Scapular Kinematics and Muscle Activities during Pushing Tasks

Abstract

Objectives: Pushing tasks are functional activities of daily living. However, shoulder complaints exist among workers exposed to regular pushing conditions. It is crucial to investigate the control of shoulder girdles during pushing tasks. The objective of the study was to demonstrate scapular muscle activities and motions on the dominant side during pushing tasks and the relationship between scapular kinematics and muscle activities in different pushing conditions. Methods: Thirty healthy adults were recruited to push a four-wheel cart in six pushing conditions. The electromyographic signals of the upper trapezius (UT) and serratus anterior (SA) muscles were recorded. A video-based system was used for measuring the movement of the shoulder girdle and scapular kinematics. Differences in scapular kinematics and muscle activities due to the effects of handle heights and weights of the cart were analyzed using two-way ANOVA with repeated measures. The relationships between scapular kinematics and muscle activities were examined by Pearson's correlation coefficients. Results: The changes in upper trapezius and serratus anterior muscle activities increased significantly with increased pushing weights in the one-step pushing phase. The UT/SA ratio on the dominant side decreases significantly with increased handle heights in the one-step pushing phase. The changes in upward rotation, lateral slide and elevation of the scapula decreased with increased pushing loads in the trunk-forward pushing phase. Conclusions: This study indicated that increased pushing loads result in decreased motions of upward rotation, lateral slide and elevation of the scapula; decreased handle heights result in relatively increased activities of the serratus anterior muscles during pushing tasks.

(J Occup Health 2013; 55: 259-266)

Introduction

Pushing tasks are functional activities of daily living. According to previous studies, pushing tasks are one of the risk factors that result in discomfort in upper extremities, and 25-41% of workers who exposed to regular pushing conditions have shoulder complaints^1-4). In the upper quarter, the scapula acts as a foundation of the shoulder girdle for multiple directions and movements of the upper extremities^5-9). Alternatively, poor positions and abnormal movements of the scapula may change the length-tension relationship of the muscles attached to on it and affect the normal functions of the shoulder girdle. Ultimately, they may also lead to shoulder-related dysfunctions⁹⁾.

There are some latent risks, such as overexertion of the musculoskeletal system, environmental factors, collisions and unstable loadings, in pushing tasks that may lead to injuries¹⁰⁾. Previous reports pointed out that truck drivers who needed to transport goods by pushing had shoulder complaints twice as often compared with the control group¹¹⁾. Another study also found that workers exposed to pushing conditions for a long time had more chances to experience shoulder-related dysfunctions than general workers (by 2.09-6.37 times)³⁾. The results of a pushing experiment among healthy subjects concluded that the peak torque value of the shoulder joint increased biomechanically with pushing weight. This finding may indicate why people had shoulder complaints while pushing for a long time or pushing with heavy weights.

The shoulder girdle consists of the glenohumeral joint, acromioclavicular joint, sternoclavicular joint and scapulothoracic articulation^12-15). The interactive movements of the glenohumeral and scapulothoracic joints affect the mobility of the shoulder girdle¹⁶⁾. Interactive movement of the glenohumeral joint and scapulothoracic joint occurs when the arm is lifted and is well known as the scapulohumeral rhythm (SHR)¹⁷⁾; Inman et al. determined that a ratio of 2 : 1 existed between abduction of the arm and upward rotation of the scapula when lifting the arm¹⁴⁾. Furthermore, the scapula was rotated upward, tilted posteriorly and rotated externally; thus the the subacromion space of the shoulder joint was increased, and this prevented impingement of the shoulder joint^{7, 17-23)}. Without the proper mechanism of movement of the scapula, an uncoordinated movement would lead to injuries, for example, increased stress in the shoulder joint^{5, 24, 25)}.

Scapular motions consist of gliding and rotation^{7, 17, 23)}. Scapular gliding includes up/down glides and medial/lateral slides; scapular rotation includes upward/downward rotation, anterior/posterior tipping and internal/external rotation^{15, 18)}. The diversity of scapular movements is controlled by muscles attached to the scapula such as the serratus anterior, trapezius, rhomboids, levator scapula and pectoralis minor. Theses scapular muscles play different roles as stabilizers or rotators in the shoulder girdle^{6, 9, 15, 24)}. The serratus anterior (SA) pulls the scapula out laterally, tips it posteriorly and rotates it upward and externally^{6, 8, 20, 24)}. The upper trapezius (UT) elevates the scapula, glides it medially and rotates it upward; the lower trapezius stabilizes the scapula against the rib cage^{20, 26)}. Hence, sufficiently strong scapular muscles can maintain an adequate scapular position in the shoulder girdle¹²⁾. In addition, a force couple made up of the SA and UT rotates the scapula upwardly and affects the function of the upper extremity^{6, 9, 13, 18, 24-28)}, and this has been widely discussed in terms of coordination and the importance of the force couple and scapular muscle strengthening^{9, 13, 26, 27, 29)}.

It has previously been reported that insufficient strength of the scapular muscles led to scapular deviations in the shoulder girdle when performing open kinetic chain exercises. The scapula tended to tip forward and rotate downward and internally after repetitive exercises of shoulder internal and external rotation³⁰⁾. It has been shown that deviation of the scapular upward rotation after a fatigue protocol resulted in an altered SHR³¹⁾. In research on subjects with shoulder-related dysfunctions, it was indicated that disorganization of scapular muscles caused scapular diskynesis¹⁰⁾. Subjects with shoulder impingement were found to have significantly lower protractor muscle strength than healthy subjects. The scapula elevated and rotated upward excessively in subjects with shoulder adhesive capsulitis, while the UT was found to be abnormally activated⁸⁾. An abnormal muscle activation pattern of the upper and lower trapezius was also found in subjects with shoulder-related dysfunctions while abducting their arms, which compensated for the lower activation of the SA muscle^{19, 20)}. The UT/SA ratio (the ratio of the UT and SA muscle activations) increased in patients with shoulder-related dysfunction, who had increased UT activations and decreased SA activation, which caused excessive scapular elevation^{12, 20)}. Occurrence of scapular substitution patterns, for example, excessive scapular tipping, elevation or upward rotation, may narrow the subacromion space or may induce tendonitis of the rotator cuff^{6-8, 12, 16, 17, 20, 21, 26, 31)}.

Previous reports in closed kinetic chain exercises revealed that the SA muscles were significantly activated in healthy subjects performing the push-up exercise^{13, 15)}; on the contrary, the UT/SA ratio rose as a result of abnormal higher activities in the upper trapezius among subjects with shoulder disabilities⁴⁾.

The pushing task belongs to a type of closed kinetic chain movements, which correspond to the following description: “the distal part of the extremity was fixed while the proximal part of the extremity moved¹⁵⁾.” Some related factors during pushing tasks have previously been discussed. The exerted hand force increases with the pushing weight¹⁾, the proper handle heights for maximum exerted pushing force is from one meter to shoulder height, and people can easily perform pushing tasks with their hands at a height of 50-60% of their shoulder height⁵⁾. Other factors that affected the hand forces, such as the direction and coefficient of the floor during pushing, were also mentioned previously^{1, 10)}. However, there is still a lacks of discussions concerning how theses related factors, for instance, the weight and height of pushing, affect scapular motion and scapular muscular control in the shoulder girdle.

The purpose of this study was to investigate scapular control as well as the relationships between scapular kinematics and muscle activities among different pushing tasks.

Subjects and Methods

The subjects were thirty healthy young adults (15 females; weight, 61.2 ± 10.48 kg; height, 1.67 ± 0.08 m; shoulder height, 90.27 ± 5.98 cm; pelvic height, 63.58 ± 4.49 cm) who were informed about the process of the experiments and signed a consent form approved by the Institutional Review Board of Chang Gung University. Subjects did not report any history of shoulder-related dysfunctions during the past of 12 months. The exclusion criteria were histories of shoulder fractures, dislocations, subluxations, impingements, and adhesive capsulitis or surgeries in the shoulder girdles.

Instrumentation

A 26-kg-cart (63 cm × 72 cm × 144 cm) with four wheels was loaded with different weights (0, 50 and 100 kg) on a fixed track. A video-based system was set up with three digital cameras (GR-DVL9800, JVC Corporation, Japan), and the sampling rate was set at 60 Hz. The Ariel Performance Analysis System (APAS) software (Ariel Dynamics Inc., Trabuco Canyon, CA, USA) was used for calculating the kinematic data. The electromyographic (EMG) signals were collected using an MP150 System (BIOPAC Systems, Inc., Goleta, CA, USA) at 1,000 Hz with and TSD 150B bipolar surface electrodes (BIOPAC Systems, Inc., Goleta, CA, USA) with a 20-mm inter electrode distance and 11.4 mm diameter.

Experimental procedures

Subjects were required to fill out an informed consent form that had been approved by Chang Gung University and then understand well the context of the experiments. The subjects' skin was prepared by shaving the hair off the electrode placement area if necessary, cleansing the skin with 70% alcohol and application of conductive gel to the skin surface. Two EMG surface electrodes were applied to both UTs parallel to the muscle fibers midway between the seventh spinous process of the cervical vertebra and acromion process. The other two electrodes were applied to both SAs parallel to the muscle fibers midway between the pectoralis major and latissimus dorsi below the axilla area. The reference electrode was applied to the left ankle on the medial malleolus^{6, 13, 15, 20, 26-28, 32)}.

To perform maximal voluntary contractions (MVC) of a subject's UT (MVC_UT), the subject sat without a back support, and the neck side-bent to the same side, rotated to the opposite side and extended as resistance is applied at the occipital and shoulder. The MVC of the SA (MVC_SA) was performed with the subject in the same seating position and with the shoulder flexed to 125 degrees as resistance was applied above the elbow while the scapula was fixed²⁸⁾.

Seven reflective markers (4-cm in diameter) were placed on bony landmarks in accordance with the International Society of Biomechanics recommendation: the seventh spinous process of the cervical vertebra (C7), right acromion (AC), trigonum spinae (TS, root of scapular spine) and angulus inferior (AI, inferior angle of scapula), eighth spinous process of the thoracic vertebra (T8), right greater trochanter (GT) and lateral epicondyle (LE)^{11, 29)}. The kinematic data of the shoulder girdle from these markers included the angles of shoulder flexion and scapular upward rotation and the distances of scapular elevation and lateral slide of the scapular inferior angle (protraction).

There were three sub-phases in the pushing task in this study. In the initial phase, subjects stood with their hands placed on the pushing bar and waited for the “Go” verbal signal (Fig. 1-a). When the subjects heard the first “Go” signal, they pushed the cart immediately one step forward and with their elbows extended (one-step pushing phase, Fig. 1-b). After the second “Go” signal, the subjects pushed with their trunk leaning forward and kept their feet fixed (trunk-forward pushing phase, Fig. 1-c). Subjects performed six pushing tasks with three different weights (0, 50 and 100 kg) and two different heights (shoulder and pelvic) randomly. The shoulder and pelvic heights were defined as 80% and 60% of the acromion height respectively¹⁾.

Data processing

In the glenohumeral joint, the angle of shoulder flexion was defined as the angle composed of the AC, LE and GT. In the scapulothoracic articulation, scapula motion was assumed to rotate and glide purely in the coronal plane. Scapular upward rotation was defined as the angle composed of the spinal column (C7-T8) and medial side of the scapula (TS-AI), which can be calculated by subtracting angle C7-AI-TS from angle T8-C7-AI (Fig. 2-a). In addition, Fig. 2-b shows that using a trigonometric method and the angle consisting of T8, C7 and AI (θ), scapular horizontal and vertical slides can be defined (the sine and cosine values of the distance between C7 and AI under the angle θ).

EMG data were processed using the AcqKnowledge software (BIOPAC Systems, Inc., Goleta, CA, USA), a band-pass filter with a frequency of 20-500 Hz and a full-wave rectifier followed by processing with a root-mean-square (RMS) algorithm with a 50-ms moving window. To normalize the RMS-EMG data, the differences between pushing phases and the initial phase were divided by the MVC_UT or MVC_SA and multiplied by 100 (presented in percentage). Both scapular kinematics and muscle activations were synchronized post-experimentally.

Statistical analysis

Statistical analysis was completed using SPSS for Windows 10.0 (SPSS Inc, Chicago, IL, USA). Descriptive statistics were used to analyze the subjects' demographic data. Differences in scapular kinematics and scapular muscle activations were calculated between the initial phase and one-step pushing phase as well as between the one-step pushing phase to trunk-forward pushing phase. These differences were examined by two within-factors (height × weight) ANOVA with repeated measures (α=0.05). If the interaction between height and weight was not significant, two-way ANOVA was used for testing the main effect; if the interaction was significant, the multiple comparisons were examined by Bonferroni adjustment. The relationships between scapular kinematics and scapular muscle activations were examined by Pearson's correlation coefficients (α=0.05).

Fig. 1-a

Initial phase: subjects placed two hands on the handles with the elbows extended.

Fig. 1-b

One-step pushing phase: subject stepped forward with the elbows extended.

Fig. 1-c

Trunk-forward pushing phase: subject pushed the cart as far as they could with their elbows extended and without raising their heels.

(The subject consented to use of these images to demonstrate the pushing task.)

Fig. 2-a

Estimation of the angle of scapular upward rotation.

Fig. 2-b

Estimation of scapular horizontal and vertical slides.

Results

From the initial phase to the one-step pushing phase

Regarding the scapular muscle activation, although no main effect of pushing weight or handle height on SA activation was found, the significant interaction between pushing weight and handle height showed that SA activation at the pelvic handle height was significantly higher than that at the shoulder handle height while pushing a 100-kg cart (F_(2,58)=3.42, p<0.05; Table 1). In addition, handle height revealed a significant main effect on UT/SA ratio under different pushing conditions, and a higher UT/SA ratio at the shoulder height was found when compared with that at the pelvic height (p<0.01; Table 2).

Regarding the scapular kinematics, Table 3 demonstrates a significant main effect of handle height on changes in scapular horizontal displacement in which the scapula slid more when pushed at the pelvic height when compared with that at the shoulder height (p<0.01). In addition, there was a significant interaction between pushing weight and handle height on scapular vertical displacement, and the displacement was increased significantly more at the pelvic handle height (F_(2,58)=16.73, p<0.01) than that at the shoulder handle height while pushing an empty cart.

Table 1 Changes in scapular muscles activations among different pushing conditions (N=30)

Pushing weight	Handle height	Changes in the UT (%MVC)		Changes in the SA (%MVC)
		One-step pushing	Trunk-forward pushing†	One-step pushing†	Trunk-forward pushing*‡
0 kg	Shoulder	2.62 ± 2.72	2.54 ± 3.48	4.16 ± 4.69	8.97 ± 17.51
	Pelvic	2.54 ± 3.34	5.10 ± 7.67	6.51 ± 8.27	8.38 ± 9.31
50 kg	Shoulder	10.93 ± 11.60	8.95 ± 10.39	11.16 ± 13.43	12.80 ± 11.06
	Pelvic	8.11 ± 8.91	7.96 ± 7.79	12.32 ± 10.37	20.17 ± 16.39
100 kg	Shoulder	19.25 ± 15.23	10.99 ± 9.77	13.80 ± 10.26	19.72 ± 19.31
	Pelvic	16.93 ± 11.35	15.58 ± 11.56#	20.04 ± 13.61#	31.39 ± 23.64

^*Pushing weight shows the main effect (p<0.05). ^‡Handle height shows the main effect (p<0.01). ^†There was significant interaction between weight and height in this phase using ANOVA with repeated measures. ^#The muscle activations were significantly higher at the pelvic handle height than at the shoulder level while pushing a 100-kg cart (p<0.05). UT: upper trapeztius. SA: serratus anterior. MVC: maximal voluntary contractions.

Table 2 UT/SA ratios among different pushing conditions (N=30)

Pushing weight	Handle height	One-step pushing*	Trunk-forward pushing
0 kg	Shoulder	1.21 ± 1.63	1.00 ± 1.68
	Pelvic	0.95 ± 1.52	0.87 ± 0.90
50 kg	Shoulder	1.57 ± 1.37	1.11 ± 1.40
	Pelvic	0.94 ± 0.88	0.67 ± 0.66
100 kg	Shoulder	2.06 ± 2.36	1.10 ± 1.19
	Pelvic	1.22 ± 1.00	0.77 ± 0.83

^*Handle height shows the main effect on one-step pushing (p<0.05). UT: upper trapezius. SA: serratus anterior.

Table 3 Changes in shoulder girdle movements among different pushing conditions (N=30)

Pushing weight	Handle height	Shoulder flexion(degrees)		Scapular upward rotation(degrees)		Scapular horizontal displacement(cm)		Scapular vertical displacement(cm)
		One-step pushing	Trunk-forward pushing	One-step pushing	Trunk-forward pushing*	One-step pushing‡	Trunk-forward pushing*	One-step pushing†	Trunk-forward pushing*
		0 kg	Shoulder	0.93 ± 5.12	32.74 ± 26.05	0.42 ± 3.10	5.93 ± 6.31	−0.05 ± 0.89	0.90 ± 1.42	−0.08 ± 0.89	−3.36 ± 3.59
Pelvic	3.28 ± 5.83		31.62 ± 23.09	0.90 ± 2.13	4.29 ± 4.77	0.32 ± 0.44	0.72 ± 1.05	−2.24 ± 2.57#	−2.44 ± 2.57
50 kg	Shoulder	−0.48 ± 7.75	15.49 ± 14.21	0.19 ± 2.33	2.89 ± 3.59	0.05 ± 0.56	0.36 ± 0.82	0.04 ± 1.24	−0.93 ± 1.81
	Pelvic	−0.22 ± 7.97	26.56 ± 22.16	0.55 ± 2.15	3.18 ± 5.01	0.23 ± 0.50	0.58 ± 1.12	0.10 ± 0.76	−1.70 ± 2.40
100 kg	Shoulder	−5.28 ± 9.18	15.85 ± 14.68	0.45 ± 3.43	2.63 ± 4.19	−0.20 ± 0.76	0.33 ± 1.02	0.45 ± 1.17	−0.50 ± 1.56
	Pelvic	−1.07 ± 7.79	18.15 ± 15.43	0.51 ± 2.98	1.07 ± 2.90	0.08 ± 0.68	−0.03 ± 0.82	0.40 ± 1.24	−0.47 ± 1.69

^*Pushing weight shows the main effect (p<0.01). ^‡Handle height shows the main effect (p<0.01). ^†There was significant interaction between weight and height in this phase using ANOVA with repeated measures. ^#The scapular vertical displacement were significantly higher at the pelvic handle height than at the shoulder level while pushing an empty (0-kg pushing weight) cart (p<0.05). The effects of handle heights were significant at the same pushing weight by multiple comparisons using Bonferroni adjustment (p<0.05).

A significant moderate correlation between SA activation and shoulder flexion was revealed while pushing a 100-kg cart at the shoulder handle height (r=0.42, p<0.05).

From the one-step pushing phase to the trunk-forward pushing phase

Regarding the scapular muscle activation, there was no main effect of pushing weight or handle height on UT activation, but the significant interaction between the two factors showed a significant increase in UT activation at the pelvic handle height compared with that at the shoulder handle height while pushing a 100-kg cart (F_(2,58)=3.86, p<0.05; Table 1). In addition, handle height and pushing weight revealed main effects on changes in SA activation (p<0.01). This indicated that SA activation increased more at the pelvic height than at the shoulder height regardless of the pushing weight. It also revealed significant differences in SA activation among the three pushing weight conditions regardless of the handle height (Table 1).

Regarding the scapular kinematics, pushing weight showed a main effect on the changes in scapular upward rotation, scapular horizontal displacement, and vertical displacement (p<0.01; Table 3). In multiple comparisons, pushing a 100-kg cart resulted in more scapular horizontal slide than pushing an empty cart (p<0.01); it also showed significant differences in scapular elevation among the three different pushing weight conditions (p<0.01).

A significant moderate correlation was revealed between SA activation and scapular horizontal displacement while pushing an empty cart at the shoulder height (r=0.37, p<0.05). While pushing a 100-kg cart at the pelvic height, there were moderate correlations of SA activation and scapular horizontal displacement (r=0.56, p<0.05) as well as SA activation and scapular vertical displacement (r=0.43, p<0.05).

Discussion

The purpose of this study was to investigate the changes in scapular kinematics and muscle activities under different pushing conditions. All the data analyzed and presented were from the subjects' dominant side. From the initial phase to the one-step pushing phase, handle height was an important factor that affected the UT/SA ratios in the pushing task. The significantly lower UT/SA ratio at the pelvic height than that at the shoulder height was due to the higher activations of the SA while pushing a 100-kg cart at the pelvic height compared with at the shoulder height (MVC 6.24% higher) (Table 1). Handle height also affected the scapular horizontal displacements significantly: the higher the handle position was, the lower the deviations of scapular horizontal displacement revealed. This was because the starting position of the scapular was set at a higher angle of upward rotation, which meant a larger horizontal displacement at shoulder handle height than at pelvic handle height. Therefore, the changes in scapular horizontal displacement would not be greater at the shoulder handle height than at the pelvic handle height. The lower deviations in scapular horizontal displacement explained the more stable condition of the scapula while pushing at the shoulder handle height. The changes in scapular vertical displacement were greater at the pelvic handle height than at the shoulder handle height, which was supported not only by the lower starting position of the scapular setting but also by the significant activations of the SA while pushing at the pelvic handle height.

From the one-step pushing phase to the trunk-forward pushing phase, handle height affected significantly the SA activation, with SA decreasing as the handle height increased. This contrasts with the proposed proportioned relation between SA activation and shoulder flexion in research performed on subjects doing wall-slide exercises³²⁾. The increased SA activations with pelvic handle height (lesser angle of shoulder flexion) may imply the cause of injuries as a result of overexertion. Pushing weight was the other factor affecting the activations of the SA: the pushing weight increased, the activations of the SA increased. This corresponded to the results of a previous study of push-up exercises in which the more difficult the push-up exercises were, the more scapular muscle activations appeared^{9, 15, 26)}. If workers are exposed to heavy pushing weights conditions for a long time, the scapular muscles are also exposed to the risk of injuries.

The angles of shoulder flexion and scapular upward rotation increased proportionally in this phase, and this corresponded to the notion of interactive scapulohumeral rhythm¹⁴⁾. Pushing weight significantly affected the changes in angles of shoulder flexion and scapular upward rotation and horizontal and vertical displacement, with increased pushing weights resulting in decreased changes in angle and displacement (Table 3). This indicated that the SA which attached scapula against the chest wall worked as a scapular stabilizer rather than a scapular rotator in closed kinetic chained tasks with difficulty¹⁵⁾.

The UT and SA are the prime movers for scapular upward rotation^{6, 9, 13, 18, 26-28)}. However, we did not find any significant correlations between the activations of the UT, SA, and changes in angles of shoulder flexion and scapular upward rotation. According to the relationship between shoulder flexion and SA activities, the activations of the SA revealed apparently 120 to 150 degrees of shoulder flexion, but the mean value of shoulder flexion during pushing tasks was merely 85.49 degrees. The reason for the insufficient SA activation may be due to the insufficient angle of shoulder flexion, which may cause a poor length-tension relationship.

In the interaction between handle height and pushing weight, the changes in SA activations while pushing a 100-kg cart at pelvic handle height were significantly greater than those at shoulder handle height. This finding may be the reason for related injuries in this pushing condition. The changes in UT activations were lesser than those in the SA across the pushing phases because the SA served as a scapular stabilizer during upper limb activities.

In summary, we demonstrated that pushing weight was the first factor that affects the scapular kinematics and muscle activities during pushing tasks. Increased pushing weight not only increased significantly the changes in scapular muscle activations, scapular upward rotation and lateral slide, but also decreased the changes in scapular upward rotation on the dominant side of the subject. Handle height was the second factor, with increased handle height resulting in decreased changes in horizontal displacement and decreased activations of the UT and SA in pushing tasks.

Limitations

The pushing task in this study was limited to represent a real situation when pushing a cart weighting less than 100 kg. However, injuries or discomfort in the shoulder occur in situations such as when pushing a cart weighing more than this. Therefore, in order to elicit the maximal exertion force of pushing as well as to standardize the method adopted in this study, we instructed subjects to perform trunk-forward pushing with their feet fixed. The pushing task was also limited due to the space constraints of our facility, so the subjects could not perform the task continuously which is similar to actual conditions. Future study focusing on sequential and dynamic pushing rather than phase-by-phase pushing will benefit the understanding of scapular control during a pushing task. In addition, this study only gathered data from the subjects' dominant side. Therefore, we were not able to examine the changes on the non-dominant side and compare the scapular kinematics between the two sides of the body. Last but not the least, there are many other scapular muscles that control the scapular kinematics and that may need to be investigated in the future.

Occupational health relevance

Even if the unusual pushing task that subjects adopted may cause changes in their shoulder region, we suggest that workers should use a proper shoulder height for pushing, which could decrease the activations of the SA under prolonged pushing or sudden exertion conditions. In addition, further investigation of the impact of scapular muscle strengthening on pushing workers is expected to alleviate shoulder complaints and prevent repeatedly-caused injuries in the shoulder girdle.

Acknowledgments: We acknowledge Dr. Yan-Ying Ju, PhD, PT, for his generous research consultation and insightful comments regarding this study. We also deeply appreciate Ms. Szu-Chieh Lee, PT, for her great assistance in subject recruitment and her permission regarding publishing of the pushing task demonstration.

References

• 1)   de Looze  MP,  van Greuningen  K,  Rebel  J,  Kingma  I,  Kuijer  PPFM. Force direction and physical load in dynamic pushing and pulling. Ergonomics 2000; 43: 377-90.
• 2)   Hoozemans  MJM,  Kuijer  PPFM,  Kingma  I, . Mechanical loading of the low back and shoulders during pushing and pulling activities. Ergonomics 2004; 47: 1-18.
• 3)   Hoozemans  MJM,  van der Beek  AJ,  Frings-Dresen  MHW,  van der Woude  LHV,  van Dijk  FJH. Pushing and pulling in association with low back and shoulder complaints. Occup Environ Med 2002; 59: 696-702.
• 4)   Schibye  B,  Sogaard  K,  Martinsen  D,  Klausen  K. Mechanical load on the low back and shoulders during pushing and pulling of two-wheeled waste containers compared with lifting and carrying of bags and bins. Clin Biomech 2001; 16: 549-59.
• 5)   Al-Eisawi  KW,  Kerk  CJ,  Congleton  JJ,  Amendola  AA,  Jendins  OC,  Gaines  WG. The effect of handle height and cart load on the initial hand forces in cart pushing and pulling. Ergonomics 1999; 42: 1099-113.
• 6)   Cools  AM,  Witvrouw  EE,  Declercq  GA,  Danneels  LA,  Cambier  DC. Scapular muscle recruitment patterns: trapezius muscle latency with and without impingement symptoms. Am J Sports Med 2003; 31: 542-9.
• 7)   Lin  J,  Hanten  WP,  Olson  SL, . Shoulder dysfunction assessment: self-report and impaired scapular movements. Phys Ther 2006; 86: 1065-74.
• 8)   Lin  JJ,  Wu  YT,  Wang  SF,  Chen  SY. Trapezius muscle imbalance in individuals suffering from frozen shoulder syndrome. Clin Rheumatol 2005; 24: 569-75.
• 9)   Moseley  JB,  Jobe  FW,  Pink  M,  Perry  J,  Tibone  J. EMG analysis of the scapular muscles during a shoulder rehabilitation program. Am J Sports Med 1992; 20: 128-34.
• 10)   Ferreira  JJ,  Boocock  MG,  Gray  MI. Review of the Risks Associated with Pushing and Pulling Heavy Loads 1st ed. Sheffield (UK): HSE Publisher; 2004.
• 11)   van der Helm  FC. A Standardized Protocol for Motion Recordings of the Shoulder. In:  Veeger  HE,  van der Helm  FC,  Rozing  PM, editors. Proceedings of the First Conference of the ISG. Netherlands: Delft University of Technology; 1997. p.7-12.
• 12)   Babyar  SR. Excessive scapular motion in individuals recovering from painful and stiff shoulders: causes and treatment strategies. Phys Ther 1996; 76: 226-38.
• 13)   Decker  MJ,  Hintermeister  RA,  Faber  KJ,  Hawkins  RJ. Serratus anterior muscle activity during selected rehabilitation exercises. Am J Sports Med 1999; 27: 784-91.
• 14)   Inman  VT,  Saunders  JB,  Abbott  LC. Observations on the function of the shoulder joint. J Bone Joint Surg 1944; 26: 1-30.
• 15)   Lear  LJ,  Gross  MT. An electromyographical analysis of the scapular stabilizing synergists during a push-up progression. J Orthop Sports Phys Ther 1998; 28: 146-57.
• 16)   McMullen  J,  Uhl  TL. A kinetic chain approach for shoulder rehabilitation. J Athl Train 2000; 35: 329-37.
• 17)   Lukasiewicz  AC,  McClure  P,  Michener L,  Pratt  N,  Snnett  B. Comparison of 3-dimensional scapular position and orientation between subject with and without shoulder impingement. J Orthop Sports Phys Ther 1999; 29: 574-86.
• 18)   Ebaugh  DD,  McClure  PW,  Karduna  AR. Three-dimensional scapulothoracic motion during active and passive arm elevation. Clin Biomech 2005; 20: 70-709.
• 19)   Lin  J,  Hanten  WP,  Olson  SL, . Functional activity characteristics of individuals with shoulder dysfunctions. J Electromyogr Kinesiol 2005; 15: 576-86.
• 20)   Ludewig  PM,  Cook  TM. Alterations in shoulder kinematics and associated muscle activity in people with symptoms of shoulder impingement. Phys Ther 2000; 80: 276-91.
• 21)   Ludewig  PM,  Cook  TM,  Nawoczenski  DA. Three-dimensional scapular orientation and muscle activity at selected positions of humeral elevation. J Orthop Sports Phys Ther 1996; 24: 57-65.
• 22)   McClure  PW,  Michener  LA,  Sennett  BJ,  Karduna  AR. Direct 3-dimensional measurement of scapular kinematics during dynamic movements in vivo. J Shoulder Elbow Surg 2001; 10: 269-77.
• 23)   Uhl  TL,  Carver  TJ,  Mattacola  CG,  Mair  SD,  Nitz  AJ. Shoulder musculature activation during upper extremity weight-bearing exercises. J Orthop Sports Phys Ther 2003; 33: 109-17.
• 24)   Cools  AM,  Witvrouw  EE,  Mahieu  NN,  Danneels  LA. Isokinetic scapular muscle performance in overhead athletes with and without impingement symptoms. J Athl Train 2005; 40: 104-10.
• 25)   Smith  J,  Kotajarvi  BR,  Padgett  DJ,  Eischen  JJ. Effect of scapular protraction and retraction on isometric shoulder elevation strength. Arch Phys Med Rehabil 2002; 83: 367-70.
• 26)   Ludewig  PM,  Hoff  MS,  Osowdki  EE,  Meschke  SA,  Rundquist  PJ. Relative balance of serratus anterior and upper trapezius muscle activity during push-up exercises. Am J Sports Med 2004; 32: 484-93.
• 27)   Ekstrom  RA,  Bifulco  KM,  Lopau  CJ,  Andersen  CF,  Gough  JR. Comparing the function of the upper and lower parts of the serratus anterior muscle using surface electromyography. J Orthop Sports Phys Ther 2004; 34: 235-43.
• 28)   Ekstrom  RA,  Soderberg  GL,  Donatelli  RA. Normalization procedures using maximum voluntary isometric contractions for the serratus anterior and trapezius muscles during surface EMG analysis. J Electromyogr Kinesiol 2005; 15: 418-28.
• 29)   Wu  G,  van der Helm  FCT,  Veeger  HEJ, . ISB recommendation on definitions of joint coordinate systems of various joints for the reporting of human joint motion—part II: shoulder, elbow, wrist and hand. J Biomech 2005; 38: 981-92.
• 30)   Tsai  NT,  McClure  PW,  Karduna  AR. Effects of muscle fatigue on 3-dimensional scapular kinematics. Arch Phys Med Rehabil 2003; 84: 1000-5.
• 31)   McQuade  KJ,  Wei  SH,  Smidt  GL. Effects of local muscle fatigue on three-dimensional scapulohumeral rhythm. Clin Biomech 1995; 10: 144-8.
• 32)   Hardwick  DH,  Beebe  JA,  McDonnell  MK,  Lang  CE. A comparison of serratus anterior muscle activation during a wall slide exercise and other traditional exercises. J Orthop Sports Phys Ther 2006; 36: 903-10.