patients, the first heart sound had
a large c component (ejection sound) in
graphic tracings. Four other patients might
have been considered as having a split first
sound by a casual observer; however, a cor-
rect statement of "loud fourth sound" was
 280 ORAVETZ ET AL.

:2R3.
2508 a !

-r
a be _.
A
I

50/66^1 II;^X1"rrW,l
Downloaded from http://ahajournals.org by on November 26, 2019

Figure 4
Case of left bundle-branch block. QRS = 0.14 sec. Phonocardiograms at second right (above)
and fifth left intercostal spaces (filter= 50). Following a fouzrth sound (IV), the first sound at
the apex shows the components (a, b, and c). The first sound (Ia) at the apex starts 90 msec
after the R rise. The Ia-Ib interval is 42 mnsec; the Ia-Ic interval is 80 m.sec. There is a wide
and reversed splitting of the second sound (interval P-A = 65 msec). The carotid rise starts
0.16 sec after the R iise.

made in these patients by our personnel prior
to recording phonocardiograms.
Discussion
The graphic stuidies on the first sound
from our laboratory'5,-18 have demonstrated
the existence of three components or groups
of vibrations within the normal first heart
sound. They demonstrated that the first two
(Ia, Ib) were of ventricular origin while the
third (Ic) was of aortic origin. They further
showed that they originated in the left heart
while the right heart contributioni to the
first sound was minimal and unrecognizable
in the normal phonocardiogram?1" These
findings were confirmed by van Bogaert and
a.ssociates;.
*In the normal heart, van Bogaert and associates21
recognized two groups of vibrations, a pre-ejec-
tional and an ejectional. The former resolves itself
to two components (Ia and lb) with high-frequency
galvanometers and filters. This explains the differ-
ence between the findings of the two laboratories.
Circuilation, Volume XXXVI, August 1967
 HEART SOUNDS
This being the case, several statements of
the past should be revised:
1. The first sound is split on auscultation
in bundle-branch block. This statement is
meaningless because the normal first sound
may be split (both on auscultation and in
phonocardiograms) due to separate audition
of the first (Ia) and second (Ib) compo-
nents. This fact has been widely recognized
since the study of Leatham22 on the first
heart sound.
2. The first sound is divided into two or
three components in the phonocardiogram
of bundle-branch block. This statement is
meaningless because either two or three com-
ponents are found in the normal first sound.
Moreover, auscultation may give the impres-
sion of a splitting of the first sound due to
two other possibilities:
a. There is a fourth sound; this plus the first
component of the first sound simulates a
splitting.
b. The third component (Ic) is louder than
normal on account of alterations of the aorta
(so-called aortic ejection sound); this, fol-
Downloaded from http://ahajournals.org by on November 26, 2019
lowing a larger component Ia (and a smaller,
inaudible component Ib), again may simu-
late splitting on auscultation. Both possibilities
were considered, and their occurrence was
demonstrated by van Bogaert and associates21
in their experimental and clinical study on
BBB.
The theoretical possibility of splitting of
the first sound as a result of BBB was based
on two assumptions: the existence of left
and right ventricular (or mitral and tricus-
pid) components within the first sound, and
the delay of activation and contraction of
one ventricle in BBB.
The first assumption has been challenged
by our studies, which showed that only the
left ventricle and aorta contributed to the
generation of the first heart sound.
The second assumption has been challenged
in LBBB (not in RBBB) by Braunwald
and associates,23' 24 following a study of the
timing of electric and mechanical events in
the normal heart.25 These authors pointed
out that lesions of the left ventricle can pro-
Circulation, Volume XXXVI, August 1967
281
duce the pattern of complete LBBB by
causing a conduction block in some of the
left ventricular branches without delay of
initiation of left ventricular contraction. For
this reason, we accepted only patients with
a typical electrocardiographic pattern of BBB
and excluded patients with congenital or
rheumatic heart disease, as well as those
with systemic or pulmonary hypertension.
This selection tried to avoid a partial "con-
duction block" with an ECG pattern of left
BBB. We also tried to avoid the occurrence
of a one-sided overload causing prolongation
of either the tension or the ejection of one
ventricle. As experimental interruption of the
left bundle branch causes a delayed con-
traction of the left ventricle,26 we have as-
sumed that this may also occur in man.
Our study of the arterial pulse showed
a normal (114 msec) Q-carotid rise interval*
in RBBB (no left ventricular delay) and a
prolonged interval (167 msec) in LBBB (left
ventricular delay). One might postulate that
this fact merely revealed a delay in left ven-
tricular ejection and not a delay in left ven-
tricular contraction. However, study of the
first sound proved to be revealing for an
understanding of the facts:
The beginning of the first sound was not
delayed over the Q wave in RBBB (55 msec
versus 59 msec for normal subjects). On the
contrary, this beginning was delayed over
Q in LBBB (91 msec versus 59 msec for
normal subjects).
In all cases, three components were ob-
served within the first sound, and no basic
difference was found in such respect between
right and left bundle-branch blocks.
The interval between the first (Ia) and the
second (lb) components (Ia minus Ib) was
only minimally longer in left versus right
BBB (44 versus 42 msec), and the difference
(2 msec) was insignificant.
The interval between the first (Ia) and
the third (Ic) components (Q-Ia minus Q-Ic)
*This interval is identical with that mentioned by
Braunwald and associates25 as representing the Q to
onset of ventricular ejection in normal man.
 282
was slightly longer in LBBB than in RBBB
(81 msec versus 78 msec), probably due to
myocardial fibrosis, but this difference was
minor.
The prolongation of the Q-Ic interval in
LBBB was similar to that of the Q-carotid
rise interval.
It was concluded, therefore, that, even
though the tension period is slightly longer
in LBBB than in either normal subjects or
patients with RBBB, the onset of left ventric-
ular contraction is delayed in LBBB.
The second sound showed the anticipated
changes, that is, wider splitting of normal
type in RBBB and reversed splitting in LBBB.
Reversed splitting was common in LBBB (17
of 20 patients) and the average interval be-
tween the P and A components in these pa-
tients (53 msec) was similar to that between
the A and P components of RBBB (57 msec).
Conclusions
The following conclusions were reached:
1. When pressure alterations, shunts, or
valvular defects causing ventricular hy-
Downloaded from http://ahajournals.org by on November 26, 2019
pertrophy or overload are excluded, ven-
tricular asynchronism is constant in
bundle-branch block.
2. The second heart sound reflects this
asynchronism in nearly all cases by re-
vealing a wide splitting in right bundle-
branch block and a reversed splitting in
left bundle-branch block.
3. The first sound, being related only to
events of the left heart and aorta, shows
a normal relationship to the Q wave of
the electrocardiogram in right bundle-
branch block, a marked delay in left
bundle-branch block.
4. The first sound, being related only to
events of the left heart and aorta, has
three components in both types of
bundle-branch block.
5. Occasional simulation of splitting of the
first sound on auscultation was explained
by incorrect evaluation of either a fourth
sound preceding the first sound or an
ejection sound. Therefore, "splitting" of
ORAVETZ ET AL.
the first sound is not a feature of bundle-
branch block.
References
1. KING, J. T.: Cliiiical recognition and physical
signs of bundle branch block. Amer Heart J
3: 505, 1928.
2. KING, J. T., AND MCEACHE:RN, D.: Nature of
the physical signs of bundle branch block.
Amer J Med Sci 183: 445, 1932.
3. Cossio, P., BRAUN-MENENDEZ, E., AND ORIAS,
O.: Quoted by Calo, 7 p. 246.
4. LAUBRY, C., AND PEZZI, C.: Quoted by Calo, 7 p.
246.
5. LEWIS, J. K.: Quoted by Calo, 7 p. 246.
6. LIAN, C., MINOT, G., AND WELTI, J. J.: Phono-
cardiographie: Auscultation Collective. Paris,
Masson et al., 1941.
7. CALO, A.: Les Bruits du Coeur et des Vaisseaux.
Paris, Masson et al., 1950.
8. SEGALL, H. N.: Cardiovascular sound in bundle-
branch block: Study of 244 cases. First Meet of
Laennec Society, Bal Harbour, October, 14,
1965. (Abstr.) Amer J Cardiol 19: 586, 1967.
9. Wolferth, C. C., and Margolies, A.: Asynchronism
in contraction of the ventricles in the so-
called common type of bundle-branch block:
Its bearing on the determination of the side
of the significant lesion and on the mechanism
of split first and second heart sounds. Amer
Heart J 10: 425, 1935.
10. BATTRO, A., BRAUN-MENENDEZ, E., AND ORIAS,
O.: Quoted by Calo, 7 p. 246.
11. CONTRO, S., AND LUISADA, A. A.: Modifications
of the heart sounds in bundle branch block.
J Mount Sinai Hosp NY 19: 70, 1952.
12. KELLY, J. J., JR.: Origin of the heart sounds.
Circulation 16: 273, 1957.
13. HABER, E., AND LEATHAM, A.: Splitting of
heart sounds from ventricular asynchrony in
bundle-branch block, ventricular ectopic beats,
and artificial pacing. Brit Heart J 27: 691
1965.
14. BRACHETTI, D., PAScuccI, E., AMBROSIONI, E.,
AND MAZZANTI, G.: II fonocardiogramma e la
meccanica cardica nei blocchi di branca com-
pleti. G Clin Med 46: 864, 1965.
15. Di BRARTOLO, G., NUNEZ-DEY, D., MUIESAN,
G., MACCANON, D. M., AND LUISADA, A. A.:
Hemodynamic correlates of the first heart
sound. Amer J Physiol 201: 888, 1961.
16. SHAH, P. M., MORI, M., MACCANON, D. M.,
AND LUISADA, A. A.: Hemodynamic correlates
of the various components of the first heart
sound. Circulation Research 12: 386, 1963.
17. SAKAMOTO, T., KUSUKAWA, R., MACCANON.
D. M., AND LUISADA, A. A.: Hemodynamic
determinants of the amplitude of the first
Circulation, Volume XXXVI, August 196 7
 HEART SOUNDS 283

leart 0so1ii(l. Circulation Research 16: 45,
1965.
18. LuJISADA, A. A., AND) MACCANON, 1D. M.: Physi-
ologic blasis of the lheart sounids. Dis Chest
49: 258, 1966.
19. LuISADA, A. A., KuRz, H., SLODKI, S. J., MAC-
CANON, D. M., AN)) KROL, B.: Normisal first
heart sounds with nonfunctional tricuispid
valve or right ventricle: Clinical and experi-
mental evidence. Circulation 35: 119, 1967.
20. VAN BOGAERT, A., ET AL.: Modifications du
premilier hruit du coeur dans le bloc de
branche: ltudle cliniiquie et expe6rinientalec
Arch Mal Coeur 56: 1253, 1963.
21. VAN' BOGAERT, A., ET AL.: Contribhution 'a 1'etude
du premiier bruit du coeuir normal. Arch Mal
Coeur 55: 368, 1962.
22. LEAThIH\\I, A.: Splittinig of the first an(1 second
heart sountdi(s. Lancet 2: 607, 19.34.
23. R311AUNWALD, E., DoN-oso, EL., SAPIN, S. O., ANlD
GR1ISHMIAN+, A.: Riglht bundle-branch block:
He-emodynamic, vectorc ardiographic and elec-
trocaLrdliographic observations. Circulation 13:
866, 1956.
24. BLIAUNWALI), E., AND MoIRRoW, A. C.: Sequience
of ventricular contraction in huLnan bundle-
branch hlock. Amier J Nled 23: 205, 1957.
25. BRAUN-WALD, E., ET AL.: Timiiing of electrical
and mechanical evenlts of the left side of the
humian heart. J Appl Physiol 8: 309, 1955.
26. BRAUN-MENEN-IEz, E., AND SOLARI, L. A.:
Ventrictular asynchron-ism in bundle branch
block. Arlh Intern Med 63: 830, 1939.

9ll

Recordings by Einthoven circa 1900

Downloaded from http://ahajournals.org by on November 26, 2019

(Couirtesy Dr. George Fahr)

t~~~~~~~~~~~~~~~~~~~~~~~....... -:-..- ::-:

-Jf- -#4
IA4AP
'U
C _1
Y'V3r1n' fT
y! 01,71.'
tH ~~T ...r i IT
Vs m Si- T-
"

Al in a

AI w> ts .1- P - . swl

744 &ttfsAs
411-11. .e'l"-.4,
iA I&A iA it

Circulation, VFolme XXXVI, Au,gust 1967