SLUTRAPPORT Dnr

SLUTRAPPORT Dnr 070104 Hörselskaderisk vid kombinerad exponering för buller och vibrationer Lage Burström Hans Pettersson Tohr Nilsson UMEÅ UNIVERSITET Folkhälsa och klinisk medicin Yrkes- och miljömedicin 2011 1

FÖRORD Denna rapport utgör slutredovisning av projektet Hörselskaderisk vid kombinerad exponering för buller och vibrationer (AFA dnr 070104) för vilket AFA beviljade medel under 2007. I arbetet har medverkat; Lage Burström (Forskare projektledare), Folkhälsa och klinisk medicin, Yrkesoch miljömedicin, Umeå universitet Tohr Nilsson (Lektor/överläkare), Folkhälsa och klinisk medicin, Yrkes- och miljömedicin, Umeå universitet samt Arbets- och miljömedicin, Sundsvalls sjukhus Hans Pettersson (Doktorand), Folkhälsa och klinisk medicin, Yrkes- och miljömedicin, Umeå universitet. INNEHÅLLSFÖRTECKNING 1. Bakgrund 3 2. Genomförande 3 3. Resultat 4 4. Diskussion 5 5. Fortsatt arbete 7 6. Referenser 7 Bilaga 1 Kostnadssammanställning Bilaga 2 Publiceringar 2

1. Bakgrund Inom många yrken exempelvis i byggnads- och tillverkningsindustrin använder anställa handhållna maskiner som alstrar både buller och vibrationer. Risken för hörselnedsättning bland dessa anställda kan bero på mer än bara bullnivåerna utan även kombinationen av de båda exponeringarna. Den ökade risken för hörselskada verkar också vara beroende av om den exponerade har tidigare diagnostiserade vibrationsskador eller besvär som påverkar blodcirkulationen i fingrarna. Den kunskap som idag finns om hörselskaderisk vid kombinerad exponering för buller och vibrationer är begränsad. I Arbetsmiljöverkets föreskrift om buller påpekas också den ökade risken vid samverkande exponering. Riktlinjer för vilka nivåer på kombinerade exponeringar som bör beaktas saknas dock. Syftet med projektet för vilket AFA beviljat medel har varit att undersöka sambandet mellan samtidig exponering för buller och vibrationer samt risken att utveckla hörselskada. Vidare är syftet att studera hur sårbarhetsfaktorer som medför försämrad blodcirkulation i fingrarna påverkar risken för hörselskada. 2. Genomförande Före igångsättande av arbetet har projektet i sin helhet granskats och godkänts av Regionala etikprövningsnämnden i Umeå (Dnr 08-151M). I undersökningen har ingått både laborativ-, kohort- och enkätstudie. Laborativstudie Projektet har utvidgats utifrån den ursprungliga projektplanen för att också innefatta en laborativ studie. Syftet med denna studie var att undersöka hur den tillfälliga hörselnedsättningen påverkades vid separat och samtidig exponering för buller respektive vibrationer. Till projektet rekryterade 22 forskningspersoner, 11 kvinnor och 11 män (medelålder 22 år), med audiometriskt verifierad fullgod hörsel samt fullgott känsel- och temperatursinne (undersökt med vibrametri och termotest). Samtliga deltagare i studien besvarade en enkät som innefattade frågor om bland annat tidigare buller/vibrationsexponering (både arbete och fritid), ålder, vikt, längd, tobaks- samt alkoholkonsumtion, mediciner samt eventuella problem i fingrar, händer, armar, nacke, rygg och axlar. Även frågor om menstruationscykeln samt preventivmedelsanvändning ingick för de kvinnliga deltagarna. Varje forskningsperson informerades om att undvika bullriga platser och att arbeta med vibrerande maskiner under dagen för respektive experiment. De skulle också undvika te, kaffe, tobak och inte utföra intensiv fysisk aktivitet en timme innan experimentet. Högst en exponering genomfördes per dygn. Experimenten utfördes i ett semi-ekofritt rum byggt med ljudabsorberande väggar. forskningspersonernas hörsel mättes före exponeringen och även från 30 s upp till 30 minuter efter exponeringen för att studera hur deras hörsel återhämtade sig. 3

Forskningspersonerna exponerades vid tre olika tillfällen för buller, vibrationer respektive en kombination av dessa. Exponeringen bestod av buller respektive vibrationer som spelats in vid användning av en vinkelslipmaskin. Varje exponering pågick under en tidsperiod av 20 min. De buller- och vibrationsnivåer som använts var 99 db(a) respektive 6.7 m/s 2. Bullerexponeringen motsvarar en 8 timmars exponering av 85 db(a) och vibrationer av 1.4 m/s 2. Exponering av buller skedde via hörlurar och vibrationerna via en vibrator utrustad med ett handtag. Utfallet var den tillfälliga hörselnedsättningen (TTS) som uppstod för höger öra vid olika frekvenser (1, 4 och 8 khz). Kohortstudien Syftet med kohortstudien var att undersöka om det föreligger ett dos - responssamband mellan samtidig exponering för buller och vibrationer samt risken att utveckla hörselskada. Undersökningen utfördes på en sedan tidigare etablerad dynamisk kohort bestående av ca 280 manliga industriarbetare bland annat svetsare, slipare, ingenjörer, försäljare och administratörer. Kohorten har följts prospektivt sedan starten 1987 och i uppföljningar 1992, 1997 samt i en enkätuppföljning 2002. Fältstudierna har skett parallellt med den 21-års-uppföljning som skett av kohorten, under våren 2008. För varje deltagare i kohorten har data sammanställts i en databas som omfattar uppgifter från olika frågeformulär, läkarundersökning och kliniska undersökningar för de olika undersökningsåren sedan starten 1987. Dessa uppgifter har kombinerats med exponeringsuppgifter för buller och vibrationer. Under uppföljningen 2008 har exponering för buller och vibrationer uppmätts med handhållna instrument för olika arbetsmoment. Totalt har mätningar genomförts för ca 120 handhållna vibrerande maskiner. Genomförda observationsstudier gav kompletterande information om exponeringstiden för de vibrerande maskiner. För buller har även personburen dosimeter använts. Vibrationer har under tidigare års uppföljningen regelbundet uppmätts. De i kohorten som exponerats för buller har undersökts regelbundet med audiometri. Audiometriska mätningar utfördes mer omfattande när kohorten startade och har minskat i omfattning under åren. Genom samarbete med företagshälsovården samt landstingets centrala arkiv i Härnösand har totalt har 1586 audiogram insamlats. För varje deltagare i kohorten har olika exponeringsdoser beräknats för varje undersökningsår. Dosmåtten bygger på kumulerade exponeringstid samt olika varianter av exponeringstiden multiplicerat med exponeringsnivån. Utfallet av kohortstudien är att undersöka om och hur dokumenterad hörselförsämring är relaterad till samtidigvibrationsexponeringen. Enkätstudie Arbetet med att ur AFA: s databas om arbetsskador selektera 400 kvinnor och män som anmält och erhållit ersättning (medicinsk invaliditet) för hörselskada har påbörjats. Arbetet är tidsmässigt planerat för igångsättande under försommaren 2011 och till de personer som erhållit ersättning sänds en kompletterande enkät med frågorna om tid i arbete i bullrig miljö, livstidsexponering för vibrationer, hörselproblem, tinnitus och livstidshistoria för köldintolerans och för vita fingrar. 4

3. Resultat Arbetet har resulterat i ett antal publikationer som beskrivs närmare nedan. Laborativstudie Artikeln The effect on the temporary threshold shift in hearing acuity from combined exposure to authentic noise and hand-arm vibration (ref 1) beskriver sambandet mellan tillfällig hörselnedsättning vid kombinerad exponering för buller och vibrationer och är under publicering i tidskriften International Archives of Occupational and Environmental Health. Kortfattat visar studien att en kombinerad exponering för buller och vibrationer inte ger någon skillnad i effekt på den tillfälliga hörselnedsättningen jämfört med enbart buller exponering. Vidare framkom att det inte förelåg någon skillnad mellan män och kvinnor vad gäller påverkan. Kohortstudie Artikel A follow-up study of welders in a heavy engineering production workshop exposed to vibration (ref 2) som publicerats i tidskriften Journal of Low frequency noise, vibration and active control beskriver hur exponering för främst vibrationer förändrats mellan år 1987 och år 2008. Resultaten visar bland annat att exponeringen under denna tidsperiod minskat från 3.9 m/s 2 till 1.9 m/s 2. Delar av materialet från kohortstudien har också presenterats vid 2nd International Conference on Human Vibration Exposure, measurement and tests, June 16-17, 2009 (ref 3). Statistiska analyser och skrivande av artikel om sambandet mellan vibrationsexponeringen och risken för hörselförsämring pågår. Preliminära resultat visar att det föreligger ett samband mellan vibrationsexponeringen och öka risk för hörselförsämring beroende på både exponeringstid och exponeringsnivå. 4. Diskussion Resultatet av våra studier både stödjer och inte stödjer hypotesen att samtidig exponering för både buller och vibrationer ökar risken för hörselnedsättning jämfört med enbart bullerexponering. Tillfällighörselnedsättning (TTS) har använts i tidigare experimentella studier för att mäta effekten av buller, vibrationer eller en kombination av dessa. Våra resultat visar att efter exponering för buller eller kombinerad exponering för buller och vibrationer gav en TTS som är stor och kräver en lång återhämtningsperiod för hörselstatus vid 4 och 8 khz. För 1 khz, var TTS att liten med en kort återhämtningsperiod och vi fann ingen skillnad i det beräknade medelvärdet för TTS mellan de tre exponeringar. För 4 och 8 khz fann vi inga signifikanta skillnader i TTS mellan bullerexponering och kombinerad exponering, ett konstaterande i enlighet med vissa studier, men inte med andra. 5

Det finns naturligtvis flera möjliga förklaringar till skillnaderna mellan denna studie och tidigare studie, både vad gäller typ av exponering som valts, forskningspersonernas hörselstatus och ålder. Resultaten från vår studie visar att kombinerad exponering för buller och vibrationer inte producera olika TTS jämfört med endast bullerexponering. Emellertid var forskningspersonerna unga och friska som inte tidigare regelbundet varit yrkesmässig exponerade för buller och vibrationer. Därför bör försiktighet iakttas vid generalisering av resultaten till de äldre som är utsatta för buller och vibrationer under långtid liksom till om den exponerade inte varit fullt frisk. Vidare är relationen mellan mätningarna av tillfällig hörselförändring producerad av buller i kombination med exponering för vibrationer och permanent hörselförsämring inte helt klarlagd. En tolkning av våra resultat kan vara att de i andra studier noterade effekten av hörselförsämring vid en kombinerad exponering beror på själva användningen av handhållna vibrerande maskiner producerar hög exponering för buller. För den studerade kohorten har en särskild studie gjorts för att undersöka exponeringens förändring över tid. Studie visar att den genomsnittliga vibrationsbelastningen bland studerade svetsare minskat under en 21-årsperiod, både uttryckt som vibrationsnivå och exponeringstid. Exponering för vibrationer orsakades i första hand från användning av slipmaskiner och slaggmaskiner. Dessa två typer av maskiner motsvarar till 96 till 98 % av den totala daglig användning av handhållna verktyg. Under studieperiod minskade maskinernas acceleration med mer än 20 % för slipmaskiner och 30 % för slaggmaskiner. Samtidigt minskade exponeringstiden för dessa maskiner med cirka 50 %. Minskningen av både storleken på accelerationen och exponeringstiden medförde att vibrationsbelastningen minskar med ca 50 %, uttryckt som den 8-timmes värde. Under den 21-åriga perioden minskade den frekvensvägda accelerationen under en 8-timmars arbetsdag från 3,9 m/s 2 år 1987 till 1,9 m/s 2 år 2008. Minskningen är också statistiskt signifikant. Den minskade vibrationsbelastning, uttryckt som 8-timmars värde, kan bero på att undersökningen i sig själv ökat företagets medvetenhet om exponeringen. Under årens lopp har företaget ersatt många maskiner med nyare mindre vibrerande samt infört regelbundet underhåll för handhållna maskiner och utvecklat rutiner för inköp av både slipskivor och mejslar. Dessutom har bättre design av produkter lett till mindre ytbehandling och förändringar i svetsprocessen har minskat behovet av mejsling av svetssömmarna. Samma minskning kan också konstateras för exponering för buller. Trots denna minskning i exponeringen och även om kohorten är relativt liten tyder resultaten på att en daglig vibrationsexponering vid insatsvärdet (2.5 m/s 2 ) att risken för hörselförsämring ökar med 4.4 % över en tioårsperiod. Dessutom visar resultaten att denna risk skiljer sig mellan olika yrkesgrupper. 6

5. Fortsatt arbete Även om projektet formellt avslutas kommer arbetet att drivas vidare. Manusframställandet av samband mellan exponering för buller och vibrationer samt den förknippade risken för hörselskada liksom sårbarhetsfaktorers (diagnosticerade besvär) inverkan kommer att färdigställas. Vidare kommer också arbetet med enkätundersökningen i samarbetet med AFA försäkring att slutföras. Samtliga delstudier kommer dessutom att ingå i Hans Petterssons avhandling som förväntas presenteras under december 2012. 6. Referenser 1. Pettersson H, Burström L, Nilsson T. The effect on the temporary threshold shift in hearing acuity from combined exposure to authentic noise and hand-arm vibration. Int Arch Occup Environ Health. 2011 Apr 16 (Epub ahead of print). 2. Burström L, Hagberg M, Liljelind I, Lundström R, Nilsson T, Pettersson H, Wahlström, J. A follow-up study of welders exposure to vibration in a heavy engineering production workshop. J Low Freq Noise, Vib Active Control 2010 29(1), 33 39. 3. Burström L, Pettersson H, Nilsson T, Wahlström J, Liljelind I, Lundström R, Hagberg M. A 21-year follow-up of the vibration load among workers in a heavy engineering production workshop. 2nd International Conference on Human Vibration Exposure, measurements and tests, Boden, Sweden, June 16th to17th 2009, pp 115-118 4. Pettersson H, Burström L, Nilsson T. Dose-response relationship between exposure to handarm vibration and noise in relation to hearing loss (manuscript). 7

Bilaga 1 Kostnadssammanställning I tabellen nedan framgår en sammanställning av den ekonomiska redovisningen där planerad budget jämförs med utfallet. Budget Utfall Löner forskare/ass 1808000 2021119 Övrigt 105000 106000 Resor 68000 72000 Omkostnadsavgift 435820 351820 Högskoleavgift 210263 80000 Summa 2627083 2630939 Kommentarer: Som framgår av budgeten överskattades kostnaderna för omkostnadsavgifter och högskoleavgifter. Därmed har förutsättningar givits för mer arbetsinsatser inom projektets ram. 8

Int Arch Occup Environ Health DOI 10.1007/s00420-011-0635-6 ORIGINAL ARTICLE The effect on the temporary threshold shift in hearing acuity from combined exposure to authentic noise and hand-arm vibration Hans Pettersson Lage Burström Tohr Nilsson Received: 9 September 2010 / Accepted: 27 March 2011 Ó Springer-Verlag 2011 Abstract Objectives This study examined and compared the effect on temporary threshold shift in hearing (TTS) in healthy subjects of noise and hand-arm vibration (HAV) combined and separately using controlled and authentic exposure conditions. This study also investigated the effect on TTS in hearing in relation to gender after such exposures. Methods Twenty-two healthy subjects (11 men/women, mean age 22 years, range 18 31 years) were exposed both separately and in combination with HAV (6.7 m/s 2 ), using vibrating handles and to noise (99dB(A)) using headphones, for 20 min. The HAV and noise were reproduced from recordings from angular grinder in operation. Hearing thresholds at 1, 4, and 8 khz were measured before and up to 30 min after exposure. Results Combined exposure to noise and HAV created significantly greater TTS in hearing than HAV exposure at 4 and 8 khz alone. After exposure to HAV, there was no significant change in hearing threshold. At 1 khz, there was a significant difference between noise and HAV exposure in TTS in hearing. There was no significant difference between combined exposure and noise exposure for any test frequency. There was no significant difference in TTS in hearing in relation to gender for 1, 4, and 8 khz for HAV, noise, or a combined exposure. H. Pettersson (&) L. Burström T. Nilsson Department of Public Health and Clinical Medicine, Occupational and Environmental Medicine, Umeå University, 901 87 Umeå, Sweden e-mail: hans.pettersson@envmed.umu.se T. Nilsson Department of Occupational and Environmental Medicine, Sundsvall Hospital, Sundsvall, Sweden Conclusions The results indicate that there is no difference in the TTS in hearing after combined exposure compared to noise exposure alone. HAV exposure did not change the hearing threshold. The TTS in hearing did not differ significantly in relation to gender after HAV, noise, or combined exposure. Keywords Hand-arm vibration (HAV) Noise Temporary threshold shift (TTS) Gender Combined exposure Introduction Noise exposures can cause both temporary hearing loss and permanent hearing loss. In a temporary threshold shift in hearing (TTS), the hearing is temporally reduced and then fully restored. If the hearing is not fully restored, the hearing loss becomes permanent i.e. it becomes a permanent threshold shift in hearing (PTS) (Quaranta et al. 1998). PTS and TTS in hearing depend on the level, frequency content, and duration of noise exposure (Quaranta et al. 1998). Several factors modify the risk of developing PTS and TTS in hearing, such as, age and medical, chemical, and genetic factors (Hodgkinson and Prasher 2006, Lee et al. 2005; Pyykkö et al. 1987; Quaranta et al. 1998). Vibration has also been identified as one of those modifying risk factors (Iki et al. 1989; Pyykkö et al. 1987, 1981; Starck et al. 1990). A mechanism behind the interaction between hand-arm vibration (HAV) and noise affecting hearing was suggested by Pyykkö et al. (1981). Their suggestion was that vibration triggered a reduction in blood flow to the cochlea due to activation of the sympathetic nervous system, leading to an increased risk of hearing loss. Exposure to vibration as a modifying risk 123

Int Arch Occup Environ Health factor has also been examined in two experimental studies, but the results produced were conflicting. Miyakita et al. (1987) found no influence on TTS in hearing in subjects after exposure to HAV and noise compared to noise exposure alone, while Zhu et al. (1997) found a larger TTS after combined exposure. These studies also lack information about different effects on TTS in relation to gender. Aim This study examines and compares the effect on TTS in hearing in healthy subjects after exposure to noise and HAV combined and separately using controlled and authentic exposure conditions. This study also investigates the effect on the TTS in hearing in relation to gender. Methods Subjects This study was approved by the Regional Board of Ethics for Medical Research in Umeå, Sweden (Dnr 08 151 M). Based on results of the study by Zhu et al. (1997), a power calculation with a power of 0.80 and a of 0.05 showed that 16 individuals should be enough to discover any differences between exposures and their effects on TTS in hearing. Twenty-two healthy subjects participated 11 men and 11 women (mean age 22 years, range 18 31 years). Four men and four women used nicotine products (cigarettes and/or snuff). All subjects were recruited by advertisement and gave their written consent to participate in the study according to the Helsinki Declaration. Inclusion criteria and questionnaire Inclusion criteria included a hearing acuity threshold of at least 10 db hearing level (dbhl) for 11 frequencies between 500 and 8,000 khz in both ears. Each subject also had to have normal thermal and tactile perception, i.e., no disorders of the peripheral nerves, for every finger on both hands. All participants had their index and little fingers on their right and left hands tested for tactile and thermal perception. To measure tactile perception, each subject placed their finger tip on a vibrating plate (VibroSense Meter, VibroSense Dynamics AB, Malmö, Sweden) using the von Bekesy method in compliance with ISO 13091-1 (2001). Thermal perception was measured on the distal phalanx of each index and little finger using a standardized method (MSA Thermotest, Somedic Sales AB, Hörby, Sweden). All subjects who met these requirements also completed a questionnaire. The questionnaire asked about their previous experience of exposure to loud noise and vibration levels during work and leisure time, their height, weight, tobacco, and alcohol consumption, if they had injured to their hands, neck, shoulders, arms, or back, and what, if any, medications they were using. Each subject was informed that they should avoid both noisy areas and work with vibrating machines on the day of the exposures. They were asked to abstain from using any type of tobacco products, drinking coffee/tea, or performing intense physical exercise less than an hour before exposure. Environmental condition The experiment was performed in a semi-noise-isolated chamber with an ambient temperature of 22 C. Background noise in the semi-isolated chamber was measured as being below 20 db(a) by a sound level meter (Brüel and Kjær 2209). The background noise level was low enough to perform audiometric tests down to -10 dbhl for 1, 4, and 8 khz according to ISO 8253-1 (1998). During the experiments, the subjects wore normal office clothing. During all exposures and hearing tests, the subject was seated in a corner of the chamber and could not see the research leader. Experimental procedure All subjects were exposed to three different conditions. Each condition lasted 20 min with at least 24 h between exposures to avoid the subject s hearing being affected by earlier exposure (Kryter et al. 1966; Zhu et al. 1997). Following a randomized block design, the order of the three exposure conditions was randomly assigned for each subject. To avoid the influence of practice or carry-over effects on subjects between exposures, the order of all three exposures was counterbalanced. The three exposure conditions were (1) noise through headphones, (2) HAV through vibrating handles, and (3) combined exposure of noise and HAV. The subject s posture during the exposures to noise through headphones and HAV through vibrating handles is shown in Fig. 1. Before and after each exposure, hearing ability for 1, 4, and 8 khz was measured in the right ear. The measurement before exposure was used as baseline data. Then, the hearing was measured 0.5, 3, 5.5, 8, 10.5, 13, 20, and 30 min after exposure. Each subject held with both hands, a feed force of 10(±5) N monitored by a wave platform (TDEA Huntleigh M1250) during all the exposures. The subject was instructed to grip the handles with a normal gripforce during the experiment. During HAV only exposure, each subject wore hearing protection. The background noise level during HAV only exposure was 73 db(a) without hearing protection and 51 db(a) with hearing protection. 123

Int Arch Occup Environ Health amplifier (Brüel and Kjær 2635). Noise exposure came through an amplifier (Sentec PA9) and headphones (Sennheiser HD 250 linear II). Exposure to vibration was produced by an electrodynamic shaker (Ling Altec 7/600) with two handles, by an amplifier (Ling Dynamic System 300) connected to a low-pass filter (Krohn Hite model 3550) and an amplifier (Sentec PA9). Figure 2 shows the noise and vibration spectra presented as functions of the frequency. Statistical analysis Fig. 1 The subject s posture during the exposures to noise through headphones and hand-arm vibration through vibrating handles Audiometric measurement A computer-based system (Diagnostic audiometer model AD229) with an audiometric headset (TDH39) was used to measure each subject s hearing threshold before and after exposure. The first measurement on each subject s hearing was carried out on both ears. Before each exposure, each subject s hearing on the right ear was measured. The fixed frequency von Békésy method (up-and-down method) with pure tone for 30 s per frequency was used for all hearing measurements. Mean value for each 30 s per frequency tracing was used as hearing the threshold (Miyakita et al. 1987; Young and Harbert 1990). Each subject was given time to examine the test method before the audiometric test. HAV and noise exposure Noise level was set to 99 db(a) inside each headphone, a level that corresponds to 8 h of equivalent noise exposure at 85 db(a), according to ISO 1999 (1990). The HAV level was 6.7 m/s 2 (rms) in a vertical direction, which corresponds to an eight-hour equivalent acceleration exposure, A(8), of 1.4 m/s 2 according to ISO 5349-1 (2001). The vibration level was controlled for using an accelerometer (Brüel and Kjær 4384) attached to the vibrating handle and a level meter (Brüel and Kjær 2513). A sound level meter (Brüel and Kjær 2209) was used for noise level control. Vibrations were recorded using an Instant replay 360 system from an angular grinder (Hitachi G23UB, rpm 6,600) by an accelerometer (Brüel and Kjær 4368) and The TTS in hearing in db was calculated by subtracting the subject s hearing threshold before and after different exposures separately for each test frequency 1, 4, and 8 khz. The data were analyzed using PASW statistics software version 18.0 (SPSS Inc., Chicago, Illinois). Q-Q plots and histograms were used to check for normality for each post-exposure measurement of hearing. The TTS in hearing after combined noise and HAV exposure was compared to exposure to noise or HAV alone using the paired sample t test for each of the eight hearing measurements after the exposures. One sample t test was used to test statistically any possible change in the hearing threshold after HAV exposure. All eight hearing measurements conducted after each exposure were used to calculate the TTS in hearing for each subject and were also used in repeated measures ANOVA. This method was used to analyze the effect of exposure and time and the interaction between them on TTS in hearing. Mauchly s sphericity test was used. If sphericity could not be assumed using repeated measurements, the Huynh Feldt correction 120 100 80 60 40 20 Noise Vibration 0 1 10 100 1000 10000 Frequency (Hz) Fig. 2 Spectra for noise and hand-arm vibration from the angular grinder as a function of the frequency 123

Int Arch Occup Environ Health was chosen. Gender differences were studied using independent t test and repeated measurements ANOVA with all eight measurements of hearing after the exposures. Holm s method was used for paired, independent, and one sample t test to adjust the p values to compensate for multiple testing (Holm 1979). A p value of \0.05 was considered significant. Results Pre-exposure The mean hearing thresholds for the subjects varied between -1.6 and -3.2 dbhl (Table 1) for the different test frequencies 1, 4, and 8 khz. Negative values indicated that the subjects had a hearing threshold better than 0 dbhl. There were no significant differences in hearing threshold between males and females for any of the tested frequencies (0.5 \ p \ 0.9). Post-exposure Table 2 showed the calculated mean TTS in hearing for the different exposure conditions presented in relation to the time after exposure for the various test frequencies. Exposure to HAV produces a small TTS in hearing for all three test frequencies, which changes marginally over time after the exposure. Regarding noise and combined exposures, the TTS in hearing is most pronounced for the test frequencies of 4 and 8 khz and shows a decline over time after the exposures. Statistical analyses (Table 3) showed that for 1 khz, there was a significant difference in TTS in hearing between the exposure to noise and HAV. At 4 and 8 khz, combined exposure (HAV and noise) and noise alone created significantly (p \ 0.01) greater TTS in hearing than exposure to HAV alone. For 4 and 8 khz, the TTS in hearing changed significantly differently after combined or noise exposure compared to after HAV exposure (p \ 0.001). There were no significant differences between combined exposure and noise exposure for any test frequency (0.11 \ p \ 0.18). Table 3 also showed the interaction between exposures and time. The TTS in hearing decreased significantly differently after combined or noise exposure compared to exposure to HAV. The TTS in hearing did not decrease significantly differently after noise exposure compared to combined exposure. The Time Factor in Table 3 was significant after combined or noise only exposure meaning that the TTS in hearing decreased after exposure. The hearing threshold did not change significantly (0.66 \ p \ 0.97) after HAV exposure for any test frequency. Gender The statistical analyses in Table 4 showed that combined exposure did not produce any significant differences in TTS in hearing in relation to gender for any of the investigated test frequencies (0.15 \ p \ 0.26). The interaction between time and gender showed that the TTS in hearing did not decrease significantly differently in relation to gender (0.06 \ p \ 0.29). After combined or noise only exposure for both genders, the TTS in hearing decreased with time. At 4 and 8 khz after noise exposure, there were no significant differences in the TTS in hearing in relation to gender. Since the interaction between time and gender was not significant, there were no differences in recovery over time in relation to gender. At 1 khz, the interaction between gender and time was significant, but the gender factor alone was not significant. Analysis of the TTS in hearing after exposure to HAV showed no significant changes over time. This could also be seen in Table 2, where the calculated mean TTS after HAV exposure was almost 0 and did not change much over time. Although the gender factor was significantly different in relation to gender at 1 and 8 khz, the interaction between time and gender showed no difference, i.e., no significant difference in decrease in the TTS in hearing in relation to gender. Discussion The TTS in hearing has been used in earlier experimental studies to measure the different effects of noise, HAV, or a combination of these (Miyakita et al. 1987; Zhu et al. 1997). TTS is measured because it is the most observable Table 1 Calculated mean threshold level (dbhl) of hearing before the exposures for test frequencies of 1, 4, and 8 khz, for all subjects as well as for men and women separately Frequency 1 khz 4 khz 8 khz Male -1.9 (-3.4 to -0.5) -2.5 (-4.9 to -0.2) -3.1 (-4.4 to -1.8) Female -1.2 (-2.7 to -0.4) -3.2 (-4.6 to -1.7) -3.2 (-4.5 to -1.9) Total -1.6 (-2.6 to -0.5) -2.8 (-4.2 to -1.5) -3.2 (-4.1 to -2.3) A total of 22 (11 women/men) subjects. The 95% confidence interval is given in parentheses 123

Int Arch Occup Environ Health Table 2 Calculated mean temporary threshold shift (TTS) in db after the different exposures and for each test frequency (1, 4, and 8 khz, respectively) measured at different times (min) after the exposures Frequency Time after exposure (min) 0.5 3 5.3 8 10.5 13 20 30 1 khz NV 0.91 (3.4) 0.45 (2.7) 0.00 (3.2) -0.45 (3.6) -0.77 (3.6) -1.0 (2.9) -1.3 (3.0) -1.2 (2.2) N 3.2 (4.0) 2.4 (2.9) 1.2 (2.7) 1.1 (3.4) 0.050 (3.6) 0.45 (3.7) -0.73 (3.4) -0.68 (3.7) V -1.1 (2.6) -1.0 (2.8) -0.77 (2.8) -1.4 (2.8) -0.68 (3.0) -1.3 (2.5) -1.1 (3.0) -1.0 (2.4) 4 khz NV 20 (9.5) 15 (6.0) 14 (5.9) 13 (7.0) 11 (5.4) 10 (5.3) 8.4 (5.3) 6.1 (5.6) N 18 (8.8) 16 (5.9) 12 (6.8) 10 (6.3) 9.7 (8.1) 8.7 (6.8) 7.7 (7.0) 5.3 (6.3) V 0.35 (2.2) 0.77 (2.3) 0.91 (2.7) 0.91 (2.8) 1.1 (2.6) 0.86 (2.4) 0.82 (2.9) 0.81 (3.8) 8 khz NV 20 (12) 18 (9.4) 17 (9.9) 16 (10) 15 (8.0) 15 (8.6) 13 (7.3) 9.5 (6.9) N 19 (9.3) 19 (7.8) 17 (7.6) 14 (7.0) 12 (6.3) 10 (6.7) 6.5 (5.6) -3.1 (3.9) V 2.0 (3.8) 1.8 (4.8) 1.6 (4.5) 1.6 (4.5) 1.8 (4.0) 2.7 (5.1) 2.9 (4.4) 2.5 (5.0) The standard deviation is given in the parentheses NV Noise and HAV Combined, N Noise, V HAV Table 3 Calculated differences (p values; Repeated measures ANOVA) for the temporary threshold shift in hearing (TTS) between the different exposure conditions in relation to the time after exposure for the different frequencies investigated as well as the interaction between exposure and time Exposure Factor 1 khz 4 khz 8 khz NV, N, V Exposure 0.158 \0.001 \0.001 Time \0.001 \0.001 \0.001 Exp 9 Time 0.001 \0.001 \0.001 NV, N Exposure 0.114 0.123 0.179 Time \0.001 \0.001 \0.001 Exp 9 Time 0.219 0.347 0.254 NV, V Exposure 0.289 \0.001 \0.001 Time 0.003 \0.001 0.001 Exp 9 Time 0.002 \0.001 \0.001 N, V Exposure 0.041 \0.001 \0.001 Time \0.001 \0.001 \0.001 Exp 9 Time \0.001 \0.001 \0.001 NV Time \0.001 \0.001 \0.001 N Time \0.001 \0.001 \0.001 V Time 0.664 0.971 0.764 NV Noise and HAV Combined, N Noise, V HAV and practical measure of noise effect on hearing acuity (Quaranta et al. 1998). Our results show that after exposure to noise or the combined exposure to noise and HAV, the TTS in hearing is large and has a long recovery period for 4 and 8 khz. For 1 khz, the TTS in hearing is small and has a short recovery period. We found that for the measured thresholds at 1 khz, there was little, or no, difference in the calculated mean Table 4 Calculated differences (p values; repeated measures ANOVA) for the temporary threshold shift in hearing (TTS) between men and women, in relation to the time after exposure for the different frequencies investigated and exposure conditions as well as the interaction between gender and time Exposure Subject effects Factor 1 khz 4 khz 8 khz NV Within Time 0.003 \0.001 \0.001 Within Time 9 Gender 0.243 0.057 0.292 Between Gender 0.178 0.259 0.150 N Within Time \0.001 \0.001 \0.001 Within Time 9 Gender 0.016 0.874 0.161 Between Gender 0.924 0.459 0.522 V Within Time 0.694 0.929 0.625 Within Time 9 Gender 0.979 0.623 0.815 Between Gender 0.025 0.902 0.018 NV Noise and HAV Combined, N Noise, V HAV TTS in hearing between the three exposures (Table 2). The TTS in hearing differs significantly after noise exposure compared to HAV only exposure (Table 3) although the difference in calculated mean TTS in hearing is very small (Table 2). For the measured thresholds at 4 and 8 khz, the TTS in hearing is significantly greater after the noise and combined exposure compared to HAV only exposure. These findings for 4 and 8 khz are in agreement with an earlier study (Zhu et al. 1997). Moreover, we found no significant differences in the TTS in hearing between noise exposure and combined exposure, a finding that agrees with the Miyakita et al. (1987) study, but not with the study carried out by Zhu et al. (1997). 123

Int Arch Occup Environ Health There are, of course, several possible explanations for the discrepancies between this study and the studies conducted by Miyakita et al. (1987) and Zhu et al. (1997). Both the Miyakita et al. (1987) study and our study used authentic exposure. In our case, the exposure was recorded from an actual running angular grinder for Miyakita et al. (1987) used an operating chainsaw during their experiment. Zhu et al. (1997), in contrast, used white noise and a sinusoidal vibration of 60 Hz. Moreover, the levels of exposure used differed between 90 db(a) and 105 db(a), and the frequency-weighted acceleration between 4.0 m/s 2 and 7.7 m/s 2. Furthermore, the exposure duration and exposure pattern varied. While in our study we used 20 min of continuous exposure before measuring the TTS in hearing, both Miyakita et al. (1987) and Zhu et al. (1997) used interrupted exposures of 5 times 2 min and 5 times 3 min, respectively. The number of subjects was approximately the same in all three studies. However, in our study we had an equal mix of genders. This could have affected the results, but studies on the effect of noise and combined exposures on TTS in relation to gender have produced inconsistent results (Loeb and Fletcher 1963; Hori et al. 1993; Smitley and Rintelmann 1971; Ward et al. 1959) showing higher, lower, and no susceptibility due to differences in gender. In the two earlier studies by Miyakita et al. (1987) and Zhu et al. (1997), one criterion for inclusion was that the subject should have had no occupational exposure to noise and vibration. In our study, the subjects had have a hearing threshold better than 10 db to be recruited. The conducting subjects in our study had very good hearing thresholds at the tested frequencies (-1.6 and -3.2 dbhl). However, they were not examined by means of otoscope in order to exclude any lesions of the outer and middle (ear-drums) sections of the ears. We also required subjects to have normal thresholds for thermal and tactile perception, both in the median and in ulnar nerves of their hands, since reduced thresholds could reduce the activation of the sympathetic nervous system. These criteria could of course have eliminated subjects who were more vulnerable to noise and vibration exposures. The results from our study indicate that combined exposure to noise and vibration does not produce different hearing thresholds compared to only noise exposure. However, the subjects were young and healthy with no history of regular occupational exposure to noise and vibration. The subjects were also recruited using a strict protocol for selecting subjects with very good hearing thresholds and no disorders of the peripheral nerves. Therefore, caution should be exercised when generalizing the results to the elderly who are exposed to noise and vibration or unhealthy subjects. Furthermore, the relation between measurements of the TTS produced by noise combined with vibration exposure and the PTS in hearing among noise/vibration-exposed workers is not fully known (Pyykkö et al. 1987; Starck et al. 1990). However, studies on subjects who have developed vibration-related disorders from HAV show an increased risk of a PTS in hearing (Iki et al. 1989; Pyykkö et al. 1981). One interpretation of our results could be that the effect noticed in combined exposure is due to the situation that the use of hand-held vibrating tools itself produces a high exposure to noise, especially if the tools used produce impacts. Acknowledgments The financial support of the AFA Insurance (Project 2007-0104) is gratefully acknowledged. Conflict of interest of interest. References The authors declare that they have no conflicts Hodgkinson L, Prasher D (2006) Effects of industrial solvents on hearing and balance: a review. Noise Health 8:114 133 Holm S (1979) A simple sequentially rejective multiple test procedure. Scand J Statist 6:65 70 Hori C, Nakashima K, Sato H (1993) A study on sex differences in temporary threshold shift (TTS) considering the menstrual cycle of women. J Hum Ergol (Tokyo) 22:131 139 Iki M, Kurumatani N, Satoh M, Matsuura F, Arai T, Ogata A et al (1989) Hearing of forest workers with vibration-induced white finger: a five-year follow-up. Int Arch Occup Environ Health 61:437 442 ISO 1999 (1990) Acoustics: determination of occupational noise exposure and estimation of noise-induced hearing impairment, International Organization for Standardization, Geneva ISO 8253-1 (1998) Acoustics audiometric test methods, International institution for standardization, Geneva ISO 13091-1 (2001) Mechanical vibration Vibrotactile perception thresholds for the assessment of nerve dysfunction. International organization for standardization, Geneva ISO 5349-1 (2001) Mechanical vibration measurement and evaluation of human exposure to hand-transmitted vibration, international organization for standardization, Geneva Kryter KD, Ward WD, Miller JD, Eldredge DH (1966) Hazardous exposure to intermittent and steady-state noise. J Acoust Soc Am 39:451 464 Lee CA, Mistry D, Uppal S, Coatesworth AP (2005) Otologic side effects of drugs. J Laryngol Otol 119:267 271. doi: 10.1258/0022215054020485 Loeb M, Fletcher JL (1963) Temporary threshold shift for normal subjects as a function of age and sex. Rep US Army Med Res Lab 3:1 9 Miyakita T, Miura H, Futatsuka M (1987) An experimental study of the physiological effects of chain saw operation. Br J Ind Med 44:41 46 Pyykkö I, Starck J, Farkkila M, Hoikkala M, Korhonen O, Nurminen M (1981) Hand-arm vibration in the aetiology of hearing loss in lumberjacks. Br J Ind Med 38:281 289 Pyykkö I, Pekkarinen J, Starck J (1987) Sensory-neural hearing loss during combined noise and vibration exposure. An analysis of risk factors. Int Arch Occup Environ Health 59:439 454 123

Int Arch Occup Environ Health Quaranta A, Portalatini P, Henderson D (1998) Temporary and permanent threshold shift: an overview. Scand Audiol Suppl 48:75 86 Smitley EK, Rintelmann WF (1971) Continuous versus intermittent exposure to rock and roll music. Effect upon temporary threshold shift. Arch Environ Health 22:413 420 Starck J, Pekkarinen J, Pyykko I (1990) Vibration as a contributing factor for the sensory-neural hearing loss in hand-arm vibration. In: Okada A, Taylor W, Dupuis H (eds) Hand-arm vibration. Kyoei Press, Japan, pp 23 27 Ward WD, Glorig A, Sklar DL (1959) Susceptibility and Sex. J Acoust Soc Am 31:1138 Young IM, Harbert F (1990) Time-intensity relations in Bekesy audiometry. Yonsei Med J 31:197 204 Zhu S, Sakakibara H, Yamada S (1997) Combined effects of handarm vibration and noise on temporary threshold shifts of hearing in healthy subjects. Int Arch Occup Environ Health 69:433 436 123

A follow-up study of welders exposure to vibration in a heavy engineering production workshop by Lage Burström, Mats Hagberg, Ingrid Liljeind, Ronnie Lundström, Tohr Nilsson, Hans Pettersson and Jens Wahlström reprinted from Journal of LOW FREQUENCY NOISE, VIBRATION AND ACTIVE CONTROL VOLUME 29 NUMBER 1 2010 MULTI-SCIENCE PUBLISHING COMPANY LTD.

JOURNAL OF LOW FREQUENCY NOISE, VIBRATION AND ACTIVE CONTROL Pages 33 39 A follow-up study of welders exposure to vibration in a heavy engineering production workshop Lage Burström 1, Mats Hagberg 2, Ingrid Liljeind 1, Ronnie Lundström 3, Tohr Nilsson 4, Hans Pettersson 1 and Jens Wahlström 3 1 Umeå University, Department of Public Health & Clinical Medicine, Occupational and Environmental Medicine, SE-901 87 Umeå, Sweden. 2 Sahlgrenska University Hospital, Department of Occupational and Environmental Medicine, Box 414, SE-405 30 Gothenburg, Sweden. 3 University Hospital of Northern Sweden, Department of Biomedical Engineering and Informatics, SE-901 85 Umeå, Sweden. 4 Sundsvall Hospital, Department of Occupational and Environmental Medicine, Sundsvall, SE-851 86 Sweden. E-mail: lage.burstrom@envmed.umu.se Received 25th February 2010 ABSTRACT Manual work involving vibrating power tools is associated with symptoms that include vascular, neurological and musculoskeletal disorders. This study examines the vibration exposure of welders to determine the change between 1987 and 2008. Vibration measurements on handheld tools were used to evaluate the acceleration and the daily exposure time was determined by subjective rating. From these data, the 8-hour equivalent vibration exposure A(8), has been calculated. During the period, the A(8) decreased from 3.9 m/s 2 to 1.9 m/s 2. It was concluded that this decrease is the result of fewer vibrating tools and a decrease in daily exposure time. Although the daily vibration exposure has decreased over the study time, for some welders the daily vibration exposure A(8) is still above the action value set by the EU directive on vibration. This means more effort should be spent to decrease vibration exposure. Keywords: Vibration, exposure, hand-arm, measurement, welders. 1. INTRODUCTION Welding joins materials by melting the work pieces and adding a filler material to form a pool of molten material (the weld pool) that cools to become a strong joint. Manual welding is common in the industry, especially in heavy engineering production workshops. Removal of welding spatters generated during the welding process as well as surface finishing often includes using percussive tools such as chipping hammers and rotary tools such as grinders. Manual work involving vibrating power tools has been associated with various symptoms, collectively named "hand-arm vibration syndrome". These symptoms, which include vascular, neurological and musculoskeletal disorders, have also been recognised as an important preventable occupational disease. For the measurements and evaluation of hand-transmitted vibration, international standards are used [1, 2]. The vibration is measured in the three directions of an orthogonal coordinate system and evaluated as the vector sum of these directions. The measured vibration acceleration is frequency-weighted on the assumption that the harmful effects of acceleration depend on the vibration frequency. Since the detrimental effects of Vol. 29 No. 1 2010 33

Follow-up of vibration load among welders vibration exposure also depend on the daily exposure time, exposure is assessed by calculating the daily energy equivalent exposure value normalised to an eight hour reference period, A(8), of the frequency-weighted acceleration values (1). T i A ( 8) = A ( i ) 8 (1) where A(8) = 8-hour equivalent acceleration [m/s 2 ]; A(i) = vector sum of the measured acceleration [m/s 2 ] and T i = duration of the acceleration A(i) [h]. The introduction of the European Machinery Directive [3] in 1989 high-lighted the need for hand-held tools to be designed and constructed in such a way that risks resulting from vibration are reduced to the lowest level. Moreover, the implementation of the European directive for hand-arm vibration [4] emphasized the effects on health of vibration resulting from vibrating machinery. The directive sets two limits: one exposure action value and one exposure limit value. If the action value is exceeded, the employer should establish and implement a programme of technical and/or organisational measures to minimize the vibration exposure. The workers shall in no cases be exposed to vibration above the exposure limit value. For hand-arm vibration, the exposure limit value, A(8), is 5 m/s 2 and the action value, A(8), is 2.5 m/s 2. Welder s exposure to vibration from hand-held vibrating tools has been studied in several investigations [5, 6] and the results show that they could be exposed to both high acceleration magnitude as well as long daily exposure times. This has also resulted in a high prevalence of vibration related diseases [7, 8]. However, few studies have followed the change of the vibration exposure over time at the same work place with employees doing the same job. This investigation studies the vibration exposure for welders who work in a heavy engineering production workshop to determine the change over time. 2. METHODS The studied cohort has been described earlier [8, 9]. In this study only the welders in the cohort have been investigated. The group of welders was employed in the heavy engineering production workshop that makes paper and pulp-mill machinery. The work task consisted mainly of welding iron and stainless steel that requires working with several hand-held vibrating tools. The production is varied and each component is produced in small numbers. The study started in 1987 and was followed up in 1992, 1997 and 2008. The criteria for inclusion were workers employed as welders (job title criteria) and actually working as such (work criteria). In 1987, 64 welders were included and for the follow up 59, 50 and 33 welders, respectively. Of the 64 welders included in 1987, 18 were followed up in 2008 (Table I). The mean age for all the welders increased from 31.3 years in 1987 to 45.3 years in 2008 (Table I). The mean age increased by about 18 years for the welders that participated in 1987 and were followed up in 2008. Table I. Mean age of subjects (y) in the study population for the different investigation years. The standard deviations are in round brackets and the number of subjects followed up in square brackets. Year 1987 1992 1997 2008 1987 31.3 (9.2) [64] 34.3 (9.0) [42] 40.0 (8.9) [34] 49.2 (8.2) [18] 1992 32.1 (9.3) [59] 37.8 (8.9) [42] 47.0 (8.5) [23] 1997 37.3 (9.4) [50] 35.6 (8.5) [24] 2008 45.3 (9.3) [33] The assessment of vibration exposure was made under normal working conditions with standardised equipment and methods [10] by measuring the intensity of vibration on a random selection of the tools used by the welders in 34 JOURNAL OF LOW FREQUENCY NOISE, VIBRATION AND ACTIVE CONTROL

Lage Burström, Mats Hagberg, Ingrid Liljeind, Ronnie Lundström, Tohr Nilsson, Hans Pettersson and Jens Wahlström accordance with international standards [1, 2]. The total number of tools included in the study was 306 and during each investigation period the quantity varied between 50% and 90% of the total number of tools used in the workshop. For hand-held tools with two handles, measurements were made on both handles and the highest measured vibration intensity was used in the analysis. The subjective assessments of daily exposure time were collected by questionnaire during the health surveillances that were conducted at the same time as the technical measurements. In the questionnaire, welders were asked to estimate the amount of time (minutes per day) they were exposed to vibration while using the different types of hand-held vibrating tools during their last working day. One-way ANOVA and repeated measures one-way ANOVA were used to test for a difference between the means of the variables. A p-value of less than 0.05 was taken as significant. All statistical analyses were performed with the statistical software package SPSS version 16.0 (SPSS, Chicago, Illinois, USA). 3. RESULTS The results from the measurements of the frequency weighted acceleration (m/s 2 ) for the different hand-held tools and investigation years are presented in Table II. The category "Others" include drills, tappers and die grinders. Table II Measured mean value and standard deviation of the frequency-weighted acceleration (m/s 2 ) for the different tools and investigation years. The standard deviations are in round brackets and the number of tools in square brackets. Year Grinders Hammers Others 1987 5.8 (2.6) [20] 11.0 (2.7) [20] 4.0 (0.9) [5] 1992 6.5 (3.2) [36] 13.5 (3.6) [16] 4.6 (3.2) [5] 1997 5.3 (2.4) [52] 10.8 (4.4) [18] 6.5 (3.4) [6] 2008 4.5 (3.2) [94] 7.6 (1.5) [26] 5.9 (2.0) [8] The measured mean frequency-weighted acceleration for the grinders varied during the different investigation years between 4.5 and 5.8 m/s 2 and for the hammers between 7.6 and 13.5 m/s 2. For the tools in the category "Others" the mean acceleration varied between 4.0 and 6.5 m/s 2. Statistical analysis shows that the change in the frequency-weighted acceleration is significant (p<0.001) for the hammers at the follow-up in 2008 compared to the other years. The variations for the rest of the tools were not significant. Table III shows the mean subjective assessment of the daily exposure time for the different years investigated. For all the investigation periods, grinders were used between 32 and 62 min per day and hammers between 14 and 42 min per day. Tools within the category "Others" were used between 1 and 4 min per day. The total daily exposure time for all vibrating tools varies between 52 min to 108 min per day. The change in the exposure time is significant for the use of the hammers and for the total daily use of hand-held tools (p=0.026, p=0.042, respectively) when comparing the results from 1987 with 2008. The changes in the exposure time for the other tools are not significant. Table IV summaries the total daily exposure time for the different years investigated. For the welders included in 1987 the total daily exposure time has decreased from 108 min in 1987 to 30 min in 2008.. the total daily exposure time is significant lower for all the welders that were followed-up in 2008 (0.010<p<0.026). The other changes of the total exposure time over the years were not significant. Moreover, a more detailed statistical analysis shows that the nine welders who started their employment after 1997 have a significantly (p<0.001) longer total exposure time compared to the other welders (113.8 min vs. 28.4 min). Vol. 29 No. 1 2010 35

Follow-up of vibration load among welders Table III Mean value and standard deviation for subjective assessment of the daily exposure time (min) for different type of tools and investigation years. The standard deviations are in brackets Year Grinders Hammers Others Total 1987 61.8 (74.2) 42.4 (65.2) 3.1 (11.8) 107.6 (102.8) 1992 57.7 (59.8) 27.6 (42.7) 2.8 (11.0) 88.0 (86.6) 1997 38.3 (41.7) 14.0 (22.4) 1.9 (2.1) 54.1 (54.6) 2008 31.5 (41.5) 19.2 (38.3) 0.9 (3.2) 51.7 (78.4) Table IV Mean value and standard deviation for subjective assessment of the total daily exposure time (min) and investigation periods. The standard deviations are in brackets. Year 1987 1992 1997 2008 1987 107.6 (102.8) 79.1 (84.0) 54.3 (55.6) 29.7 (19.2) 1992 88.0 (86.6) 58.5 (57.1) 29.0 (19.4) 1997 54.1 (54.6) 28.4 (19.1) 2008 51.7 (78.4) Table V shows the total frequency-weighted equivalent acceleration for eight hours calculated in accordance with the EU directive. The frequency-weighted energy-equivalent acceleration for eight hours decreased from 3.9 m/s 2 in 1987 to 1.9 m/s 2 in 2008. The decrease is also significant (p<0.001) and correspond to a decrease by 9.5% per year. For the welders included in 1987, the A(8) changed from 3.9 m/s 2 in 1987 and 1992 to 2.4 m/s 2 in 1997 and 1.4 m/s 2 in 2008. There was no significant difference in the A(8) between 1987 and 1992. For the other follow-ups in 1997 and 2008, all changes were significant (0.001<p<0.029). The nine welders who started their employment after 1997 have a statistically (p<0.001) higher A(8) compared with the other welders (3.2 m/s 2 vs. 1.4 m/s 2 ). Table V Mean value (M) and standard deviation (SD) for calculated 8-hour energy-equivalent acceleration (m/s 2 ) for the different investigation periods. Year 1987 1992 1997 2008 1987 3.9 (1.6) 3.9 (1.7) 2.4 (1.0) 1.4 (0.4) 1992 4.0 (1.6) 2.5 (1.0) 1.4 (0.4) 1997 2.4 (1.0) 1.4 (0.4) 2008 1.9 (1.1) 4. DISCUSSION This study shows that the mean vibration load among the group of studied welders decreased over a 21-year period, both expressed as the vibration acceleration and exposure time. The welders exposure to vibration is primarily from the use of grinders and hammers. These two types of tools correspond to 96 to 98% of the total daily use of hand-held tools. During this period, the tool acceleration decreased for more than 20% for the grinders and 30% for the hammers. At the same time, the exposure times for these tools decreased by about 50%. The decrease in both the magnitude of the acceleration and the exposure time resulted in the welders vibration load decreasing by about 50%, described as the 8-hour energy equivalent A(8). 36 JOURNAL OF LOW FREQUENCY NOISE, VIBRATION AND ACTIVE CONTROL

Lage Burström, Mats Hagberg, Ingrid Liljeind, Ronnie Lundström, Tohr Nilsson, Hans Pettersson and Jens Wahlström The measured vibration magnitude show that grinders have a larger variability in measured magnitudes compared to hammers. Grinders include many models and tools from different manufacturers while the tools in the hammer category are more or less of the same type. Moreover, the grinders are equipped with a variety of different grinding wheels (flap discs, cut-off wheels, grindstones, etc.) that influence the vibration level. Although the average magnitude measured for each tool is representative of the typical use of the hand-held power tool, the result depends on the welder s actual task. Furthermore, the measurement of the tool acceleration was conducted on a random selection of tools used by the welders. Therefore, the results should be interpreted carefully since the vibration magnitude only reflects the situation during the measurements. The changes in the vibration magnitudes over time depend on the chosen tools. During the investigation, many tools were replaced with newer ones. For every 5- year period, between 50 and 65% of the tools were replaced and for every 10-year period 60 to 90% were replaced. Over the 21 years, 95% of the tools were replaced. The most frequently used tools used by welders grinders and hammers were replaced. Tools such as drills and thread tappers are less frequently replaced. The result shows that the vibration magnitudes were more or less the same for the measurements conducted in 1987 and 1992. After that the magnitudes for the grinders and hammers decreased. However, the decrease was only significant for the hammers at the follow-up in 2008 compared to the other years. In this investigation, the welders daily exposure times were subjectively assessed and the large standard deviation indicates that the precision of the average duration of exposure is rather low. The large variation is probably due to the diversified work as well as the question used for assessing the exposure time. Palmer et al found that worker estimates of the durations of daily exposures to handtransmitted vibration tend to be greater than found by observation of tool users [11]. However, an earlier investigation of the same group of welders showed that the same tendency of large individual deviation and evidence for a good agreement between self-reported and measured exposure time on a group basis [12]. During the 21-year period, the frequency-weighted energy-equivalent acceleration for a period of 8-hours decreased from 3.9 m/s 2 in1987 to 1.9 m/s 2 2008. This decrease is also statistically significant.. The relation establish for calculating A(8) (Equation 1) suggests that vibration magnitude has greater impact than the duration of exposure. For example, if the magnitude is halved, the A(8) is reduced by 50%; if the exposure time is halved, the A(8) is reduced by only 30%. For the welders in this study, the halved exposure time had a greater impact on the A(8) compared to the corresponding reduction (about 20%) of the vibration magnitudes. Looking at the welders who have been followed since 1987, the A(8) decreased by 65% in 2008 compared with the average decrease of 50%. This difference could be interpreted that welders with a longer period of employment have received work that includes less use of vibrating hand-held tools. However, this tendency is only statistically significant at the follow-up in 2008. Therefore, the changes seem to more relate to the job title than to the period of employment. The general change in "job title" exposure has also been found in others studies. For example, Suttinen et al. [13] found that among forestry workers both the exposure time and the vibration acceleration decreased over a 20-year period resulting in a decrease of the A(8) from 5.1 m/s 2 to 1.5 m/s 2. The decreased vibration load, expressed as the A98), among the welders could be because the investigation increased the company s awareness of the exposure to vibration. Over the years, the company have replaced many tools with newer less vibrating ones as well as introduced regular maintenance for the hand-held tools and settled purchase routines for both grinding discs and chisels. Moreover, better design of the products has lead to less surface finishing and changes in the welding process have reduced the need for chiselling of the weld pool. Furthermore, the results also indicate that the manufacturers of the tools have succeeded in meeting the demands to reduce the vibration emissions of tools. This trend has also been noticed in others studies [14]. Vol. 29 No. 1 2010 37

Follow-up of vibration load among welders For the welders the present mean 8-hour energy equivalent acceleration was found to be 1.9 m/s 2. This value is below the action limit value stated in the EU directive on vibration but the variation is large and some workers are still exposed to A(8) that exceeds the action value. For this reason more effort should be spent to decrease vibration exposure. Since the use of the hammers has 30% greater impact on the A(8) compared to the grinders, it is most important to reduce the use of the hammers. 5. CONCLUSIONS Regular surveillances of exposure and health have significantly reduced the welders exposure to vibration and thereby their risk for vibration related injuries. Notwithstanding that the vibration exposure among the welders has decreased over the study time, for some welders the daily vibration exposure A(8) is still above the exposure action value. For this reason, more effort should be spent to decrease vibration exposure. ACKNOWLEDGEMENT The financial support of the Swedish Council for Work Life Research (Project 2006-0968) and AFA Insurance (Project 2007-0104) is gratefully acknowledged. REFERENCES [1] International Organization for Standardization ISO 5349-1, Mechanical vibration Measurement and evaluation of human exposure to handtransmitted vibration part 1: general guidelines, 2001. [2] International Organization for Standardization ISO 5349-2, Mechanical vibration Measurement and evaluation of human exposure to handtransmitted vibration part 2: practical guidance for measurement at the workplace, 2001. [3] European Council, Directive 89/392/EEC of 14 June 1989 on the approximation of the laws of the Member States relating to machinery. Off J Europe Communities L 183 (1989) 9-32. [4] European Council, Directive 2002/44/EC of the European parliament and of the Council of 25 June 2002 on the minimum health and safety requirements regarding the exposure of worker to the risks rising from physical agents (vibration) (sixteenth individual Directive within the meaning of Article 16(1) of Directive 89/391/EEC). Off J Europe Communities; L 177 (2002) 13-19. [5] Behrens, V. Taylor, W. Wilcox, T. Miday, R Spaeth, S. Burg, J. Wasserman, D. Reynolds, D. Doyle, T. Carlson, W. Smith, R. Samueloff, S. Howie, G. and Rappaport, M., Vibration syndrome in chipping and grinding workers. Journal of Occupational Medicine, 1984, 26(10, 765-788. [6] Bovenzi, M. Petronio, L. And DiMarino, F, Epidemiological survey of shipyard workers exposed to hand-arm vibration. International Archives of Occupational and Environmental Health, 1980, 46, 251-266. [7] McGeoch, K. And Gilmour, H., Cross sectional study of a workforce exposed to hand-arm vibration: with objective tests and the Stockholm workshop scales. Occupational and Environmental Medicine, 2000, 57, 35-42. [8] Nilsson, T. Burström, L. And Hagberg, M., Risk assessment of vibration exposure and white finger among platers. International Archives of Occupational and Environmental Health, 1989, 61(7), 473-481. 38 JOURNAL OF LOW FREQUENCY NOISE, VIBRATION AND ACTIVE CONTROL

Lage Burström, Mats Hagberg, Ingrid Liljeind, Ronnie Lundström, Tohr Nilsson, Hans Pettersson and Jens Wahlström [9] Hagberg, M. Burström, L. Lundström, R. And Nilsson, T., Incidence of Raynaud s phenomenon in relation to hand-arm vibration exposure among male workers at an engineering plant a cohort study. Journal of Occupational Medicine and Toxicology, 2008, 16, 3-13. [10] Burström, L. Pettersson, H. Nilsson, T. Wahlström, J. Liljelind, I. Lundström, R. And Hagberg, M., A 21-year follow-up of the vibration load among workers in a heavy engineering production workshop. In Dahlhiem. S., ed Proceeding of the 2nd International Conference on Human Vibration Exposure, measurements and tests, Centek, Luleå, 2009, 115-118. [11] Palmer, K.T. Haward, B. Griffin, M.J. Bendall, H. And Coggon, D., Validity of self reported occupational exposure to hand-transmitted and whole body vibration. Occupational and Environmental Medicine, 2000, 57, 237-241. [12] Burström, L. Lundström, R. Hagberg, M. and Nilsson, T., Comparison of different measures for hand-arm vibration exposure. Safety Science, 1998, 28, 3-14. [13] Sutinen, P. Toppila, E. Starck, J. Brammer, A. Zou, J. and Pyykkö, I., Handarm vibration syndrome with use of anti-vibration chain saws: 19-year followup study of forestry workers. International Archives of Occupational and Environmental Health, 2006, 79(8), 665-671. [14] Brereton, P. and Hewitt, S., Fulfilling the hand-arm vibration requirements of the EU Machinery Directive. Journal of Low Frequency Noise, Vibration and Active Control, 2000, 19(3), 115-121. Vol. 29 No. 1 2010 39