- Email: [email protected]
Upper limb joint kinetics during manual wheelchair propulsion in patients with different levels of spinal cord injury
Journal of Biomechanics 43 (2010) 2508–2515
Contents lists available at ScienceDirect
Journal of Biomechanics journal homepage: www.elsevier.com/locate/jbiomech www.JBiomech.com
Upper limb joint kinetics during manual wheelchair propulsion in patients with different levels of spinal cord injury Angel Gil-Agudo a,n, Antonio Del Ama-Espinosa a, Enrique Pe´rez-Rizo a, Soraya Pe´rez-Nombela a, Luis Pablo Rodrı´guez-Rodrı´guez b a b
Biomechanics and Technical Aids Unit, Physical Medicine and Rehabilitation Department, National Hospital for Spinal Cord Injury, SESCAM, Toledo, Spain Physical Medicine and Rehabilitation Department, Complutense University of Madrid, Spain
a r t i c l e in f o
a b s t r a c t
Article history: Received 18 May 2010
The purpose of this study was to compare the forces and moments of the whole upper limb, analyzing forces and moments at the shoulder, elbow and wrist joints simultaneously during manual wheelchair propulsion of persons with different levels of spinal cord injury (SCI) on a treadmill. Fifty-one people participated in this study and were grouped by their level of SCI: C6 tetraplegia (G1), C7 tetraplegia (G2), high paraplegia (G3), and low paraplegia (G4). An inverse dynamic model was defined to compute net joint forces and moments from segment kinematics, the forces acting on the pushrim, and subject anthropometrics. Right side, upper limb kinematic data were collected with four camcorders (Kinescan–IBV). Kinetic data were recorded by replacing the wheels with SmartWheels (Three Rivers Holdings, LLC). All participants propelled the wheelchair at 3 km/h for 1 min. The most noteworthy findings in both our tetraplegic groups in relation to paraplegic groups were increased superior joint forces in the shoulder (G1 and G2 vs G3 p o 0.001; G1 and G2 vs G4 po 0.01), elbow (G1 vs G3 po 0.001; G1 vs G4 po 0.05) and wrist (G1 vs G4 p o 0.001), an increased adduction moment in the shoulder (G1 vs G3 po 0.001; G1 vs G4 po 0.01; G2 vs G3 and G4 p o 0.05) and the constancy of the moments of force of the wrist the fact that they reached their lowest values in the tetraplegic groups. This pattern may increase the risk of developing upper limb overuse injuries in tetraplegic subjects. & 2010 Elsevier Ltd. All rights reserved.
Keywords: Wheelchairs Biomechanics Kinetics Spinal cord injury
1. Introduction Interest in the biomechanical analysis of manual wheelchair propulsion has increased as earlier studies have reported a high incidence of upper limb pathology (Gellman et al., 1988; Sie et al., 1992; Silfverskiold and Waters, 1991; Subbarao et al., 1995). Biomechanical analysis of wheelchair propulsion yields pertinent information for identifying the factors that predispose to upper limb injuries. Only a few investigations focusing on the biomechanical patterns of manual wheelchair propulsion have taken people with tetraplegia into account (Newsam et al., 1996, 1999; Dallmeijer et al., 1994, 1998; Kulig et al., 2001; van Drongelen et al., 2005). One study revealed that a greater proportion of individuals with tetraplegia experience shoulder pain as compared with subjects with paraplegia (Curtis et al., 1999). Specific topics, such as pushrim force application (Dallmeijer et al., 1998)
n Correspondence to: Unidad de Biomeca´nica y Ayudas Te´cnicas, Hospital Nacional de Paraple´jicos, Finca la Peraleda s/n, 45071 Toledo, Spain. Tel.: + 34925247763; fax: + 34925247745. E-mail address: [email protected] (A. Gil-Agudo).
0021-9290/$ - see front matter & 2010 Elsevier Ltd. All rights reserved. doi:10.1016/j.jbiomech.2010.05.021
or shoulder joint kinetics (Kulig et al., 1998, 2001), and general aspects, such as the temporal characteristics of wheelchair propulsion (Newsam et al., 1996) or upper-limb kinematics (Newsam et al., 1999), have been studied in wheelchair users with different levels of SCI, including cervical level injuries. These studies suggest that the level of the subject’s SCI could influence the biomechanics of wheelchair propulsion. However, little information has been reported on the whole upper limb kinetics pattern. Only one study of upper limb kinetics during wheelchair propulsion in a group with upper limb impairment has been published but the subjects analyzed had heterogeneous types of disability (Finley et al., 2004). We hypothesized that upper limb joint kinetics during manual wheelchair propulsion would vary among individuals with different levels of SCI. In a kinematic previous study, main differences between tetraplegic and paraplegic population were found in the distal upper limb joints (Newsam et al., 1999). This could also occur in kinetic evaluation since cervical spine injuries affect hand muscles. The purpose of the present investigation was to compare the forces and moments of whole upper limb, analyzing forces and moments at the shoulder, elbow and wrist joints in unison during manual wheelchair propulsion of persons with four different levels of SCI (two tetraplegic and two
A. Gil-Agudo et al. / Journal of Biomechanics 43 (2010) 2508–2515
paraplegic) on a treadmill. Simultaneous analysis of what occurs in the 3 upper limb joints will allow us to identify the primary alterations and those that derive from compensatory mechanisms.
2. Methods
2509
0,5. Vishay Revere Transducers BV, Breda, The Netherlands). After 2-min adaptation period, participants propelled the wheelchair at 3 km/h during 1 min. We used a digital slope meter (Solatronic EN 17, Fisco Tools Limited, Brook Road, Rayleigh, Essex, UK) to verify that the treadmill surface remained parallel to the floor at all times. A spotter at the front of the treadmill controlled the safety tether.
2.5. Data analysis
2.1. Participants Fifty-one people participated in this study. Subjects under 18 and over 65 years were excluded. The level of SCI should include between C6 and L3. All subjects were classified as ASIA A or B according to the American Spinal Injury Association (Maynard et al., 1997). Participants were daily users of manual wheelchair. None of the subjects participated in competitive sports and reported no upper extremity in the past. Time since onset of injury was at least of 1 year old. Subjects were categorized into four groups according to the neurological level of their lesion: C6 tetraplegia (G1, n¼ 12), C7 tetraplegia (G2, n ¼ 8), paraplegia between T1 and T10, also known as ‘‘high paraplegia’’ (G3, n¼ 17) and paraplegia between T11 and L3, or ‘‘low paraplegia’’ (G4, n¼14). The demographic characteristics of these patients are summarized in Table 1. 2.2. Kinematics Kinematic right upper limb data were collected at 50 Hz with four camcorders (Kinescan–IBV, Instituto de Biomeca´nica de Valencia, Valencia, Spain) (Page et al., 2007). Spatial marker coordinates were smoothed out using a procedure of mobile means. Reflective markers were positioned following ISB recommendations (Wu et al., 2005) to define local reference systems on the hand, forearm, and arm (Fig. 1). The axes of this reference system have been described in detail elsewhere (Gil-Agudo et al., 2010). 2.3. Kinetics The wheels of the chair were replaced by two SmartWheels (Three Rivers Holdings, LLC, Mesa, AZ, USA) (Asato et al., 1993). A synchronization pulse from the Kinescan–IBV was used to trigger the start of kinetic and kinematic collection. Kinetic data were recorded at a frequency of 240 Hz and filtered using a Butterworth, fourth-order, low-pass filter with a cutoff frequency of 20 Hz and a zero phase lag. Spatial marker coordinates were interpolated by cubic spline to synchronize with the kinetic data. Kinematic and kinetic systems were fitted by splines, then downsampled to 50 Hz. All subjects were right-hand dominant. The data recorded with the right wheel were used for the kinetic analysis. The left wheel also was replaced to balance the inertial characteristics of both axes and thus ensure symmetrical propulsion.
2.5.1. Biomechanical model We used an inverse dynamic model described in an earlier publication (GilAgudo et al., 2010). The model was used to compute net joint forces and moments from segment kinematics, the forces acting on the pushrim, and subject anthropometrics (Clauser et al., 1969). Net joint forces and moments were calculated on the global reference system and then expressed on the proximal reference system of the joint (Mercer et al., 2006; Cooper et al., 1999). For the shoulder joint complex, the analysis was focused on the glenohumeral joint; the movements of the scapula, clavicle and thoracic spine were not considered. The forces reported constituted the reaction forces on the joint, expressed on the proximal reference system of the joint. Moments were reported as the action moments and also expressed on the proximal reference system of the joint. For the hand segment, the point of contact on the pushrim was determined using the 3D position of the kinematic markers: in the paraplegic group, this point was assumed to be at the midpoint between metacarpophalangeal joints 3 and 5 (Cooper et al., 1997), and in the tetraplegic group, the point of contact with the pushrim was assumed to be the proximal part of the palmar face of the hand, due to the weaker grip of these subjects.
2.5.2. Data simplification Data were collected in the middle 20-s interval to avoid the effect of acceleration and braking. Five consecutive cycles were selected from the 20-s data recording. The cycles were normalized from 0% to 100% since the time spent in each cycle varied across individuals and across cycles. The push phase started/
Seventh vertebral cervicae Left and Right acromicolavicular Joint
Technical Markers Third metacarpal joint
External Epicondile Ulnar Styloid
2.4. Data compilation A standard adjustable wheelchair, the Action3 Invacare (Invacare Corp, Elyria, OH, USA) was used to obtain a common sitting position for all the subjects (Bregman et al., 2009), and placed on a treadmill (Bonte Zwolle B.V., BO Systems, Netherlands). The seat was 0.42 m wide and deep. The width of the back of the wheelchair was 0.42 m, and the height 0.40 m. Seat angle to the horizontal was 101 and the angle of the back to the vertical was 51. The camber of the wheels was set at 01. The subject position was identical in all subjects so that elbow flexion with the hand on the highest point of the pushrim would be 1001 and the rear axle would be situated 2 cm ahead of the humeral head in each case. Power output was determined by a drag test in which the drag force of the wheelchair-user system was measured (van der Woude et al., 1986) with a force transducer (Revere ALC
Fifth mecatarpal joint
Wheel axle marker Fig. 1. Marker placement: external and internal epicondyles of the elbow joint, radial, and ulnar styloid process, first phalanx of fingers 2, 3, and 5. Six support markers were used, three on the upper part of the arm and three on a support placed on the distal forearm. Two support markers were positioned to identify the head of the humerus.
Table 1 Subject demographics. Group
Number and gender of subjects
Time since injury (months)
Age (years)
Height (m)
Weight (kg)
G1 (C6)
12 9 Men 8 5 Men 17 16 Men 14 10 Men
39.4 (41.7)
26.6 (8.7)
1.8 (0.1)
68.1 (11.5)
85.0 (94.8)
29.0 (6.7)
1.7 (0.1)
65.9 (9.8)
91.4 (97.6)
39.9 (14.0)
1.8 (0.1)
72.7 (14.4)
55.1 (68.8)
41.4 (14.0)
1.7 (0.1)
66.3 (12.1)
G2 (C7) G3 (T1–T10) G4 (T11–L3)
Data are expressed as means (SD if applicable).
2510
A. Gil-Agudo et al. / Journal of Biomechanics 43 (2010) 2508–2515
finished at the instant at which the propulsive moment exerted by the user during hand contact with the pushrim was higher/lower than 1 N m (Fig. 2). The peaks were determined for each stroke individually, and then averaged over 5 cycles. We analyzed temporospatial variables (push time, recovery time, push/recovery ratio, cadence, stroke distance, contact angle, release angle, propulsion angle). In addition, the output variables of the biomechanical model were the time-varying 3D joint net forces and moments, determined over the entire propulsion cycle. The upper limb joint kinetics during the recovery phase was calculated setting handrim forces to zero, so the inverse dynamics model only considered the weight of the arm, as well as upper-extremity motion resulting from the chosen recovery pattern (Gil-Agudo et al., 2010). We used the following sign convention: Forces
Fx: +anterior, posterior. Fy: +superior, inferior. Fz: +lateral, medial.
normal distribution of the variables was needed. Analyses were made with SPSS 12.0 (SPSS Inc., Chicago, IL, USA). This research was approved by the Local Ethics Committee. The ethical principles for experimentation in humans established by the Declaration of Helsinki were adhered to. All participants gave informed consent before the study began.
3. Results The external power output yielded by the treadmill during the drag test was 18.3 (2.7) W. Differences were found in the push/ recovery ratio, which was higher in both tetraplegic groups (p o0.05), but stroke distance was lower in G1 than in G4 (p o0.05) (Table 2).
Moments
Mx: + adduction, abduction (+ cubital deviation, radial deviation for wrist). My: + pronation, supination (+ internal rotation, external rotation for shoulder)
Mz: + flexion, extension. All the necessary equations and calculations were processed with Mathlab (The Mathworks Inc., Natick, MA, USA).
2.5.3. Statistical analysis Descriptive analyses were made of joint variables (median 7 interquartile range). Due to the small size sample, we used the Kruskal–Wallis test to check differences among groups. A p value of less than 0.05 was considered statistically significant. When differences among groups were detected, we performed a Mann–Whitney test with the Bonferroni correction (p value set at 0.008 for pairwise comparisons). As a non-parametric test was made, no hypothesis on the
3.1. Shoulder In the case of G1 and G2, the peak force was superior, whereas in G3 and G4 it was inferior (Fig. 3a). Differences between the overall tetraplegic group and overall paraplegic group were statistically significant (Table 3). The maximum value of the moment acting on the x-axis was in adduction direction in every case (Fig. 4). The adduction moment was higher in G1 than in G3 and G4 (p o0.001, p o0.01) and was also higher in G2 than in G3 and G4 (p o0.05) (Table 3). 3.2. Elbow The magnitude of the vertical forces in superior direction was greater in G1 than in G3 (po0.001) and G4 (p o0.05) (Fig. 3b), as also occurred with the medial forces on the z-axis in G1 compared to G3 and G4. On the x-axis, the anterior forces were lower in G1 than in G3 (po0.05) (Table 4). The peak adductor moment was greater in the tetraplegic than in the paraplegic groups (G1 vs. G3 and G4, p o0.001; G2 vs. G3 and G4, p o0.01) (Table 4). 3.3. Wrist
Fig. 2. Description of the variables contact angle, release angle and propulsion angle.
On the y-axis, the magnitude of the vertical forces in superior direction was greater in G1 than in G4 (p o0.01). The minimum values on the y-axis were inferior and of lower magnitude in G1 than in the two paraplegic groups (po0.001) and in G2 vs. G3 (p o0.05) (Table 5). Significant differences were found in the peak moments of force on all three axes in all of the comparisons between the two tetraplegic and the two paraplegic groups. The peak cubital deviation, pronation, and palmar flexion moments of force were lower in both G1 and G2 compared to G3 and G4 (Table 5). The value of the wrist moments remained almost constant throughout the cycle (Fig. 5).
Table 2 Temporospatial variables.
G1 G2 G3 G4
Push time (s)
Recovery time (s)
Push/recovery
0.6 0.5 0.4 0.4
0.5 0.5 0.5 0.5
1.3 1.1 0.8 0.8
(0.3) (0.1) (0.5) (0.1)
(0.1) (0.1) (0.2) (0.2)
Data are expressed as medians (interquartile ranges). a b
Difference vs. G3, po 0.05 with Bonferroni correction. Difference vs. G4, p o 0.05 with Bonferroni correction.
(0.5)a, (0.4)a, (0.2) (0.2)
b b
Cadence (cycles/s)
Stroke distance (m)
0.9 1.1 1.1 1.2
0.6 0.7 0.8 0.7
(0.3) (0.3) (0.4) (0.3)
(0.3)b (0.5) (0.3) (0.2)
Contact angle (deg.)
Release angle (deg.)
Propulsion angle (deg.)
108.2 111.3 115.7 110.4
51.3 52.2 54.0 48.1
62.5 58.6 64.5 57.5
(18.5) (20.6) (17.5) (17.5)
(14.2) (16.1) (16.4) (13.4)
(16.1) (29.0) (21.1) (13.8)
A. Gil-Agudo et al. / Journal of Biomechanics 43 (2010) 2508–2515
2511
Fig. 3. Illustration of the mean cycle of the vertical forces in the shoulder (3A), elbow (3B), and wrist (3C) throughout the cycle, in both the push phase and the recovery phase for each group of patients. Data are the means obtained from all the subjects. Vertical lines represent the division between push and recovery phases for each group.
4. Discussion In this comprehensive analysis of the upper limb kinetics during manual wheelchair propulsion of persons with levels of SCI
from C6 tetraplegia to low paraplegia, our working hypothesis was confirmed: differences were found between persons with paraplegia and tetraplegia that appear related in part to differences in wrist kinetics. The most noteworthy findings in
2512
A. Gil-Agudo et al. / Journal of Biomechanics 43 (2010) 2508–2515
Table 3 Peak forces and moments acting on shoulder joint over the course of the entire propulsion cycle. Group
Fx (N) Max
G1 G2 G3 G4
Group
31.0 29.2 39.2 34.7
Min (22.2) (9.0)a (14.0) (18.0)
7.1 5.0 0.3 0.7
(5.4)b,c (4.1)a,d (4.0) (2.8)
Fz (N)
Max
34.7 35.1 27.1 30.8
(15.0) (22.1) (23.5) (28.2)
Mx (N m) Max
G1 G2 G3 G4
Fy (N) Min
7.3 10.2 15.0 12.7
(42.7)b,c (17.4)b,c (13.6) (10.7)
Max
35.4 36.5 46.9 42.3
(30.4)a (12.6)a (8.6) (17.8)
My (N m) Min 4.0 4.4 5.8 4.7
(1.7)a (2.6) (2.6) (2.2)
(23.6) (12.1) (6.6) (4.7)
8.5 10.0 8.3 9.5
(10.1) (8.1) (1.9) (7.9)
Mz (N m)
Max
Min
1.5 2.3 3.2 3.5
2.1 1.8 0.3 0.7
(2.7) (2.1) (2.6) (2.4)
10.9 11.8 10.8 9.1
Min
Max (2.3)b,e (2.4)c,f (0.6) (0.9)
10.7 8.0 7.7 6.4
Min (4.6) (3.2) (4.8) (4.6)
3.9 4.8 6.9 4.9
(4.5) (2.3) (2.5) (2.6)
Data are expressed as medians (interquartile ranges). a
Difference vs. G3, po 0.05 with Bonferroni correction. Difference vs. G3, p o 0.001 with Bonferroni correction. c Difference vs. G4, p o0.01 with Bonferroni correction. d Difference vs. G4, p o0.05 with Bonferroni correction. e Difference vs. G4, po 0.001 with Bonferroni correction. f Difference vs. G3, p o 0.01 with Bonferroni correction. b
Fig. 4. The mean cycle of the shoulder joint moments in the frontal plane throughout the cycle in both the push phase and the recovery phase. Data are the means obtained from all the subjects. Vertical lines represent the division between push and recovery phases for each group.
our tetraplegic population were increased upward joint forces in the shoulder, elbow and wrist, an increased adduction moment in the shoulder and an almost constant value of the moments of force on the wrist, which reached their lowest values in the 2 tetraplegic groups. No differences were found between the two tetraplegic groups, probably because, in contrast with earlier studies of manual propulsion over different surfaces and in diverse situations (Newsam et al., 1996), our propulsion conditions were not very demanding so
that all the groups could execute them, i.e., slow velocity (3 km/h) and an unramped treadmill. The presence or absence of abdominal musculature in our two paraplegic groups did not alter any of the kinetics recorded in the upper limb, as previously reported in kinematic upper limb analysis (Newsam et al., 1999). The characteristics of our four SCI groups were the same as in previous reports (Kulig et al., 2001; Newsam et al., 1996, 1999). Comparisons are often difficult because of different testing procedures, units of measurement, equipment employed, and
A. Gil-Agudo et al. / Journal of Biomechanics 43 (2010) 2508–2515
2513
Table 4 Peak forces and moments acting on elbow joint over the course of the entire propulsion cycle. Group
Fx (N)
Fy (N)
Max G1 G2 G3 G4 Group
Min (2.5)a (3.7) (7.7) (6.2)
11.4 13.8 18.5 15.3
G1 G2 G3 G4
Max
49.8 47.3 35.7 36.3
(17.2)b (15.7) (20.6) (14.8)
Mx (N m) Max 1.5 1.3 0.3 0.4
(1.9) (0.8)e,f (0.6) (0.5)
19.7 14.5 4.8 10.4
Min (13.6)c,b (14.6) (7.6) (5.3)
Max
12.8 15.0 21.0 17.2
(10.8)a (8.3) (7.1) (6.7)
My (N m) Min
c,d
Fz (N)
(2.5) (1.8) (1.3) (1.6)
1.8 1.8 0.3 0.7
(9.5)c,d (9.5)d,e (4.0) (4.6)
9.0 6.8 1.1 0.8
10.7 14.7 15.8 16.3
(11.1) (12.6) (11.6) (16.8)
Mz (N m)
Max
1.3 1.6 1.1 1.5
Min
Min c,d
(1.9) (2.4)e,f (0.6) (0.9)
2.1 2.3 3.2 3.5
(3.6) (2.1) (2.6) (2.4)
Max
Min
4.6 4.9 6.0 4.6
1.6 0.9 0.5 0.6
(9.9) (2.4) (1.7) (2.0)
(1.3)e,b (1.0) (0.9) (0.9)
Data are expressed as medians (interquartile ranges). a
Difference vs. G3, po 0.05 with Bonferroni correction. Difference vs. G4, p o 0.05 with Bonferroni correction. Difference vs. G3, p o 0.001 with Bonferroni correction. d Difference vs. G4, p o 0.001 with Bonferroni correction. e Difference vs. G3, po 0.01 with Bonferroni correction. f Difference vs. G4, p o 0.01 with Bonferroni correction. b c
Table 5 Peak forces and moments acting on wrist joint over the course of the entire propulsion cycle. Group
Fx (N)
Fy (N)
Max
Min
G1 G2 G3 G4
6.6 9.8 17.8 17.4
(14.6) (7.5) (13.6) (10.0)
Group
Mx (N m) Max
G1 G2 G3 G4
0.0 0.0 0.2 0.3
Max
12.3 9.7 4.3 4.8
(11.8)a,b (9.7)c,d (4.8) (5.1)
(0.0) (0.1)b,c (0.3) (0.3)
60.9 53.1 39.9 41.6
Min (19.1)d (15.8) (26.7) (15.4)
1.6 2.7 6.8 6.9
(2.6)b,e (4.5)c (3.2) (3.0)
My (N m) Min
e,b
Fz (N)
0.1 0.2 0.9 0.7
Max e,b
(0.1) (0.1)e,b (0.6) (0.6)
0.0 0.0 0.6 0.4
(0.0) (0.0)e,b (0.3) (0.4)
Min
5.6 5.4 2.4 2.7
22.7 19.0 19.5 17.7
(6.7) (4.1) (4.1) (4.2)
(19.3) (5.8) (11.9) (7.1)
Mz (N m) Min
e,b
Max
0.0 0.0 0.1 0.1
Max e,b
(0.0) (0.0)e,b (0.1) (0.1)
0.0 0.0 0.6 0.6
Min e,b
(0.0) (0.0)e,b (0.4) (0.6)
0.1 0.2 0.4 0.4
(0.1)b,c (0.2) (0.6) (0.6)
Data are expressed as medians (interquartile ranges). a
Difference vs. G3, po 0.01 with Bonferroni correction. Difference vs. G4, p o 0.001 with Bonferroni correction. Difference vs. G3, p o 0.05 with Bonferroni correction. d Difference vs. G4, p o 0.01 with Bonferroni correction. e Difference vs. G3, po 0.001 with Bonferroni correction. b c
characteristics of the sample studied (Finley et al., 2004). Most studies report that net forces and moments depend strongly on the propulsion speed (Koontz et al., 2002; Veeger et al., 2002; van Drongelen et al., 2005; Collinger et al., 2008). A uniform velocity for all subjects was chosen in this study. We chose 3 km/h to approximate low-load daily wheelchair propulsion and to optimize test performance in the tetraplegic group (Bregman et al., 2009). This allowed group differences to be determined (Finley et al., 2004) and ensured a submaximal exercise level for all subjects (van Drongelen et al., 2005). The predominant force in people with tetraplegia is applied to the pushrim downward on the vertical axis. This force of action on contact with the pushrim elicits an opposite force of reaction that is transmitted to all the upper limb joints, so that there is a clear predominance of upward vertical forces during the push phase in every joint. This situation predisposes to the compression of
structures like the median nerve in the carpal tunnel or the rotator cuff in the subacromial space. The net joint moments of the glenohumeral joint correlate closely with the glenohumeral joint compression forces (Praagman et al., 2000; Mercer et al., 2006) and pushrim forces have been related to carpal tunnel syndrome (Gellman et al., 1998; Boninger et al., 1999). In an earlier study, no increase in the articular compression forces was found in people with upper limb impairment, probably because the propulsion conditions were not uniform for all the groups (Finley et al., 2004). The predominance of the adductor moments of the shoulder forces during the push phase is due to similar mechanisms as the increased lateromedial forces on the pushrim reported in other studies (Dallmeijer et al., 1998). Both mechanisms allow people incapable of actively extending the elbow and with impaired hand strength to bring the upper limb closer to the pushrim.
2514
A. Gil-Agudo et al. / Journal of Biomechanics 43 (2010) 2508–2515
Fig. 5. The mean cycle of the wrist joint moments in radial-cubital deviation movement throughout the cycle in both the push phase and the recovery phase. Data are the means obtained from all the subjects. Vertical lines represent the division between push and recovery phases for each group.
Most wrist kinetic differences between people with tetraplegia and paraplegia can be attributed to the point of force application of the hand on the pushrim, which influences the calculation of hand torque (Linden et al., 1996). In the case of people with paraplegia, the point of application of force is located at the head of the third metacarpal. However, people with tetraplegia lack full hand muscle function and it is more difficult for them to grasp the pushrim. Consequently, the point of application of the force is shifted to the proximal part of the hand. This involves a change in the model with backward displacement of the point of application of force, which originates relevant differences in the moments of force on the wrist. In the tetraplegic groups, the articular moments remained practically constant throughout the cycle. The value of the articular moments depends on inertia and muscular action. Since the muscular action is practically nonexistent in people with tetraplegia, the final result depends on inertia alone, which in turn depends mainly on weight, making it an almost constant value. It is difficult to compare wheelchair propulsion pattern of our subjects with previous reported (Dallmeijer et al., 1998; Finley et al., 2004) because we did not analyze pushrim kinetics or upper limb kinematics. Subjects with tetraplegia propelled the wheelchair with an increased push/recovery time ratio, but achieved less distance with each stroke. In contrast with other authors (Finley et al., 2004), we found that the push/recovery phase ratio was higher in our tetraplegic groups and the cadence increased as the injury level went from high cervical to low paraplegia. The low resistance of the treadmill compared to other devices (Richter et al., 2007) and the relatively slow propulsion velocity probably allowed the patients with tetraplegia to keep their hands in contact with the pushrim for more time. A long, smooth stroke that maximizes contact with the pushrim and avoids abrupt changes in direction is recommended to reduce cadence and peak force (Boninger et al., 2005). It has been found recently that the tangential force production should be discouraged in the training of wheelchair users because it may not be less efficient, but may also lead to an increase in the prevalence of shoulder and wrist
complaints (Bregman et al., 2009). The propulsion technique found in our tetraplegic population could be a compensatory mechanism to circumvent the predominance of the vertical forces in superior direction that compress the wrist and drive the humerus upward toward the acromion. 5. Conclusion Patients with tetraplegia have upper limb impairment but they can successfully complete the task of manual wheelchair propulsion with relevant adaptations in the kinetic pattern. At a velocity of 3 km/ h, there were no differences in kinetic pattern between the two tetraplegic groups or the two paraplegic groups. The most noteworthy findings in our tetraplegic population were increased superior joint forces in the shoulder, elbow and wrist, an increased adduction moment in the shoulder and the constancy of the moments of force of the wrist, which reached their lowest values in the tetraplegic group. These patterns may increase the risk of developing upper limb overuse injuries in this population. Further studies that include pushrim kinetics, upper limb, and trunk kinematics are needed to pinpoint the kinetic joint data related to pathology. Conflict of interest statement None of the authors of this paper have any conflict of interest in relation to any sources of any kind pertinent to this study.
Acknowledgments We thank Dr. Antonio Sa´nchez-Ramos (Head of Department of Physical Medicine and Rehabilitation) for facilitating our work. We would like to thank Jose´ Luis Rodrı´guez-Martı´n for his critical review of the manuscript and methodology recommendations and to Barbara Thomas for the revision of this manuscript in English.
A. Gil-Agudo et al. / Journal of Biomechanics 43 (2010) 2508–2515
This work was part of a project financed by the Consejerı´a de Sanidad de Castilla-La Mancha, which does not have any commercial interest in the results of this investigation. (Ref: 06006-00).
References Asato, K.T., Cooper, R.A., Robertson, N.R., Ster, D.F., 1993. SMARTWheels: development and testing of a system for measuring manual wheelchair propulsion dynamics. IEEE Trans. Biomed. Eng. 40 (12), 1320–1324. Bregman, D.J.J., van Drongelen, S., Veeger, H.E.J., 2009. Is effective force application in handrim wheelchair propulsion also efficient? Clin. Biomech. 24 13–19. Boninger, M.L., Cooper, R.A., Baldwin, M.A., Shimada, S.D., Koontz, A.L., 1999. Wheelchair pushrim kinetics: body weight and median nerve conduction. Arch. Phys. Med. Rehabil. 80, 910–915. Boninger, M., Koontz, A., Sisto, S., et al., 2005. Pushrim biomechanics and injury prevention in spinal cord injury: recommendations based on CULP-SCI investigations. J. Rehabil. Res. Dev. 42 (Suppl. 3), 9–20. Clauser, C.E., McConville, J.T., Young, J.W., 1969. Weight, volume and center of mass of segments of the human body, AMRL-TR-69-70. Wright-Patterson Air Force Base, Yellow Springs, Ohio. Collinger, J.L., Boninger, M.L., Koontz, A.M., Price, R., Sisto, S.A., Tolerico, M.L., Cooper, R.A., 2008. Shoulder biomechanics during the push phase of wheelchair propulsion: a multisite study of persons with paraplegia. Arch. Phys. Med. Rehabil. 89, 667–676. Cooper, R.A., Robertson, R.N., VanSickle, D.P., Bonninger, M.L., Shimada, S.D., 1997. Methods for determining three-dimensional wheelchair pushrim forces and moments—a technical note. J. Rehabil. Res. Dev. 34, 435–446. Cooper, R., Boninger, M., Shimada, S., Lawrence, B., 1999. Glenohumeral joint kinematics and kinetics for three coordinate system representation during wheelchair propulsion. Am. J. Phys. Med. Rehabil. 78, 435–446. Curtis, K.A., Drysdale, G.A., Lanza, R.D., Kolber, M., Vitolo, R.S., West, R., 1999. Shoulder pain in wheelchair users with tetraplegia and paraplegia. Arch. Phys. Med. Rehabil. 76, 360–367. Dallmeijer, A.J., Kappe, Y.J., Veeger, D.H., Janssen, T.W., van der Woude, L.H., 1994. Anaerobic power output and propulsion technique in spinal cord injures subjects during wheelchair ergometry. J. Rehabil. Res Dev. 31, 120–128. Dallmeijer, A.J., van der Woude, L.H.V., Veeger, H.E.J., Hollander, A.P., 1998. Effectiviness of force application in manual wheelchair propulsion in persons with spinal cord injuries. Am. J. Phys. Med. Rehabil. 77, 213–221. Finley, M., Rasch, E., Keyser, R., Rodgers, M., 2004. The biomechanics of wheelchair propulsion in individuals with and without upper-limb impairment. J. Rehabil. Res. Dev. 41, 395–402. Gellman, H., Sie, I., Waters, R.L., 1988. Late complications of the weight-bearing upper-extremity in the paraplegic patient. Clin. Orthop. Relat. Res. 233, 132–135. Gellman, H., Chandler, D.R., Petrasek, J., Sie, I., Adkins, R., Waters, R.L., 1998. Carpal tunnel syndrome in paraplegic patients. J. Bone Jt. Surg. 70-A, 517–519. Gil-Agudo, A., Del Ama-Espinosa, A., Pe´rez-Rizo, E., Pe´rez-Nombela, S., Crespo-Ruiz, B., 2010. Shoulder joint kinetics during wheelchair propulsion on a treadmill at two different speeds in spinal cord injury patients. Spinal Cord 48, 290–296.
2515
Koontz, A.M., Cooper, R.A., Boninger, M.L., Souza, A.L., Fay, B.T., 2002. Shoulder kinematics and kinetics during two speeds of wheelchair propulsion. J. Rehabil. Res. Dev. 39, 635–650. Kulig, K., Rao, S.S., Mulroy, S.J., Newsam, C.J., Gronley, J.K., Bontrager, E.L., Perry, J., 1998. Shoulder joint kinetics during the push phase of wheelchair propulsion. Clin. Orthop. 354, 132–143. Kulig, K., Newsam, C.J., Mulroy, S.J., Rao, S., Gronley, J.K., Bontrager, E.L., Perry, J., 2001. The effect of level spinal cord injury on shoulder joint kinetics during manual wheelchair propulsion. Clin. Biomech. 16, 744–751. Linden, M.L., van Valent, L., Veeger, H.E., van der Woude, L.H.V., 1996. The effect of wheelchair handrim tube diameter on propulsion efficiency and force application. IEEE Trans. Rehabil. Eng. 4, 123–132. Maynard, F.M., Bracken, M.B., Creasey, G., 1997. International standards for neurological and functional classification of spinal cord injury. American Spinal Injury Association. Spinal Cord 5, 266–274. Mercer, J.L., Boninger, M., Koontz, A., Ren, D., Dyson-Hudson, T., Cooper, R., 2006. Shoulder joint kinetics and patology in manual wheelchair users. Clin. Biomech. 21, 781–789. Newsam, C.J., Mulroy, S.J., Gronley, J.K., Bonstrager, E.L., Perry, J., 1996. Temporal– spatial characteristics of wheelchair propulsion. Effects of level spinal cord injury, terrain and propulsion rate. Am. J. Phys. Med. Rehabil. 75, 292–299. Newsam, C.J., Rao, S.S., Mulroy, S.J., Gronley, J.K., Bontrager, E.L., Perry, J., 1999. Three dimensional upper extremity motion during manual wheelchair propulsion in men with different levels of spinal cord injury. Gait Posture 19, 223–232. Page, A., Mata, V., Hoyos, J.V., Porcar, R., 2007. Experimental determination of instantaneous screw axis in human motions. Error analysis. Mech. Mach. Theory 42 (4), 429–441. Praagman, M., Stokdijk, M., Veeger, H.E.J., Visser, B., 2000. Predicting mechanical load of the glenohumeral joint using net joint moments. Clin. Biomech. 15, 315–321. Richter, W.M., Rodriguez, R., Woods, K.R., Axelson, P.W., 2007. Stroke pattern and handrim biomechanics for level and uphill wheelchair propulsion at selfselected speeds. Arch. Phys. Med. Rehabil. 88, 81–87. Sie, I.H., Waters, R.L., Adkins, R.H., Gellman, H., 1992. Upper extremity pain in the postrehabilitation spinal cord injury patient. Arch. Phys. Med. Rehabil. 73, 44–48. Silfverskold, J., Waters, R.L., 1991. Shoulder pain and functional disability in spinal cord injury patients. Clin. Orthop. 272, 141–145. Subbarao, J., Klopfstein, J., Turpin, R., 1995. Prevalence and impact of wrist and shoulder pain in patients with spinal cord injury. J. Spinal Cord Med. 18, 9–13. van der Woude, L.H.V., De Groot, G., Hollander, A.P., van Ingen Schenau, G.J., Rozendal, R.H., 1986. Wheelchair ergonomics and physiology testing of prototypes. Ergonomics 29, 1561–1573. van Drongelen, S., Van der Woude, L.H.V., Janssen, T.W., Angenot, E., Chadwick, E.K., Veeger, D.H., 2005. Mechanical load on the upper extremity during wheelchair activities. Arch. Phys. Med. Rehabil. 86, 1214–1220. Veeger, H.E.J., Rozendaal, L.A., van der Helm, F.C.T., 2002. Load on the shoulder on low intensity wheelchair propulsion. Clin. Biomech. 17, 211–218. Wu, G., van der Helm, F., Veeger, H.E.J., Makhsous, M., Van Roy, P., Anglin, C., Nagels, J., Karduna, A., McQuade, K., Wang, X., Werner, F., Bucholz, B., 2005. ISB recommendation on definitions of joint coordinate systems of varius joints for the reporting of human joint motion—Part II: shoulder, elbow, wrist and hand. J. Biomech. 38, 981–992.
Contents lists available at ScienceDirect
Journal of Biomechanics journal homepage: www.elsevier.com/locate/jbiomech www.JBiomech.com
Upper limb joint kinetics during manual wheelchair propulsion in patients with different levels of spinal cord injury Angel Gil-Agudo a,n, Antonio Del Ama-Espinosa a, Enrique Pe´rez-Rizo a, Soraya Pe´rez-Nombela a, Luis Pablo Rodrı´guez-Rodrı´guez b a b
Biomechanics and Technical Aids Unit, Physical Medicine and Rehabilitation Department, National Hospital for Spinal Cord Injury, SESCAM, Toledo, Spain Physical Medicine and Rehabilitation Department, Complutense University of Madrid, Spain
a r t i c l e in f o
a b s t r a c t
Article history: Received 18 May 2010
The purpose of this study was to compare the forces and moments of the whole upper limb, analyzing forces and moments at the shoulder, elbow and wrist joints simultaneously during manual wheelchair propulsion of persons with different levels of spinal cord injury (SCI) on a treadmill. Fifty-one people participated in this study and were grouped by their level of SCI: C6 tetraplegia (G1), C7 tetraplegia (G2), high paraplegia (G3), and low paraplegia (G4). An inverse dynamic model was defined to compute net joint forces and moments from segment kinematics, the forces acting on the pushrim, and subject anthropometrics. Right side, upper limb kinematic data were collected with four camcorders (Kinescan–IBV). Kinetic data were recorded by replacing the wheels with SmartWheels (Three Rivers Holdings, LLC). All participants propelled the wheelchair at 3 km/h for 1 min. The most noteworthy findings in both our tetraplegic groups in relation to paraplegic groups were increased superior joint forces in the shoulder (G1 and G2 vs G3 p o 0.001; G1 and G2 vs G4 po 0.01), elbow (G1 vs G3 po 0.001; G1 vs G4 po 0.05) and wrist (G1 vs G4 p o 0.001), an increased adduction moment in the shoulder (G1 vs G3 po 0.001; G1 vs G4 po 0.01; G2 vs G3 and G4 p o 0.05) and the constancy of the moments of force of the wrist the fact that they reached their lowest values in the tetraplegic groups. This pattern may increase the risk of developing upper limb overuse injuries in tetraplegic subjects. & 2010 Elsevier Ltd. All rights reserved.
Keywords: Wheelchairs Biomechanics Kinetics Spinal cord injury
1. Introduction Interest in the biomechanical analysis of manual wheelchair propulsion has increased as earlier studies have reported a high incidence of upper limb pathology (Gellman et al., 1988; Sie et al., 1992; Silfverskiold and Waters, 1991; Subbarao et al., 1995). Biomechanical analysis of wheelchair propulsion yields pertinent information for identifying the factors that predispose to upper limb injuries. Only a few investigations focusing on the biomechanical patterns of manual wheelchair propulsion have taken people with tetraplegia into account (Newsam et al., 1996, 1999; Dallmeijer et al., 1994, 1998; Kulig et al., 2001; van Drongelen et al., 2005). One study revealed that a greater proportion of individuals with tetraplegia experience shoulder pain as compared with subjects with paraplegia (Curtis et al., 1999). Specific topics, such as pushrim force application (Dallmeijer et al., 1998)
n Correspondence to: Unidad de Biomeca´nica y Ayudas Te´cnicas, Hospital Nacional de Paraple´jicos, Finca la Peraleda s/n, 45071 Toledo, Spain. Tel.: + 34925247763; fax: + 34925247745. E-mail address: [email protected] (A. Gil-Agudo).
0021-9290/$ - see front matter & 2010 Elsevier Ltd. All rights reserved. doi:10.1016/j.jbiomech.2010.05.021
or shoulder joint kinetics (Kulig et al., 1998, 2001), and general aspects, such as the temporal characteristics of wheelchair propulsion (Newsam et al., 1996) or upper-limb kinematics (Newsam et al., 1999), have been studied in wheelchair users with different levels of SCI, including cervical level injuries. These studies suggest that the level of the subject’s SCI could influence the biomechanics of wheelchair propulsion. However, little information has been reported on the whole upper limb kinetics pattern. Only one study of upper limb kinetics during wheelchair propulsion in a group with upper limb impairment has been published but the subjects analyzed had heterogeneous types of disability (Finley et al., 2004). We hypothesized that upper limb joint kinetics during manual wheelchair propulsion would vary among individuals with different levels of SCI. In a kinematic previous study, main differences between tetraplegic and paraplegic population were found in the distal upper limb joints (Newsam et al., 1999). This could also occur in kinetic evaluation since cervical spine injuries affect hand muscles. The purpose of the present investigation was to compare the forces and moments of whole upper limb, analyzing forces and moments at the shoulder, elbow and wrist joints in unison during manual wheelchair propulsion of persons with four different levels of SCI (two tetraplegic and two
A. Gil-Agudo et al. / Journal of Biomechanics 43 (2010) 2508–2515
paraplegic) on a treadmill. Simultaneous analysis of what occurs in the 3 upper limb joints will allow us to identify the primary alterations and those that derive from compensatory mechanisms.
2. Methods
2509
0,5. Vishay Revere Transducers BV, Breda, The Netherlands). After 2-min adaptation period, participants propelled the wheelchair at 3 km/h during 1 min. We used a digital slope meter (Solatronic EN 17, Fisco Tools Limited, Brook Road, Rayleigh, Essex, UK) to verify that the treadmill surface remained parallel to the floor at all times. A spotter at the front of the treadmill controlled the safety tether.
2.5. Data analysis
2.1. Participants Fifty-one people participated in this study. Subjects under 18 and over 65 years were excluded. The level of SCI should include between C6 and L3. All subjects were classified as ASIA A or B according to the American Spinal Injury Association (Maynard et al., 1997). Participants were daily users of manual wheelchair. None of the subjects participated in competitive sports and reported no upper extremity in the past. Time since onset of injury was at least of 1 year old. Subjects were categorized into four groups according to the neurological level of their lesion: C6 tetraplegia (G1, n¼ 12), C7 tetraplegia (G2, n ¼ 8), paraplegia between T1 and T10, also known as ‘‘high paraplegia’’ (G3, n¼ 17) and paraplegia between T11 and L3, or ‘‘low paraplegia’’ (G4, n¼14). The demographic characteristics of these patients are summarized in Table 1. 2.2. Kinematics Kinematic right upper limb data were collected at 50 Hz with four camcorders (Kinescan–IBV, Instituto de Biomeca´nica de Valencia, Valencia, Spain) (Page et al., 2007). Spatial marker coordinates were smoothed out using a procedure of mobile means. Reflective markers were positioned following ISB recommendations (Wu et al., 2005) to define local reference systems on the hand, forearm, and arm (Fig. 1). The axes of this reference system have been described in detail elsewhere (Gil-Agudo et al., 2010). 2.3. Kinetics The wheels of the chair were replaced by two SmartWheels (Three Rivers Holdings, LLC, Mesa, AZ, USA) (Asato et al., 1993). A synchronization pulse from the Kinescan–IBV was used to trigger the start of kinetic and kinematic collection. Kinetic data were recorded at a frequency of 240 Hz and filtered using a Butterworth, fourth-order, low-pass filter with a cutoff frequency of 20 Hz and a zero phase lag. Spatial marker coordinates were interpolated by cubic spline to synchronize with the kinetic data. Kinematic and kinetic systems were fitted by splines, then downsampled to 50 Hz. All subjects were right-hand dominant. The data recorded with the right wheel were used for the kinetic analysis. The left wheel also was replaced to balance the inertial characteristics of both axes and thus ensure symmetrical propulsion.
2.5.1. Biomechanical model We used an inverse dynamic model described in an earlier publication (GilAgudo et al., 2010). The model was used to compute net joint forces and moments from segment kinematics, the forces acting on the pushrim, and subject anthropometrics (Clauser et al., 1969). Net joint forces and moments were calculated on the global reference system and then expressed on the proximal reference system of the joint (Mercer et al., 2006; Cooper et al., 1999). For the shoulder joint complex, the analysis was focused on the glenohumeral joint; the movements of the scapula, clavicle and thoracic spine were not considered. The forces reported constituted the reaction forces on the joint, expressed on the proximal reference system of the joint. Moments were reported as the action moments and also expressed on the proximal reference system of the joint. For the hand segment, the point of contact on the pushrim was determined using the 3D position of the kinematic markers: in the paraplegic group, this point was assumed to be at the midpoint between metacarpophalangeal joints 3 and 5 (Cooper et al., 1997), and in the tetraplegic group, the point of contact with the pushrim was assumed to be the proximal part of the palmar face of the hand, due to the weaker grip of these subjects.
2.5.2. Data simplification Data were collected in the middle 20-s interval to avoid the effect of acceleration and braking. Five consecutive cycles were selected from the 20-s data recording. The cycles were normalized from 0% to 100% since the time spent in each cycle varied across individuals and across cycles. The push phase started/
Seventh vertebral cervicae Left and Right acromicolavicular Joint
Technical Markers Third metacarpal joint
External Epicondile Ulnar Styloid
2.4. Data compilation A standard adjustable wheelchair, the Action3 Invacare (Invacare Corp, Elyria, OH, USA) was used to obtain a common sitting position for all the subjects (Bregman et al., 2009), and placed on a treadmill (Bonte Zwolle B.V., BO Systems, Netherlands). The seat was 0.42 m wide and deep. The width of the back of the wheelchair was 0.42 m, and the height 0.40 m. Seat angle to the horizontal was 101 and the angle of the back to the vertical was 51. The camber of the wheels was set at 01. The subject position was identical in all subjects so that elbow flexion with the hand on the highest point of the pushrim would be 1001 and the rear axle would be situated 2 cm ahead of the humeral head in each case. Power output was determined by a drag test in which the drag force of the wheelchair-user system was measured (van der Woude et al., 1986) with a force transducer (Revere ALC
Fifth mecatarpal joint
Wheel axle marker Fig. 1. Marker placement: external and internal epicondyles of the elbow joint, radial, and ulnar styloid process, first phalanx of fingers 2, 3, and 5. Six support markers were used, three on the upper part of the arm and three on a support placed on the distal forearm. Two support markers were positioned to identify the head of the humerus.
Table 1 Subject demographics. Group
Number and gender of subjects
Time since injury (months)
Age (years)
Height (m)
Weight (kg)
G1 (C6)
12 9 Men 8 5 Men 17 16 Men 14 10 Men
39.4 (41.7)
26.6 (8.7)
1.8 (0.1)
68.1 (11.5)
85.0 (94.8)
29.0 (6.7)
1.7 (0.1)
65.9 (9.8)
91.4 (97.6)
39.9 (14.0)
1.8 (0.1)
72.7 (14.4)
55.1 (68.8)
41.4 (14.0)
1.7 (0.1)
66.3 (12.1)
G2 (C7) G3 (T1–T10) G4 (T11–L3)
Data are expressed as means (SD if applicable).
2510
A. Gil-Agudo et al. / Journal of Biomechanics 43 (2010) 2508–2515
finished at the instant at which the propulsive moment exerted by the user during hand contact with the pushrim was higher/lower than 1 N m (Fig. 2). The peaks were determined for each stroke individually, and then averaged over 5 cycles. We analyzed temporospatial variables (push time, recovery time, push/recovery ratio, cadence, stroke distance, contact angle, release angle, propulsion angle). In addition, the output variables of the biomechanical model were the time-varying 3D joint net forces and moments, determined over the entire propulsion cycle. The upper limb joint kinetics during the recovery phase was calculated setting handrim forces to zero, so the inverse dynamics model only considered the weight of the arm, as well as upper-extremity motion resulting from the chosen recovery pattern (Gil-Agudo et al., 2010). We used the following sign convention: Forces
Fx: +anterior, posterior. Fy: +superior, inferior. Fz: +lateral, medial.
normal distribution of the variables was needed. Analyses were made with SPSS 12.0 (SPSS Inc., Chicago, IL, USA). This research was approved by the Local Ethics Committee. The ethical principles for experimentation in humans established by the Declaration of Helsinki were adhered to. All participants gave informed consent before the study began.
3. Results The external power output yielded by the treadmill during the drag test was 18.3 (2.7) W. Differences were found in the push/ recovery ratio, which was higher in both tetraplegic groups (p o0.05), but stroke distance was lower in G1 than in G4 (p o0.05) (Table 2).
Moments
Mx: + adduction, abduction (+ cubital deviation, radial deviation for wrist). My: + pronation, supination (+ internal rotation, external rotation for shoulder)
Mz: + flexion, extension. All the necessary equations and calculations were processed with Mathlab (The Mathworks Inc., Natick, MA, USA).
2.5.3. Statistical analysis Descriptive analyses were made of joint variables (median 7 interquartile range). Due to the small size sample, we used the Kruskal–Wallis test to check differences among groups. A p value of less than 0.05 was considered statistically significant. When differences among groups were detected, we performed a Mann–Whitney test with the Bonferroni correction (p value set at 0.008 for pairwise comparisons). As a non-parametric test was made, no hypothesis on the
3.1. Shoulder In the case of G1 and G2, the peak force was superior, whereas in G3 and G4 it was inferior (Fig. 3a). Differences between the overall tetraplegic group and overall paraplegic group were statistically significant (Table 3). The maximum value of the moment acting on the x-axis was in adduction direction in every case (Fig. 4). The adduction moment was higher in G1 than in G3 and G4 (p o0.001, p o0.01) and was also higher in G2 than in G3 and G4 (p o0.05) (Table 3). 3.2. Elbow The magnitude of the vertical forces in superior direction was greater in G1 than in G3 (po0.001) and G4 (p o0.05) (Fig. 3b), as also occurred with the medial forces on the z-axis in G1 compared to G3 and G4. On the x-axis, the anterior forces were lower in G1 than in G3 (po0.05) (Table 4). The peak adductor moment was greater in the tetraplegic than in the paraplegic groups (G1 vs. G3 and G4, p o0.001; G2 vs. G3 and G4, p o0.01) (Table 4). 3.3. Wrist
Fig. 2. Description of the variables contact angle, release angle and propulsion angle.
On the y-axis, the magnitude of the vertical forces in superior direction was greater in G1 than in G4 (p o0.01). The minimum values on the y-axis were inferior and of lower magnitude in G1 than in the two paraplegic groups (po0.001) and in G2 vs. G3 (p o0.05) (Table 5). Significant differences were found in the peak moments of force on all three axes in all of the comparisons between the two tetraplegic and the two paraplegic groups. The peak cubital deviation, pronation, and palmar flexion moments of force were lower in both G1 and G2 compared to G3 and G4 (Table 5). The value of the wrist moments remained almost constant throughout the cycle (Fig. 5).
Table 2 Temporospatial variables.
G1 G2 G3 G4
Push time (s)
Recovery time (s)
Push/recovery
0.6 0.5 0.4 0.4
0.5 0.5 0.5 0.5
1.3 1.1 0.8 0.8
(0.3) (0.1) (0.5) (0.1)
(0.1) (0.1) (0.2) (0.2)
Data are expressed as medians (interquartile ranges). a b
Difference vs. G3, po 0.05 with Bonferroni correction. Difference vs. G4, p o 0.05 with Bonferroni correction.
(0.5)a, (0.4)a, (0.2) (0.2)
b b
Cadence (cycles/s)
Stroke distance (m)
0.9 1.1 1.1 1.2
0.6 0.7 0.8 0.7
(0.3) (0.3) (0.4) (0.3)
(0.3)b (0.5) (0.3) (0.2)
Contact angle (deg.)
Release angle (deg.)
Propulsion angle (deg.)
108.2 111.3 115.7 110.4
51.3 52.2 54.0 48.1
62.5 58.6 64.5 57.5
(18.5) (20.6) (17.5) (17.5)
(14.2) (16.1) (16.4) (13.4)
(16.1) (29.0) (21.1) (13.8)
A. Gil-Agudo et al. / Journal of Biomechanics 43 (2010) 2508–2515
2511
Fig. 3. Illustration of the mean cycle of the vertical forces in the shoulder (3A), elbow (3B), and wrist (3C) throughout the cycle, in both the push phase and the recovery phase for each group of patients. Data are the means obtained from all the subjects. Vertical lines represent the division between push and recovery phases for each group.
4. Discussion In this comprehensive analysis of the upper limb kinetics during manual wheelchair propulsion of persons with levels of SCI
from C6 tetraplegia to low paraplegia, our working hypothesis was confirmed: differences were found between persons with paraplegia and tetraplegia that appear related in part to differences in wrist kinetics. The most noteworthy findings in
2512
A. Gil-Agudo et al. / Journal of Biomechanics 43 (2010) 2508–2515
Table 3 Peak forces and moments acting on shoulder joint over the course of the entire propulsion cycle. Group
Fx (N) Max
G1 G2 G3 G4
Group
31.0 29.2 39.2 34.7
Min (22.2) (9.0)a (14.0) (18.0)
7.1 5.0 0.3 0.7
(5.4)b,c (4.1)a,d (4.0) (2.8)
Fz (N)
Max
34.7 35.1 27.1 30.8
(15.0) (22.1) (23.5) (28.2)
Mx (N m) Max
G1 G2 G3 G4
Fy (N) Min
7.3 10.2 15.0 12.7
(42.7)b,c (17.4)b,c (13.6) (10.7)
Max
35.4 36.5 46.9 42.3
(30.4)a (12.6)a (8.6) (17.8)
My (N m) Min 4.0 4.4 5.8 4.7
(1.7)a (2.6) (2.6) (2.2)
(23.6) (12.1) (6.6) (4.7)
8.5 10.0 8.3 9.5
(10.1) (8.1) (1.9) (7.9)
Mz (N m)
Max
Min
1.5 2.3 3.2 3.5
2.1 1.8 0.3 0.7
(2.7) (2.1) (2.6) (2.4)
10.9 11.8 10.8 9.1
Min
Max (2.3)b,e (2.4)c,f (0.6) (0.9)
10.7 8.0 7.7 6.4
Min (4.6) (3.2) (4.8) (4.6)
3.9 4.8 6.9 4.9
(4.5) (2.3) (2.5) (2.6)
Data are expressed as medians (interquartile ranges). a
Difference vs. G3, po 0.05 with Bonferroni correction. Difference vs. G3, p o 0.001 with Bonferroni correction. c Difference vs. G4, p o0.01 with Bonferroni correction. d Difference vs. G4, p o0.05 with Bonferroni correction. e Difference vs. G4, po 0.001 with Bonferroni correction. f Difference vs. G3, p o 0.01 with Bonferroni correction. b
Fig. 4. The mean cycle of the shoulder joint moments in the frontal plane throughout the cycle in both the push phase and the recovery phase. Data are the means obtained from all the subjects. Vertical lines represent the division between push and recovery phases for each group.
our tetraplegic population were increased upward joint forces in the shoulder, elbow and wrist, an increased adduction moment in the shoulder and an almost constant value of the moments of force on the wrist, which reached their lowest values in the 2 tetraplegic groups. No differences were found between the two tetraplegic groups, probably because, in contrast with earlier studies of manual propulsion over different surfaces and in diverse situations (Newsam et al., 1996), our propulsion conditions were not very demanding so
that all the groups could execute them, i.e., slow velocity (3 km/h) and an unramped treadmill. The presence or absence of abdominal musculature in our two paraplegic groups did not alter any of the kinetics recorded in the upper limb, as previously reported in kinematic upper limb analysis (Newsam et al., 1999). The characteristics of our four SCI groups were the same as in previous reports (Kulig et al., 2001; Newsam et al., 1996, 1999). Comparisons are often difficult because of different testing procedures, units of measurement, equipment employed, and
A. Gil-Agudo et al. / Journal of Biomechanics 43 (2010) 2508–2515
2513
Table 4 Peak forces and moments acting on elbow joint over the course of the entire propulsion cycle. Group
Fx (N)
Fy (N)
Max G1 G2 G3 G4 Group
Min (2.5)a (3.7) (7.7) (6.2)
11.4 13.8 18.5 15.3
G1 G2 G3 G4
Max
49.8 47.3 35.7 36.3
(17.2)b (15.7) (20.6) (14.8)
Mx (N m) Max 1.5 1.3 0.3 0.4
(1.9) (0.8)e,f (0.6) (0.5)
19.7 14.5 4.8 10.4
Min (13.6)c,b (14.6) (7.6) (5.3)
Max
12.8 15.0 21.0 17.2
(10.8)a (8.3) (7.1) (6.7)
My (N m) Min
c,d
Fz (N)
(2.5) (1.8) (1.3) (1.6)
1.8 1.8 0.3 0.7
(9.5)c,d (9.5)d,e (4.0) (4.6)
9.0 6.8 1.1 0.8
10.7 14.7 15.8 16.3
(11.1) (12.6) (11.6) (16.8)
Mz (N m)
Max
1.3 1.6 1.1 1.5
Min
Min c,d
(1.9) (2.4)e,f (0.6) (0.9)
2.1 2.3 3.2 3.5
(3.6) (2.1) (2.6) (2.4)
Max
Min
4.6 4.9 6.0 4.6
1.6 0.9 0.5 0.6
(9.9) (2.4) (1.7) (2.0)
(1.3)e,b (1.0) (0.9) (0.9)
Data are expressed as medians (interquartile ranges). a
Difference vs. G3, po 0.05 with Bonferroni correction. Difference vs. G4, p o 0.05 with Bonferroni correction. Difference vs. G3, p o 0.001 with Bonferroni correction. d Difference vs. G4, p o 0.001 with Bonferroni correction. e Difference vs. G3, po 0.01 with Bonferroni correction. f Difference vs. G4, p o 0.01 with Bonferroni correction. b c
Table 5 Peak forces and moments acting on wrist joint over the course of the entire propulsion cycle. Group
Fx (N)
Fy (N)
Max
Min
G1 G2 G3 G4
6.6 9.8 17.8 17.4
(14.6) (7.5) (13.6) (10.0)
Group
Mx (N m) Max
G1 G2 G3 G4
0.0 0.0 0.2 0.3
Max
12.3 9.7 4.3 4.8
(11.8)a,b (9.7)c,d (4.8) (5.1)
(0.0) (0.1)b,c (0.3) (0.3)
60.9 53.1 39.9 41.6
Min (19.1)d (15.8) (26.7) (15.4)
1.6 2.7 6.8 6.9
(2.6)b,e (4.5)c (3.2) (3.0)
My (N m) Min
e,b
Fz (N)
0.1 0.2 0.9 0.7
Max e,b
(0.1) (0.1)e,b (0.6) (0.6)
0.0 0.0 0.6 0.4
(0.0) (0.0)e,b (0.3) (0.4)
Min
5.6 5.4 2.4 2.7
22.7 19.0 19.5 17.7
(6.7) (4.1) (4.1) (4.2)
(19.3) (5.8) (11.9) (7.1)
Mz (N m) Min
e,b
Max
0.0 0.0 0.1 0.1
Max e,b
(0.0) (0.0)e,b (0.1) (0.1)
0.0 0.0 0.6 0.6
Min e,b
(0.0) (0.0)e,b (0.4) (0.6)
0.1 0.2 0.4 0.4
(0.1)b,c (0.2) (0.6) (0.6)
Data are expressed as medians (interquartile ranges). a
Difference vs. G3, po 0.01 with Bonferroni correction. Difference vs. G4, p o 0.001 with Bonferroni correction. Difference vs. G3, p o 0.05 with Bonferroni correction. d Difference vs. G4, p o 0.01 with Bonferroni correction. e Difference vs. G3, po 0.001 with Bonferroni correction. b c
characteristics of the sample studied (Finley et al., 2004). Most studies report that net forces and moments depend strongly on the propulsion speed (Koontz et al., 2002; Veeger et al., 2002; van Drongelen et al., 2005; Collinger et al., 2008). A uniform velocity for all subjects was chosen in this study. We chose 3 km/h to approximate low-load daily wheelchair propulsion and to optimize test performance in the tetraplegic group (Bregman et al., 2009). This allowed group differences to be determined (Finley et al., 2004) and ensured a submaximal exercise level for all subjects (van Drongelen et al., 2005). The predominant force in people with tetraplegia is applied to the pushrim downward on the vertical axis. This force of action on contact with the pushrim elicits an opposite force of reaction that is transmitted to all the upper limb joints, so that there is a clear predominance of upward vertical forces during the push phase in every joint. This situation predisposes to the compression of
structures like the median nerve in the carpal tunnel or the rotator cuff in the subacromial space. The net joint moments of the glenohumeral joint correlate closely with the glenohumeral joint compression forces (Praagman et al., 2000; Mercer et al., 2006) and pushrim forces have been related to carpal tunnel syndrome (Gellman et al., 1998; Boninger et al., 1999). In an earlier study, no increase in the articular compression forces was found in people with upper limb impairment, probably because the propulsion conditions were not uniform for all the groups (Finley et al., 2004). The predominance of the adductor moments of the shoulder forces during the push phase is due to similar mechanisms as the increased lateromedial forces on the pushrim reported in other studies (Dallmeijer et al., 1998). Both mechanisms allow people incapable of actively extending the elbow and with impaired hand strength to bring the upper limb closer to the pushrim.
2514
A. Gil-Agudo et al. / Journal of Biomechanics 43 (2010) 2508–2515
Fig. 5. The mean cycle of the wrist joint moments in radial-cubital deviation movement throughout the cycle in both the push phase and the recovery phase. Data are the means obtained from all the subjects. Vertical lines represent the division between push and recovery phases for each group.
Most wrist kinetic differences between people with tetraplegia and paraplegia can be attributed to the point of force application of the hand on the pushrim, which influences the calculation of hand torque (Linden et al., 1996). In the case of people with paraplegia, the point of application of force is located at the head of the third metacarpal. However, people with tetraplegia lack full hand muscle function and it is more difficult for them to grasp the pushrim. Consequently, the point of application of the force is shifted to the proximal part of the hand. This involves a change in the model with backward displacement of the point of application of force, which originates relevant differences in the moments of force on the wrist. In the tetraplegic groups, the articular moments remained practically constant throughout the cycle. The value of the articular moments depends on inertia and muscular action. Since the muscular action is practically nonexistent in people with tetraplegia, the final result depends on inertia alone, which in turn depends mainly on weight, making it an almost constant value. It is difficult to compare wheelchair propulsion pattern of our subjects with previous reported (Dallmeijer et al., 1998; Finley et al., 2004) because we did not analyze pushrim kinetics or upper limb kinematics. Subjects with tetraplegia propelled the wheelchair with an increased push/recovery time ratio, but achieved less distance with each stroke. In contrast with other authors (Finley et al., 2004), we found that the push/recovery phase ratio was higher in our tetraplegic groups and the cadence increased as the injury level went from high cervical to low paraplegia. The low resistance of the treadmill compared to other devices (Richter et al., 2007) and the relatively slow propulsion velocity probably allowed the patients with tetraplegia to keep their hands in contact with the pushrim for more time. A long, smooth stroke that maximizes contact with the pushrim and avoids abrupt changes in direction is recommended to reduce cadence and peak force (Boninger et al., 2005). It has been found recently that the tangential force production should be discouraged in the training of wheelchair users because it may not be less efficient, but may also lead to an increase in the prevalence of shoulder and wrist
complaints (Bregman et al., 2009). The propulsion technique found in our tetraplegic population could be a compensatory mechanism to circumvent the predominance of the vertical forces in superior direction that compress the wrist and drive the humerus upward toward the acromion. 5. Conclusion Patients with tetraplegia have upper limb impairment but they can successfully complete the task of manual wheelchair propulsion with relevant adaptations in the kinetic pattern. At a velocity of 3 km/ h, there were no differences in kinetic pattern between the two tetraplegic groups or the two paraplegic groups. The most noteworthy findings in our tetraplegic population were increased superior joint forces in the shoulder, elbow and wrist, an increased adduction moment in the shoulder and the constancy of the moments of force of the wrist, which reached their lowest values in the tetraplegic group. These patterns may increase the risk of developing upper limb overuse injuries in this population. Further studies that include pushrim kinetics, upper limb, and trunk kinematics are needed to pinpoint the kinetic joint data related to pathology. Conflict of interest statement None of the authors of this paper have any conflict of interest in relation to any sources of any kind pertinent to this study.
Acknowledgments We thank Dr. Antonio Sa´nchez-Ramos (Head of Department of Physical Medicine and Rehabilitation) for facilitating our work. We would like to thank Jose´ Luis Rodrı´guez-Martı´n for his critical review of the manuscript and methodology recommendations and to Barbara Thomas for the revision of this manuscript in English.
A. Gil-Agudo et al. / Journal of Biomechanics 43 (2010) 2508–2515
This work was part of a project financed by the Consejerı´a de Sanidad de Castilla-La Mancha, which does not have any commercial interest in the results of this investigation. (Ref: 06006-00).
References Asato, K.T., Cooper, R.A., Robertson, N.R., Ster, D.F., 1993. SMARTWheels: development and testing of a system for measuring manual wheelchair propulsion dynamics. IEEE Trans. Biomed. Eng. 40 (12), 1320–1324. Bregman, D.J.J., van Drongelen, S., Veeger, H.E.J., 2009. Is effective force application in handrim wheelchair propulsion also efficient? Clin. Biomech. 24 13–19. Boninger, M.L., Cooper, R.A., Baldwin, M.A., Shimada, S.D., Koontz, A.L., 1999. Wheelchair pushrim kinetics: body weight and median nerve conduction. Arch. Phys. Med. Rehabil. 80, 910–915. Boninger, M., Koontz, A., Sisto, S., et al., 2005. Pushrim biomechanics and injury prevention in spinal cord injury: recommendations based on CULP-SCI investigations. J. Rehabil. Res. Dev. 42 (Suppl. 3), 9–20. Clauser, C.E., McConville, J.T., Young, J.W., 1969. Weight, volume and center of mass of segments of the human body, AMRL-TR-69-70. Wright-Patterson Air Force Base, Yellow Springs, Ohio. Collinger, J.L., Boninger, M.L., Koontz, A.M., Price, R., Sisto, S.A., Tolerico, M.L., Cooper, R.A., 2008. Shoulder biomechanics during the push phase of wheelchair propulsion: a multisite study of persons with paraplegia. Arch. Phys. Med. Rehabil. 89, 667–676. Cooper, R.A., Robertson, R.N., VanSickle, D.P., Bonninger, M.L., Shimada, S.D., 1997. Methods for determining three-dimensional wheelchair pushrim forces and moments—a technical note. J. Rehabil. Res. Dev. 34, 435–446. Cooper, R., Boninger, M., Shimada, S., Lawrence, B., 1999. Glenohumeral joint kinematics and kinetics for three coordinate system representation during wheelchair propulsion. Am. J. Phys. Med. Rehabil. 78, 435–446. Curtis, K.A., Drysdale, G.A., Lanza, R.D., Kolber, M., Vitolo, R.S., West, R., 1999. Shoulder pain in wheelchair users with tetraplegia and paraplegia. Arch. Phys. Med. Rehabil. 76, 360–367. Dallmeijer, A.J., Kappe, Y.J., Veeger, D.H., Janssen, T.W., van der Woude, L.H., 1994. Anaerobic power output and propulsion technique in spinal cord injures subjects during wheelchair ergometry. J. Rehabil. Res Dev. 31, 120–128. Dallmeijer, A.J., van der Woude, L.H.V., Veeger, H.E.J., Hollander, A.P., 1998. Effectiviness of force application in manual wheelchair propulsion in persons with spinal cord injuries. Am. J. Phys. Med. Rehabil. 77, 213–221. Finley, M., Rasch, E., Keyser, R., Rodgers, M., 2004. The biomechanics of wheelchair propulsion in individuals with and without upper-limb impairment. J. Rehabil. Res. Dev. 41, 395–402. Gellman, H., Sie, I., Waters, R.L., 1988. Late complications of the weight-bearing upper-extremity in the paraplegic patient. Clin. Orthop. Relat. Res. 233, 132–135. Gellman, H., Chandler, D.R., Petrasek, J., Sie, I., Adkins, R., Waters, R.L., 1998. Carpal tunnel syndrome in paraplegic patients. J. Bone Jt. Surg. 70-A, 517–519. Gil-Agudo, A., Del Ama-Espinosa, A., Pe´rez-Rizo, E., Pe´rez-Nombela, S., Crespo-Ruiz, B., 2010. Shoulder joint kinetics during wheelchair propulsion on a treadmill at two different speeds in spinal cord injury patients. Spinal Cord 48, 290–296.
2515
Koontz, A.M., Cooper, R.A., Boninger, M.L., Souza, A.L., Fay, B.T., 2002. Shoulder kinematics and kinetics during two speeds of wheelchair propulsion. J. Rehabil. Res. Dev. 39, 635–650. Kulig, K., Rao, S.S., Mulroy, S.J., Newsam, C.J., Gronley, J.K., Bontrager, E.L., Perry, J., 1998. Shoulder joint kinetics during the push phase of wheelchair propulsion. Clin. Orthop. 354, 132–143. Kulig, K., Newsam, C.J., Mulroy, S.J., Rao, S., Gronley, J.K., Bontrager, E.L., Perry, J., 2001. The effect of level spinal cord injury on shoulder joint kinetics during manual wheelchair propulsion. Clin. Biomech. 16, 744–751. Linden, M.L., van Valent, L., Veeger, H.E., van der Woude, L.H.V., 1996. The effect of wheelchair handrim tube diameter on propulsion efficiency and force application. IEEE Trans. Rehabil. Eng. 4, 123–132. Maynard, F.M., Bracken, M.B., Creasey, G., 1997. International standards for neurological and functional classification of spinal cord injury. American Spinal Injury Association. Spinal Cord 5, 266–274. Mercer, J.L., Boninger, M., Koontz, A., Ren, D., Dyson-Hudson, T., Cooper, R., 2006. Shoulder joint kinetics and patology in manual wheelchair users. Clin. Biomech. 21, 781–789. Newsam, C.J., Mulroy, S.J., Gronley, J.K., Bonstrager, E.L., Perry, J., 1996. Temporal– spatial characteristics of wheelchair propulsion. Effects of level spinal cord injury, terrain and propulsion rate. Am. J. Phys. Med. Rehabil. 75, 292–299. Newsam, C.J., Rao, S.S., Mulroy, S.J., Gronley, J.K., Bontrager, E.L., Perry, J., 1999. Three dimensional upper extremity motion during manual wheelchair propulsion in men with different levels of spinal cord injury. Gait Posture 19, 223–232. Page, A., Mata, V., Hoyos, J.V., Porcar, R., 2007. Experimental determination of instantaneous screw axis in human motions. Error analysis. Mech. Mach. Theory 42 (4), 429–441. Praagman, M., Stokdijk, M., Veeger, H.E.J., Visser, B., 2000. Predicting mechanical load of the glenohumeral joint using net joint moments. Clin. Biomech. 15, 315–321. Richter, W.M., Rodriguez, R., Woods, K.R., Axelson, P.W., 2007. Stroke pattern and handrim biomechanics for level and uphill wheelchair propulsion at selfselected speeds. Arch. Phys. Med. Rehabil. 88, 81–87. Sie, I.H., Waters, R.L., Adkins, R.H., Gellman, H., 1992. Upper extremity pain in the postrehabilitation spinal cord injury patient. Arch. Phys. Med. Rehabil. 73, 44–48. Silfverskold, J., Waters, R.L., 1991. Shoulder pain and functional disability in spinal cord injury patients. Clin. Orthop. 272, 141–145. Subbarao, J., Klopfstein, J., Turpin, R., 1995. Prevalence and impact of wrist and shoulder pain in patients with spinal cord injury. J. Spinal Cord Med. 18, 9–13. van der Woude, L.H.V., De Groot, G., Hollander, A.P., van Ingen Schenau, G.J., Rozendal, R.H., 1986. Wheelchair ergonomics and physiology testing of prototypes. Ergonomics 29, 1561–1573. van Drongelen, S., Van der Woude, L.H.V., Janssen, T.W., Angenot, E., Chadwick, E.K., Veeger, D.H., 2005. Mechanical load on the upper extremity during wheelchair activities. Arch. Phys. Med. Rehabil. 86, 1214–1220. Veeger, H.E.J., Rozendaal, L.A., van der Helm, F.C.T., 2002. Load on the shoulder on low intensity wheelchair propulsion. Clin. Biomech. 17, 211–218. Wu, G., van der Helm, F., Veeger, H.E.J., Makhsous, M., Van Roy, P., Anglin, C., Nagels, J., Karduna, A., McQuade, K., Wang, X., Werner, F., Bucholz, B., 2005. ISB recommendation on definitions of joint coordinate systems of varius joints for the reporting of human joint motion—Part II: shoulder, elbow, wrist and hand. J. Biomech. 38, 981–992.