Sound exposures and hearing thresholds of symphony orchestra musicians Julia Doswell Royster EnvironmentalNoiseConsultants, Inc., P.O. Box30698,Raleigh,North Carolina27622-0698

LarryH. Royster DepartmentofMechanicaland•4erosœace Engineering, North CarolinaState University, Raleigh,North Carolina27695-7910 Mead C. Killion

EtymoticResearch,Elk GroveVillage,Illinois 60007

(Received28 March 1990;acceptedfor publication21 January1991)

To assess the risk of noise-induced hearinglossamongmusiciansin the ChicagoSymphony Orchestra,personaldosimeterssetto the 3-dB exchangerate were usedto obtain 68 noise

exposure measurements duringrehearsals andconcerts. Themusicians' Leqvaluesrangedfrom 79-99 dB A-weightedsoundpressurelevel [dB(A) ], with a meanof 89.9 dB(A). Basedon 15

h of on-the-job exposure perweek,thecorresponding 8-hdailyLeq(excluding off-the-job practiceand playing) rangedfrom 75-95 dB(A) with a meanof 85.5 dB(A). Mean hearing threshold levels (HTLs) for 59 musicianswere better than those for an unscreened

nonindustrialnoise-exposed population(NINEP), and only slightlyworsethan the 0.50 fractile data for the ISO 7029 (1984) screenedpresbycusispopulation.However, 52.5% of individualmusiciansshowednotchedaudiogramsconsistentwith noise-inducedhearing damage.Violinistsand violistsshowedsignificantlypoorerthresholdsat 3-6 kHz in the left ear than in the right ear, consistentwith the left ear'sgreaterexposurefrom their instruments. After HTLs were correctedfor ageand sex,HTLs were found to be significantlybetter for both earsof musiciansplayingbass,cello, harp, or piano and for the right earsof violinistsand violists than for their left ears or for both ears of other musicians. For 32 musicians for whom

bothHTLs andLeqwereobtained, HTLs at 3-6 kHz werefoundto becorrelated withtheLeq measured.

PACS numbers:43.50.Qp, 43.50.Yw, 43.66.Ed, 43.75.Cd

I. INTRODUCTION

II. RELEVANT

In response to musicians' concerns aboutthe possibility of developingnoise-induced hearingloss,soundexposure measurements and hearingthresholddata wereobtainedfor musiciansin the ChicagoSymphonyOrchestrato determine whether the playersface a significantrisk of developing noise-inducedpermanentthresholdshift (NIPTS) in their profession. Two approaches weretakento determinethe risk of hearingdamage.

A. Sound exposures

( 1) Soundexposuresfor musicianswere measuredand evaluatedin termsof theirpotentialto causeNIPTS, according to the modelpresentedin ISO 1999 (1990). ß (2) Measured hearing thresholdsfor musicianswere comparedto referenceage-effectpopulationsto determine whetherthe groupof musiciansshowedany more hearing lossthanexpectedfrom agingandnonoccupational hearing hazards.

LITERATURE

Both the instantaneous sound levels measured for a mu-

sicianand the resultingequivalentcontinuoussoundlevel

(Leq)forthemeasurement perioddependontheinstrument playedandthemusicbeingperformed,aswellastheposition of themusicianonthestagein relationto otherplayers.The microphonelocationis also important, especiallyin instanceswhereoneear receivesa highersoundexposureeither from the player'sown instrument(suchasthe left ear for a violinist)or fromotherinstruments. Literaturereports of soundlevelsandexposures of musicians havenot always described their measurement conditionsadequatelyto permit comparisonof resultsamongauthors. WestmoreandEversden( 1981) recordedA-weighted soundpressurelevels[dB(A)] usinga tripod-mounted

soundlevelmeterat various stagepositions duringdifferent

The studydid not attemptto addresstheissueof annoyanceamongmusiciansand/or players'difficultyin monitoringtheirowninstruments duringperiodsof highsoundlevelson stage.Neither did it attemptto addressthe occurrence of tinnitusor anyotherauditorydysfunctions exceptelevat-

Karlsson(1983),usingtripod-mounted microphones, measuredLeqvalues93.1dB(A) for heavyexposure positions and88.9dB(A) for lightexposure positions duringperfor-

ed pure tone thresholds.

mances. Axelsson andLindgren ( 1981 ) reported Leq'S mea-

2793

J. Acoust.Soc. Am. 89 (6), June 1991

symphonicworks;levelsexceeded 90 dB(A) 17.5% of the

playingtime,butno Leqvaluesweregiven.Jansson and

0001-4966/91/062793-11500.80

@ 1991 AcousticalSociety of America

2793

Redistribution subject to ASA license or copyright; see http://acousticalsociety.org/content/terms. Download to IP: 160.36.178.25 On: Fri, 19 Dec 2014 06:59:10

sured using tripod-mountedmicrophonesduring perfor-

mances. The Leq'Sfor concerts rangedfrom 83.3to 91.6 dB(A). Based on a workweek with 25 h of rehearsaland

performance, theestimated dailyequivalent Leq'Sfor musiciansrangedfrom 78.5 dB(A) to 88.4 dB(A). In a contract report,CampandHorstman(1987) described personaldosimetryresultson musiciansusingthe 5-dB exchangerate specifiedby OSHA; the averageLoshavaluesrangedfrom 82.4 dB(A) for bassplayersto 96.0 dB(A) for horn players duringrehearsalsand performances of the Gotterdammerung from Wagner'sRing Cycle, a work which containsexcessivelyloudpassages. (Note: In the currentstudytheLosha

isestimated to beabout2 dBlowerthantheLeq.) Woolford et al. (1988) summarizea dosimetrystudy of 30 musicians by Schacke(1987). Averagesoundlevelsfor brassand for woodwindplayersboth rangedfrom 87-96 dB (A), with the

exposure for a typicaldaycalculated asan 8-hLeqof 87.7 dB(A). The rangesof averagesoundlevelswere lower for violins and violas [86-93 dB(A) ] and for cellosand basses [81-87 dB(A)]. In summary, the available studieson orchestralsound exposuresindicate that at least some orchestramembers

havetypicaldaily8-hLeqvaluesexceeding 85dB(A) from their on-the-jobplaying time alone. Therefore, it would be expectedthat somemusiciansmay face a risk of hearing damagefrom on-the-jobnoise,dependingon their longterm noiseexposureand their own susceptibilityto noise-induced hearing loss. The predicted risk would increasefor some playersif the total noiseexposurewerecalculatedto include off-the-jobexposuressuchas solitarypractice,teachingstudent musicians,and moonlightingin other musicalgroups.

Axelssonand Lindgren (1981) studiedthe hearingof 139 musicians,many of whom had previousmilitary noise

exposure. Binaurallyaveragedhearingthresholds showeda notchedshapewith thepoorestthresholdat 6000Hz, butthe averagedegreeof losswassmall.However,theseaverages might havebeenmuch higherif thresholdshad not been measureddown to -- 15 dB. The greatesthearing losses were found in musiciansplaying bassoon,horn, trumpet, andtrombone.Stringplayersshowedbetteraveragehearing thanexpectedfor their agein the right ear,but left earHTLs werepoorer.Karlssonet al. ( 1983) measuredHTLs (down to a minimum of + 10 dB) for 392 musicians. Median and

25thpercentiledatawereequivalentto Spoor's(1967) referencevalues,indicatingno signof NIPTS for the overallpopulation.However,mediandata did showa 6-kHz dip in the audiometricconfiguration. Janssonet al. (1986) attempted to reconcilethe differencesin findingsbetweenthe two precedingSwedishstudies.Throughreanalysis of the data,the researchers agreedthat althoughmedianHTLs of musicians werecloseto ISO 7029 (1984) presbycusis referencedata, a higherpercentof individualmusiciansdid showpoorHTLs (worsethan the 0.05 fractile of the referencedata) than in a non-noise-exposed population. Ostri et al. (1989) measuredthe HTLs of 96 musicians. Half the males and 13 % of the females showed notched au-

diograms.The 75th percentile,median,and 25th percentile HTL datafor musicianswerepoorerby up to 10-15 dB than matchingpercentilesfor ISO 7029 presbycusis reference dataat 3-8 kHz. The medianaudiogramfor malesshoweda maximum dip at 6 kHz. Violinists showedsignificantly poorerhearingin the left ear than in the right. To summarize the available studies of musicians' hear-

B. Hearing levels

Another way of assessing the risk of NIPTS amongmusiciansisto evaluatethehearingthresholdlevels(HTLs) for orchestramembers.If musiciansas a group do not show morehearinglossthanexpectedfor theirage,thentheriskof NIPTS for their profession may be assumedto be insignificant.

Westmore and Eversden (1981 ) obtainedaudiograms for 34 orchestralmusiciansand reportedthat 34% of the earsshowedaudiometricpatternsconsistentwith noise-induced hearing loss (not defined,but assumedto mean a notchedshape);however,the degreeof losswasslight.Rabinowitz et al. (1982) reportedaudiometricresultsfor 110 orchestralmusicians.Only 52 individualshad any HTLs exceeding20 dB; basedon historyinformation,the authors concludedthat the hearinglosscouldbe attributedsolelyto music in 22 musicians,while it was impossibleto evaluate the proportionof music-relatedlossin the other 30 cases. Johnsonet al. ( 1985, 1986) measuredhearingthresholdsof 60 orchestramembers;they reportedthe averagedifferences betweenthe measuredthresholdsand the expectedage-effectthresholdsfrom Spoor's(1967) data aslessthan 10 dB, but theyalludedto a notchshapewith poorerHTLs at 6 kHz than 8 kHz. In a pilot studyof the hearingthresholdsof 13 orchestral musicians, Woolford et al. (1988) found HTLs no worse than 25 dB in half the ears, while 11 ears showed sensorineural loss and 2 ears showed conductive loss.

2794

J.Acoust. Soc.Am.,Vol.89,No.6, June1991

ing, all investigators havefound hearinglossin a pattern consistentwith noise-inducedhearing damage in some players,but the majorityof musiciansshowhearinglevels within the normalrange.This consistency suggests that the degreeof hearinghazardislow,but that susceptible individualsmaybe at risk.Resultsshowgreaterhearinglossin the left earthan in the right earfor violinists,indicatingthat the apparentNIPTS isrelatedto the greaterexposure of the left ear.

III. METHOD

A. Subjects

All membersof the ChicagoSymphonyOrchestrawere encouraged to participatein thestudyby voluntarilysigning up to takean audiogramand/or to weara dosimeterto measuresoundexposureduringa rehearsalor performance.Audiogramsweregivenpriorto theorchestra's playingin order to avoid contaminationof HTLs by temporarythreshold shift.Of the approximately100musiciansin the orchestra, 59 voluntarily receivedaudiograms. A total of 68 dosimetrysampleswere obtainedon 44 musicians.Data were collectedduring two separateweeks,

severalmonthsapart,in orderto increasethe varietyof musicalselectionsbeingrehearsedand performed.Two dosimeterswereplacedin differentpositionson the stageduring eachmeasurementperiodas a checkon the personaldosiRoyster etal.:Sound exposures andhearing thresholds

2794

Redistribution subject to ASA license or copyright; see http://acousticalsociety.org/content/terms. Download to IP: 160.36.178.25 On: Fri, 19 Dec 2014 06:59:10

metry readings,but theseresultswere not includedin the analysissincethey were not actual exposuremeasurements for individualplayers. B. Instrumentation

Portable audiometers were used with Etymotic ResearchER-3A insertearphonesin order to allow testingin availablequiet roomsin OrchestraHall. The calibrationof eachaudiometerwascheckedwith a specificpair of ER-3A insertearphonesaccordingto the interim calibrationdata in ANSI

S3.6-1989

to obtain exact correction

factors with re-

spectto soundpressurelevelsfor audiometriczero at each frequencyin each channel. During the first week of data

microphonewell toward the front of the right collar in order to minimize the shadowingof the microphoneby the musician'shead and body. Two musiciansparticipatedin exploratorytestingof the differencein level betweena microphonepositionnear the left ear versusnear the right ear of a violinistor violist. The measured differencewas 6-8 dB(A) with the musician hold-

ing his headcloseto erect.If the player leanshis left ear into the violin or viola more strongly,a greaterexposuredifference between

ears occurs.

IV. RESULTS

A. Sound exposure measurements

collection, the audiometers used were Medical Dimensions

Summary statistics forthemeasurement-period Leqval-

Otomatic units. During the secondweek, one Beltone 9D

ues are presentedin the first column of Table I, and a fre-

audiometer

and one Monitor

MI

5000 unit were used. In all

caseshearing thresholdswere measuredmanually in 5-dB stepsdown to 0 dB hearinglevel usingthe Hughson-Westlake ascending-descending procedure. Biological calibration checks of audiometer

function were carried out before

each day of use. The roomsusedfor audiometrictestingwereonesmall-

grouppracticeroom and oneoffice.The backgroundsound pressurelevelsweremonitoredusinga Rion NA61 precision sound level meter with a type NX-01A octaveband filter unit. Soundlevelsin the practiceroom were within allowed specifications re: ANSI S3.1-1977at 1 kHz and higher,but exceededtheselevelsby lessthan 10 dB at 0.5 kHz. Levelsin the officeexceededthe allowedspecifications by 14 dB at 0.5 kHz and 5 dB at 1 kHz. However, the ER-3A insert ear-

phonesprovideover 32 dB of meanattenuationof ambient soundat each octaveband from 125-8000 Hz (Berger and Killion, 1989), thereby reducingthe effectivelevelsat the ear to within ANSI S3.1-1977 specifications.As an indication that backgroundlevelswere acceptablefor testingto

audiometric zero,ttiesubjects usedforbiologicalcalibration checkscould detect tonesat 0 dB hearing level at all test frequencies from 0.5 to 8 kHz in bothtestingrooms,ascould many of the studysubjects. Widely acceptedpracticesfor soundsurveysand individual exposuremonitoringwere followed (Royster et al. 1986). Larson Davis 700 integratingsoundlevel meter/do-

simeters wereusedto measure themusicians' Leqvalues(3dB exchangerate). The computercapabilityof theseinstrumentsallowsa completetime historyof soundlevelsduring the measurementperiodto be recordedand statisticallyanalyzed. The dosimeterwas supportedaround the musician's waist by a belt and was worn on the sideor back of the hip. The microphoneof eachdosimeterwasclippedontothe musician's collar.

When

the musician

indicated

chestra,the lowestLeqvalueswereobtainedduringthese samples. Someorchestramembersexpressedconcernthat even

thoughoverallLeqexposures mightnot be excessive, the presenceof very high soundpressurelevelsfor brief periods of time mightstill harmtheir hearing.WhereasOSHA setsa limit of 140 dB peak soundpressurelevel for exposureto impactor impulsenoises,thereis very little evidenceto support this limit accordingto Ward (1986). The safe peak soundpressurelevelincreases asthe durationof the impulse or impact decreases.Summary statisticsfor the maximum A-weighted peak SPLs recorded during dosimetry are shown in the second column of Table I. In 82% of the sam-

plesthe maximumpeaklevelwas 130 dB(A) or below,and only two samples (3%) included peaks exceeding140

that one ear

would receivemore noiseexposurethan the other, the microphonewas placed on the side expectedto receivethe greaterexposure.However, in the caseof violinistsand violists,it wasnecessary to placethe microphoneon the left side of the back of the collar in order not to interfere with the left-

shoulderplaying positionof the instrument.For someviolinistsor violistswho preferredthat the microphonebe kept awayfrom the left side,it wasplacedon the right front collar instead.In thesecases,an attempt was made to place the 2795

quency histogram of theLeqdistribution isshownasFig. 1. The sampledurationswereusually2 to 3 hs.The 41 readings made during the first week of data collectionrangedfrom 79-99 dB(A) [mean = 89.3 dB(A)], and the 27 readings made during the secondweek ranged from 83-97 dB(A) [mean = 90.5 dB(A) ]. Although the meansand the maximum values for the two data collection periods are very close,the lower limit for the earlier period is 4 dB lessthan for the later period,probablydue to the particularworksof music being played and rehearsed.During the first period the programincludedone strings-onlywork, Haydn's Symphony No. 85, "La Reine," plus two full-orchestraworks, Britten's Simple Symphonyand Shostakovich'sSymphony No. 5, Opus47. During the secondperiodthe programconsistedof Ran's Concerto for Orchestra and Berlioz' SymphonieFantastique(Opus 14). As a few of the dosimetry readingsfrom the first data collectionperiod were from rehearsalswhich includedonly the string sectionsof the or-

J. Acoust.Soc. Am., Vol. 89, No. 6, June 1991

TABLE I. Summarystatistics for soundexposure measurements, dB(A). Max.

Mean s.d. Median Minimum Maximum

Leq

Peak

89.8 4.7 90.0 79.0 99.0

124.9 6.4 124.0 112.0 143.5

rms 106.4 5.0 106.8 95.5 115.5

Royster eta/.: Sound exposuresand hearingthresholds

8-h

Leq 85.5 4.7 85.7 74.7 94.7

2795

Redistribution subject to ASA license or copyright; see http://acousticalsociety.org/content/terms. Download to IP: 160.36.178.25 On: Fri, 19 Dec 2014 06:59:10

and percussion(group2) and for thoseinstrumentssituated in front of them (group 3) tend to be in the upperportionof

the overallrangeof 79 to 99 dB(A). The Leqvaluesfor violinsandviolas(group 1) arespreadthroughoutthe entire range,while valuesfor group4 fall in the lower half of the range.

C. Daily equivalent 8-h exposures

TheLeqvaluesobtained forsampling periods hadto be

z

convertedto equivalentdaily 8-h.exposuresin order to use the ISO 1999 (1990) modelto evaluatethe potentialNIPTS from orchestralmusic. The maximum on-the-jobrehearsal and performancetime for the ChicagoSymphonyOrchestra musiciansis 15 hoursper week. Therefore,each measured

Leqvaluewasconverted to a dailyequivalent exposure by assumingthat the measuredvaluewasrepeated15 hoursper 40-h work-week.The resultingeffectis to reduceeachmea-

suredLeqvalueby4.3dB;summary statistics fortheequivalentdaily8-hLeq'SareshownasthefourthcolumnofTable I.

I-EQs

D. Hearing threshold levels

DB(•)

FIG. 1. Frequency histogram of Leqvalues measured for dosimetry sam-

plingperio•ds.

PresentedasTable II are the meanbinaurally averaged HTLs and standarddeviationsfor all subjectsand for varioussubgroups,as well as the numberof subjectsper group and their mean age. E. Hearing thresholds by gender group

dB(A). It is possiblethat thesehighestpeaksmay haveoccurred when the dosimetermicrophonewas accidentally bumped.The greatmajority of peakSPLswere in the range from 115-129 dB(A). (Note that although A-weighted peak valueswere reported,the valueswould be expectedto increaseonly 1-2 dB if unweightedmeasurements had been made.) If thesepeaklevelsare representative, the chanceof acoustictrauma occurringduring orchestralmusicis quite small.

Perhapsa moremeaningfulindicatorof thesoundintensity duringvery loud musicalpassages is the maximumrms sound level measuredduring the dosimetry sample. Summary statisticsfor the maximum rms values recorded are shown in the third column of Table I. The maximum

rms

As expected, female musiciansshow better average HTLs than malesespeciallyat 3-8 Hz. However,both males and femalesshowgoodaverageHTLs for their groupmean age.In Fig. 3 the meanHTLs for maleand femalemusicians are plotted togetherwith the expectedHTLs for two agematched and sex-matched reference populations: the screenedISO 7029 (1984) referencedata representingpres-. bycusisalone, and the unscreenednonindustrial noise-exposedpopulation (NINEP) assumedto representtypical hearing for citizensof the U.S.A. without industrial noise exposurebut includingthe influencesof other hearinghazards suchas military serviceand other non-occupational noiseexposure,plusmedicalpathology(RoysterandThomas, 1979). The musiciansas a group show better average

equivalentsoundlevelswere 110dB(A) or belowfor 76% of the samples,and ehevery highestvaluerecordedwas 115.5

hearing thantheunscr•ened NINEP,andtheirhearing is

dB(A).

nateeffects fromanyhearing haz'ard except aging.

B. Exposures for different instrument sections

F. Audiogram patterns

To assess whethersoundexposuresdifferedfor various

only slightly poorerthan that of subjectsscreenedto elimi-

The data plottedin Fig. 3 suggestthat there is a contri-

instrument groups, themeasured Leqvalueswereseparated bution from NIPTS in the HTLs of the musicians,as evident into the.followinggroups: group 1: violin and viola (23 samples); group 2: horn, trumpet and trombone( 13 samples); group 3: clarinet, flute, bassoon,and percussion(17 sampies); and

group4: bass,cello,harp, and piano ( 15 samples).

Theresulting measured Leqdistributions foreachgroup arepresented asFig.2. NotethattheLeqvaluesforthebrass 2796

J. Acoust. Soc.Am.,Vol.89,No.6, June1991

from the shapeof their averageaudiograms,with relatively greaterhearinglossat 2-4 kHz than at 6-8 kHz. All audiograms were examinedto tabulate the percentageof ears showingpatternsof hearinglosswhich are classicallyassociatedwith NIPTS: (a) a dip or notchpatternwith HTLs at 3, 4, and/or 6 kHz being10 dB or moreworsethan adjacent lower and higherfrequencies,or (b) a dip or notchof 10 dB or more superimposed on a slopinghigh-frequency-emphasisloss.Of the total of 118 ears,53 (44.9%) showeda dip Royster etaL:Soundexposures andhearing thresholds

2796

Redistribution subject to ASA license or copyright; see http://acousticalsociety.org/content/terms. Download to IP: 160.36.178.25 On: Fri, 19 Dec 2014 06:59:10

1,8

GROUP Xi

GROUP

•

:

8 : :

•

6

It

4

Z

..•............. i............. ;............. i............. i............. •............ .:.-

'" .......................... (............. ' ............. ' ............. ' ............ -:--

• 4

2

2

....

I

'76

88

84

88

LEQ•

92

96

1ee

,

,

,

'78

I

,

,

.

88

I

,

,

,

84

DB (I:I)

-

!

.

,

.

88

L.EQ•

I

.

.

,

92

i

.

.

.

98

I

188

DB (R)

FIG. 2. Frequency histogramsof sam-

pling-period Leqvalues measured forfour instrumentgroups.

GROUP

3

GROUP 4• : :

8

Z IU O'

It

4

-=4

2

2

I

'76

,

,

.

I

88

,

,

,

I

,

,

84

L.EQ•

,

I

88

i

,

i

i

92

,

i

i

i

96

i

i

i

i

zee

DB(R)

!

?8

,

,

,

I

88

,

,

,

I

,

,

84

LEQ•

and 9 (7.6%) showeda dip superimposed on a slopingloss, for a total of 52.5% earsshowingsignsof NIPTS. In termsof individualmusiciansrather than ears, 42 persons(71%) showedaudiometricpatternsconsistentwith NIPTS in one

,

I

,

,

,

88

I

92

,

,

,

I

98

,

,

,

I

188

DB(R)

H. Hearing thresholds by instrument group

As shown in Fig. 5, the mean HTLs for instrument groupsappear grosslysimilar to each other, although the or both ears. meanagesof the groupsdifferby up to about8 yr. In orderto evaluate HTL differenceswhile accounting for age, the G. Hearing thresholds by age group HTLs of each musicianwere age-correctedby subtracting Shownin Fig. 4 arethemeanbinaurallyaveragedHTLs the median yaluesfrom the ISO 7029 (1984) presbycusis for musiciansin four agebrackets(30-39 years,40-49 years, data for the matching age and sex. The mean binaurally 50-59 years,and 60-69 years) in comparisonto the screened averaged age-correctedHTLs of musiciansby instrument ISO 7029 (1984) and unscreened NINEP reference data for groupare presentedasTable III, alongwith the total group the matchingageandthe matchingproportionof malesand and the sameother subgroupsfor which data without age females.[The two musiciansunder age 30 were dropped correctionswere shownin Table II. The age-correctedrefromthisanalysis. ] In generaleachagegroup'smeanHTLs sultsby instrumentgroupare plottedasFig. 6. Mean HTLs fall betweenthe two referencepopulations,with slight to are similarfor the instrumentgroupsat 0.5, 1, 2, and 8 kHz, mild notchedshapeswhichare consistent with the presence but group 4 (playing the quieter instruments--bass,cello, of somenoise-induced hearingloss.The notchedshapeis piano,and harp) showsthe besthearingat the more noisemost pronouncedfor the 40-49-year-oldage group.Howsusceptible frequencies of 3, 4, and 6 kHz. Resultsfor group ever, even in this case the musicians' mean HTLs are about 1 (violin and viola) fall about halfway betweenthosefor the sameasthe unscreened NINEP meanHTLs at 3-4 kHz group4 and thosefor groups2 and 3. in spiteof the suggestion of someappar. ent NIPTS, and muTo examinethe possibilitythat interauraldifferencesfor sicians'HTLs are similarto or betterthan the presbycusis violinistsand violistsmight be obscuredby averagingthe median HTLs at 6-8 kHz. HTLs of the two ears,the age-correctedHTLs of each ear 2797

J. Acoust.Soc. Am., Vol. 89, No. 6, June 1991

Royster eta/.: Soundexposuresand hearingthresholds

2797

Redistribution subject to ASA license or copyright; see http://acousticalsociety.org/content/terms. Download to IP: 160.36.178.25 On: Fri, 19 Dec 2014 06:59:10

TABLE II. MeanbinaurallyaveragedHTLs for musicians andstandarddeviations, dB. Hearing thresholdlevel, dB Audiometrictestfrequency,kHz

Mean age,

Group

N

years

Total

59

52.4

Statistic mean

0.5

1

2

3

4

6

8

11.0

8.6 9.2

13.9 11.4

21.0 15.1

25.8 18.2

24.7 16.6

24.4 20.1

9.5 10.0

14.3 11.1

23.2 15.4

29.0 18.6

27.1 17.2

25.5 20.8

5.6 4.5

12.5 12.6

13.5 11.5

14.6 11.9

16.0 11.0

20.4 17.9

6.7 6.1

8.1 7.5

9.8 8.5

12.3 10.1

12.9 9.5

11.8 13.6

3.5 3.6

9.2 7.0

23.1 19.4

24.2 18.9

17.7 12.8

8.5 9.2

10.0 8.0

16.4 13.3

21.4 13.3

24.0 12.9

25.6 13.2

26.2 12.3

12.3 10.6

19.6 11.0

28.8 13.6

42.1 17.8

39.1 17.1

40.2 20.3

12.5 16.7

18.8 15.2

25.0 13.3

29.2 14.7

33.3 12.9

46.7 15.9

s.d.

Males

46

53.2

9.3

mean

12.1

s.d.

Females

13

Ages 30-39

Ages 40-49 Ages 50-59 Ages 60-69

49.5

12 35.3

12 46.3 13 55.1 14 64.6

9.8

mean

7.1

s.d.

6.1

mean

7.7

s.d.

6.1

mean

6.3

s.d.

4.5

mean

13.5

s.d.

12.5

mean

12.3

s.d.

Ages 70 and up Group 1: violin

6 27

72.8 54.9

8.2

mean

19.2

s.d.

12.1

mean

11.0

s.d.

9.2

10.1 11.2

15.3 12.1

22.4 15.0

27.7 16.7

26.4 16.0

25.9 20.5

7 46.7

mean s.d.

7.5 7.1

5.0 6.1

10.4 8.1

17.9 14.5

21.8 16.9

21.8 13.8

17.1 16.2

Group 3: clarinet, flute, bassoon,percussion

12 48.3

mean ' s.d.

11.5 12.7

7.7 8.0

11.7 13.5

22.9 18.7

26.5 22.1

25.0 19.7

21.7 17.7

Group 4: bass, cello, piano, harp

13 53.9

mean s.d.

11.0 6.9

6.5 4.5

12.9 7.7

15.0 10.3

20.2 17.4

19.2 15.3

24.2 22.5

and viola

Group 2: horn and trumpet

were plotted by instrumentgroup, as shownin Fig. 7. Althoughthe left earsof group1showsimilarHTLs to group2 and 3, their right earsare similar to thoseof group4. A least-squares analysisof variancewas performedon HTL asa functionof instrumentgroup,ear, and frequency category(the noise-susceptible frequencies3, 4, and 6 kHz versus0.5, 1,2, and 8 kHz. The least-squares adjustedmeans (adjustedfor the unequalgroupsizes)at 3, 4, and 6 kHz in the left ear weresignificantlybetterfor group4 than for the othergroups,whichwereequivalent.In the right ear,groups

! ..: .......................

1 and 4 both showsignificantlybetter mean HTLs than groups2 and 3.

I. Hearingthresholdsby left and right ear

Theresultsindicatingpoorerhearingin theleftearthan in therightfor group1canprobablybeattributedto a differencein soundexposurebetweenears.The left earof a violin-

istor violistreceives greaterexposure thantherightbecause it facesthesounding boardand (for someplayers)istilted verycloseto theinstrument, whiletherightearissomewhat

: .............

:

: : ............

: .............

:

:

:..............

: ..

:

"i.............i.............::...........................i"•::.......•' ............""

FIG. 3. Mean hearingthresholds of male and femalemusicianscomparedto data

fromtworeference populations matched by sexandage:the screened ISO 7029 (1984) presbycusis population,and the unscreened

whitenonindustrial noise-exposed popula-

+ Zso '7eae o. •USZC-taNS

tion (NINEP).

ßx. I•IHTTE:: N1'NBP

ß6

.1.

2

4

FREQUENCYs

2798

J. Acoust. Soc.Am.,Vol.89,No.6, June1991

8

KHZ

Royster eta/.:Soundexposures andhearing thresholds

2798

Redistribution subject to ASA license or copyright; see http://acousticalsociety.org/content/terms. Download to IP: 160.36.178.25 On: Fri, 19 Dec 2014 06:59:10

ß

:

..:

.................... ,•"•'•'i';'•i ............ !'"".: .....::•...............