Muscle Force and Gait Performance: Relationships After Spinal Cord Injury

General Procedures and Participants 

The study was part of the European Multicenter Study about Human Spinal Cord Injury (EMSCI) and was carried out in 7 European SCI rehabilitation centers over a period of 20 months. The protocol of the study was approved by the local ethics committee and all participants were informed and gave written consent.

In the 7 SCI rehabilitation centers, all subjects with traumatic or ischemic SCI who had been referred for primary rehabilitation were examined with the same clinical assessments according to a fixed timeframe: within the first week after the injury and subsequently after 1, 2, 6, and 12 months.16 The results of these assessments were stored anonymously in a central electronic database. For this retrospective study, we performed a query of this database in June 2005. The query included a key identifier, age, sex, level of lesion, date of the lesion, test dates, results of the Lower Extremity Motor Score (LEMS), Walking Index for Spinal Cord Injury (WISCI), and gait speed using the 10-meter walking test (10MWT) 1 and 6 months after injury. We excluded participants who were aged under 15 or over 70 or with an incomplete database record. In addition, subjects with SCI who exceeded 75% of a scale (for LEMS: 0.75×50=37; for WISCI, 0.75×20=15) 1 month after SCI were excluded. Chances for further improvement in such patients are small. Moreover, ceiling effects cannot be excluded. Such ceiling effects are likely to occur with both the WISCI and the LEMS but less with the self-selected walking speed using the 10MWT.

For the comparison between the improvement of ambulatory function and motor impairment, we chose a longitudinal approach with 2 time points. To pool patients with comparable potential to recover voluntary muscle force and ambulatory function, we categorized patients to 1 of 3 groups based on the assessment made 1 month after SCI: (1) group A (motor complete and nonwalking): patients with neither voluntary muscle contraction in the lower extremity nor walking function (LEMS, 0; WISCI, 0); (2) group B (motor incomplete and nonwalking): patients with some preserved voluntary muscle strength but no walking function (LEMS range, 1–37; WISCI, 0); and (3) group C (motor incomplete and standing or walking): patients with preserved voluntary muscle contraction and walking function (LEMS range, 1–37; WISCI range, 1–15).

In addition, we performed a cross-sectional analysis of patients who regained ambulatory function 6 months after SCI. The goal of this analysis was to compare the differences in LEMS, WISCI, and gait speed between patients with paraplegia and tetraplegia in 2 separate groups. Patients who achieved only limited walking function with the help of walking aids (WISCI <20) were categorized to group 1. All patients who regained unrestricted ambulatory function (WISCI, 20) were categorized in group 2.

Outcome Measures 

Examinations were performed 1, 3, 6, and 12 months after the injury by experienced physicians (ASIA assessment) and physical therapists (assessment of walking). For this analysis, results of the 1-month and 6-month examinations were used. To standardize the testing procedure, regular workshop meetings and additional training were organized for the examiners.

The voluntary muscle strength of 5 key muscles (hip flexors, knee extensors, ankle dorsiflexors, toe extensors, ankle plantarflexors) of both lower extremities (LEMS) was tested in accordance with the standard neurologic assessment developed by ASIA.5 Each muscle was given a value between 0 and 5 according to the strength of voluntary muscle contraction. Maximum and minimum LEMS were 50 and 0, respectively.

To assess walking performance, we used the revised version of the Walking Index for Spinal Cord Injury (WISCI II)17, 18 and the 10MWT.19 The WISCI II describes the walking status of a patient based on the requirements of assistance and/or bracing and/or walking aids. The ordinal scale ranges from 0 (neither standing nor walking function) to 20 (independent walking). The 10MWT reflects the time necessary for walking 10m. Subjects were instructed to walk in a straight line at a comfortable self-selected pace over the distance of 14m. Walking time was taken after the subject walked 2m and was stopped 2m before the finish line to account for potential acceleration and deceleration effects. Subjects could use their preferred assistive device, including minimal physical assistance. Gait speed (in m/s) was calculated from the results of the 10MWT.

Data Analysis 

Given the fact that LEMS and WISCI are ordinal-scaled variables, we applied nonparametric tests and medians are given. Gait speed is presented as mean ± standard deviation (SD). SPSSa for Windows was used for the statistical analysis.

Course of Motor Scores and Walking Function 

In group A (motor complete and nonwalking 1 month after the injury, n=117), 29 (24.8%) patients improved their LEMS from 0 to a median of 10. The most frequent improvement was from 0 to 1. Eighty-eight (75.2%) patients remained at a 0 value, which means complete paralysis. Nine (7.7%) patients improved in the WISCI from 0 to a median of 8, and in the 10MWT to an average gait speed of .14±.10m/s. In 108 (92.3%) patients, walking function (ie, WISCI, gait speed) did not change. Although the changes in LEMS and WISCI had no clinical impact, the Wilcoxon signed-rank test revealed a significant improvement of the total group (for LEMS, P<.001; for WISCI, P<.01).

In group B (motor incomplete and nonwalking, n=46), LEMS improved in 43 (93.5%) patients, from a median of 14 to 28, that is, a relative improvement of 24%±18%. However, in 3 (6.5%) patients, LEMS decreased.

WISCI improved in 39 (84.8%) patients, from 0 to a median of 10, that is, relative improvement of 52%±33%. In 7 (15.2%) patients, WISCI did not change.

Gait speed improved in 35 (76.1%) patients, from 0 to an average value of .41±.45m/s, that is, relative improvement of 28%±31%. No change in gait speed was observed in 11 (23.9%) patients.

For the total group, the overall improvement in LEMS, WISCI, and gait speed was significant (P<.001). WISCI improved to a larger extent compared with LEMS (P<.001). No difference was evident between the relative improvements of LEMS and gait speed (P=.54).

In group C (motor incomplete and standing or walking, n=15), all patients improved in LEMS, WISCI, and gait speed. The median LEMS 1 month after the injury was 29 and improved to 41 at the sixth month after the SCI, which corresponds to a relative improvement of 23%±12%. The corresponding values for WISCI were 8 one month postinjury and 16 five months later. This reflects a relative improvement of 46%±20%. Gait speed improved from .36±.29m/s to .88±.44m/s, that is, a relative improvement of 36%±23%. The overall improvements of LEMS, WISCI, and gait speed was significant (P=.001). WISCI (P<.01) and gait speed (P<.05) showed a greater relative improvement than LEMS. See also table 2 and figure 1.

Table 2. LEMS, WISCI, and Gait Speed at the First and Sixth Months After SCI of Patients in Groups A, B, and C

		LEMS			WISCI			Gait Speed
	Group	1 Month	6 Months	RI (%)	1 Month	6 Months	RI (%)	1 Month	6 Months	RI (%)
	A	0	0	6±12	0	0	3±11	0	.01±.05	0.7±3
	B	14	28	24±18	0	10	52±33	0	.41±.45	28±31
	C	29	41	23±12	8	16	46±20	.36±.29	.88±.44	36±23

NOTE. Values are median and mean ± SD. Abbreviation: RI, relative improvement.

View Large Image Download to PowerPoint

Fig 1. Relative improvements of LEMS, WISCI, and gait speed (difference between measurement at 6 months and at 1 month divided by the respective normative value). (A) Motor incomplete patients who were not able to stand or walk 1 month after SCI; and (B) patients who were motor incomplete and able to stand or walk 1 month after SCI. *P<.05; †P<.01; ‡P<.001.

Outcome of Subjects With Tetraplegia Versus Paraplegia 

Subjects with tetraplegia in group 1 who achieved only limited walking function 6 months after the injury (WISCI <20, n=16) showed a median LEMS of 31.5, a median WISCI of 8, and a mean walking speed of 0.4±0.3m/s. By comparison, the corresponding values of patients with paraplegia (n=33) were 23 for the LEMS (P<.01), 12 for the WISCI (P=.6), and 0.4±0.3m/s for the walking speed (P=.68).

Subjects with tetraplegia in group 2 (n=24), who could walk without restrictions (WISCI, 20), showed a median LEMS of 48, a median WISCI of 20, and a mean walking speed of 1.4±0.4m/s. Patients with paraplegia (n=13) showed a median LEMS of 45 (P<.05), a median WISCI of 20 (P=1.0), and a mean walking speed of 1.4±0.3m/s (P=.89). See also table 3 and figure 2.

Table 3. Comparison of LEMS, WISCI, and Gait Speed Between Patients With Tetraplegia and Paraplegia in 2 Groups

		Group 1		Group 2
	Test	Tetraplegia (n=16)	Paraplegia (n=33)	Tetraplegia (n=24)	Paraplegia (n=130)
	LEMS	31.5	23	48	45
	WISCI	8	12	20	20
	Gait speed	0.4±0.3	0.4±0.3	1.4±0.4	1.4±0.3

NOTE. Values are median and mean ± SD. Group 1 is patients with limited walking function (WISCI <20); group 2 is patients with unrestricted walking function (WISCI, 20).

View Large Image Download to PowerPoint

Fig 2. Comparison of LEMS, WISCI, and gait speed between patients with tetraplegia (Tetra) and paraplegia (Para) who achieved (A) limited and (B) unrestricted walking function 6 months after the injury. *P<.05; †P<.01.

Course of LEMS and Locomotor Function 

As has been shown in this study, an increase in motor score is not always associated with an improvement in function. Conversely, an improvement in locomotor function between 1 and 6 months after SCI in patients with both incomplete tetraplegia and incomplete paraplegia is associated with a variable increase in LEMS. Along the same line, Wirz et al21 showed that locomotor training by a robotic device (Lokomat) in patients with chronic SCI resulted in a significant improvement of locomotor ability, which was not associated with an increase in LEMS. This might be due to an extra-activation of lower-extremity muscles by the automatic proprioceptive feedback during locomotion.22 In addition, locomotor training affects many muscles and activates them in a functional way, resulting in a better coordination of synergistic muscles.22, 23 This might lead to a better functional result than any other physical therapy because training effects were shown to be specific to the focus of the therapy.24

Another study,25 which focused on the efficacy of the application of high doses of methylprednisolone, has shown a small increment in ASIA score compared with controls. However, in this study it remained open whether the observed improvements on the level of the body functions were associated with an improved activity, ambulatory function. Recently, this treatment has been questioned (eg, Hurlbert26). According to our observations, the effect of this treatment on outcome, that is, a small increase of ASIA score, might not reflect improvement in function in all patients.

Difference Between Outcome in Subjects With Paraplegia and Tetraplegia 

Patients with paraplegia and tetraplegia who became ambulatory 6 months after SCI demonstrated a similar locomotor function according to WISCI and walking speed. However, in patients with tetraplegia, all lower-extremity “key” muscles had to reach a functional level of muscle contraction to permit locomotor function. In contrast, patients with paraplegia achieved this performance already at lower LEMS. This became evident in patients with limited walking function (WISCI <20). This difference is suggested to be due to the additional weakness of trunk and upper-limb muscles in patients with tetraplegia, which requires substantially more lower-extremity muscle force to compensate for the postural instability. Once this stability had been provided, a further increase in LEMS was associated with a steep improvement in locomotor ability. Consequently, ambulatory patients with tetraplegia achieved a higher level of walking function than patients with paraplegia. In contrast, if the LEMS remains below a certain level, only patients with paraplegia become ambulatory while patients with tetraplegia remain wheelchair-bound.

Methodologic Considerations 

This study has some limitations. First, incomplete data records are inherent to a multicenter database, reflected by a high exclusion rate. Second, on the one hand, the multicenter nature of this study enables access to a substantial number and broad spectrum of subjects. However, reliability of the procedures is hard to ensure due to multiple examiners. This applies in particular to the clinical, nonapparative assessments. To keep random and systematic error as small as possible, regular workshop meetings were organized and datasets were tested for plausibility by an independent person. The assessment of reliability was not a focus of this study and was therefore not further explored. Third, for the assessment of voluntary muscle force, a limitation and source of bias might be the use of clinical motor scores rather than objective measures (eg, dynamometry). Fourth, despite the straightforward aim of this study (comparing muscle force and functional walking scores), no unambiguous statistical method is available to directly compare the different variables.