Journal Logo

Original Articles

Gait characteristics of post-stroke hemiparetic patients with different walking speeds

Wang, Yijia,,b,,c; Mukaino, Masahikoa; Ohtsuka, Keid; Otaka, Yoheia; Tanikawa, Hirokid; Matsuda, Fumihirod; Tsuchiyama, Kazuhirod; Yamada, Junyae; Saitoh, Eiichia

Author Information
International Journal of Rehabilitation Research: March 2020 - Volume 43 - Issue 1 - p 69-75
doi: 10.1097/MRR.0000000000000391
  • Open

Abstract

Introduction

Gait disorder is a common clinical problem for stroke survivors and is among the prevalent physical limitations contributing to stroke-related disability that impacts performance of activities of daily living. Gait disorder is therefore a major target for post-stroke rehabilitation. Many studies have investigated the characteristics and mechanism of gait with various biomechanical evaluation methods, including evaluation of spatiotemporal, kinematic, and kinetic parameters (Nadeau et al., 2013; Balaban and Tok, 2014). Among the aforementioned parameters, spatiotemporal parameters are the simplest to analyze. Spatiotemporal parameter data are easy to obtain using affordable systems such as simplified gait analysis or wearable systems. Thus, deeper understanding of the spatiotemporal patterns of gait disorders could contribute to the improvement of the quality of evaluation and intervention on gait in rehabilitation clinics.

Hemiparetic gait is characterized by specific spatiotemporal patterns, including decreased cadence, prolonged swing duration on the paretic side, prolonged stance duration on the nonparetic side, and step length asymmetry, compared with the gait parameters of healthy subjects (Roth et al., 1997; Chen et al., 2005b; Patterson et al., 2010). However, as the gait speed of healthy subjects is usually higher than that of stroke patients, the differences in spatiotemporal patterns between stroke patients and healthy subjects could be influenced by gait speed (Chen et al., 2005b; Wonsetler and Bowden, 2017). Thus, the speed-matched comparison of gait patterns should be meaningful to understand the features of hemiparetic gait, eliminating the effect of gait speed. Several studies examined speed-matched comparisons between stroke patients and healthy controls and found differences in spatiotemporal patterns of gait, although sample sizes were small (Titianova and Tarkka, 1995; Chen et al., 2005b; Rinaldi and Monaco, 2013). However, inconsistencies were noted among the studies. One study found that the swing time on the affected side was prolonged (Titianova and Tarkka, 1995), whereas other studies presented no significant differences between patients and controls (Chen et al., 2005b; Rinaldi and Monaco, 2013). These inconsistencies may be related to differences in gait speed. For example, the asymmetry in step length and swing time is a strong feature of hemiparetic gait (Chen et al., 2005b), but this may only be seen in patients with lower gait speed (Titianova et al., 2008).

To further understand the feature of spatiotemporal characteristics of hemiparetic gait in patients with high- and low-gait speeds, gait speed-based stratified comparison of gait parameters between hemiparetic patients and healthy controls would be meaningful. Thus, this study aimed to retrospectively investigate the characteristics of the spatiotemporal gait parameters of stroke patients by performing a stratified gait speed matching comparison using the database on three-dimensional gait analysis of stroke patients and healthy subjects.

Methods

Participants

Spatiotemporal data during treadmill gait of 136 individuals with cerebrovascular event and resultant hemiparesis and who underwent three-dimensional gait analysis measurement from January 2015 to September 2017 were extracted from the clinical gait analysis database of the Fujita Health University Hospital.

Hemiparetic subjects who normally walked with a cane or ankle-foot orthosis (AFO) during daily living were allowed to use a handrail and/or AFO while walking on the treadmill. Inclusion criteria were those aged 20–69 years and with unilateral hemiparesis due to stroke. The exclusion criteria were presence of orthopedic disease, severe cardiopulmonary disease limiting gait ability, and unstable medical condition. Each subject’s walking ability and lower extremity functional motor level were quantified using Brunnstrom’s Motor Recovery Stage and Stroke Impairment Assessment Set, which scores lower limb motor function from 0 to 15 (Chino et al., 1994).

Data of healthy controls were drawn from the Fujita Health University Hospital database, which was developed for a previous study on developing gait analysis methodology (Itoh et al., 2012; Tanikawa et al., 2016; Hishikawa et al., 2018; Mukaino et al., 2018). The database included gait analysis data of 560 trials of 112 individuals aged 20–69 years who volunteered to be measured at walking speeds of 1, 2, 3, 4, and 5 km/h as control speeds. Data of 136 stroke patients were matched with those of healthy controls by age (within ±2 years) and height (within ±5 cm). If there were no control data matching either or both of age and height, these data were excluded from the analysis. After matching, 130 pairs of data in total were analyzed (Fig. 1). The stroke patients and controls were then grouped into the following five categories according to their walking speed: 0.5–1.4, 1.5–2.4, 2.5–3.4, 3.5–4.4, and 4.5–5.5 km/h.

Fig. 1
Fig. 1:
Flow diagram for data extraction and matching.

Procedure

Data were extracted from the database as follows. The details of the measurement method are described elsewhere (Mukaino et al., 2016; Mukaino et al., 2018). In brief, a three-dimensional motion capture system with force plate measurement (KinemaTracer, Kissei Comtec Co., Ltd., Matsumoto, Japan) was utilized. A simplified set including 12 markers was placed on both sides of the shoulder, pelvis, hip, knee, ankle, and 5th metatarsal head (Mukaino et al., 2016; Mukaino et al., 2018). The participants’ subjectively comfortable gait speed was determined based on a 10-m walk test.

Before measurement, the patients walked on the treadmill to get accustomed to treadmill gait for 2 min. After achieving a steady state, data were collected for 20 s and data for at least five complete gait cycles were collected from each subject. Videos were recorded at a sampling frequency of 60 Hz and measurement time of 20 s. Heel-strike and toe-off events were determined automatically by the system, and two experienced physical therapists checked the accuracy of the timing and adjusted if there was an error. The step length, stance, swing, and double stance time were recorded from these events. The double stance of the paretic side was defined as the double stance before paretic swing, whereas double stance of the nonparetic side was defined as that after the paretic swing.

Outcome measures and statistics

The step length, stance, swing, and double stance time were compared between the patients and controls. The values of controls were the averages for both the left and right sides. Asymmetries in spatial and temporal parameters were quantified using the raw and absolute values of symmetry index (SI) (Robinson et al., 1987), which was calculated as follows: raw value of SI (RSI) = (Vparetic − Vnonparetic)/0.5 (Vparetic + Vnonparetic) × 100%, where Vparetic is the value of a gait parameter recorded for the paretic leg of the patient or the left leg of the control, and Vnonparetic is the corresponding value for the nonparetic leg of the patient or right leg of the control. Absolute values of SI (ASI) were employed to evaluate amplitude of the asymmetry, which could vary in direction (Roerdink and Beek, 2011). To evaluate gait parameter variability, coefficient of variation (CV: SD/average) was used.

All statistical analyses were performed using SPSS version 18.0 for Windows (SPSS Inc., Chicago, Illinois, USA). The quantitative variables were tested using the single sample Kolmogorov–Smirnov test if the variables were normally distributed. Student’s paired t-test as the parametric test and the Wilcoxon signed-rank test as the nonparametric test were used for the comparison. Statistical significance was set at P < 0.05.

Ethics

This study was approved by the Research Ethics Board of Fujita Health University. All participants provided written informed consent.

Results

Demographic variables

Participants’ characteristics are presented in Table 1. The final study sample included 130 patients and 130 controls. No significant differences were found between the demographic characteristics (e.g. age, height, and velocity) of the stroke and control groups, except for sex, which presented significant differences between the 0.5 and 1.4 and 1.5–2.4 km/h groups.

Table 1
Table 1:
Demographic and clinical measures of stroke patients and healthy controls

Stride length, step length, and cadence

The stride length, cadence, and step length on the paretic and nonparetic sides of stroke patients are presented in Table 2. Overall, there was no significant difference between stroke patients and controls in stride length, cadence, and step lengths (effect size: 0.09, 0.34, and 0.07, respectively). The stratified comparison with gait speed revealed a significant difference between stroke patients and controls at 0.5–2.4 km/h; stride length and step length were significantly longer (0.5–1.4 km/h) and the cadence was significantly lower in the hemiparetic group than in controls (0.5–2.4 km/h). Although there was no significant difference observed between stroke patients and controls in RSI of step length (effect size: 0.34), ASI of step length was significantly higher and the effect size was high (0.98). The significantly high ASI was also observed in the stratified comparison, except in the 3.5–4.4 km/h group. Given the previous reports showing that handrail use affects stride length (Abe et al., 2009; IJmker et al., 2015), comparison of stride and cadence between patients and matched controls without the handrail was also performed (Table 3) and showed no difference was found in the handrail-free condition.

Table 2
Table 2:
Step length, stride length, and cadence of stroke patients and healthy controls
Table 3
Table 3:
Stride length and cadence of stroke patients walking without handrail and healthy controls

Temporal gait parameters

The temporal parameters in all patients and matched controls are shown in Table 4. Overall, stance time for the paretic and nonparetic sides of patients were significantly longer (effect size: 0.24 and 0.53, respectively); the stratified comparison revealed significantly longer paretic stance time at 0.5–1.4 km/h and nonparetic stance time at 1.5–3.4 km/h. The difference in RSI and ASI of stance time between stroke patients and controls was significant (effect size: 1.16 and 1.29, respectively) at 0.5–3.4 km/h.

Table 4
Table 4:
Temporal parameters of stroke patients and healthy controls

The paretic swing time was significantly longer (effect size: 0.61) and nonparetic swing time was shorter (effect size: 0.57) than those controls. Both RSI and ASI were higher in stroke patients (effect size: 1.20 and 1.28, respectively). In the stratified comparison, paretic swing lengthening was significant at 0.5–3.4 km/h, whereas nonparetic swing time shortening was observed at 0.5–2.4 km/h. RSI was significantly higher in stroke patients at 0.5–3.4 km/h, and ASI was significantly higher in all gait speed groups.

The double stance time of the paretic and nonparetic sides was significantly longer in hemiparetic patients than in controls (effect size: 0.43 and 0.36, respectively). ASI was significantly higher in stroke patients (effect size: 0.69); however, there was no significant difference in RSI (effect size: 0.10). The longer paretic and nonparetic double stance time and higher RSI were observed at 0.5–2.4 km/h. ASI was significantly high in gait speed <3.4 km/h.

Variability

CVs of step length and temporal parameters are presented in Table 5. CVs of step length, stance time, and swing time of the paretic and nonparetic sides were larger in stroke patients [effect size (paretic/nonparetic): step length 0.59/0.56, stance time 0.72/0.63, swing time 0.72/0.79]. CVs of stance time were larger in patients in the paretic side at gait speeds <3.4 km/h. CVs of swing time for both sides at speeds <3.4 km/h were larger in patients than in controls. CVs of swing time of the paretic side at 3.5–4.4 km/h were larger in patients than in controls. There was no significant difference in CVs of paretic double support time between controls and stroke patients.

Table 5
Table 5:
Variability of spatiotemporal parameters of stroke patients and matched controls

Discussion

Our study revealed differences in the spatiotemporal parameters between stroke patients and speed-, age-, and height-matched controls with its gradation in different gait speed groups. The longer stride length and step length, and lower cadence were evident in individuals with a very low-gait speed (<1.4 km/h or <2.4 km/h). This might relate to the high rate of handrail use in patients. In the previous studies, handrail use was shown to lengthen the stride length (Abe et al., 2009; IJmker et al., 2015). In this study, a large number of the patients in the low-gait speed group (<2.4 km/h) used a handrail. Consistently, no significant differences between controls and patients walking without a handrail were observed (Table 2).

Although the averaged step length and RSI of stroke patients were similar to the control groups, ASI of step length was significantly high in the stroke patients, indicating that the direction of the step length asymmetry varied in all gait speed groups. Previous studies have indicated that the step length asymmetry would be determined by the ability for propulsive force generation(Balasubramanian et al., 2007) and balance with swing capacity or compensatory strategy (Roerdink and Beek, 2011; Allen et al., 2011). Thus, the present results may reflect the variety in gait ability and compensatory strategy among the stroke patients.

Moreover, the changes in temporal parameters were observed in low-gait speed groups, which may reflect the compensatory response of patients when walking: instability on the paretic limb could cause compensatory shortening of paretic stance time, as this is considered to reflect balance ability (Patterson et al., 2008); leg stiffness due to the impaired paretic limb causes compensatory prolonged swing time (Nadeau et al., 1999).

This typical temporal pattern of gait abnormality in hemiparetic patients with prolonged paretic swing time and shortened nonparetic swing reflects paretic limb function impairment (Brandstater et al., 1983). Thus, swing time symmetry that strongly correlates stance time symmetry (Lauziere et al., 2014) has been used as the representative temporal parameter to describe post-stroke gait. In the present study, both prolonged paretic swing and shortened paretic single stance, and subsequent increase in swing asymmetry was observed. Although the prolonged paretic swing and double stance were less evident in the without handrail condition, the increase in RSI and ASI was still evident.

Interestingly, the abnormal increase in ASI of swing time was also seen in the high-gait speed groups who presented no significant increase in RSI, indicating that the extent of asymmetry was increased similar to that of the low-gait speed groups, but the direction was varied, suggesting patients with asymmetry in nontypical direction were included. Previous studies have shown the strong relationship between the extent of swing time asymmetry and balance ability (Lewek et al., 2014; Hendrickson et al., 2014). The asymmetry in nontypical direction observed in high-gait speed groups may relate more to the balance ability of patients than to hemiparesis.

Variability of gait pattern indices is considered to reflect the instability of gait (Maki, 1997), which is influenced by various impairments. For example, poorer strength, balance, and processing speed are reported to be associated with greater stance time variability (Hausdorff et al., 2001a, 2001b; Brach et al., 2008; Lamoth et al., 2011), poorer strength, and processing speed with greater step length variability (Kang and Dingwell, 2008; Brach et al., 2008). Considering these relationships with various impairments and the fact that the abnormality was seen in all gait speeds in this study, the variability indices should be sensitive indices for gait abnormality.

Additionally, the asymmetry and variability in gait indices were shown to be related to fall risk (Kressig et al., 2008; Verghese et al., 2009; Parker et al., 2013); thus it could also be used concurrently for risk management. As these indices are easily acquired in outpatient clinics using clinical measurement tools, including simple systems such as accelerometer systems or carpet-type walkway systems, it is feasible to measure these indices in daily clinical settings for gait disorder assessment.

There are several limitations in this study. The sample includes patients who used a handrail during the assessments. As discussed previously, the lengthening of stride in low-gait speed patients was due to the high handrail usage rate. Handrail use has also been shown to increase nonparetic swing time and improve swing time asymmetry (Chen et al., 2005a). However, the nonparetic swing time was significantly shorter in low-gait speed groups despite the high handrail use rate. Additionally, the temporal asymmetry and CVs were also significantly high (<0.05) in stroke patients without handrail use (data not shown). Thus, the overall tendency seen in this study seemed to be robust. Another limitation is the significant difference in sex ratio between patients and controls, which might have affected the results. However, the influence is expected to be significantly reduced by the height matching.

In conclusion, our data showed the changes in spatiotemporal pattern of hemiparetic gait with the gradation of different walking speeds, in comparison with the matched control data, which may serve as a reference to evaluate gait abnormality. The asymmetry and variability indices presented as sensitive indicators of gait abnormality, which may also serve as fall risk indicators. Further investigation into the underlying mechanisms and detailed relationships to fall risks may facilitate the utility of spatiotemporal parameters for daily practices.

Acknowledgements

This study was supported by the Fujita Health University fund (grant number 2015100341).

Conflicts of interest

There are no conflicts of interest.

References

Abe H, Michimata A, Sugawara K, Sugaya N, Izumi S. Improving gait stability in stroke hemiplegic patients with a plastic ankle-foot orthosis. Tohoku J Exp Med. 2009; 218:193–199
Allen JL, Kautz SA, Neptune RR. Step length asymmetry is representative of compensatory mechanisms used in post-stroke hemiparetic walking. Gait Posture. 2011; 33:538–543
Balaban B, Tok F. Gait disturbances in patients with stroke. PM R. 2014; 6:635–642
Balasubramanian CK, Bowden MG, Neptune RR, Kautz SA. Relationship between step length asymmetry and walking performance in subjects with chronic hemiparesis. Arch Phys Med Rehabil. 2007; 88:43–49
Brach JS, Studenski S, Perera S, VanSwearingen JM, Newman AB. Stance time and step width variability have unique contributing impairments in older persons. Gait Posture. 2008; 27:431–439
Brandstater ME, de Bruin H, Gowland C, Clark BM. Hemiplegic gait: analysis of temporal variables. Arch Phys Med Rehabil. 1983; 64:583–587
Chen G, Patten C, Kothari DH, Zajac FE. Gait deviations associated with post-stroke hemiparesis: improvement during treadmill walking using weight support, speed, support stiffness, and handrail hold. Gait Posture. 2005a; 22:57–62
Chen G, Patten C, Kothari DH, Zajac FE. Gait differences between individuals with post-stroke hemiparesis and non-disabled controls at matched speeds. Gait Posture. 2005b; 22:51–56
Chino N, Sonoda S, Domen K, Saitoh E, Kimura A. Stroke Impairment Assessment Set (SIAS). A new evaluation instrument for stroke patients. Jpn J Rehabil Med. 1994; 31:119–125
Hausdorff JM, Nelson ME, Kaliton D, Layne JE, Bernstein MJ, Nuernberger A, Singh MA. Etiology and modification of gait instability in older adults: a randomized controlled trial of exercise. J Appl Physiol (1985). 2001a; 90:2117–2129
Hausdorff JM, Rios DA, Edelberg HK. Gait variability and fall risk in community-living older adults: a 1-year prospective study. Arch Phys Med Rehabil. 2001b; 82:1050–1056
Hendrickson J, Patterson KK, Inness EL, McIlroy WE, Mansfield A. Relationship between asymmetry of quiet standing balance control and walking post-stroke. Gait Posture. 2014; 39:177–181
Hishikawa N, Tanikawa H, Ohtsuka K, Mukaino M, Inagaki K, Matsuda F, et al. Quantitative assessment of knee extensor thrust, flexed-knee gait, insufficient knee flexion during the swing phase, and medial whip in hemiplegia using three-dimensional treadmill gait analysis. Top Stroke Rehabil. 2018; 25:548–553
IJmker T, Lamoth C J, Houdijk H, Tolsma M, van der Woude LH, Daffertshofer A, Beek PJ. Effects of handrail hold and light touch on energetics, step parameters, and neuromuscular activity during walking after stroke. J Neuroeng Rehabil. 2015; 12:70
Itoh N, Kagaya H, Saitoh E, Ohtsuka K, Yamada J, Tanikawa H, et al. Quantitative assessment of circumduction, hip hiking, and forefoot contact gait using Lissajous figures. Jpn J Compr Rehabil Sci. 2012; 3:78–84
Kang HG, Dingwell JB. Separating the effects of age and walking speed on gait variability. Gait Posture. 2008; 27:572–577
Kressig RW, Herrmann FR, Grandjean R, Michel JP, Beauchet O. Gait variability while dual-tasking: fall predictor in older inpatients?. Aging Clin Exp Res. 2008; 20:123–130
Lamoth CJ, van Deudekom FJ, van Campen JP, Appels BA, de Vries OJ, Pijnappels M. Gait stability and variability measures show effects of impaired cognition and dual tasking in frail people. J Neuroeng Rehabil. 2011; 8:2
Lauziere S, Betschart M, Aissaoui R, Nadeau S. Understanding spatial and temporal gait asymmetries in individuals post stroke. Int J Phys Med Rehabil. 2014; 2:201
Lewek MD, Bradley CE, Wutzke CJ, Zinder SM. The relationship between spatiotemporal gait asymmetry and balance in individuals with chronic stroke. J Appl Biomech. 2014; 30:31–36
Maki BE. Gait changes in older adults: predictors of falls or indicators of fear. J Am Geriatr Soc. 1997; 45:313–320
Mukaino M, Ohtsuka K, Tanikawa H, Matsuda F, Yamada J, Itoh N, et al. Clinical-oriented three-dimensional gait analysis method for evaluating gait disorder. J Vis Exp. 2018; 133:57063
Mukaino M, Ohtsuka K, Tsuchiyama K, Matsuda F, Inagaki K, Yamada J, et al. Feasibility of a simplified, clinically oriented, three-dimensional gait analysis system for the gait evaluation of stroke patients. Prog Rehabil Med. 2016; 1:20160001
Nadeau S, Arsenault AB, Gravel D, Bourbonnais D. Analysis of the clinical factors determining natural and maximal gait speeds in adults with a stroke. Am J Phys Med Rehabil. 1999; 78:123–130
Nadeau S, Betschart M, Bethoux F. Gait analysis for poststroke rehabilitation: the relevance of biomechanical analysis and the impact of gait speed. Phys Med Rehabil Clin N Am. 2013; 24:265–276
Parker K, Hanada E, Adderson J. Gait variability and regularity of people with transtibial amputations. Gait Posture. 2013; 37:269–273
Patterson KK, Gage WH, Brooks D, Black SE, McIlroy WE. Changes in gait symmetry and velocity after stroke: a cross-sectional study from weeks to years after stroke. Neurorehabil Neural Repair. 2010; 24:783–790
Patterson KK, Parafianowicz I, Danells CJ, Closson V, Verrier MC, Staines WR, et al. Gait asymmetry in community-ambulating stroke survivors. Arch Phys Med Rehabil. 2008; 89:304–310
Rinaldi LA, Monaco V. Spatio-temporal parameters and intralimb coordination patterns describing hemiparetic locomotion at controlled speed. J Neuroeng Rehabil. 2013; 10:53
Robinson RO, Herzog W, Nigg BM. Use of force platform variables to quantify the effects of chiropractic manipulation on gait symmetry. J Manipulative Physiol Ther. 1987; 10:172–176
Roerdink M, Beek PJ. Understanding inconsistent step-length asymmetries across hemiplegic stroke patients: impairments and compensatory gait. Neurorehabil Neural Repair. 2011; 25:253–258
Roth EJ, Merbitz C, Mroczek K, Dugan SA, Suh WW. Hemiplegic gait. Relationships between walking speed and other temporal parameters. Am J Phys Med Rehabil. 1997; 76:128–133
Tanikawa H, Ohtsuka K, Mukaino M, Inagaki K, Matsuda F, Teranishi T, et al. Quantitative assessment of retropulsion of the hip, excessive hip external rotation, and excessive lateral shift of the trunk over the unaffected side in hemiplegia using three-dimensional treadmill gait analysis. Top Stroke Rehabil. 2016; 23:311–317
Titianova EB, Peurala SH, Pitkänen K, Tarkka IM. Gait reveals bilateral adaptation of motor control in patients with chronic unilateral stroke. Aging Clin Exp Res. 2008; 20:131–138
Titianova EB, Tarkka IM. Asymmetry in walking performance and postural sway in patients with chronic unilateral cerebral infarction. J Rehabil Res Dev. 1995; 32:236–244
Verghese J, Holtzer R, Lipton RB, Wang C. Quantitative gait markers and incident fall risk in older adults. J Gerontol A Biol Sci Med Sci. 2009; 64:896–901
Wonsetler EC, Bowden MG. A systematic review of mechanisms of gait speed change post-stroke. Part 1: spatiotemporal parameters and asymmetry ratios. Top Stroke Rehabil. 2017; 24:435–446
Keywords:

gait analysis; hemiparesis; stroke

Copyright © 2020 Wolters Kluwer Health, Inc. All rights reserved.