Wheelchair Axle Position Effect on the Propulsion Performance of Persons With C7 Tetraplegia: A Repeated-Measures Study

Abstract

Objectives

To assess the changes in speed, stroke frequency, acceleration, and shoulder range of motion (ROM) associated with different wheelchair axle positions in people with chronic C7 tetraplegia.

Methods

This repeated-measures study was conducted at the Chronic Spinal Cord Injury Unit, FLENI Escobar, Argentina. The speed, stroke frequency, acceleration, and shoulder ROM during wheelchair propulsion were measured in nine participants with C7 spinal cord injury (SCI) in four different axle positions (forward and up, forward and down, backward and down, backward and up). Two strokes performed at maximum speed were analyzed on a smooth level vinyl floor in a motion analysis laboratory. Data were analyzed for significant statistical differences using the Friedman test and the Wilcoxon signed rank test.

Results

Our study showed significant differences in the speed with axle position 1 (1.57 m/s) versus 2 (1.55 m/s) and position 2 (1.55 m/s) versus 4 (1.52 m/s). The shoulder ROM showed a significant difference in the sagittal plane in position 2 (59.34 degrees) versus 3 (61.64 degrees), whereas the stroke frequency and the acceleration parameters showed no statistically significant differences with the different rear axle positions.

Conclusions

Our study showed that modifying the rear axle position can improve the propulsion speed and produce changes in the shoulder ROM in the wheelchair propulsion of individuals with C7 SCI.

Introduction

Many persons with spinal cord injury (SCI) are dependent on a wheelchair for their primary daily mobility. Different research studies have examined propulsion patterns,¹ upper limb kinematics,^2,3 electrical stimulation during wheelchair propulsion,^4,5 and forces applied on the wheelchair pushrim^6,7 to optimize the wheelchair propulsion.

In recent years, the wheelchair design, the frame composition, and the use of accessories led to improvement in the performance of the users.^8,9 Wheelchair configuration is key to optimize the gross mechanical efficiency and to improve speed, acceleration, and stroke frequency.¹⁰ The guidelines for upper limb preservation recommend adjusting the rear axle as far forward as possible without compromising the stability of the user with an angle between the upper arm and forearm of 100 to 120 degrees¹¹ when the hand is placed at the top dead-center position on the push rim to improve the propulsion biomechanics¹¹ and reduce the metabolic cost and muscle stress.¹² Kotajarvi et al.¹³ examined the effect of the wheelchair rear axle position on handrim biomechanics in people with paraplegia and found that low seat positions resulted in an improvement in the timing variables, though no significant differences were observed in the speed variable. Boninger et al.,¹⁴ in a group of 40 paraplegic manual wheelchair users, found that a reduced vertical distance between the axle and the shoulder and a more forward axle position were correlated with improvements in the wheelchair propulsion biomechanics. Masse et al.¹⁵ investigated six seating positions in five participants with paraplegia and found that the middle and backward position had the lowest pushing frequency.

In this way, Freixes et al.¹⁶ assessed changes in the speed, acceleration, stroke frequency, and shoulder range of motion (ROM) associated with different wheelchair axle positions in people with chronic C6 tetraplegia. This research determined that the up and forward axle position improved the speed and acceleration but increased the stroke frequency.

Individuals with C7 tetraplegia represent a special subset of wheelchair users. Their shoulder muscle activity is similar to that of individuals with paraplegia, but the hand grip is similar in function to those individuals with C6 tetraplegia (i.e., having to propel with gloves and predominance of axial forces).¹⁷ Moreover, the mean propulsion velocity in participants with C7 motor complete tetraplegia was significantly lower than in the group with paraplegia but significantly higher than in the group with C6 tetraplegia.¹⁷

To the best of our knowledge, there are no publications that have studied the relationship between wheelchair axle configuration and propulsion kinematics in individuals with a C7 SCI.

The purpose of this study was to analyze the kinematic changes during wheelchair manual propulsion associated with different rear axle positions in users with C7 tetraplegia. Our hypothesis is that the up and forward axle position increases both speed and acceleration with lower stroke frequency.

Methods

The present study was approved by the institutional review board and ethics committee of FLENI Institute. All the participants signed a written consent form to participate in the trial. Wheelchair users from our spinal cord injuries rehabilitation unit patient database were contacted via telephone and prospectively included to take part in the study. The inclusion criteria were the following: people with chronic (≥12 months) SCI C7 AIS A and B in accordance with the American Spinal Injury Association Impairment Scale (AIS),¹⁸ without grasping capacity, and who had completed a standard rehabilitation program that included the training of wheelchair skills and had normal trunk, upper limb, and pelvic ROM.

All participants had been using rigid frame manual wheelchairs. Wheelchair axle position varied among the study population. We did not include individuals who presented chronic upper limb pain, respiratory diseases, or heart failure or those who were unable to give informed consent owing to cognitive impairment.

Nine participants, eight with C7 AIS A SCI and one with C7 AIS B SCI, were studied. All of them were dependent on a wheelchair for their primary daily mobility. The characteristics of participants are detailed in Table 1.

Table 1.

Participants’ characteristics

Participants	Age	Sex	ASIA motor level	UEMS R and L	Time since injury, months
1	44	M	C7 A	28	28
2	26	M	C7 A	29	26
3	22	F	C7 A	30	33
4	54	F	C7 A	27	83
5	22	M	C7 A	30	35
6	37	F	C7 A	28	26
7	27	M	C7 A	28	12
8	25	M	C7 B	27	25
9	57	M	C7 A	26	20

Note: F = female; L = left; M = male; R = right; UEMS = upper extremity motor score.

The participants were evaluated in a motion analysis laboratory (ELITE – BTS, Italy). Six infrared cameras were located surrounding a capture volume of 600 × 150 × 180 cm. The system collected the coordinates of retroreflective markers attached to the right acromial end, seventh cervical vertebrae, right olecranon, styloid process of the ulna and radius, third metacarpal head, sacrum, and the wheel axle. To avoid optical occlusion, the sacral marker was placed on the wheelchair back. With the aim of locating the right position of the sacrum, one virtual marker was generated in processing time and referred to the original sacral marker at a distance equivalent to the chair back width. Two extra markers were placed on the wheels at a fixed distance from the axis. The markers were placed only on the right side of the participant and chair. Marker paths were acquired at 100 Hz and filtered with a second-order zero-lag Butterworth low-pass filter (fc = 6 Hz). We used the method described by Rab et al.¹⁹ to calculate the upper limb angular kinematics. Data acquisition and processing was performed using Biomech-Analyzer software (BTS, Italy).

Adjustable wheelchair

The study was performed using a Quickie R2 rigid frame ultra-lightweight wheelchair (Sunrise Medical, Longmont, CO) equipped with a 16 in. (41 cm) high JAY 2 medium backrest and with 26 in. (66 cm) diameter rear wheels, 23 in. (59 cm) plastic-coated handrim, 16 in. (41 cm) seat width, 16 in. (41 cm) seat depth, 4 in. (10 cm) polyurethane front wheels, and 0 degree camber angle. The rear axle position was modified in the vertical (Y) and the longitudinal direction (X). We established four rear axle positions according to the most frequently prescribed configurations in our rehabilitation center: forward and up axle (position 1), forward and down axle (position 2), backward and down axle (position 3), and backward and up axle (position 4). Moving the axle forward generates a seat unit that is posterior in relation to the rear wheels, whereas moving the axle backward results in a seat unit that is anterior relative to the rear wheels. Moving the posterior seat down produces a seat unit that is lower in relation to the rear wheels, whereas moving the posterior seat up results in a seat unit that is higher relative to the rear wheels. In the four axle positions, the backrest angle was adjusted so that it was always the same with respect to the horizontal plane (16 degrees). The position of the participants was standardized to the chair from a marker on the shoulder to axle position 4 on a 0.752 m distance (SD ± 0.024). Axle position was then modified anteriorly 0.25 in. (0.63 cm) and 1.75 in. (4.44 cm) in relation to the back seat and downward 4.5 in. (11.4 cm) and 6.5 in. (16.5 cm) with respect to the seat (Figure 1). The backrest and footrest heights of the wheelchair were adjusted in such a way that their position was similar to the usual configuration of the participant’s personal chair.

Figure 1.

Seat position matrix. 1 = position 1; 2 = position 2; 3 = position ;, and 4 = position 4.

Data collection

Data were collected as participants propelled the instrumented wheelchair on a 10-m section of smooth level vinyl floor. Each participant had a 5-minute warm up session before data gathering. Participants were requested to propel the wheelchair for five trials at maximum speed with each axle position. The first two trials served as practice sessions so the participants could become familiar with the test conditions, and the last three trials were used for data recording. There were 5-minute breaks between trials. Each trial consisted of one ramp up stroke from a still starting position (which was not analyzed because the retroreflective markers could not be detected by the infrared cameras) and two full strokes that were analyzed. All strokes were performed at maximum speed. A propulsion cycle was defined as the period between the initial contact of the user’s hand on the wheelchair hand rim until the next initial contact. The initial contact was obtained by visual cues using the vertical coordinates from the center of rotation of the wrist.

Axle positions were modified at random with a random numbers table for each participant and throughout the different participants along the study. To prevent falls or potential injuries after sudden decelerations, all participants were fastened with a safety strap at the middle trunk that did not prevent them from leaning forward during the trials.

Data analysis

The speed, stroke frequency, acceleration, and shoulder ROM during wheelchair propulsion were analyzed. Speed and acceleration were calculated as the average of the instant values during the entire trial. Speed was calculated as the average value of the linear velocity of the sacral marker, and acceleration was obtained as a first temporal derivative of velocity. The stroke frequency was defined as the number of propulsion cycles per minute. The shoulder ROM was obtained as the difference between the maximum and the minimum degrees of the shoulder rotation curves in the three different planes. The final value is the average of the two strokes of the three trials analyzed for each participant.

Data were analyzed for significant statistical differences using the Friedman test and the Wilcoxon signed rank test. Statistical significance was set at a p value of ≤.05.

Results

Speed

Statistically significant differences were observed in the speed at different axle positions. When analyzing the comparison of related pairs, only position 1 (1.57 m/s) versus 2 (1.55 m/s) and position 2 (1.55 m/s) versus 4 (1.52 m/s) showed significant differences (Tables 2 and 3).

Table 2.

Speed, stroke frequency, and acceleration at different axle positions

Kinematic parameters	Axle positions				Friedman test
	1. Forward and up	2. Forward and down	3. Backward and down	4. Backward and up	p value
Speed, m/s
Median (Min–Max)	1.57 (0.97–2.21)	1.55 (0.91–2.12)	1.55 (0.83–2.04)	1.52 (0.95–2.26)	.14*
Stroke frequency (circles/s)
Median (Min–Max)	1.42 (0.92–1.9)	1.41 (0.9–1.83)	1.35 (0.92–1.94)	1.35 (0.94–1.88)	.67
Acceleration, m/s2
Median (Min–Max)	0.25 (0.08–1.13)	0.25 (0.05–0.79)	0.21 (0.04–0.85)	0.32 (0.08–0.7)	.124

Note: Max = maximum; Min = minimum; s = seconds.

*Significance level set at p < .05.

Table 3.

Comparison of related pairs by Wilcoxon signed rank test

	1 vs. 2	1 vs. 3	1 vs. 4	2 vs. 3	2 vs. 4	3 vs. 4
Speed, p values	.03*	.18	.18	1.0	.03*	.18

Note: Axel positions: 1, forward and up; 2, forward and down; 3, backward and down; 4, backward and up.

*Significance level set at p < .05.

Stroke frequency

No statistically significant differences were observed in the stroke frequency for the different axle positions (Table 2).

Acceleration

No statistically significant differences were observed in the acceleration for the different axle positions (Table 2).

Shoulder ROM

No statistically significant shoulder ROM differences were observed in the coronal and transverse planes, although significant differences were found in the sagittal plane. Decreased shoulder ROM in the sagittal plane was found in position 2 in comparison to position 3 (Tables 4 and 5).

Table 4.

Shoulder range of motion at different axle positions

Kinematic parameter	Results (axle position)				Friedman test
	1 (FU)	2 (FD)	3 (BD)	4 (BU)	p values
Frontal (°)
Median	19.10	19.34	17.20	22.17	.758
Min–Max	6.55–43.21	7.46–41.67	8.54–39.34	9.16–62.09
Transverse (°)
Median	22.60	21.27	23.19	25.37	.098
Min–Max	8.73–28.1	11.1–30.82	12.03–33.99	12.16–33.24
Sagittal (°)
Median	59.72	59.34	61.64	54.96	.04*
Min–Max	42.51–67.77	46.76–64.24	47.76–64.85	10.71–63.44

Note: BD = backward and down axle; BU = backward and up axle; FD = forward and down axle; FU = forward and up axle; Max = maximum; Min = minimum.

*Significance level set at p < .05.

Table 5.

Shoulder range of motion at different axle positions by Wilcoxon signed rank test

Kinematic parameter, p values	Comparison of related pairs (axle position)
	1 vs. 2	1 vs. 3	1 vs. 4	2 vs. 3	2 vs. 4	3 vs. 4
Frontal	.95	.59	.85	.76	.67	.31
Transverse	.26	.16	.07	.26	.09	.16
Sagittal	.13	.85	.26	*.008	.21	.26

Note: Axel positions: 1, forward and up; 2, forward and down; 3, backward and down; 4, backward and up.

*Significance level set at p < .05.

Axle positions

Our results show that at

1. position 1 (forward and up axle position), the participants reached the highest speed and stroke frequency in comparison to the remaining positions.

2. position 2 (forward and down axle position), stroke frequency was lower than position 1 but higher than positions 3 and 4. In the sagittal plane, significant decreased shoulder ROM difference was found versus position 3.

3. position 3 (backward and down axle position), the participants reached the lowest acceleration.

4. position 4 (backward and up axle position), the participants reached the lowest speed and shoulder ROM in the sagittal plane. Stroke frequency was the same as position 4 and lower than positions 1 and 2. Acceleration was the highest in comparison to the remaining positions.

Discussion

To our knowledge, this is the first study analyzing the impact of four different axle positions on the speed, stroke frequency, acceleration, and shoulder ROM in participants with C7 tetraplegia on a smooth level vinyl floor.

With respect to the kinematic parameters, our results show no significant differences in the acceleration, unlike the study by Freixes et al.¹⁶ that determined that the up and forward axle position increases acceleration in the C6 propulsion group on the same type of surface. Kotajarvi et al.¹³ and Masse et al.¹⁵ studied different axle positions in participants with paraplegia on an ergonometer. On the other hand, Bertolaccini et al.⁹ examined the influence of axle position on the activity of upper limb muscles. None of these three studies measured acceleration. It is important to highlight that although mass has a greater influence during acceleration, acceleration torque also increases when less weight is placed upon the casters (55% configuration).²⁰ This could explain our results because the changes we made did not modify the mass distribution of the wheelchair-user system. Stroke frequency for participants with C7 motor complete SCI also did not show significant differences with different axle positions during propulsion. Stroke frequency was studied by many authors with different results. Freixes et al.¹⁶ determined that the up and forward axle position increases the stroke frequency in the C6 propulsion group. On an ergonometer, Boninger et al.¹⁴ found that with participants with paraplegia positioned further back in the wheelchair, the frequency of propulsion decreased, but these researchers did not compare different axle positions within participants. Kotajarvi et al.¹³ did not find significant differences in the stroke frequency with nine different seat positions in a group with paraplegia also assessed on an ergonometer. In our studied population, the lack of a significant difference in this last variable may be explained by the fact that we only applied small changes to the rear axle position in participants with normal shoulder muscular activity.

We know that the up and forward axle position improved the speed in a group with C6 tetraplegia¹⁶ in contrast with Kotajarvi et al.¹³ who found that nine different axle positions had no effect on the speed in participants with paraplegia. Our research shows that the median speed of position 1 (forward and up) was the fastest, even though the difference was minimal with respect to the other positions. A lower rolling resistance and an optimal elbow angle (close to 120 degrees) in this axle position are factors that can explain this result. Finally, there exists a relationship between stroke frequency and the sagittal shoulder ROM: When the hand contact rim increases, the stroke frequency decreases. Although the ROM in the sagittal plane showed that position 2 is significantly lower in comparison to position 3, we are not able to determine whether this difference can have an impact on the daily lives of wheelchair users. Therefore, we suggest placing the elbow in an angle of between 100 and 120 degrees when the hand is placed at the top dead-center position of the push rim and the axle as forward as possible without compromising rear stability.²¹

Participants were asked to propel at maximum speed because we wanted to analyze the effect the four axle positions had on the kinematic parameters of propulsion. If a user exerts maximum effort in each trial, the momentum of each stroke will be similar, thus making it possible to examine how much impact an axle position has on the variables studied.

Even though participants were asked to propel at maximum speed, no transient tips occurred during the trials. This is probably related to the axle positions studied. They were chosen in accordance with the most frequently prescribed configurations in daily practice. Wheelchairs were neither extremely stable nor extremely unstable. This latter, consequently, could have minimized the risk of instability or falls.

When defining the best axle position for wheelchair users with C7 motor complete tetraplegia, it would be important to take into account that even though the significant differences in speed and ROM found in our study are not large, they can play a positive role in their lives. However, it is essential to consider the user′s preference as well. Future studies are needed to determine the effect of rear axle position changes on the kinetic parameters in individuals with C7 SCI.

Study Limitations

There were limitations to this study. Our study was aimed at approximating the performance of participants with C7 SCI when propelling a wheelchair with different rear axle positions. Sample size was not justified and was small (n = 9). The small sample size is explained by the difficulty in finding participants who met the strict inclusion criteria established (people with chronic SCI C7 AIS A and B, with no grasping capacity, who had completed a standard rehabilitation program that included the training of wheelchair skills and who had normal trunk, upper limb, and pelvic ROM).

The seat width was not adjustable (although the width of the wheels was), and the backrest was the same for all participants (no complaints were reported in relation to its comfort). The analysis was based on only two strokes because persons with C7 SCI do not usually propel for long distances. The markers were only placed on the right side of the patients. The rationale for this is that it was not our intention to measure any parameter related to the symmetry of the propulsion. Another limitation of this study worth mentioning is that handrim and joint forces were not collected (i.e., kinetic parameters); this prevented us from determining whether forces changed with the axle positions as well. Finally, participants were not asked to express a preference for an axle position.

Conclusion

Our study showed that modifying the rear axle position can improve the propulsion speed and produce changes in the shoulder ROM in the wheelchair propulsion of individuals with C7 spinal cord injury.