- Email: [email protected]
Gait variability and disability in multiple sclerosis
Gait & Posture 38 (2013) 51–55
Contents lists available at SciVerse ScienceDirect
Gait & Posture journal homepage: www.elsevier.com/locate/gaitpost
Gait variability and disability in multiple sclerosis Michael J. Socie a, Robert W. Motl b, John H. Pula c, Brian M. Sandroff b, Jacob J. Sosnoff b,* a
Department of Mechanical Science and Engineering, University of Illinois, Urbana, IL, United States Department of Kinesiology and Community Health, University of Illinois, Urbana, IL, United States c Illinois Neurological Institute, University of Illinois College of Medicine, Peoria, IL, United States b
A R T I C L E I N F O
A B S T R A C T
Article history: Received 19 June 2012 Received in revised form 4 October 2012 Accepted 20 October 2012
Gait variability is clinically relevant in some populations, but there is limited documentation of gait variability in persons with multiple sclerosis (MS). This investigation examined average and variability of spatiotemporal gait parameters in persons with MS and healthy controls and subsequent associations with disability status. 88 individuals with MS (age 52.4 11.1) and 20 healthy controls (age 50.9 8.7) performed two self-paced walking trials on a 7.9-m electronic walkway to determine gait parameters. Disability was indexed by the Expanded Disability Status Scale (EDSS) and ranged between 2.5 and 6.5. Gait variability was indexed by standard deviation (SD) and coefficient of variation (CV = SD/mean) of step time, step length, and step width. Average gait parameters were significantly correlated with EDSS (r = 0.756–0.609) and were significantly different in individuals with MS compared to controls (p 0.002). Also, step length (p < 0.001) and step time (p < 0.001) variability were both significantly greater in MS compared to controls. EDSS was positively correlated with step length variability and individuals with MS who used assistive devices to walk had significantly greater step length variability than those who walked independently (p’s < .05). EDSS was correlated with step time and length variability even when age was taken into account. Additionally, Fisher’s z test of partial correlations revealed that average gait parameters were more closely related to disability status than gait variability in individuals with MS. This suggests that focusing on average gait parameters may be more important than variability in therapeutic interventions in MS. ß 2012 Elsevier B.V. All rights reserved.
Keywords: Walking impairment EDSS Motor function Spatiotemporal gait parameters
1. Introduction Multiple sclerosis (MS) is a neurologic disease that affects an estimated 400,000 American adults and 2.5 million adults worldwide. MS results in demyelination and axonal loss in the central nervous system (CNS). Such CNS damage may result in physical disability (potentially including weakness, sensory loss, and/or ataxia) and gait impairment [1,2]. Disability in MS can be indexed by the Expanded Disability Status Scale (EDSS) [3,4], which is influenced by ambulatory status. EDSS scores range from 0 to 10, where 0 represents no impairment, 4 represents the onset of significant walking impairment, 6 represents onset of assistive device use during ambulation, and 10 represents death due to MS [4]. In addition to the EDSS, which is a clinical standard, disability level in MS has been characterized by use of assistive devices
* Corresponding author at: Department of Kinesiology and Community Health, University of Illinois at Urbana-Champaign, 301 Freer Hall, 906 S. Goodwin Ave., Urbana, IL 61801, United States. Tel.: +1 217 333 9472, fax: +1 217 244 7322. E-mail addresses: [email protected] (M.J. Socie), [email protected] (R.W. Motl), [email protected] (J.H. Pula), [email protected] (B.M. Sandroff), [email protected] (J.J. Sosnoff). 0966-6362/$ – see front matter ß 2012 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.gaitpost.2012.10.012
during ambulation [5]. Greater than 50% of individuals with MS require assistive devices within 15 years of disease onset [1]. Gait impairment is one of the most frequent consequences of MS, and walking dysfunction is considered by the majority of patients as the most challenging, life-altering aspect of the disease [6]. Persons with MS walk slower, take shorter, wider, and slower steps, and spend a greater percent of the gait cycle in doublesupport compared with healthy controls, even early in the disease course [7–9]. Such gait impairment have also been directly associated with disability status as a marker of disease progression [7,10]. Gait variability (i.e. fluctuations in gait parameters between steps) is predictive of mobility impairment and falls in older adults and other neurological populations [11–14]. Movement variability is further associated with health in biological systems and is a marker of motor control function [15,16]. Gait variability might be more sensitive to dysfunction in MS than traditional parameters such as walking velocity [17]. However, we are unaware of research testing that possibility and this is important in forming therapeutic interventions for MS. There is some evidence that persons with MS have elevated gait variability compared to controls. For example, one study reported that persons with MS exhibit greater kinematic variability at the
52
M.J. Socie et al. / Gait & Posture 38 (2013) 51–55
hip, knee, and ankle during ambulation than healthy controls [18]. Other studies have reported that persons with MS who had mild impairment demonstrated greater variability of step time and single-support time [8] and step length [17] than controls. Data from a small sample (N = 10) indicates that gait variability increases with disability in MS [19]. That study reported that only persons with MS who had walking impairment (EDSS > 4.0) had greater variability of stride length compared to healthy controls, but this could be driven by low statistical power associated with the small sample and should be replicated with a larger sample. Consequently, further documentation of gait variability as a function of disability is a necessary step toward investigating the clinical relevance of gait variability in MS. This investigation examined average and variability of spatiotemporal gait parameters in persons with MS with a wide range of disability levels and healthy controls as well as associations between average gait parameters, gait variability and disability in persons with MS. We hypothesized that individuals with MS would exhibit greater gait variability than controls. Second, we hypothesized that gait variability would be greater in persons with MS with greater disability. Third, we hypothesized that associations between gait variability and disability would be stronger than associations between average gait parameters and disability in MS.
between individuals with MS who used assistive devices to walk and those who walked independently were determined using a mixed-model ANCOVA while controlling for age and gender. 2.3.3. Average vs. variability parameters A Fisher’s z-test was used to determine differences between the strength of correlations between average gait parameters and EDSS compared to gait variability and EDSS.
3. Results 3.1. Participant characteristics Eighty-eight participants with MS and 20 healthy controls participated in this investigation (Table 1). The MS group had an average age of 52.4 years, average duration of MS of 11.8 years, and median EDSS score of 4.5. 83% of the MS group was female, 32% used unilateral assistive devices (e.g. canes), and 6% used bilateral assistive devices (e.g. walkers). The control group had an average age of 50.9 years and 80% of the group was female. There were no differences between groups in demographic characteristics (p > 0.05). 3.2. MS vs. controls
2. Methods The procedures for this investigation were approved by a University Institutional Review Board and all participants provided written informed consent prior to data collection. 2.1. Participants Persons with and without MS participated in this investigation. Inclusion criteria for participants with MS required a neurologist-confirmed diagnosis, the ability to walk without or with an assistive device (e.g. a cane or walker) and be relapse-free for at least 30 days prior to testing. Inclusion criteria for controls required no gait impairment, no assistive device use, and no medical condition causing significant morbidity such as neurological or cardiovascular disease. 2.2. Procedures Upon arrival at the testing facility, participants with MS underwent an examination by a neurologist to generate an EDSS score [4]. To determine spatiotemporal gait parameters and gait variability, all participants performed two walking trials across a 7.9-m electronic walkway (GAITRiteTM, CIR Systems Inc., Haverton, PA, USA) at self-selected, comfortable speed. Participants began walking 1.5 m in front of he GAITRiteTM and continued to walk 1.5 m beyond the mat to insure that steady state gait was measured. On average the MS group took 13.3 steps, whereas the control group took 10.7 steps per trial. Use of assistive devices was permitted during testing. Average walking velocity, step length, step time, and step width were recorded. Variability of step length, step width, and step time was indexed by the standard deviation (SD) as well as the coefficient of variation (CV = SD/mean). The SD is operationalized as an absolute measure of variability, whereas CV is operationalized as a relative measure of variability. Gait parameters were calculated individually for each pass over the walkway; those values were then averaged across trials to produce final values. 2.3. Statistical analysis Statistical analyses were performed using SPSS software version 19.0 (SPSS Inc., Chicago, IL, USA). Statistical significance was assumed for p < 0.05. Group differences in age and gender distribution were determined using an independent sample t-test and chi-square test, respectively. 2.3.1. MS vs. controls Differences in average and variability of gait parameters between individuals with MS and controls were determined using independent sample t-test. The magnitudes of group differences were indexed by Cohen’s d effect sizes [20]. 2.3.2. Gait variability and disability Associations between gait variability and disability level were examined in the context of EDSS as well as assistive device use. Pearson’s and Spearman’s correlations were performed for average gait parameters, gait variability, age, and EDSS in the MS group. Partial correlations were also performed for gait parameters and EDSS while controlling for age and gender. Differences in gait parameters
The MS group demonstrated lower walking velocity and step length and greater step time and step width than the control group (Table 2). Step length and step time SD and CV were significantly greater in MS vs. controls (p < 0.05). The effect sizes of these differences were moderate to large in magnitude (d’s ranged from 0.4 to 1.4). The MS group also had a greater step width SD, but not CV, compared to the control group. 3.3. Gait variability and disability EDSS was significantly correlated with all average gait parameters, gait variability parameters (except step width SD), and age in the MS group (Table 3). Given that age was correlated with gait variability in our sample and that gait variability has been previously associated with advanced age [12] and gender [21], we examined correlations between EDSS and gait in the MS group while controlling for age and gender. Partial correlations between EDSS and all average parameters and gait variability remained significant in individuals with MS while controlling for age and gender (Table 3). After controlling for age and gender, individuals with MS who used assistive devices during ambulation had significantly greater Table 1 Participant characteristics. Characteristic
MS (N = 88)
Controls (N = 20)
Age (years) SD Range
52.4 11.1 30–78
50.9 8.7 31–62
MS duration (years) SD Range
11.8 9.9 0–43
N/A
EDSS IQR Range
4.5 3.0 2.0–6.5
N/A
Gender (% female)
83.0
80.0
Unilateral assist (% users)
31.8
0.0
5.7
0.0
Bilateral assist (% users)
Note: MS, multiple sclerosis; N, sample size; SD, standard deviation; EDSS, Expanded Disability Status Scale; IQR, inter-quartile range.
M.J. Socie et al. / Gait & Posture 38 (2013) 51–55 Table 2 Spatiotemporal gait parameters.
Table 4 Gait parameters according to assistive device use.
Parameter
MS
Control
p
d
Walking velocity (m/s) Step length (cm) Step time (ms) Step width (cm) Step length SD (cm) Step time SD (ms) Step width SD (cm) Step length CV (%) Step time CV (%) Step width CV (%)
1.0 0.3 58.0 12.5 603 101 12.6 4.3 2.7 1.0 3.0 2.0 2.1 0.8 5.1 2.7 4.7 2.4 19.2 11.9
1.4 0.1 73.8 7.4 530 34 8.6 2.5 2.2 1.3 1.0 0.03 1.5 0.5 2.0 0.8 1.8 0.5 18.5 5.2
<0.001 <0.001 <0.001 0.002 0.04 <0.001 0.003 <0.001 <0.001 0.803
1.3 1.5 0.8 0.9 0.4 1.4 0.9 1.1 1.2 0.1
Note: Values are mean standard deviation; p, statistical significance of group difference; d, Cohen’s d effect size; CV, coefficient of variation.
EDSS scores and step length variability, and significantly lower walking velocity and average step length than individuals with MS who walked independently (Table 4). The effect sizes for these differences were large (d 0.8). There were no differences between assistive device groups in step width, step time, step width SD or CV, and step time CV (p > 0.05). 3.4. Average vs. variability parameters A Fisher’s z-test demonstrated that the partial correlation between average step time and EDSS was significantly greater than the partial correlation between step time variability (indexed with either SD or CV) and EDSS (p = 0.003). There were no significant differences between strength of correlations between EDSS and step length vs. step length variability (p = 0.184) nor step width vs. step width variability (p = 0.170). The effect sizes for significant assistive device group differences in walking velocity (d = 1.3) and average step length (d = 0.9) were seemingly greater than the effect size for step length variability (d = 0.8). 4. Discussion Gait impairment is highly prevalent in individuals with MS and has been characterized by mean spatiotemporal gait parameters, but less so by variability in those parameters [1,2,6,7]. This investigation examined differences in average and variability of spatiotemporal gait parameters in a relatively large group of persons with MS (N = 88), and healthy controls. Additionally, we investigated associations among variability and average metrics of gait and disability, indexed by EDSS and use of assistive devices
Table 3 Pearson’s, Spearman’s and partial correlation coefficients in persons with MS. Partial
Pearson/Spearman Age
EDSS
r
r Age Walking velocity Step length Step time Step width Step length SD Step time SD Step width SD Step length CV Step time CV Step width CV
0.73* 0.61* 0.55* 0.39* 0.36* 0.39* 0.09 0.51* 0.35* 0.27*
53
r
r 0.44* 0.76* 0.63* 0.61* 0.40* 0.38* 0.52* 0.13 0.57* 0.40* 0.35*
EDSS
0.33 0.36 0.01 0.14 0.26 0.07 0.07 0.32 0.12 0.07
rp – 0.36* 0.42* 0.19 0.11 0.23* 0.22* 0.02 0.37* 0.22* 0.07
– 0.67 0.49 0.57 0.35 0.26 0.41 0.12 0.42 0.34 0.26
p – <0.001 <0.001 <0.001 0.001 0.01 <0.001 0.28 <0.001 0.001 0.014
Note: r, Pearson’s correlation coefficient; r, Spearman correlation coefficient; rp, partial correlation coefficient. * p < .05.
N Age (years) EDSS Walking velocity (m/s) Step length (cm) Step time (ms) Step width (cm) Step length SD (cm) Step time SD (ms) Step width SD (cm) Step length CV (%) Step time CV (%) Step width CV (%)
Independently ambulatory
Ambulatory with assistance
p
d
55 49.2 11.3 3.6 1.2 1.2 0.3 62.3 10.9 563 70 11.5 4.0 2.5 0.9 2.5 1.4 2.2 0.8 4.3 2.2 4.3 2.0 21.8 13.2
33 57.7 8.6 5.9 0.7 0.8 0.2 50.8 11.9 669 110 14.5 7.9 3.1 1.0 3.6 2.4 2.2 0.8 6.5 3.0 5.4 2.7 15.0 7.9
– – 0.001 0.029 0.003 0.364 0.805 0.006 0.008 0.50 0.019 0.745 0.763
– – 1.5 1.3 0.9 1.1 0.7 –0.6 –0.6 0.0 0.8 0.5 0.6
Note: Values are mean standard deviation. p, statistical significance of group difference; d, Cohen’s d effect size; N, sample size; EDSS, Expanded Disability Status Scale; CV, coefficient of variation.
during ambulation in persons with MS. Finally, we examined the possibility that gait variability is more closely related to dysfunction in MS than average gait parameters. The first observation in this investigation was that the MS group had significantly greater variability of step length and step time, but no difference in step width variability (as indexed by CV) compared with the healthy control group. This is consistent with previous investigations, which demonstrated significantly greater variability (CV) of step time [8], step length [17], and stride length [19] in persons with MS vs. controls. The unique aspect of the current observation is that we demonstrate differences in step length and time variability in the same sample, whereas previous investigations identified differences in one or the other, but not both. This is likely due to the larger sample size and wide range of disability levels in the MS group in the current study. Additionally, the current investigation examined gait variability in persons with MS utilizing both absolute (SD) and relative (CV) variability measures. The use of both metrics corresponds to the lack of a gold standard in quantifying gait variability [16]. The only difference between metrics was greater absolute step width variability in persons with MS compared to controls, whereas there was no difference in relative variability. This discrepancy has potential clinical import given that relative step width variability has been linked to falls in other clinical populations [11]. The third novel observation of this investigation was that variability of step length and step time correlated positively with EDSS in individuals with MS, i.e. persons with greater disability exhibit greater step length variability and step time variability. Step length variability was also greater in individuals with MS who used assistive devices compared to those who did not. These associations remained significant when controlling for age and gender, thereby suggesting that changes in gait variability in individuals with MS over a range of disability levels are not driven solely by age-related declines in locomotion, but rather disease specific dysfunction. There are multiple possible explanations why step length and step time variability are related to disability in persons with MS. One might be muscle quality (ratio of muscle strength to lean muscle mass, i.e. functional muscle strength) given that this parameter is associated with gait variability in healthy older adults [22]. Another study demonstrated that decreased muscle strength as well as deficits in balance and proprioception are related to gait variability in elderly individuals [23]. Given that persons with MS have decreased muscle strength, proprioception, and balance and that these characteristics are related to gait impairment [24,25],
54
M.J. Socie et al. / Gait & Posture 38 (2013) 51–55
changes in these factors as a function of disability may contribute to gait variability in MS. This suggests that progressive resistance training might be an effective strategy for changing metrics of gait variability in MS. Another potential explanation for the increase in gait variability among the more disabled individuals with MS is an increase in noise within the neuromuscular system. Walking models have demonstrated that an increase in neuromuscular signal noise leads to increased gait variability [26]. Indeed, elevated neuromuscular noise has been previously suggested to contribute to impairment in MS [27]. However, specific evidence linking gait variability to neuromuscular signal activity in MS has not been reported. Use of assistive devices may be another explanation for increasing gait variability with increasing disability [28]. This study observed greater step length variability in individuals with MS who walk with assistive devices compared to those who walk independently. The additional burden of coordinating body movement along with that of an assistive device could further tax limited neuromotor resources. Perhaps better fitting of devices to users or increased physical activity and use of assistive devices could alleviate some of the impact of assistive devices on gait variability. Our results highlight differences in the associations between step length and step time variability and disability compared to step width variability and disability in persons with MS. Step length and step time variability increased with disability, whereas step width variability decreased as a function of disability. Additionally, there is a difference in magnitude between observed values of step time/length CV (5%) and step width CV (20%). The approximately fourfold increase in magnitude of step width variability suggests that the mediolateral plane is less tightly controlled than other parameters, which has been demonstrated in postural control literature [29]. One potential explanation for such observations is the possibility that stability gait parameters (e.g. step width) and propulsion gait parameters (e.g. step length, step time) are controlled by separate neural circuits that could be differently influenced by disability [30,31]. In contrast to step length and step time variability, step width variability was negatively correlated with EDSS. Again, one possible explanation for lower step width variability in more disabled individuals is the use of assistive devices. One report asserts that the use of a cane leads to significant reduction of gait variability in MS [28]. However, the lack of difference in step width variability between the independently ambulatory and ambulatory with assistance groups in the current investigation makes this possibility less likely. Another explanation for the negative correlation between step width variability and EDSS could be that individuals with MS who have greater disability walk with a more rigid gait pattern. Previous research demonstrated that individuals with MS exhibit lower approximate entropy of step width than controls, which is suggested to indicate a more rigid gait pattern [19]. This more rigid gait pattern could be characterized by reduced step width variability. Improvements in flexibility and balance through intervention could potentially reduce this rigid gait pattern. Gait variability has been hypothesized to be more sensitive to dysfunction than average gait parameters in MS [17,19]. Collectively, our results do not support this possibility. For example, the partial correlation between average step time and EDSS was stronger than the correlation between absolute and relative step time variability and EDSS and there were no significant differences in strength of correlations between average step length/width and variability of step length/width and EDSS. Furthermore, the effect sizes for differences between individuals with MS who used assistive devices to walk and those who walked independently were larger for walking velocity, step length, step time, and step
width (d = 1.3, 0.9, 1.1, 0.7, respectively) than differences in variability of step length, time, and width (d = 0.8, 0.5, 0.6, respectively). However, we note that the current metrics of gait variability (SD and CV) quantifies only magnitude and does not consider structure of variability in time [15,16]. Other metrics (e.g. detrended fluctuation analysis) could be investigated to further characterize gait variability in MS and to determine the utility of that information relative to average gait parameters. The clinical significance of gait variability in MS is unclear. Interventions have been successful in increasing function in MS [32], but associations between functional interventions and gait variability are unknown. The specific targets of interventions (e.g. leg strength or balance) and subsequent changes to gait variability could help define the underlying factors driving gait variability in MS. From the clinical perspective of fall risk, gait variability has been associated with falls in other clinical populations [12,13], but this association has not been examined in MS. Given that one of the factors contributing to falls in MS is disability [33] and that the current results demonstrate an association between step width variability and disability, gait variability may be connected to falls in persons with MS. Despite the novel observations, there were limitations in the current investigation. The cross-sectional nature of this study does not provide information about changing gait variability in an individual with MS over time. Another potential limitation was the relatively small number of steps collected to determine gait variability. Previous research has reported on variability of spatiotemporal parameters over similar time series [8,11,12,23], however some research has suggested that a greater number of steps are required to investigate gait variability in certain situations [34]. Lastly, it is not clear how participants perceived their walking impairment. Future work should examine how perceived walking impairment is related to gait variability in persons with MS [35]. In summary, results of the current study illustrate that persons with MS have greater variability of step length and step time than individuals without MS. Additionally, step length and step time variability correlate positively with disability while step width variability correlates negatively with EDSS. We have also demonstrated that average gait parameters are a better marker of disability than the current metric (CV) of gait variability. Despite this fact, further research considering gait variability in MS is warranted to investigate associations between gait variability, motor control function, assistive device usage, and falls in MS. Future research should also consider alternative metrics of gait variability that quantify structure in addition to metrics of variability magnitude. Acknowledgements This investigation was funded in part by the OSF Foundation, who took no role in experimental design or manuscript preparation. The authors would like to thank members of the Motor Control Research and Exercise Neuroscience Laboratories at the University of Illinois for their contributions towards data collection. Conflict of interest statement The authors declare that they have no conflict of interest. References [1] Noseworthy JH, Lucchinetti C, Rodriguez M, Weinshenker BG. Multiple sclerosis. New England Journal of Medicine 2000;343:938–52. [2] Pearson OR, Busse ME, van Deursen RW, Wiles CM. Quantification of walking mobility in neurological disorders. QJM 2004;97:463–75. [3] Kurtzke JF. Historical and clinical perspectives of the Expanded Disability Status Scale. Neuroepidemiology 2008;31:1–9.
M.J. Socie et al. / Gait & Posture 38 (2013) 51–55 [4] Kurtzke JF. Rating neurologic impairment in multiple sclerosis: an Expanded Disability Status Scale (EDSS). Neurology 1983;33:1444–52. [5] Souza A, Kelleher A, Cooper R, Cooper RA, Iezzoni LI, Collins DM. Multiple sclerosis and mobility-related assistive technology: systematic review of literature. Journal of Rehabilitation Research and Development 2010;47:213–23. [6] Larocca NG. Impact of walking impairment in multiple sclerosis: perspectives of patients and care partners. Patient 2011;4:189–201. [7] Givon U, Zeilig G, Achiron A. Gait analysis in multiple sclerosis: characterization of temporal–spatial parameters using GAITRite functional ambulation system. Gait and Posture 2009;29:138–42. [8] Sosnoff JJ, Sandroff BM, Motl RW. Quantifying gait abnormalities in persons with multiple sclerosis with minimal disability. Gait and Posture 2012;36:154–6. [9] Kelleher KJ, Spence W, Solomonidis S, Apatsidis D. The characterisation of gait patterns of people with multiple sclerosis. Disability and Rehabilitation 2010;32:1242–50. [10] Sosnoff JJ, Weikert M, Dlugonski D, Smith DC, Motl RW. Quantifying gait impairment in multiple sclerosis using GAITRite technology. Gait and Posture 2011;34:145–7. [11] Brach JS, Berlin JE, VanSwearingen JM, Newman AB, Studenski SA. Too much or too little step width variability is associated with a fall history in older persons who walk at or near normal gait speed. Journal of Neuroengineering and Rehabilitation 2005;2:21. [12] Brach JS, Studenski SA, Perera S, VanSwearingen JM, Newman AB. Gait variability and the risk of incident mobility disability in community-dwelling older adults. Journals of Gerontology A Biological Sciences and Medical Sciences 2007;62:983–8. [13] Hausdorff JM. Gait variability: methods, modeling and meaning. Journal of Neuroengineering and Rehabilitation 2005;2:19. [14] Lord S, Baker K, Nieuwboer A, Burn D, Rochester L. Gait variability in Parkinson’s disease: an indicator of non-dopaminergic contributors to gait dysfunction? Journal of Neurology 2011;258:566–72. [15] Newell KM, Mayer-Kress G, Liu YT. Aging, time scales, and sensorimotor variability. Psychology and Aging 2009;24:809–18. [16] Stergiou N, Harbourne R, Cavanaugh J. Optimal movement variability: a new theoretical perspective for neurologic physical therapy. Journal of Neurologic Physical Therapy 2006;30:120–9. [17] Flegel M, Knox K, Nickel D. Step-length variability in minimally disabled women with multiple sclerosis or clinically isolated syndrome. International Journal of Multiple Sclerosis Care 2012;14:26–30. [18] Crenshaw SJ, Royer TD, Richards JG, Hudson DJ. Gait variability in people with multiple sclerosis. Multiple Sclerosis 2006;12:613–9. [19] Kaipust JP, Huisinga JM, Filipi M, Stergiou N. Gait variability measures reveal differences between multiple sclerosis patients and healthy controls. Motor Control 2012;16:229–44.
55
[20] Cohen J. Statistical power analysis for the behavioral sciences, 3rd ed., New York: Academic Press; 1988. [21] Barrett R, Noordegraaf MV, Morrison S. Gender differences in the variability of lower extremity kinematics during treadmill locomotion. Journal of Motor Behavior 2008;40:62–70. [22] Shin S, Valentine RJ, Evans EM, Sosnoff JJ. Lower extremity muscle quality and gait variability in older adults. Age and Ageing 2012;41:595–9. [23] Callisaya ML, Blizzard L, McGinley JL, Schmidt MD, Srikanth VK. Sensorimotor factors affecting gait variability in older people—a population-based study. Journals of Gerontology A Biological Sciences and Medical Sciences 2010;65:386–92. [24] Thoumie P, Lamotte D, Cantalloube S, Faucher M, Amarenco G. Motor determinants of gait in 100 ambulatory patients with multiple sclerosis. Multiple Sclerosis 2005;11:485–91. [25] Broekmans T, Gijbels D, Eijnde BO, Alders G, Lamers I, Roelants M, et al., The relationship between upper leg muscle strength and walking capacity in persons with multiple sclerosis. Multiple Sclerosis 2012, in press. [26] Roos PE, Dingwell JB. Influence of simulated neuromuscular noise on movement variability and fall risk in a 3D dynamic walking model. Journal of Biomechanics 2010;43:2929–35. [27] Kail R. Speed of information processing in patients with multiple sclerosis. Journal of Clinical and Experimental Neuropsychology 1998;20:98–106. [28] Gianfrancesco MA, Triche EW, Fawcett JA, Labas MP, Patterson TS, Lo AC. Speed- and cane-related alterations in gait parameters in individuals with multiple sclerosis. Gait and Posture 2011;33:140–2. [29] O’Connor SM, Kuo AD. Direction-dependent control of balance during walking and standing. Journal of Neurophysiology 2009;102:1411–9. [30] Al-Yahya E, Dawes H, Smith L, Dennis A, Howells K, Cockburn J. Cognitive motor interference while walking: a systematic review and meta-analysis. Neuroscience and Biobehavioral Reviews 2011;35:715–28. [31] Moe-Nilssen R, Aaslund MK, Hodt-Billington C, Helbostad JL. Gait variability measures may represent different constructs. Gait and Posture 2010;32: 98–101. [32] Kraft GH. Rehabilitation still the only way to improve function in multiple sclerosis. Lancet 1999;354:2016–7. [33] Sosnoff JJ, Socie MJ, Boes MK, Sandroff BM, Pula JH, Suh Y, et al. Mobility, balance and falls in persons with multiple sclerosis. PLoS One 2011;6:e28021. [34] Hollman JH, Childs KB, McNeil ML, Mueller AC, Quilter CM, Youdas JW. Number of strides required for reliable measurements of pace, rhythm and variability parameters of gait during normal and dual task walking in older individuals. Gait and Posture 2010;32:23–8. [35] Pilutti LA, Dlugonski D, Sandroff BM, Suh Y, Pula JH, Sosnoff JJ, et al., Further validation of Multiple Sclerosis Walking Scale-12 scores based on spatiotemporal gait parameters. Archives of Physical Medicine and Rehabilitation 2012, in press.
Contents lists available at SciVerse ScienceDirect
Gait & Posture journal homepage: www.elsevier.com/locate/gaitpost
Gait variability and disability in multiple sclerosis Michael J. Socie a, Robert W. Motl b, John H. Pula c, Brian M. Sandroff b, Jacob J. Sosnoff b,* a
Department of Mechanical Science and Engineering, University of Illinois, Urbana, IL, United States Department of Kinesiology and Community Health, University of Illinois, Urbana, IL, United States c Illinois Neurological Institute, University of Illinois College of Medicine, Peoria, IL, United States b
A R T I C L E I N F O
A B S T R A C T
Article history: Received 19 June 2012 Received in revised form 4 October 2012 Accepted 20 October 2012
Gait variability is clinically relevant in some populations, but there is limited documentation of gait variability in persons with multiple sclerosis (MS). This investigation examined average and variability of spatiotemporal gait parameters in persons with MS and healthy controls and subsequent associations with disability status. 88 individuals with MS (age 52.4 11.1) and 20 healthy controls (age 50.9 8.7) performed two self-paced walking trials on a 7.9-m electronic walkway to determine gait parameters. Disability was indexed by the Expanded Disability Status Scale (EDSS) and ranged between 2.5 and 6.5. Gait variability was indexed by standard deviation (SD) and coefficient of variation (CV = SD/mean) of step time, step length, and step width. Average gait parameters were significantly correlated with EDSS (r = 0.756–0.609) and were significantly different in individuals with MS compared to controls (p 0.002). Also, step length (p < 0.001) and step time (p < 0.001) variability were both significantly greater in MS compared to controls. EDSS was positively correlated with step length variability and individuals with MS who used assistive devices to walk had significantly greater step length variability than those who walked independently (p’s < .05). EDSS was correlated with step time and length variability even when age was taken into account. Additionally, Fisher’s z test of partial correlations revealed that average gait parameters were more closely related to disability status than gait variability in individuals with MS. This suggests that focusing on average gait parameters may be more important than variability in therapeutic interventions in MS. ß 2012 Elsevier B.V. All rights reserved.
Keywords: Walking impairment EDSS Motor function Spatiotemporal gait parameters
1. Introduction Multiple sclerosis (MS) is a neurologic disease that affects an estimated 400,000 American adults and 2.5 million adults worldwide. MS results in demyelination and axonal loss in the central nervous system (CNS). Such CNS damage may result in physical disability (potentially including weakness, sensory loss, and/or ataxia) and gait impairment [1,2]. Disability in MS can be indexed by the Expanded Disability Status Scale (EDSS) [3,4], which is influenced by ambulatory status. EDSS scores range from 0 to 10, where 0 represents no impairment, 4 represents the onset of significant walking impairment, 6 represents onset of assistive device use during ambulation, and 10 represents death due to MS [4]. In addition to the EDSS, which is a clinical standard, disability level in MS has been characterized by use of assistive devices
* Corresponding author at: Department of Kinesiology and Community Health, University of Illinois at Urbana-Champaign, 301 Freer Hall, 906 S. Goodwin Ave., Urbana, IL 61801, United States. Tel.: +1 217 333 9472, fax: +1 217 244 7322. E-mail addresses: [email protected] (M.J. Socie), [email protected] (R.W. Motl), [email protected] (J.H. Pula), [email protected] (B.M. Sandroff), [email protected] (J.J. Sosnoff). 0966-6362/$ – see front matter ß 2012 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.gaitpost.2012.10.012
during ambulation [5]. Greater than 50% of individuals with MS require assistive devices within 15 years of disease onset [1]. Gait impairment is one of the most frequent consequences of MS, and walking dysfunction is considered by the majority of patients as the most challenging, life-altering aspect of the disease [6]. Persons with MS walk slower, take shorter, wider, and slower steps, and spend a greater percent of the gait cycle in doublesupport compared with healthy controls, even early in the disease course [7–9]. Such gait impairment have also been directly associated with disability status as a marker of disease progression [7,10]. Gait variability (i.e. fluctuations in gait parameters between steps) is predictive of mobility impairment and falls in older adults and other neurological populations [11–14]. Movement variability is further associated with health in biological systems and is a marker of motor control function [15,16]. Gait variability might be more sensitive to dysfunction in MS than traditional parameters such as walking velocity [17]. However, we are unaware of research testing that possibility and this is important in forming therapeutic interventions for MS. There is some evidence that persons with MS have elevated gait variability compared to controls. For example, one study reported that persons with MS exhibit greater kinematic variability at the
52
M.J. Socie et al. / Gait & Posture 38 (2013) 51–55
hip, knee, and ankle during ambulation than healthy controls [18]. Other studies have reported that persons with MS who had mild impairment demonstrated greater variability of step time and single-support time [8] and step length [17] than controls. Data from a small sample (N = 10) indicates that gait variability increases with disability in MS [19]. That study reported that only persons with MS who had walking impairment (EDSS > 4.0) had greater variability of stride length compared to healthy controls, but this could be driven by low statistical power associated with the small sample and should be replicated with a larger sample. Consequently, further documentation of gait variability as a function of disability is a necessary step toward investigating the clinical relevance of gait variability in MS. This investigation examined average and variability of spatiotemporal gait parameters in persons with MS with a wide range of disability levels and healthy controls as well as associations between average gait parameters, gait variability and disability in persons with MS. We hypothesized that individuals with MS would exhibit greater gait variability than controls. Second, we hypothesized that gait variability would be greater in persons with MS with greater disability. Third, we hypothesized that associations between gait variability and disability would be stronger than associations between average gait parameters and disability in MS.
between individuals with MS who used assistive devices to walk and those who walked independently were determined using a mixed-model ANCOVA while controlling for age and gender. 2.3.3. Average vs. variability parameters A Fisher’s z-test was used to determine differences between the strength of correlations between average gait parameters and EDSS compared to gait variability and EDSS.
3. Results 3.1. Participant characteristics Eighty-eight participants with MS and 20 healthy controls participated in this investigation (Table 1). The MS group had an average age of 52.4 years, average duration of MS of 11.8 years, and median EDSS score of 4.5. 83% of the MS group was female, 32% used unilateral assistive devices (e.g. canes), and 6% used bilateral assistive devices (e.g. walkers). The control group had an average age of 50.9 years and 80% of the group was female. There were no differences between groups in demographic characteristics (p > 0.05). 3.2. MS vs. controls
2. Methods The procedures for this investigation were approved by a University Institutional Review Board and all participants provided written informed consent prior to data collection. 2.1. Participants Persons with and without MS participated in this investigation. Inclusion criteria for participants with MS required a neurologist-confirmed diagnosis, the ability to walk without or with an assistive device (e.g. a cane or walker) and be relapse-free for at least 30 days prior to testing. Inclusion criteria for controls required no gait impairment, no assistive device use, and no medical condition causing significant morbidity such as neurological or cardiovascular disease. 2.2. Procedures Upon arrival at the testing facility, participants with MS underwent an examination by a neurologist to generate an EDSS score [4]. To determine spatiotemporal gait parameters and gait variability, all participants performed two walking trials across a 7.9-m electronic walkway (GAITRiteTM, CIR Systems Inc., Haverton, PA, USA) at self-selected, comfortable speed. Participants began walking 1.5 m in front of he GAITRiteTM and continued to walk 1.5 m beyond the mat to insure that steady state gait was measured. On average the MS group took 13.3 steps, whereas the control group took 10.7 steps per trial. Use of assistive devices was permitted during testing. Average walking velocity, step length, step time, and step width were recorded. Variability of step length, step width, and step time was indexed by the standard deviation (SD) as well as the coefficient of variation (CV = SD/mean). The SD is operationalized as an absolute measure of variability, whereas CV is operationalized as a relative measure of variability. Gait parameters were calculated individually for each pass over the walkway; those values were then averaged across trials to produce final values. 2.3. Statistical analysis Statistical analyses were performed using SPSS software version 19.0 (SPSS Inc., Chicago, IL, USA). Statistical significance was assumed for p < 0.05. Group differences in age and gender distribution were determined using an independent sample t-test and chi-square test, respectively. 2.3.1. MS vs. controls Differences in average and variability of gait parameters between individuals with MS and controls were determined using independent sample t-test. The magnitudes of group differences were indexed by Cohen’s d effect sizes [20]. 2.3.2. Gait variability and disability Associations between gait variability and disability level were examined in the context of EDSS as well as assistive device use. Pearson’s and Spearman’s correlations were performed for average gait parameters, gait variability, age, and EDSS in the MS group. Partial correlations were also performed for gait parameters and EDSS while controlling for age and gender. Differences in gait parameters
The MS group demonstrated lower walking velocity and step length and greater step time and step width than the control group (Table 2). Step length and step time SD and CV were significantly greater in MS vs. controls (p < 0.05). The effect sizes of these differences were moderate to large in magnitude (d’s ranged from 0.4 to 1.4). The MS group also had a greater step width SD, but not CV, compared to the control group. 3.3. Gait variability and disability EDSS was significantly correlated with all average gait parameters, gait variability parameters (except step width SD), and age in the MS group (Table 3). Given that age was correlated with gait variability in our sample and that gait variability has been previously associated with advanced age [12] and gender [21], we examined correlations between EDSS and gait in the MS group while controlling for age and gender. Partial correlations between EDSS and all average parameters and gait variability remained significant in individuals with MS while controlling for age and gender (Table 3). After controlling for age and gender, individuals with MS who used assistive devices during ambulation had significantly greater Table 1 Participant characteristics. Characteristic
MS (N = 88)
Controls (N = 20)
Age (years) SD Range
52.4 11.1 30–78
50.9 8.7 31–62
MS duration (years) SD Range
11.8 9.9 0–43
N/A
EDSS IQR Range
4.5 3.0 2.0–6.5
N/A
Gender (% female)
83.0
80.0
Unilateral assist (% users)
31.8
0.0
5.7
0.0
Bilateral assist (% users)
Note: MS, multiple sclerosis; N, sample size; SD, standard deviation; EDSS, Expanded Disability Status Scale; IQR, inter-quartile range.
M.J. Socie et al. / Gait & Posture 38 (2013) 51–55 Table 2 Spatiotemporal gait parameters.
Table 4 Gait parameters according to assistive device use.
Parameter
MS
Control
p
d
Walking velocity (m/s) Step length (cm) Step time (ms) Step width (cm) Step length SD (cm) Step time SD (ms) Step width SD (cm) Step length CV (%) Step time CV (%) Step width CV (%)
1.0 0.3 58.0 12.5 603 101 12.6 4.3 2.7 1.0 3.0 2.0 2.1 0.8 5.1 2.7 4.7 2.4 19.2 11.9
1.4 0.1 73.8 7.4 530 34 8.6 2.5 2.2 1.3 1.0 0.03 1.5 0.5 2.0 0.8 1.8 0.5 18.5 5.2
<0.001 <0.001 <0.001 0.002 0.04 <0.001 0.003 <0.001 <0.001 0.803
1.3 1.5 0.8 0.9 0.4 1.4 0.9 1.1 1.2 0.1
Note: Values are mean standard deviation; p, statistical significance of group difference; d, Cohen’s d effect size; CV, coefficient of variation.
EDSS scores and step length variability, and significantly lower walking velocity and average step length than individuals with MS who walked independently (Table 4). The effect sizes for these differences were large (d 0.8). There were no differences between assistive device groups in step width, step time, step width SD or CV, and step time CV (p > 0.05). 3.4. Average vs. variability parameters A Fisher’s z-test demonstrated that the partial correlation between average step time and EDSS was significantly greater than the partial correlation between step time variability (indexed with either SD or CV) and EDSS (p = 0.003). There were no significant differences between strength of correlations between EDSS and step length vs. step length variability (p = 0.184) nor step width vs. step width variability (p = 0.170). The effect sizes for significant assistive device group differences in walking velocity (d = 1.3) and average step length (d = 0.9) were seemingly greater than the effect size for step length variability (d = 0.8). 4. Discussion Gait impairment is highly prevalent in individuals with MS and has been characterized by mean spatiotemporal gait parameters, but less so by variability in those parameters [1,2,6,7]. This investigation examined differences in average and variability of spatiotemporal gait parameters in a relatively large group of persons with MS (N = 88), and healthy controls. Additionally, we investigated associations among variability and average metrics of gait and disability, indexed by EDSS and use of assistive devices
Table 3 Pearson’s, Spearman’s and partial correlation coefficients in persons with MS. Partial
Pearson/Spearman Age
EDSS
r
r Age Walking velocity Step length Step time Step width Step length SD Step time SD Step width SD Step length CV Step time CV Step width CV
0.73* 0.61* 0.55* 0.39* 0.36* 0.39* 0.09 0.51* 0.35* 0.27*
53
r
r 0.44* 0.76* 0.63* 0.61* 0.40* 0.38* 0.52* 0.13 0.57* 0.40* 0.35*
EDSS
0.33 0.36 0.01 0.14 0.26 0.07 0.07 0.32 0.12 0.07
rp – 0.36* 0.42* 0.19 0.11 0.23* 0.22* 0.02 0.37* 0.22* 0.07
– 0.67 0.49 0.57 0.35 0.26 0.41 0.12 0.42 0.34 0.26
p – <0.001 <0.001 <0.001 0.001 0.01 <0.001 0.28 <0.001 0.001 0.014
Note: r, Pearson’s correlation coefficient; r, Spearman correlation coefficient; rp, partial correlation coefficient. * p < .05.
N Age (years) EDSS Walking velocity (m/s) Step length (cm) Step time (ms) Step width (cm) Step length SD (cm) Step time SD (ms) Step width SD (cm) Step length CV (%) Step time CV (%) Step width CV (%)
Independently ambulatory
Ambulatory with assistance
p
d
55 49.2 11.3 3.6 1.2 1.2 0.3 62.3 10.9 563 70 11.5 4.0 2.5 0.9 2.5 1.4 2.2 0.8 4.3 2.2 4.3 2.0 21.8 13.2
33 57.7 8.6 5.9 0.7 0.8 0.2 50.8 11.9 669 110 14.5 7.9 3.1 1.0 3.6 2.4 2.2 0.8 6.5 3.0 5.4 2.7 15.0 7.9
– – 0.001 0.029 0.003 0.364 0.805 0.006 0.008 0.50 0.019 0.745 0.763
– – 1.5 1.3 0.9 1.1 0.7 –0.6 –0.6 0.0 0.8 0.5 0.6
Note: Values are mean standard deviation. p, statistical significance of group difference; d, Cohen’s d effect size; N, sample size; EDSS, Expanded Disability Status Scale; CV, coefficient of variation.
during ambulation in persons with MS. Finally, we examined the possibility that gait variability is more closely related to dysfunction in MS than average gait parameters. The first observation in this investigation was that the MS group had significantly greater variability of step length and step time, but no difference in step width variability (as indexed by CV) compared with the healthy control group. This is consistent with previous investigations, which demonstrated significantly greater variability (CV) of step time [8], step length [17], and stride length [19] in persons with MS vs. controls. The unique aspect of the current observation is that we demonstrate differences in step length and time variability in the same sample, whereas previous investigations identified differences in one or the other, but not both. This is likely due to the larger sample size and wide range of disability levels in the MS group in the current study. Additionally, the current investigation examined gait variability in persons with MS utilizing both absolute (SD) and relative (CV) variability measures. The use of both metrics corresponds to the lack of a gold standard in quantifying gait variability [16]. The only difference between metrics was greater absolute step width variability in persons with MS compared to controls, whereas there was no difference in relative variability. This discrepancy has potential clinical import given that relative step width variability has been linked to falls in other clinical populations [11]. The third novel observation of this investigation was that variability of step length and step time correlated positively with EDSS in individuals with MS, i.e. persons with greater disability exhibit greater step length variability and step time variability. Step length variability was also greater in individuals with MS who used assistive devices compared to those who did not. These associations remained significant when controlling for age and gender, thereby suggesting that changes in gait variability in individuals with MS over a range of disability levels are not driven solely by age-related declines in locomotion, but rather disease specific dysfunction. There are multiple possible explanations why step length and step time variability are related to disability in persons with MS. One might be muscle quality (ratio of muscle strength to lean muscle mass, i.e. functional muscle strength) given that this parameter is associated with gait variability in healthy older adults [22]. Another study demonstrated that decreased muscle strength as well as deficits in balance and proprioception are related to gait variability in elderly individuals [23]. Given that persons with MS have decreased muscle strength, proprioception, and balance and that these characteristics are related to gait impairment [24,25],
54
M.J. Socie et al. / Gait & Posture 38 (2013) 51–55
changes in these factors as a function of disability may contribute to gait variability in MS. This suggests that progressive resistance training might be an effective strategy for changing metrics of gait variability in MS. Another potential explanation for the increase in gait variability among the more disabled individuals with MS is an increase in noise within the neuromuscular system. Walking models have demonstrated that an increase in neuromuscular signal noise leads to increased gait variability [26]. Indeed, elevated neuromuscular noise has been previously suggested to contribute to impairment in MS [27]. However, specific evidence linking gait variability to neuromuscular signal activity in MS has not been reported. Use of assistive devices may be another explanation for increasing gait variability with increasing disability [28]. This study observed greater step length variability in individuals with MS who walk with assistive devices compared to those who walk independently. The additional burden of coordinating body movement along with that of an assistive device could further tax limited neuromotor resources. Perhaps better fitting of devices to users or increased physical activity and use of assistive devices could alleviate some of the impact of assistive devices on gait variability. Our results highlight differences in the associations between step length and step time variability and disability compared to step width variability and disability in persons with MS. Step length and step time variability increased with disability, whereas step width variability decreased as a function of disability. Additionally, there is a difference in magnitude between observed values of step time/length CV (5%) and step width CV (20%). The approximately fourfold increase in magnitude of step width variability suggests that the mediolateral plane is less tightly controlled than other parameters, which has been demonstrated in postural control literature [29]. One potential explanation for such observations is the possibility that stability gait parameters (e.g. step width) and propulsion gait parameters (e.g. step length, step time) are controlled by separate neural circuits that could be differently influenced by disability [30,31]. In contrast to step length and step time variability, step width variability was negatively correlated with EDSS. Again, one possible explanation for lower step width variability in more disabled individuals is the use of assistive devices. One report asserts that the use of a cane leads to significant reduction of gait variability in MS [28]. However, the lack of difference in step width variability between the independently ambulatory and ambulatory with assistance groups in the current investigation makes this possibility less likely. Another explanation for the negative correlation between step width variability and EDSS could be that individuals with MS who have greater disability walk with a more rigid gait pattern. Previous research demonstrated that individuals with MS exhibit lower approximate entropy of step width than controls, which is suggested to indicate a more rigid gait pattern [19]. This more rigid gait pattern could be characterized by reduced step width variability. Improvements in flexibility and balance through intervention could potentially reduce this rigid gait pattern. Gait variability has been hypothesized to be more sensitive to dysfunction than average gait parameters in MS [17,19]. Collectively, our results do not support this possibility. For example, the partial correlation between average step time and EDSS was stronger than the correlation between absolute and relative step time variability and EDSS and there were no significant differences in strength of correlations between average step length/width and variability of step length/width and EDSS. Furthermore, the effect sizes for differences between individuals with MS who used assistive devices to walk and those who walked independently were larger for walking velocity, step length, step time, and step
width (d = 1.3, 0.9, 1.1, 0.7, respectively) than differences in variability of step length, time, and width (d = 0.8, 0.5, 0.6, respectively). However, we note that the current metrics of gait variability (SD and CV) quantifies only magnitude and does not consider structure of variability in time [15,16]. Other metrics (e.g. detrended fluctuation analysis) could be investigated to further characterize gait variability in MS and to determine the utility of that information relative to average gait parameters. The clinical significance of gait variability in MS is unclear. Interventions have been successful in increasing function in MS [32], but associations between functional interventions and gait variability are unknown. The specific targets of interventions (e.g. leg strength or balance) and subsequent changes to gait variability could help define the underlying factors driving gait variability in MS. From the clinical perspective of fall risk, gait variability has been associated with falls in other clinical populations [12,13], but this association has not been examined in MS. Given that one of the factors contributing to falls in MS is disability [33] and that the current results demonstrate an association between step width variability and disability, gait variability may be connected to falls in persons with MS. Despite the novel observations, there were limitations in the current investigation. The cross-sectional nature of this study does not provide information about changing gait variability in an individual with MS over time. Another potential limitation was the relatively small number of steps collected to determine gait variability. Previous research has reported on variability of spatiotemporal parameters over similar time series [8,11,12,23], however some research has suggested that a greater number of steps are required to investigate gait variability in certain situations [34]. Lastly, it is not clear how participants perceived their walking impairment. Future work should examine how perceived walking impairment is related to gait variability in persons with MS [35]. In summary, results of the current study illustrate that persons with MS have greater variability of step length and step time than individuals without MS. Additionally, step length and step time variability correlate positively with disability while step width variability correlates negatively with EDSS. We have also demonstrated that average gait parameters are a better marker of disability than the current metric (CV) of gait variability. Despite this fact, further research considering gait variability in MS is warranted to investigate associations between gait variability, motor control function, assistive device usage, and falls in MS. Future research should also consider alternative metrics of gait variability that quantify structure in addition to metrics of variability magnitude. Acknowledgements This investigation was funded in part by the OSF Foundation, who took no role in experimental design or manuscript preparation. The authors would like to thank members of the Motor Control Research and Exercise Neuroscience Laboratories at the University of Illinois for their contributions towards data collection. Conflict of interest statement The authors declare that they have no conflict of interest. References [1] Noseworthy JH, Lucchinetti C, Rodriguez M, Weinshenker BG. Multiple sclerosis. New England Journal of Medicine 2000;343:938–52. [2] Pearson OR, Busse ME, van Deursen RW, Wiles CM. Quantification of walking mobility in neurological disorders. QJM 2004;97:463–75. [3] Kurtzke JF. Historical and clinical perspectives of the Expanded Disability Status Scale. Neuroepidemiology 2008;31:1–9.
M.J. Socie et al. / Gait & Posture 38 (2013) 51–55 [4] Kurtzke JF. Rating neurologic impairment in multiple sclerosis: an Expanded Disability Status Scale (EDSS). Neurology 1983;33:1444–52. [5] Souza A, Kelleher A, Cooper R, Cooper RA, Iezzoni LI, Collins DM. Multiple sclerosis and mobility-related assistive technology: systematic review of literature. Journal of Rehabilitation Research and Development 2010;47:213–23. [6] Larocca NG. Impact of walking impairment in multiple sclerosis: perspectives of patients and care partners. Patient 2011;4:189–201. [7] Givon U, Zeilig G, Achiron A. Gait analysis in multiple sclerosis: characterization of temporal–spatial parameters using GAITRite functional ambulation system. Gait and Posture 2009;29:138–42. [8] Sosnoff JJ, Sandroff BM, Motl RW. Quantifying gait abnormalities in persons with multiple sclerosis with minimal disability. Gait and Posture 2012;36:154–6. [9] Kelleher KJ, Spence W, Solomonidis S, Apatsidis D. The characterisation of gait patterns of people with multiple sclerosis. Disability and Rehabilitation 2010;32:1242–50. [10] Sosnoff JJ, Weikert M, Dlugonski D, Smith DC, Motl RW. Quantifying gait impairment in multiple sclerosis using GAITRite technology. Gait and Posture 2011;34:145–7. [11] Brach JS, Berlin JE, VanSwearingen JM, Newman AB, Studenski SA. Too much or too little step width variability is associated with a fall history in older persons who walk at or near normal gait speed. Journal of Neuroengineering and Rehabilitation 2005;2:21. [12] Brach JS, Studenski SA, Perera S, VanSwearingen JM, Newman AB. Gait variability and the risk of incident mobility disability in community-dwelling older adults. Journals of Gerontology A Biological Sciences and Medical Sciences 2007;62:983–8. [13] Hausdorff JM. Gait variability: methods, modeling and meaning. Journal of Neuroengineering and Rehabilitation 2005;2:19. [14] Lord S, Baker K, Nieuwboer A, Burn D, Rochester L. Gait variability in Parkinson’s disease: an indicator of non-dopaminergic contributors to gait dysfunction? Journal of Neurology 2011;258:566–72. [15] Newell KM, Mayer-Kress G, Liu YT. Aging, time scales, and sensorimotor variability. Psychology and Aging 2009;24:809–18. [16] Stergiou N, Harbourne R, Cavanaugh J. Optimal movement variability: a new theoretical perspective for neurologic physical therapy. Journal of Neurologic Physical Therapy 2006;30:120–9. [17] Flegel M, Knox K, Nickel D. Step-length variability in minimally disabled women with multiple sclerosis or clinically isolated syndrome. International Journal of Multiple Sclerosis Care 2012;14:26–30. [18] Crenshaw SJ, Royer TD, Richards JG, Hudson DJ. Gait variability in people with multiple sclerosis. Multiple Sclerosis 2006;12:613–9. [19] Kaipust JP, Huisinga JM, Filipi M, Stergiou N. Gait variability measures reveal differences between multiple sclerosis patients and healthy controls. Motor Control 2012;16:229–44.
55
[20] Cohen J. Statistical power analysis for the behavioral sciences, 3rd ed., New York: Academic Press; 1988. [21] Barrett R, Noordegraaf MV, Morrison S. Gender differences in the variability of lower extremity kinematics during treadmill locomotion. Journal of Motor Behavior 2008;40:62–70. [22] Shin S, Valentine RJ, Evans EM, Sosnoff JJ. Lower extremity muscle quality and gait variability in older adults. Age and Ageing 2012;41:595–9. [23] Callisaya ML, Blizzard L, McGinley JL, Schmidt MD, Srikanth VK. Sensorimotor factors affecting gait variability in older people—a population-based study. Journals of Gerontology A Biological Sciences and Medical Sciences 2010;65:386–92. [24] Thoumie P, Lamotte D, Cantalloube S, Faucher M, Amarenco G. Motor determinants of gait in 100 ambulatory patients with multiple sclerosis. Multiple Sclerosis 2005;11:485–91. [25] Broekmans T, Gijbels D, Eijnde BO, Alders G, Lamers I, Roelants M, et al., The relationship between upper leg muscle strength and walking capacity in persons with multiple sclerosis. Multiple Sclerosis 2012, in press. [26] Roos PE, Dingwell JB. Influence of simulated neuromuscular noise on movement variability and fall risk in a 3D dynamic walking model. Journal of Biomechanics 2010;43:2929–35. [27] Kail R. Speed of information processing in patients with multiple sclerosis. Journal of Clinical and Experimental Neuropsychology 1998;20:98–106. [28] Gianfrancesco MA, Triche EW, Fawcett JA, Labas MP, Patterson TS, Lo AC. Speed- and cane-related alterations in gait parameters in individuals with multiple sclerosis. Gait and Posture 2011;33:140–2. [29] O’Connor SM, Kuo AD. Direction-dependent control of balance during walking and standing. Journal of Neurophysiology 2009;102:1411–9. [30] Al-Yahya E, Dawes H, Smith L, Dennis A, Howells K, Cockburn J. Cognitive motor interference while walking: a systematic review and meta-analysis. Neuroscience and Biobehavioral Reviews 2011;35:715–28. [31] Moe-Nilssen R, Aaslund MK, Hodt-Billington C, Helbostad JL. Gait variability measures may represent different constructs. Gait and Posture 2010;32: 98–101. [32] Kraft GH. Rehabilitation still the only way to improve function in multiple sclerosis. Lancet 1999;354:2016–7. [33] Sosnoff JJ, Socie MJ, Boes MK, Sandroff BM, Pula JH, Suh Y, et al. Mobility, balance and falls in persons with multiple sclerosis. PLoS One 2011;6:e28021. [34] Hollman JH, Childs KB, McNeil ML, Mueller AC, Quilter CM, Youdas JW. Number of strides required for reliable measurements of pace, rhythm and variability parameters of gait during normal and dual task walking in older individuals. Gait and Posture 2010;32:23–8. [35] Pilutti LA, Dlugonski D, Sandroff BM, Suh Y, Pula JH, Sosnoff JJ, et al., Further validation of Multiple Sclerosis Walking Scale-12 scores based on spatiotemporal gait parameters. Archives of Physical Medicine and Rehabilitation 2012, in press.