- Email: [email protected]
Bilateral coordination and gait symmetry after body-weight supported treadmill training for persons with chronic stroke
Clinical Biomechanics 28 (2013) 448–453
Contents lists available at SciVerse ScienceDirect
Clinical Biomechanics journal homepage: www.elsevier.com/locate/clinbiomech
Bilateral coordination and gait symmetry after body-weight supported treadmill training for persons with chronic stroke Stephanie A. Combs a, b,⁎, Eric L. Dugan c, Elicia N. Ozimek d, Amy B. Curtis b a
Krannert School of Physical Therapy, University of Indianapolis, 1400 East Hanna Avenue, Indianapolis, IN 46227, USA Interdisciplinary Health Sciences Program, Western Michigan University, Kalamazoo, MI 49008, USA Department of Kinesiology, Center for Orthopaedic and Biomechanics Research, Boise State University, Boise, ID 83725, USA d Biomechanics Laboratory, School of Physical Education, Sport and Exercise Science, Ball State University, Muncie, IN 47306, USA b c
a r t i c l e
i n f o
Article history: Received 22 July 2012 Accepted 5 February 2013 Keywords: Stroke Rehabilitation Gait symmetry Coordination
a b s t r a c t Background: Locomotor interventions are commonly assessed using functional outcomes, but these outcomes provide limited information about changes toward recovery or compensatory mechanisms. The study purposes were to examine changes in gait symmetry and bilateral coordination following body-weight supported treadmill training in individuals with chronic hemiparesis due to stroke and to compare findings to participants without disability. Methods: Nineteen participants with stroke (>6 months) who ambulated between 0.4 and 0.8 m/s and 22 participants without disability were enrolled in this repeated-measures study. The stroke group completed 24 intervention sessions over 8 weeks with 20 minutes of walking/session. The non-disabled group served as a comparison for describing changes in symmetry and coordination. Bilateral 3-dimensional motion analysis and gait speed were assessed across 3 time points (pre-test, immediate post-test, and 6-month retention). Continuous relative phase was used to evaluate bilateral coordination (thigh–thigh, shank–shank, foot–foot) and gait symmetry was assessed with spatiotemporal ratios (step length, swing time, stance time). Findings: Significant improvements in continuous relative phase (shank–shank and foot–foot couplings) were found at post-test and retention for the stroke group. Significant differences in spatiotemporal symmetry ratios were not found over time. Compared to the non-disabled group, changes in bilateral coordination moved in the direction of normal recovery. Most measures of continuous relative phase were more responsive to change after training than the spatiotemporal ratios. Interpretations: After body-weight supported treadmill training, the stroke group made improvements toward recovery of normal bilateral coordination. Bilateral coordination and gait symmetry measures may assess different aspects of gait. © 2013 Elsevier Ltd. All rights reserved.
1. Introduction Distinguishing between intervention induced changes that represent recovery or reinforced compensatory mechanisms is a challenge when evaluating the efficacy of rehabilitation protocols for persons following stroke. In the context of gait, recovery refers to the restitution of pre-stroke gait patterns while compensation refers the use of different gait patterns to accomplish the task (Levin et al., 2009). Metrics such as gait velocity are often used to assess changes in gait function; however, gait velocity does not provide insight into underlying impairments or quality of gait performance (Olney et al., 1994;
⁎ Corresponding author at: Krannert School of Physical Therapy, University of Indianapolis, 1400 East Hanna Avenue, Indianapolis, IN 46227, USA. E-mail addresses: [email protected] (S.A. Combs), [email protected] (E.L. Dugan), [email protected] (E.N. Ozimek), [email protected] (A.B. Curtis). 0268-0033/$ – see front matter © 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.clinbiomech.2013.02.001
Patterson et al., 2008). This led Patterson et al. (2008) to suggest that gait symmetry should be a focus of gait function assessments (with gait velocity as an adjunct) due to its potential to better describe impairments in post-stroke gait. However, spatiotemporal symmetry measures have been reported to be inadequate for distinguishing asymmetry in persons with stroke from persons without known pathology (Hodt-Billington et al., 2008). Therefore, in order to discriminate between recovery of pre-stroke gait function and adaptation of compensatory gait patterns, additional metrics are necessary. Recent work suggests that gait symmetry is only one aspect of the bilateral function of the limbs during gait (Meijer et al., 2011; Plotnik and Hausdorff, 2008). As described by Plotnik and Hausdorff (2008), gait symmetry and bilateral coordination may be separate aspects of the bilateral function of the limbs during gait. While gait symmetry (as measured by swing times of the right and left legs) and bilateral coordination (as measured by phase coordination index which takes into account accuracy and consistency of stepping) are not correlated in healthy adults, they have been reported to be strongly correlated in
S.A. Combs et al. / Clinical Biomechanics 28 (2013) 448–453
individuals with stroke (Meijer et al., 2011). The reasons for this strong association between gait symmetry and bilateral coordination in patients with stroke are not fully understood. Meijer et al. (2011) suggest that the relationship between gait symmetry and bilateral coordination may be due to the asymmetric nature of motor and sensory impairments after stroke. The differing relationships between spatiotemporal asymmetry and bilateral limb coordination in healthy adults as compared to individuals with stroke support the need to incorporate both types of measures when evaluating rehabilitation protocols. Gait velocity, gait symmetry, and bilateral coordination provide different (and perhaps in some cases overlapping) insights regarding gait function. These differences are important in the context of evaluating the efficacy of rehabilitation protocols. In particular, it is important to develop a better understanding of the mechanisms for how interventions lead to recovery of and/or compensatory adaptations in the gait pattern. To this end, the overall aims of the current study were to examine changes in gait velocity, gait symmetry, and bilateral coordination following body weight supported treadmill training (BWSTT) in persons with hemiparesis due to stroke and to compare these findings to participants without disability. BWSTT is a means of locomotor training that can increase overground gait speed after stroke (Sullivan et al., 2002, 2007). As gait symmetry deteriorates over time in persons following stroke (Patterson et al., 2010a), reported improvements in stance and swing time symmetry during BWSTT (Chen et al., 2005a; Hassid et al., 1997; Hesse et al., 1999) suggest that the repetitive and task-oriented nature of the intervention may also promote carryover of improved gait symmetry to overground environments. We hypothesized that BWSTT would lead to immediate and long-term changes in gait velocity, gait symmetry, and bilateral coordination for the group of participants with chronic hemiparesis due to the repetitive, task-specific nature of BWSTT, and that these changes would trend toward the values found by a group of nondisabled adults. Measures of bilateral coordination are novel concepts for quantifying changes in gait function in persons with stroke; therefore, we compared their responsiveness to change after BWSTT to more common spatiotemporal symmetry ratios.
449
Additional inclusion criteria for the stroke group were: 1) at least 6-months post stroke; 2) had only experienced one stroke; 3) ambulatory with or without the use of an assistive device or orthosis; 4) able to ambulate up to 14 m without the use of an assistive device with supervision; 5) able to ambulate at a self-selected gait velocity between 0.4 m/s and 0.8 m/s (Perry et al., 1995); 6) currently not receiving physical therapy services or lower extremity botulinum toxin injections; 7) medically stable with a physician release stating approval to enter an exercise program; 8) able to follow at least two-step verbal instructions. Non-disabled participants were additionally screened according to the following inclusion criteria: 1) able to walk independently during home and community activities; 2) no known neurological condition or deficits, such as prior stroke. The protocol was approved by Institutional Review Boards at participating institutions, and participants provided informed consent prior to initiating the study. Eligibility determination during the initial visit for participants with stroke included two comfortable 10-meter walk tests (CWT) to verify a mean gait velocity between 0.4 and 0.8 m/s. 2.3. Intervention Participants in the stroke group were treated for 24 sessions of BWSTT over 8 weeks with 20 minutes of total walking each session (Sullivan et al., 2002, 2007). Rest breaks were allowed if requested, but were not included in the overall walking time. During BWSTT verbal and manual assistance were provided by 1–2 researchers (SC and trained student physical therapists) to facilitate an optimal gait pattern primarily through weight shifting at the pelvis and/or paretic limb movement during the gait cycle. Participants did not wear an ankle foot orthosis or use handrails during training. Treadmill speed was increased by 0.2 mph increments each subsequent training session until age and gender norms were achieved and body weight support was decreased from 30% over the course of the intervention according to a standardized protocol. The progression parameters of the BWSTT protocol have been previously described (Combs et al., 2010). 2.4. Data acquisition
2. Methods 2.1. Design A prospective repeated measures design was used. Participants with chronic hemiparesis due to stroke (stroke group) were evaluated three times: one-week before and after the intervention period and sixmonths following completion of the intervention (pre-test, post-test and, retention). Participants without disability (non-disabled group) were evaluated three times during the same timeframes as the stroke group, but did not participate in the intervention. Measurements were conducted by members of the research team who were not involved with the intervention. 2.2. Participants The convenience sample of 41 participants included 19 participants with chronic hemiparesis due to stroke and 22 participants without disability. Participants with stroke were recruited from local stroke support groups and clinicians. Participants without disability were recruited from local university communities. All participants were screened according to the following inclusion criteria: 1) between the ages of 40 and 80 years; 2) no current musculoskeletal conditions or recent orthopedic surgeries (outside of typical age-related changes) within 6 months of the study; 3) no complications from other health conditions that could influence walking; 4) able to travel to and from research sessions.
All participants were evaluated using the CWT and bilateral three-dimensional motion analysis. Participants with stroke did not use assistive devices during data collection; however, a research assistant closely monitored these participants by walking next to them, if necessary. All participants wore shoes during data collection and two participants in the stroke group required the use of their ankle foot orthosis on the paretic leg for improved safety during the walking trials. The CWT is a valid and reliable (ICC = 0.94) measure of shortdistance, self-selected overground gait velocity in persons with chronic stroke (Flansbjer et al., 2005). It was conducted in an open room along a 14-meter walkway. Participants were timed for the middle 10 m using electronic timing gates (Fitness Technologies, 21 Bishop Street, Adelaide-Skye, Australia, 5072). The mean of three trials was used for data analysis (Flansbjer et al., 2005). Bilateral three-dimensional motion data were captured with a 10-camera Vicon MX motion capture system (Vicon, 5419 McConnel Avenue, Los Angeles, CA, USA 90066) (120 Hz). Participants wore 39 retro-reflective markers attached to specific anatomical landmarks based on Vicon's Plug-in-Gait marker set. Participants in the stroke group walked along a 6-meter walkway with embedded force plates (Model BP600900-6-1000, Advanced Medical Technology, Inc., Watertown, MA, USA) at a self-selected comfortable pace for two practice trials, followed by 10 trials for data collection. Participants in the non-disabled group were instructed to walk along the 6-meter walkway with embedded force plates at a slow walking speed that was matched to the average comfortable walking speed of the stroke
450
S.A. Combs et al. / Clinical Biomechanics 28 (2013) 448–453
group established at pre-test with the CWT. Non-disabled group participants practiced walking at the slow pace for two trials using a metronome to assist in establishing the slower pace, followed by 10 trials without the metronome for data analysis. 2.5. Data reduction The first three walking trials per participant that included a clean force plate strike and all markers present were averaged and used for subsequent data analysis. A clean force plate strike was defined as a trial where the entire foot, and no part of the contra-lateral foot, made contact with the force plate. Force plate data were used to identify foot-strike and toe-off. All kinematic data were filtered using a 2nd order low-pass Butterworth filter with a cutoff frequency of 6 Hz. The kinematic data were normalized to 100% of the gait cycle, defined as foot strike to foot strike of the same limb. Spatiotemporal variables including step length, stance time, and swing time were calculated for the paretic and non-paretic legs of the stroke group and for the right and left legs of the non-disabled group. Gait symmetry was evaluated with spatiotemporal symmetry ratios defined as the ratio of paretic to nonparetic (Patterson et al., 2010b) values for each variable for the stroke group and the ratio of right to left values for the non-disabled group. Bilateral coordination was examined through continuous relative phase (CRP) analysis. First, mean sagittal plane segmental angular positions and velocities were used to construct normalized phase plots for the thigh, shank, and foot segments from each participant (Barela et al., 2000; Peters et al., 2003). Phase angles were then calculated as the angle between the right horizontal and the line generated from the origin to each point on the phase plot. The phase angles were calculated at 1% intervals of the gait cycle. CRP for the interlimb couplings (thigh–thigh, shank–shank, and foot–foot) was defined as the difference between the normalized phase angles of contralateral segments, right minus left for the control group and non-paretic minus paretic for the stroke group (Haddad et al., 2006). The mean CRP for each interlimb coupling was displayed as a continuous curve averaged at each 1% interval of the complete gait cycle. A CRP curve near 0° represents an in-phase relationship between the two limb segments (Stergiou et al., 2001). CRP values that deviate from 0° indicate a more out-of-phase relationship between the two segment motions. The CRP curves were also represented as a single value by averaging the mean absolute CRP (MARP) over each 1% interval of the gait cycle (Kwakkel and Wagenaar, 2002; Stergiou et al., 2001). 2.6. Statistical analysis A threshold of at least 80% completion of intervention sessions (at least 20 out of 24 sessions) was required for participant outcomes to be included in the final data analysis. Data analysis was performed using Statistical Package for the Social Sciences version 17.0 (IBM SPSS, 1 New Orchard Road, Armonk, NY, USA 10504). Demographic characteristics and dependent variables at each measurement time point (pre-test, post-test and, retention) were examined for normality (Shapiro–Wilk, P > .05). Demographics were compared between groups by independent t-test (age) and Fisher's exact test (gender). A one-way repeated measures analysis of variance was used to examine within-group differences for all dependent measures over the three measurement periods. Statistically significant differences were determined using a 95% confidence interval adjusted for repeated measures. Post hoc analysis for the CWT and spatiotemporal symmetry ratios were conducted using the Bonferroni correction on eligible pairwise comparisons. Post hoc analyses for MARP were conducted using confidence bands constructed around the mean differences between CRP curves from pre- to post-test and pre-test to retention to determine whether
differences existed using a non-parametric bootstrap procedure (Duhamel et al., 2004; Lenhoff et al., 1999). Use of confidence bands has been recommended as a preferred test for evaluating differences between mean continuous gait curves (Duhamel et al., 2004; Lenhoff et al., 1999) and is a complementary follow-up procedure to the repeated measures analysis of variance (Duhamel et al., 2004). The confidence bands encompassed the region of the true mean difference between CRP curves at 95% probability. The customized bootstrapping procedure, developed with MATLAB (MathWorks, 3 Apple Hill Drive, Natick, MA, USA 01760), involved simultaneous re-sampling of the CRP curves at each measurement period (L = 500) in order to estimate the variability present in the population. Points where the confidence bands did not contain 0 (horizontal axis) indicated a statistically significant difference between the two CRP curves and therefore a significant change in coordination at that point. The number of points that did not contain 0 was totaled to reveal the overall percentage of the gait cycle that showed a significant difference in interlimb coordination. This analysis allowed for a description of how much of the gait cycle changed as well as where in the gait cycle the changes occurred. Responsiveness to change (Stratford et al., 2002) of the MARP and spatiotemporal symmetry variables were assessed with effect sizes (ES) and standardized response means (SRM). While responsiveness to change does not necessarily determine the meaningfulness of the change to the patient, this analysis is important as it allows for interpretation of how well the variable measures change over time (Stratford et al., 2002). ES was calculated using the difference in mean scores from pre- to post-test and from pre-test to retention divided by the respective pooled standard deviations (Dunlap et al., 1996). SRM was calculated using the difference in mean scores from pre- to post-test and from pre-test to retention divided by the respective standard deviation of the change scores. Responsiveness based on the ES and SRM of the measures was classified as small (d = 0.2), medium (d = 0.5), or large (d = 0.8) (Cohen, 1992). 3. Results Three participants in the stroke group did not complete the study due to changes in health status unrelated to the study. Motion analysis data from one participant in the stroke group and 3 participants in the non-disabled group were not used due to data collection error. We were unable to locate one participant from the non-disabled group at the 6-month testing session. These participants were subsequently excluded from the analysis. The fifteen participants in the stroke group who were included in the final data analysis all completed more than 80% of the intervention sessions. No statistically significant differences were found between groups for age and gender (P > .05). See Table 1 for group demographics. There were also no significant differences over time for the spatiotemporal symmetry ratios and MARP variables for the non-disabled group (P > .05). (Table 2) However, there was a statistically significant main effect difference (P = .02) in slow gait velocity for the non-disabled group with pretest gait speed significantly differing from post-test (P = .03) gait speed (Table 2). For the stroke group, statistically significant main effect differences were found over time for the CWT and MARP (shank/shank and foot/
Table 1 Participant demographics. Groups
Years of age, mean(SD) Gender, male/female Years post stroke, mean (SD) Side of Paresis, right/left
P
Stroke n = 15
Non-disabled n = 18
59.9 (11.2) 4/11 3.8 (3.2) 6/9
57.4 (7.6) 8/10 n/a n/a
P ≤ 0.05; independent t-test used for age; Fisher's exact test used for gender.
.45 .47
S.A. Combs et al. / Clinical Biomechanics 28 (2013) 448–453
451
Table 2 Mean gait velocity, spatiotemporal symmetry ratios and MARP over time. Variables
Stroke group (n = 15)
Non-disabled group (n = 18)
Mean (SD) Pre-test Gait velocity CWT (m/s) Spatiotemp. ratios Step Length Sym Stance time sym Swing time sym MARP Thigh–thigh (°) Shank–shank (°) Foot–foot (°)
Post-test
Retention
Main effects
Mean (SD)
F
P
Pre-test
Main effects Post-test
Retention
F
P
0.63 (0.15)
0.75 (0.19)a
0.80 (0.25)b
11.41
.00
0.73 (0.07)
0.68 (0.05)c
0.69 (0.06)
4.32
.02
1.27 (0.26) 0.90 (0.07) 1.32 (0.24)
1.26 (0.44) 0.89 (0.07) 1.24 (0.15)
1.34 (0.53) 0.89 (0.06) 1.30 (0.19)
0.82 0.69 1.15
.45 .46 .32
1.01 (0.05) 1.01 (0.05) 1.01 (0.08)
1.00 (0.06) 1.00 (0.04) 1.00 (0.07)
0.99 (0.06) 1.02 (0.04) 1.00 (0.10)
1.31 1.24 0.02
.28 .30 .98
0.74 25.87 21.20
.49 .00 .00
−0.60 (6.03) 0.72 (7.12) 2.04 (6.52)
0.39 (6.89) 0.58 (7.94) 0.35 (5.91)
−0.01 (5.82) −0.93 (8.47) −0.41 (5.88)
0.21 0.29 0.72
.82 .75 .49
−11.45 (15.91) −43.72 (21.07) −52.77 (35.60)
−9.37 (19.49) −24.18 (23.79)a −26.31 (26.07)a
−7.66 (14.62) −29.06 (25.28)b −28.97 (35.51)b
SD = standard deviation, CWT = 10-meter comfortable walk test, spatiotemporal ratios = paretic/nonparetic variables, sym = symmetry, MARP = mean absolute continuous relative phase. a Indicates significant difference at b.01 from pre-test to post-test. b Indicates significant difference at b.01 from pre-test to retention. c Indicates significant difference at b.05 from pre-test to post-test.
The primary purposes of this study were to examine changes in gait symmetry and bilateral coordination following BWSTT in persons with chronic hemiparesis due to stroke and to compare the findings to persons without disability in order to identify direction of change toward recovery or compensation mechanisms. Although no significant differences were found over time in spatiotemporal symmetry, bilateral coordination based on CRP analysis significantly improved toward more in-phase relationships between the paretic and nonparetic limbs immediately following BWSTT partially supporting our hypothesis. Changes in bilateral coordination were still apparent 6-months after BWSTT; however, only a portion of the improvements in coordination were maintained by the stroke group. While the BWSTT protocol employed in this study led to lasting changes in gait velocity, it may not have been sufficient to bring about permanent changes in bilateral coordination. Alternative application of BWSTT protocol parameters or a longer duration of training may be necessary to induce permanent changes in coordination along with any changes in symmetry of step length, stance time, and swing time. Determination of optimal intervention parameters for persons following stroke will be an important direction of future research. In the present study, a greater percentage of statistically significant differences in bilateral coordination were found during the latter part of the CRP curves signifying that most of the changes occurred during swing phase. Impaired swing initiation has been identified as
30
CRP Degrees
20 10 0 -10
1
11
21
31
41
51
61
71
81
91
101
-20 -30 -40 -50
Percent of Gait Cycle
B
10 0 -10
CRP Degrees
4. Discussion
A
1
11
21
31
41
51
61
71
81
91
81
91
101
-20 -30 -40 -50 -60 -70
Percent of Gait Cylce
-80
C 0 1
11
21
31
41
51
61
71
101
-20
CRP Degrees
foot) variables (P = .00). Post hoc analyses revealed statistically significant increases in self-selected gait velocity from pre- to post-test (P = .00) and pre-test to retention (P b .01) (Table 2). Main effects did not demonstrate statistically significant differences over time for the spatiotemporal symmetry ratios or MARP (thigh/thigh) variables for the stroke group (P > .05) (Table 2, Fig. 1A). CRP post-hoc comparisons using confidence bands indicated statistically significant differences (P b .05) for 33% of the gait cycle from pre- to post-test for the shank–shank interlimb CRP curves (Fig. 1B). At retention, 17% of this significant difference was maintained. Statistically significant differences (P b .05) of 55% of the gait cycle were found for the foot–foot interlimb CRP curves from pre- to post-test with 24% maintained at retention (Fig. 1C). Changes in interlimb CRP corresponded primarily to swing phase. Mean thigh–thigh, shank– shank, and foot–foot interlimb CRP curves for the non-disabled group are included in Fig. 1A–C respectively. Responsiveness to change assessed by ES and SRM of the MARP and spatiotemporal symmetry variables for the stroke group is presented in Table 3.
-40 -60 -80
-100 -120 -140
Percent of Gait Cycle
Fig. 1. Mean interlimb CRP curves across gait cycle. (A) Thigh–thigh; (B) shank–shank; (C) foot–foot. Stroke group mean CRP displayed with black lines: solid black = baseline; large dashes = post-test; small dashes = retention. Gray areas between curves indicate statistically significant difference from pre- to post-test; P ≤ .05; cross-hatched areas between curves (XXX) indicate statistically significant difference from pre-test to retention; P ≤ .05; non-disabled group mean CRP displayed with dotted gray line.
452
S.A. Combs et al. / Clinical Biomechanics 28 (2013) 448–453
Table 3 Responsiveness to change in the stroke group after BWSTT (n = 15). Variables
Spatiotemporal ratios Step length sym Stance time sym Swing time sym MARP Thigh–thigh Shank–shank Foot–foot
Pre-to post-test
Pre-test to retention
ES
SRM
ES
SRM
.07 .23 .42
.30 .20 .29
.04 .14 .09
.09 .20 .12
.12 .87 .85
.18 2.16 1.58
.25 .63 .67
.28 1.31 1.35
ES = effect size (Cohen's d); SRM= standardized response mean; sym = symmetry; MARP = mean absolute continuous relative phase. Responsiveness: small (d = 0.2), medium (d = 0.5), large (d = 0.8).
a major contributor to differences in gait kinematics of persons with stroke compared to those without disability (Chen et al., 2005b). Barela et al. (2000) also reported that the intralimb coordination of paretic segments was most different from adults without disability during the swing phase of the gait cycle. Given these consistent deviations from normal during swing after stroke, it stands to reason that a greater potential for change may be possible during swing as was found after the intervention in this study. Based on the training needs of our group of participants, manual facilitation was most often provided to the paretic leg during late stance and swing to improve active ankle dorsiflexion and passive knee flexion, induced by active hip flexion. While the progressively increased treadmill speed and reduction in body weight support cannot be ruled out as potential parameters that influenced bilateral coordination in our participants, the manual facilitation to achieve more normal stepping patterns may have played a role in the improved swing phase coordination after BWSTT. Future studies will need to separate intervention-based control parameters in order to determine their different effects on coordination. Examination of the shank–shank and foot–foot couplings revealed that the paretic leg remained ahead of the non-paretic in the phase trajectory based on the negative CRP values during all three measurement periods. However, at post-test and retention these values were significantly less negative meaning the limb segments were moving toward a more in-phase relationship with improved coordination. Although the stroke group did not achieve the level of bilateral coordination demonstrated by the group without disability during overground walking (Fig. 1), they showed the ability to change movement patterns in the direction of normal recovery (Levin et al., 2009) after BWSTT. Change primarily at the interlimb level (as opposed to intralimb level) has been demonstrated in previous work with healthy adults in response to additional load applied unilaterally during comfortable walking (Haddad et al., 2006). The results of the current study lend support to these findings and indicate that with or without neurological dysfunction, gait is adaptable at the interlimb level. Future exploration of changes in coordination at the intralimb level within the paretic and non-paretic limbs after BWSTT may expand our understanding of the system's capacity to modulate coordination patterns after stroke. Regardless of efforts to standardize the slow walking pace of the non-disabled group during testing, significant differences in gait velocities were found across time. However, the non-disabled participants maintained stable patterns of spatiotemporal symmetry and bilateral coordination over the testing periods even with differences in their slowed walking pace. This was not the case with the stroke group that also demonstrated significant differences in gait speeds as well as improved bilateral coordination. Speed has been proposed as a limitation to the use of CRP analysis in persons with stroke (Barela et al., 2000) with increased walking speed associated with changes in intralimb coordination during mid-swing of the non-paretic limb (Hutin et al., 2012). For the stroke group in this study, change from pre-test in gait speed was not associated with change in bilateral
coordination. Meijer et al. (2011) similarly reported that gait speed and measures of bilateral coordination are independent walking properties. Thus, the alteration in coordination after locomotor training in this study may be attributed to a training parameter within the BWSTT modality. Despite reports of enhanced spatiotemporal symmetry while walking on a treadmill with body weight support (Chen et al., 2005a; Hassid et al., 1997; Hesse et al., 1999), these training enhancements did not transfer to overground walking for this group with stroke. Comparatively, the MARP measures of bilateral coordination (shank– shank and foot–foot) were more responsive to change after BWSTT than the spatiotemporal symmetry ratios. Plotnik et al. (2008) reported that bilateral coordination may characterize the rhythmic properties between limbs during stepping which differs from spatiotemporal symmetry measures (i.e. swing time and step length) that more likely reflect properties of propulsion from each limb (Balasubramanian et al., 2007; Plotnik et al., 2008). Accordingly, the BWSTT intervention employed in this study may have induced greater bilateral coordination between limbs due to the rhythmic, continuous motion of the treadmill. As a collective variable (Barela et al., 2000; Stergiou et al., 2001) that takes into consideration both angular position and velocity across strides, CRP may be more responsive after interventions such as BWSTT due to a more specific representation of the organization of neuromuscular control processes (Stergiou et al., 2001). In contrast, the use of body weight support could have interfered with propulsion due to reduced loading through the lower extremities and in turn reduced the responsiveness to changes in spatiotemporal symmetry. This finding is consistent with our recent report of a lack of change in kinetic symmetry (i.e. relative propulsion of the paretic leg and joint work) in persons following stroke after BWSTT (Combs et al., 2012). Therefore, using multiple metrics to evaluate rehabilitation interventions may be important in order to address different characteristics of gait as well as variable effects of interventions. While CRP was more responsive to changes in bilateral coordination, the clinical meaningfulness of those changes is unknown and should to be investigated in future studies. The small sample size may limit the generalizability of the findings. Lack of a true control group for comparison of the BWSTT to an alternative intervention limits the interpretation of the results. Other possible limitations were that sagittal plane motions only were used and that two participants required an AFO during the data collection. Although the use of an AFO has been shown to have little effect on intralimb coordination (Barela et al., 2000) and kinetic symmetry (Combs et al., 2012), its contribution to interlimb coordination is unknown.
5. Conclusions Because evidence indicates that gait asymmetry worsens over time in patients following stroke (Patterson et al., 2010a), the findings of this study have direct clinical implications for rehabilitation. The improvements made by the stroke group toward recovery of normal bilateral coordination after BWSTT justifies the importance of continued rehabilitation for persons in the chronic stage of stroke. CRP analysis is a promising measure of bilateral coordination of gait and should be explored in future studies.
Acknowledgments This project was supported in part by the University of Indianapolis Research/Creative Endeavor's Grant program. The researchers of this study would like to acknowledge our participants; Jeff Frame, MS at the Biomechanics Lab at Ball State University; and graduates from the Krannert School of Physical Therapy, Miranda Passmore, PT, DPT, Cara Riesner, PT, DPT, Dana Whipker, PT, DPT, and Elizabeth Yingling PT, DPT.
S.A. Combs et al. / Clinical Biomechanics 28 (2013) 448–453
References Balasubramanian, C.K., Bowden, M.G., Neptune, R.R., Kautz, S.A., 2007. Relationship between step length asymmetry and walking performance in subjects with chronic hemiparesis. Arch. Phys. Med. Rehabil. 88, 43–49. Barela, J.A., Whitall, J., Black, P., Clark, J.E., 2000. An examination of constraints affecting the intralimb coordination of hemiparetic gait. Hum. Mov. Sci. 19, 251–273. Chen, G., Patten, C., Kothari, D.H., Zajac, F.E., 2005a. Gait deviations associated with post-stroke hemiparesis: improvement during treadmill walking using weight support, speed, support stiffness, and handrail hold. Gait Posture 22, 57–62. Chen, G., Patten, C., Kothari, D.H., Zajac, F.E., 2005b. Gait differences between individuals with post-stroke hemiparesis and non-disabled controls at matched speeds. Gait Posture 22, 51–56. Cohen, J., 1992. A power primer. Psychol. Bull. 112, 155–159. Combs, S.A., Dugan, E.L., Passmore, M., Riesner, C., Whipker, D., Yingling, E., et al., 2010. Balance, balance confidence, and health-related quality of life in persons with chronic stroke after body weight-supported treadmill training. Arch. Phys. Med. Rehabil. 91, 1914–1919. Combs, S.A., Dugan, E.L., Ozimek, E.N., Curtis, A.B., 2012. Effects of body weight supported treadmill training on kinetic symmetry in persons with chronic stroke. Clin. Biomech. 27, 887–892. Duhamel, A., Bourriez, J.L., Devos, P., Krystkowiak, P., Destee, A., Derambure, P., et al., 2004. Statistical tools for clinical gait analysis. Gait Posture 20, 204–212. Dunlap, W.P., Cortina, J.M., Vaslow, J.B., Burke, M.J., 1996. Meta-analysis of experiments with matched groups or repeated measures designs. Psychol. Methods 1, 170–177. Flansbjer, U.B., Holmback, A.M., Downham, D., Patten, C., Lexell, J., 2005. Reliability of gait performance tests in men and women with hemiparesis after stroke. J. Rehabil. Med. 37, 75–82. Haddad, J.M., Emmerik, R.E.A.v, Whittlesey, S.N., Hamill, J., 2006. Adaptations in interlimb and intralimb coordination to asymmetrical loading in human walking. Gait Posture 23. Hassid, E., Rose, D., Commisarow, J., Guttry, M., Dobkin, B.H., 1997. Improved gait symmetry in hemiparetic stroke patients induced during body weight-supported treadmill stepping. J. Neurol. Rehabil. 11, 21–26. Hesse, S., Konrad, M., Uhlenbrock, D., 1999. Treadmill walking with partial body weight support versus floor walking in hemiparetic subjects. Arch. Phys. Med. Rehabil. 80, 421–427. Hodt-Billington, C., Helbostad, J.L., Moe-Nilssen, R., 2008. Should trunk movement or footfall parameters quantify gait asymmetry in chronic stroke patients? Gait Posture 27, 552–558. Hutin, E., Pradon, D., Barbier, F., Bussel, B., Gracies, J.M., Roche, N., 2012. Walking velocity and lower limb coordination in hemiparesis. Gait Posture 36, 205–211.
453
Kwakkel, G., Wagenaar, R.C., 2002. Effect of duration of upper- and lower-extremity rehabilitation sessions and walking speed on recovery of interlimb coordination in hemiplegic gait. Phys. Ther. 82, 432–448. Lenhoff, M.W., Santner, T.J., Otis, J.C., Peterson, M.G.E., Williams, B.J., Backus, S.I., 1999. Bootstrap prediction and confidence bands: a superior statistical method for analysis of gait data. Gait Posture 9, 10–17. Levin, M.F., Kleim, J.A., Wolf, S.L., 2009. What do motor “recovery” and “compensation” mean in patients following stroke? Neurorehabil. Neural. Repair. 23, 313–319. Meijer, R., Plotnik, M., Zwaaftink, E.G., van Lummel, R.C., Ainsworth, E., Martina, J.D., et al., 2011. Markedly impaired bilateral coordination of gait in post-stroke patients: is this deficit distinct from asymmetry? A cohort study. J. Neuroeng. Rehabil. 8. Olney, S.J., Griffin, M.P., McBride, I.D., 1994. Temporal, kinematic, and kinetic variables related to gait speed in subjects with hemiplegia: a regression approach. Phys. Ther. 74, 872–885. Patterson, K.K., Parafianowicz, I., Danells, C.J., Closson, V., Verrier, M.C., Staines, W.R., et al., 2008. Gait asymmetry in community-ambulating stroke survivors. Arch. Phys. Med. Rehabil. 89, 304–310. Patterson, K.K., Gage, W.H., Brooks, D., Black, S.E., McIlroy, W.E., 2010a. Changes in gait symmetry and velocity after stroke: a cross-sectional study from weeks to years after stroke. Neurorehabil. Neural. Repair. 24, 783–790. Patterson, K.K., Gage, W.H., Brooks, D., Black, S.E., McIlroy, W.E., 2010b. Evaluation of gait symmetry after stroke: a comparison of current methods and recommendations for standardization. Gait Posture 31, 241–246. Perry, J., Garrett, M., Gronley, J.K., Mulroy, S.J., 1995. Classification of walking handicap in the stroke population. Stroke 26, 982–989. Peters, B.T., Haddad, J.M., Heiderscheit, B.C., Van Emmerik, R.E.A., Hamill, J., 2003. Limitations in the use and interpretation of continuous relative phase. J. Biomech. 36, 271–274. Plotnik, M., Hausdorff, J.M., 2008. The role of gait rhythmicity and bilateral coordination of stepping in the pathophysiology of freezing of gait in parkinson's disease. Mov. Disord. 23 (Suppl. 2), S444–S450. Plotnik, M., Giladi, N., Hausdorff, J.M., 2008. Bilateral coordination of walking and freezing of gait in parkinson's disease. Eur. J. Neurosci. 27, 1999–2006. Stergiou, N., Scholten, S.D., Jensen, J.L., Blanke, D., 2001. Intralimb coordination following obstacle clearance during running: the effect of obstacle height. Gait Posture 13, 210–220. Stratford, P.W., Spadoni, G., Kennedy, D., Westaway, M.D., Alcock, G.K., 2002. Seven points to consider when investigating a measure's ability to detect change. Physiother. Can. 54, 16–24. Sullivan, K.J., Knowlton, B.J., Dobkin, B.H., 2002. Step training with body weight support: effect of treadmill speed and practice paradigms on poststroke locomotor recovery. Arch. Phys. Med. Rehabil. 83, 683–691. Sullivan, K.J., Brown, D.A., Klassen, T., Mulroy, S., Ge, T., Azen, S.P., et al., 2007. Effects of task-specific locomotor and strength training in adults who were ambulatory after stroke: results of the steps randomized clinical trial. Phys. Ther. 87, 1580–1602.
Contents lists available at SciVerse ScienceDirect
Clinical Biomechanics journal homepage: www.elsevier.com/locate/clinbiomech
Bilateral coordination and gait symmetry after body-weight supported treadmill training for persons with chronic stroke Stephanie A. Combs a, b,⁎, Eric L. Dugan c, Elicia N. Ozimek d, Amy B. Curtis b a
Krannert School of Physical Therapy, University of Indianapolis, 1400 East Hanna Avenue, Indianapolis, IN 46227, USA Interdisciplinary Health Sciences Program, Western Michigan University, Kalamazoo, MI 49008, USA Department of Kinesiology, Center for Orthopaedic and Biomechanics Research, Boise State University, Boise, ID 83725, USA d Biomechanics Laboratory, School of Physical Education, Sport and Exercise Science, Ball State University, Muncie, IN 47306, USA b c
a r t i c l e
i n f o
Article history: Received 22 July 2012 Accepted 5 February 2013 Keywords: Stroke Rehabilitation Gait symmetry Coordination
a b s t r a c t Background: Locomotor interventions are commonly assessed using functional outcomes, but these outcomes provide limited information about changes toward recovery or compensatory mechanisms. The study purposes were to examine changes in gait symmetry and bilateral coordination following body-weight supported treadmill training in individuals with chronic hemiparesis due to stroke and to compare findings to participants without disability. Methods: Nineteen participants with stroke (>6 months) who ambulated between 0.4 and 0.8 m/s and 22 participants without disability were enrolled in this repeated-measures study. The stroke group completed 24 intervention sessions over 8 weeks with 20 minutes of walking/session. The non-disabled group served as a comparison for describing changes in symmetry and coordination. Bilateral 3-dimensional motion analysis and gait speed were assessed across 3 time points (pre-test, immediate post-test, and 6-month retention). Continuous relative phase was used to evaluate bilateral coordination (thigh–thigh, shank–shank, foot–foot) and gait symmetry was assessed with spatiotemporal ratios (step length, swing time, stance time). Findings: Significant improvements in continuous relative phase (shank–shank and foot–foot couplings) were found at post-test and retention for the stroke group. Significant differences in spatiotemporal symmetry ratios were not found over time. Compared to the non-disabled group, changes in bilateral coordination moved in the direction of normal recovery. Most measures of continuous relative phase were more responsive to change after training than the spatiotemporal ratios. Interpretations: After body-weight supported treadmill training, the stroke group made improvements toward recovery of normal bilateral coordination. Bilateral coordination and gait symmetry measures may assess different aspects of gait. © 2013 Elsevier Ltd. All rights reserved.
1. Introduction Distinguishing between intervention induced changes that represent recovery or reinforced compensatory mechanisms is a challenge when evaluating the efficacy of rehabilitation protocols for persons following stroke. In the context of gait, recovery refers to the restitution of pre-stroke gait patterns while compensation refers the use of different gait patterns to accomplish the task (Levin et al., 2009). Metrics such as gait velocity are often used to assess changes in gait function; however, gait velocity does not provide insight into underlying impairments or quality of gait performance (Olney et al., 1994;
⁎ Corresponding author at: Krannert School of Physical Therapy, University of Indianapolis, 1400 East Hanna Avenue, Indianapolis, IN 46227, USA. E-mail addresses: [email protected] (S.A. Combs), [email protected] (E.L. Dugan), [email protected] (E.N. Ozimek), [email protected] (A.B. Curtis). 0268-0033/$ – see front matter © 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.clinbiomech.2013.02.001
Patterson et al., 2008). This led Patterson et al. (2008) to suggest that gait symmetry should be a focus of gait function assessments (with gait velocity as an adjunct) due to its potential to better describe impairments in post-stroke gait. However, spatiotemporal symmetry measures have been reported to be inadequate for distinguishing asymmetry in persons with stroke from persons without known pathology (Hodt-Billington et al., 2008). Therefore, in order to discriminate between recovery of pre-stroke gait function and adaptation of compensatory gait patterns, additional metrics are necessary. Recent work suggests that gait symmetry is only one aspect of the bilateral function of the limbs during gait (Meijer et al., 2011; Plotnik and Hausdorff, 2008). As described by Plotnik and Hausdorff (2008), gait symmetry and bilateral coordination may be separate aspects of the bilateral function of the limbs during gait. While gait symmetry (as measured by swing times of the right and left legs) and bilateral coordination (as measured by phase coordination index which takes into account accuracy and consistency of stepping) are not correlated in healthy adults, they have been reported to be strongly correlated in
S.A. Combs et al. / Clinical Biomechanics 28 (2013) 448–453
individuals with stroke (Meijer et al., 2011). The reasons for this strong association between gait symmetry and bilateral coordination in patients with stroke are not fully understood. Meijer et al. (2011) suggest that the relationship between gait symmetry and bilateral coordination may be due to the asymmetric nature of motor and sensory impairments after stroke. The differing relationships between spatiotemporal asymmetry and bilateral limb coordination in healthy adults as compared to individuals with stroke support the need to incorporate both types of measures when evaluating rehabilitation protocols. Gait velocity, gait symmetry, and bilateral coordination provide different (and perhaps in some cases overlapping) insights regarding gait function. These differences are important in the context of evaluating the efficacy of rehabilitation protocols. In particular, it is important to develop a better understanding of the mechanisms for how interventions lead to recovery of and/or compensatory adaptations in the gait pattern. To this end, the overall aims of the current study were to examine changes in gait velocity, gait symmetry, and bilateral coordination following body weight supported treadmill training (BWSTT) in persons with hemiparesis due to stroke and to compare these findings to participants without disability. BWSTT is a means of locomotor training that can increase overground gait speed after stroke (Sullivan et al., 2002, 2007). As gait symmetry deteriorates over time in persons following stroke (Patterson et al., 2010a), reported improvements in stance and swing time symmetry during BWSTT (Chen et al., 2005a; Hassid et al., 1997; Hesse et al., 1999) suggest that the repetitive and task-oriented nature of the intervention may also promote carryover of improved gait symmetry to overground environments. We hypothesized that BWSTT would lead to immediate and long-term changes in gait velocity, gait symmetry, and bilateral coordination for the group of participants with chronic hemiparesis due to the repetitive, task-specific nature of BWSTT, and that these changes would trend toward the values found by a group of nondisabled adults. Measures of bilateral coordination are novel concepts for quantifying changes in gait function in persons with stroke; therefore, we compared their responsiveness to change after BWSTT to more common spatiotemporal symmetry ratios.
449
Additional inclusion criteria for the stroke group were: 1) at least 6-months post stroke; 2) had only experienced one stroke; 3) ambulatory with or without the use of an assistive device or orthosis; 4) able to ambulate up to 14 m without the use of an assistive device with supervision; 5) able to ambulate at a self-selected gait velocity between 0.4 m/s and 0.8 m/s (Perry et al., 1995); 6) currently not receiving physical therapy services or lower extremity botulinum toxin injections; 7) medically stable with a physician release stating approval to enter an exercise program; 8) able to follow at least two-step verbal instructions. Non-disabled participants were additionally screened according to the following inclusion criteria: 1) able to walk independently during home and community activities; 2) no known neurological condition or deficits, such as prior stroke. The protocol was approved by Institutional Review Boards at participating institutions, and participants provided informed consent prior to initiating the study. Eligibility determination during the initial visit for participants with stroke included two comfortable 10-meter walk tests (CWT) to verify a mean gait velocity between 0.4 and 0.8 m/s. 2.3. Intervention Participants in the stroke group were treated for 24 sessions of BWSTT over 8 weeks with 20 minutes of total walking each session (Sullivan et al., 2002, 2007). Rest breaks were allowed if requested, but were not included in the overall walking time. During BWSTT verbal and manual assistance were provided by 1–2 researchers (SC and trained student physical therapists) to facilitate an optimal gait pattern primarily through weight shifting at the pelvis and/or paretic limb movement during the gait cycle. Participants did not wear an ankle foot orthosis or use handrails during training. Treadmill speed was increased by 0.2 mph increments each subsequent training session until age and gender norms were achieved and body weight support was decreased from 30% over the course of the intervention according to a standardized protocol. The progression parameters of the BWSTT protocol have been previously described (Combs et al., 2010). 2.4. Data acquisition
2. Methods 2.1. Design A prospective repeated measures design was used. Participants with chronic hemiparesis due to stroke (stroke group) were evaluated three times: one-week before and after the intervention period and sixmonths following completion of the intervention (pre-test, post-test and, retention). Participants without disability (non-disabled group) were evaluated three times during the same timeframes as the stroke group, but did not participate in the intervention. Measurements were conducted by members of the research team who were not involved with the intervention. 2.2. Participants The convenience sample of 41 participants included 19 participants with chronic hemiparesis due to stroke and 22 participants without disability. Participants with stroke were recruited from local stroke support groups and clinicians. Participants without disability were recruited from local university communities. All participants were screened according to the following inclusion criteria: 1) between the ages of 40 and 80 years; 2) no current musculoskeletal conditions or recent orthopedic surgeries (outside of typical age-related changes) within 6 months of the study; 3) no complications from other health conditions that could influence walking; 4) able to travel to and from research sessions.
All participants were evaluated using the CWT and bilateral three-dimensional motion analysis. Participants with stroke did not use assistive devices during data collection; however, a research assistant closely monitored these participants by walking next to them, if necessary. All participants wore shoes during data collection and two participants in the stroke group required the use of their ankle foot orthosis on the paretic leg for improved safety during the walking trials. The CWT is a valid and reliable (ICC = 0.94) measure of shortdistance, self-selected overground gait velocity in persons with chronic stroke (Flansbjer et al., 2005). It was conducted in an open room along a 14-meter walkway. Participants were timed for the middle 10 m using electronic timing gates (Fitness Technologies, 21 Bishop Street, Adelaide-Skye, Australia, 5072). The mean of three trials was used for data analysis (Flansbjer et al., 2005). Bilateral three-dimensional motion data were captured with a 10-camera Vicon MX motion capture system (Vicon, 5419 McConnel Avenue, Los Angeles, CA, USA 90066) (120 Hz). Participants wore 39 retro-reflective markers attached to specific anatomical landmarks based on Vicon's Plug-in-Gait marker set. Participants in the stroke group walked along a 6-meter walkway with embedded force plates (Model BP600900-6-1000, Advanced Medical Technology, Inc., Watertown, MA, USA) at a self-selected comfortable pace for two practice trials, followed by 10 trials for data collection. Participants in the non-disabled group were instructed to walk along the 6-meter walkway with embedded force plates at a slow walking speed that was matched to the average comfortable walking speed of the stroke
450
S.A. Combs et al. / Clinical Biomechanics 28 (2013) 448–453
group established at pre-test with the CWT. Non-disabled group participants practiced walking at the slow pace for two trials using a metronome to assist in establishing the slower pace, followed by 10 trials without the metronome for data analysis. 2.5. Data reduction The first three walking trials per participant that included a clean force plate strike and all markers present were averaged and used for subsequent data analysis. A clean force plate strike was defined as a trial where the entire foot, and no part of the contra-lateral foot, made contact with the force plate. Force plate data were used to identify foot-strike and toe-off. All kinematic data were filtered using a 2nd order low-pass Butterworth filter with a cutoff frequency of 6 Hz. The kinematic data were normalized to 100% of the gait cycle, defined as foot strike to foot strike of the same limb. Spatiotemporal variables including step length, stance time, and swing time were calculated for the paretic and non-paretic legs of the stroke group and for the right and left legs of the non-disabled group. Gait symmetry was evaluated with spatiotemporal symmetry ratios defined as the ratio of paretic to nonparetic (Patterson et al., 2010b) values for each variable for the stroke group and the ratio of right to left values for the non-disabled group. Bilateral coordination was examined through continuous relative phase (CRP) analysis. First, mean sagittal plane segmental angular positions and velocities were used to construct normalized phase plots for the thigh, shank, and foot segments from each participant (Barela et al., 2000; Peters et al., 2003). Phase angles were then calculated as the angle between the right horizontal and the line generated from the origin to each point on the phase plot. The phase angles were calculated at 1% intervals of the gait cycle. CRP for the interlimb couplings (thigh–thigh, shank–shank, and foot–foot) was defined as the difference between the normalized phase angles of contralateral segments, right minus left for the control group and non-paretic minus paretic for the stroke group (Haddad et al., 2006). The mean CRP for each interlimb coupling was displayed as a continuous curve averaged at each 1% interval of the complete gait cycle. A CRP curve near 0° represents an in-phase relationship between the two limb segments (Stergiou et al., 2001). CRP values that deviate from 0° indicate a more out-of-phase relationship between the two segment motions. The CRP curves were also represented as a single value by averaging the mean absolute CRP (MARP) over each 1% interval of the gait cycle (Kwakkel and Wagenaar, 2002; Stergiou et al., 2001). 2.6. Statistical analysis A threshold of at least 80% completion of intervention sessions (at least 20 out of 24 sessions) was required for participant outcomes to be included in the final data analysis. Data analysis was performed using Statistical Package for the Social Sciences version 17.0 (IBM SPSS, 1 New Orchard Road, Armonk, NY, USA 10504). Demographic characteristics and dependent variables at each measurement time point (pre-test, post-test and, retention) were examined for normality (Shapiro–Wilk, P > .05). Demographics were compared between groups by independent t-test (age) and Fisher's exact test (gender). A one-way repeated measures analysis of variance was used to examine within-group differences for all dependent measures over the three measurement periods. Statistically significant differences were determined using a 95% confidence interval adjusted for repeated measures. Post hoc analysis for the CWT and spatiotemporal symmetry ratios were conducted using the Bonferroni correction on eligible pairwise comparisons. Post hoc analyses for MARP were conducted using confidence bands constructed around the mean differences between CRP curves from pre- to post-test and pre-test to retention to determine whether
differences existed using a non-parametric bootstrap procedure (Duhamel et al., 2004; Lenhoff et al., 1999). Use of confidence bands has been recommended as a preferred test for evaluating differences between mean continuous gait curves (Duhamel et al., 2004; Lenhoff et al., 1999) and is a complementary follow-up procedure to the repeated measures analysis of variance (Duhamel et al., 2004). The confidence bands encompassed the region of the true mean difference between CRP curves at 95% probability. The customized bootstrapping procedure, developed with MATLAB (MathWorks, 3 Apple Hill Drive, Natick, MA, USA 01760), involved simultaneous re-sampling of the CRP curves at each measurement period (L = 500) in order to estimate the variability present in the population. Points where the confidence bands did not contain 0 (horizontal axis) indicated a statistically significant difference between the two CRP curves and therefore a significant change in coordination at that point. The number of points that did not contain 0 was totaled to reveal the overall percentage of the gait cycle that showed a significant difference in interlimb coordination. This analysis allowed for a description of how much of the gait cycle changed as well as where in the gait cycle the changes occurred. Responsiveness to change (Stratford et al., 2002) of the MARP and spatiotemporal symmetry variables were assessed with effect sizes (ES) and standardized response means (SRM). While responsiveness to change does not necessarily determine the meaningfulness of the change to the patient, this analysis is important as it allows for interpretation of how well the variable measures change over time (Stratford et al., 2002). ES was calculated using the difference in mean scores from pre- to post-test and from pre-test to retention divided by the respective pooled standard deviations (Dunlap et al., 1996). SRM was calculated using the difference in mean scores from pre- to post-test and from pre-test to retention divided by the respective standard deviation of the change scores. Responsiveness based on the ES and SRM of the measures was classified as small (d = 0.2), medium (d = 0.5), or large (d = 0.8) (Cohen, 1992). 3. Results Three participants in the stroke group did not complete the study due to changes in health status unrelated to the study. Motion analysis data from one participant in the stroke group and 3 participants in the non-disabled group were not used due to data collection error. We were unable to locate one participant from the non-disabled group at the 6-month testing session. These participants were subsequently excluded from the analysis. The fifteen participants in the stroke group who were included in the final data analysis all completed more than 80% of the intervention sessions. No statistically significant differences were found between groups for age and gender (P > .05). See Table 1 for group demographics. There were also no significant differences over time for the spatiotemporal symmetry ratios and MARP variables for the non-disabled group (P > .05). (Table 2) However, there was a statistically significant main effect difference (P = .02) in slow gait velocity for the non-disabled group with pretest gait speed significantly differing from post-test (P = .03) gait speed (Table 2). For the stroke group, statistically significant main effect differences were found over time for the CWT and MARP (shank/shank and foot/
Table 1 Participant demographics. Groups
Years of age, mean(SD) Gender, male/female Years post stroke, mean (SD) Side of Paresis, right/left
P
Stroke n = 15
Non-disabled n = 18
59.9 (11.2) 4/11 3.8 (3.2) 6/9
57.4 (7.6) 8/10 n/a n/a
P ≤ 0.05; independent t-test used for age; Fisher's exact test used for gender.
.45 .47
S.A. Combs et al. / Clinical Biomechanics 28 (2013) 448–453
451
Table 2 Mean gait velocity, spatiotemporal symmetry ratios and MARP over time. Variables
Stroke group (n = 15)
Non-disabled group (n = 18)
Mean (SD) Pre-test Gait velocity CWT (m/s) Spatiotemp. ratios Step Length Sym Stance time sym Swing time sym MARP Thigh–thigh (°) Shank–shank (°) Foot–foot (°)
Post-test
Retention
Main effects
Mean (SD)
F
P
Pre-test
Main effects Post-test
Retention
F
P
0.63 (0.15)
0.75 (0.19)a
0.80 (0.25)b
11.41
.00
0.73 (0.07)
0.68 (0.05)c
0.69 (0.06)
4.32
.02
1.27 (0.26) 0.90 (0.07) 1.32 (0.24)
1.26 (0.44) 0.89 (0.07) 1.24 (0.15)
1.34 (0.53) 0.89 (0.06) 1.30 (0.19)
0.82 0.69 1.15
.45 .46 .32
1.01 (0.05) 1.01 (0.05) 1.01 (0.08)
1.00 (0.06) 1.00 (0.04) 1.00 (0.07)
0.99 (0.06) 1.02 (0.04) 1.00 (0.10)
1.31 1.24 0.02
.28 .30 .98
0.74 25.87 21.20
.49 .00 .00
−0.60 (6.03) 0.72 (7.12) 2.04 (6.52)
0.39 (6.89) 0.58 (7.94) 0.35 (5.91)
−0.01 (5.82) −0.93 (8.47) −0.41 (5.88)
0.21 0.29 0.72
.82 .75 .49
−11.45 (15.91) −43.72 (21.07) −52.77 (35.60)
−9.37 (19.49) −24.18 (23.79)a −26.31 (26.07)a
−7.66 (14.62) −29.06 (25.28)b −28.97 (35.51)b
SD = standard deviation, CWT = 10-meter comfortable walk test, spatiotemporal ratios = paretic/nonparetic variables, sym = symmetry, MARP = mean absolute continuous relative phase. a Indicates significant difference at b.01 from pre-test to post-test. b Indicates significant difference at b.01 from pre-test to retention. c Indicates significant difference at b.05 from pre-test to post-test.
The primary purposes of this study were to examine changes in gait symmetry and bilateral coordination following BWSTT in persons with chronic hemiparesis due to stroke and to compare the findings to persons without disability in order to identify direction of change toward recovery or compensation mechanisms. Although no significant differences were found over time in spatiotemporal symmetry, bilateral coordination based on CRP analysis significantly improved toward more in-phase relationships between the paretic and nonparetic limbs immediately following BWSTT partially supporting our hypothesis. Changes in bilateral coordination were still apparent 6-months after BWSTT; however, only a portion of the improvements in coordination were maintained by the stroke group. While the BWSTT protocol employed in this study led to lasting changes in gait velocity, it may not have been sufficient to bring about permanent changes in bilateral coordination. Alternative application of BWSTT protocol parameters or a longer duration of training may be necessary to induce permanent changes in coordination along with any changes in symmetry of step length, stance time, and swing time. Determination of optimal intervention parameters for persons following stroke will be an important direction of future research. In the present study, a greater percentage of statistically significant differences in bilateral coordination were found during the latter part of the CRP curves signifying that most of the changes occurred during swing phase. Impaired swing initiation has been identified as
30
CRP Degrees
20 10 0 -10
1
11
21
31
41
51
61
71
81
91
101
-20 -30 -40 -50
Percent of Gait Cycle
B
10 0 -10
CRP Degrees
4. Discussion
A
1
11
21
31
41
51
61
71
81
91
81
91
101
-20 -30 -40 -50 -60 -70
Percent of Gait Cylce
-80
C 0 1
11
21
31
41
51
61
71
101
-20
CRP Degrees
foot) variables (P = .00). Post hoc analyses revealed statistically significant increases in self-selected gait velocity from pre- to post-test (P = .00) and pre-test to retention (P b .01) (Table 2). Main effects did not demonstrate statistically significant differences over time for the spatiotemporal symmetry ratios or MARP (thigh/thigh) variables for the stroke group (P > .05) (Table 2, Fig. 1A). CRP post-hoc comparisons using confidence bands indicated statistically significant differences (P b .05) for 33% of the gait cycle from pre- to post-test for the shank–shank interlimb CRP curves (Fig. 1B). At retention, 17% of this significant difference was maintained. Statistically significant differences (P b .05) of 55% of the gait cycle were found for the foot–foot interlimb CRP curves from pre- to post-test with 24% maintained at retention (Fig. 1C). Changes in interlimb CRP corresponded primarily to swing phase. Mean thigh–thigh, shank– shank, and foot–foot interlimb CRP curves for the non-disabled group are included in Fig. 1A–C respectively. Responsiveness to change assessed by ES and SRM of the MARP and spatiotemporal symmetry variables for the stroke group is presented in Table 3.
-40 -60 -80
-100 -120 -140
Percent of Gait Cycle
Fig. 1. Mean interlimb CRP curves across gait cycle. (A) Thigh–thigh; (B) shank–shank; (C) foot–foot. Stroke group mean CRP displayed with black lines: solid black = baseline; large dashes = post-test; small dashes = retention. Gray areas between curves indicate statistically significant difference from pre- to post-test; P ≤ .05; cross-hatched areas between curves (XXX) indicate statistically significant difference from pre-test to retention; P ≤ .05; non-disabled group mean CRP displayed with dotted gray line.
452
S.A. Combs et al. / Clinical Biomechanics 28 (2013) 448–453
Table 3 Responsiveness to change in the stroke group after BWSTT (n = 15). Variables
Spatiotemporal ratios Step length sym Stance time sym Swing time sym MARP Thigh–thigh Shank–shank Foot–foot
Pre-to post-test
Pre-test to retention
ES
SRM
ES
SRM
.07 .23 .42
.30 .20 .29
.04 .14 .09
.09 .20 .12
.12 .87 .85
.18 2.16 1.58
.25 .63 .67
.28 1.31 1.35
ES = effect size (Cohen's d); SRM= standardized response mean; sym = symmetry; MARP = mean absolute continuous relative phase. Responsiveness: small (d = 0.2), medium (d = 0.5), large (d = 0.8).
a major contributor to differences in gait kinematics of persons with stroke compared to those without disability (Chen et al., 2005b). Barela et al. (2000) also reported that the intralimb coordination of paretic segments was most different from adults without disability during the swing phase of the gait cycle. Given these consistent deviations from normal during swing after stroke, it stands to reason that a greater potential for change may be possible during swing as was found after the intervention in this study. Based on the training needs of our group of participants, manual facilitation was most often provided to the paretic leg during late stance and swing to improve active ankle dorsiflexion and passive knee flexion, induced by active hip flexion. While the progressively increased treadmill speed and reduction in body weight support cannot be ruled out as potential parameters that influenced bilateral coordination in our participants, the manual facilitation to achieve more normal stepping patterns may have played a role in the improved swing phase coordination after BWSTT. Future studies will need to separate intervention-based control parameters in order to determine their different effects on coordination. Examination of the shank–shank and foot–foot couplings revealed that the paretic leg remained ahead of the non-paretic in the phase trajectory based on the negative CRP values during all three measurement periods. However, at post-test and retention these values were significantly less negative meaning the limb segments were moving toward a more in-phase relationship with improved coordination. Although the stroke group did not achieve the level of bilateral coordination demonstrated by the group without disability during overground walking (Fig. 1), they showed the ability to change movement patterns in the direction of normal recovery (Levin et al., 2009) after BWSTT. Change primarily at the interlimb level (as opposed to intralimb level) has been demonstrated in previous work with healthy adults in response to additional load applied unilaterally during comfortable walking (Haddad et al., 2006). The results of the current study lend support to these findings and indicate that with or without neurological dysfunction, gait is adaptable at the interlimb level. Future exploration of changes in coordination at the intralimb level within the paretic and non-paretic limbs after BWSTT may expand our understanding of the system's capacity to modulate coordination patterns after stroke. Regardless of efforts to standardize the slow walking pace of the non-disabled group during testing, significant differences in gait velocities were found across time. However, the non-disabled participants maintained stable patterns of spatiotemporal symmetry and bilateral coordination over the testing periods even with differences in their slowed walking pace. This was not the case with the stroke group that also demonstrated significant differences in gait speeds as well as improved bilateral coordination. Speed has been proposed as a limitation to the use of CRP analysis in persons with stroke (Barela et al., 2000) with increased walking speed associated with changes in intralimb coordination during mid-swing of the non-paretic limb (Hutin et al., 2012). For the stroke group in this study, change from pre-test in gait speed was not associated with change in bilateral
coordination. Meijer et al. (2011) similarly reported that gait speed and measures of bilateral coordination are independent walking properties. Thus, the alteration in coordination after locomotor training in this study may be attributed to a training parameter within the BWSTT modality. Despite reports of enhanced spatiotemporal symmetry while walking on a treadmill with body weight support (Chen et al., 2005a; Hassid et al., 1997; Hesse et al., 1999), these training enhancements did not transfer to overground walking for this group with stroke. Comparatively, the MARP measures of bilateral coordination (shank– shank and foot–foot) were more responsive to change after BWSTT than the spatiotemporal symmetry ratios. Plotnik et al. (2008) reported that bilateral coordination may characterize the rhythmic properties between limbs during stepping which differs from spatiotemporal symmetry measures (i.e. swing time and step length) that more likely reflect properties of propulsion from each limb (Balasubramanian et al., 2007; Plotnik et al., 2008). Accordingly, the BWSTT intervention employed in this study may have induced greater bilateral coordination between limbs due to the rhythmic, continuous motion of the treadmill. As a collective variable (Barela et al., 2000; Stergiou et al., 2001) that takes into consideration both angular position and velocity across strides, CRP may be more responsive after interventions such as BWSTT due to a more specific representation of the organization of neuromuscular control processes (Stergiou et al., 2001). In contrast, the use of body weight support could have interfered with propulsion due to reduced loading through the lower extremities and in turn reduced the responsiveness to changes in spatiotemporal symmetry. This finding is consistent with our recent report of a lack of change in kinetic symmetry (i.e. relative propulsion of the paretic leg and joint work) in persons following stroke after BWSTT (Combs et al., 2012). Therefore, using multiple metrics to evaluate rehabilitation interventions may be important in order to address different characteristics of gait as well as variable effects of interventions. While CRP was more responsive to changes in bilateral coordination, the clinical meaningfulness of those changes is unknown and should to be investigated in future studies. The small sample size may limit the generalizability of the findings. Lack of a true control group for comparison of the BWSTT to an alternative intervention limits the interpretation of the results. Other possible limitations were that sagittal plane motions only were used and that two participants required an AFO during the data collection. Although the use of an AFO has been shown to have little effect on intralimb coordination (Barela et al., 2000) and kinetic symmetry (Combs et al., 2012), its contribution to interlimb coordination is unknown.
5. Conclusions Because evidence indicates that gait asymmetry worsens over time in patients following stroke (Patterson et al., 2010a), the findings of this study have direct clinical implications for rehabilitation. The improvements made by the stroke group toward recovery of normal bilateral coordination after BWSTT justifies the importance of continued rehabilitation for persons in the chronic stage of stroke. CRP analysis is a promising measure of bilateral coordination of gait and should be explored in future studies.
Acknowledgments This project was supported in part by the University of Indianapolis Research/Creative Endeavor's Grant program. The researchers of this study would like to acknowledge our participants; Jeff Frame, MS at the Biomechanics Lab at Ball State University; and graduates from the Krannert School of Physical Therapy, Miranda Passmore, PT, DPT, Cara Riesner, PT, DPT, Dana Whipker, PT, DPT, and Elizabeth Yingling PT, DPT.
S.A. Combs et al. / Clinical Biomechanics 28 (2013) 448–453
References Balasubramanian, C.K., Bowden, M.G., Neptune, R.R., Kautz, S.A., 2007. Relationship between step length asymmetry and walking performance in subjects with chronic hemiparesis. Arch. Phys. Med. Rehabil. 88, 43–49. Barela, J.A., Whitall, J., Black, P., Clark, J.E., 2000. An examination of constraints affecting the intralimb coordination of hemiparetic gait. Hum. Mov. Sci. 19, 251–273. Chen, G., Patten, C., Kothari, D.H., Zajac, F.E., 2005a. Gait deviations associated with post-stroke hemiparesis: improvement during treadmill walking using weight support, speed, support stiffness, and handrail hold. Gait Posture 22, 57–62. Chen, G., Patten, C., Kothari, D.H., Zajac, F.E., 2005b. Gait differences between individuals with post-stroke hemiparesis and non-disabled controls at matched speeds. Gait Posture 22, 51–56. Cohen, J., 1992. A power primer. Psychol. Bull. 112, 155–159. Combs, S.A., Dugan, E.L., Passmore, M., Riesner, C., Whipker, D., Yingling, E., et al., 2010. Balance, balance confidence, and health-related quality of life in persons with chronic stroke after body weight-supported treadmill training. Arch. Phys. Med. Rehabil. 91, 1914–1919. Combs, S.A., Dugan, E.L., Ozimek, E.N., Curtis, A.B., 2012. Effects of body weight supported treadmill training on kinetic symmetry in persons with chronic stroke. Clin. Biomech. 27, 887–892. Duhamel, A., Bourriez, J.L., Devos, P., Krystkowiak, P., Destee, A., Derambure, P., et al., 2004. Statistical tools for clinical gait analysis. Gait Posture 20, 204–212. Dunlap, W.P., Cortina, J.M., Vaslow, J.B., Burke, M.J., 1996. Meta-analysis of experiments with matched groups or repeated measures designs. Psychol. Methods 1, 170–177. Flansbjer, U.B., Holmback, A.M., Downham, D., Patten, C., Lexell, J., 2005. Reliability of gait performance tests in men and women with hemiparesis after stroke. J. Rehabil. Med. 37, 75–82. Haddad, J.M., Emmerik, R.E.A.v, Whittlesey, S.N., Hamill, J., 2006. Adaptations in interlimb and intralimb coordination to asymmetrical loading in human walking. Gait Posture 23. Hassid, E., Rose, D., Commisarow, J., Guttry, M., Dobkin, B.H., 1997. Improved gait symmetry in hemiparetic stroke patients induced during body weight-supported treadmill stepping. J. Neurol. Rehabil. 11, 21–26. Hesse, S., Konrad, M., Uhlenbrock, D., 1999. Treadmill walking with partial body weight support versus floor walking in hemiparetic subjects. Arch. Phys. Med. Rehabil. 80, 421–427. Hodt-Billington, C., Helbostad, J.L., Moe-Nilssen, R., 2008. Should trunk movement or footfall parameters quantify gait asymmetry in chronic stroke patients? Gait Posture 27, 552–558. Hutin, E., Pradon, D., Barbier, F., Bussel, B., Gracies, J.M., Roche, N., 2012. Walking velocity and lower limb coordination in hemiparesis. Gait Posture 36, 205–211.
453
Kwakkel, G., Wagenaar, R.C., 2002. Effect of duration of upper- and lower-extremity rehabilitation sessions and walking speed on recovery of interlimb coordination in hemiplegic gait. Phys. Ther. 82, 432–448. Lenhoff, M.W., Santner, T.J., Otis, J.C., Peterson, M.G.E., Williams, B.J., Backus, S.I., 1999. Bootstrap prediction and confidence bands: a superior statistical method for analysis of gait data. Gait Posture 9, 10–17. Levin, M.F., Kleim, J.A., Wolf, S.L., 2009. What do motor “recovery” and “compensation” mean in patients following stroke? Neurorehabil. Neural. Repair. 23, 313–319. Meijer, R., Plotnik, M., Zwaaftink, E.G., van Lummel, R.C., Ainsworth, E., Martina, J.D., et al., 2011. Markedly impaired bilateral coordination of gait in post-stroke patients: is this deficit distinct from asymmetry? A cohort study. J. Neuroeng. Rehabil. 8. Olney, S.J., Griffin, M.P., McBride, I.D., 1994. Temporal, kinematic, and kinetic variables related to gait speed in subjects with hemiplegia: a regression approach. Phys. Ther. 74, 872–885. Patterson, K.K., Parafianowicz, I., Danells, C.J., Closson, V., Verrier, M.C., Staines, W.R., et al., 2008. Gait asymmetry in community-ambulating stroke survivors. Arch. Phys. Med. Rehabil. 89, 304–310. Patterson, K.K., Gage, W.H., Brooks, D., Black, S.E., McIlroy, W.E., 2010a. Changes in gait symmetry and velocity after stroke: a cross-sectional study from weeks to years after stroke. Neurorehabil. Neural. Repair. 24, 783–790. Patterson, K.K., Gage, W.H., Brooks, D., Black, S.E., McIlroy, W.E., 2010b. Evaluation of gait symmetry after stroke: a comparison of current methods and recommendations for standardization. Gait Posture 31, 241–246. Perry, J., Garrett, M., Gronley, J.K., Mulroy, S.J., 1995. Classification of walking handicap in the stroke population. Stroke 26, 982–989. Peters, B.T., Haddad, J.M., Heiderscheit, B.C., Van Emmerik, R.E.A., Hamill, J., 2003. Limitations in the use and interpretation of continuous relative phase. J. Biomech. 36, 271–274. Plotnik, M., Hausdorff, J.M., 2008. The role of gait rhythmicity and bilateral coordination of stepping in the pathophysiology of freezing of gait in parkinson's disease. Mov. Disord. 23 (Suppl. 2), S444–S450. Plotnik, M., Giladi, N., Hausdorff, J.M., 2008. Bilateral coordination of walking and freezing of gait in parkinson's disease. Eur. J. Neurosci. 27, 1999–2006. Stergiou, N., Scholten, S.D., Jensen, J.L., Blanke, D., 2001. Intralimb coordination following obstacle clearance during running: the effect of obstacle height. Gait Posture 13, 210–220. Stratford, P.W., Spadoni, G., Kennedy, D., Westaway, M.D., Alcock, G.K., 2002. Seven points to consider when investigating a measure's ability to detect change. Physiother. Can. 54, 16–24. Sullivan, K.J., Knowlton, B.J., Dobkin, B.H., 2002. Step training with body weight support: effect of treadmill speed and practice paradigms on poststroke locomotor recovery. Arch. Phys. Med. Rehabil. 83, 683–691. Sullivan, K.J., Brown, D.A., Klassen, T., Mulroy, S., Ge, T., Azen, S.P., et al., 2007. Effects of task-specific locomotor and strength training in adults who were ambulatory after stroke: results of the steps randomized clinical trial. Phys. Ther. 87, 1580–1602.