Effects of Lateral Trunk Support on Scoliotic Spinal Alignment in Persons With Spinal Cord Injury: A Radiographic Study

Methods 

Ten men and 7 women participated voluntarily in the present study after we obtained informed consents. Table 1 lists the main demographic features of the enrolled subjects. Subjects met the following criteria: (1) SCI between C4 and T12, (2) onset over 1 year, (3) sitting on the wheelchair more than 4 hours a day, (4) diagnosed thoracic or lumbar scoliosis through anteroposterior (AP) radiographs, (5) flexible scoliosis, confirmed by supine and side-lying bending radiographs. The SCI was neurologically complete, as defined by the American Spinal Injury Association.26 All subjects gave written informed consent, as approved by the Medical Ethics Committee at the National Taiwan University Hospital.

Table 1. Demographic Data of the Subjects

	Demographic Data	Value
	Subjects (n)	17
	Mean age (range) (y)	35.4 (18.3–51.7)
	Men/women (n)	10/7
	Mean body weight ± SE (kg)	58.0±2.7
	Mean body height ± SE (cm)	161.9±2.8
	Time from onset (range) (y)	17.6 (2.9–35.0)

Abbreviation: SE, standard error.

In a radiography room, the subject sat relaxed on an adjustable seating systema without any extra support except for a seat belt and foot rests (fig 1). The seat plane was level with a 90° backrest. We adjusted the dimensions of the chair according to each subject’s anthropometry. The seat depth and foot rest height were adjusted according to the thigh and leg lengths. These steps allowed the customization of the adjusted seating system to the specific subject. The backrest of the chair was made of transparent acrylic boards to prevent the degradation of the quality of the radiographic images. During the experiment, we instructed subjects to put their hands on their thighs for support and to sit as much upright as possible. Radiographs of the thoracic and lumbosacral spine in the AP and lateral directions were taken with a computerized radiography system.b The radiation dose was about 1mSv in 1 exposure. Apart from the radiographic measurements, the spinal alignment was also photographed from AP and lateral views using a digital camera. The shape of the spine was described by stickers attached to the spinous processes of the vertebrae via careful palpation. To examine the effects of LTS, 3-point lateral trunk supports made of transparent acrylic boards were then applied to the subject by an experienced occupational therapist, based on the shape of the subject’s spinal curve from the radiographs. Radiographs in the AP and lateral directions were then taken in the same way as in the without-LTS condition described above. It took about 1 hour to complete the experiment.

View Large Image Download to PowerPoint

Fig 1. The experimental setup. Each subject was seated on an adjustable seating system in place for radiographic imaging.

To quantify the changes of the spinal curve with LTS in the frontal plane, we measured Cobb angles2, 3 and scoliotic indices3 from the radiographic images in both the with- and without-LTS conditions. The Cobb method has been commonly used to measure and describe the severity of scoliosis in clinical research and practice because it is easy and quick to calculate.1 However, because only the angle of curvature defined by the ends of the scoliotic curve is obtained, the change of the whole spine is not included in the Cobb method. The scoliotic index proposed by Greenspan overcomes the limitation by considering the deviation of each involved vertebra (fig 2).3 It is calculated as follows.

View Large Image Download to PowerPoint

Fig 2. Scoliotic index for measuring scoliosis severity.

In the sagittal plane, we measured the thoracic angle on the lateral radiograph between the upper endplate of the T6 vertebra and the lower endplate of the T12 vertebra using the Cobb method, while the lumbar angle was measured from the upper endplate of the L1 vertebra to the lower endplate of the L5 vertebra.16 These 2 variables, together with the Cobb angles and scoliotic indices, were all measured by an orthopedic surgeon (TMW) using a house-developed software system in Matlab, version 6.5.c The software loaded the image and then corrected the image distortion using the lead grid before assisting the user to draw lines for the calculation of the Cobb angles and scoliotic indices.

We analyzed the measured data using a paired t test with a significance level of .05 to compare the differences between the without- and with-LTS conditions. Relative change in angle (RCA) with LTS was defined as the percent change of the Cobb angle relative to the Cobb angle in the without-LTS condition. The correlation of the Cobb angles and scoliotic indices was examined with the Pearson product moment correlation. SPSSd was used for all the statistical analyses.

Results 

The main clinical features of all 17 subjects before and after applying LTS are listed in table 2. The AP radiographic images of a subject (C5, class A) with line segments drawn for the calculation of the Cobb angles without and with LTS are given in figure 3. The means of the Cobb angles ± standard error (SE) of the subjects in the frontal plane without and with LTS were 39.57°±5.24° and 30.45°±4.64°, respectively. Seating with LTS gave significantly smaller Cobb angles than without LTS (P<.001), the mean difference ± SE being 9.12°±1.64°. An average RCA of 26.16%±4.23% was noted after applying LTS. The mean scoliotic indices of the subjects without and with LTS were .55±.09 and .46±.07, respectively. The improvement of the spinal curve with LTS was also indicated by a significant mean reduction of .09±.04 in the scoliotic indices (P=.027). The Cobb angles and scoliotic indices for all subjects correlated highly (r=.783, P<.001) (fig 4). There was a weak correlation between the severity of scoliosis (Cobb angles) and the corresponding RCA (r=−.347, P=.172).

Table 2. Cobb Angles, Lateral Thoracic and Lumbar Angles, and Corresponding RCA Values of All Subjects Without and With LTS

	Subject No.	Level of SCI	Cobb Angles			Lateral Thoracic Angles			Lateral Lumbar Angles
			Without LTS (deg)	With LTS (deg)	RCA⁎ (%)	Without LTS (deg)	With LTS (deg)	RCA⁎ (%)	Without LTS (deg)	With LTS (deg)	RCA⁎ (%)
	1	T9	65.83	62.06	5.74	48.12	42.07	12.57	11.43	3.60	68.50
	2	T3	21.59	12.87	40.38	19.91	35.26	−77.10	3.40	2.49	26.69
	3	T7	28.63	21.24	25.83	30.36	36.16	−19.10	13.71	8.72	36.41
	4	T5	52.95	30.75	41.93	50.86	52.82	−3.85	53.05	51.49	2.93
	5	T3	54.13	39.76	26.55	42.43	31.31	26.21	33.96	22.09	34.95
	6	T9	18.12	7.59	58.13	48.88	70.69	−44.62	72.30	66.82	7.58
	7	C6	20.99	14.36	31.59	55.23	35.58	35.58	35.13	33.03	5.99
	8	T10	14.37	5.48	61.90	21.10	17.58	16.68	26.42	16.32	38.21
	9	T9	63.52	57.35	9.72	14.87	1.47	90.11	72.35	62.43	13.71
	10	T11	83.77	59.14	29.40	30.09	39.91	−32.64	34.92	26.36	24.51
	11	C5	40.94	25.90	36.73	47.68	55.85	−17.14	14.56	15.82	−8.65
	12	T5	17.88	14.85	16.96	34.99	31.80	9.12	11.65	6.32	45.78
	13	T5	51.72	43.74	15.43	72.71	30.96	57.42	66.16	64.77	2.11
	14	T7	50.42	49.20	2.41	5.46	11.13	−103.85	19.26	15.80	17.99
	15	C6	15.39	13.18	14.41	38.22	31.13	18.55	26.16	21.08	19.42
	16	C5	53.90	43.03	20.16	37.40	6.68	82.14	3.91	0.99	74.80
	17	C6	18.58	17.20	7.45	46.83	56.90	∼21.50	3.16	3.12	1.27
	Mean ± SE		39.57±5.24	30.45±4.64	26.16±4.23	37.95±4.03	34.54±4.47	1.68±12.40	29.50±5.73	24.78±5.56	24.24±5.68
	Mean difference ± SE		9.12±1.64†			3.40±3.99			4.72±.93†

⁎ Relative change in angle (%) = (without LTS angle − with LTS angle)/without LTS angle × 100%. † P<.05 (significant difference between angles without and with LTS).

View Large Image Download to PowerPoint

Fig 3. Posterior photographs of a subject with a complete C5 lesion (A) without and (B) with LTS. (C) and (D) are the corresponding mirror-imaged AP radiographs on which lines are drawn for the calculation of the Cobb angles.

View Large Image Download to PowerPoint

Fig 4. Correlation of Cobb angles and scoliotic indices (Pearson r=.783, P<.001).

The mean of the lumbar angles without and with LTS were 29.50°±5.73° and 24.78°±5.56°, respectively. The lumbar angle was significantly decreased with LTS (P<.001), the mean reduction and mean RCA being 4.72°±0.93° and 24.24%±5.68%, respectively. However, there was no significant difference in the thoracic angles between the 2 seating configurations (without LTS: 37.95°±4.03°; with LTS: 34.54° ±4.47°; mean difference: 3.40°±3.99°; P=.41; mean RCA: 1.68%±12.4%), although a decreasing trend after using LTS was present. As often observed in clinical practice, a better posture alignment with LTS, including a more symmetric and erect trunk and head-neck posture in both the frontal and sagittal planes, could be seen from photographs of a subject (fig 5).

View Large Image Download to PowerPoint

Fig 5. Lateral photographs of a subject with a complete C5 lesion (A) without and (B) with LTS. (C) and (D) are the corresponding radiographs.

Discussion 

Special seating with LTS is among the methods found most acceptable by persons with SCI and clinicians for managing scoliosis after SCI.8, 21 However, as indicated in the introduction, no study has quantified the influence of LTS in a special seating system on the spinal alignment in persons with SCI. One reason for this is the difficulties in measuring or assessing the spine of persons with SCI, including the inability of photographic methods to measure skeletal alignment directly,4, 5, 27 and technical difficulties in taking spine radiographs of persons with SCI who have lost the ability to sit independently,16, 28 while they are sitting in special seating or a wheelchair. Holmes et al22 investigated the effects of lateral support pads on the correction rate for the scoliotic spinal curve in subjects with CP by using photography. Because only the spinous processes could be palpated and identified when using photography, they had to use the spinous process angle to represent the severity of the scoliosis. This may lead to errors in the measurements because as the degree of rotation increases, the spinous process migrates toward the convexity of the curve.3, 29 Most radiographs of the spine were taken in standing or supine positions,3 and few were taken in a sitting position.30 The experimental method developed in this study overcame these difficulties, thus enabling the first documentation of the effect of special seating with LTS on the changes of the frontal and sagittal scoliotic spinal alignment in persons with SCI.

The LTS in the special seating significantly improved the spinal alignment of the SCI persons with SCI in the frontal plane with reduced Cobb angles and scoliotic indices. Because the subjects had flexible scoliotic spines, the spinal curvature at the apex would be expected to increase owing to gravitational action, especially when the line of action of the gravitational force is further away from the center of the base of the spinal curve. Persistent bending as a result of this eccentric loading may accelerate the progression of the scoliosis.14, 25 LTS according to the 3-point force configuration aims to provide a moment that counterbalances the bending effect of the eccentric gravitational loading on the scoliotic spine, reducing the tension on the convex side and the compression on the concave side of the apex of the spinal curve. A similar biomechanic principle was also used in orthotic devices for the nonoperative treatment of idiopathic scoliosis.23 It is noted that the RCA in the frontal plane was not associated with the severity of scoliosis, because there existed only a weak correlation between the RCA and initial Cobb angle. This suggests that LTS in special seating is effective in improving spinal alignment in the frontal plane, regardless of the severity of the scoliosis. Apart from this immediate effect, the application of LTS to support the trunk and to counterbalance the gravitational bending moment may have a long-term effect on the progression of scoliosis. The radiographic technique would need to be applied longitudinally to a population of persons with SCI who had been treated with special seating interventions to explore the long-term effect of LTS on the progression of scoliosis.

LTS not only significantly improved the spinal alignment, but also reduced the lumbar angle in the sagittal plane, resulting in a more erect seating posture. This phenomenon has not been described in the literature before. A possible explanation is that LTS provides a derotational effect on the spinal column while applying a bending moment on the spine in the frontal plane.24 This is because a scoliotic spine is often coupled with rotational deformities in the other 2 planes, the amount of which depends on the constraints from both the bony and ligamentous structures. Axial rotational deformity can also cause the spine to appear deformed in the sagittal plane. A bending moment in the body’s frontal plane may also produce moment components in the spinal transverse and sagittal planes at the same time, which may help reduce the lumbar angle in the subject by improving the bony alignment and removing abnormal tension in the surrounding ligamentous structures.25 This effect was also found for thoracic curves, but with less statistical significance, possibly because the restraint from the rib cage that was further restricted by the specialized back support limited the derotational effects.31 Without LTS a subject’s trunk tended to deviate from the vertical because of the spinal deformity and loss of muscle control (see Fig 3, Fig 5). A compensatory movement of the head and neck to position the head close to an upright and center posture was noted. This is clearly shown by the photographic data. LTS helped maintain the upright and center position of the head by correcting and supporting the vertical alignment of the trunk, thus reducing the required effort for the head and neck muscles. It seems that LTS not only helped in improving the spinal alignment, but also enhanced the alignment of the head and trunk with less muscular effort.

The radiographic method combined with the adjustable seating system developed in the present study allows the AP and lateral radiographs of the thoracic and lumbosacral spine to be taken in a seated position. This technique can be very useful in monitoring changes of spinal alignment in patients who have to remain in a seated position in daily living, which cannot be achieved by traditional approaches such as photographs4, 5, 27 or radiographs taken in standing or supine positions.3 The method was used to demonstrate the immediate effects of LTS on the improvement of the spinal alignment in the frontal plane and reduction of the lumbar angles that could be obtained with a special seating system with LTS. However, because scoliosis is a complex deformity in 3 dimensions involving both bending and rotation of the vertebrae,25, 32 often together with rib deformation,31 further development of the present method to obtain transverse plane data from the bi-planar radiographs such as using model-based 2- and 3-dimensional registration techniques33 is needed. Apart from LTS, further research effort will be needed for the quantification of the long-term effects of other components of special seating, such as pelvic support, on the alignment of the spine.

Conclusions 

A radiographic method combined with an adjustable seating system was developed which can be used to measure the frontal and sagittal spinal alignment of a person in a seated position. The method was used to quantify the immediate effects of special seating with LTS on the changes of the frontal and sagittal spinal alignment in persons with SCI. The results showed that LTS significantly improved the spinal alignment in the frontal plane, regardless of the severity of scoliosis. LTS also reduced the lumbar angle in the sagittal plane, resulting in a more erect seating posture. Because scoliosis is a complex deformity in 3 dimensions, further investigation using a 3-dimensional measurement technique is necessary to provide additional data in the transverse plane. The quantification of the long-term effects of LTS as well as other components of special seating also requires further study.