Ergonomic Risk Factors for the Wrists of Hairdressers. Taiwan. Taiwan

Transcription

1 Ergonomic Risk Factors for the Wrists of Hairdressers Hsieh-Ching Chen 1, Cha-Mei Chang 1, Yung-Ping Liu 1, Chih-Yong Chen 2 1 Department of Industrial Engineering and Management, Chaoyang University of Technology, Taiwan 2 Institute of Occupational Safety and Health, Council of Labor Affairs, Executive Yuan, Taiwan Address correspondence to: Hsieh-Ching Chen Department of Industrial Engineering and Management Chaoyang University of Technology No. 168 Jifong E. Rd., Wufong Township Taichung County 41349; Taiwan Tel: ext 4255 Fax: hcchen@cyut.edu.tw

2 Abstract This study utilized a portable data logger to measure the wrist angles and forearm flexor and extensor electromyography (EMG) of 21 hairstylists. The hairstylists were divided into two groups, one with 11 barbers (9 males and 2 females) specializing in men s hairdressing, and one with 10 hairdressers (2 males and 8 females) specializing in women s hairdressing. The standard haircut task was divided into three subtasks: hair cutting, washing and blow-drying. The mechanical exposures of the overall task and subtasks were quantified to compare how subtasks, occupational groups, and gender groups differ. Experimental results show that the average time to finish a woman s haircut (51.4 min) is significantly longer than that for a man s haircut (35.6 min) (p<0.005). Female hairstylists had significantly greater EMG activity than male hairstylists did (p<0.001). The non-dominant hands of hairdressers have significantly higher overall wrist velocity than those of barbers (p<0.005). Analytical results suggest that the relatively higher force exertion and wrist velocity of female hairstylists combined with prolonged exposure may account for the higher rate of hand/wrist pain in female hairdressers than in male barbers. Keywords: Workload, musculoskeletal disorder, ergonomic risk factors 1. Introduction Musculoskeletal discomfort, pain or injury among hairdressers and barbers is common and results in reduced job performance and productivity, increased time off work, and even early retirement (New Zealand DOL, 2007). Following asthma and hand eczema, musculoskeletal disorders are the primary cause of premature departure from the hairdresser profession in Finland (Leino et al., 1999). In the US, hairdressing is a high-risk occupation associated with back pain that results in morbidity and reduced production (Guo, 1995). English et al. (1995) examined 221 UK females with upper-limb soft-tissue disorders seeking treatment at orthopedic clinics and identified that hairdressers have significantly more 1

3 shoulder and ganglia injuries than other occupations. Kang et al. (1999) evaluated work-related symptom prevalence among hairdressers. The exposed group comprised 184 hair salon employees in 6 districts of Pusan city, and non-exposed group comprised 119 people living in the same areas. They reported the prevalence of musculoskeletal symptoms among Korean hairdressers was neck(59.9%) shoulder(76.6%), upper back(41.2%), lower 刪 除 : were back (72.2%), arm and elbow(31.3%), wrist(44.2%), finger(35.0%), leg(71.1%). The highest age-adjusted odd ratio (4.83) for exposed group compare to non-exposed group was for fingers. According to data from Taiwan s Bureau of Labor Statistics, more than 50,000 workers were employed at salons or barbershops as hairdressers and barbers (Taiwan DGBAS, 2001). Although the total number of hairstylists is limited compare to other industry populations, the hairdressing industry accounted for approximately 24% of compensation cases for work-related hand-wrist morbidity between Jan 2003 and June 2006 (Taiwan IOSH, 2006). Furthermore, the number of such insurance claims has increased annually since 1998, with most musculoskeletal disorder cases from professions associated with women s hairdressing. A previous study (Taiwan IOSH, 2004), which analyzed compensation cases in in the database generated by the Council of Labor Affairs, found that 260 new work-related cases were for hand-arm or neck-shoulder disorders. Among these cases, 22 cases (20 for females and 2 for males) were hairstylists. Undesirable postures and movements in activities such as hair cutting, combing, washing, and blow-drying must be eliminated to minimize risk of injury among hairdressers (New Zealand DOL, 2007). Few studies have focused on neck/shoulder problems of hairdressers (Hagberg, 1984; Nevala-Puranen et al., 1998). Arokoski et al. (1998) performed an intervention on 21 female hairdressers with neck/shoulder or back pain. After approximately 4 weeks of rehabilitation and vocational courses during 1 year, a positive 2

4 outcome was achieved; that is, physical strain and pain symptoms were reduced. Veiersted et al. (2008) quantitatively analyzed the biomechanical workload of the neck and shoulder region of female hairdressers before and after an intervention. Their study showed that hairdressers typically worked with their arms elevated at 60 for approximately 13% of total work time. Although strain injuries to wrists or elbows are major reasons why hairdressers leave the profession (Leino et al., 1999), knowledge regarding the mechanical exposures of hairdressers remains limited. This study assesses and compares the mechanical exposure of the wrists of barbers and hairdresser while performing regular haircutting tasks. 2. Method 2.1 Subjects Eleven barbers (9 males and 2 females) who specialized in men s hairdressing, and 10 hairdressers (2 males and 8 females) who specialized in women s hairdressing, were recruited from 2 barbershops and 3 hair salons in Taichung City, Taiwan. The sample size from each salon or barbershop was roughly 50 60% of the working population in the salons and barbershops. All subjects possessed professional certificates, a minimum of five years of experience, and less than two weeks sick leave due to musculoskeletal pain associated with any body part during the previous 12 months. Personal data of subjects, such as age, stature, experience, clinical records, work information, and discomfort in the musculoskeletal system, were collected by an experimenter who interviewed individual subjects. The age and years of experience was not significantly different between barbers (age, 31.7±2.4 yrs; experience, 13.2±3.7 yrs) and hairdressers (age, 28.4±4.4 yrs; experience, 10.5±5.3 yrs). However, barbers (height, 167.1± 5.6 cm; weight, 63.2±8.8 kg) were significantly taller and heavier than the hairdressers (height, 160.3±4.1 cm; 3

5 weight, 52.7±4.3 kg) (p<0.01, t-test). All subjects were right-handed and worked approximately 12 hours a day, 6 days a week. A barber on average handled 7 8 cases (customers) per day; each standard haircut took roughly 35 minutes. Periods between customers were rest periods. A hairdresser on average handled 6 7 customers per day; each standard haircut took approximately 50 minutes. The most common tasks during haircuts were hair cutting, washing and blow-drying. Among the barbers, only 1 complained about wrist pain and the others had only experienced minor soreness in the neck, shoulders, lower back, and lower legs due to working for extended periods. No barber had sought rehabilitation therapy or any alternative medical treatments for persistent pain. However, 50% of the hairdressers experienced pain in their dominant hand and/or wrist, and 30% of the hairdressers experienced pain in their non-dominant hand and/or wrist. About 50% of hairdressers have sought rehabilitation therapy or a particular Chinese medical treatment (acupuncture or naprapathy) for pain relief. 2.2 Experimental procedures To assess mechanical exposure associated with posture, forceful exertion and repetitive motion of hands, a data logger (Liu, 2006) and camcorder were synchronized to record the haircut process. Two biaxial goniometers (SX65, BioMetrics Ltd., UK) were employed to measure right and left dorsal-palmer flexion and radial-ulnar deviation of wrist angles. The goniometers were encased in saran wrap to prevent them from getting wet during hair washing. Four EMG electrodes with 20mm intra-distance bipolar surface electrodes (SX230, BioMetrics Ltd., UK; gain: 1000; band-pass: Hz; input impedance > Ω) were placed over the right and left flexor digitorum superficialis and extensor digitorum comunis of subjects to record relevant muscular activities. These muscles were chosen for their relevance when performing tasks, such as grasping a tool, cutting and washing hair, and for the ability 4

6 of researchers to position surface electrodes over muscle bellies. Before electrodes were placed, skin was cleaned with acetone and rubbed gently with fine abrasive paper. Each bipolar surface electrode was placed along the muscle with the electrode center at the point of recommended insertion of needle electrode (Delagi et al., 1994). Figure 1 shows the experimental setup for a barber cutting hair. Figure 1. Experimental setup of a barber cutting hair Subjects were interviewed before calibration thereby providing about 10 min for stabilizing EMG electrode impedance. The resting EMGs of each subject were recorded for 30 seconds immediately after the interview while subject relaxed and stood with arms hanging down. A series of calibrations were then conducted to obtain an individual reference for wrist angle and maximal voluntary capacity of each muscle. Angular measurements were calibrated in relation to individual maximal voluntary flexion-extension and radial-ulnar deviation of the wrist. All maximal voluntary contractions were performed for both arms while standing in order of right extensor, right flexor, left extensor, and left flexor. A 5-sec maximal voluntary muscle contraction on each muscle group was measured with 2-minute rest between each maximum attempt. To test maximal voluntary contractions of the wrist flexor/extensor muscles, subjects were asked to flex/extend their wrists maximally against an external load applied to their wrist and forearm while the elbow was at 90 flexion. The recorded EMG signals were later utilized to normalize the EMG signals recorded during task performance by expressing these signals as a percentage of maximal voluntary contraction (%MVC). For details of the calibration procedure, see Chen et al. (2005). All subjects were instructed in the standard operating procedure before each experiment. No pauses or rest periods were allowed during a task with the exception of answering a 5

7 personal phone call. Each subject carried out a standard haircut task on a randomly assigned customer. The standard task was first cutting hair, then washing, and last, blow-drying the hair. Task duration was defined as the time from when the subject put a bib on a customer to when the bib was taken off prior to payment. The tools used during the task were scissors, combs, hairdryers, razors and a clipper (barber). The barber chairs and styling chairs were all height-adjustable. A hairstylist s stool was allowed for hairdresser group according to custom. The basin type used backwash or bend over was based on customer preference. The logger was carried on a belt around the subject s waist while performing the haircut tasks. Bilateral EMG and wrist angles were recorded for each complete task. The logger analog-to-digital (A/D) interface converted the EMG and wrist angle signals with a sampling rate of 1000Hz per channel and stored the data on a compact flash memory card for further analyses. Recorded data were later downloaded to a personal computer and imported into Viewlog, a signal processing and analysis program, developed by the Institute of Occupational Safety and Health, Taiwan (Chen et al., 2005). 2.3 Data processing and statistical analysis For each measured task, Viewlog calculated the mean wrist angle and root mean square (RMS) value of EMG signals for intervals of 0.2s to characterize wrist movement and muscular activity. Each goniometer signal was first converted into an engineering unit of degree by applying calibration parameters obtained in individual maximal voluntary flexion-extension and radial-ulnar deviation tasks. Each EMG signal, including the EMG signals obtained during maximal voluntary contraction, subtracted individual rest level from the original RMS EMG. Each RMS EMG was then normalized to %MVC by dividing the magnitude of individual RMS EMG obtained in maximal voluntary contraction. The amplitude probability distribution function (APDF) was then analyzed for each calibrated wrist angle and EMG signals (Jonsson, 1982). Wrist angle during the test period was 6

8 characterized by rigorous positions, 10th and 90th percentiles of the APDF, and median (50th percentile) position. This study defines wrist angle of flexion and ulnar deviation as positive and the angle of extension and radial deviation as negative. That is, the 10th percentile of the APDF represents rigorous positions of wrist extension or radial deviation angle, whereas the 90th percentile of the APDF represents rigorous positions of wrist flexion or ulnar deviation angle. The 10th, 50th, and 90th percentiles of APDF of RMS EMG were used to describe low, median, and high muscular exertion levels, respectively. Mean angular velocity of each wrist angle was computed numerically by first differentiating wrist angle filtered by a fourth-order Butterworth digital filter with a 5Hz cut-off frequency, and then taking the RMS of differentiated result. Mean power frequency (MPF) was adopted as an index of wrist repetitiveness and was computed according to descriptions offered by Hansson et al. (1996) with angular data down-sampled to a 20Hz sampling rate. The down-sampling procedure was applied such that MPF computation was identical to those in other studies (Hansson et al., 1996, 2000; Åkesson et al., 1997; Arvidsson et al., 2003; Juul-Kristensen et al., 2001, 2002; Stål et al., 1999, 2000, 2003). An experimenter visually inspected the video recording and synchronized data to divide the overall task into the following subtasks: hair cutting, washing, and blow-drying. All parameters were reprocessed for each subtask. Task period was determined for all subjects and subtasks using video data analyzed by an experimenter. The SPSS version 15.0 for Windows was used for statistical analyses. Occupational and gender differences in overall tasks for all measurements (angle, EMG, velocity and MPF of wrist joints) were assessed using independent t-tests. Repeated-measures ANOVA was employed to compare differences between subtasks, sides, and occupational or gender groups. For repeated-measures ANOVA, occupation (barbers and hairdressers) or gender (males and females) was the between-subject factor, and subtask (cutting, washing and blow-drying) and 7

9 side (dominant and non-dominant) were within-subject factors. When applying repeated-measures ANOVA, either occupation or gender was taken as the only between-subject factor. A difference was considered significant at p< Experimental Results 3.1 Task duration Average duration of men s haircuts was 35.6min with 20.1, 9.2, and 6.3min spent in haircutting, washing, and blow-drying, respectively (Table 1). Mean duration of women s haircuts was approximately 15 minutes longer than that for men s haircuts (p<0.005, t-test); haircutting and blow-drying subtasks for females were 5min and 10min longer than those for males, respectively (p<0.001, t-test). Table 1. Overall and subtask durations (mean±sd) of a standard haircut (unit: min) 3.2 Wrist position For overall task, the hairdresser or the female group exhibited significantly greater wrist flexion in their non-dominant hands than the barber or the male group, respectively, for the median and 90th percentile (Table 2). Repeated-measures ANOVA results demonstrate that no significant within-subject effects exist for subtask and side, or between-subject effects exist for occupation or gender. Nevertheless, while using occupation as a between-subject factor, significant interactive effects of side occupation (p<0.001) and side subtask (p<0.005) in extension-flexion existed at the median and 90th percentile flexion (Table 2). Additionally, a significant and complex side subtask occupation interactive effect was observed under extension-flexion at the median and 10th percentile extension (p<0.001), and in radial-ulnar at the median (p<0.001) and 90th percentile ulnar (p<0.005) deviations. With gender as a 8

10 between-subject factor, only a slightly significant side subtask effect (p<0.005) existed in the median extension-flexion and in the 90th percentile ulnar deviation. For both occupational groups, subject bilateral range of motion (ROM), counted from the 10 90th percentile, was about 60 for wrist flexion-extension and 30 for radial-ulnar deviation. Table 2. Mean (SD) wrist positions of barbers (B, n=11) and hairdressers (H, n=10) during overall task and subtasks (unit: deg) 3.3 Muscle activity According to overall measured muscle activity, hairdressers exhibited greater non-dominant extensor EMGs at all 3 APDF levels (p<0.001, t-test) and greater dominant flexor EMG at the median APDF (p<0.005, t-test) than barbers did (Table 3). However, occupational groups did not significantly differ in EMGs, as indicated by repeated-measures ANOVA. Only significant subtask effects (p<0.001) were found for both extensor and flexor EMG at the median and 90th-percentile APDF, and significant side subtask and subtask occupation interactive effects (p<0.005) were identified for particular combinations of muscles and APDF levels (Table 3). Notably, EMG differences between gender groups were more prominent than occupational groups. When gender is applied as a between-subject factor, repeated-measures ANOVA results demonstrate that female subjects exhibited significantly greater EMGs than male subjects at all combinations of muscles and APDF levels (Table 4). Moreover, a significant subtask effect (p<0.001) was observed in all statistical tests, except that for the extensor muscle at the 10th-percentile APDF. This significance was demonstrated by the fact that both gender groups have the greatest and lowest flexor EMGs on both sides during hair washing and cutting, respectively, at the median and 90th-percentile APDF. Similarly, a greater non-dominant extensor EMG was observed for both gender groups during 9

11 blow-drying subtask than during cutting and washing subtasks at the median and 90th-percentile APDF. The greatest mean 90th percentile flexor EMG was observed during the washing subtask in both female subjects (dominant, 36.9%MVC; non-dominant, 33.8%MVC) and male subjects (dominant, 19.2%MVC; non-dominant, 12.8%MVC). The highest mean 90th percentile extensor EMG existed mostly during the blow-drying subtask with 15.2 and 43.5%MVC in non-dominant hands of male and female subjects, respectively, and 38.0%MVC in the dominant of female subjects. Table 3. Mean (SD) EMG activity of the wrist flexor and extensor muscles of barbers (B, n=11) and hairdressers (H, n=10) in the overall task and subtasks (unit, %MVC) Table 4. Mean (SD) EMG activity of the wrist flexor and extensor muscles of male subjects (M, n=11) and female subjects (F, n=10) in the overall task and subtasks (unit, %MVC) 3.4 Velocity and repetitiveness Female subjects exhibited significantly faster overall extension-flexion speed in their non-dominant hands than male subjects did (p<0.001, t-test) (Table 5). This group difference also applied to hairdressers and barbers (p<0.005, t-test). Surprisingly, no gender effect or any gender-associated effect was identified by repeated-measures ANOVA. Comparatively, in taking occupation group as a between-subject factor, a significant subtask effect (p<0.005) and subtask side occupation interaction (p<0.001) were observed for both extension-flexion and radial-ulnar velocities, and a significant subtask occupation interactive effect (p<0.001) was noted for extension-flexion velocity (Table 5). The highest mean angular velocity (92.7 /s) was in the flexion-extension direction of the hairdressers non-dominant wrists during the blow-drying subtask. Overall mean velocities of the bilateral 10

12 wrists of hairdressers were and /s in flexion and deviation (data not shown), roughly 10 20% greater than those of barbers. Notably, no significant effect of gender, occupation, subtask, side, or interaction of these factors existed in MPF. The overall MPF of bilateral wrists of hairdresser were and Hz in flexion and deviation (not in tables), respectively, and were generally equivalent to those of barbers. Table 5. Average wrist velocity and mean power frequency (mean ± SD) in the overall task and subtasks 4. Discussion Musculoskeletal symptoms are more prevalent among female than male hairstylists in Taiwan (Taiwan IOSH, 2004). This outcome may reflect the fact that more females worked as hairdressers than as barbers in Taiwan and due to differences between barber and hairdresser tasks. This study demonstrates that female subjects had significantly higher EMG activity in all statistical tests, and significantly faster overall extension-flexion velocity in their non-dominant hands than male subjects. In addition to significant gender differences in EMG activity, significant side occupation and side subtask occupation interactive effects existed in angle position, and a significant subtask occupation effect was identified for extension-flexion velocity. These significant interactive effects associated with occupational factor are indicative of the task differences between barbers and hairdressers. According to onsite observations, barbers and hairdressers used their non-dominant hands to comb, hold hair with their fingers and wave/curl hair while cutting and/or blow-drying. Handling the long hair of female customers typically required higher exertion and possibly faster hand 11

13 movement of the non-dominant hand than that experienced by barbers handling short hair. Comparatively, barbers had fewer fast movements while combing hair, holding hair with their fingers, and blow-drying men s hair than hairdressers for the same tasks. Since elevated dynamic EMGs result from increased exertion and fast muscle contraction, the slow movements of barbers may account for the significantly lower EMGs in the non-dominant hands of barbers than those of hairdressers. This observation may explain the higher overall extensor EMG (Table 3) and faster overall flexion velocity (Table 5) demonstrated by non-dominant hands of hairdressers than barbers. Another possible cause of higher complaint rate among hairdressers than barbers may be their longer continual exposure, although both occupation groups exhibited relatively similar working postures and EMG activity in their dominant hands. The average duration of a woman s haircut was 15 minutes longer than a man s haircut haircutting was 5min longer and blow-drying was 10min longer. During the haircutting and blow-drying subtasks, hairdressers generally applied more force and higher velocity with their non-dominant hands than did barbers. Hairdressers are therefore less likely to share the workload with their dominant hands by using non-dominant hands when performing auxiliary tasks during rest periods. Additionally, the muscles, tendons and joints of hairdressers require longer time to recover from rapid movement than do those of barbers. Thus, the combination of fast movement and prolonged exposure may account for hairdressers having a greater complaint rate than barbers. Table 6 lists the wrist exposure of workers right or dominant hand in several industrial studies. According to previous studies, numerous guidelines for work methods focus on exposure level. Reducing the level of internal biomechanical exposure is recommended when exposure exceeds a threshold level of 30 50%MVC (dynamic/occasional contraction) (Westgaard and Winkel, 1996). In this study, the 90th percentile of overall muscular loads 12

14 measured in dominant flexors (19 24%MVC) and extensors (21 27%MVC) of barbers and hairdressers were <30%MVC, and less than those values for machine-milking (Stål, 2000) and poultry-deboning tasks (Juul-Kristensen, 2002). Nevertheless, when only female hairstylists were considered, their mean dominant flexor EMG (29.4%MVC) and extensor EMG (34%MVC) were higher than those values for poultry-deboning tasks (Juul-Kristensen, 2002). According to Kilbom (1994) and Malchaire (1997), 15% MVC is the mean acceptable contraction intensity when working over prolonged periods. Comparing this limit and that for several industrial tasks (Table 6), this study identified low muscular loading in male subjects and moderate muscular loading in female subjects during haircutting tasks. Table 6. Wrist exposure of worker right or dominant hand from selected industrial studies and this study Malchaire et al. (1997) demonstrated that high angular velocity of the wrist predicted musculoskeletal wrist injuries (OR=1.46) and the risk increased when flexion velocity exceeded 50 /s. The movements of hairdressers were not forceful; however, the rate of movement was very rapid. In this study, mean flexion velocities (barber=67 /s, hairdresser=74 /s) were considerably greater than 50 /s, and even higher than those values for repetitive tasks with a high risks of hand disorders such as tasks in fish (Hansson, 1996) and poultry processing (Juul-Kristensen, 2001, 2002), machine-milking (Stål, 1999, 2000, 2003), 刪 除 : machine and dentistry (Åkesson, 1997) (Table 6). If median velocity is estimated by dividing mean velocity by 1.5 (Hansson, 1996), the hairdressing task may also have greater velocity than repetitive tasks in the lamination industry (Hansson, 2000) and non-forceful quality control operators who have a high prevalence of wrist/hand disorders (Arvidsson, 2003). Hairdressing includes tasks with repetitive movements and this repetitiveness is 13

15 considered moderately high (MPF= Hz in flexion and Hz in deviation) when compared with MPF values obtained by previous studies (Table 6). This study observed slightly lower MPF values than those reported for fish (Hansson, 1996) and poultry processing (Juul-Kristensen, 2001, 2002), non-forceful or repetitive laminate industry tasks (Arvidsson, 2003; Hansson, 2000) (range, Hz in flexion and Hz in deviation). Nevertheless, the repetitiveness of barbers and hairdressers were close to those values for machine-milking ( Hz in flexion and Hz in deviation) (Stål, 刪 除 : machine 1999, 2003), and much higher than those in dentistry (Åkesson, 1997). Moreover, all these studies reported a high prevalence of wrist and hand disorders. Several studies have indicated that hand positions and perceived symptoms in the wrists and hands are correlated (Åkesson, 1997; Stål, 1999). This study shows that wrist ROM in flexion (60 ) was comparable to or higher than those values for fish and poultry processing and all other listed industrial works (Table 6). Furthermore, the hairdressing task had greater ulnar deviation than all other tasks listed except for fish and poultry processing (Hansson, 1996; Juul-Kristensen, 2001). Not only do flexion and extension of the wrist contribute to carpal tunnel syndrome (CTS), radial and ulnar deviations have been implicated (Tanaka et al., 1995). A previous study that obtained in vivo pressure measurements inside the carpal tunnel showed that these pressures increased parabolically as radial and ulnar deviations increased (Weiss et al., 1995). Bergamasco et al. (1998) suggested that wrist angle exceeding 24 in ulnar deviation and 15 in radial deviation were unacceptable. In this study, hairdressers held their hands at an ulnar deviation of >24 for more than 10% of the task time and, thus, may be subject to injury risk via undesirable wrist postures. Experimental results are limited by finite evaluated tasks and gender-unmatched groups. In short, barbers and hairdressers carry out several tasks other than hair cutting, washing, and blow-drying. Furthermore, the hairdresser group also shampoo, color, and perm hair more 14

16 than barbers. Different exposure percentages for tasks involving a particular tool or working manner can result in a difference in mechanical exposure between barbers and hairdressers. This study only compared exposure differences between barber and hairdresser groups when performing the most common tasks. The results may vary according to hairdressing tasks. Further investigation and quantitative analysis of other hairdressing tasks are required to accurately assess the workloads of barbers and hairdressers. The percentage of female subjects in the barber group in this study may be lower than that in the working population in Taiwan s hairdressing industry. Testing groups with different genders was partially due to biased background populations in barber and hairdresser industries. However, females are reportedly more vulnerable to CTS than males. Armstrong and Chaffin (1979) and Matias et al. (1998) demonstrated that anthropometric factors play a role in the development of CTS when they are associated with prolonged duration of risk and awkward postures. The anatomical differences between males and females (e.g., wrist circumferences and radial bone size) may be a factor for increased risk for CTS in females. This study was unable to distinguish clearly between the effects of gender and occupation due to the limited number of samples and gender-unmatched samples. Future studies with genderand age-matched groups are required to justify further the gender effect on musculoskeletal prevalence among hairdressers. The manual tasks in hairdressing have persisted for decades and are likely not easily replaced or changed. The high number of claims associated with musculoskeletal morbidity in the hairdressing industry in many countries may not be easily reduced without major reductions in the mechanical exposure of hairdressers. This study quantified the exposures of two similar tasks with different prevalence rates to locate substantial risk factors for musculoskeletal disorders. Despite aforementioned study limitations, analytical results of this study identified possible causes of the high incident rate of work-related musculoskeletal 15

17 disorders in hairdressers. Additionally, the quantitative assessment in this study serves as a reference for comparing mechanical exposures of various tasks with high risk for musculoskeletal disorders. 16

18 Acknowledgement The authors wish to thank the National Science Council and the Institute of Occupational Safety, Taiwan for providing financial support for this research study (NSC E , IOSH 95-H102). Reference Åkesson, I., Hansson, G.-Å., Balogh, H., Moritz, U., Skerfving, S., Quantifying work load in neck, shoulders and wrists in female dentists. Int. Arch. J Occup. Environ. Health 69, Armstrong, T.J., Chaffin, D.B., Carpal tunnel syndrome and selected personal attributes. J. Occup. Med. 21, Arokoski, J.P.A., Nevala-Puranen, N., Danner, R., Halonen, M., Tikkanen, R., Occupationally oriented medical rehabilitation and hairdressers work techniques A one-and-a-half-year follow-up, Int. J. Occup. Saf. Ergon. 4, Arvidsson, I., Åkesson, I., Hansson G.-Å., Wrist movements among females in a repetitive, non-forceful work. Appl. Ergon. 34, Bergamasco, R., Girola, C., Colombini, D., Guidelines for designing jobs featuring repetitive tasks. Ergonomics. 4, Chen, H.C., Sheng, C.F., Chen C.Y., Worksite monitoring technology for investigating repetitive strain injuries in upper limbs. J. Occup. Saf. Health 13, (in Chinese) Delagi, E.F., Iazzetty, J., Perotto, A., Morrison, D., Anatomical guide for the electromyographer: the limbs and trunk. Springfield: Charles C. Thomas. English, C.J., Maclaren, W.M., Court-Brown, C., Hughes, S.P.F., Porter, R.W., Wallace, 17

19 W.A., Graves, R.J., Pethick, A.J., Soutar, C.A., Relations between upper limb soft tissue disorders and repetitive movements at work, Am. J. Ind. Med. 27, Guo, H.R., Tanaka, S., Cameron, L.L., Seligman, P.J., Behrens, V.J., Back pain among workers in the United States: National estimates and workers at high risk, Am. J. Ind. Med. 28, Hagberg, M., 1984, Occupational musculoskeletal stress and disorders of the neck and shoulder: a review of possible pathophysiology, Int. Arch. Occup. Environ. Health 53, Hansson G.-Å, Balogh I, Ohlsson, K., Rylander, L., Skerfving, S., Goniometer measurement and computer analysis of wrist angles and movements applied to occupational repetitive work. J. Electromyogr. Kines. 6, Hansson G.-Å, Balogh I, Ohlsson, K., Pålsson, B., Rylander, L., Skerfving, S., 2000, Impact of physical exposure on neck and upper limb disorders in female workers. Appl. Ergon. 31, Jonsson, B., Measurement and evaluation of local muscular strain on the shoulder during constrained work. J. Hum. Ergol. 11, Juul-Kristensen, B., Hansson, G-Å., Fallentin, N., Andersen, J.H., Ekdahl, C., Assessment of work postures and movements using a video-based observation method and direct technical measurements, Appl. Ergon. 32, Juul-Kristensen, B, Fallentin N., Hansson, G-Å., Madeleine, P., Andersen, J.H., Ekdahl, C., Physical workload during manual and mechanical deboning of poultry. Int. J. Ind. Ergo. 39, Kang, D.M., Lee, J.T., Kang, M.S., Park, S.H., Urm, S.H., Kim, S.J., Jeong K.W., Shon H.S., Park, B.J., Prevalence on dermatologic, respiratory and musculoskeletal symptoms 18

20 among hairdressers, Korean J. Occup. Environ. Med. 11, Kilbom, Å., Repetitive work of the upper extremity: part II The scientific basis (knowledge base) for the guide. Int. J. Ind. Ergon. 14, Leino, T., Tuomi, K., Paakkulainen, H., Klockars, M., Health reasons for leaving the profession as determined among Finnish hairdressers in Int. Arch. Occup. Environ. Health 72, Liu, Y.P., Chen, H.C., Chen, C.Y., Portable data logger for worksite measurement of physical workload. J. Med. Biol. Eng. 26, Malchaire, J.B., Cock, N.A., Piette, A., Leao, R.D., Lara, M., Amaral, F., 1997, Relationship between work constraints and the development of musculoskeletal disorders of the wrist: A prospective study. Int. J. Ind. Ergon.. 19, Matias, A.C., Salvendy, G., Kuczek, T., Predictive models of carpal tunnel syndrome causation among VDT operators. Ergonomics 41, Nevala-Puranen, N., Halonen, M., Tikkanen, R., Arokoski, J.P.A., Changes in hairdressers work techniques and physical capacity during rehabilitation, Occup. Ergonom. 1, New Zealand DOL, 2007, Health and safety in hairdressing, Department of Labour (DOL), Wellington, New Zealand, ISBN Stål, M., Hansson, G.-Å., Moritz, U., Wrist positions and movements as possible risk factors during machine milking. Appl. Ergon. 30, Stål, M., Hansson, G.-Å., Moritz, U., Upper extremity muscular load during machine milking. Int. J. Ind. Ergo., 26, Stål, M., Pinzke, S., Hansson, G.-Å., Kolstrup, C., Highly repetitive work operations in a modern milking system. A case study of wrist positions and movements in a rotary 19

21 system. Ann. Agric. Environ. Med. 10: Taiwan DGBAS, 2001, General report on industry, commerce and service census in Taiwan-Fuchien area, Directorate-General of Budget, Accounting and Statistics (DGBAS), Taiwan. Taiwan IOSH, 2004, A study of the exposure to musculoskeletal hazards in the manufacturing industry, Institute of Occupational Safety and Health (IOSH), Council of Labor Affairs, Taiwan, IOSH93-H103. (in Chinese) Taiwan IOSH, 2006, Investigation of risk factor exposure in work-related hand-wrist musculoskeletal disorders, Institute of Occupational Safety and Health (IOSH), Council of Labor Affairs, Taiwan, IOSH95-H102. (in Chinese) Tanaka, S., Wild, D.K., Seligman, P.J., Halperin, W.E., Behrens, V.J., Putz-Anderson, V., Prevalence and work-relatedness of self-reported carpal tunnel syndrome among US workers: analysis of the occupational health supplement data of 1988 National Health Interview Survey, Am. J. Ind. Med. 27, Veiersted, K.B., Gould, K.S., Østerås, N., Hansson, G.-Å., Effect of an intervention addressing working technique on the biomechanical load of the neck and shoulders among hairdressers, Appl. Ergon. 39, Weiss N.D., Gordon, L., Bloom, T., So, Y., Rempel, D.M., Position of the wrist associated with the lowest carpal tunnel pressure: implications for splint design, J. Bone Joint Surg. 77, Westgaard, R.H., Winkel, J., Guidelines for occupational musculoskeletal load as a basis for intervention: a critical review. Appl. Ergon. 27, Figure captions Figure 1. Experiment setup of a barber cutting hair 20

22 Table captions Table 1. Overall and subtask durations (mean±sd) of a standard haircut (unit: min) Table 2. Mean (SD) wrist positions of barbers (B, n=11) and hairdressers (H, n=10) during overall task and subtasks (unit: deg) Table 3. Mean (SD) EMG activity of the wrist flexor and extensor muscles of barbers (B, n=11) and hairdressers (H, n=10) in the overall task and subtasks (unit, %MVC) Table 4. Mean (SD) EMG activity of the wrist flexor and extensor muscles of male (M, n=11) and female (F, n=10) subjects in the overall task and subtasks (unit, %MVC) Table 5. Average wrist velocity and mean power frequency (mean ± SD) in the overall task and subtasks Table 6. Wrist exposure of worker right or dominant hand from selected industrial studies and this study 21

23 Table 1. Basic data of subjects (mean±sd) Table 1. Overall and subtask durations (mean±sd) of a standard haircut (unit: min) Group Overall Sub-task Task* cutting washing blow-drying** Barber (n=11) 35.6± ± ± ±1.8 Hairdresser (n=10) 51.4± ± ± ±3.7 *p<0.005, **p<0.001 significant occupation group difference(t-test) 22

24 Table 2. Mean (SD) wrist positions of barbers (B, n=11) and hairdressers (H, n=10) during overall task and subtasks (unit: deg) Angle Task 10th percentile 50th percentile 90th percentile B H B H B H O -36(8) -42(8) -10(8) -14(8) 23(9) 18(6) dominant C -31(11) -42(8) -6(9) -12(7) 28(12) 21(8) W -38(9) -46(10) -15(8) -23(10) 13(9) 12(11) B -41(8) -37(9) -14(9) -12(8) 18(8) 17(8) O -44(8) -36(7) -19(6)** -7(8) 11(6)* 26(12) non-dominant C -45(9) -31(7) -23(8) -2(8) 8(6) 26(14) W -38(9) -39(10) -12(10) -5(16) 13(10) 22(16) B -42(10) -39(9) -15(8) -10(10) 14(8) 24(14) p 1 <0.001 p 1 <0.001 p-value p 3 <0.001 P 2 <0.005 p 2 <0.005 p 3 <0.001 p a <0.005 Angle Task 10th percentile 50th percentile 90th percentile B H B B H B O -3(9) -7(6) 10(9) 9(7) 23(9) 27(9) dominant C -2(9) -7(5) 12(9) 8(6) 24(8) 23(7) W -6(10) -10(7) 8(10) 7(9) 19(11) 21(8) B -4(8) -3(11) 8(8) 14(12) 21(8) 28(10) O -10(8) -8(7) 3(6) 7(9) 16(7) 23(10) non-dominant C -10(9) -8(7) 0(8) 7(8) 13(7) 21(8) W -8(8) -9(8) 4(7) 7(8) 16(8) 23(9) B -8(7) -7(8) 7(6) 8(11) 19(8) 23(12) p-value n.s. p 3 <0.001 p 3 <0.005 p a <0.005 O: overall task; C: cutting subtask; W: washing subtask; B: blow-drying subtask *p<0.005, **p<0.001 significant occupation group difference in overall task(t-test) extension(-)/flexion(+) radial(-)/ulnar(+) p<0.001 significant gender difference in overall task(t-test) p 1, p 2,p 3 denote significant interaction effects of side occupation, side subtask, and side subtask occupation, respectively, with occupation as between-subject factor (repeated-measures ANOVA). p a denotes significant effect of side subtask with gender as between-subject factor (repeated-measures ANOVA). 刪 除 : repeated 刪 除 : repeated 23

25 Table 3. Mean (SD) EMG activity of the wrist flexor and extensor muscles of barbers (B, n=11) and hairdressers (H, n=10) in the overall task and subtasks (unit: %MVC) Extensor dominant non-dominant Flexor dominant non-dominant Task 10th percentile 50th percentile 90th percentile B H B H B H O 3.0(3.2) 4.5(1.5) 9.2(9.4) 12.9(3.6) 21(20) 27(6) C 3.3(3.1) 4.7(1.6) 9.3(9.4) 12(3) 20(20) 25(6) W 3.0(3.3) 4.3(1.6) 9.2(9.4) 12(5) 22(20) 27(11) B 3.1(3.6) 4.9(1.8) 9.2(10) 15(4) 23(23) 30(7) O 1.7(1.3)** 4.5(1.6) 6.9(4.6)** 15(5) 16(11)** 33(10) C 2.0(1.6) 4.3(2.0) 5.7(4.2) 13(4) 14(10) 27(8) W 1.7(1.1) 4.9(1.6) 6.8(5.8) 16(6) 19(13) 33(11) B 1.6(1.2) 6.4(2.4) 6.1(4.3) 19(6) 18(14) 40(13) p 1 <0.001 p 1 <0.001 p value n.s. p 2 <0.005 p 3 <0.001 Task 10th percentile 50th percentile 90th percentile B H B H B H O 1.4(0.7) 2.1(1.0) 4.9(3.3)* 9.6(3.1) 19(12) 24(9) C 1.4(0.6) 2.0(0.9) 4.0(2.7) 8.5(2.4) 14(9) 20(5) W 1.7(1.1) 2.2(1.4) 7.7(5.5) 13(6) 25(14) 30(15) B 1.6(0.8) 3.1(2.0) 5.8(4.6) 11(6) 22(16) 24(12) O 1.8(1.2) 2.7(2.2) 4.0(3.7) 9.7(7.4) 13(10) 24(18) C 1.8(1.2) 2.6(2.1) 3.6(3.3) 8.3(5.7) 9.2(6.7) 20(15) W 1.9(1.3) 2.4(1.7) 5.8(5.7) 13(10) 17(15) 29(21) B 1.7(1.1) 3.5(2.7) 4.1(3.6) 11(8) 14(13) 26(17) p p 1 <0.001 p 1 <0.001 value p 3 <0.001 O: overall task; C: cutting subtask; W: washing subtask; B: blow-drying subtask *p<0.005, **p<0.001 significant occupation group difference in overall task(t-test) p 1 denotes significant within-subject effect of subtask (repeated-measures ANOVA). p 2, p 3 denote significant interaction effects of side subtask and subtask occupation, respectively, with occupation as between-subject factor (repeated-measures ANOVA). 刪 除 :, 刪 除 : repeated 刪 除 : repeated 24

26 Table 4. Mean (SD) EMG activity of the wrist flexor and extensor muscles of male (M, n=11) and female (F, n=10) subjects in the overall task and subtasks (unit: %MVC) Extensor dominant non-dominant Flexor dominant non-dominant Task 10th percentile 50th percentile 90th percentile M F M F M F O 2.3(2.0) 5.3(2.4) 7.0(6.3) 15.6(6.0) 15.7(13.2)* 34.0(12.0) C 2.6(2.1) 5.5(2.3) 7.2(6.3) 14.9(6.2) 15.2(12.8) 31.2(12.9) W 2.3(2.2) 5.1(2.5) 6.8(6.6) 14.7(6.6) 16.3(13.9) 33.7(13.8) B 2.2(1.9) 6.0(2.8) 6.7(6.3) 18.4(5.5) 16.0(14.5) 38.0(13.5) O 1.8(1.5)** 4.5(1.5) 5.5(3.7)** 15.6(4.1) 13.5(7.7)** 36.0(8.5) C 1.9(1.6) 4.3(1.9) 5.3(3.6) 13.4(4.0) 12.0(7.1) 28.5(7.3) W 1.7(1.4) 4.9(1.4) 5.5(3.2) 17.2(4.7) 15.1(7.7) 36.9(10.0) B 2.0(1.7) 5.9(2.9) 6.1(4.3) 19.3(6.3) 15.2(9.2) 43.5(10.6) p 1 <0.001 p 1 <0.001 p value p 2 <0.001 p 2 <0.001 p 3 <0.001 p 3 <0.001 p 3 <0.001 Task 10th percentile 50th percentile 90th percentile M F M F M F O 1.3(0.5) 2.3(1.0) 4.2(2.2)** 10.4(2.7) 14.5(6.9)** 29.4(8.6) C 1.3(0.6) 2.1(0.9) 3.9(2.4) 8.8(2.3) 11.5(6.3) 23.6(4.8) W 1.4(0.8) 2.6(1.4) 5.7(2.8) 15.5(4.9) 19.2(9.6) 36.9(13.2) B 1.3(0.5) 3.4(1.8) 4.3(1.8) 12.8(5.2) 15.7(9.9) 31.6(13.1) O 1.5(1.1) 3.1(2.1) 3.0(2.1)* 10.9(7.0) 9.7(5.3)* 27.7(17.1) C 1.4(1.0) 3.0(2.0) 2.8(2.0) 9.3(5.3) 7.6(4.1) 21.6(14.6) W 1.7(1.2) 2.8(1.6) 4.4(3.1) 14.3(9.8) 12.8(7.6) 33.8(21.9) B 1.5(1.0) 3.7(2.5) 3.3(2.4) 12.2(7.3) 11.0(9.2) 29.7(16.6) p 1 <0.001 p 1 <0.001 p 1 <0.001 p 2 <0.001 p 2 <0.001 p 3 <0.005 p 3 <0.001 p 3 <0.001 p value O: overall task; C: cutting subtask; W: washing subtask; B: blow-drying subtask *p<0.005, **p<0.001 significant gender difference in overall task(t-test) p 1 denotes significant within-subject effect of subtask (repeated-measures ANOVA). p 2 denote significant interaction effect of subtask gender (repeated-measures ANOVA). p 3 denotes significant between-subject effect of gender (repeated-measures ANOVA). 刪 除 :, 刪 除 : repeated 刪 除 : repeated 刪 除 : repeated 25

27 Table 5. Average wrist velocity and mean power frequency (mean ± SD) in the overall task and subtasks Wrist Parameter Non-dominant Dominant p movement Task barber hairdresser barber hairdresser value O 64.2±11.5* 82.6± ± ±7.3 C 59.8± ± ± ±8.6 ext-flex W 69.8± ± ± ±4.5 Angular velocity (deg/s) Mean power frequency (Hz) radial-ulnar ext-flex radial-ulnar B 63.7± ± ± ±12.7 O 33.5± ± ± ±8.2 C 29.5± ± ± ±8.8 W 37.5± ± ± ±8.8 B 35.2± ± ± ±9.4 O 0.36± ± ± ±0.16 C 0.37± ± ± ±0.16 W 0.40± ± ± ±0.12 B 0.36± ± ± ±0.16 O 0.41± ± ± ±0.11 C 0.44± ± ± ±0.19 W 0.52± ± ± ±0.15 B 0.53± ± ± ±0.20 O: overall task; C: cutting subtask; W: washing subtask; B: blow-drying subtask *p<0.005 significant occupation difference in overall task(t-test) p<0.001 significant gender difference in overall task(t-test) p 1 denotes significant within-subject effect of subtask (repeated-measures ANOVA). p 2 denotes significant interaction effect of subtask occupation (repeated-measures ANOVA). p 3 denotes significant subtask side occupation interaction effect (repeated-measures ANOVA). p 1 <0.005 p 2 <0.001 p 3 <0.001 p 1 <0.005 p 3 <0.001 n.s. n.s. 刪 除 : repeated 刪 除 : repeated 刪 除 : repeated 26

28 Table 6. Wrist exposure of worker right or dominant hand from selected industrial studies and this study Wrist angle (deg) 10 th 90 th percentile EMG (%MVC) 10 th 90 th percentile MPF (Hz) Angular velocity ( /s) mean/median*(high) Reference Task (no. of subjects) ext(-)/flx(+) radi(-)/ulna(+) Flexor Extensor Flexion Deviation Flexion Deviation this study Barber (n=11) this study Hairdresser (n=10) Hansson, 1996 Fish processing (n=32) n.a. n.a /41*(142) 36/25*(84) Juul-Kristensen, 2001 Poultry processing (n=21) n.a. n.a /38*(130) 34/23*(76) Juul-Kristensen, Poultry deboning 2002 (mechanical, n=13) /36*(127) 28/18*(64) Stål, 1999, 2000 Machine-Milking (loose-housing, n=11) /28*(147) 16*(76) Machine-Milking (tethering, n=11) /24*(136) 15*(72) Stål, 2003 Machine-Milking (rotary, n=13) n.a. n.a /36*(155) 29/17*(74) Arvidsson, 2003 Non-forceful QC work (n=12) n.a. n.a /29*(108) 20*(67) Hansson, 2000 Repetitive industry work (laminate, n=81) n.a. n.a. n.a n.a. 23*(110) n.a. Åkesson, 1997 Female dentist (n=12) /6.2*(45) 8.2/2.7*(26) * denotes the median (50th percentile) velocity () denotes the high (90th percentile) velocity 刪 除 : Machine 刪 除 : Machine 刪 除 : Machine 27