Motion analysis of wheelchair propulsion movements in hemiplegic patients: effect of a wheelchair cushion on suppressing posterior pelvic tilt.

[Purpose] This study sought to ascertain whether, in hemiplegic patients, the effect of a wheelchair cushion to suppress pelvic posterior tilt when initiating wheelchair propulsion would continue in subsequent propulsions. [Subjects] Eighteen hemiplegic patients who were able to propel a wheelchair in a seated position participated in this study. [Methods] An adjustable wheelchair was fitted with a cushion that had an anchoring function, and a thigh pad on the propulsion side was removed. Propulsion movements from the seated position without moving through three propulsion cycles were measured using a three-dimensional motion analysis system, and electromyography was used to determine the angle of pelvic posterior tilt, muscle activity of the biceps femoris long head, and propulsion speed.
[Results] Pelvic posterior tilt could be suppressed through the three propulsion cycles, which served to increase propulsion speed. Muscle activity of the biceps femoris long head was highest when initiating propulsion and decreased thereafter.
[Conclusion] The effect of the wheelchair cushion on suppressing pelvic posterior tilt continued through three propulsion cycles.

INTRODUCTION

Hemiplegic patients can acquire wheelchair propulsion movements relatively quickly, and therefore those who have ambulatory difficulties or who are receiving training before being able to walk again use wheelchairs. However, hemiplegics are prone to posterior tilting of the pelvis due to factors such as wheelchair depth, hamstring contraction, and sitting for long periods1), and these factors are difficult for many people to change. To date, few studies have investigated the effect of long-term pelvic posterior tilt on wheelchairs2,3,4), and compared with research on therapy that aims to reestablish ambulation, research on wheelchair posture tends to receive less attention. Kinose et al. noted wheelchair problems as a reason for this5); that is, most wheelchairs in hospitals and therapeutic facilities are likely to be standard models with dimensions and a seat that cannot be adjusted for individuals. Less than an optimal posture while propelling a wheelchair can easily aggravate abnormal muscle tension and associated responses, which can then become major clinical problems in themselves6).

The dimensions of wheelchairs and their seats have been noted as possible causes of pelvic posterior tilt7), and the use of a wheelchair cushion is therefore recommended8). The effect of using a wheelchair cushion with anchor support6) and notches for the thigh section on the propulsion side9) have been reported for hemiplegic subjects. The importance of heel-ground contact7), footrest height10), and seat height11) in wheelchair propulsion by hemiplegic subjects has been previously reported. However, these reports often use propulsion speed as an evaluation indicator, and there are no reports that have quantitatively evaluated the angle of pelvic posterior tilt. We previously found that pelvic posterior tilt could be suppressed by using a wheelchair cushion in hemiplegics under the conditions of being seated without moving up to initiation of proplusion12). However, this effect of the cushion needs to be examined in more detailed by investigating how the angle of pelvic posterior tilt changes under the conditions of being seated through initiation and continuation of propulsions. Thus, in this study, we investigated the continuity of the effect of using a cushion to suppress pelvic posterior tilt through three propulsion cycles. In addition, because hemiplegics use the lower limb on the propulsion side to control both propulsion and steering when propelling a wheelchair7), the wheelchair propulsion movement examined in this study was one-sided leg propulsion.

SUBJECTS AND METHODS

The subjects were 18 hemiplegics (17 men and 1 woman; age range, 44–73) who could propel a wheelchair while in a seated position. Subjects were excluded if they had severe spinal deformation, marked sensory impairment, or severe higher brain dysfunction and could not understand an explanation of the study. The physical functions and disease characteristics of the subjects are shown in Table 1. Nine subjects had right hemiplegia, and 9 had left hemiplegia (112.5±39.8 days from disease onset to day of measurement). Physical function was assessed using the leg Brunnstrom Recovery Stage and the Functional Independence Measure. We assumed that age and sex differences would have a negligible effect on the results. Although such effects are a possibility, previous studies that were mixed sex and that considered a broad range of ages have not reported age or sex differences11, 12).

Table 1.

Physical functions and disease characteristics of the hemiplegic subjects

Subject	Sex	Age	Diagnosis	Paralyzed side	Period from onset (days)	Height (cm)	Weight (kg)	BRS score	FIM score
a	Male	51	Cerebral infarction	Left	111	168.2	74.0	IV	125
b	Female	72	Multiple cerebran infarction	Left	166	156.0	44.8	IV	105
c	Male	53	Cerebral hemorrhage	Right	89	173.4	63.0	IV	99
d	Male	60	Cerebral infarction	Left	141	166.7	64.0	IV	113
e	Male	61	Cerebral hemorrhage	Right	106	160.3	63.4	III	95
f	Male	62	Cerebral hemorrhage	Left	83	168.1	50.3	III	66
g	Male	55	Thalamic hemorrhage	Left	63	161.3	51.0	III	109
h	Male	63	Cerebral hemorrhage	Right	116	168.4	78.0	IV	100
i	Male	44	Cerebral hemorrhage	Right	141	162.5	51.0	III	114
j	Male	62	Thalamic hemorrhage	Right	106	165.2	65.5	IV	122
k	Male	63	Cerebral hemorrhage	Left	147	165.3	66.9	III	79
l	Male	49	Cerebral hemorrhage	Right	62	167.1	54.3	II	76
m	Male	73	Thalamic hemorrhage	Left	95	158.0	44.7	III	100
n	Male	67	Thalamic hemorrhage	Right	140	160.4	48.4	II	76
o	Male	65	Putamen hemorrhage	Right	211	163.5	50.0	IV	66
p	Male	67	Thalamic hemorrhage	Right	116	160.7	52.3	III	113
q	Male	72	Cerebral infarction	Left	91	160.3	45.1	III	121
r	Male	69	Cerebral hemorrhage	Left	41	173.4	68.5	II	99

BRS: Brunnstrom Recovery Stage; FIM: Functional Independence Measure

Measurements were obtained under two conditions: 1) subjects seated without moving in an adjustable wheelchair and 2) from the seated position through three propulsion cycles of travel in a straight line under one-sided leg propulsion. Maximum isometric contraction of the long head of the biceps femoris was measured for 5 s in the prone position. The initial ground contact of the propelling foot through to the next ground contact of the same foot was defined as the initiation of propulsion.

We used the same wheelchair cushion in this study as in our previous study12). The cushion had an anchoring function, and the thigh pad on the propulsion side was removed. The thickness of the cushion was 8 cm (difference in elevation of 2.5 cm). The cushion comprised a polyethylene foam pad underneath a low-rebound, high-density urethane pad and a polyethylene cushion cover.

To measure kinematic data, we used a three-dimensional (3D) motion analysis system (Vicon MX), consisting of 8 infrared cameras, and an electromyograph (DKH). All patients used the same adjustable wheelchair (Nissin Medical Industries Co., Ltd.). The camera sampling frequency was 120 Hz. The propulsion pathway was 3.6 m, and the measurements were made starting at the 1.8 m mark and continued thereafter. Infrared reflective markers were pasted at the following 15 locations: bilateral acromion, spinous process of the second thoracic vertebra, bilateral anterior superior iliac spine (ASIS), bilateral posterior superior iliac spine (PSIS), hip joint, knee joint, external malleolus, and head of the fifth metatarsal. To calculate the pelvic angle, the bilateral ASIS and PSIS markers were used to define a pelvic coordinate system.

Electromyography was used to analyze muscle activity during the propulsion movements and was synchronized to the 3D motion analysis system. Data were sampled at 1,080 Hz and input into a PC after A/D conversion. The waveform of the electromyogram was processed with a 20–450 Hz band-pass filter and subjected to full-wave rectification to determine the integrated electromyogram (IEMG). The muscle measured was the biceps femoris long head, and the electrodes were attached at the positions recommended by Aldo et al13).

The adjustable wheelchair had 24-inch propulsion wheels and 5-inch casters, and the height difference of the seat from front to back was set at 0 cm. The bottom of the back support was raised so that the markers attached to the pelvis were visible from behind. The five adjustable parameters of the wheelchair were the seat height, depth, and width and the heights of the arm and back supports. Adjustment of the wheelchair to patient dimensions was conducted by referring to the Wheelchair SIG workshop text (2003).

The parameters analyzed were the pelvic posterior tilt angle when initiating propulsion through three propulsion cycles, muscle activity of the biceps femoris long head, and propulsion speed. The pelvic posterior tilt angle around the X-axis was calculated using Vicon Body Builder Ver. 3.6 after inputting the marker positions measured by the VICON motion analysis system into a PC. As in a previous study14), we used as the measurement point the peak bending moment of the hip (peak bending) on the propulsion side initially when seated without moving and then during each propulsion cycle. The hip joint angle was calculated using the DIFF Gait, Wave Eyes software from the Clinical Gait Analysis Forum of Japan.

For the muscle activity of the biceps femoris long head, the IEMG was determined from the ground contact of the propelling foot to the peak bending moment in each propulsion cycle. As a procedure to normalize the results between subjects, the IEMG during the maximum isometric contraction of the biceps femoris long head was taken as 100%, and the results were converted to an IEMG proportion (%IEMG). Maximum isometric contraction was measured for 5 s, but only 3 s were used for the IEMG calculation.

The value resulting from dividing the amount of movement from the ground contact of the propelling foot through to the next ground contact of the same foot by the time taken to complete the movement was defined as the propulsion speed. The distance was derived from the marker position at the head of the fifth metatarsal and the ankle measured by the Vicon system and was calculated from the required time.

The pelvic posterior tilt angle, the biceps femoris long head %IEMG, and the propulsion speed were investigated by one-way analysis of variance with time as a factor and by the multiple comparison method (Dunnett’s test). Statistical significance was set at p<0.05. Statistical analysis was performed in SPSS ver. 15.0 J for Windows.

This study was approved by the Ethics Review Committee of the International University of Health and Welfare (09-122) and was conducted after obtaining consent from each subject and his or her attending physician.

RESULTS

The pelvic posterior tilt angle from the seated position without moving through three propulsion cycles, the propulsion speed for each propulsion cycle, and the biceps femoris long head %IEMG are shown in Table 2. A significant difference was found in the pelvic posterior tilt angle between being seated without moving and the initiation of propulsion (p<0.05), between being seated without moving and the second propulsion cycle (p<0.05), and between being seated without moving and the third propulsion cycle (p<0.05). Significant differences were found in the propulsion speed and biceps femoris long head %IEMG between the initiation of propulsion and the second propulsion cycle (p<0.01) and between the initiation of propulsion and the third propulsion cycle (p<0.01).

Table 2.

Measurement values from sitting through three wheelchair propulsion cycles

Parameter	Stationary	Start of propulsion	Second propulsion cycle	Third propulsion cycle
Pelvic posterior tilt angle (°)	11.4±3.9	13.5±4.0	13.6±4.3	13.7±4.0
BFLH %IEMG (%)		20.2±8.3	11.9±6.7	8.7±5.1
Propulsion speed (m/s)		0.29±0.06	0.40±0.08	0.46±0.10

Values indicate the mean±SD of the 18 subjects.

DISCUSSION

This study revealed significant differences in the pelvic posterior tilt angle and propulsion speed between being seated without moving and initiating propulsion and between being seated without moving and the second and third propulsion cycles. In addition, the propulsion speed had increased by the third cycle. The pelvic posterior tilt angle changed over time, and was 11° when sitting still; it was later found to be 13°. Thus, the results of anchoring the cushion and removing the thigh pad from the propulsion side in the present study showed results similar to those of the studies by Takeda et al.6), Cron9), and Engstrom15). Furthermore, the effect of a wheelchair cushion suppressing pelvic posterior tilt when sitting still as investigated by Kawada et al.12) was shown to continue through the third propulsion cycle. We also found that the effect continued through to the third propulsion cycle, and it was possible to quantitatively express the pelvic posterior tilt angle over time. Since the pelvis tilted posteriorly from the seated position through to initiation of propulsion and did not change greatly after that, the time from the seated position without moving to initiation of propulsion is important for us to consider when attempting to suppress pelvic posterior tilting in one-sided leg propulsion by a hemiplegic patient. This novel finding is a potential point of focus when observing one-sided leg propulsion movements in the clinical setting.

A significant difference in the biceps femoris long head %IEMG was found between initiation of propulsion and the second and third propulsion cycles. This suggests that high muscle activity in the biceps femoris long head is necessary for moving the combined mass of the wheelchair and the patient when initiating propulsion, and with the acceleration obtained, the wheelchair could subsequently be moved by exerting a lower amount of force16). So, the patient could move forward using a lower force from the second propulsion cycle onward. As the propulsion speed increased, it is likely that the effect of the wheelchair cushion on suppressing pelvic posterior tilt continued through the third propulsion cycle.

To conclude, analysis of one-sided leg propulsion revealed continuity of the effect of a wheelchair cushion on suppressing pelvic posterior tilt from the seated position without moving through three propulsion cycles. Future research should compare cushions of differing materials and shapes, verify their long-term effects, and investigate potential application to wheelchair cushion design.