Innocent murmurs andthird heart sounds Black …heart.bmj.com/content/heartjnl/43/1/67.full.pdf · and 14 years of age had a systolic murmur. In Johannesburg, a survey on 200 White

Br Heart J 1980; 43: 67-73

Innocent murmurs and third heart soundsin Black schoolchildren*MARGARET J McLARENt, ANTHONY S LACHMANt, WENDY A POCOCK,JOHN B BARLOW

From the Cardiovascular Research Unit, Department of Medicine, University of the Witwatersrand; andthe Cardiac Clinic, General Hospital, Jtohannesburg, South Africa

SUMMARY Normal auscultatory findings were studied during a heart survey in which 12 050 Blackschoolchildren, aged 2 to 18 years, were examined by cardiologists. Physiological third heart soundswere detected in 96 per cent of children, innocent systolic murmurs in 72 per cent, and innocentmid-diastolic murmurs in 0-27 per cent. The term 'innocent systolic murmur' was used for vibratorysystolic murmurs (70%) and pulmonary ejection systolic murmurs (4.2%) but distinct separation ofthese two murmurs was often difficult. Vibratory systolic murmurs were present throughout the age range.

Important features in differentiating innocent systolic murmurs from those caused by mild organicheart disease included the intonation, site of maximal intensity, timing in systole, and behaviour withpostural change. Innocent mid-diastolic murmurs are short murmrrurs occurring immediately afterthe third heart sound in children, with no supportive evidence of organic heart disease.

Cardiology has made considerable progress overthe past half century and improvement in diagnosticand surgical techniques has led to an increasedunderstanding of many forms of heart disease andtheir management. Of equal importance is thedefinition of the auscultatory features of thehealthy heart and the recognition of normalvariations. In 1909, George Still' gave his classicaldescription of the vibratory systolic murmur and,nearly two decades later, Mackenzie2 still cam-paigned for the widespread medical acceptance ofinnocent systolic murmurs. It was not possible atthat time to differentiate these from abnormalmurmurs except by the exclusion of other evidenceof organic heart disease. Levine3 4 introduced aclassification of systolic murmurs based on intensityalone. This classification is excellent for recordpurposes but is of limited value in predicting theinnocent or pathological nature of a systolic mur-mur.5 Leatham6 classified systolic murmurs intoejection and regurgitant types on the basis ofanatomical and haemodynamic considerations andhe later clarified7 some of the auscultatory features

*Supported by the South African Medical Research Council, Anglo-American Corporation of South Africa Ltd, South Afrcan BreweriesInstitute, and the Stella and Paul Loewenstein Cardiac Fund of theUniversity of the Witwatersrand.tPresent address: Cardiology Unit, New Britain General Hospital,New Britain, Connecticut, USA.Received for publication 23 July 1979

of the normal heart. There have been subsequentcontributions,8-'0 and today the clinical recognitionof an innocent murmur usually presents littledifficulty to the skilled auscultator." This recogni-tion is of the utmost importance'2-'4 since innocentpraecordial murmurs occur most frequently inchildren and young adults and are commonlydetected when auscultation is performed as partof routine examinations at schools or during un-related illnesses. In such instances, young patientsand their parents need positive reassurance thatthere is no organic cardiac lesion in order to avoidincorrect management or even the development of acardiac neurosis.15-7

Several studies on the incidence of innocentsystolic murmurs have been reported.8 12 18 In oneseries,'8 96 per cent of healthy children between 3and 14 years of age had a systolic murmur. InJohannesburg, a survey on 200 White school-children,'2 aged 2 to 12 years, yielded a prevalencerate for innocent murmurs of 84 per cent. In May1972, a survey' was undertaken on Black school-children of the South Western Townships nearJohannesburg (Soweto) with the primary objectiveof determining the incidence of rheumatic heartdisease. The large number of children who hadto be examined provided the opportunity of makingother cardiological observations, including theincidence of non-ejection systolic clicks and as-

67

on 24 May 2018 by guest. P

rotected by copyright.http://heart.bm

j.com/

Br H

eart J: first published as 10.1136/hrt.43.1.67 on 1 January 1980. Dow

nloaded from

http://heart.bmj.com/

McLaren, Lachman, Pocock, Barlow

sociated mitral systolic murmurs20 as well ascongenital heart disease.2' We now present anddiscuss data on the incidence of innocent murmursand physiological third heart sounds.

Subjects and methods

A stratified random sample of 12 050 children fromcreches and schools of Soweto were examined duringthe period of May to October 1972. A detailedaccount of the sampling technique and methodshas been given previously19 but relevant featureswill be repeated. The ages of the children rangedfrom 2 to 18 years (Fig. 1), and the number of boysand girls was equal.

Fig. 1 Age distribution ofschoolchildren in the Sowetoheart survey. The age was notstated (NS) on the forms ofabout 1 per cent.

10

)u-cs

60

to08

-4'0

ca0~.)2q}

when it was heard in the absence of any abnormalcardiac findings7 14. Thus third heart soundsassociated with non-ejection clicks or late systolicmurmurs were excluded from the analysis.

Innocent systolic murmursThe two innocent systolic murmurs assessed werethe vibratory systolic murmur' and the pulmonaryejection systolic murmur. Though some authors'6 16have suggested that the term 'innocent systolicmurmur' be used for all unimportant murmurs, weexcludedmurmurs resulting from haemodynamicallyinsignificant congenital anomalies such as minuteventricular septal defects22. Such murmurs are notcompletely innocuous since the lesions which give

m Postured children

[I Li 5m L = LINS 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18

Aae (years)

EXAMINATIONThe medical team consisted of five cardiologistsand five physicians who had all received at least sixmonths' training in this cardiac unit. Threeobservers, at least one of whom was a seniorcardiologist, were present at each session. Auscul-tation was performed on all children in the supineand left lateral positions. A random 10 per cent(Fig. 1) of the sample was also examined in the erectand squatting postures. Children with normalhearts were examined by one observer only, butif an abnormal murmur was suspected or detected,the children were examined independently by allthree observers.'9 The finding of an 'innocent'mid-diastolic murmur had to be confirmed by atleast one other observer and an electrocardiogramwas also recorded. Examinations were carried outon three days a week over a five-month period. Theaverage rate of auscultation of children with normal

hearts was 25 an hour for each observer.

DIAGNOSTIC CRITERIA

Physiological third heart sounds

A third heart sound was regarded as physiological

rise to them may still act as a potential nidus forinfective endocarditis.10The diagnostic criteria for vibratory and pul-

monary ejection murmurs were those previouslydescribed.7 9 23 A vibratory systolic murmur is ashort crescendo-decrescendo murmur with acharacteristic low-frequency musical or buzzingintonation. It is commonly of maximal intensityjust inside the apex or at the left sternal border, butis frequently audible over the entire praecordium.It has a characteristic behaviour with posture,becoming softer or disappearing on standing and re-appearingon squatting.'3 24 The innocent pulmonaryejection systolic murmur is high-pitched and has ablowing quality. It is usually loudest at the secondleft interspace, but may also be heard over theaortic area, left sternal border, apex, and the neck,especially on the left side.'2 13 25 The second heartsound must be normally split and the componentsof normal intensity.7 9 26

'Innocent' mid-diastolic murmursWe have applied this term to the vibrations whichare heard as a short murmur immediately following

68



j.com/

Br H


nloaded from


Innocent murtmurs and third heart sounds

a third heart sound in some apparently normalyoung subjects (Fig. 2) with no supportive evidenceof organic heart disease.'4 27 28

ANALYSISThe variation in prevalence of these auscultatoryfindings with age, sex, and individual observerexperience was assessed. A x2 test for homogeneitywas used to assess differences. However, the large

STD

sample size necessary for the estimation of theprevalence of rheumatic heart disease renderedsignificance tests of little value in the analysis ofcommon findings such as third heart sounds andvibratory systolic murmurs. In many instances,small variations in prevalence of no clinical im-portance were shown to be highly significant(P < 0 0001). The behaviour of third heart soundsand vibratory systolic murmurs with posturalchange was noted.

(P < 0 0001), the range of 93 to 98 per cent is notclinically meaningful (Fig. 3). Excluding oneobserver (observer D, Fig. 4), who recorded aprevalence rate of 89 per cent, there was minimalvariation in the prevalence rates among the observers.Of the third heart sounds, 26 per cent occurred

as an isolated finding and 74 per cent were associatedwith an innocent murmur. In the routinely pos-tured group, 1107 (94.3%) of the 1174 children

Fig. 2 Phonocardiogramrecorded on medium frequency(MF) at the mitral area (MA)and let sternal border (LSB) inan apparently normal9-year-old child. Regular, low

:M .: frequency vibrations of avibratory systolic murmur

iX. i (VIB.SM), a third heartT 7T sound (3), and a soft short

mid-diastolic murmur (MDM)areX11 ~~shown.

had third heart sounds. All of these were presentin the supine position. In 86 per cent, the thirdheart sound disappeared on standing but was againheard on squatting, in 9 per cent it remained in allthree postures, and in 5 per cent it was present inthe supine position only.

INNOCENT SYSTOLIC MURMURSAn innocent systolic murmur was present in 8475children (72-1%) with normal hearts.

Results

Of the total of 12 050 children, 11 754 (97.5%)had normal hearts. Rheumatic heart disease'9 wasdiagnosed in 80 children (6X9 per 1000), mitralvalve prolapse20 in 168 (13X4 per 1000), and con-genital heart disease2' in 48 (3 9 per 1000). Theoverall prevalence of the auscultatory features inchildren with normal hearts was 96 per cent forthird heart sounds, 72 per cent for innocent systolicmurmurs, and 0X27 per cent for innocent mid-diastolic murmurs. In only 246 children (2-1 %)was neither a third heart sound nor an innocentmurmur detected.

PHYSIOLOGICAL THIRD HEART SOUNDSA physiological third heart sound was present in11 289 children (96 0%) and occurred with equalfrequency in both sexes. Though significantvariation in prevalence with age was shown

v-2 4 6 8 10 12 14 16 18

Age (years)

Fig. 3 Variation in prevalence of third heart sounds(3HS) and vibratory systolic murmurs (VIB.SM)with age.

100 -

C

'a 90

880-u

0)0 00.E0 70-C)

C)

a)" 60-

A 3 HS

* VIB. SM

69



j.com/

Br H


nloaded from



Vibratory systolic murmursVibratory systolic murmurs were detected in 70 3per cent (8264) of children with normal hearts andwere of equal prevalence in both sexes. Thoughthe relation between prevalence and age was againstatistically significant (P < 0.0001), in clinical termsthis was not impressive apart from a small peak in5- to 6-year-old children (Fig. 3). There was aclinically significant variation in the prevalenceamong the individual auscultators (Fig. 4), and thiscould not be explained by differences in the agesof the children examined by them. The meanprevalence for the more experienced observers(A, B, D, F, G, and I) was 74 per cent and withinthis group there was minimal variation (73 to 75%),whereas the mean prevalence in children examinedby the less-skilled auscultators (C, E, and H) andobserver J who examined only 65 children was 66per cent.

100 -

90 -

Z 80-

~0.

o 70-0

a,w, 60-C

0

GE 50 -0-

20 -

10-

0

0 . .

A I

* I

00

0

*

n I i-7-7A B C D E F G H

A A*_x

A

3 HS

0S \VIR SM

1757 2076 1024 711 2009rotal 1108 855 1100 909 65

Observer

Fig. 4 Individual observer (A-J) prevalence rates ofphysiological third heart sounds (3HS), vibratorysystolic murmurs (VIB.SM), and pulmonary ejectionsystolic murmurs (PESM) in Soweto schoolchildren.

In 99 per cent of children with vibratory systolicmurmurs, a physiological third heart sound was

also present. The most frequent site of maximalintensity of the murmur was the left sternal border

(62%), followed by both the left sternal border andapex (17%), and the apex alone (11%). The murmurwas loudest at the second left interspace in 1 percent only and in the remainder (9%) at variouscombinations of these three sites.Of the routinely postured children, 768 (65 4%)

had vibratory systolic murmurs. The prevalence ofthese murmurs in the supine, standing, andsquatting positions was 64-7, 15-9, and 57 per cent,respectively. Sixty-three per cent of the murmurswere present in the supine position, disappeared onstanding, and were again heard on squatting; 24 percent were heard in all three postures; 12 per centin the supine position only, and 1 per cent onsquatting only. In no instance was a vibratorysystolic murmur detected only in the standingposition. In the supine position, 95 per cent of themurmurs were of grade 2 intensity or less, 4 9 percent were of grade 3 intensity, and 0 1 per cent (11children) the murrnur was of grade 4 intensity;99 per cent of the murmurs still audible in thestanding position were then grade 2 or less andnone was grade 4. On squatting, 96 per cent weregrade 2 or less and the remainder were grade 3.

Pulmonary ejection systolic murmursInnocent pulmonary ejection systolic murmurswere recorded in 492 of the children with normalhearts (4 2%). They were more frequent in girlswith a ratio of 1-5:1 and the 15- to 18-year agegroup had the highest prevalence (Fig. 5). Therewas a significant variation in the prevalence ratesamong the individual observers and this was stillpresent when the less experienced observers wereexcluded from the analysis (P<0.001). The meanprevalence of the more experienced group was 3-64per cent, whereas the less experienced auscultatorsrecorded a higher rate of 5-28 per cent.

INNOCENT MID-DIASTOLIC MURMURSA short mid-diastolic murmur was heard in 32(0 27%) of the children who had no other evidenceof heart disease and were classified as 'normal'.Twenty-one of these children were boys. Theprevalence was 0-32 per cent in 2- to 6-year-oldchildren, 0-32 per cent in those aged 7 to 10 years,0-16 per cent in 11 to 14-year-olds, and 0-39 percent in the 15- to 18-year age group. The youngestchild with this murmur was 3 years of age. All butthree of the murmurs were detected by the moreexperienced auscultators. With only one exception,a vibratory systolic murmur was always an associatedfeature. However, none of them had an associatedpulmonary ejection systolic murmur. The diastolicmurmur was best heard at the apex in 25 subjects,at the left sternal border in six, and at both these

70

v IL.._.[V



j.com/

Br H


nloaded from


Innocent murmurs and third heart sounds

sites in one. The electrocardiograms were normalin all of the children with this murmur, there wereno obvious chest deformities, and the childrenappeared well.

12 -

10 -

8 -

6 -

4I-

2-

0 I I I I I I

2 4 6 8 10 12 14 16

Age (years)Fig. 5 Variation in prevalence of innocent pulmonaryejection systolic murmurs with age (P < 0 01).

DiscussionPhysiological third heart sounds and innocentsystolic murmurs are common auscultatory findingsin schoolchildren. In this survey, the vibratorysystolic murmur was the most frequent of theinnocent systolic murmurs. It was prevalentthroughout the 2- to 18-year age group and, incontrast to the findings of other workers,'0 11 29there was no significant fall in prevalance among

adolescents. The differentiation of vibratory andpulmonary ejection systolic murmurs is often ill-defined'0 1114 and, especially in an epidemiologicalstudy, is an academic exercise of little practicalimportance. In most instances, the innocentsystolic murmur had the quality of both murmurs

at different sites. Thus at the apex or lower leftsternal border, there was the typical 'vibratory'intonation whereas at the second or third left inter-

space it had more of the blowing character of apulmonary ejection murmur. This led to confusionin the interpretation of these murmurs as shownby the pronounced variation in the observers'prevalence rates. Some observers considered suchmurmurs to be vibratory only, whereas the in-terpretation of others was that both a vibratoryand a pulmonary ejection murmur were presentin the same child. We are aware that the classicalStill's murmur, which has typical uniform lowfrequency vibrations on a phonocardiogram (Fig. 2),has been shown by its response to a Valsalvamanoeuvre9 and by other techniques30 to have aleft-sided origin. Nevertheless, the use of the term'innocent systolic murmur' to include both pulmon-ary ejection and vibratory systolic murmurs inchildren is now preferred."With a background of a high prevalence of rheu-

matic heart disease in this community,'9 therecognition of the innocent nature of these murmursis of special importance. Both overdiagnosis andunderdiagnosis of rheumatic heart disease areharmful. Overdiagnosis may condemn some childrento a prolonged antibiotic regimen, unnecessarylimitation of activities, an overprotective parentalattitude, difficulty in obtaining employment, andheavily loaded insurance policies.'0 3' On the otherhand, underdiagnosis has the obvious danger ofearly rheumatic heart disease being missed and thusallowed to progress. Medical practitioners involvedin the primary care of children in family practice orcommunity health services should receive specialisedtraining in the recognition of these murmurs. Thissurvey suggested that even among observers trainingin auscultation, experience was of value in theirdetection. The features of most importance indifferentiating innocent systolic murmurs fromthose caused by organic heart disease were theirquality and characteristic behaviour with posture.The vibratory systolic murmur disappears orbecomes softer on standing and this behaviour maypartly explain why this murmur is less frequentlyheard in routine school examination programmes'2in which children are auscultated in the standing

L position only. A vibratory systolic murmur waspresent in only 15-9 per cent of Soweto children inthe standing position. The intensity of the murmurwas not meaningful since many murmurs (63%)resulting from rheumatic mitral regurgitation were

I less than grade 3,19 whereas some vibratory systolicmurmurs were loud. The length and timing of the

t systolic murmur are also of limited value. Thougha late or pansystolic murmur always indicates

t underlying organic heart disease,9 a murmur in mid-or early systole may be innocent or abnormal.5 9 18The site of maximal intensity provides an important

-c

0

a1)a.a1)0)ca1)0)

a1)C-

71



j.com/

Br H


nloaded from



additional clue to the origin of the murmur particu-larly in small muscular ventricular septal defects,11but site alone cannot always be used to differentiateinnocent from pathological murmurs. Though anychild suspected of having an abnormal systolicmurmur was examined independently by threeobservers and a diagnosis of organic heart diseaseaccepted only after agreement or discussion,20 21 wecannot exclude the possibility that a few subjectswhom we have classified as normal did, in fact,have minor cardiac pathology. It is difficult, forexample, on clinical examination alone, andprobably impossible during an epidemiologicalsurvey, to be certain of detecting all cases of smallatrial septal defect, anomalous pulmonary venousdrainage, or very mild mitral regurgitation reflectedby an early, sometimes intermittent, apical mur-mur.5 20A mid-diastolic murmur, which satisfied the

criteria for being innocent, was a rare auscultatoryfinding. This entity has not received generalrecognition in textbooks on cardiology. Referencehas been made to a ventricular filling murmurbeing mimicked by a left- and right-sided thirdsound occurring one after the other.'4 26 We supportthe observations of a few workers 27 28 that shortvibrations do occur in diastole in some apparentlynormal young individuals and may be heard as amurmur (Fig. 2). The origin of these vibrations isuncertain but they may be a result of rapid flowacross the mitral or tricuspid valves. It has alsobeen suggested28 that the vibrations arise duringdistension of the left ventricular wall, since theyfrequently start with a loud third heart sound. Wehave followed other children with this murmurand it has sometimes disappeared with time despiteno rheumatic fever prophylaxis. In at least onechild, however, a loud non-ejection systolic clicklater developed which suggested that there was, infact, a minimal abnormality of the mitral valve.Since a short mid-diastolic murmur may be theonly sign of minimal organic mitral stenosis, thediagnosis of an innocent mid-diastolic murmurshould only be made after a careful but non-invasive cardiac assessment to exclude this moreserious alternative. In addition, children with thismurmur should be followed to observe its naturalhistory.

We thank Drs K R Bloom, D M Bramwell-Jones,E Cohen, G E Gale, K Kanarek, and J B Lakierfor their assistance in the auscultation of thechildren, and Professor D M Hawkins and Mr BThomas (Department of Applied Mathematics,University of the Witwatersrand), and Mr RGarrison (Epidemiology Branch, National Heart,

Lung and Blood Institute, Bethesda, Maryland) forhelp with the statistical analyses.

References

'Still GF. Common disorders and diseases of childhood.1st ed. London: H. Frowde, 1909.2Mackenzie J. Diseases of the heart. London: OxfordUniversity Press, 1925.3Levine SA. The systolic murmur. Its clinical signi-ficance. JAMA 1933; 101: 436-8.4Levine SA, Harvey WP. Clinical auscultation of theheart. Philadelphia, London: W. B. Saunders, 1949.5Barlow JB, Pocock WA. The problem of non-ejectionsystolic clicks and associated mitral systolic murmurs:emphasis on the billowing mitral leaflet syndrome.Am HeartJ 1975; 90: 636-55.6Leatham A. A classification of systolic murmurs. BrHeart J7 1955; 17: 574.7Leatham A. Auscultation of the heart. Lancet 1958;2: 703-8.8Fogel DH. The innocent systolic murmur in children:a clinical study of its incidence and characteristics.Am Heart J 1960; 59: 844-55."Barlow JB, Pocock WA. The isolated systolic murmur.S Afr MedJ 1965; 39: 909-18.

'°Friedman S. Some thoughts about functional orinnocent murmurs. Clin Pediatr (Phila) 1973; 12:678-9.

"Tavel ME. The systolic murmur-innocent or guilty?Am J Cardiol 1977; 39: 757-9.

"Barlow JB, Pocock WA. The significance of aorticejection systolic murmurs. Am Heart J 1962; 64:149-58.

"Humphries JO, McKusick VA. The differentiation oforganic and 'innocent' systolic murmurs. Prog Cardio-vasc Dis 1962; 5: 152-71.

'4Leatham A. Auscultation of the heart and phono-cardiography. London: J. A. Churchill, 1970.

'Coleman EN, Doig WB. Diagnostic problems withinnocent murmurs in children. Lancet 1970; 2: 228-32.

"Levy AM. Innocent murmurs. Cardiovasc Clin 1972;4: No 3, 18-26.

"Br MedJ Editorial. Innocent praecordial murmurs inchildren. 1974; 1: 529-30.

'8Lessof M, Brigden W. Systolic murmurs in healthychildren and in children with rheumatic fever. Lancet1957; 2: 673-4.

'9McLaren MJ, Hawkins DM, Koornhof HJ, et al.Epidemiology of rheumatic heart disease in blackschoolchildren of Soweto, Johannesburg. Br Med J1975; 3: 474-8.

20McLaren MJ, Hawkins DM, Lachman AS, Lakier JB,Pocock WA, Barlow JB. Non ejection systolic clicksand mitral systolic murmurs in black schoolchildren ofSoweto, Johannesburg. Br Heart Y 1976; 38: 718-24.

2"McLaren MJ, Lachman AS, Barlow JB. Prevalence ofcongenital heart disease in black schoolchildren ofSoweto, Johannesburg. Br Heart J 1979; 41: 554-8.

22Van der Hauwaert L, Nadas AS. Auscultatory findingsin patients with a small ventricular septal defect.Circulation 1961; 23: 886-91.

72



j.com/

Br H


nloaded from


Innocent murmurs and third heart sounds

23Caceres CA, Perry LW. The innocent murmur. Aproblem in clinical practice. Boston: Little, Brown, 1967.

24Rhodes PH. Diagnosis of innocent heart murmurs inchildren. Gen Pract 1955; 12: no. 1, 69-71.

25Harris TN. Phonocardiographic study of pulmonic-systolic murmurs in children. Am Heart 7 1955; 50:805-15.

26Leatham A, Segal B, Shafter H. Auscultatory andphonocardiographic findings in healthy children withsystolic murmurs. Br Heart 1963; 25: 451-9.

27Liebman J, Sood S. Diastolic murmurs in apparentlynormal children. Circulation 1968; 38: 755-62.

28Luisada AA, Dayem MKA. Functional diastolicmurmurs. Am Heart 1972; 84: 265-72.

'9de Monchy C, Van der Hoeven GMA, Beneken JEW.Studies on innocent praecordial murmurs in children.III. Follow up study of children with an innocent

praecordial vibratory murmur. Br Heart 7 1973; 35:685-90.

30Wennevold A. The origin of the innocent 'vibratory'murmur studied with intra-cardiac phonocardiography.Acta Med Scand 1967; 181: 1-5.

3'Bergman AB, Stamm SJ. The morbidity of cardiacnondisease in schoolchildren. N EnglJ Med 1967; 276:1008-13.

32Stuckey D, Dowd B, Walsh H. Cardiac murmurs inschoolchildren. Med Aust 1957; 1: 36-8.

Requests for reprints to Professor J B Barlow,Cardiac Unit, Department of Medicine, Universityof the Witwatersrand, Hospital Street, Johannes-burg 2001, South Africa.

73



j.com/

Br H


nloaded from


Documents

Innocent murmurs andthird heart sounds Black …heart.bmj.com/content/heartjnl/43/1/67.full.pdf · and 14 years of age had a systolic murmur. In Johannesburg, a survey on 200 White