Ultrasonographic Evaluation of the Shoulder in Elite Wheelchair Tennis Players

Transcription

1 Journal of Sport Rehabilitation, 2010, Human Kinetics, Inc. Ultrasonographic Evaluation of the Shoulder in Elite Wheelchair Tennis Players In-Ho Jeon, Hemanshu Kochhar, Jong-Min Lee, Hee-Soo Kyung, Woo-Kie Min, Hwan-Sung Cho, Ho-Wug Wee, Dong-Joo Shin, and Poong-Taek Kim Context: Wheelchair tennis has been identified as a high-risk sport for shoulder injury, so understanding shoulder pathology in these athletes is important. Objective: This study investigated the incidence and pattern of shoulder injuries in wheelchair tennis players using high-resolution ultrasonography. Design: Descriptive study. Setting: International Wheelchair Tennis Open. Participants: 33 elite-level wheelchair tennis players. Outcome Measures: Wheelchair tennis players completed a self-administered questionnaire, and shoulders of each athlete were investigated using high-resolution ultrasonography (linear probe 7.5 MHz). Results: The most common pathology in the dominant shoulder was acromioclavicular pathology, in 21 players (63.6%). Full-thickness rotator-cuff tears involving the supraspinatus were found in 8 dominant shoulders and 6 nondominant shoulders. There were no correlations between identified shoulder pathology and the different variables studied, such as age, training time per day, length of wheelchair use, and length of career as a wheelchair tennis player. Conclusion: High prevalence of rotator-cuff and acromioclavicular pathology was found by ultrasonographic examination in the elite wheelchair tennis players in both dominant and nondominant shoulders. A high index of suspicion of these pathologies in wheelchair athletes is required. Keywords: ultrasonography, rotator cuff, acromioclavicular pathology A high prevalence of upper limb pain has been reported in individuals who rely on manual wheelchairs for mobility. 1 Up to 78% of individuals with spinalcord injury have reported experiencing shoulder pain. 2 Wheelchair use requires continuous use of the upper extremities, not only for mobility but also for transfers, weight-relief lifts, and reaching activities. 3,4 Wheelchair tennis has been identified as a sport associated with a high risk of shoulder injury, but little information is available on the incidence or types of shoulder injury sustained during the sport. 3 Jeon, Kochhar, and Min are with the Dept of Orthopedic Surgery, and Lee, Kyung, and Cho, the Dept of Diagnostic Radiology, Kyungpook National University, Daegu, Korea. Wee is with the Dept of Orthopedic Surgery, Daegu Veterans Hospital, Daegu, Korea. Shin and Kim are with the Dept of Orthopedic Surgery, Daegu Fatima Hospital, Daegu, Korea. 1

2 2 Jeon et al Ultrasonography is one of the most important imaging methods for evaluation of clinical shoulder problems. 5,6 Although its diagnostic accuracy in detecting rotator-cuff tears varies according to the size of the lesion, it has proved to be successful in the diagnosis of complete tears of the supraspinatus, with good sensitivity (93 97%) and specificity (91 100%). 6 In athletes with repeated overhead activities, rotator-cuff abnormalities and degenerative changes in the joint, labrum, or ligament have been reported. 7,8 However, few studies have focused on the shoulder joint of wheelchair athletes, particularly tennis players, who are involved in a substantial amount of overhead activity. The high incidence of rotator-cuff injuries seen in wheelchair tennis players has been explained by the continuous repetition of wheelchair propulsion and repeated overhead activities associated with tennis. Muscle imbalances around the shoulder joint, 9 whether a product of injury or a result of functional movement patterns, 10 have also been identified as contributing to the development of shoulder disorders in wheelchair users. 11 The purpose of this study was to investigate the incidence and pattern of shoulder injuries with ultrasonographic evaluation in wheelchair tennis players playing at a competitive level. Participants Methods After approval from an internal review board, 33 high-level wheelchair tennis players who participated in the International Wheelchair Tennis Open were recruited for this study. All study participants were supplied informational sheets that explained the purpose and method of the study. Written consent was obtained in all cases. Inclusion criteria included use of a manual wheelchair for at least 1 year for at least 50% of home and community mobility. All participants were paraplegic. Design One of 3 orthopedic surgeons specializing in shoulder surgery performed a clinical examination on each study participant per routine protocol of the upper extremity unit of the hospital. The physician administered a questionnaire to collect basic information including age, sex, years involved in wheelchair tennis, training time per day, years spent in a wheelchair, and presence and duration of shoulder pain. Subjective shoulder pain was graded as mild, moderate, or severe. The dominant shoulder was defined as the arm used for both the tennis racket and wheel control; the nondominant arm was used only for wheel control (Figure 1). Sonographic Evaluation Sonographic examinations were performed by a single musculoskeletal radiologist. All sonograms were obtained in real time with the use of an IU 22 (Philips Med, Bothel) with a 7.5-MHz broadband linear-array transducer. All participants underwent sonograms on both shoulders. The long head of the biceps tendon, acromioclavicular joint, and rotator-cuff tendons (supraspinatus, infraspinatus, and subscapularis) were investigated under the modified protocol of Naredo et al. 12

3 Shoulder Ultrasound in Wheelchair Tennis Players 3 Figure 1 Wheelchair tennis players on the court. (A) Dominant arm for racket and nondominant arm for wheelchair control. (B) Wheelchair player while serving impingement position. Long Head of the Biceps. The long head of the biceps tendon was scanned at the level of the bicipital groove. The participants were seated with their arm at their side in a neutral position, the elbow flexed to 90, and the forearm positioned supinely on the thigh. The transducer was placed at the level of the lateral surface of the coracoid process and moved downward to the point where the biceps tendon passes through the bicipital groove. The biceps tendon was first viewed transversely, and then the transducer was rotated 90 to examine it longitudinally. Acromioclavicular Joint. The acromioclavicular joint was examined by longitudinal scanning of the joint with the arm positioned as in the biceps test. Rotator-Cuff Tendons. The supraspinatus, infraspinatus, and subscapularis of the rotator cuff were examined. The supraspinatus tendon was scanned with the study participant seated on a stool with the arm positioned in internal rotation and hyperextension. Longitudinal views of the supraspinatus were taken with the transducer at the level of the anterolateral edge of the acromion, parallel to the long axis of the tendon, positioned approximately 45 between the frontal and sagittal planes. The transverse view was at a right angle to this position. The maximal thickness of the supraspinatus tendon was measured in the transverse view just below the anterolateral edge of the acromion. In addition, a dynamic exam was conducted to evaluate subacromial gliding motion and morphological changes of the anterior edge of the supraspinatus while abducting the shoulder in neutral rotation. Images of the infraspinatus were

4 4 Jeon et al taken with the hand of the participant placed on the contralateral shoulder. The infraspinatus was scanned with the transducer placed parallel to the long axis of the tendon. The scan range was from the level of the posterior surface of the acromion to the inferior edge of infraspinatus tendon. The scan covered the lateral end of the infraspinatus tendon. The subscapularis was first scanned with the arm in neutral rotation and then in external rotation. The rotator-cuff lesions were graded as degeneration, partial tear, or full-thickness tear (the diagnostic criteria described in Table 1). The size of each rotator-cuff tear was defined as small (<1 cm), medium (1 3 cm), large (3 5 cm), or massive (>5 cm) at the onset of the study. Table 1 Ultrasonographic Diagnostic Criteria of Shoulder Abnormalities Abnormality Class Type Definition Rotator-cuff tears I II III Degenerative change Partial-thickness tear Full-thickness tear Tendon hypoechogenicity or thickening with or without internal hypoechoic or hyperechoic foci or greater tuberosity showing an irregularity on the insertion site of the tendon Hypoechoic fiber discontinuity involving the bursal or articular surface or intrasubstance hypoechoic defect or focal tendon thinning (on the articular or bursal side) Nonvisualization of tendon or fiber discontinuity completely from the humeral head to the subacromial subdeltoid bursa or superior convexity instead of concavity Lesions of the long head of biceps I Effusion Thickness of the hypoechoic hole of fluid surrounding the biceps long head tendon greater than 2 mm II Partial Partial interruption of the tendon fibers, rupture separation of the ends, and hypoechoic fluid filling the defect III Complete rupture Lesions of the acromioclavicular joint Effusion Adapted from Naredo et al Osteoarthritis Nonvisualization of tendon in the bicipital groove with complete interruption of the tendon fibers Effusion in the acromioclavicular joint only Cortical irregularities or osteophytes, usually accompanied by intra-articular hypoechoic fluid displacing joint capsule, ie, effusion, or joint bulging by longitudinal scanning

5 Shoulder Ultrasound in Wheelchair Tennis Players 5 Statistical Analysis The statistical software package SPSS for Windows (version 12.0) was used for the analysis of all data. Statistical analysis was performed to compare clinical factors, supraspinatus tear, acromioclavicular-joint pathology, and biceps lesion with gender, age, training time (h), wheelchair use (y), and length of career as a tennis player (y). The comparison of all clinical variables with gender was done with a chi-square test (P =.05). For age, training time (h), wheelchair use (y), and length of career as a tennis player (y), a Kruskal Wallis test was used because a normality test was not valid. Clinical Results Results There were 26 male and 7 female athletes recruited for the study. The average age of the recruits was 36.2 years (range 25 45). The average years involved in wheelchair tennis was 7.5 (range 5 15). The time for tennis practice per day ranged from 4 to 7 hours, with an average of 5.6 hours. The average number of years spent in a wheelchair was 12.5 (range 6 20). Twenty-three athletes (69.6%) reported shoulder pain on the dominant side at the time of the study. Of these, 3 described their pain as severe, 10 as moderate, and another 10 as mild. Twenty-three patients (50%) complained of opposite-side shoulder pain during wheelchair propulsion. The other 10 cases (30.3%) had no pain. Sonographic Results The most common pathologic finding identified in the dominant shoulder of wheelchair tennis players by ultrasonography was acromioclavicular pathology in 21 recruits (63.6%), which consisted of 15 cases with osteoarthritic changes (45.4%) and 6 (18.1%) with joint effusion (Figure 2). Full-thickness rotator-cuff tears were found in 8 dominant shoulders (24.2%) and 6 nondominant shoulders (18.1%). No bilateral full-thickness tear was noted. All full-thickness tears occurred in the supraspinatus tendon (Figure 3). In the dominant shoulders, medium-size full-thickness tears and small partial-thickness tears were more common, whereas small full-thickness tears and medium-size partial-thickness tears were more common in nondominant shoulders (Figure 4). Subscapularis degenerative lesions were found in 4 cases, in both dominant and nondominant shoulders. However, there was no infraspinatus-tendon pathology identified with sonographic examination. Overall, there were no full-thickness tears identified in the infraspinatus and subscapularis tendons (Table 2). Biceps-tendon pathology was found in 7 dominant (21.2%) and 6 nondominant shoulders (18.2%). Subacromial and subdeltoid effusion was found in 11 dominant shoulders (33.3%) and in 6 nondominant shoulders (18.2%). Ultrasonographic findings organized by the subjective pain ratings in the shoulder joint are presented in the Table 3. Among 10 asymptomatic participants, acromioclavicular pathology was the most common finding, followed by supraspinatus pathology, which included 2 degenerative lesions, 2 partial tears, and 1

6 6 Jeon et al Figure 2 Ultrasonographic findings of acromioclavicular joint using 7.5- MHz broadband linear-array transducer. (A) Longitudinal scan of right acromioclavicular joint reveals irregular bony contour (white arrows) and expansion of joint capsule (black arrows) as a result of effusion. These findings suggest degenerative osteoarthritis. (B) Longitudinal scan of right acromioclavicular joint shows normal features of bone and softtissue structures. Figure 3 (A) Longitudinal oblique coronal scan of right supraspinatus tendon discloses remarkable and irregular defect (arrows) at distal end of the tendon suggesting full-thickness tear. (B) Transverse scan of right supraspinatus tendon shows irregular defect (arrows) along the anterior margin of the tendon. Because the defect occupies half of the supraspinatus tendon s width, the full-thickness tear was regarded as medium-sized. Figure 4 (A) Longitudinal scan shows small defect (arrows) at distal end of right supraspinatus tendon suggesting partial-thickness tear. (B) In transverse scan of right supraspinatus tendon, small partial-thickness tear (arrows) is noted

7 Table 2 Degeneration and Tears of Rotator Cuff, Acromioclavicular Joint, Biceps, and Subacromial Bursa at Ultrasonography in Dominant and Nondominant Shoulder of 33 Wheelchair Tennis Players Dominant Nondominant Supraspinatus Small Medium Large Small Medium Large I (n = 3) I (n = 5) II (n = 8) II (n = 5) III (n = 8) III (n = 6) Subscapularis Small Medium Large Small Medium Large I (n = 4) I (n = 4) II (n = 0) II (n = 1) III (n = 0) III (n = 0) Infraspinatus Small Medium Large Small Medium Large I (n = 0) I (n = 0) II (n = 0) II (n = 0) III (n = 0) III (n = 0) Biceps I II III I II III Acromioclavicular joint Effusion Osteoarthritis Effusion Osteoarthritis Subacromial-subdeltoid bursa Effusion Effusion (+) (+)

8 8 Jeon et al full-thickness tear. Three athletes with severe pain presented with 3 osteoarthritic changes in the acromioclavicular joint and 2 full-thickness supraspinatus tears. When the relationship of supraspinatus and acromioclavicular pathology with clinical variables was analyzed, average age, training time per day, length of wheelchair use, and length of career as a wheelchair tennis player were not statistically significantly different (Tables 4 and 5). Table 3 Ultrasonographic Findings in the Symptomatic Shoulders of Wheelchair Tennis Players According to Subjective Pain Pain None n = 10 Mild n = 10 Moderate n = 10 Severe n = 3 Supraspinatus I 2 II III Subscapularis I 1 II III ACJ effusion ACJ arthritis LHB lesion I II III Subacromial bursitis Abbreviations: ACJ, acromioclavicular joint; LHB, long head of biceps. Table 4 Statistical Analysis of Supraspinatus Pathology With Clinical Variables Supraspinatus Variable P Sex.784 a female, n (%) 4 (57.1) 2 (28.6) 1 (14.3) male, n (%) 13 (50.0) 6 (23.1) 7 (26.9) Age (y), mean (SD) (7.49) (7.59) (4.47).403 b Training time per 3.94 (1.95) 4.50 (1.20) 3.38 (1.85).338 b day (h), mean (SD) Wheelchair use (y), 9.24 (3.31) (4.40) 9.50 (3.78).334 b mean (SD) Length of career (y), mean (SD) 7.91 (8.26) 8.44 (3.70) 5.75 (3.54).337 b a Chi-square test. b Kruskal Wallis test.

9 Shoulder Ultrasound in Wheelchair Tennis Players 9 Table 5 Statistical Analysis of Acromioclavicular Pathology With Clinical Variables Acromioclavicular Joint Variable P Sex.976 a female, n (%) 3 (42.9) 1 (14.3) 3 (42.9) male, n (%) 12 (46.2) 3 (11.5) 11 (42.3) Age (y), mean (SD) (8.70) (2.00) (5.72).646 b Training time per 4.07 (1.83) 5.00 (2.45) 3.50 (1.45).473 b day (h), mean (SD) Wheelchair use (y), (3.60) 9.50 (3.51) 9.43 (4.00).814 b mean (SD) Length of career (y), mean (SD) 8.47 (8.67) 5.25 (3.40) 7.14 (3.78).655 b a Chi-square test. b Kruskal Wallis test. Discussion The objective of this study was to characterize shoulder pathology in wheelchair tennis players. The cohort in this study comprised wheelchair tennis players playing at a high or very high level of competition. The prevalence of rotator-cuff and acromioclavicular-joint pathologies found with ultrasonographic examination in this study was relatively high (rotator cuff 24.2%, acromioclavicular joint 63.6%) in wheelchair tennis players. Shoulder problems in the weight-bearing shoulder are not uncommon, 4,13,14 but athletes using wheelchairs are more prone to injuries, particularly those participating in overhead sports. Factors that have been mentioned as contributors to the development of shoulder complaints in these individuals are the relatively high load and high frequency of this load on the shoulder during wheelchair propulsion and sports participation. Fast propulsion has been shown to create greater forces in the shoulder joint than free propulsion in wheelchair athletes. 15 During routine transfer from wheelchair to a bed, the weight of the body is transferred from the trunk through the clavicle and scapula across the subacromial soft tissue to the humerus. During wheelchair propulsion, the shoulder is forced through an arc of motion against resistance and, eventually, the rotator cuff because of its increased transmission of forces, as well as during overhead tennis play, is susceptible to injuries. 2 Full-thickness rotator-cuff tears were found more in the dominant shoulder, which is used for both overhead activity in tennis and forceful propulsion and control of the wheelchair. In dominant shoulders, medium-size full-thickness tears and small partial-thickness tears were the more common type, whereas small full-thickness tears and medium-size partial-thickness tears were more common in nondominant shoulders. This implies that the dominant shoulder is more prone to full-thickness tears of the supraspinatus because of repeated overhead activities in addition to being the weight-bearing shoulder. The nondominant shoulders of these athletes are equally exposed to injury from wheel control and propulsion.

10 10 Jeon et al All full-thickness tears involved the supraspinatus tendon, and none of the athletes in this study presented with a complete rupture of the subscapularis or infraspinatus. Subscapularis degenerative lesions were found in equal (n = 4) numbers in dominant and nondominant shoulders. However, there was no infraspinatus-tendon pathology identified on ultrasonographic evaluation. This finding is similar to the observations reported by Brasseur et al 16 in ultrasonographic rotator-cuff evaluation in 150 veteran tennis players playing at the competitive level. Biomechanical analyses have identified multiple tensile regions in each tendinous unit of the rotator cuff. 17 The principal units identified are the anterior third of the supraspinatus, midsuperior and inferior portions of the infraspinatus, and superior and midsuperior portions of the subscapularis. The supraspinatus middle and posterior thirds failed at significantly lower tensile loads than the infraspinatus and subscapularis. This could be the explanation for the infrequent injury of infraspinatus and subscapularis tendons in our study. Acromioclavicular pathology was found in 63.6% of the wheelchair athletes in this study, the most common pathology identified in the scanned shoulders. Uneven joint surfaces; incomplete, thin, fragile intra-articular discs; and transmission of high compressive forces through the clavicle during weight bearing in these subjects contribute to degenerative disease of the joint. 1,18 Weight lifts, bench press, and abduction exercise can provoke pain at the acromioclavicular joint. Although acromioclavicular-joint arthritis is a very common pathologic ultrasonographic finding in wheelchair athletes, this was not always symptomatic in these athletes. Pathogenesis of shoulder symptoms in these wheelchair athletes is multifactorial. 1,9,11 Scapular kinematics and muscle imbalance may predispose manual wheelchair users to the development of shoulder pathology. 19 Scapular dyskinesis has been identified in wheelchair-bound individuals 20 and also in overhead athletes. The combination of being wheelchair-bound and an overhead athlete can cause serious alterations in the motion and position of the scapula. We speculate that these alterations can then lead to the impingement and rotator-cuff tears that we observed with the ultrasound imaging. We did not examine scapulothoracic kinematics in this study, but studies on scapulothoracic dyskinesia in this group of athletes are required in the future to enhance these clinical findings. With repetitive upward axial loading of the shoulder that occurs with upper limb weight bearing, inadequate depressor strength appears to be the most important predisposing factor. 3,21 Chronic overuse and repetitive impingement positioning 1 seem to be the other contributing factors. Ultrasonographic study of the acute biceps tendon showed that edema was the first sign of overuse injury, and more playing time suggested a larger increase in tendon diameter. In this study, however, wheelchair tennis players did not present a statistically significant difference when the relationship of supraspinatus and acromioclavicular pathology with clinical variables was analyzed, such as average age, training time per day, and length of wheelchair use and wheelchair tennis career. There has not been a diagnostic-imaging study of the shoulder in asymptomatic and symptomatic wheelchair athletes, although there has been a comparative study on the onset and prevalence of shoulder pathology in athletic and nonathletic wheelchair users. 4 Knowledge of this shoulder pathology can help both treating physician and wheelchair athletes modify training schedules and avoid injury during sports. We need to formulate guidelines for early detection, classification, and treatment of shoulder injuries in wheelchair athletes to provide them more pain-free years.

11 Shoulder Ultrasound in Wheelchair Tennis Players 11 A weakness of this study was that there was no well-controlled group that could determine the relationship between age and rotator-cuff pathology. In addition, the number of participants was limited. The diagnostic criteria of the pathological findings on the target structures were adopted from Naredo et al, 12 who considered tendon-sheath effusion a type I lesion of the biceps long head tendon (Table 1). Because the biceps long head tendon sheath is connected to the shoulder s articular space, there could be a debate over whether the effusion in this space is a surrogate marker of early pathological change of the biceps long head tendon. However, in this study, effusion in the long head of biceps-tendon sheath was included in the diagnostic criteria because the ultrasonographic differentiation between effusion from expansion of glenohumeral-joint pathology and effusion from tenosynovitis was not always concrete. Conclusion In conclusion, we performed high-resolution ultrasonography to investigate shoulder pathology in high-level wheelchair tennis athletes. We found a high prevalence of rotator-cuff and acromioclavicular pathology in the athletes in both the dominant and nondominant shoulders. Although acromioclavicular-joint arthritis is a very common pathologic finding in wheelchair athletes, this was not always symptomatic in these athletes. A high index of suspicion for these pathologies in wheelchair athletes is required. Acknowledgment No financial benefits in any form have been received from any commercial party directly or indirectly related to the subject of this article. References 1. Bayley JC, Cochran TP, Sledge CB. The weight-bearing shoulder. the impingement syndrome in paraplegics. J Bone Joint Surg Am. 1987;69: Fullerton HD, Borckardt JJ, Alfano AP. Shoulder pain: a comparison of wheelchair athletes and nonathletic wheelchair users. Med Sci Sports Exerc. 2003;35: Vanlandewijck Y, Theisen D, Daly D. Wheelchair propulsion biomechanics: implications for wheelchair sports. Sports Med. 2001;31: Van Drongelen S, Van der Woude LH, Janssen TW, Angenot EL, Chadwick EK, Veeger DH. Mechanical load on the upper extremity during wheelchair activities. Arch Phys Med Rehabil. 2005;86: Zanetti M, Hodler J. Imaging of degenerative and posttraumatic disease in the shoulder joint with ultrasound. Eur J Radiol. 2000;35: Naredo E, Iagnocco A, Valesini G, Uson J, Beneyto P, Crespo M. Ultrasonographic study of painful shoulder. Ann Rheum Dis. 2003;62: Abrams JS. Special shoulder problems in the throwing athlete: pathology, diagnosis, and nonoperative management. Clin Sports Med. 1991;10: McFarland EG, Wasik M. Epidemiology of collegiate baseball injuries. Clin J Sport Med. 1998;8: Burnham RS, May L, Nelson E, Steadward R, Reid DC. Shoulder pain in wheelchair athletes. The role of muscle imbalance. Am J Sports Med. 1993;21:

12 12 Jeon et al 10. Miyahara M, Sleivert GG, Gerrard DF. The relationship of strength and muscle balance to shoulder pain and impingement syndrome in elite quadriplegic wheelchair rugby players. Int J Sports Med. 1998;19: Sinnott KA, Milburn P, McNaughton H. Factors associated with thoracic spinal cord injury, lesion level and rotator cuff disorders. Spinal Cord. 2000;38: Naredo E, Aguado P, De Miguel E, et al. Painful shoulder: comparison of physical examination and ultrasonographic findings. Ann Rheum Dis. 2002;61: Finley MA, Rodgers MM. Prevalence and identification of shoulder pathology in athletic and nonathletic wheelchair users with shoulder pain: a pilot study. J Rehabil Res Dev. 2004;41: Samuelsson KA, Tropp H, Gerdle B. Shoulder pain and its consequences in paraplegic spinal cord-injured, wheelchair users. Spinal Cord. 2004;42: Taylor D, Williams T. Sports injuries in athletes with disabilities: wheelchair racing. Paraplegia. 1995;33: Brasseur JL, Lucidarme O, Tardieu M, et al. Ultrasonographic rotator-cuff changes in veteran tennis players: the effect of hand dominance and comparison with clinical findings. Eur Radiol. 2004;14: Huang CY, Wang VM, Pawluk RJ, et al. Inhomogeneous mechanical behavior of the human supraspinatus tendon under uniaxial loading. J Orthop Res. 2005;23: Worcester JN, Jr, Green DP. Osteoarthritis of the acromioclavicular joint. Clin Orthop Relat Res. 1968;58: Finley MA, McQuade KJ, Rodgers MM. Scapular kinematics during transfers in manual wheelchair users with and without shoulder impingement. Clin Biomech (Bristol, Avon). 2005;20: Nawoczenski DA, Clobes SM, Gore SL, et al. Three-dimensional shoulder kinematics during a pressure relief technique and wheelchair transfer. Arch Phys Med Rehabil. 2003;84: Ambrosio F, Boninger ML, Souza AL, Fitzgerald SG, Koontz AM, Cooper RA. Biomechanics and strength of manual wheelchair users. J Spinal Cord Med. 2005;28: