EXPOSURE TO EXCESSIVE SOUNDS AND HEARING STATUS IN ACADEMIC CLASSICAL MUSIC STUDENTS

Interntionl Journl of Occuptionl Medicine nd Environmentl Helth 2017;30(1):55 75 https://doi.org/10.13075/ijomeh.1896.00709 EXPOSURE TO EXCESSIVE SOUNDS AND HEARING STATUS IN ACADEMIC CLASSICAL MUSIC STUDENTS MAŁGORZATA PAWLACZYK-ŁUSZCZYŃSKA, MAŁGORZATA ZAMOJSKA-DANISZEWSKA, ADAM DUDAREWICZ, nd KAMIL ZABOROWSKI Nofer Institute of Occuptionl Medicine, Łódź, Polnd Deprtment of Physicl Hzrds Abstrct Objectives: The im of this study ws to ssess hering of music students in reltion to their exposure to excessive sounds. Mteril nd Methods: Stndrd pure-tone udiometry (PTA) ws performed in 168 music students, ged 22.5±2.5 yers. The control group included 67 subjects, non-music students nd non-musicins, ged 22.8±3.3 yers. Dt on the study subjects musicl experience, instruments in use, time of weekly prctice nd dditionl risk fctors for noise-induced hering loss (NIHL) were identified by mens of questionnire survey. Sound pressure levels produced by vrious groups of instruments during solo nd group plying were lso mesured nd nlyzed. The music students udiometric hering threshold levels (HTLs) were compred with the theoreticl predictions clculted ccording to the Interntionl Orgniztion for Stndrdiztion stndrd ISO 1999:2013. Results: It ws estimted tht the music students were exposed for 27.1±14.3 h/week to sounds t the A-weighted equivlent-continuous sound pressure level of 89.9±6.0 db. There were no significnt differences in HTLs between the music students nd the control group in the frequency rnge of 4000 8000 Hz. Furthermore, in ech group HTLs in the frequency rnge 1000 8000 Hz did not exceed 20 db HL in 83% of the exmined ers. Nevertheless, high frequency notched udiogrms typicl of the noise-induced hering loss were found in 13.4% nd 9% of the musicins nd non-musicins, respectively. The odds rtio (OR) of notching in the music students incresed significntly long with higher sound pressure levels (OR = 1.07, 95% confidence intervl (CI): 1.014 1.13, p < 0.05). The students HTLs were worse (higher) thn those of highly screened non-noise-exposed popultion. Moreover, their hering loss ws less severe thn tht expected from sound exposure for frequencies of 3000 Hz nd 4000 Hz, nd it ws more severe in the cse of frequency of 6000 Hz. Conclusions: The results confirm the need for further studies nd development of hering conservtion progrm for music students. Int J Occup Med Environ Helth 2017;30(1):55 75 Key words: Noise-induced hering loss, Music students, Exposure to excessive sounds, Pure-tone udiometry, Hering threshold levels, High-frequency notches INTRODUCTION Noise-induced hering loss (NIHL) is the second most prevlent sensorineurl hering loss, preceded only by presbycusis. The potentil risk of hering loss in musicins hs been extensively investigted since 1960s. The mjority of studies hve focused on professionl musicins, people working in music venues nd generl public listening to loud music for long time. It hs been shown tht plyers, especilly professionl orchestrl musicins, cn develop NIHL nd suffer from other Funding: the study ws supported by the Ministry of Science nd Higher Eduction of Polnd grnt IMP 18.2/2012 2013 Assessment of exposure to excessive sounds nd hering sttus in students enrolled in cdemic music eduction. Grnt mnger: Młgorzt Pwlczyk-Łuszczyńsk, Ph.D. with hbilittion, NIOM professor. Received: June 12, 2015. Accepted: Jnury 4, 2016. Corresponding uthor: M. Pwlczyk-Łuszczyńsk, Nofer Institute of Occuptionl Medicine, Deprtment of Physicl Hzrds, św. Teresy 8, 91-348 Łódź, Polnd (e-mil: Mlgorzt.Pwlczyk@imp.lodz.pl). Nofer Institute of Occuptionl Medicine, Łódź, Polnd 55

M. PAWLACZYK-ŁUSZCZYŃSKA ET AL. hering symptoms such s tinnitus or hypercusis, which cn influence their work bilities more severely thn the hering loss itself. However, becuse of insufficient udiometric evidence of hering loss cused purely by music exposure, there is still disgreement nd specultion concerning the risk of hering loss from music exposure lone [1 10]. According to the literture dt, musicins, in prticulr clssicl orchestrl musicins, re often exposed to sounds t levels exceeding the upper exposure ction vlues from the Noise Directive 2003/10/EC [11] s well s the Polish mximum dmissible intensity (MAI) vlues [12]. College music students re potentil future employees of orchestrs, so they constitute group t higher risk of hering loss. However, dt on noise exposure conditions nd hering sttus in college music students re limited. Wht is more, music students re not covered by the sme regultory frmework s employees. For exmple, Phillips nd Mce [13] hve mesured sound pressure levels (SPLs) mong students of University School of Music in USA nd found tht singers nd brss, wind nd string plyers during individul music clsses were exposed to sounds t verged A-weighted SPLs of 87 95 db. Such levels re comprble to those mesured in professionl orchestrl musicins [14 17]. Furthermore, Fern [18] hs estimted tht students spent on verge 10 35 h/week on plying musicl instruments nd dditionlly performed in orchestrs for 2 3 h 56 times/yer. For comprison, Pwlczyk-Łuszczyńsk et l. [10] hve found tht professionl orchestrl musicins were usully exposed to music for 7 70 h/week (verge 30 h/week) due to both on-the-job nd off-the-job plying. Therefore, it is not surprising tht Phillips et l. [19], when nlyzing the prevlence of hering impirment in student musicins (ged 18 25 yers), hve found typicl NIHL notches (occurring minly t 6000 Hz) in 45% of 329 students. On the other hnd, in Brlow s study [20], 44% of 50 young people studying populr music showed evidence of udiometric notch t 4000 6000 Hz, nd 16% were clssified under the UK Occuptionl Helth nd Sfety guidelines s exhibiting mild hering loss. The prevlence of notched udiogrms ws considerbly higher thn the one reported by the studies on the generl popultion but ws round the sme level or lower thn tht reported from the studies of trditionl music courses nd conservtoires. Furthermore, students exposed to music t high sound pressure levels were found to hve moderte temporry threshold shifts, which correlted with their history of personl exposure, most significntly t 4000 Hz. Moreover, trnsient-evoked otocoustic emission (TEOAE) responses in those young musicins showed overll lower mplitudes compred to the students with less exposure to music [21]. The overll objective of this study ws to nlyze hering sttus of undergrdute music students in reltion to their exposure to sounds during university eduction. In prticulr, it hs been ttempted to: evlute the totl sound exposure of music students, including their vrious ctivities ssocited with plying instruments, nd on tht bsis, to determine the expected noise-induced permnent threshold shifts, ccording to the model described in the Interntionl Orgniztion for Stndrdiztion stndrd ISO 1999:2013 [22]; compre udiometric hering thresholds in the music students to the ge-relted reference dt from the otologiclly norml popultion ( highly screened ) in ccordnce with ISO 7029:2000 [23], tht is dtbse A from ISO 1999:2013, s well s to the unscreened, gedrelted norml control group comprising young people who were non-music students nd non-musicins; nlyze the ssocition between music exposure nd some hering symptoms, including the presence of high frequency notched udiogrms nd tinnitus. MATERIAL AND METHODS Exposure to excessive sounds nd udiometric hering threshold levels were determined in university music students. Dt on the students musicl experience, 56 IJOMEH 2017;30(1)

HEARING STATUS IN CLASSICAL MUSIC STUDENTS ORIGINAL PAPER instruments in use, time of weekly prctice nd dditionl risk fctors for noise-induced hering loss were identified by mens of questionnire survey. In ddition, their hering bility ws ssessed using the (modified) Amsterdm Inventory for Auditory Disbility nd Hndicp ((m)aiadh) [24]. Audiometric hering threshold levels (HTLs) were compred with the theoreticl predictions clculted ccording to ISO 1999:2013 [22]. The study comprised 168 undergrdute clssicl music students, ged 19.3 31.7 yers from 2 universities, i.e., Acdemy of Music in Łódź nd Acdemy of Music in Gdńsk, Polnd. The comprison group consisted of 67 subjects, ged 18 31 yers, who were non-music students nd non-musicins nd who were not exposed to noise t work. In both groups, individuls with middle er pthology were excluded from the study. The subjects received finncil compenstion for their prticiption in the study. They were recruited by dvertisement. The study protocol ws pproved by the Ethic Committee of the Nofer Institute of Occuptionl Medicine, Łódź, Polnd (decision No. 8/2013) nd ll the subjects gve their written informed consent for prticiption in the exmintions. Questionnire survey All the music students filled in questionnire developed to enble identifiction of risk fctors for NIHL. In prticulr, the questionnire included items relted to: music exposure (i.e., yers in school (plying instruments), instruments plyed, hours of individul nd group prctice per dy/week currently/in the pst, ensemble prticiption), helth sttus nd medicl history (pst middle-er diseses, nd surgery, etc.), physicl fetures (body weight, height, skin pigmen- ttion), lifestyle (smoking, noisy hobbies, listening to personl medi plyer, ttending disco/brs, rock concerts, etc.), use of individul hering protectors, self-ssessment of hering sttus. The subjects from the control group were lso interviewed using similr questionnire, but without questions on music exposure. In ddition, ll the subjects completed (modified) Amsterdm Inventory for Auditory Disbility nd Hndicp. This questionnire consists of 30 questions, including 2 control questions not included in the ssessment. The questions re divided into 5 prts (subscles) seprtely ssessing: bility of discrimintion (differentition) of sounds (subscle I), uditory locliztion (subscle II), understnding speech in noise (subscle III), intelligibility in quiet (subscle IV), detection of sounds (subscle V). The respondents reported how often they were ble to her effectively in the situtions specified bove. The 4 nswer ctegories were s follows: lmost never, occsionlly, frequently nd lmost lwys. Responses to ech question were coded on scle from 0 to 3; the higher the score, the smller the perceived hering difficulties. The totl score per subject ws obtined by dding the scores for 28 questions. Mximum totl score for the questionnire ws 84. Additionlly, the nswers for ech subscle were summed up (mximum score for subscle I ws 24, while for other subscles the totl ws 15) [24]. Hering exmintions The stndrd pure-tone udiometry (PTA) ws performed, using the Audio Trveller Audiometer type 222 (Intercoustics, USA) with TDH 39 hedphones. Hering threshold levels (HTLs) for ir conduction t frequencies 250 8000 Hz were determined using the scending descending technique in 5-dB steps. In ddition, trnsient-evoked otocoustic emissions (TEOAE) nd IJOMEH 2017;30(1) 57

M. PAWLACZYK-ŁUSZCZYŃSKA ET AL. distortion-product otocoustic emissions (DPOAE) were determined. However, results of the ltter tests will be described elsewhere. Before the hering tests, otoscopy ws performed in order to screen for conditions tht would exclude n exmined subject from the study. The hering exmintions were crried out in sound-proof room or in quiet rooms locted in universities buildings where the A-weighted equivlent-continuous sound pressure level of bckground noise did not exceed 35 db. where: L Aeq,T1 the energy verge of collected smples of the A-weighted equivlent-continuous SPL produced by respective instrument during collective plying, in db, L Aeq,T2 the energy verge of collected smples of the A-weighted equivlent-continuous SPL produced by respective instrument during individul plying, in db, T 1, T 2 the declred time of collective nd individul prctice per week, in hours, T o the reference durtion equlled 40 h. Music exposure evlution To evlute the students exposure to music (excessive sounds), sound pressure level (SPL) ws mesured during individul nd collective rehersls s well s during lessons with techer nd concerts. The surveys included vrious instruments nd diverse repertoire. The mesurements were performed ccording to the Polish stndrds PN-N-01307:1994 nd PN- EN ISO 9612:2011 (ISO 9612:2009) [25 27] using personl sound exposure meters (i.e., personl logging dosimeters type 4436 nd 4443 (Brüel & Kjær, Denmrk)) worn by the students plying instruments. The distnce between the microphone nd the ers (0.1 0.4 m) ws s short s prcticlly possible (without disturbing the plyers). Ech single mesurement period usully corresponded to the durtion of rehersl, lesson or concert. In generl, results of 294 mesurement smples (covering in totl pprox. 231 h) were collected. For ech study subject, the A-weighted weekly noise exposure level (L EX,w ) ws clculted from the vlues of the A-weighted equivlent-continuous sound pressure levels produced by the respective instrument (e.g., violin or trumpet) nd time of weekly prctice obtined from the questionnire, using the following eqution (1): 1 0.1 L Aeq,T1 0.1 L Aeq,T2 L EX,w 10 log T1 10 T2 10 (1) T o Additionlly, the students exposures were described by noise immission level (L IM ), i.e., mesure of the cumultive sound (noise) energy to which n individul ws exposed over time given by the formul (2): L IM = L EX,w +10 log T (2) where: L EX,w A-weighted weekly noise exposure level, in db, T time of exposure to excessive sounds (plying instrument) fter ge of 18 yers, in yers. Prediction of noise-induced hering loss The music students udiometric hering threshold levels were compred with the theoreticl predictions clculted in ccordnce with ISO 1999:2013 [22]. The foresid stndrd specifies the method for determining sttisticl distribution of hering threshold levels in dult popultions fter exposure to noise bsed on 4 prmeters: ge, gender, noise exposure level (i.e., A-weighted equivlent continuous SPL normlized over 8-h working dy or 40-h working week) nd time (durtion) of noise exposure (in yers). However, to compre predictions obtined for the students of different gender, ge, time nd exposure, the soclled stndrdized hering threshold levels (SHTL) were determined using the following formuls (3) nd (4): 58 IJOMEH 2017;30(1)

HEARING STATUS IN CLASSICAL MUSIC STUDENTS ORIGINAL PAPER SHTL = 1.282 (HTL PHTL Q50 )/ (PHTL Q10 PHTL Q50 ) for HTL PHTL Q50 (3) SHTL = 1.282 (HTL PHTL Q50 )/ (PHTL Q90 PHTL Q50 ) for HTL < PHTL Q50 (4) where: HTL the ctul hering threshold level, in db HL, PHTL Q50/Q10/Q90 medin, 10th nd 90th percentiles of predicted distribution of hering threshold level l, in db HL. These clcultions were pplied to the students udiogrms twice, i.e., their HTLs were compred to the HTLs of the highly screened (otologiclly norml) non-noiseexposed popultion (dtbse A from ISO 1999:2013) nd n equivlent (ccording to ge, gender, L EX,w nd T) noise-exposed popultion. Stndrdized hering threshold levels were lso determined for the control group. However, in the ltter cse, udiogrms were only relted to the dt from otologiclly norml persons not occuptionlly exposed to noise. Sttisticl nlysis Differences in vribles verges between subgroups of the study subjects (i.e., the music students nd control group s well s between the stringed, wind nd percussion instrument plyers) were nlyzed using the t-test for independent dt or the Mnn-Whitney U test, where pplicble. The reltions between HTLs nd sound exposure prmeters were evluted using the Spermn s rnk correltion coefficient. The stndrdized hering threshold levels were nlyzed using the one-smple t-test. Answers to the questionnire nd frequency of some outcomes (e.g., prevlence of the high-frequency notched udiogrms) were presented s proportions with 95% confidence intervls (95% CI). Differences in proportions between vrious pirs of the subgroups (e.g., the music students nd control group or vrious instruments plyers) were nlyzed using the Fisher s exct test. The binry logistic regression with Wld test ws used for the nlysis of ssocition between the prevlence of hering-relted outcomes (e.g., high-frequency notched udiogrms) nd sound exposure level, ge, gender, time of plying instrument nd other fctors. Sttisticl nlysis ws crried out with n ssumed level of significnce p = 0.05. However, when compring pirs of vrious subgroups of the subjects, to void the risk of mss significnce, p-vlue divided by the number (N) of possible comprisons (p = 0.05/N) ws set s limit for sttisticl significnce. The sttisticl nlysis employed Sttistic (version 9.1.) (SttSoft, Inc., USA) softwre pckge. RESULTS The study groups chrcteristics nd questionnire dt Among 168 music students, 60.1% plyed stringed instruments, 33.3% plyed wind instruments nd 6.5% plyed percussion (Figure 1). Their men (M) ge ws 22.8 yers (with stndrd devition (SD) of 2.5 yers, medin (Me) of 22.4 yers). Regrding gender, 48.8% (82) of them were femle students nd 51.2% (86) were mle students. Time of musicl prctice vried from 6 to 20 yers (M±SD = 12.5±3.2 yers, Me = 13 yers). Severl students plyed more thn one musicl instruments. pino 38% hrp 2% cello 15% double bss 6% guitr 3% violin 30% viol 6% percussion 6.5% (N = 11) stringed instruments 60.1% (N = 101) wind instruments 33.3% (N = 56) tub 4% Fig. 1. Students plying vrious instruments horn 9% orgn 5% trombone 7% ccordin 13% trumpet 18% bssoon 2% flute 12% oboe 12% clrinet 5% sxophone 13% IJOMEH 2017;30(1) 59

M. PAWLACZYK-ŁUSZCZYŃSKA ET AL. Mjority (85.4%) of the 67 subjects from the control group were non-music students. About one-qurter (25.4%) of them were occsionlly exposed to noise t their workplce or during internship or pprenticeship. In the control group, the men ge ws lso 22.8 yers (SD = 3.3 yers, Me = 22.3 yers). In terms of gender, women were little more numerous thn men (56.2% vs. 44.8%). However, in generl, there were no significnt differences concerning ge nd gender between the music students nd controls. Similr reltionships were observed when nlyzing medicl history, physicl fetures nd some spects of lifestyle, such s: smoking hbits nd noisy hobbies (shooting, pintbll, motor sports, etc.) (Tbles 1 nd 2). However, greter frctions of the music students compred to the control group reported listening to personl medi ply- Tble 1. Prevlence of noisy hobbies in the music students nd in the control group Noisy hobby music students Respondents [% (95% CI)] control group Listening to PMPs 86.9 (80.9 91.2) 77.6 (66.10 86.00) Using PMPs t lest 1 h dily 62.5 (55.0 69.5) 40.3 (29.40 52.30) every dy 57.1 (49.6 64.4) 35.8 (25.40 47.80) Frequent ttending clubs, pubs, etc. never 7.1 (4.1 12.2) 6.0 (2.00 14.90) nowdys 12.5 (8.3 18.5) 4.5 (1.10 13.00) in the pst 11.9 (7.8 17.8) 1.5 (0.01 8.80) Frequent ttendnce t loud music concerts, etc. never 14.3 (9.8 20.5) 20.9 (12.80 32.30) nowdys 4.8 (2.3 9.3) 1.5 (0.01 8.80) in the pst 4.8 (2.3 9.3) 1.5 (0.01 8.80) Prcticing noisy motor sports (quds) never 80.4 (73.6 85.7) 76.1 (64.50 84.80) nowdys 13.1 (8.8 19.1) 16.4 (9.30 27.30) in the pst 6.0 (3.2 10.8) 6.0 (2.00 14.90) Shooting never 81.5 (74.9 86.7) 70.1 (58.30 79.80) nowdys 14.3 (9.8 20.5) 22.4 (14.00 33.90) in the pst 4.2 (1.9 8.6) 6.0 (2.00 14.90) Using noisy tools never 44.6 (37.3 52.2) 43.3 (32.10 55.20) nowdys 26.8 (20.7 34.0) 11.9 (6.00 22.20) in the pst 28.0 (21.7 35.2) 41.8 (30.80 53.70) PMPs personl medi plyers; CI confidence intervl. Significnt differences (p < 0.05). 60 IJOMEH 2017;30(1)

HEARING STATUS IN CLASSICAL MUSIC STUDENTS ORIGINAL PAPER Tble 2. Prevlence of risk fctors other thn noise exposure for the noise-induced hering loss (NIHL) in the music students nd in the control group Smoking Risk fctor music students Respondents [% (95% CI)] control group nowdys 35.7 (28.90 43.20) 32.8 (22.8 44.8) in the pst 38.1 (31.10 45.60) 25.4 (16.5 37.1) Elevted blood pressure 3.6 (1.50 7.80) 4.5 (1.1 13.0) Dibetes 0.6 (0.01 3.60) 0.0 (0.0 6.7) Rynud s phenomenon 3.0 (1.10 7.00) 3.0 (0.3 11.0) Light skin pigmenttion 28.0 (21.70 35.20) 23.9 (15.2 35.5) Ototoxic ntibiotic tretment 5.4 (2.70 10.10) 7.5 (2.9 16.8) Body mss index (BMI) > 25 20.2 (14.90 27.00) 26.9 (17.7 38.6) CI confidence intervl. ers (PMPs) (for t lest n hour) every dy (57.1% vs. 35.8%, p < 0.05) nd frequent ttending music clubs in the pst (11.9% vs. 1.5%, p < 0.05). On the other hnd, subjects in the control group declred frequent use of noisy tools in the pst more often thn the music students (41.8% vs. 28%, p < 0.05), while the opposite reltion ws observed nowdys (11.9% vs. 26.8%, p < 0.05) (Tble 1). It is worth noting tht only 8.9% (95% CI: 5.4 14.3%) of the music students declred using hering protective devices (minly erplugs) t present or in the pst, while 47% (95% CI: 39.6 54.6%) plyers intended to use them in the future. Hering protective devices were minly worn during rehersls, including group rehersls (27.9%) nd solo rehersls t school (18.6%) nd t home (16.3%). Regrding the prevlence of other NIHL risk fctors such s: smoking, elevted blood pressure, dibetes, white-finger syndrome, light skin pigmenttion, ototoxic ntibiotic tretments nd overweight (body mss index (BMI) > 25), there were no significnt differences between the study subjects nd the comprtive group (Tble 2). Assessment of exposure to music Tble 3 summrizes sound pressure levels mesured in the students plying different instruments during solo nd group rehersls, lessons nd concerts. According to the collected dt, the music students were exposed to sounds t: the equivlent continuous A-weighted SPL (L Aeq,T ) of 80 98 db (10 90th percentile), mximum A-weighted SPL (L Amx ) of 94 113 db, pek C-weighted SPL (L Cpek ) of 115 137 db. There ws considerble diversity in sound exposure mong the music students plying vrious instruments, prtly due to the vribility in the repertoire, kind of lessons nd plce of testing, etc. The highest L Aeq,T levels were mesured mong the percussion, brss instruments (sxophone, trumpet, trombone, tube, horn) nd wood-wind (bssoon, flute, oboe, clrinet) plyers (Figure 2). According to the responses included in the questionnire, the music students plyed instruments on verge for 27.1±14.3 h/week, including 16.4±8.2 h of solo prcticing nd 7.1±6.1 h of group plying. IJOMEH 2017;30(1) 61

M. PAWLACZYK-ŁUSZCZYŃSKA ET AL. Tble 3. Sound pressure levels (SPLs) mesured in the music students during solo nd group prcticing, lessons with techer, concerts, etc. Instrument Sound pressure level [db] A-weighted equivlent-continuous A-weighted mximum C-weighted pek M±SD (L) 10 50 90th percentile M±SD 10 50 90th percentile M±SD 10 50 90th percentile Violin (N = 37) 85.2±3.9 (86.7) 79.9 86.0 91.0 98.0±3.4 94.2 98.0 102.5 120.2±14.6 103.8 124.7 132.1 Viol (N = 16) 85.0±2.9 (85.9) 81.0 84.4 89.4 98.4±3.2 94.1 98.7 102.7 125.1±8.1 16.1 126.1 133.3 Cello (N = 19) 80.7±3.7 (82.5) 76.7 79.9 87.2 96.2±4.8 90.1 96.6 104.0 125.3±8.5 110.1 124.8 137.2 Double bss (N = 15) 80.9±5.5 (84.7) 74.7 80.4 88.9 95.8±5.7 89.5 96.2 99.4 123.4±7.4 110.5 124.5 132.6 Clrinet (N = 15) 90.6±2.6 (91.5) 87.9 90.1 93.2 104.0±3.6 98.8 103.9 109.5 122.6±5.5 115.5 123.1 127.5 Oboe (N = 11) 86.5±1.8 (86.8) 84.8 86.5 88.1 100.2±3.2 97.4 98.9 103.3 127.0±9.5 116.3 128.7 135.9 Bssoon (N = 8) 91.6±4.4 (92.9) 83.5 93.6 95.4 100.2±3.3 94.9 101.0 105.2 116.2±2.5 114.0 115.3 121.2 Flute (N = 15) 91.2±5.4 (94.1) 82.3 91.6 97.6 104.4±5.4 95.6 105.6 110.1 122.2±7.0 114.7 121.8 131.1 Horn (N = 8) 93.9±2.8 (94.5) 87.7 95.0 96.1 110.6±8.4 106.0 107.7 131.3 123.9±3.2 120.2 123.6 129.9 Trumpet (N = 24) 96.6±3.2 (97.7) 91.5 96.8 99.4 112.3±7.7 104.4 110.7 124.1 127.3±9.4 117.0 126.0 138.7 Trombone (N = 18) 95.3±4.4 (97.1) 88.2 95.3 101.6 104.9±24.7 98.3 108.4 125.0 129.0±8.4 120.1 127.2 145.1 Tub (N = 12) 93.6±2.1 (94.0) 92.1 93.5 94.9 109.2±4.3 105.2 109.7 114.0 132.2±8.4 124.2 129.7 145.1 Sxophone (N = 19) 97.5±2.7 (98.3) 94.2 97.5 100.5 112.8±9.0 104.9 110.3 130.7 126.4±7.7 117.8 124.8 138.3 Percussion (N = 28) 96.0±7.7 (105.2) 87.7 94.3 109.1 108.4±12.2 101.3 109.8 119.5 134.8±6.2 126.8 134.8 143.8 Hrp (N = 4) 85.5±0.6 (85.5) 84.8 85.5 86.1 99.8±3.0 95.5 100.8 102.2 123.7±4.7 119.8 122.3 130.3 Pino (N = 25) 84.4±4.7 (88.2) 80.4 83.3 89.7 97.6±6.5 89.4 96.5 105.3 125.7±8.9 115.4 125.4 140.6 Orgn (N = 1) 83.5±0.0 (83.5) 83.5 83.5 83.5 94.5±0.0 94.5 94.5 94.5 106.7±0.0 106.7 106.7 106.7 Accordion (N = 10) 84.1±5.7 (86.9) 76.7 83.2 91.6 104.3±10.2 96.1 102.6 119.8 127.0±10.4 114.9 126.0 142.3 Guitr (N = 9) 77.5±8.9 (89.2) 70.0 73.4 98.5 92.9±7.0 86.5 90.1 110.3 122.3±7.0 108.7 122.3 133.4 Totl (N = 294) 89.1±7.5 (97.0) 79.9 89.1 98.1 103.0±10.6 93.9 102.6 112.8 125.6±9.8 115.2 126.0 137.0 M men; SD stndrd devition.; L energy verge of the N smples of mesured A-weighted equivlent-continuous SPL. 62 IJOMEH 2017;30(1)

HEARING STATUS IN CLASSICAL MUSIC STUDENTS ORIGINAL PAPER L [db] Aeq,T 115 110 collective plying (Me) solo plying (Me) 105 100 95 90 85 80 75 70 65 violin viol cello double bss clrinet oboe bssoon flute horn trumpet trombone tub sxophone percussion pino ccordin guitr totl Instrument L Aeq,T A-weighted equivlent-continuous sound pressure level; Me medin. Whiskers rnge 10 90th percentile. Fig. 2. A-weighted equivlent-continuous sound pressure levels mesured in the music students plying vrious instruments during solo nd collective prctice For ech subject, weekly noise exposure level (L EX,w ) ws determined from the equivlent-continuous A- weighted SPLs produced by the respective instrument nd the declred time of weekly prctice (eqution 1). Since some students plyed mny instruments, such evlutions were bsed on the dt concerning the min instrument. The weekly noise exposure levels clculted from those dt rnged between 75 106 db (M±SD = 86.8±6.3 db, Me = 84.9 db) (Figure 3), while the corresponding noise immission levels vried from 85 db to 114 db (M±SD = 94.4±6.2 db, Me = 92.6 db). There were significnt differences in sound exposure between the students plying instruments belonging to different groups. The highest vlues of the L EX,w levels were observed in Cses [n] 180 160 140 120 100 80 60 40 20 0 0 (0%) 15 (9%) 86 (51%) 129 (77%) 166 (99%)168 (100%) 157 (93%) 146 (87%) 70 75 80 85 90 95 100 105 110 L EX,w [db] L EX,w A-weighted weekly noise exposure level. Fig. 3. Cumultive distribution of the A-weighted weekly noise exposure level (L EX,w ) in the music students IJOMEH 2017;30(1) 63

M. PAWLACZYK-ŁUSZCZYŃSKA ET AL. L [db] EX,w 105 100 95 90 85, b, c b, c the students plying percussion, while the lowest in those plying stringed instruments (Figure 4). Nerly hlf (48.8%) of the students were exposed to excessive sounds t levels exceeding the Polish mximum dmissible intensity (MAI) vlue for occuptionl noise (L EX,w = 85 db) [12], while 29.2% of them were exposed to the L EX,w levels bove 87 db (Figure 3), i.e., exposure limit vlue specified by the 2003/10/EC Directive [11]. 80 L EX,w A-weighted weekly noise exposure level. Significnt differences between the stringed nd wind instrument plyers (p < 0.05/3). stringed wind percussion Instrument b Significnt differences between the stringed nd percussion instrument plyers (p < 0.05/3). c Significnt differences between the wind nd percussion instrument plyers (p < 0.05/3). Dt re given s men vlues with 95% confidence intervls. Fig. 4. A-weighted weekly noise exposure levels (L EX,w ) in the music students plying stringed, wind nd percussion instruments Self-ssessment of hering cpbility Generlly, lmost ll the music students (92.9%) nd individuls in the control group (98.5%) ssessed their hering s good. Nevertheless, some of them complined of vrious hering-relted symptoms (Tble 4). In prticulr, greter proportions of the music students, compred to the control group, reported hering impirment (31.5% vs. 14.9%, p < 0.05) nd complined of difficulty in speech intelligibility in noisy environment (45.8% vs. 29.9%, p < 0.05), constnt or temporry tinnitus (11.3% vs. 4.5%, p > 0.05) nd hypercusis (36.3% vs. 11.9%, p < 0.05). Tble 4. Prevlence of self-reported hering-relted symptoms in the music students nd in the control group Symptom music students Respondents [% (95% CI)] control group Good hering 92.9 (87.80 95.90) 98.5 (91.10 100.50) Hering impirment self-reported 31.5 (25.00 38.90) 14.9 (8.20 25.60) noticed by fmily 8.9 (5.40 14.30) 10.4 (4.90 20.40) in the right er 3.6 (1.50 7.80) 1.5 (0.01 8.80) in the left er 9.5 (5.90 15.00) 4.5 (1.10 13.00) in both ers 16.1 (11.30 22.50) 9.0 (3.90 18.60) sudden 4.8 (2.30 9.30) 0.0 (0.00 6.70) incresing from yer to yer 3.0 (1.10 7.00) 3.0 (0.30 11.00) incresing from month to month 11.3 (7.30 17.10) 6.0 (2.00 14.90) incresing in different mnner 10.7 (6.80 16.40) 3.0 (0.30 11.00) Hering loss in the fmily 44.0 (36.80 51.60) 44.8 (33.50 56.60) 64 IJOMEH 2017;30(1)

HEARING STATUS IN CLASSICAL MUSIC STUDENTS ORIGINAL PAPER Tble 4. Prevlence of self-reported hering-relted symptoms in the music students nd in the control group cont. Symptom Difficulties with hering/understnding music students Respondents [% (95% CI)] control group whisper 9.5 (5.90 15.00) 19.4 (11.60 30.60) norml speech 2.4 (0.70 6.20) 1.5 (0.01 8.80) speech in noisy environment 45.8 (38.50 53.40) 29.9 (20.20 41.70) trebles 0.6 (0.01 3.60) 0.0 (0.00 6.70) Use of higher rdio nd TV volume settings 13.7 (9.30 19.80) 7.5 (2.90 16.80) Tinnitus 11.3 (7.30 17.10) 4.5 (1.10 13.00) constnt tinnitus 2.4 (0.70 6.20) 3.0 (0.30 11.00) trnsient (periodic) tinnitus 8.9 (5.40 14.30) 4.5 (1.10 13.00) Hypercusis 36.3 (29.40 43.80) 11.9 (6.00 22.20) CI confidence intervl. Significnt differences (p < 0.05). The music students nd non-music student controls exmined using the (m)aiadh obtined the men totl score pproximtely 88±9% of the mximum vlue (84), which suggests no substntil hering problems (Tble 5). Moreover, significnt differences neither in the totl score nor in the scores in vrious subscles were noted between the foresid groups. However, comprison of the (m)aiadh outcomes in vrious groups of plyers reveled tht reltively low (< 70% of mximum vlue) scores were most frequent in the percussion plyers (Tble 6). In prticulr, significntly higher proportions of the students plying percussion thn those plying stringed or wind instruments scored reltively low in subscle III (evluting intelligibility in Tble 5. Hering bility in terms of score in the (modified) Amsterdm Inventory for Auditory Disbility nd Hndicp ((m)aiadh) in the music students nd in the control group (m)aiadh score music students Respondents control group M±SD 10 50 90th percentile M±SD 10 50 90th percentile Totl 73.7±7.2 63 75 81 73.8±7.3 63 75.5 81 Subscle I Distinction of sounds 21.9±2.0 19 22 24 21.9±2.1 19 22 24 II Auditory locliztion 12.5±2.2 9 13 15 12.6±2.2 9 13 15 III Intelligibility in noise 12.5±1.5 11 12 14 12.4±1.4 11 12 14 IV Intelligibility in quiet 13.3±1.6 11 14 15 13.3±1.5 11 14 15 V Detection of sounds 13.5±1.8 11 14 15 13.5±1.8 11 14 15 Abbrevitions s in Tble 3. IJOMEH 2017;30(1) 65

M. PAWLACZYK-ŁUSZCZYŃSKA ET AL. Tble 6. Hering bility in terms of scores below 70% of the mximum vlues in the (m)aiadh in the music students ccording to the plyed instrument (m)aiadh score music students (totl) stringed instrument plyers Respondents [% (95% CI)] wind instrument plyers percussion plyers Totl < 70% 3.1 (1.2 7.2) 1.0 (0.01 5.90) 3.8 (0.40 13.60) 18.2 (4.30 49.00) Subscle < 70% I Distinction of sounds 1.8 (0.4 5.6) 2.0 (0.20 7.60) 0.0 (0.00 7.70) 9.1 (0.01 40.00) II Auditory locliztion 17.8 (12.7 24.5) 16.2 (10.10 24.80) 15.1 (7.70 27.40) 45.5 (21.40 71.90) III Intelligibility in noise 6.7 (3.7 11.8) 6.1 (2.60 12.90) 3.8 (0.40 13.60) b 27.3 (9.50 57.20),b IV Intelligibility in quiet 5.5 (2.8 10.3) 5.1 (1.90 11.70) 1.9 (0.01 10.30) b 27.3 (9.50 57.20),b V Detection of sounds 6.7 (3.7 11.8) 4.0 (1.30 10.30) 7.5 (2.60 18.50) 27.3 (9.50 57.20) Abbrevitions s in Tbles 1 nd 5. Significnt differences between the percussion nd stringed instrument plyers (p < 0.05/3). b Significnt differences between the percussion nd wind instrument plyers (p < 0.05/3). noise) nd subscle IV (evluting intelligibility in quiet) (p < 0.05/3). Results of hering tests Audiometric hering threshold levels determined in the 168 music students (335 ers, one student hd single sided defness) nd in 67 subjects (134 ers) from the control group re shown in Figure 5. Mjority of the study subjects hd norml hering (HTLs in the frequency rnge 1000 8000 Hz 20 db HL). Furthermore, both speech-frequency hering loss (men threshold t 500 Hz, 1000 Hz, 2000 Hz nd 4000 Hz > 20 db HL) nd high-frequency hering loss (men threshold t 3000 Hz, 4000 Hz nd 6000 Hz > 20 db HL) were observed only in few percent of the nlyzed udiogrms (Tble 7). Nevertheless, typicl NIHL notches t 4000 Hz or 6000 Hz of t lest 15 db depth (reltive to both the best preceding threshold occurring t frequencies 1000 3000 Hz (or 1000 4000 Hz) nd the threshold t 8000 Hz) were observed in 13.4% nd 9% of udiogrms in the music students nd the control group, respectively (Tble 8). Most of them (97.8% nd 91.7%) occurred t 6000 Hz. The portion with bilterl notching t ny frequency ws 4.2% nd 1.5% in the music students nd control group, respectively. Generlly, the music students pure tone ir-conduction thresholds were better (lower) in the frequency HTL [db HL] 0 5 10 15 20 Frequency [Hz] 1 000 1 500 2 000 3 000 4 000 6 000 8 000 control group music students Significnt differences between the music students nd the control group (p < 0.05). Dt re given s men vlues with 95% confidence intervls. Fig. 5. Audiometric hering threshold levels (HTL) determined in the music students (N = 168; 335 ers) nd in the control group (N = 67; 134 ers) 66 IJOMEH 2017;30(1)

HEARING STATUS IN CLASSICAL MUSIC STUDENTS ORIGINAL PAPER Tble 7. Prevlence of pure tone udiometry bnorml results in the music students nd in the control group Hering threshold level Pure tone udiometry music students Respondents [% (95% CI)] control group 0.25 8 khz > 20 db HL 24.2 (19.90 29.10) 26.1 (19.4 34.2) 1 8 khz > 20 db HL 17.0 (13.40 21.40) 17.2 (11.7 24.5) 0.5 kh, 1 kh, 2 kh nd 4 khz > 20 db HL 8.1 (5.60 11.50) 6.7 (3.5 12.5) 3 6 khz > 20 db HL 10.4 (7.60 14.20) 11.9 (7.4 18.7) Men hering threshold level in frequency 0.25 8 khz > 20 db HL 0.3 (0.01 1.90) 3.0 (0.9 7.8) 1 8 khz > 20 db HL 0.9 (0.20 2.80) 3.0 (0.9 7.8) 0.5 kh, 1 kh, 2 kh nd 4 khz > 20 db HL 0.6 (0.02 2.30) 3.0 (0.9 7.8) 3 6 khz > 20 db HL 2.1 (0.90 4.40) 3.7 (1.4 8.7) CI confidence intervl. Significnt differences between the music students nd the control group (p < 0.05). Tble 8. Prevlence of high-frequency notched udiogrms in the music students nd in the control group Totl High-frequency notching music students Respondents [% (95% CI)] control group 6 khz 13.1 (9.90 17.20) 8.2 (4.50 14.30) 4 khz 0.3 (0.01 1.90) 0.7 (0.01 4.20) 4 khz or 6 khz 13.4 (10.20 17.50) 9.0 (5.10 15.20) Right er 6 khz 13.2 (8.80 19.20) 6.0 (2.00 14.90) 4 khz 0.0 (0.00 2.80) 0.0 (0.00 6.70) 4 khz or 6 khz 13.2 (8.80 19.20) 6.0 (2.00 14.90) Left er 6 khz 13.1 (8.80 19.10) 10.4 (4.90 20.40) 4 khz 0.6 (0.01 3.60) 1.5 (0.06 5.20) 4 khz or 6 khz 13.7 (9.30 19.80) 11.9 (6.00 22.20) Bilterl notching 6 khz 4.2 (1.90 8.60) 1.5 (0.01 8.80) 4 khz 0.0 (0.00 2.80) 0.0 (0.00 6.70) 4 khz or 6 khz 4.2 (1.90 8.60) 1.5 (0.01 8.80) CI confidence intervl. IJOMEH 2017;30(1) 67

M. PAWLACZYK-ŁUSZCZYŃSKA ET AL. rnge 250 3000 Hz (Figure 5) s compred to the control group (p < 0.05). Furthermore, lower percentge of udiogrms with speech-frequency hering loss ws observed in the musicins. However, there were no significnt differences in hering threshold levels between the students nd control group in the frequency rnge 4000 8000 Hz (p > 0.05). Moreover, neither proportion of high-frequency losses (Tble 7) nor proportion of NIHL notches in udiogrms (Tble 8) differed significntly in both groups. SHTL 4 3 2 1 0 1 control group music students 500 1 000 2 000 3 000 4 000 6 000 8 000 Frequency [Hz] Comprison of ctul nd predicted hering threshold levels Figure 6 presents the stndrdized hering threshold levels in the music students. It is worth noting tht the closer to zero vlue of SHTL, the better the prediction of hering loss ccording to ISO 1999:2013 [22]. On the other hnd, positive vlue of SHTL indictes tht ctul hering threshold level is higher thn the predicted one, while negtive vlue denotes lower-thn-predicted level. As cn be seen, hering loss in the music students ws higher (worse) thn the one predicted (from their weekly noise exposure level) t 6000 Hz, nd lower (better) for 3000 Hz SHTL 4 3 2 1 0 1 2 noise-exposed popultion non-noise-exposed popultion 500 1 000 2 000 3 000 4 000 6 000 Frequency [Hz] All vlues of SHTLs differed significntly from 0 (p < 0.05). Dt re given s men vlues with stndrd devition. Fig. 6. Stndrdized hering threshold levels (SHTL) (reltive to reference non-noise-exposed otologiclly norml nd noise-exposed popultion ccording to ISO 1999:2013 [22]) in the music students (N = 168; 335 ers) (equtions 3 nd 4) Significnt differences between the music students nd the control group (p < 0.05). All vlues of SHTLs differed significntly from 0 (p < 0.05). Dt re given s men vlues with stndrd devition. Fig. 7. Stndrdized hering threshold levels (SHTL) (reltive to reference non-noise-exposed otologiclly norml popultion ccording to ISO 1999:2013 [22]) in the music students (N = 168; 335 ers) nd in the control group (N = 67; 134 ers) (equtions 3 nd 4) nd 4000 Hz (p < 0.05) (Figure 6). Furthermore, the ctul hering threshold levels were significntly higher thn those predicted for 500 Hz, 1000 Hz nd 2000 Hz. Comprison of the music students to the otologiclly norml (highly screened) non-noise-exposed popultion (dtbse A from ISO 1999:2013 [22]) reveled tht the music students hering losses (in the frequency rnge 1000 8000 Hz) were higher thn those expected (p > 0.05) (Figure 6). Similr reltionships were noted when nlysing hering losses in the control group (Figure 7). Furthermore, there were no significnt differences between the stndrdized hering threshold levels (reltive to the non-noise-exposed popultion) t 4000 Hz, 6000 Hz nd 8000 Hz between both groups, while SHTLs in the frequency rnge 500 3000 Hz were lower in the music students compred to the control group. Assocition between music exposure nd hering Wek but sttisticlly significnt reltionships were noted between udiometric hering threshold levels in the frequency rnge 3000 6000 Hz nd some of the sound 68 IJOMEH 2017;30(1)

HEARING STATUS IN CLASSICAL MUSIC STUDENTS ORIGINAL PAPER Tble 9. Reltions between udiometric hering threshold levels (HTL) nd sound exposure prmeters in the music students Audiometric HTL frequency [Hz] solo nd group plying A-weighted equivlent continuous SPL group plying Spermn s rnk correltion coefficient solo plying weekly noise exposure level noise immission level 1000 0.032 0.048 0.002 0.009 0.005 2000 0.011 0.045 0.019 0.009 0.024 3000 0.043 0.022 0.072 0.045 0.129 4000 0.046 0.003 0.135 0.073 0.066 6000 0.002 0.001 0.115 0.009 0.080 8000 0.070 0.082 0.053 0.055 0.054 SPL sound pressure levels. Sttisticlly significnt correltions re given in bold. exposure prmeters (Tble 9). In prticulr, HTLs t 4000 Hz nd 6000 Hz were positively correlted with the A-weighted equivlent-continuous SPL during solo plying, while t 3000 Hz with noise immission level. Furthermore, there were differences in high-frequency hering losses between the students plying stringed, HTL [db HL] 0 5 10 15 20 25 Frequency [Hz] 1 000 1 500 2 000 3 000 4 000 6 000 8 000 wind instrument plyers stringed instrument plyers percussion plyers Significnt differences between the percussion nd wind instrument plyers (p < 0.05/3). b Significnt differences between the percussion nd stringed instrument plyers (p < 0.05/3). Dt re given s men vlues with 95% confidence intervls. Fig. 8. Audiometric hering threshold levels (HTL) in the music students plying stringed, wind nd percussion instruments b b b b b b wind nd percussion instruments (Figure 8). The percussion plyers, who were exposed to the highest L EX,w levels, hd significntly greter hering losses thn the lowerexposed students plying stringed nd wind instruments (p < 0.05/3). However, there were no significnt differences between the stringed nd wind instrument plyers. In the percussion plyers (22.7%, 95% CI: 9.9 44%) compred to the subjects plying stringed (13.4%, 95% CI: 9.4 18.9%) nd wind (11.6%, 95% CI: 6.8 19%) instruments the prevlence of notched udiogrms ws lso greter; however, it ws sttisticlly insignificnt (Figure 9). Since the prevlence of high-frequency notched udiogrms cn vry with ge, gender, noise exposure nd other fctors, to nlyze their influence, the binry logistic regression ws pplied with the logistic model expressed s follows: ln[p/(1 p)] = b 0 +b 1 x 1 +b 2 x 2 +... +b n x n (5) where: p probbility of the outcome, i.e., x 1, x 2,, x n re the independent vribles included in the model (i.e., ge, gender, listening to PMPs (yes/no), period of plying instrument, in yers, estimted weekly noise exposure level, in db), b 0, b 1,, b n regression coefficients (i.e., the logrithmic vlues of the odds rtio). IJOMEH 2017;30(1) 69

M. PAWLACZYK-ŁUSZCZYŃSKA ET AL. Notched udiogrms [%] 50 40 30 20 10 0 stringed instrument wind instrument percussion No significnt differences between the pirs of subgroups of plyers (p > 0.05/3). Dt re given s men vlues with 95% confidence intervls. Fig. 9. Prevlence of high-frequency notched udiogrms in the music students plying stringed, wind nd percussion instruments An odds rtio with 95% CI > 1 indictes positive correltion between the dependent vrible (i.e., prevlence of highfrequency notches in udiogrm) nd explntory vrible (e.g., ge), while vlue < 1 indictes negtive correltion between the dependent vrible nd explntory vrible. Plyers The binry logistic regression reveled tht the prevlence of high-frequency notched udiogrms ws positively ssocited only with the students weekly noise exposure levels (OR = 1.07, 95% CI: 1.014 1.13, p < 0.05) (Tble 10). However, neither sound exposure relted to music eduction nor frequent listening to personl music plyers vi hedphones hd impct on the prevlence of tinnitus in the music students. DISSCUSION The study ws designed to determine if clssicl music students re t higher risk of hering loss compred to non-music students nd non-musicins. It lso ttempted to nswer if hering sttus of music students reflects their exposure to excessive sounds during university eduction. In this study, the students exposure ws evluted from the sound pressure levels mesured during vrious students ctivities using personl noise dosimeters nd the declred time of weekly prctice obtined from the questionnire survey. It ws evluted tht the music students were exposed for 10 42 h (10 90th percentile) per week to sounds t the A-weighted equivlent-continuous sound pressure level of 80 97 db (10 90th percentile). Thus, bout 49% nd 29% of the students were exposed to sounds t weekly noise exposure level (L EX,w ) exceeding the upper ction limit vlue (85 db) nd the exposure limit vlue (87 db), respectively. These outcomes re similr to the results of erlier studies both, those Tble 10. Assocition between high-frequency notched udiogrms or tinnitus in the music students (dependent binry vrible) nd weekly noise exposure level nd other individul fctors (independent vribles) tested using the logistic regression Vrible Notched udiogrm Tinnitus b OR (95% CI) p OR (95% CI) p Age 0.851 (0.719 1.007) 0.061 0.936 (0.742 1.182) 0.579 Gender 0.579 (0.279 1.204) 0.143 1.405 (0.489 4.033) 0.525 Listening to PMP 2.069 (0.598 7.165) 0.250 1.242 (0.435 3.542) 0.684 Period of plying instrument 1.098 (0.989 1.220) 0.081 1.072 (0.924 1.244) 0.359 Weekly noise exposure level 1.070 (1.014 1.130) 0.014 0.961 (0.877 1.054) 0.401 PMP personl medi plyer; OR odds rtio; CI confidence intervl. Chi 2 = 12.798, df = 5, p = 0.0253641. b Chi 2 = 2.215, df = 5, p = 0.8186489. Sttisticlly significnt ssocition is given in bold. 70 IJOMEH 2017;30(1)

HEARING STATUS IN CLASSICAL MUSIC STUDENTS ORIGINAL PAPER concerning young musicins nd professionl orchestrl musicins [13 17]. In this study, pure-tone ir-conduction thresholds nd hering-relted symptoms in the clssicl music students were compred to the control group. Since HTLs of the study subjects were lso compred to the ge-relted reference dt from highly screened otologiclly norml persons (dtbse A from ISO 1999:2013), the control group comprised norml (unscreened) young people, minly non-music students, mong whom bout one-qurter were occsionlly exposed to noise during internship or pprenticeship. Generlly, there were no significnt differences between the music students nd the control group in terms of ge, gender, medicl history, physicl fetures nd some spects of lifestyle such s noisy hobby. However, greter frctions of the music students frequently listened to personl medi plyers, while greter proportion of the subjects in the control group often used noisy tools in the pst. Individul susceptibility (or vulnerbility) to noise, long with the degree of hering loss, vries gretly mong people. It is believed tht NIHL is complex disese resulting from interction between intrinsic nd environmentl fctors. Besides, well-known environmentl fctors contributing to NIHL, such s exposure to noise nd some other fctors, my lso ply role. They include co-exposures to ototoxic substnces (orgnic solvents nd hevy metls), co-exposure to noise nd vibrtion, ototoxic drugs (minoglycosides) nd het. Associtions hve lso been observed between severl individul fctors nd NIHL, including: smoking, elevted blood pressure, dibetes, cholesterol levels, skin pigmenttion, gender nd ge, nd genetic predisposition s suggested by clinicl knowledge nd guidelines [28]. Additionl NIHL risk fctors were rre in both groups of the subjects. Furthermore, there were no significnt differences between the music students nd the comprtive group concerning the prevlence of smoking, elevted blood pressure, dibetes, white-finger syndrome, light skin pigmenttion, ototoxic ntibiotic tretments nd overweight (BMI > 25). Hering loss is permnent threshold shift exceeding predefined limit. The limit vlue hs not been stndrdized so fr. In this study, the hering thresholds up to 20 db HL were considered s norml since some studies of NIHL prevlence in dults hve defined NIHL in terms of bsolute thresholds levels bove 20 db HL [29]. Thus, mjority of the music students nd control subjects were found s hving norml hering becuse in both groups in the cse of 83% of ers HTLs (in the frequency rnge 1000 8000 Hz) did not exceed 20 db HL. Only few of them hd high-frequency hering loss (in the frequency rnge of 3000 6000 Hz) greter thn 20 db HL, but there were no significnt differences in the proportion of bnorml udiogrms between the groups. However, the music students pure-tone ir-conduction thresholds were better in the frequency rnge 250 3000 Hz s compred to the control group, while no significnt differences were observed in higher frequency rnge (4000 8000 Hz). Comprison of the music students nd control subjects to the highly screened otologiclly norml non-noiseexposed popultion (dtbse A from ISO 1999:2013) reveled tht their hering losses (in the frequency rnge of 1000 8000 Hz) were higher thn those expected (Figure 7). Moreover, there were no significnt differences between the stndrdized hering threshold levels (reltive to the non-noise-exposed popultion) t 4000 Hz, 6000 Hz nd 8000 Hz between both groups, while SHTLs in the frequency rnge of 500 3000 Hz were lower in the music students compred to the control group, i.e., closer to hering threshold levels in the highly screened otologiclly norml non-noise-exposed popultion. Furthermore, ctul hering loss in the music students ws significntly higher (worse) thn tht predicted from the sound exposure level (ccording to ISO 1999:2013) IJOMEH 2017;30(1) 71

M. PAWLACZYK-ŁUSZCZYŃSKA ET AL. t 6000 Hz, nd lower (better) t 3000 Hz nd 4000 Hz (Figure 6). Recently, Lüders et l. [30] hve nlyzed hering threshold levels 250 16 000 Hz in group of 42 music students in comprison to non-musicin group in order to determine whether high-frequency udiometry is useful tool in erly detection of hering impirment. When tested using conventionl udiometry, likewise in our study, mjority of the subjects (92.9%) hd hering thresholds within norml limits. However, contrry to our results, both conventionl nd high-frequency udiometry reveled sttisticlly significnt differences when compring the udiometric thresholds of the music students group nd control group, with the worst threshold noted in the group of music students. The most significnt differences were found in the evlution of high frequencies, which llows for the inference tht spordic high-frequency threshold ssessment cn be useful in erly detection of hering loss in musicins. Since in our study mjority of the subjects hd hering thresholds within norml limits, in order to identify erly signs of NIHL, the prevlence of high-frequency notches (i.e., shrp drop in the hering sensitivity t 4000 Hz or 6000 Hz) in udiogrms ws nlyzed. In our study, high-frequency notches (minly t 6000 Hz) were found in 13.4% nd 9% of udiogrms in the musicins nd non-musicins, respectively. Moreover, no significnt differences were noted between both groups. The prevlence of notches t 6000 Hz in the control group is not surprising since ccording to the Nord-Trøndelg Hering Loss Study [31] such notches were observed both in unscreened nd screened for history of noise exposure nd er-relted disorders nd diseses popultions of men nd women up to 40 yers old. Moreover, some erlier studies in musicins, including professionl orchestrl musicin, hve lso shown high frequency notches (minly t 6000 Hz) in udiogrms [7,9,10]. Similr outcomes hve been lso obtined for younger (student) musicins [18,19]. For exmple, Phillips et l. [19] hve nlyzed hering cpbility in music students (N = 329) nd found typicl NIHL notches in 45% of them, with 78% of notches occurring t 6000 Hz. Proportion of the totl popultion with bilterl notching t ny frequency ws 11.5%, mostly occurring t 6000 Hz. There ws significnt increse in the frequency of notching in students who reported more thn 2 h/dy of personl prctice. However, no significnt ssocitions were observed for instrument group or other noise exposures. Generlly, vrious definitions of NIHL notches re pplied. Notches were conservtively defined (from clinicl stndpoint) s t lest 10 db drop in the threshold from 1000 Hz, 2000 Hz, or 3000 Hz to 4000 Hz or from 1000 Hz, 2000 Hz, 3000 Hz, or 4000 Hz to 6000 Hz, with t lest 5 db recovery t 8000 Hz. In the foresid study, notches were defined s t lest 15 db drop t 4000 Hz or 6000 Hz reltive to the best preceding frequencies bck to 1000 Hz [19]. On the other hnd, in this study we nlyzed the presence of notches t lest of 15 db depth reltive to both the best preceding threshold from 1000 Hz, nd the threshold t 8000 Hz. Thus, it is not surprising tht we found smller percentge of notched udiogrms compred to the erlier observtions (13% vs. 45%) [19]. Nevertheless, in this study the odds rtio of notching in the music students incresed significntly long with higher weekly noise exposure level (OR = 1.07, 95% CI: 1.014 1.13, p < 0.05). When pplying less severe criteri, the proportion of notched udiogrms incresed to 22.7%, but the ssocition between music exposure nd their incidence ws no longer vlid. It is worth noting tht there were differences (but not sttisticlly significnt) in the prevlence of high-frequency notched udiogrms mong the young musicins plying stringed, wind nd percussion instruments. The highest incidence of notches ws observed in the percussion plyers, who were exposed to the highest L EX,w levels. 72 IJOMEH 2017;30(1)