- Email: [email protected]
Effects of work experience and exertion height on static lifting strengths and lift strategies of experienced and novice female participants
International Journal of Industrial Ergonomics 44 (2014) 607e614
Contents lists available at ScienceDirect
International Journal of Industrial Ergonomics journal homepage: www.elsevier.com/locate/ergon
Effects of work experience and exertion height on static lifting strengths and lift strategies of experienced and novice female participants Yi-Lang Chen* Department of Industrial Engineering and Management, Ming Chi University of Technology, 84, Gungjuan Rd., Taishan Dist., New Taipei 24301, Taiwan, ROC
a r t i c l e i n f o
a b s t r a c t
Article history: Received 21 June 2013 Received in revised form 12 June 2014 Accepted 13 June 2014 Available online
In this study, 46 female experienced workers and inexperienced novices (23 each) were recruited to determine their maximum lifting strength at 15 exertion heights between 10 and 150 cm from the floor. The results revealed that the experienced workers' strengths at all 15 heights exhibited relatively little fluctuation, and were approximately 50e70 N lower than those of novices when heights were 50 cm. No differences in strengths were observed at 60e150 cm between the groups. The experienced workers tended to adopt a consistently deep squat at lower heights (50 cm) and a more erect posture with stiffened arms at higher heights (70 cm), resulting in lower L4/L5 disc compression forces and shoulder moments than in novices, respectively. In contrast to the lifting techniques adopted by experienced workers to effectively avoid overloading, the findings suggest that novice female workers who lack experience should be cautious and trained for performing lifting tasks. Relevance to industry: This study demonstrated that a discrepancy exists in lifting strengths between experienced and inexperienced female handlers at various heights. Findings from this study offers a reference for lifting task design and effective posture training on lifting, regarding the weighting of various shelves. © 2014 Elsevier B.V. All rights reserved.
Keywords: Lifting postures Experienced workers and novices Lifting heights
1. Introduction Manual handling remains prevalent in Taiwan, and over 500,000 workers are employed in jobs related to warehousing (CEPD, 2011). Most tasks in warehouses, distribution centers, and various retail establishments involve manual materials handling and strength exertion at various heights of lift. These tasks typically involve submaximal weights; in other words, the weights handled are light to moderate and therefore do not require only male employees. A recent study of Taiwanese warehouse workers revealed that the lower back, shoulders, and arms were regions of bodily discomfort for both male and female workers, particularly for novice workers (Chen, 2008). This might have resulted from the warehousing tasks, which were highly repetitive but diverse in lifting heights, that these workers were asked to perform. One factor affecting lifting strength is the exertion height; in other words, the height at which the lift is initiated (Yates et al., 1980; Mital et al., 1993; Lee, 2004; Chen et al., 2011). This is
* Tel.: þ886 2 29089899x3103; fax: þ886 2 29085900. E-mail address: [email protected]. http://dx.doi.org/10.1016/j.ergon.2014.06.004 0169-8141/© 2014 Elsevier B.V. All rights reserved.
particularly true in service warehouse stores, which economize and maximize storage space. Lifting strength capacity most likely changes with height and thus placing a person at higher percentages of their maximum in some locations than others (Yates et al., 1980). The closer they are to their maximum, the greater their potential risk for injury (Mital et al., 1993). Investigations on lifting strength have typically been confined to specific height levels (e.g., floor, knuckle, and shoulder); however, this is not applicable to warehousing tasks. Height of lifting also influences posture, which, in turns, affects internal loads on potentially vulnerable tissues (Chen, 2000; Plamondon et al., 2012). Therefore, it is critical to understand the effect of height on lifting capacity and the corresponding body loads. Much of the literature has reported that the handling techniques of highly skilled workers differ substantially from those of novices (Authier et al., 1996; Gagnon et al., 1996). Field studies have also revealed that workers use methods other than those traditionally recommended (Kuorinka et al., 1994; Baril-Gingras and Lortie, 1995; Authier et al., 1996). However, certain researchers have determined that experienced workers possess advantages; for a given task, experienced workers generally employ lower biomechanical spinal loads (Marras et al., 2006), fewer back-muscle
608
Y.-L. Chen / International Journal of Industrial Ergonomics 44 (2014) 607e614
activations (Keir and MacDonell, 2004), a narrower lumbar-spine range of motion (Hodder et al., 2010), higher subjective discomfort thresholds (Parakkat et al., 2007) and psychophysically accepted lifting weights (Mital, 1987), and a greater knee flexion during ground-level task (Plamondon et al., 2010, 2012) than novices. Conversely, Mital and Manivasagan (1983) indicated that task variables similarly influenced the maximum lifting capacities of experienced workers and college students, and controversy in this regard might be attributable to the distinct experimental settings adopted by these studies. Warehouses in Taiwan have high employee turnover rates, and recruit new full-time and part-time workers weekly (Chen, 2008). These inexperienced novice workers require more attentiveness to avoid injury resulting from daily tasks. Recently, Chen et al. (2011) systemically assessed the lifting strength of male experienced workers and novices at full-range heights, and observed that the average strength lifted upward by novices was 4.57e7.61 kg lower than that of experienced workers when lifting at 100e120 cm heights. The workers' upward strengths, contrary to expectations, remained nearly unchanged throughout all 15 full-range heights. In that study, the postures adopted by the workers were also highly differentiated from that of the novices when performing near-floor positions, but their strengths were equivalent. Experienced workers tended to adopt a safer (i.e., more flexed knees) and more skillful technique than novices did to generate force, which might result in lower spinal loads during lifting. However in that study, only male experienced workers and inexperienced university students were compared. Previous studies have demonstrated that the inherent difference in muscular characteristics between sexes (Xiao et al., 2005; Agrawal et al., 2009). Garg and Ayoub (1980) noted that the maximum lifting capacity of women was 60e70% that of men, and an earlier study by Murrel (1971) also suggested that women were able to exert approximately two-thirds the force of men. However, the ratio of isometric strength between Chinese men and women was nearly 50% (Xiao et al., 2005). Miller et al. (1993) observed that sex differences in muscle strength were primarily confined to muscles in the upper extremities, particularly the shoulders. Hooftman et al. (2009) indicated that even if men and women perform the same tasks, they do not perform these tasks in the same manner, which might also lead to differences in exposure. This can be attributed to a poorer ergonomic fit of the workplace to women than to men, or because men and women choose to perform the same task using distinct strategies. Lifting tasks distributed to male and female workers in supermarkets were observed to be identical regarding shelf heights and load weights. This might imply a higher risk for women than for men, because these tasks place the women at a disadvantage compared with men when lifting. We therefore examined the effect of height on lifting capacity and the corresponding body loads for experienced and inexperienced female participants recruited from a large supermarket and a university, respectively. We hypothesized that experienced female participants' strengths and lifting strategies at 15 lifting heights would differ from inexperienced female participants' because of the daily workplace practices where the experienced group had worked. We also examined whether the results regarding the effects of these tasks on men were also applicable to women.
volunteered for this study, and received an hourly wage for completing all of the test conditions. Twenty-three workers with a minimum of 2-years of experience were selected as experienced participants. They were selected from a large supermarket, and their duties consisted primarily of varied shelf-replenishing tasks. Another 23 novice participants (with no manual material-handling experience) were recruited from among university students, based on previous studies (Mital, 1987; Marras et al., 2006; Parakkat et al., 2007). Informed consent was obtained from all participants in accordance with the regulations. The Ethics Committee of Chang Gung Memorial Hospital of Taiwan approved this study. The mean (standard deviation; SD) age, stature, and body mass were 28.6 (6.4) years, 158.9 (5.3) cm, and 58.3 (10.2) kg for the experienced group, and 22.3 (3.2) years, 160.2 (4.1) cm, and 51.3 (6.5) kg for the inexperienced group, respectively. Detailed information, anthropometric data, and isometric muscular strengths for the 46 participants are listed in Table 1. The isometric strength measurement procedure was based on the studies of Ayoub et al. (1978) and Mital (1987). 2.2. Experimental apparatus Lifting strengths were measured using the Static-Lifting Strength Tester (SLST), as shown in Fig. 1. The SLST, with incremental height settings for measuring strengths, consists of a standing platform, a steel frame with 15 positioning holes (ranging from 10 to 150 cm in increments of 10 cm), a sliding height stopper along the frame, and a 55-cm handle bar (diameter: 3.5 cm) attached to the stopper. Force applied vertically to the handle bar was measured by a load cell connected to the bar that was firmly attached to the stopper; the strength signal (40 Hz) was then transferred to an A-D converter and a digital readout unit (Jackson Strength Evaluation System, JSES Model 32628, USA). The A-D converter was calibrated prior to the strength testing against known weights. 2.3. Body posture recording and biomechanical analysis The testing postures of randomly sampled participants (12 workers and 12 novices) were videotaped in this study. Six adhesive reflective markers were attached to the right side of their ulna
Table 1 Participant data for the experienced workers and novices (23 in each group). Variables
Mean Age (years)* Stature (cm) Body mass (kg)* Acromial height (cm) Elbow height (cm) Knuckle height (cm) Hip height (cm) Knee height (cm) Hand length (cm) Hand width (cm) Isometric strengths (N)a Composite strength Back strength Shoulder strength Arm strength
2. Methods 2.1. Participants *
Forty-six experienced female workers and novice participants (23 of each), with no history of musculoskeletal disorders,
Experienced workers
28.6 158.9 58.3 127.8 95.5 68.9 81.7 42.1 14.9 7.1 599.8 497.8 282.2 218.5
Novices S.D. 6.4 5.3 10.2 5.03 3.81 4.62 6.30 3.33 0.64 0.5 132.3 128.4 68.6 51.9
Mean 22.3 160.2 51.3 128.0 96.9 69.4 83.0 43.6 14.9 7.0 611.5 521.4 267.5 231.3
S.D. 3.2 4.1 6.5 4.88 2.63 5.03 6.15 2.32 0.67 0.4 138.2 136.2 80.4 49.0
Significantly different at p < .05. a Testing procedure was based on the studies of Ayoub et al. (1978) and Mital (1987).
Y.-L. Chen / International Journal of Industrial Ergonomics 44 (2014) 607e614
609
2.5. Experimental procedure
Fig. 1. Device used to test lifting strength and schematic testing posture in the study, at a height of 20 cm.
condyle, the lateral epicondyle of the humerus, the acromial shelf, the femoral greater trochanter, the lateral femoral condyle, and the lateral malleolus of the fibula. In the study, the joint angles were all defined as the inter-joint angles between the two adjacent segments (Fig. 1). The digital video camera (Qualisys MacReflex, Sweden) was placed at a distance of 5 m from the participants, and was used to record the positions of each joint marker when participants performed the strength tests. The camera heights were consistently adjusted to orthogonally align with the participants' hip positions to eliminate as many distortion errors as possible resulting from the varied testing heights. Image-processing software (Qualisys MacReflex, Sweden) was then used to calculate the relevant joint angles. When the participants performed the test tasks, the L4/L5 disc compression forces and shoulder reactive moments were estimated by the 3D Static Strength Prediction Program™ (3DSSPP, V6.0.5), which was developed by the Center for Ergonomics at the University of Michigan. Participants' joint angles, anthropometric data, and strength values were used as inputs in the program to estimate the lumbar compression forces and shoulder moments. Anthropometric data collected from Taiwanese women were input in the system to account for anthropometric factors in the population. To aid in posture entry, an inverse kinematics algorithm was developed using 3DSSPP, based on the preferred postures of people manipulating loads by using known hand positions. In this study, the participants' maximal strength was considered the load on their hands, and L4/L5 disc compression forces and shoulder reactive moments were thus calculated. 2.4. Experimental design This study collected the lifting strengths of 46 participants (23 workers and 23 novices), using 15 exertion heights. The heights, ranging from 10 to 150 cm and spaced 10 cm apart, were established using the SLST. Consequently, 2070 strength data (46 participants 15 heights 3 repetitions) were collected. We also estimated the compression forces on the L4/L5 disc and shoulder reactive moments based on randomly sampled participants (12 workers and 12 novices) to clarify the posture strategies based on experience and exerting height.
All participants were familiarized with the experimental procedures; they warmed up for at least 10 min before the data were collected. The participants engaged in simple warm-up stretches, as suggested by Knudson and Noffal (2005), to avoid injury. During the experiment, the participants wore light clothing and rubber gym shoes. They were requested to perform all maximum strengths in a random height arrangement with freestyle but symmetrical lifting postures. The strength-testing procedure was performed according to the methodology used by Ayoub et al. (1978) and Chaffin (1975). All participants were asked to perform each task combination by using their maximum vertical lifting strength for a minimum of three repetitions. Each strength measurement was repeated until three readings were obtained and were consistent within a range of 10%. The maximum strength among the three readings and the corresponding posture data were analyzed. A period of 5 s was chosen as sufficient time to build and maintain a constant force, and was sufficiently brief to endure without perceiving muscular fatigue. A minimum rest period of 2 min was required between successive trials, and each participant was tested for approximately 5e12 min for three repetitions. The participants were also restricted to performing the tests 2 h every half day to avoid a carry-over effect. During the test, participants were asked to grasp the handle bar with their palms upwards, and their hands spaced approximately 48 cm apart. The horizontal-projection distance between the tip of the shoes of the participants and the handle bar was set at 15 cm. As observed in a recent field study, this foot position is the most advantageous testing posture to apply force for participants for full-range heights (Chen, 2008; Chen et al., 2011). 2.6. Statistical analysis In this study, a factorial repeated-measures two-way analysis of variance (ANOVA) was used to assess the differences in muscular strengths, body joint angles, and estimated body loads, between work experience (experienced workers and novices) and exertion heights (10e150 cm in increments of 10 cm). Duncan's multiplerange test (MRT) was used for post hoc comparisons. Each participant was considered a block. To understand how strength value generated by each participant affected the body loads, the estimated spinal load and shoulder moment were normalized using corresponding strength at each lifting height. Because of the differences in body mass between the two groups, the strength values of all participants were also adjusted according to body mass. For identifying the differences in strength values, body joint angles and corresponding biomechanical loads at each height level between the two groups, an independent t-test was used. The experimental data were analyzed using SPSS 19.0 at a significance level of .05. 3. Results We determined the lifting strength of 46 female experienced and inexperienced participants at 15 exertion heights, between 10 and 150 cm. Except for body mass and age, the two groups did not differ anthropometrically or in standard static strength. Among the 46 participants, the testing postures of randomly sampled participants (12 experienced workers and 12 novices) were videotaped. The mean (SD) age, stature, and body mass were 27.3 (5.6) years, 158.8 (5.9) cm, and 59.3 (10.2) kg for the experienced subgroup, and 22.4 (2.9) years, 161.1 (3.9) cm, and 51.9 (6.1) kg for the inexperienced subgroup, respectively. Neither experienced nor inexperienced participants exhibited results that differed from the group
610
Y.-L. Chen / International Journal of Industrial Ergonomics 44 (2014) 607e614
data. A power analysis revealed that the 24 subgroup participants exhibited a statistical power of approximately 0.95, with a 0.05 alpha level for the test combinations used in the study. 3.1. ANOVA on strength values The maximal lifting strength values were significantly influenced by work experience (F (1, 630) ¼ 39.05, p < .001) and exerting height (F (14, 630) ¼ 5.27, p < .001), and the two-way interaction (experience height: F (14, 630) ¼ 3.62, p < .001) also exerted a significant effect. A one-way ANOVA further revealed that lifting height influenced only the force exerted by novices (F (14, 308) ¼ 10.78, p < .001). Conversely, experienced workers' strength was unrelated to lifting heights (F (14, 308) ¼ 0.58, p > .05). The results of Duncan's MRT are presented in Table 2. 3.2. Differences in strength between experienced workers and novices
Fig. 2. Mean (SD) lifting strength values at 15 exertion heights for both experienced worker and novice participants (*p < .05, **p < .01, ***p < .001).
The strength profiles at various exertion heights are presented in Fig. 2. The strength averaged across 15 heights was 162.9 ± 59.8 N and 187.0 ± 57.9 N for experienced workers and novices, respectively. Differences in strengths between the experienced workers and novices were observed at lower heights, whereas the strengths were nearly identical at higher heights. Fig. 2 also revealed the ttest results of maximum strengths at various exertion heights between the two groups. As presented in the figure, at lower heights (50 cm), experienced worker' strengths were lower than novices' strengths (with differences of 50.7 Ne71.1 N, and t ranged from 3.18 to 3.95; all p < .05). When the heights were higher than 50 cm, no significant differences in strength were observed between the groups. When the strengths were adjusted based on the body mass of each participant, the discrepancy in strength profiles between the two groups remained similar (with differences of 0.88e1.78 N/kg when heights were 60 cm, and t ranged from 3.26 to 5.34; all p < .005).
angles at lower heights (i.e., 10e50 cm) in the novices' tests (Fig. 3). Fig. 4 presented the postures for exerting at heights of 10, 50, 70, and 120 cm of one representative participant from each group. The independent t-test results revealed that, when lifting at 10 cm and 50 cm, the knees, elbows, and shoulders of the experienced workers were more flexed than those of the novices. At 70 cm, the knees, hips, and elbows of the experienced workers were more extended than those of the novices (all p < .05). At 120 cm, the experienced workers flexed their upper limbs more than the novices did. Posture analysis revealed that 7 out of 12 sampled participants at a lifting height of 60 cm, and 5 out of 12 participants at a lifting height of 70 cm, dramatically changed their lifting postures from a squat to a standing arm lift. The joint angles at the height of 60 cm presented in Fig. 3 represent the average posture of the 12 sampled workers. None of the workers adopted the posture associated with these angles.
3.3. Posture analyses
3.4. The estimated L4/L5 disc compression forces and shoulder reactive moments
The ANOVA results revealed that four joint angles were all significantly affected by work experience (F (1, 630) ¼ 31.3 e 54.2, all p < .001) and exerting height (F (14, 630) ¼ 10.4 e 27.1, all p < .001). Comparisons of the postures adopted when lifting at maximum strength indicated larger knee, hip, shoulder, and elbow
Table 2 Results of Duncan's MRT on mean (SD) lifting strengths at 15 exertion heights in the two participant groups. Heights (cm)
Lifting strengths (N)
10 20 30 40 50 60 70 80 90 100 110 120 130 140 150
170.3 162.9 153.1 151.6 164.5 179.0 184.1 175.6 164.7 157.3 153.6 161.3 164.0 150.9 151.0
Experienced workers (60.1) (58.8) (57.6) (66.6) (63.0) (58.7) (55.2) (65.2) (57.0) (59.4) (59.3) (62.3) (67.0) (56.1) (57.4)
Duncan groups
Novices
241.4 230.1 223.1 217.6 215.2 202.7 197.1 187.5 167.6 153.7 159.7 163.8 155.3 143.7 144.9
(62.0) (61.3) (63.3) (56.1) (55.5) (54.2) (52.0) (47.3) (43.2) (38.3) (37.8) (42.5) (30.1) (31.7) (42.4)
Duncan groups
The ANOVA results revealed that work experience significantly influenced the compression forces (F (1, 630) ¼ 40.5, p < .001) and shoulder moments (F (1, 630) ¼ 34.9, p < .001). Exerting height also revealed similar effects (compression force: F (14, 630) ¼ 6.1, p < .001; shoulder moment: F (14, 630) ¼ 4.7, p < .001). Figs. 5 and 6 present the comparisons of estimated compression forces on the L4/L5 disc and shoulder reactive moments between the two groups, respectively. The mean workers' compression forces on the L4/L5 disc at 15 heights were all less than 1600 N, whereas those of the novices ranged from 460 N to 5114 N. The results of the t-test revealed that, when the exertion height was 70 cm, the compression force of the novices was higher than that of the workers (all p < .01), with differences of 952 Ne3575 N. With the exception of the height range of 60e90 cm, which exhibited identical shoulder moments between the two groups, the experienced workers exhibited lower moment amounts (all p < .05), as displayed in Fig. 6. Fig. 7 further compares the ratios of L4/L5 disc compression forces and shoulder moments normalized by corresponding strengths between the two groups at the 15 exertion heights. At heights 70 cm, the experienced workers maintained compression force per strength value (kN/N) at lower levels than the novices did. A similar trend was observed when handling upper arm loads (N-m/N ¼ m, or moment-arm) at heights >100 cm.
Y.-L. Chen / International Journal of Industrial Ergonomics 44 (2014) 607e614
Elbow 200
200
160
150
120
611
Hip
100
Joint angles (degrees)
80
50
40
0
0 10 20 30 40
50 60 70 80 90 100 110 120 130 140 150
Shoulder 100 80
10 20
30 40 50 60
70 80 90 100 110 120 130 140 150
Knee
200 150
60 100
40
Novices
50
20
Experienced workers
0
0 10
20
30
40
50
60
70
80
90 100 110 120 130 140 150
Exertion heights (cm)
10
20 30
40
50
60 70
80
90 100 110 120 130 140 150
Exertion heights (cm)
Fig. 3. Comparison of the lifting postures of experienced workers and novices during testing at 15 exertion heights.
4. Discussion This study is the first to investigate the lifting strength associated with a full range of exertion heights between female experienced and inexperienced participants. As hypothesized, experienced workers' strengths at 15 lifting heights differed from inexperienced novices' and the results revealed that a discrepancy in strength existed during lifting at lower heights (50 cm)
120 cm
between the two groups. When the heights were >50 cm, no significant differences in strength were observed. This might explain the significant effect of the two-way interaction (i.e., experience height). The novices' strength values reduced with the height increase (ranging from 241.4 N at 10 cm to 143.7 N at 140 cm), whereas the experienced workers' strength values revealed relatively little fluctuation (ranging from 184.1 N at 70 cm to 150.9 N at 140 cm). Certain studies have suggested that maximum acceptable weight of lift (MAWL) patterns are similar between experienced workers and inexperienced novices (Garg and Saxena, 1979; Mital and Manivasagan, 1983). These results might not be completely comparable to our study, particularly at certain lifting heights. The MAWL might be more relevant to workers repeatedly performing submaximal lifts, rather than a one-time maximum effort. It is difficult to fully elucidate why a dramatic decrease in strength was observed among the novices but not among the experienced workers in this study. Our hypothesis of difference
70 cm
50 cm
10 cm
Experienced worker
Novice
Fig. 4. Comparison of the schematic exerting postures of a representative participant from each group at heights of 10, 50, 70 and 120 cm.
Fig. 5. Mean (SD) compression forces on the L4/L5 discs of experienced workers and novices during testing (*p < .05, **p < .01, ***p < .001).
612
Y.-L. Chen / International Journal of Industrial Ergonomics 44 (2014) 607e614
Fig. 6. Mean (SD) estimated shoulder moments of experienced workers and novices during testing (*p < .05, **p < .01, ***p < .001).
existing in lifting strategies between the two participant groups was verified and might explain why differing strength patterns of the two groups were observed. This might be the so-called survivor effect (Marras et al., 2006). A worker with a skilled technique and at least 2 years of experience with handling was considered an experienced worker, and was selected as a participant in our study. By contrast, 23 university students were recruited in this study as novices without lifting experience. Workers with daily work experience might have learned a safety criterion by shifting their postures to avoid overloading when exerting maximum force. As shown in Fig. 3, experienced workers' knee and hip joints were more flexed at heights of 10e50 cm and more extended at heights of 70e100 cm than novices' joints were. When exerting at lower heights, workers chose a nearly knee-flexion strategy, whereas the
Fig. 7. Comparison of the ratios of the L4/L5 disc compression forces and shoulder moments normalized by corresponding strengths at 15 exertion heights for both groups (*p < .05, **p < .01, ***p < .001).
novices tended to choose a more stooped strategy (Fig. 4). Plamondon et al. (2010) compared groups of expert and novice handlers during a horizontal transferring task, and observed that their lifting postures differed; experts bent their lumbar regions less but bent their knees more. However, the external peak L5/S1 moments generally did not differ between the groups (Plamondon et al., 2010). Our result was somewhat similar to that of Plamondon et al.; the spinal loading differed between the groups in this study. A stooped lifting posture, which can result in the generation of greater forces, imposes more compressive force on the lumbarspinal discs (Anderson and Chaffin, 1986; Straker, 2003). Fig. 4 indicates that the moment-arm from the handlebars to the novice participants' lower backs was visibly longer than that of the experienced workers. This might have led the experienced workers to be more safety-conscious than the novices (Fig. 5), resulting in lower strength values and corresponding spinal loads for the workers. Unlike the experienced workers, who adopted considerably different postures between lifts at heights of 50e70 cm, the novice participants adopted postures that varied gradually with increased height (Fig. 4). Once the experienced workers could not adopt a deep-squat strategy, they adopted a nearly erect posture. The semisquat technique that has often been suggested by previous researchers (Garg and Saxena, 1979; Straker, 2003) was not used by any experienced worker in our study. Our results indicated that none of the experienced workers adopted postures associated with the averaged angles shown in Fig. 3. Lortie (2002) indicated that experienced workers adopted moderate-knee flexion to avoid unbalancing the body. This might reflect the actual daily shelf-work exposure of these workers. Most handling tasks in warehouses are typically conducted at submaximal weights to repetitively replenish numerous goods on shelves in a workday. A more stable and less strenuous posture becomes more critical to fit the demands of the task. This might have resulted in experienced workers having exhibited consistent strengths when the heights were increased during the tests. Once the positions were higher than 100 cm, the lower body posture (i.e., the knee and hip joint angles) adopted by the experienced workers and novices was relatively similar (Fig. 3). In other words, with this nearly erect body posture, their strength was determined using their upper extremities (Kumar, 1991). Chen (2000) observed that experienced workers tended to flex and fix their upper arms and move their trunks forward when lifting at higher positions. This stiffened upper-limbs method might reduce the stress on the shoulder and elbow joints by shortening the moment-arm (i.e., the distance from the load to the joints). As presented in Figs. 3 and 4, the experienced workers adopted a stiffer strategy with more flexed elbows and shoulders during lifting. Although no differences in strength at the higher heights were observed between the two groups, the workers inevitably sustained lower loads on the upper limbs than the novices did by shortening their moment-arms. To further elucidate the effects of the posture strategy adopted by the experienced workers for the lifting, the biomechanical loads were normalized using the strength values for each height. When the L4/L5 disc compressions and shoulder moments were individually divided by the corresponding strengths, the ratios reflected the differences in the posture strategies. As shown in Fig. 7, when heights were <90 cm, the experienced workers tended to adopt a safe posture to minimize the load pressure on the lower back; in other words, to effectively maintain the compression force per strength value (KN/N) at lower heights. Marras et al. (2006) suggested that the increased spinal load levels exhibited by novices are due to underdeveloped motor control programming, and that the biomechanical risk is greatly reduced with experience. The
Y.-L. Chen / International Journal of Industrial Ergonomics 44 (2014) 607e614
experienced workers' body mass was heavier than that of the novices (mean 7.0 kg), which increased the spinal load because of the contribution from the upper body mass. When the exertion heights were higher than 90 cm, the workers attempted to minimize the unitary upper arms loads. This result might be due to the experienced workers adopting safer strategies to protect themselves when they were confronted with the tasks. The result might be critical in training novices to use a safe strategy. Although the body masses among the experienced workers were heavier than those among the student participants in this study, the experienced workers exhibited relatively low strengths at low heights. When the strengths were adjusted using the body mass of each participant, the discrepancy in strength profiles between the two groups remained similar. Singh et al. (2009) indicated that obesity does not affect the MAWL and suggested that existing MAWL data can be used to accommodate both general and obese workers. However, further studies based on the biomechanical and physiological approaches are required to provide more complete understanding of whether heavier body masses affected the posture that workers adopted. In this study, we also examined whether results regarding the effects of these task variables on men were also applicable to women. The strength averaged across 15 heights was 162.9 N and 187.0 N for female experienced workers and novices, respectively. These values were approximately 52.5% of those of male experienced workers and 62.7% of those of male novices in the study of Chen et al. (2011) and were consistent with the results presented in other studies (Garg and Ayoub, 1980; Xiao et al., 2005). The testing protocol adopted in this study was identical to the study of Chen et al. (2011), but the results differed. In this study, when lifting at higher positions, no differences in strength were observed between the two groups (Fig. 2). This result did not match that of the Chen et al. study (2011) of male participants. One possible reason was the posture strategy. Regarding men, differences in lifting strengths between the male experienced workers and novices only existed at heights of 100e120 cm, in which the novices were 4.57e7.61 kg (44.8e74.7 N) weaker than the experienced workers. Differences between these studies might be attributable to the fact that women are generally shorter (Chen et al., 2011) and weaker than men, and therefore, exertion heights relative to the participants' heights differed. Sex, lifting height, and work experience might have interactively influenced the participants' strength and the posture strategy. The result also indicated that variations in lifting strengths and postures between male and female novice participants might be attributable to their lack of experience. These results warrant further clarification. This study had limitations. The assumptions used by 3DSSPP were used to determine the biomechanical measures, as well as the fact that this analysis was only performed on a subset of participants (i.e., 12 workers and 12 novices). Twenty-three university students were recruited in this study as novice participants without any lifting experience. Moreover, warehouse tasks involve various degrees of muscular efforts, but this study examined only maximum lifting strengths. Under distinct task conditions, criteria for reducing the physical strain of lifters might differ. Why experienced workers adopted particular strategies based on their posture changes was not entirely clear in this study. They might have a host of expectations and strategies to cope with their particular jobs. This simulated experimental test might approximate the whole work experience in an overall evaluation. Because lifting tasks might vary between repeatedly submaximal and nearmaximal (such as a female employee lifting a box containing a dozen beverages), the experienced workers might have learned to protect themselves from injury or minimize fatigue. Furthermore, the experienced workers adopted a flexed knee position, which
613
might have resulted from a considerable amount of time working with low shelving, leading to the deep squat postures adopted by the workers at heights 50 cm. The experienced workers at heights 50 cm were either unable or unwilling to lift as hard as the novices were, even though they were asked to exert maximal force. This resulted in lower loading on the back of the experienced workers compared with the novices. Although lifting is a dynamic activity, the analysis of static lifting strengths provided by this study offers a valuable reference for lifting task design and employee training, regarding the weighting of various shelves and effective posture training on lifting. 5. Conclusion We demonstrated that a discrepancy exists in lifting strengths between experienced and inexperienced female participant groups at various heights. Lifting strengths of experienced workers were significantly lower than those of novices when heights were 50 cm and also exhibited relatively little fluctuation compared with that of the novices. The workers tended to adopt a consistently deep squat at lower heights and a nearly erect and stiffened-arm posture at higher heights. However, the results of this study were dissimilar to those of Chen et al. (2011), who collected data on Chinese men. These results implied that the experienced workers might have learned to change their postures to effectively avoid overloading. Acknowledgment This study was partially supported by funding from the National Science Council, Taiwan, ROC, grant no. NSC97-2221-E-131-019. This funding is gratefully acknowledged. References Anderson, C.K., Chaffin, D.B., 1986. A biomechanical evaluation of five lifting techniques. Appl. Ergon. 17, 2e8. Agrawal, K.N., Singh, R.K.P., Satapathy, K.K., 2009. Isometric strength of agricultural workers of Meghalaya: a case study of an Indian population. Int. J. Ind. Ergon. 39, 919e923. Authier, M., Lortie, M., Gagnon, M., 1996. Manual handling techniques: comparing novices and experts. Int. J. Ind. Ergon. 17, 419e429. Ayoub, M.M., Bethea, N.J., Deivanayagam, S., Asfour, S.S., Bakken, G.M., Liles, D., Mital, A., Sherif, M., 1978. Determination and Modeling of Lifting Capacity. Final Report. HEW [NIOSH], Cincinnati, OH. Baril-Gingras, G., Lortie, M., 1995. The handling of objects other than boxes: univariate analysis of handling techniques in a large transport company. Ergonomics 38, 905e925. Chaffin, D.B., 1975. Ergonomics guide for the assessment of human static strength. AIHA J. 36, 505e511. Chen, C.J., 2008. Investigation of Musculoskeletal Disorders for the Warehouse Workers in Taiwan. Master Dissertation. Ming Chi University of Technology, Taipei, Taiwan. Chen, Y.L., 2000. Optimal lifting techniques adopted by Chinese men when determining their maximum acceptable weight of lifting. AIHA J. 61, 642e648. Chen, Y.L., Lee, Y.C., Chen, C.J., 2011. Differences in lifting strength profiles between experienced workers and novices at various exertion heights. Int. J. Ind. Ergon. 41, 53e58. Council for Economic Planning and Development (CEPD), 2011. Guidelines and Action Plans for Service Industry Development. Policy Report, Taipei, Taiwan, R.O.C. Gagnon, M., Plamondon, A., Gravel, D., Lortie, M., 1996. Knee movement strategies differentiate expert from novice workers in asymmetrical manual materials handling. J. Biomech. 29, 1445e1453. Garg, A., Ayoub, M.M., 1980. What criteria exist for determining how much load can be lifted safely? Hum. Factors 22, 475e486. Garg, A., Saxena, U., 1979. Effects of lifting frequency and technique on physical fatigue with special reference to psychophysical methodology and metabolic rate. AIHA J. 49, 894e903. Hodder, J.N., MacKinnon, S.N., Ralhan, A., Keir, P.J., 2010. Effects of training and experience on patient transfer biomechanics. Int. J. Ind. Ergon. 40, 282e288. Hooftman, W.E., van der Beek, A.J., van de Wal, B.G., Knol, D.L., Bongers, P.M., Burdorf, A., van Mechelen, W., 2009. Equal task, equal exposure? Are men and
614
Y.-L. Chen / International Journal of Industrial Ergonomics 44 (2014) 607e614
women with the same tasks equally exposed to awkward working postures? Ergonomics 52, 1079e1086. Keir, P.J., MacDonell, C.W., 2004. Muscle activity during patient transfers: a preliminary study on the influence of lift assists and experience. Ergonomics 47, 296e306. Knudson, D., Noffal, G., 2005. Time course of stretch-induced isometric strength deficits. Eur. J. Appl. Physiol. 94, 348e351. Kumar, S., 1991. Arm lift strength in work space. Appl. Ergon. 22, 317e328. Kuorinka, I., Lortie, M., Gautreau, M., 1994. Manual handling in warehouses: the illusion of correct working postures. Ergonomics 37, 655e661. Lee, T.H., 2004. Static lifting strengths at different exertion heights. Int. J. Ind. Ergon. 34, 263e269. cision d'action. Le Travail Lortie, M., 2002. Manutention: prise d'information et de Humain 65, 193e216. Marras, W.S., Parakkat, J., Chany, A.M., Yang, G., Burr, D., Lavender, S.A., 2006. Spine loading as a function of lifting frequency, exposure duration, and work experience. Clin. Biomech. 21, 345e352. Miller, A.E.J., MacDougall, J.D., Tarnopolsky, M.A., Sale, D.G., 1993. Gender differences in strength and muscle fiber characteristics. Eur. J. Appl. Physiol. 66, 254e262. Mital, A., 1987. Patterns of differences between the maximum weights of lift acceptable to experienced and inexperienced materials handlers. Ergonomics 30, 1137e1147. Mital, A., Garg, A., Karwowski, W., Kumar, S., Smith, J.L., Ayoub, M.M., 1993. Status in human strength research and application. IIE Trans. 25, 57e69.
Mital, A., Manivasagan, I., 1983. Maximum acceptable weight of lift as a function of material density, center of gravity location, hand preference, and frequency. Hum. Factors 25, 33e42. Murrel, K.F.H., 1971. Ergonomics: Man and His Working Environment. Chapman and Hall, London. Parakkat, J., Yang, G., Chany, A.M., Burr, D., Marras, W.S., 2007. The influence of lift frequency, lift duration and work experience on discomfort reporting. Ergonomics 50, 396e409. re, C., Salazar, E., 2010. Biomechanical Plamondon, A., Denis, D., Delisle, A., Larivie differences between expert and novice workers in a manual material handling task. Ergonomics 53, 1239e1253. Plamondon, A., Lariviere, C., Delisle, A., Denis, D., Gagnon, D., 2012. Relative importance of expertise, lifting height and weight lifted on posture and lumbar external loading during a transfer task in manual material handling. Ergonomics 55, 87e102. Singh, D., Park, W., Levy, M.S., 2009. Obesity does not reduce maximum acceptable weights of lift. Appl. Ergon. 40, 1e7. Straker, L.M., 2003. A review of research on techniques for lifting low-lying objects: 2. Evidence for a correct technique. Work 20, 83e96. Xiao, G., Lei, L., Dempsey, P.G., Lu, B., Liang, Y., 2005. Isometric muscle strength and anthropometric characteristics of a Chinese sample. Int. J. Ind. Ergon. 35, 674e679. Yates, J.W., Kamon, E., Rodgers, S.H., Champney, P.C., 1980. Static lifting strength and maximal isometric voluntary contractions of back, arm and shoulder muscles. Ergonomics 23, 37e47.
Contents lists available at ScienceDirect
International Journal of Industrial Ergonomics journal homepage: www.elsevier.com/locate/ergon
Effects of work experience and exertion height on static lifting strengths and lift strategies of experienced and novice female participants Yi-Lang Chen* Department of Industrial Engineering and Management, Ming Chi University of Technology, 84, Gungjuan Rd., Taishan Dist., New Taipei 24301, Taiwan, ROC
a r t i c l e i n f o
a b s t r a c t
Article history: Received 21 June 2013 Received in revised form 12 June 2014 Accepted 13 June 2014 Available online
In this study, 46 female experienced workers and inexperienced novices (23 each) were recruited to determine their maximum lifting strength at 15 exertion heights between 10 and 150 cm from the floor. The results revealed that the experienced workers' strengths at all 15 heights exhibited relatively little fluctuation, and were approximately 50e70 N lower than those of novices when heights were 50 cm. No differences in strengths were observed at 60e150 cm between the groups. The experienced workers tended to adopt a consistently deep squat at lower heights (50 cm) and a more erect posture with stiffened arms at higher heights (70 cm), resulting in lower L4/L5 disc compression forces and shoulder moments than in novices, respectively. In contrast to the lifting techniques adopted by experienced workers to effectively avoid overloading, the findings suggest that novice female workers who lack experience should be cautious and trained for performing lifting tasks. Relevance to industry: This study demonstrated that a discrepancy exists in lifting strengths between experienced and inexperienced female handlers at various heights. Findings from this study offers a reference for lifting task design and effective posture training on lifting, regarding the weighting of various shelves. © 2014 Elsevier B.V. All rights reserved.
Keywords: Lifting postures Experienced workers and novices Lifting heights
1. Introduction Manual handling remains prevalent in Taiwan, and over 500,000 workers are employed in jobs related to warehousing (CEPD, 2011). Most tasks in warehouses, distribution centers, and various retail establishments involve manual materials handling and strength exertion at various heights of lift. These tasks typically involve submaximal weights; in other words, the weights handled are light to moderate and therefore do not require only male employees. A recent study of Taiwanese warehouse workers revealed that the lower back, shoulders, and arms were regions of bodily discomfort for both male and female workers, particularly for novice workers (Chen, 2008). This might have resulted from the warehousing tasks, which were highly repetitive but diverse in lifting heights, that these workers were asked to perform. One factor affecting lifting strength is the exertion height; in other words, the height at which the lift is initiated (Yates et al., 1980; Mital et al., 1993; Lee, 2004; Chen et al., 2011). This is
* Tel.: þ886 2 29089899x3103; fax: þ886 2 29085900. E-mail address: [email protected]. http://dx.doi.org/10.1016/j.ergon.2014.06.004 0169-8141/© 2014 Elsevier B.V. All rights reserved.
particularly true in service warehouse stores, which economize and maximize storage space. Lifting strength capacity most likely changes with height and thus placing a person at higher percentages of their maximum in some locations than others (Yates et al., 1980). The closer they are to their maximum, the greater their potential risk for injury (Mital et al., 1993). Investigations on lifting strength have typically been confined to specific height levels (e.g., floor, knuckle, and shoulder); however, this is not applicable to warehousing tasks. Height of lifting also influences posture, which, in turns, affects internal loads on potentially vulnerable tissues (Chen, 2000; Plamondon et al., 2012). Therefore, it is critical to understand the effect of height on lifting capacity and the corresponding body loads. Much of the literature has reported that the handling techniques of highly skilled workers differ substantially from those of novices (Authier et al., 1996; Gagnon et al., 1996). Field studies have also revealed that workers use methods other than those traditionally recommended (Kuorinka et al., 1994; Baril-Gingras and Lortie, 1995; Authier et al., 1996). However, certain researchers have determined that experienced workers possess advantages; for a given task, experienced workers generally employ lower biomechanical spinal loads (Marras et al., 2006), fewer back-muscle
608
Y.-L. Chen / International Journal of Industrial Ergonomics 44 (2014) 607e614
activations (Keir and MacDonell, 2004), a narrower lumbar-spine range of motion (Hodder et al., 2010), higher subjective discomfort thresholds (Parakkat et al., 2007) and psychophysically accepted lifting weights (Mital, 1987), and a greater knee flexion during ground-level task (Plamondon et al., 2010, 2012) than novices. Conversely, Mital and Manivasagan (1983) indicated that task variables similarly influenced the maximum lifting capacities of experienced workers and college students, and controversy in this regard might be attributable to the distinct experimental settings adopted by these studies. Warehouses in Taiwan have high employee turnover rates, and recruit new full-time and part-time workers weekly (Chen, 2008). These inexperienced novice workers require more attentiveness to avoid injury resulting from daily tasks. Recently, Chen et al. (2011) systemically assessed the lifting strength of male experienced workers and novices at full-range heights, and observed that the average strength lifted upward by novices was 4.57e7.61 kg lower than that of experienced workers when lifting at 100e120 cm heights. The workers' upward strengths, contrary to expectations, remained nearly unchanged throughout all 15 full-range heights. In that study, the postures adopted by the workers were also highly differentiated from that of the novices when performing near-floor positions, but their strengths were equivalent. Experienced workers tended to adopt a safer (i.e., more flexed knees) and more skillful technique than novices did to generate force, which might result in lower spinal loads during lifting. However in that study, only male experienced workers and inexperienced university students were compared. Previous studies have demonstrated that the inherent difference in muscular characteristics between sexes (Xiao et al., 2005; Agrawal et al., 2009). Garg and Ayoub (1980) noted that the maximum lifting capacity of women was 60e70% that of men, and an earlier study by Murrel (1971) also suggested that women were able to exert approximately two-thirds the force of men. However, the ratio of isometric strength between Chinese men and women was nearly 50% (Xiao et al., 2005). Miller et al. (1993) observed that sex differences in muscle strength were primarily confined to muscles in the upper extremities, particularly the shoulders. Hooftman et al. (2009) indicated that even if men and women perform the same tasks, they do not perform these tasks in the same manner, which might also lead to differences in exposure. This can be attributed to a poorer ergonomic fit of the workplace to women than to men, or because men and women choose to perform the same task using distinct strategies. Lifting tasks distributed to male and female workers in supermarkets were observed to be identical regarding shelf heights and load weights. This might imply a higher risk for women than for men, because these tasks place the women at a disadvantage compared with men when lifting. We therefore examined the effect of height on lifting capacity and the corresponding body loads for experienced and inexperienced female participants recruited from a large supermarket and a university, respectively. We hypothesized that experienced female participants' strengths and lifting strategies at 15 lifting heights would differ from inexperienced female participants' because of the daily workplace practices where the experienced group had worked. We also examined whether the results regarding the effects of these tasks on men were also applicable to women.
volunteered for this study, and received an hourly wage for completing all of the test conditions. Twenty-three workers with a minimum of 2-years of experience were selected as experienced participants. They were selected from a large supermarket, and their duties consisted primarily of varied shelf-replenishing tasks. Another 23 novice participants (with no manual material-handling experience) were recruited from among university students, based on previous studies (Mital, 1987; Marras et al., 2006; Parakkat et al., 2007). Informed consent was obtained from all participants in accordance with the regulations. The Ethics Committee of Chang Gung Memorial Hospital of Taiwan approved this study. The mean (standard deviation; SD) age, stature, and body mass were 28.6 (6.4) years, 158.9 (5.3) cm, and 58.3 (10.2) kg for the experienced group, and 22.3 (3.2) years, 160.2 (4.1) cm, and 51.3 (6.5) kg for the inexperienced group, respectively. Detailed information, anthropometric data, and isometric muscular strengths for the 46 participants are listed in Table 1. The isometric strength measurement procedure was based on the studies of Ayoub et al. (1978) and Mital (1987). 2.2. Experimental apparatus Lifting strengths were measured using the Static-Lifting Strength Tester (SLST), as shown in Fig. 1. The SLST, with incremental height settings for measuring strengths, consists of a standing platform, a steel frame with 15 positioning holes (ranging from 10 to 150 cm in increments of 10 cm), a sliding height stopper along the frame, and a 55-cm handle bar (diameter: 3.5 cm) attached to the stopper. Force applied vertically to the handle bar was measured by a load cell connected to the bar that was firmly attached to the stopper; the strength signal (40 Hz) was then transferred to an A-D converter and a digital readout unit (Jackson Strength Evaluation System, JSES Model 32628, USA). The A-D converter was calibrated prior to the strength testing against known weights. 2.3. Body posture recording and biomechanical analysis The testing postures of randomly sampled participants (12 workers and 12 novices) were videotaped in this study. Six adhesive reflective markers were attached to the right side of their ulna
Table 1 Participant data for the experienced workers and novices (23 in each group). Variables
Mean Age (years)* Stature (cm) Body mass (kg)* Acromial height (cm) Elbow height (cm) Knuckle height (cm) Hip height (cm) Knee height (cm) Hand length (cm) Hand width (cm) Isometric strengths (N)a Composite strength Back strength Shoulder strength Arm strength
2. Methods 2.1. Participants *
Forty-six experienced female workers and novice participants (23 of each), with no history of musculoskeletal disorders,
Experienced workers
28.6 158.9 58.3 127.8 95.5 68.9 81.7 42.1 14.9 7.1 599.8 497.8 282.2 218.5
Novices S.D. 6.4 5.3 10.2 5.03 3.81 4.62 6.30 3.33 0.64 0.5 132.3 128.4 68.6 51.9
Mean 22.3 160.2 51.3 128.0 96.9 69.4 83.0 43.6 14.9 7.0 611.5 521.4 267.5 231.3
S.D. 3.2 4.1 6.5 4.88 2.63 5.03 6.15 2.32 0.67 0.4 138.2 136.2 80.4 49.0
Significantly different at p < .05. a Testing procedure was based on the studies of Ayoub et al. (1978) and Mital (1987).
Y.-L. Chen / International Journal of Industrial Ergonomics 44 (2014) 607e614
609
2.5. Experimental procedure
Fig. 1. Device used to test lifting strength and schematic testing posture in the study, at a height of 20 cm.
condyle, the lateral epicondyle of the humerus, the acromial shelf, the femoral greater trochanter, the lateral femoral condyle, and the lateral malleolus of the fibula. In the study, the joint angles were all defined as the inter-joint angles between the two adjacent segments (Fig. 1). The digital video camera (Qualisys MacReflex, Sweden) was placed at a distance of 5 m from the participants, and was used to record the positions of each joint marker when participants performed the strength tests. The camera heights were consistently adjusted to orthogonally align with the participants' hip positions to eliminate as many distortion errors as possible resulting from the varied testing heights. Image-processing software (Qualisys MacReflex, Sweden) was then used to calculate the relevant joint angles. When the participants performed the test tasks, the L4/L5 disc compression forces and shoulder reactive moments were estimated by the 3D Static Strength Prediction Program™ (3DSSPP, V6.0.5), which was developed by the Center for Ergonomics at the University of Michigan. Participants' joint angles, anthropometric data, and strength values were used as inputs in the program to estimate the lumbar compression forces and shoulder moments. Anthropometric data collected from Taiwanese women were input in the system to account for anthropometric factors in the population. To aid in posture entry, an inverse kinematics algorithm was developed using 3DSSPP, based on the preferred postures of people manipulating loads by using known hand positions. In this study, the participants' maximal strength was considered the load on their hands, and L4/L5 disc compression forces and shoulder reactive moments were thus calculated. 2.4. Experimental design This study collected the lifting strengths of 46 participants (23 workers and 23 novices), using 15 exertion heights. The heights, ranging from 10 to 150 cm and spaced 10 cm apart, were established using the SLST. Consequently, 2070 strength data (46 participants 15 heights 3 repetitions) were collected. We also estimated the compression forces on the L4/L5 disc and shoulder reactive moments based on randomly sampled participants (12 workers and 12 novices) to clarify the posture strategies based on experience and exerting height.
All participants were familiarized with the experimental procedures; they warmed up for at least 10 min before the data were collected. The participants engaged in simple warm-up stretches, as suggested by Knudson and Noffal (2005), to avoid injury. During the experiment, the participants wore light clothing and rubber gym shoes. They were requested to perform all maximum strengths in a random height arrangement with freestyle but symmetrical lifting postures. The strength-testing procedure was performed according to the methodology used by Ayoub et al. (1978) and Chaffin (1975). All participants were asked to perform each task combination by using their maximum vertical lifting strength for a minimum of three repetitions. Each strength measurement was repeated until three readings were obtained and were consistent within a range of 10%. The maximum strength among the three readings and the corresponding posture data were analyzed. A period of 5 s was chosen as sufficient time to build and maintain a constant force, and was sufficiently brief to endure without perceiving muscular fatigue. A minimum rest period of 2 min was required between successive trials, and each participant was tested for approximately 5e12 min for three repetitions. The participants were also restricted to performing the tests 2 h every half day to avoid a carry-over effect. During the test, participants were asked to grasp the handle bar with their palms upwards, and their hands spaced approximately 48 cm apart. The horizontal-projection distance between the tip of the shoes of the participants and the handle bar was set at 15 cm. As observed in a recent field study, this foot position is the most advantageous testing posture to apply force for participants for full-range heights (Chen, 2008; Chen et al., 2011). 2.6. Statistical analysis In this study, a factorial repeated-measures two-way analysis of variance (ANOVA) was used to assess the differences in muscular strengths, body joint angles, and estimated body loads, between work experience (experienced workers and novices) and exertion heights (10e150 cm in increments of 10 cm). Duncan's multiplerange test (MRT) was used for post hoc comparisons. Each participant was considered a block. To understand how strength value generated by each participant affected the body loads, the estimated spinal load and shoulder moment were normalized using corresponding strength at each lifting height. Because of the differences in body mass between the two groups, the strength values of all participants were also adjusted according to body mass. For identifying the differences in strength values, body joint angles and corresponding biomechanical loads at each height level between the two groups, an independent t-test was used. The experimental data were analyzed using SPSS 19.0 at a significance level of .05. 3. Results We determined the lifting strength of 46 female experienced and inexperienced participants at 15 exertion heights, between 10 and 150 cm. Except for body mass and age, the two groups did not differ anthropometrically or in standard static strength. Among the 46 participants, the testing postures of randomly sampled participants (12 experienced workers and 12 novices) were videotaped. The mean (SD) age, stature, and body mass were 27.3 (5.6) years, 158.8 (5.9) cm, and 59.3 (10.2) kg for the experienced subgroup, and 22.4 (2.9) years, 161.1 (3.9) cm, and 51.9 (6.1) kg for the inexperienced subgroup, respectively. Neither experienced nor inexperienced participants exhibited results that differed from the group
610
Y.-L. Chen / International Journal of Industrial Ergonomics 44 (2014) 607e614
data. A power analysis revealed that the 24 subgroup participants exhibited a statistical power of approximately 0.95, with a 0.05 alpha level for the test combinations used in the study. 3.1. ANOVA on strength values The maximal lifting strength values were significantly influenced by work experience (F (1, 630) ¼ 39.05, p < .001) and exerting height (F (14, 630) ¼ 5.27, p < .001), and the two-way interaction (experience height: F (14, 630) ¼ 3.62, p < .001) also exerted a significant effect. A one-way ANOVA further revealed that lifting height influenced only the force exerted by novices (F (14, 308) ¼ 10.78, p < .001). Conversely, experienced workers' strength was unrelated to lifting heights (F (14, 308) ¼ 0.58, p > .05). The results of Duncan's MRT are presented in Table 2. 3.2. Differences in strength between experienced workers and novices
Fig. 2. Mean (SD) lifting strength values at 15 exertion heights for both experienced worker and novice participants (*p < .05, **p < .01, ***p < .001).
The strength profiles at various exertion heights are presented in Fig. 2. The strength averaged across 15 heights was 162.9 ± 59.8 N and 187.0 ± 57.9 N for experienced workers and novices, respectively. Differences in strengths between the experienced workers and novices were observed at lower heights, whereas the strengths were nearly identical at higher heights. Fig. 2 also revealed the ttest results of maximum strengths at various exertion heights between the two groups. As presented in the figure, at lower heights (50 cm), experienced worker' strengths were lower than novices' strengths (with differences of 50.7 Ne71.1 N, and t ranged from 3.18 to 3.95; all p < .05). When the heights were higher than 50 cm, no significant differences in strength were observed between the groups. When the strengths were adjusted based on the body mass of each participant, the discrepancy in strength profiles between the two groups remained similar (with differences of 0.88e1.78 N/kg when heights were 60 cm, and t ranged from 3.26 to 5.34; all p < .005).
angles at lower heights (i.e., 10e50 cm) in the novices' tests (Fig. 3). Fig. 4 presented the postures for exerting at heights of 10, 50, 70, and 120 cm of one representative participant from each group. The independent t-test results revealed that, when lifting at 10 cm and 50 cm, the knees, elbows, and shoulders of the experienced workers were more flexed than those of the novices. At 70 cm, the knees, hips, and elbows of the experienced workers were more extended than those of the novices (all p < .05). At 120 cm, the experienced workers flexed their upper limbs more than the novices did. Posture analysis revealed that 7 out of 12 sampled participants at a lifting height of 60 cm, and 5 out of 12 participants at a lifting height of 70 cm, dramatically changed their lifting postures from a squat to a standing arm lift. The joint angles at the height of 60 cm presented in Fig. 3 represent the average posture of the 12 sampled workers. None of the workers adopted the posture associated with these angles.
3.3. Posture analyses
3.4. The estimated L4/L5 disc compression forces and shoulder reactive moments
The ANOVA results revealed that four joint angles were all significantly affected by work experience (F (1, 630) ¼ 31.3 e 54.2, all p < .001) and exerting height (F (14, 630) ¼ 10.4 e 27.1, all p < .001). Comparisons of the postures adopted when lifting at maximum strength indicated larger knee, hip, shoulder, and elbow
Table 2 Results of Duncan's MRT on mean (SD) lifting strengths at 15 exertion heights in the two participant groups. Heights (cm)
Lifting strengths (N)
10 20 30 40 50 60 70 80 90 100 110 120 130 140 150
170.3 162.9 153.1 151.6 164.5 179.0 184.1 175.6 164.7 157.3 153.6 161.3 164.0 150.9 151.0
Experienced workers (60.1) (58.8) (57.6) (66.6) (63.0) (58.7) (55.2) (65.2) (57.0) (59.4) (59.3) (62.3) (67.0) (56.1) (57.4)
Duncan groups
Novices
241.4 230.1 223.1 217.6 215.2 202.7 197.1 187.5 167.6 153.7 159.7 163.8 155.3 143.7 144.9
(62.0) (61.3) (63.3) (56.1) (55.5) (54.2) (52.0) (47.3) (43.2) (38.3) (37.8) (42.5) (30.1) (31.7) (42.4)
Duncan groups
The ANOVA results revealed that work experience significantly influenced the compression forces (F (1, 630) ¼ 40.5, p < .001) and shoulder moments (F (1, 630) ¼ 34.9, p < .001). Exerting height also revealed similar effects (compression force: F (14, 630) ¼ 6.1, p < .001; shoulder moment: F (14, 630) ¼ 4.7, p < .001). Figs. 5 and 6 present the comparisons of estimated compression forces on the L4/L5 disc and shoulder reactive moments between the two groups, respectively. The mean workers' compression forces on the L4/L5 disc at 15 heights were all less than 1600 N, whereas those of the novices ranged from 460 N to 5114 N. The results of the t-test revealed that, when the exertion height was 70 cm, the compression force of the novices was higher than that of the workers (all p < .01), with differences of 952 Ne3575 N. With the exception of the height range of 60e90 cm, which exhibited identical shoulder moments between the two groups, the experienced workers exhibited lower moment amounts (all p < .05), as displayed in Fig. 6. Fig. 7 further compares the ratios of L4/L5 disc compression forces and shoulder moments normalized by corresponding strengths between the two groups at the 15 exertion heights. At heights 70 cm, the experienced workers maintained compression force per strength value (kN/N) at lower levels than the novices did. A similar trend was observed when handling upper arm loads (N-m/N ¼ m, or moment-arm) at heights >100 cm.
Y.-L. Chen / International Journal of Industrial Ergonomics 44 (2014) 607e614
Elbow 200
200
160
150
120
611
Hip
100
Joint angles (degrees)
80
50
40
0
0 10 20 30 40
50 60 70 80 90 100 110 120 130 140 150
Shoulder 100 80
10 20
30 40 50 60
70 80 90 100 110 120 130 140 150
Knee
200 150
60 100
40
Novices
50
20
Experienced workers
0
0 10
20
30
40
50
60
70
80
90 100 110 120 130 140 150
Exertion heights (cm)
10
20 30
40
50
60 70
80
90 100 110 120 130 140 150
Exertion heights (cm)
Fig. 3. Comparison of the lifting postures of experienced workers and novices during testing at 15 exertion heights.
4. Discussion This study is the first to investigate the lifting strength associated with a full range of exertion heights between female experienced and inexperienced participants. As hypothesized, experienced workers' strengths at 15 lifting heights differed from inexperienced novices' and the results revealed that a discrepancy in strength existed during lifting at lower heights (50 cm)
120 cm
between the two groups. When the heights were >50 cm, no significant differences in strength were observed. This might explain the significant effect of the two-way interaction (i.e., experience height). The novices' strength values reduced with the height increase (ranging from 241.4 N at 10 cm to 143.7 N at 140 cm), whereas the experienced workers' strength values revealed relatively little fluctuation (ranging from 184.1 N at 70 cm to 150.9 N at 140 cm). Certain studies have suggested that maximum acceptable weight of lift (MAWL) patterns are similar between experienced workers and inexperienced novices (Garg and Saxena, 1979; Mital and Manivasagan, 1983). These results might not be completely comparable to our study, particularly at certain lifting heights. The MAWL might be more relevant to workers repeatedly performing submaximal lifts, rather than a one-time maximum effort. It is difficult to fully elucidate why a dramatic decrease in strength was observed among the novices but not among the experienced workers in this study. Our hypothesis of difference
70 cm
50 cm
10 cm
Experienced worker
Novice
Fig. 4. Comparison of the schematic exerting postures of a representative participant from each group at heights of 10, 50, 70 and 120 cm.
Fig. 5. Mean (SD) compression forces on the L4/L5 discs of experienced workers and novices during testing (*p < .05, **p < .01, ***p < .001).
612
Y.-L. Chen / International Journal of Industrial Ergonomics 44 (2014) 607e614
Fig. 6. Mean (SD) estimated shoulder moments of experienced workers and novices during testing (*p < .05, **p < .01, ***p < .001).
existing in lifting strategies between the two participant groups was verified and might explain why differing strength patterns of the two groups were observed. This might be the so-called survivor effect (Marras et al., 2006). A worker with a skilled technique and at least 2 years of experience with handling was considered an experienced worker, and was selected as a participant in our study. By contrast, 23 university students were recruited in this study as novices without lifting experience. Workers with daily work experience might have learned a safety criterion by shifting their postures to avoid overloading when exerting maximum force. As shown in Fig. 3, experienced workers' knee and hip joints were more flexed at heights of 10e50 cm and more extended at heights of 70e100 cm than novices' joints were. When exerting at lower heights, workers chose a nearly knee-flexion strategy, whereas the
Fig. 7. Comparison of the ratios of the L4/L5 disc compression forces and shoulder moments normalized by corresponding strengths at 15 exertion heights for both groups (*p < .05, **p < .01, ***p < .001).
novices tended to choose a more stooped strategy (Fig. 4). Plamondon et al. (2010) compared groups of expert and novice handlers during a horizontal transferring task, and observed that their lifting postures differed; experts bent their lumbar regions less but bent their knees more. However, the external peak L5/S1 moments generally did not differ between the groups (Plamondon et al., 2010). Our result was somewhat similar to that of Plamondon et al.; the spinal loading differed between the groups in this study. A stooped lifting posture, which can result in the generation of greater forces, imposes more compressive force on the lumbarspinal discs (Anderson and Chaffin, 1986; Straker, 2003). Fig. 4 indicates that the moment-arm from the handlebars to the novice participants' lower backs was visibly longer than that of the experienced workers. This might have led the experienced workers to be more safety-conscious than the novices (Fig. 5), resulting in lower strength values and corresponding spinal loads for the workers. Unlike the experienced workers, who adopted considerably different postures between lifts at heights of 50e70 cm, the novice participants adopted postures that varied gradually with increased height (Fig. 4). Once the experienced workers could not adopt a deep-squat strategy, they adopted a nearly erect posture. The semisquat technique that has often been suggested by previous researchers (Garg and Saxena, 1979; Straker, 2003) was not used by any experienced worker in our study. Our results indicated that none of the experienced workers adopted postures associated with the averaged angles shown in Fig. 3. Lortie (2002) indicated that experienced workers adopted moderate-knee flexion to avoid unbalancing the body. This might reflect the actual daily shelf-work exposure of these workers. Most handling tasks in warehouses are typically conducted at submaximal weights to repetitively replenish numerous goods on shelves in a workday. A more stable and less strenuous posture becomes more critical to fit the demands of the task. This might have resulted in experienced workers having exhibited consistent strengths when the heights were increased during the tests. Once the positions were higher than 100 cm, the lower body posture (i.e., the knee and hip joint angles) adopted by the experienced workers and novices was relatively similar (Fig. 3). In other words, with this nearly erect body posture, their strength was determined using their upper extremities (Kumar, 1991). Chen (2000) observed that experienced workers tended to flex and fix their upper arms and move their trunks forward when lifting at higher positions. This stiffened upper-limbs method might reduce the stress on the shoulder and elbow joints by shortening the moment-arm (i.e., the distance from the load to the joints). As presented in Figs. 3 and 4, the experienced workers adopted a stiffer strategy with more flexed elbows and shoulders during lifting. Although no differences in strength at the higher heights were observed between the two groups, the workers inevitably sustained lower loads on the upper limbs than the novices did by shortening their moment-arms. To further elucidate the effects of the posture strategy adopted by the experienced workers for the lifting, the biomechanical loads were normalized using the strength values for each height. When the L4/L5 disc compressions and shoulder moments were individually divided by the corresponding strengths, the ratios reflected the differences in the posture strategies. As shown in Fig. 7, when heights were <90 cm, the experienced workers tended to adopt a safe posture to minimize the load pressure on the lower back; in other words, to effectively maintain the compression force per strength value (KN/N) at lower heights. Marras et al. (2006) suggested that the increased spinal load levels exhibited by novices are due to underdeveloped motor control programming, and that the biomechanical risk is greatly reduced with experience. The
Y.-L. Chen / International Journal of Industrial Ergonomics 44 (2014) 607e614
experienced workers' body mass was heavier than that of the novices (mean 7.0 kg), which increased the spinal load because of the contribution from the upper body mass. When the exertion heights were higher than 90 cm, the workers attempted to minimize the unitary upper arms loads. This result might be due to the experienced workers adopting safer strategies to protect themselves when they were confronted with the tasks. The result might be critical in training novices to use a safe strategy. Although the body masses among the experienced workers were heavier than those among the student participants in this study, the experienced workers exhibited relatively low strengths at low heights. When the strengths were adjusted using the body mass of each participant, the discrepancy in strength profiles between the two groups remained similar. Singh et al. (2009) indicated that obesity does not affect the MAWL and suggested that existing MAWL data can be used to accommodate both general and obese workers. However, further studies based on the biomechanical and physiological approaches are required to provide more complete understanding of whether heavier body masses affected the posture that workers adopted. In this study, we also examined whether results regarding the effects of these task variables on men were also applicable to women. The strength averaged across 15 heights was 162.9 N and 187.0 N for female experienced workers and novices, respectively. These values were approximately 52.5% of those of male experienced workers and 62.7% of those of male novices in the study of Chen et al. (2011) and were consistent with the results presented in other studies (Garg and Ayoub, 1980; Xiao et al., 2005). The testing protocol adopted in this study was identical to the study of Chen et al. (2011), but the results differed. In this study, when lifting at higher positions, no differences in strength were observed between the two groups (Fig. 2). This result did not match that of the Chen et al. study (2011) of male participants. One possible reason was the posture strategy. Regarding men, differences in lifting strengths between the male experienced workers and novices only existed at heights of 100e120 cm, in which the novices were 4.57e7.61 kg (44.8e74.7 N) weaker than the experienced workers. Differences between these studies might be attributable to the fact that women are generally shorter (Chen et al., 2011) and weaker than men, and therefore, exertion heights relative to the participants' heights differed. Sex, lifting height, and work experience might have interactively influenced the participants' strength and the posture strategy. The result also indicated that variations in lifting strengths and postures between male and female novice participants might be attributable to their lack of experience. These results warrant further clarification. This study had limitations. The assumptions used by 3DSSPP were used to determine the biomechanical measures, as well as the fact that this analysis was only performed on a subset of participants (i.e., 12 workers and 12 novices). Twenty-three university students were recruited in this study as novice participants without any lifting experience. Moreover, warehouse tasks involve various degrees of muscular efforts, but this study examined only maximum lifting strengths. Under distinct task conditions, criteria for reducing the physical strain of lifters might differ. Why experienced workers adopted particular strategies based on their posture changes was not entirely clear in this study. They might have a host of expectations and strategies to cope with their particular jobs. This simulated experimental test might approximate the whole work experience in an overall evaluation. Because lifting tasks might vary between repeatedly submaximal and nearmaximal (such as a female employee lifting a box containing a dozen beverages), the experienced workers might have learned to protect themselves from injury or minimize fatigue. Furthermore, the experienced workers adopted a flexed knee position, which
613
might have resulted from a considerable amount of time working with low shelving, leading to the deep squat postures adopted by the workers at heights 50 cm. The experienced workers at heights 50 cm were either unable or unwilling to lift as hard as the novices were, even though they were asked to exert maximal force. This resulted in lower loading on the back of the experienced workers compared with the novices. Although lifting is a dynamic activity, the analysis of static lifting strengths provided by this study offers a valuable reference for lifting task design and employee training, regarding the weighting of various shelves and effective posture training on lifting. 5. Conclusion We demonstrated that a discrepancy exists in lifting strengths between experienced and inexperienced female participant groups at various heights. Lifting strengths of experienced workers were significantly lower than those of novices when heights were 50 cm and also exhibited relatively little fluctuation compared with that of the novices. The workers tended to adopt a consistently deep squat at lower heights and a nearly erect and stiffened-arm posture at higher heights. However, the results of this study were dissimilar to those of Chen et al. (2011), who collected data on Chinese men. These results implied that the experienced workers might have learned to change their postures to effectively avoid overloading. Acknowledgment This study was partially supported by funding from the National Science Council, Taiwan, ROC, grant no. NSC97-2221-E-131-019. This funding is gratefully acknowledged. References Anderson, C.K., Chaffin, D.B., 1986. A biomechanical evaluation of five lifting techniques. Appl. Ergon. 17, 2e8. Agrawal, K.N., Singh, R.K.P., Satapathy, K.K., 2009. Isometric strength of agricultural workers of Meghalaya: a case study of an Indian population. Int. J. Ind. Ergon. 39, 919e923. Authier, M., Lortie, M., Gagnon, M., 1996. Manual handling techniques: comparing novices and experts. Int. J. Ind. Ergon. 17, 419e429. Ayoub, M.M., Bethea, N.J., Deivanayagam, S., Asfour, S.S., Bakken, G.M., Liles, D., Mital, A., Sherif, M., 1978. Determination and Modeling of Lifting Capacity. Final Report. HEW [NIOSH], Cincinnati, OH. Baril-Gingras, G., Lortie, M., 1995. The handling of objects other than boxes: univariate analysis of handling techniques in a large transport company. Ergonomics 38, 905e925. Chaffin, D.B., 1975. Ergonomics guide for the assessment of human static strength. AIHA J. 36, 505e511. Chen, C.J., 2008. Investigation of Musculoskeletal Disorders for the Warehouse Workers in Taiwan. Master Dissertation. Ming Chi University of Technology, Taipei, Taiwan. Chen, Y.L., 2000. Optimal lifting techniques adopted by Chinese men when determining their maximum acceptable weight of lifting. AIHA J. 61, 642e648. Chen, Y.L., Lee, Y.C., Chen, C.J., 2011. Differences in lifting strength profiles between experienced workers and novices at various exertion heights. Int. J. Ind. Ergon. 41, 53e58. Council for Economic Planning and Development (CEPD), 2011. Guidelines and Action Plans for Service Industry Development. Policy Report, Taipei, Taiwan, R.O.C. Gagnon, M., Plamondon, A., Gravel, D., Lortie, M., 1996. Knee movement strategies differentiate expert from novice workers in asymmetrical manual materials handling. J. Biomech. 29, 1445e1453. Garg, A., Ayoub, M.M., 1980. What criteria exist for determining how much load can be lifted safely? Hum. Factors 22, 475e486. Garg, A., Saxena, U., 1979. Effects of lifting frequency and technique on physical fatigue with special reference to psychophysical methodology and metabolic rate. AIHA J. 49, 894e903. Hodder, J.N., MacKinnon, S.N., Ralhan, A., Keir, P.J., 2010. Effects of training and experience on patient transfer biomechanics. Int. J. Ind. Ergon. 40, 282e288. Hooftman, W.E., van der Beek, A.J., van de Wal, B.G., Knol, D.L., Bongers, P.M., Burdorf, A., van Mechelen, W., 2009. Equal task, equal exposure? Are men and
614
Y.-L. Chen / International Journal of Industrial Ergonomics 44 (2014) 607e614
women with the same tasks equally exposed to awkward working postures? Ergonomics 52, 1079e1086. Keir, P.J., MacDonell, C.W., 2004. Muscle activity during patient transfers: a preliminary study on the influence of lift assists and experience. Ergonomics 47, 296e306. Knudson, D., Noffal, G., 2005. Time course of stretch-induced isometric strength deficits. Eur. J. Appl. Physiol. 94, 348e351. Kumar, S., 1991. Arm lift strength in work space. Appl. Ergon. 22, 317e328. Kuorinka, I., Lortie, M., Gautreau, M., 1994. Manual handling in warehouses: the illusion of correct working postures. Ergonomics 37, 655e661. Lee, T.H., 2004. Static lifting strengths at different exertion heights. Int. J. Ind. Ergon. 34, 263e269. cision d'action. Le Travail Lortie, M., 2002. Manutention: prise d'information et de Humain 65, 193e216. Marras, W.S., Parakkat, J., Chany, A.M., Yang, G., Burr, D., Lavender, S.A., 2006. Spine loading as a function of lifting frequency, exposure duration, and work experience. Clin. Biomech. 21, 345e352. Miller, A.E.J., MacDougall, J.D., Tarnopolsky, M.A., Sale, D.G., 1993. Gender differences in strength and muscle fiber characteristics. Eur. J. Appl. Physiol. 66, 254e262. Mital, A., 1987. Patterns of differences between the maximum weights of lift acceptable to experienced and inexperienced materials handlers. Ergonomics 30, 1137e1147. Mital, A., Garg, A., Karwowski, W., Kumar, S., Smith, J.L., Ayoub, M.M., 1993. Status in human strength research and application. IIE Trans. 25, 57e69.
Mital, A., Manivasagan, I., 1983. Maximum acceptable weight of lift as a function of material density, center of gravity location, hand preference, and frequency. Hum. Factors 25, 33e42. Murrel, K.F.H., 1971. Ergonomics: Man and His Working Environment. Chapman and Hall, London. Parakkat, J., Yang, G., Chany, A.M., Burr, D., Marras, W.S., 2007. The influence of lift frequency, lift duration and work experience on discomfort reporting. Ergonomics 50, 396e409. re, C., Salazar, E., 2010. Biomechanical Plamondon, A., Denis, D., Delisle, A., Larivie differences between expert and novice workers in a manual material handling task. Ergonomics 53, 1239e1253. Plamondon, A., Lariviere, C., Delisle, A., Denis, D., Gagnon, D., 2012. Relative importance of expertise, lifting height and weight lifted on posture and lumbar external loading during a transfer task in manual material handling. Ergonomics 55, 87e102. Singh, D., Park, W., Levy, M.S., 2009. Obesity does not reduce maximum acceptable weights of lift. Appl. Ergon. 40, 1e7. Straker, L.M., 2003. A review of research on techniques for lifting low-lying objects: 2. Evidence for a correct technique. Work 20, 83e96. Xiao, G., Lei, L., Dempsey, P.G., Lu, B., Liang, Y., 2005. Isometric muscle strength and anthropometric characteristics of a Chinese sample. Int. J. Ind. Ergon. 35, 674e679. Yates, J.W., Kamon, E., Rodgers, S.H., Champney, P.C., 1980. Static lifting strength and maximal isometric voluntary contractions of back, arm and shoulder muscles. Ergonomics 23, 37e47.