MANUAL WHEELCHAIR USERS are commonly diagnosed

Transcription

1 1141 Relation Between Median and Ulnar Nerve Function and Wrist Kinematics During Wheelchair Propulsion Michael L. Boninger, MD, Bradley G. Impink, BSE, Rory A. Cooper, PhD, Alicia M. Koontz, PhD ABSTRACT. Boninger ML, Impink BG, Cooper RA, Koontz AM. Relation between median and ulnar nerve function and wrist kinematics during wheelchair propulsion. Arch Phys Med Rehabil 2004;85: Objective: To investigate the relation between median and ulnar nerve health and wrist kinematics in wheelchair users. Design: Case series. Setting: Biomechanics laboratory and electrodiagnostic laboratory at a Veterans Health Administration medical center and a university hospital, respectively. Participants: Thirty-five people with spinal cord injury who use manual wheelchairs. Intervention: Subjects propelled their own wheelchair on a dynamometer at 0.9 and 1.8m/s. Bilateral biomechanic data were obtained by using force and moment sensing pushrims and a kinematic system. Bilateral median and ulnar nerve conduction studies were also completed. Main Outcome Measures: Wrist flexion, extension, radial and ulnar deviation peaks, and ranges of motion (ROMs) as related to median and ulnar motor and sensory amplitudes. A secondary analysis included peak pushrim forces and moments and stroke frequency. Results: There was a significant, positive correlation between flexion and extension ROM and both ulnar motor amplitude (r.383, P.05) and median motor amplitude (r.361, P.05). Conclusions: Contrary to our hypothesis, subjects using a greater ROM showed better nerve function than subjects propelling with a smaller ROM. Subjects using a larger ROM used less force and fewer strokes to propel their wheelchairs at a given speed. It is possible that long, smooth strokes may benefit nerve health in manual wheelchair users. Key Words: Kinematics; Median nerve; Rehabilitation; Ulnar nerve; Wheelchairs; Wrist by the American Congress of Rehabilitation Medicine and the American Academy of Physical Medicine and Rehabilitation From the Human Engineering Research Laboratories, Department of Physical Medicine and Rehabilitation, School of Medicine (Boninger, Impink, Cooper, Koontz), Department of Rehabilitation Science and Technology, School of Health and Rehabilitation Sciences (Boninger, Cooper, Koontz), and Department of Bioengineering, School of Engineering (Boninger, Impink, Cooper, Koontz), University of Pittsburgh; and VA Pittsburgh Health Care System, Center of Excellence in Wheelchairs and Related Technology (Boninger, Impink, Cooper, Koontz), Pittsburgh, PA. Supported by the US Department of Veterans Affairs Rehabilitation Research & Development Services (project no. B2290T), National Institute for Disability and Rehabilitation Research (grant nos. H133A011107, H133N000019), National Center for Medical Rehabilitation Research of the National Institutes of Health (grant no. 1 P01 HD33989), and the Paralyzed Veterans of America and Eastern Paralyzed Veterans of America. No commercial party having a direct financial interest in the results of the research supporting this article has or will confer a benefit upon the authors(s) or upon any organization with which the author(s) is/are associated. Reprint requests to Michael L. Boninger, MD, Human Engineering Research Laboratories, VA Pittsburgh Health Care System, 7180 Highland Dr, 151R-1, Pittsburgh, PA 15206, Boninger@pitt.edu /04/ $30.00/0 doi: /j.apmr MANUAL WHEELCHAIR USERS are commonly diagnosed with carpal tunnel syndrome (CTS) as well as ulnar nerve injury. The incidence of CTS in this population ranges from 49% to 63%. 1-6 In addition to CTS, ulnar nerve damage has been reported by several investigators. 3,5,7 These injuries can have significant consequences because wheelchair users rely on their arms for mobility, transfers, and most activities of daily living. The consequences of upper-limb injury are so significant that Sie et al 6 stated that having upper-limb pain is equivalent to a spinal cord injury (SCI) of a higher level. Upper-limb pain has been associated with lower quality of life and increased dependence on helpers. 8,9 Researchers have long hypothesized that there is a link between wheelchair propulsion and CTS. This thought was partially derived from the ergonomics literature, which established that CTS is more prevalent in workers who perform high-force, high-repetition tasks Our group 15 confirmed this hypothesis when we reported a direct link between median nerve injury and wheelchair propulsion biomechanics. That study found a correlation between median nerve function and cadence of propulsion, as well as magnitude and rate of rise of forces exerted. Higher cadence and forces correlated with worse median nerve function. Research has also found a link between CTS and the awkward or extreme wrist postures required in some work. 10,16 Several investigators have studied the association between CTS and wrist posture by measuring the pressure in the carpal canal in various positions. 2,17-19 Tanzer 19 apparently was the first to investigate carpal tunnel pressure changes with respect to wrist position. While performing CTS surgery, he found that pressure increased in the proximal portion of the tunnel during flexion and extension, whereas only extension increased pressure in the distal portion. Since this initial study, other investigators 17,18,20 have confirmed that extremes of wrist flexion and extension can greatly increase the pressure within the carpal tunnel, more so in patients with CTS. In 1 study 2 specific to wheelchair users, 18 subjects with paraplegia had manometric studies performed with their wrist in various positions, they had higher pressures in wrist extension than did control subjects with CTS but no paralysis. Other investigators have examined wrist posture in a work setting as a risk factor for CTS. Armstrong and Chaffin 16 examined posture and force in 36 women, half of whom had CTS. Using cinematography to measure position, they found that the women with CTS were in a position of wrist extension more frequently than those without CTS. Werner et al 10 used videotapes to rate posture and force exerted during repetitive tasks. They found that abnormal posture was associated with symptoms and nerve conduction evidence of CTS. The purpose of this study was to investigate whether wrist motion during wheelchair propulsion is related to both median and ulnar nerve function. We hypothesized that greater peak flexion, extension, and radial and ulnar deviation would correlate with decreased median and ulnar nerve function. We further hypothesized that increased absolute range of motion (ROM) would correlate with worse median and ulnar nerve function.

2 1142 WHEELCHAIR KINEMATICS AND NERVE FUNCTION, Boninger METHODS Participants Subjects were recruited from 2 primary sources: wheelchair vendors and discharge records from a large inpatient SCI unit. Letters were sent to the wheelchair users inviting them to participate in the study. The letter did not mention that pain or CTS was the focus of the study. Subjects were recruited in this manner to identify all persons with SCI, not just those currently being followed through regular clinic visits. To be included in the study, subjects had to have a traumatic SCI at the second thoracic level or below that occurred more than 1 year before the study. Participants needed to use a manual wheelchair for mobility and to have no history of trauma to the shoulder or wrist. Before the experiment, each subject provided written informed consent as approved by our institutional review board. Each subject was weighed in his/her wheelchair, with a load cell under each wheel. Their chair was weighed separately and the difference between the 2 measures was used as the subject s weight. Height was self-reported. Nerve Conduction Studies Each subject had bilateral upper-extremity nerve conduction studies (NCSs), as described previously. 15 The following is an abbreviated review of those studies. Skin temperature was maintained at more than 34 C throughout the study. All NCSs were done with a Nicolet Viking II a by 1 of 3 technicians certified by the American Board of Electrodiagnostic Medicine who were blinded to other subject information. Hand-held bipolar stimulation, using individually required voltage and pulse duration, was delivered to provide a supramaximal stimulus. One-centimeter disk-recording electrodes were used for motor studies, and loop ring recording electrodes were used for sensory studies. Antidromic sensory latencies were measured as time from stimulation to the negative peak, and orthodromic motor latencies were measured as time from stimulation to the initiation of the negative potential. Sensory amplitudes were measured from negative peak to positive trough. Motor amplitudes were measured from baseline to negative peak. The median and ulnar sensory responses were recorded from the second and fifth digits, respectively. The distance from the stimulator to the active electrode was 14cm. The median and ulnar motor responses were recorded from the abductor pollicis brevis muscle and the abductor digiti minimi muscle, respectively. The distance from the stimulator to the active electrode was 8cm. All stimulation was performed at the wrist. Subjects with peripheral polyneuropathy, as determined by abnormal median and ulnar motor and sensory latencies or amplitudes, were dropped from the analysis. Kinematic and Kinetic Data The experimental setup and kinetic data collection methods were presented previously. 21 A brief description is presented here. An Optotrak b camera was used to collect kinematic data. A 2-camera system was used to visualize both arms. Data were collected at a rate of 60Hz. Markers were placed bilaterally on specific bony landmarks of each subject. These landmarks included the lateral epicondyle, ulnar styloid, radial styloid, and the third metacarpophalangeal joint (MCJ). Once all markers were placed, subjects were asked to sit in a set position while data were collected for 20 seconds. In this position, the subject sat upright with upper arms straight down at the side anda90 bend at the elbow. Subjects were also asked to keep their wrists straight and hands open with palms facing each other. This position was used to offset any nonneutral initial angles from marker positioning. The SMART Wheel device, c used for kinetic data collection, was placed on both sides of every subject s wheelchair. The SMART Wheel has a precision of 2N and a resolution of 0.2N at a collection rate of 240Hz. 22 The SMART Wheel did not alter the camber, diameter, or tire size of the subject s normal wheels. Once the devices were attached, each subject s wheelchair was secured on a dynamometer and subjects were instructed to propel their wheelchair to become acclimated to the dynamometer. Subjects were then asked to propel at 2 different speeds: 0.9m/s (2mph) and 1.8m/s (4mph). Once the subject reached the desired steadystate speed, both kinematic and kinetic data were collected for 20 seconds. Subjects were given approximately 1 minute to rest between trials. The SMART Wheel was used to determine the forces and moments applied to the pushrim, as well as the stroke frequency. Specifically, we analyzed the total force applied to the pushrim and the cadence. 23 Data Analysis Each 20-second trial produced a data file containing the position data of each marker. By using this data, a local coordinate system for the wrist was developed (fig 1). That system is described in our earlier work. 24 The first axis was defined as a line pointing from the midpoint between the radial and ulnar styloids to the third MCJ. The second axis was defined as perpendicular to axis 1 and pointed out of the back of the hand and wrist. Rotations about this axis represented ulnar and radial deviation. The third axis was perpendicular to the plane formed by the first 2 axes and pointed from the midpoint of the wrist through the radial styloid. Movements about this axis represented flexion and extension. This local coordinate system was used to determine the flexion and extension and ulnar and radial deviation of every subject for both speeds. The motion data for each subject were plotted and 5 representative strokes were selected for both speeds (0.9, 1.8m/s). Representative strokes were selected through visual inspection by one of the investigators. This step assured that errors resulting from lost markers were not reflected in the data analysis. Only strokes with obvious kinematic abnormalities were not used. Variables were averaged over these 5 strokes. Data Reduction Research has found a correlation in propulsion technique between the right and left side as well as between various speeds of propulsion. 15 Also, NCSs have found that right and left sides correlate. 15 All left and right side propulsion data were significantly related (r.458, P.01), therefore, the 2 sides were averaged together. We used these variables to determine differences between speeds. To further reduce our biomechanic variables, we performed correlations between the biomechanic data at 0.9 and 1.8m/s. These data correlated significantly (r.758, P.01). Therefore, average variables representative of left and right sides as well as both speeds were calculated for use in our regression. For the NCS data, the correlation between left and right was significant (r.473, P.01), and therefore the left and right side data were combined for each NCS variable. Statistical Analysis Distributions of the data were examined and means and standard deviations (SDs) calculated. All statistical calculations were completed with the SPSS. d Changes in kinematics variables across speeds were investigated by using a paired t test. Pearson correlations were computed as an exploratory analysis of the relation between motion and median and ulnar

3 WHEELCHAIR KINEMATICS AND NERVE FUNCTION, Boninger 1143 Fig 1. Axes of local coordinate system. Rotation around axis 3 represents flexion and extension. Rotation around axis 2 represents radial and ulnar deviation. nerve function. Pearson correlations were also used to investigate the relation between subject characteristics (height, weight, age, years since SCI) and nerve function, as well as kinematics. Relationships with a P value less than 0.1 were then entered into a stepwise linear regression analysis, with nerve function variables being the dependent variables and subject characteristics and motions being the independent variables. Because of unexpected findings, we completed a secondary analysis to investigate a possible correlation between wrist ROM and pushrim kinetic variables of frequency, resultant force, and nonplanar moment. We averaged the highly correlated left and right side kinetic data and then averaged the highly correlated speed data. RESULTS Subject Characteristics Thirty-five subjects (27 men, 8 women) were recruited for the study. They were between the ages of 18 and 65 years (avg age, 38.9y). The average time from injury data to test date was 12.2 years (range, ). Subjects average weight was kg and height was cm. No subject was eliminated because of abnormalities in the nerves tested. Not surprisingly, age was significantly (P.05) related to median sensory amplitude (r.471) and latency (r.461) and to ulnar sensory amplitude (r.473) and latency (r.648) but not to motor latencies or amplitudes. Years with SCI was not related to NCS. Height was significantly related to ulnar sensory amplitude (r.541), ulnar motor latency (r.446), and median sensory amplitude (r.419). Weight was significantly related to median sensory amplitude (r.456) and latency (r.515); to ulnar sensory amplitude (r.377) and latency (r.380); and to median motor latency (r.403). Wrist Kinematics Mean peak angles and ROMs are shown in table 1. The paired t test found no significant differences between speeds for each of these measurements. However, the difference between mean peak wrist flexion approached significance (P.061), with less motion at the faster speed. The mean ROMs were compared with subject characteristics. We found significant negative correlations between wrist flexion and extension ROM and age (r.351, P.05) and weight (r.547, P.01). Wrist Kinematics and Nerve Function No significant correlation was found between ulnar or median latencies and wrist kinematics. A significant positive correlation was found between wrist flexion and extension ROM and median motor amplitude (r.361, P.05; fig 2), as well as ulnar motor amplitude (r.383, P.05; fig 3). Based on this finding and the correlational analysis of subject characteristics, Table 1: Mean Wrist Angles for Each Speed Measured in Degrees 0.9m/s Mean SD 1.8m/s Mean SD Peak wrist flexion* Peak wrist extension Peak ulnar deviation Peak radial deviation Flexion/extension ROM Ulnar/radial deviation ROM *There were no significant differences between speeds, but peak wrist flexion approached significance with P.061.

4 1144 WHEELCHAIR KINEMATICS AND NERVE FUNCTION, Boninger Table 2: Linear Stepwise Regression Results Dependent Variable Variables Entered Into Model r P Median motor amplitude Flexion/extension ROM Stature and flexion/ extension ROM Ulnar motor amplitude Flexion/extension ROM Fig 2. Median motor amplitude versus flexion and extension ROM. Scatterplot of data with regression line. separate regressions were conducted with median motor amplitude and ulnar motor amplitude as the dependent variables. For both regressions, the independent variables were wrist ROM, subject age, height, and weight. The results of this analysis are presented in table 2. Regression analysis showed that higher ranges of wrist motion were associated with increased median and ulnar motor amplitudes. Wrist Kinematics and Pushrim Kinetics Contrary to our hypotheses, we found positive correlations between wrist ROM and median and ulnar nerve amplitude. We therefore conducted an analysis to investigate the relation between wrist ROM and the pushrim kinetic variables of stroke frequency, resultant force, and nonplanar moment. The nonplanar moment is the sum of the moments exerted on the handrim that are not about the axle of the wheel and therefore do not lead to forward motion. The mean resultant force was 90.3N and the mean stroke frequency was 1.0Hz. Note that these variables represent a combination of the left and right side and 2 speeds. We found significant negative correlations between wrist flexion and extension ROM and stroke frequency (r.372, P.05; fig 4) and resultant force (r.362, P.05). DISCUSSION To our knowledge, this is the largest study that has investigated kinematics at the wrist during wheelchair propulsion, and the only study that has examined the relation between kinematics and nerve function. Our kinematic results are in general agreement with results of previous studies. 21,24-27 Direct comparison with other studies is difficult because of differences in marker systems and computational techniques. As seen in our previous study, 21 there was a trend toward reduced ROM at higher speeds. As seen in other studies, 21,25-27 the overall ROM in flexion and extension is similar to the overall ROM in ulnar and radial deviation. The most interesting finding in this study was the relation between wrist ROM and both median and ulnar nerve amplitudes. The positive correlations indicate that larger ROM was actually related to better nerve health. This finding was contrary to our hypotheses, which were informed by ergonomics literature that has found larger ROM to be associated with greater nerve injury. 17,18,20 Our hypotheses were supported by the studies indicating that greater wrist flexion and extension increased pressure in the carpal canal. 2,19 When we investigated the effect of ROM on other propulsion biomechanics, we discovered that wheelchair users who had a larger ROM also pushed with a slower cadence or contacted the pushrim less frequently to go at the same speed. In addition, the large ROM was associated with lower peak force. This finding is somewhat intuitive. The users who had increased ROM propelled with a longer smooth stroke, and this likely resulted in the observed positive effects. It is possible that wheelchair users with less ROM are injuring their nerves because of the higher force and cadence they use when propelling their wheelchairs. Note that amplitude was related to kinematics and not to latency. Although latency is thought to be an earlier marker of nerve injury in CTS, amplitude changes may signify more significant nerve damage. This study s results add to the mounting evidence that wheelchair users should Fig 3. Ulnar motor amplitude versus flexion and extension ROM. Scatterplot of data with regression line. Fig 4. Average frequency versus flexion and extension ROM. Scatterplot of data with regression line. Frequency or cadence is in strokes per second.

5 WHEELCHAIR KINEMATICS AND NERVE FUNCTION, Boninger 1145 consider using long, smooth strokes to propel, even if ROM at the wrist is increased. This type of propulsive stroke may reduce the chances of developing CTS. It is possible that our studies with wheelchair users have a broader application to ergonomics and factory workers. The results of this study, combined with our previous work, indicate that higher forces and frequency have a greater potential to cause nerve damage than does ROM. 15 Our studies of wheelchair users have certain advantages over ergonomic studies of factory workers. Wheelchair propulsion is a highly repetitive task that likely has less variation than many industrial jobs. With the SMART Wheel, we can precisely measure forces. Most ergonomic studies estimate force by job site evaluation, 10 through electromyography, 11,12 or through knowledge of the mass of the tool. 13 Finally, a dynamometer captures repeatable kinematic data on multiple strokes. This study is correlative in nature and does not prove cause and effect. Also, the study is limited in that our sample size did not allow us to fully factor in all variables such as subject symptoms and physical examination findings. It is possible that increased wrist ROM is associated with a third factor, such as wrist joint architecture or connective tissue flexibility, that leads to healthier nerves. Although a significant relationship was found, the ROM only explained some of the variation in NCSs. It is likely that transfers and other subject characteristics have an impact on injury risk. Although this is among the largest wheelchair biomechanics studies conducted, a larger sample size that would allow for more complex statistical modeling is warranted. In addition, future studies should follow subjects longitudinally so that more insight into cause and effect can be determined. CONCLUSIONS We found that larger ROM at the wrist was related to larger amplitudes in median and ulnar motor nerve responses. We further found that larger wrist ROM was associated with lower forces, moments, and cadence during the propulsive stroke. Clinicians should consider advising manual wheelchair users to take long smooth strokes when propelling a wheelchair as a possible way to reduce the risk of injury. Further study is needed with large sample sizes followed over time. In addition, studies are needed to determine the ideal means of training a wheelchair user on how to push correctly. References 1. Aljure J, Eltorai I, Bradley WE, Lin JE, Johnson B. Carpal tunnel syndrome in paraplegic patients. Paraplegia 1985;23: Gellman H, Chandler DR, Petrasek J, Sie I, Adkins R, Waters RL. Carpal tunnel syndrome in paraplegic patients. J Bone Joint Surg Am 1988;70: Tun CG, Upton J. The paraplegic hand: electrodiagnostic studies and clinical findings. J Hand Surg [Am] 1988;13: Davidoff G, Werner R, Waring W. Compressive mononeuropathies of the upper extremity in chronic paraplegia. Paraplegia 1991;29: Burnham RS, Steadward RD. Upper extremity peripheral nerve entrapments among wheelchair athletes: prevalence, location, and risk factors. Arch Phys Med Rehabil 1994;75: Sie IH, Waters RL, Adkins RH, Gellman H. Upper extremity pain in the postrehabilitation spinal cord injured patient. Arch Phys Med Rehabil 1992;73: Stefaniwsky L, Bilowit DS, Prasad SS. Reduced motor conduction velocity of the ulnar nerve in spinal cord injured patients. Paraplegia 1980;18: Lundqvist C, Siosteen A, Blomstrand C, Lind B, Sullivan M. Spinal cord injuries. clinical, functional, and emotional status. Spine 1991;16: Dalyan M, Cardenas DD, Gerard B. Upper extremity pain after spinal cord injury. Spinal Cord 1999;37: Werner RA, Franzblau A, Albers JW, Armstrong TJ. Median mononeuropathy among active workers are there differences between symptomatic and asymptomatic workers? Am J Ind Med 1998;33: Silverstein BA, Fine LJ, Armstrong TJ. Occupational factors and carpal tunnel syndrome. Am J Ind Med 1987;11: Loslever P, Ranaivosoa A. Biomechanical and epidemiological investigation of carpal tunnel syndrome at workplaces with high risk factors. Ergonomics 1993;36: Roquelaure Y, Mechali S, Dano C, et al. Occupational and personal risk factors for carpal tunnel syndrome in industrial workers. Scand J Work Environ Health 1997;23: Keir PJ, Wells RP, Ranney DA, Lavery W. The effects of tendon load and posture on carpal tunnel pressure. J Hand Surg [Am] 1997;22: Boninger ML, Cooper RA, Baldwin MA, Shimada SD, Koontz A. Wheelchair pushrim kinetics: body weight and median nerve function. Arch Phys Med Rehabil 1999;80: Armstrong TJ, Chaffin DB. Carpal tunnel syndrome and selected personal attributes. J Occup Med 1979;21: Gelberman RH, Hergenroeder PT, Hargens AR, Lundborg GN, Akeson WH. The carpal tunnel syndrome. A study of carpal canal pressures. J Bone Joint Surg Am 1981;63: Werner CO, Elmqvist D, Ohlin P. Pressure and nerve lesion in the carpal tunnel. Acta Orthop Scand 1983;54: Tanzer RC. The carpal-tunnel syndrome. J Bone Joint Surg Am 1959;41: Lundborg G, Gelberman RH, Minteer-Convery M, Lee YF, Hargens AR. Median nerve compression in the carpal tunnel functional response to experimentally induced controlled pressure. J Hand Surg [Am] 1982;7: Boninger ML, Cooper RA, Robertson RN, Shimada SD. Wrist biomechanics during wheelchair propulsion: a full description using a local coordinate system in 3-D space [abstract]. Arch Phys Med Rehabil 1996;77: Cooper RA, Robertson RN, VanSickle DP, Boninger ML, Shimada SD. Methods for determining three-dimensional wheelchair pushrim forces and moments: a technical note. J Rehabil Res Dev 1997;34: Boninger ML, Cooper RA, Robertson RN, Shimada SD. Threedimensional pushrim forces during two speeds of wheelchair propulsion. Am J Phys Med Rehabil 1997;76: Shimada SD, Cooper RA, Boninger ML, Koontz AM, Corfman TA. Comparison of three different models to represent the wrist during wheelchair propulsion. IEEE Trans Neural Syst Rehabil Eng 2001;9: Rao SS, Bontrager EL, Gronley JK, Newsam CJ, Perry J. Threedimensional kinematics of wheelchair propulsion. IEEE Trans Rehabil Eng 1996;4: Newsam CJ, Rao SS, Mulroy SJ, Gronley JK, Bontrager EL, Perry J. Three dimensional upper extremity motion during manual wheelchair propulsion in men with different levels of spinal cord injury. Gait Posture 1999;10: Veeger HE, Meershoek LS, van der Woude LH, Langenhoff JM. Wrist motion in handrim wheelchair propulsion. J Rehabil Res Dev 1998;35: Suppliers a. Nicolet Instrument Corp, 5225 Verona Rd, Madison, WI b. Northern Digital Inc, 103 Randall Dr, Waterloo, ON N2V 1C5. c. Three Rivers Holdings, 1826 West Broadway Rd, Ste 43, Mesa, AZ d. SPSS Inc, 233 S Wacker Dr, 11th Fl, Chicago, IL