Evaluation of the Wisconsin Gait Scale in Patients With Multiple Sclerosis and Spastic Hemiplegia

Introduction

Multiple sclerosis (MS) is the most prevalent neurological condition related to the central nervous system (CNS), leading not only to motor dysfunction but also to cognitive impairment in young and middle-aged adults [1]. With the highest incidence between the ages of 20 and 50 years, MS is a major problem for both the national healthcare systems and society, since it is among the leading causes of neurological disability in young people [2]. The prevalence of MS varies in different parts of the world, and it is estimated that in 2020 there were approximately 2.8 million individuals with MS worldwide [2].

Evidence in the literature shows a high prevalence of gait dysfunction in MS and its impact on the functional status and quality of life of patients and their families [3]. Mobility impairment in MS varies depending on the disease subtype but is reflected in spatiotemporal and kinematic gait characteristics. Mobility impairment is most commonly associated with decreased gait velocity, walking endurance, cadence, step length, and single support time, along with increased step width and limited joint range of motion, including reduced knee flexion during the swing phase and reduced ankle dorsiflexion at initial contact. Moreover, walking in individuals with MS is associated with increased energy expenditure and is therefore inefficient and metabolically uneconomical [3]. All these characteristics are also observed in the hemiparetic gait of individuals with stroke, which is also linked to CNS damage, like in MS [3]. Researchers highlight the importance of comprehensive gait assessment in individuals with MS [4], therefore, it should also involve comprehensive gait analysis, including spatiotemporal and kinematic gait parameters. Advanced instrumented 3-dimensional methods enable detailed and accurate analysis of the aforementioned characteristics [5]. These, however, are highly demanding in terms of cost, know-how, equipment, and time [6]. On the other hand, there are few tools for cost-effective and quick observational assessment of walking ability in patients with MS. The validated clinical outcome measures applied in evaluation of gait in MS most importantly include the Timed 25-Foot Walk [7], the 6-Minute and 2-Minute Walk Test [8,9], and the 12-Item Multiple Sclerosis Walking Scale (MSWS-12) [10]. However, none of these tools allow for a comprehensive observational, descriptive assessment of both spatiotemporal and kinematic gait parameters.

The Wisconsin Gait Scale (WGS) is an observational tool with excellent psychometric properties that is effectively used to evaluate hemiparetic gait following stroke. It allows for descriptive, multifactorial, and detailed assessment of spatiotemporal parameters, including stance time, step length, and stance width, as well as kinematic parameters of the hip, knee, ankle, and pelvis across the sagittal, transverse, and frontal planes [11–15]. The WGS has been shown to have excellent intra-rater and inter-rater reliability in the case of individuals with hemiparesis due to stroke [11–14]. Furthermore, the pediatric version of the tool designed for conducting qualitative observational gait assessment of children with spastic hemiplegic cerebral palsy has also been shown to have very high inter- and intra-observer reliability [15].

Earlier research conducted by our team focused on the applicability of the WGS in assessment of walking abilities in patients with stroke [11,12]. One of the studies (2020) was designed to compare the WGS scores with the symmetry indexes acquired using a 3-dimensional gait analysis. The findings showed that WGS scores consistently corresponded to gait pattern characteristics identified using advanced biomechanical tools [12]. Another study (2022) compared results acquired using an application software designed to compute the scores on the WGS with the results of a 3-dimensional gait analysis, and the findings showed high agreement between the 2 measures and consequently confirmed the usefulness of the former tool for the clinical practice [11]. One more study (2023) assessed inter- and intra-rater reliability of the computerized version of the pediatric WGS used in children with spastic hemiplegic cerebral palsy, confirming the high repeatability of the tool [15]. All of the above studies involved populations of patients with either post-stroke hemiparesis [11,12] or cerebral palsy [15], that is, individuals with unilateral damage to the CNS leading to a typical pattern of hemiparetic gait. Further, these studies were designed to assess the agreement between this simple observational tool and the time-consuming and costly 3-dimensional analysis [11,12]. The present study differs from the earlier research since it is the first attempt to assess the feasibility and reliability of the WGS in a population of patients with spastic hemiplegia due to MS. In fact, hemiplegia is observed less frequently in patients with MS, and the neurological changes are often more complex and scattered in nature. Nevertheless, some patients with MS present this type of gait pattern. Therefore, in this study, we aimed to determine whether the WGS can also be effectively used in this population as a tool enabling a comprehensive, descriptive assessment of gait impairment, taking into account spatiotemporal and kinematic parameters.

Notably, individuals with MS most commonly present with bilateral weakness (paraparesis) and paraplegia, resulting from spinal cord lesions. Pure motor hemiplegia does not typically occur in MS and even if one side of the body is visibly weaker, the other side is also affected by the condition, albeit to a lesser degree [16]. Importantly, studies involving patients with hemiparetic gait patterns after stroke have also reported altered kinematics and temporal characteristics of gait, as well as compensatory strategies on the unaffected side in these patients, even though in this case the damage is linked to the brain rather than the spinal cord. Consequently, the sides of the body in patients after stroke are described as indirectly involved (or unaffected) and directly involved (affected) [17–19]. Moreover, it appears that gait patterns after stroke and in the hemiplegic subtype of MS are similar to a degree, which may be explained by the fact that in both cases the damage occurs in certain CNS structures, namely the pyramidal tracts [20]. In view of the above, we decided to apply the WGS, originally designed to assess hemiparetic gait patterns after stroke and enable observational assessment of gait pattern asymmetry [11–15], and to investigate the feasibility of the tool in MS patients with hemiplegia. The objective of the study design was to provide an additional tool for clinical practice, enabling assessment of gait in individuals with spastic hemiplegia in MS, which, unlike the commonly applied tools, can be used to perform a comprehensive observational evaluation of both the spatiotemporal and kinematic parameters of gait.

In this study, we also aimed to assess the intra-rater and inter-rater reliability, as well as the concurrent criterion validity, of the WGS in the evaluation of gait in individuals with MS-related hemiplegia. Finally, we aimed to investigate the feasibility of using the WGS in daily clinical practice with hemiplegic patients with MS.

Material and Methods

PARTICIPANTS:

The study included 30 individuals with the hemiplegic type of MS (13 men and 17 women, with mean age of 43.7±5.2 years, and mean Expanded Disability Status Scale [EDSS] of 4.95±0.55). The group characteristics are presented in Table 1. The trials were conducted in the biomechanics laboratory at the University of Rzeszów, Poland.

All patients were assessed by a physiotherapist with many years of experience in the diagnosis and rehabilitation of individuals with hemiparetic gait patterns. Only patients with an asymmetrical gait pattern resulting from unilateral muscle weakness and spasticity, typical of a hemiparetic pattern, were enrolled in the study. Hemiplegic gait pattern was identified based on medical records (diagnosis of MS with spastic hemiplegia), physical examination (assessment of muscle tone, muscle strength, and gait pattern), and clinical observation of the gait pattern, which identified clear asymmetry in the lower limbs, including step length, gait phase duration, and range of motion in the joints, and in pelvic alignment.

The eligibility criteria for the study group included clinically confirmed spastic hemiplegia associated with MS, an EDSS score between 4 and 6, a hemiplegic gait pattern, age between 30 and 60 years, and the ability to walk without assistance from another person (walking aids were permitted). Participants were also required to have a stable medical condition, no cognitive impairment affecting their ability to follow instructions, no orthopedic disorders of the lower limbs, and no other neurological conditions (eg, ataxia, paraparesis, or diffuse symptoms) that could produce a more symmetrical or ataxic gait. The experimental protocol was approved by the local Bioethics Commission at the University of Rzeszów (approval no. 2017/12/13), and the research project was registered with the Australian New Zealand Clinical Trials Registry (ACTRN12618000494235). The study design complied with the Declaration of Helsinki. All the study participants provided written informed consent.

ASSESSMENT OF GAIT USING THE WGS:

During the trial, participants walked a distance of 10 m, and 2 synchronized digital cameras were used to record their gait. The cameras were arranged to make the recording in both the frontal and sagittal planes. For this purpose, one camera was aligned with the direction of the gait in the frontal plane, whereas the second camera, arranged to capture the image in the sagittal plane, was placed at a distance of 2 m from and halfway up the walking path. The cameras were set up to record 3 trials to evaluate the affected side and 3 trials focusing on the unaffected side, for a total of 6 trials. The participants were instructed to walk the distance at a comfortable, self-selected pace, with the support of the orthopedic aids normally used, as needed.

The WGS used in the study consists of 14 submeasures related to the specific aspects of the gait pattern during stance, toe off, swing, and heel strike phases, considering spatiotemporal and kinematic parameters. All these items are rated on a 3-point scale, except for 2 submeasures that are rated on a 5-point and a 4-point scale, respectively. The total score ranges from 13.35 to 42 points, with lower scores corresponding to better quality of gait pattern [11].

Three raters, working separately, interpreted the recordings and evaluated the participants’ gait using the WGS [11]. All raters had expert knowledge of gait disorders associated with the hemiplegic type of MS, and they all had been trained and were experienced in using and interpreting the WGS. Each rater entered the results for the gait parameters assessed on the WGS, assigning scores in each of the 14 WGS categories in an Excel table. Each video recording was assessed twice by each of the 3 raters at 2-week intervals. Then, statistical analyses of the results were conducted, and the scores acquired during the 2 examinations and by the 3 raters were compared.

INTRA-RATER AND INTER-RATER RELIABILITY:

To verify the reliability of measurements performed using the WGS, the assessment was conducted for 2 key aspects: agreement between different examiners (inter-rater reliability) and consistency of assessments by a single examiner over time (intra-rater reliability). The inter-rater reliability assessment was based on a comparison of the scores obtained by 3 independent examiners who individually reviewed the same set of video recordings. This way it was possible to verify the objectivity of the scale and its resistance to subjective interpretation. In turn, to enable intra-rater reliability assessment, each examiner evaluated the same recordings twice, 2 weeks apart (measurement 1 and measurement 2). Subsequently, results of these 2 evaluations were compared to determine the stability of the measurement over time and the consistency of a single examiner, while minimizing the potential impact of the memory effect.

GAIT ASSESSMENT WITH THE 10-METER WALK TEST, 2-MINUTE WALK TEST, AND TIMED UP AND GO TEST: CONCURRENT CRITERION VALIDITY: To determine concurrent criterion validity, a comparative analysis was conducted, taking into account the mean scores on the WGS and the results of the established clinical gait assessment tools: the 10-Meter Walk Test, the 2-Minute Walk Test, and the Timed Up and Go Test, which have been validated and are commonly recognized and widely used to assess gait and functional mobility in MS populations [21–23].

During the 10-Meter Walk Test, intended to assess the gait velocity (m/s), the participants were instructed to walk a distance of 10 m at a comfortable speed. Further analyses used mean scores acquired during 2 trials. During the 2-Minute Walk Test, walking distance was measured (in meters), and the score was an indicator of walking efficiency. The participants were instructed to walk for 2 minutes, between 2 points situated 30 m apart, with a self-selected speed. During the trials, the participants could use their own orthopedic aids. The Timed Up and Go Test was used to measure the time (in seconds) needed by the participants, receiving no assistance, to get up from a chair (with no armrests), walk a distance of 3 m, come back, and sit down. The test assessed independent mobility and dynamic balance during the activity of walking [21–23]. All the tests were preceded by brief instruction and demonstration trials. Participants were allowed to rest between the tests to avoid fatigue that could potential affect the results.

DATA ANALYSES:

All the statistical analyses were computed using R software (version 4.4.2), with a significance level (α) defined as 0.05. The analyses aimed to determine the reliability of measurements performed using the WGS and assessed 2 aspects: agreement between the scores acquired by 3 examiners (inter-rater reliability), and consistency of the scores reported by each examiner at different times (intra-rater reliability). The main measure applied to quantify both types of reliability was the intraclass correlation coefficient (ICC) type 2, as classified by Shrout and Fleiss, which is appropriate for the research model used. Additionally, for intra-rater reliability, the Bland-Altman method was used to assess the agreement between measurement 1 and measurement 2 for each examiner.

Concurrent criterion validity was determined by comparing the WGS with established measures described above. The analyses investigated the correlations between the mean WGS score and the results acquired using all the clinical tests, which have been validated and are recognized and commonly used to assess gait and functional mobility in MS populations [21–23]. The mean of the scores reported by all 3 raters was calculated to acquire a single, more reliable and consistent measure for each patient. This approach minimized the effect of possible errors or subjectivity of a single rater and provided a score which more closely reflected the “true” rating of the patient’s performance. A normal distribution, assessed using the Shapiro-Wilk test, was found for all 3 tests and for the mean WGS score; therefore, the Pearson correlation coefficient (r) was applied.

The dedicated PLUS Module of Statistica 13.3 software was used to compute the minimum sample size for the relevant population, with the factors of the fraction size 0.3, a maximum error of 5%, and a 95% confidence level. Ultimately, a sample size of 28 participants was determined, and 30 participants were enrolled for the study.

The sample size calculation was based on estimating the proportion of individuals in the population who met the specified inclusion criterion. In determining the minimum sample size, we considered the annual number of patients with MS hospitalized in the rehabilitation ward, of whom approximately 60% had an EDSS score of 4 to 6. Based on the data from our hospital, we estimated that 30% of these patients presented with spastic hemiplegia associated with MS. In the absence of definitive epidemiological data, this 30% estimate was derived from our clinical observations within this specific patient population.

We also estimated the statistical power. To justify the sample size, the minimal clinically important difference was used as a reference point, as it allows the sample size to be linked to a meaningful, patient-centered improvement – an essential consideration in the design of clinical research. The published study entitled “The Wisconsin Gait Scale – The minimal clinically important difference” [24] has provided the data for estimating the effect size for the WGS by showing a very strong correlation (R2=0.63) between the change in the gait rating and the improvement in overall performance. The effect corresponding to the correlation coefficient r=0.79 provided the basis for the analysis of power. To identify such high association with a power of 80% and significance level alpha=0.05, a sample of only 10 participants was required. In view of this, the sample of n=30 was considered to be fully sufficient to ensure very high statistical power.

Results

TOTAL SCORE OF THE WGS:

The total WGS score was calculated for each study participant twice by each of the 3 raters. Table 2 includes descriptive statistics for the scores on the WGS reported by the 3 raters in each series of measurements.

INTRA-RATER RELIABILITY:

Comparison of the results in measurement 1 and measurement 2 performed by the same rater showed excellent agreement in each case, reflected by an ICC of 0.991 for rater 1, ICC of 0.991 for rater 2, and ICC of 0.994 for rater 3 (Table 3). Bland-Altman analysis showed 95% limits of agreement ranging from −1.37 to 1.10 for rater 1, −1.33 to 0.99 for rater 2, and −1.07 to 0.74 for rater 3. Bland-Altman plots comparing scores from measurement 1 and measurement 2 for each rater are presented in Figure 1.

INTER-RATER RELIABILITY:

A comparative analysis of the results of the assessments performed by the 3 raters showed excellent agreement in measurement 1 and measurement 2, with ICC values of 0.988 and 0.978, respectively (Table 4).

CONCURRENT CRITERION VALIDITY:

The analyses showed very strong and significant correlations (P<0.001) between the mean scores in the WGS and in all the clinical tests applied in this study to assess gait. There was a strong positive correlation with the score in the Timed Up and Go Test (r=0.91). The findings also showed strong negative correlations with gait speed in the 10-Meter Walk Test (r=−0.94) and with distance covered in the 2-Minute Walk Test (r=−0.94), as shown in Table 5.

Discussion

LIMITATIONS:

The present study has several limitations. First, the sample included only 30 individuals with the spastic hemiplegia type MS. This suggests that further research involving a larger population is warranted; however, it should be noted that the minimum sample size calculation for this study indicated that a sample of 30 participants was adequately powered for a feasibility study. Second, our study focused exclusively on individuals with spastic hemiplegic MS, which inherently limits the generalizability of our findings to the broader MS population. However, this was a deliberate choice, as the WGS is specifically designed to assess hemiparetic gait pattern after stroke. Because gait pattern in spastic hemiplegia type MS is similar to hemiparetic gait after stroke, we decided to investigate the feasibility of the WGS in assessing the gait of individuals with this specific subtype of MS. Therefore, the eligibility criteria were defined to ensure a homogenous sample, meaning that individuals with other MS subtypes, namely those presenting with cerebellar ataxia, bilateral involvement, or more scattered symptoms, were not enrolled. The WGS relies on evaluating asymmetry by comparing gait parameters between the affected and unaffected sides, making it unsuitable for individuals with other MS subtypes, including ataxia or bilateral involvement, in which gait deficits are of a different nature, being symmetrical or largely linked to impaired coordination rather than spasticity. Given that spastic hemiplegia involves asymmetric motor impairment and often presents with specific patterns of muscle tone, specific gait disturbance, and functional compensation, the results of this study are most applicable to individuals with the spastic hemiplegic subtype of MS. Nevertheless, it should also be pointed out that although bilateral weakness (paraparesis) and paraplegia resulting from spinal cord lesions are most commonly observed, and cases of spastic hemiplegic MS are not typical, they can provide valuable diagnostic information, particularly in regard to changes located in the brain. Future studies should focus on modifying the WGS or developing new gait assessment tools that are valid for a more diverse representation of MS subtypes. Expanding such tools would enhance the inclusivity of research in this area and support the development of tailored interventions for a broader MS population. It is also necessary to continue the research assessing the construct validity and sensitivity of the WGS to changes resulting from gait rehabilitation in individuals with spastic hemiplegia associated with MS.

Conclusions

The findings present excellent intra-rater and inter-rater reliability and confirm the concurrent criterion validity of the WGS in individuals with the hemiplegia type of MS. The study shows that the WGS, designed to assess hemiparetic gait after stroke, can effectively be applied as a tool for observational evaluation of walking abilities in patients with spastic hemiplegia due to MS. Consequently, owing to the above evidence, clinical practice can gain an additional tool for the assessment of gait in individuals with hemiplegic spasticity in MS, which unlike the commonly applied scales and tests, can be used to perform a comprehensive observational evaluation of both the spatiotemporal and kinematic parameters of gait.