- Email: [email protected]
Obstacle crossing following stroke improves over one month when the unaffected limb leads, but not when the affected limb leads
Gait & Posture 39 (2014) 213–217
Contents lists available at SciVerse ScienceDirect
Gait & Posture journal homepage: www.elsevier.com/locate/gaitpost
Obstacle crossing following stroke improves over one month when the unaffected limb leads, but not when the affected limb leads§ Catherine M. Said a,b,*, Mary Galea c, Noel Lythgo d a
Physiotherapy, Melbourne School of Health Sciences, University of Melbourne, Parkville, Victoria, Australia Physiotherapy Department, Austin Health, Heidelberg West, Victoria, Australia c Department of Medicine, University of Melbourne, c/- Royal Melbourne Hospital, Parkville, Victoria, Australia d School of Medical Sciences, RMIT University, Victoria, Australia b
A R T I C L E I N F O
A B S T R A C T
Article history: Received 15 January 2013 Received in revised form 24 May 2013 Accepted 8 July 2013
While it is well established that obstacle crossing is impaired following stroke, it is not known whether obstacle crossing improves as gait improves following stroke. The purpose of this study was to determine whether obstacle crossing changed over a one month time period in people with a recent stroke. Twenty participants receiving rehabilitation following a recent stroke were tested on two occasions one month apart. Participants received usual care rehabilitation, including physiotherapy, between the tests. The main outcome measure was obstacle crossing speed as participants stepped over a 4-cm high obstacle. Secondary measures were spatiotemporal variables. Data were collected via a three dimensional motion analysis system. When leading with the affected limb no changes in obstacle crossing speed or spatiotemporal variables were observed over the one month period. When leading with the unaffected limb, crossing speed significantly increased (p = .002), and affected trail limb swing time (p = .03) and crossing step double support time reduced (p = .016). While not significant, the lead and trail limb preobstacle distance increased (p = .08), and lead swing time (p = .052) reduced. Change in obstacle crossing speed did not correlate with change in level gait speed. Obstacle crossing does not necessarily improve over a one month time period in people receiving rehabilitation following stroke. These findings suggest that there may be a need for more targeted training of obstacle crossing, particularly when leading with the affected limb. ß 2013 Elsevier B.V. All rights reserved.
Keywords: Stroke Gait disorder Neurologic Rehabilitation Obstacle crossing
Obstacle crossing is impaired for many people following a stroke [1,2]. Said et al. found over 50% of people able to walk and receiving rehabilitation following stroke either contacted the obstacle or lost balance when attempting to clear a small obstacle [1]. People with stroke also utilised different movement patterns to clear an obstacle compared to unimpaired participants [3,4]. They had significantly slower gait speeds as they crossed the obstacle, which accounted for some differences in the gait pattern such as lead limb placement before the obstacle [3]. However, speed did not account for all gait adjustments. Compared with healthy participants walking at matched speed, people with stroke placed the unaffected lead limb and affected and unaffected trail limb closer to the obstacle after crossing [3]. They also positioned their
§ This data was presented in part at the Australian Physiotherapy Conference Week, Brisbane, 2011. * Corresponding author at: Melbourne School of Health Sciences, The University of Melbourne, Parkville 3010, Australia. Tel.: +61 3 9496 2055. E-mail addresses: [email protected], [email protected] (C.M. Said).
0966-6362/$ – see front matter ß 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.gaitpost.2013.07.008
centre of mass closer to the base of support when leading with the affected limb [4]. Many of these studies assessed people undergoing rehabilitation within six months of stroke, thus some people may not have attained optimal walking recovery [3–5]. While obstacle crossing is more challenging than level ground walking [6–8], it is fundamentally a locomotor task. It is reasonable to expect that if walking ability improves, obstacle crossing may also improve. Gait speed is an important clinical marker of gait improvement [9], and there are valid rationales for anticipating that increased level over-ground gait speed would lead to improvements in obstacle crossing. Some movement deficits during obstacle crossing following stroke were partly speed associated [3,4] and people who fail an obstacle crossing task cross the obstacle more slowly compared with people who pass [10]. Thus if gait speed improves following stroke, aspects of obstacle crossing performance may also normalise. In addition, for some participants, the obstacle crossing task may have been ‘novel’ following their stroke, as their exposure to complex walking tasks may have been limited at this stage of recovery. With repeated exposure to complex locomotor
214
C.M. Said et al. / Gait & Posture 39 (2014) 213–217
tasks, including obstacle crossing, people with subacute stroke may change their obstacle crossing strategy. We therefore wanted to investigate whether obstacle crossing changes over time in people with subacute stroke. While many variables have been examined during obstacle crossing, the primary outcome of interest for this study was gait speed during obstacle crossing, as this differentiates between people with stroke and controls [3]. It is also associated with failure on an obstacle crossing task [10]. To explore mechanisms behind observed changes, temporal and sagittal plane spatial variables were examined. The purpose of this study was to determine whether obstacle crossing gait speed improves over time in people recovering from a recent stroke. It was hypothesised that if level over-ground (unobstructed) walking speed increased following stroke, it would be associated with an increase in obstacle crossing gait speed. It was also anticipated that spatial and temporal variables would change and approach values achieved by unimpaired adults. 1. Method This was an observational study. Ethics approval was obtained and participants provided informed consent. Obstacle crossing performance and falls, obstacle crossing performance and spatiotemporal characteristics, and variability in spatiotemporal characteristics in these participants have been previously reported [10,11]. 1.1. Participants Data were collected from 20 participants recruited from two hospital physiotherapy departments. A sample of 20 was sufficient to detect a large difference between the two tests (d = 0.8), with a power of 70% and a two tailed alpha of 0.05. Participants were receiving ongoing physiotherapy for a gait or balance disorder following a recent stroke, and capable of walking 10 m without a gait aid or physical assistance. Participants were excluded if they had other medical, musculoskeletal or neurological conditions that may have impacted walking. Mean age was 61.1 years (SD = 15), mean height 167.9 cm (SD = 9.23) and participants were first tested a median of 60.7 days (SD = 47.2) post stroke. Participant characteristics are provided in Table 1. 1.2. Apparatus Data were recorded by a six or eight camera VICON 612 3D motion system1 and AMTI forceplate.2 During the obstructed trials a red coloured balsa wood obstacle measuring 40 mm high 1.5 mm thick 600 mm long was positioned after the forceplate, approximately 5 m from the start. Data processing utilised Vicon BodyBuilder1 Version 3.55 (build 136).1 1.3. Procedure Participants wore loose fitting shorts, walking shoes and any prescription eyewear. Anthropometric measures were obtained (Vicon Plug-in Gait Product Guide1) and twenty-one 14 mm passive reflective markers were placed on the lower limbs, [12,13] acromions and obstacle as previously described [3]. Participants performed four unobstructed walking trials at selfselected speed, followed by eight trials with the obstacle. They were instructed to walk at self-selected speed and step over the 1 Oxford Metrics Ltd., 14 Minns Estate, West Way, Oxford, OX2OJB, United.ingdom. 2 Advanced Mechanical Technology, Inc.: Watertown, MA.
obstacle without contacting it or losing balance. Participants were accompanied by a therapist, who walked behind and to the side of the subject lightly holding a safety belt. Assistance was only provided if required. Participants repeated the test procedure one month later (mean 29.5 days, SD = 5.4). During this time they received their usual physiotherapy. No attempt was made to standardise or prescribe the type of treatment received; however, as ‘receiving physiotherapy for a gait or balance disorder’ was a criterion for inclusion, it was assumed that a portion of therapy was directed towards these issues. 1.4. Gait variables: data processing For each participant, one trial leading with the affected limb and one trial leading with the unaffected limb were analysed for each test. The first trial with adequate data (minimal marker occlusion and clean forceplate strike, if available) was selected. The gait pattern utilised by people with stroke to cross an obstacle was fairly consistent over three attempts within a single session [11], so the selection of one trial for analysis was justified. Trials were excluded from motion analysis if the participant required assistance to maintain balance. Data were filtered using Woltring filtering routine with a predicted Mean-Squared-Error value of 20. BodyBuilder11 was used to create ‘virtual markers’ on the shoe at the most distal point of the toe and heel. Data were exported to Microsoft Excel to calculate obstacle crossing gait speed (trail heel contact preobstacle to trail heel contact post-obstacle), lead and trail limb preobstacle horizontal distance (lead or trail heel position before the obstacle to the obstacle), vertical toe clearance (top of the obstacle to the lead or trail toe) and post-obstacle horizontal distance (lead or trail heel position after the obstacle to the obstacle). Foot contact and toe off events were obtained from BodyBuilder1 by visual inspection of the position of the virtual markers on the heel and toe. Lead limb swing time was from lead limb toe off pre-obstacle to lead limb foot contact post-obstacle. Trail limb swing time was from trail limb toe off pre-obstacle to foot contact post-obstacle. Crossing double support time was from lead foot contact postobstacle to trail toe off pre-obstacle. More details on data processing have been previously reported [3]. 1.5. Statistical analysis Statistical analyses were undertaken using IBM SPSS Version 20 for Windows.3 Paired t tests were used to determine whether level over-ground walking speed had changed between the two test sessions. To determine whether obstacle crossing gait speed or spatiotemporal variables changed over the one month period, data from trials leading with the affected limb and unaffected limb were analysed separately. While most spatial data were normally distributed, temporal data were skewed. Therefore paired t tests were used for spatial data and Wilcoxon Signed Rank tests were used for temporal data. Effect sizes were calculated for all data. For parametric tests, effect sizes of 0.8, 0.5 and 0.2 were interpreted as large, medium and small respectively [14]. For nonparametric tests, effect sizes of 0.5, 0.3 and 0.1 were interpreted as large, medium and small respectively [15]. To determine whether changes in obstacle crossing speed were associated with changes in level over-ground walking speed, correlations between the difference scores for the two tests were calculated using Pearson’s r. As this was an exploratory study and risks associated with a Type I error were low, no corrections for multiple tests were applied. 3
IBM SPSS Incß IBM Corporation.
C.M. Said et al. / Gait & Posture 39 (2014) 213–217
215
Table 1 Characteristics of participants. Participant no.
Age (y)
Lesion site
Days post-stroke
Gait aida
Gait speed (m/s)
1 2 3 4 5 6 7 8b 9c,d 10 11 12 13 14 15 16b 17d 18 19 20
75 33 57 65 60 26 85 49 74 70 52 71 73 70 62 43 68 76 55 58
R subcortical infarct L frontal infarct R basal ganglia haemorrhage R thalamic haemorrhage L corona radiata infarct R MCA infarct R MCA infarct R subcortical stroke L post limb internal capsule infarct Brainstem stroke R pontine, L corona radiata L MCA infarct Multiple infarcts both hemispheres and cerebellum L MCA infarct L basal ganglia haemorrhage L putamen/corona radiata lacunar infarct L pontine infarct R MCA/ACA infarct R occipitoparietal haemorrhage L medullary infarct L MCA infarct
10 42 34 28 12 155 48 58 97 27 62 55 50 50 77 207 72 35 66 28
Nil Nil Nil Nil Nil Nil SPS AFO, SPS AFO, SPS Nil Nil 4WF 2WF 4WF SPS SAB, SPS Nil Nil Nil Nil
1.1 1.1 0.8 0.9 0.8 0.9 0.4 0.2 0.2 0.8 1.1 0.6 0.6 0.1 0.4 0.8 0.6 1.2 1.3 1.1
Adapted from Said et al. [10], with permission of the American Physical Therapy Association. ß2013 American Physical Therapy Association. Abbreviations: R, right; L, left; MCA, middle cerebral artery; ACA, anterior cerebral artery; SPS, single point stick; 2WF, 2 wheeled frame; 4WF, 4 wheeled frame; AFO, ankle foot orthosis; SAB, Swedo ankle brace. a Gait Aid is the aid the participant reported using for walking at home/in the community. No SPS/frames were used during testing. For safety participants were allowed to use AFO or SAB. b Did not lead with the unaffected limb for any trials on Test 1 or Test 2. c Unable to complete eight trials on Test 1 due to fatigue. d Did not lead with unaffected limb for any trials on Day 2.
2. Results All 20 participants attended testing on both occasions. All participants led with their affected limb for at least one trial on both tests. Two participants did not lead with their unaffected limb on either test and two participants did not lead with their unaffected limb on Test 2, therefore data for this condition are only available for 16 participants. Six of the 20 participants failed at least one obstacle crossing trial at Test 1 and five failed at least one trial at Test 2; five participants contacted the obstacle with their affected lead limb, three participants contacted the obstacle with their unaffected lead limb and three participants required therapist assistance to maintain balance. Unobstructed gait speed increased from 0.76 m/s (SD = 0.36) at Test 1 to 0.85 m/s (SD = 0.36) at Test 2 (t(19) = 2.373, p = .028). Table 2 illustrates that when leading with the affected limb no significant changes in obstacle crossing speed were observed, although the effect size was medium. In contrast, Table 3 illustrates that when leading with the unaffected limb crossing speed
significantly increased (t(15) = 3.790, p = .002), with a large effect size. No correlations between change in unobstructed gait speed and change in obstacle crossing speed were detected when leading with either the affected (r = .092) or unaffected limb (r = .227). Inspection of scatterplots revealed one outlier when leading with the affected limb (participant 15) and two outliers when leading with the unaffected limb (participants 10 and 15). While correlations improved with outliers removed, they were not significant (affected r = .406, p = .084); unaffected r = .498, p = .070). Further inspection of the spatial and temporal variables confirmed no significant changes to the gait pattern during obstacle crossing when the affected limb led (Table 2). Inspection of effect sizes revealed a medium reduction in trail (unaffected) limb clearance. There were small to medium increases in lead and trail limb pre-obstacle distance and trail post-obstacle distance and a small to medium reduction in lead post-obstacle distance. Remaining variables showed small effect sizes. Data were
Table 2 Gait speed and spatiotemporal variables during obstacle crossing for Test 1 and Test 2 for the affected lead limb (n = 20). Variable
Crossing speed (m/s) Lead pre-obstacle distance (cm) Trail pre-obstacle distance (cm) Lead toe clearance (cm) Trail toe clearance (cm) Lead post-obstacle distance (cm) Trail post-obstacle distance (cm) Lead swing time (s) Trail swing time (s) Double support time (s)
Unimpaireda (n = 12) [3,4]
Stroke
Mean self-selected speed (SD)
Mean matched speed (SD)
Mean Test 1 (SD)
Mean Test 2 (SD)
ES
1.19 (0.21) 116.8 (15.9) 49.9 (4.7) 11.6 (4.8) 9.0 (3.9) 28.4 (8.6) 97.7 (18.0) 0.51 (0.02) 0.46 (0.02) 0.13 (0.03)
0.83 (0.25) 105.3 (13.4) 47.5 (6.3) 11.4 (5.3) 8.4 (5.3) 20.3 (10.7) 77.3 (21.4) 0.63 (0.10) 0.56 (0.07) 0.19 (0.07)
0.72 (0.32) 93.5 (26.7) 47.2 (7.8) 11.3 (4.2) 10.1 (5.8) 13.6 (7.3) 60.5 (24.7) 0.57b (0.23) 0.47b (0.19) 0.15b (0.10)
0.76 (0.34) 96.2 (29.3) 48.8 (9.2) 11.4 (2.9) 9.1 (4.4) 12.5 (8.0) 62.7 (24.6) 0.57b (0.20) 0.47b (0.13) 0.15b (0.15)
0.63 0.25 0.25 0.05 0.42 0.32 0.38 0.04 0.12 0.00
Abbreviations: ES, effect size. a Unimpaired data from previous publications [3,4]. b Median (Interquartile range).
Mean Change (SD) 0.04 (0.09) 2.7 (15.0) 1.5 (9.1) 0.1 (3.0) 1.0 (3.4) 1.1 (4.9) 2.1 (7.7) 0.01b (0.12) 0.00b (0.07) 0.00b (0.05)
p Value .074 .427 .457 .828 .204 .337 .231 .778 .452 1.000
C.M. Said et al. / Gait & Posture 39 (2014) 213–217
216
Table 3 Gait speed and spatiotemporal variables during obstacle crossing for Test 1 and Test 2 for the unaffected lead limb (n = 16). Variable
Crossing speed (m/s) Lead pre-obstacle distance (cm) Trail pre-obstacle distance (cm) Lead toe clearance (cm) Trail toe clearance (cm) Lead post-obstacle distance (cm) Trail post-obstacle distance (cm) Lead swing time (s) Trail swing time (s) Double support time (s)
Unimpaireda (n = 12) [3,4]
Stroke
Mean self-selected speed (SD)
Mean slow speed (SD)
Mean Test 1 (SD)
Mean Test 2 (SD)
ES
1.18 (0.27) 109.1 (19.8) 45.7 (8.8) 11.2 (4.0) 9.9 (4.4) 31.4 (9.2) 101.4 (18.8) 0.50 (0.05) 0.46 (0.04) 0.13 (0.04)
0.86 (0.25) 109.8 (14.0) 47.9 (8.0) 10.6 (4.3) 11.3 (9.7) 20.0 (8.5) 80.4 (19.1) 0.63 (0.12) 0.56 (0.07) 0.19 (.06)
0.78 (0.27) 98.5 (23.7) 48.6 (8.3) 12.6 (4.7) 8.6 (6.0) 13.9 (8.1) 64.0 (24.7) 0.54b (0.17) 0.53b (0.10) 0.16b (0.08)
0.88 (0.29) 105.8 (29.4) 52.5 (9.6) 12.7 (4.4) 8.5 (3.5) 14.3 (6.3) 67.8 (21.6) 0.53b (0.12) 0.49b (0.13) 0.15b (0.06)
1.29 0.66 0.67 0.06 0.04 0.11 0.63 0.34 0.38 0.42
Mean Change (SD) 0.10 (0.11) 7.3 (15.8) 3.9 (8.5) 0.1 (2.6) 0.1 (4.5) 0.4 (5.4) 3.8 (9.4) 0.02b (0.11) 0.02b (0.09) 0.02b (0.04)
p Value .002 .083 .082 .836 .897 .759 .129 .052 .033 .016
Abbreviation: ES, effect size. a Unimpaired data from previous publications [3,4]. b Median (Interquartile range).
re-analysed excluding participants who did not lead with the unaffected limb (i.e. only including participants able to lead with both limbs) with similar results. When leading with the unaffected limb (Table 3), affected trail swing time (Z = 2.135, p = .03) and crossing step double support time (Z = 2.401, p = .016) were significantly reduced, with medium effect sizes. Lead swing time was also reduced, however this did not reach significance (Z = 1.942, p = .052). There were nonsignificant increases of medium effect size for lead and trail limb pre-obstacle distance, and trail post-obstacle distance. 3. Discussion Results demonstrated that while obstacle crossing when leading with the unaffected limb improves over one month with usual rehabilitation, there was no significant improvement when leading with the affected limb. Furthermore, changes in obstacle crossing speed did not correlate with changes in level over-ground gait speed; even with outliers removed the correlations were only of moderate strength and were non-significant. Thus improvements observed in obstacle crossing speed when leading with the unaffected limb cannot be attributed to improvements in level over-ground walking speed. The lack of association between improvement in level ground walking and obstructed walking probably reflect differences between the tasks. Obstacle crossing places additional demands on both limb control and balance compared with unobstructed over-ground walking [6–8]. While this study did not measure changes in limb control or balance, future studies should consider whether improvement in obstacle crossing is associated with changes in limb control or balance. While no significant spatial or temporal differences in obstacle crossing performance were observed when leading with the affected limb, the increase in crossing speed had a medium effect size. Thus more participants or a longer follow-up period may have resulted in a significant finding. Increases of small effect size were also noted for foot placement before and after the obstacle, suggesting that increases in speed were accomplished by modifying step length. In contrast differences in temporal variables had a very small effect size, suggesting that as a group there was little change on these variables. When leading with the unaffected limb there was a significant increase in obstacle crossing speed, accompanied by reductions in affected trail limb swing time and double support. While there were not significant changes in foot placement before or after the obstacle, effect size calculations showed a moderate increase in lead and trail pre-obstacle distance and trail post-obstacle distance. These changes ‘normalise’ deficits previously observed following stroke (Table 3), and contribute to the increase in
obstacle crossing speed. Interestingly, there was no change in lead post-obstacle distance. It has been observed that people with stroke contact the obstacle with the heel of the lead limb [3,10]. Placement of the lead limb closer to the obstacle increases the risk of lead limb contact with the obstacle on landing [3,10], although once lead limb post-obstacle distance is normalised for gait speed, it did not differ between people who fail and those who pass an obstacle crossing task [10]. It is therefore of concern that while other features of the obstacle crossing task improved when leading with the unaffected limb, the potentially risky foot placement did not improve. It is not surprising that lead limb clearance did not change when leading with either the affected or unaffected limb. While some studies found lead limb clearance is increased following stroke [1,16], a previous study using a similar methodology showed no difference between people with stroke and controls [3] and lead clearances observed in the current study are similar to previously reported values. While no significant changes in trail limb clearance were observed, two points should be considered. First, previous research has shown that affected trail limb clearance is reduced in people with stroke [3], which may place them at risk of falls. It would have been desirable to see an increase in affected trail toe clearance, however inspection of data (Table 3) showed no change. Second, there was a moderate reduction in unaffected trail limb clearance at the second test (Table 2). While it is not clear whether this reduction is clinically significant, as reduced toe clearance may be associated with a trip, this finding should be monitored in future studies. The findings suggest that obstacle crossing may require taskrelated training following stroke to address specific identified deficits [17] and to train leading with both the affected and unaffected limb. While obstacle crossing did improve when leading with the unaffected limb, two participants did not lead at all with the unaffected limb and two participants did not lead with the unaffected limb on the second test. While it is not clear why these participants preferred to lead with the affected limb, reluctance to cross an obstacle with the unaffected limb first may reflect a more limited ability to modify the gait pattern. In addition, as previously noted, while there was a trend for most spatial parameters to increase in line with changes in gait speed, unaffected lead post obstacle distance did not increase. There may be a role for training this parameter, particularly in people with very low gait speeds after stroke. There is preliminary evidence that obstacle crossing can be trained in people with stroke and other neurological disorders. Repeated practice of obstacle crossing may be beneficial in people with stroke [18,19] and Parkinson’s Disease [20], although these studies did not include control groups that underwent conventional walking training. Strength training
C.M. Said et al. / Gait & Posture 39 (2014) 213–217
[21,22], Tai Chi [23,24] and exercise that includes training on an obstacle course [25,26] can improve obstacle crossing performance in older people without stroke. Auditory feedback about spatial parameters has been used to train young (mean age 26 years) healthy participants to clear an obstacle on a treadmill with minimal toe clearance [27]. While more research into effective programmes to train obstacle crossing following stroke is required, these studies suggest that obstacle crossing can be trained and provide treatment paradigms that could be explored. The use of safety harnesses and body weight support systems are also being more widely used in clinical practice and could be used to enable people with stroke to practice these more challenging tasks in a safe manner. 3.1. Study limitations The main limitation of this study is the small sample size. This is particularly an issue when leading with the unaffected limb, as four subjects did not lead with the unaffected limb on one or both tests. In addition, a longer time period between test sessions may have resulted in greater improvements in obstacle crossing between the two tests. The one month time period was chosen to reduce the risk of attrition during the study and to ensure participants continued with ‘usual care’ rehabilitation during the study period. Participants were also tested at varying times poststroke, however all were still receiving physiotherapy and thus the treating physiotherapist had determined they were still making improvements in their walking. It is also acknowledged that as rehabilitation during the one month time period was not regulated, it is not known what specific training each participant received. There are currently no recommended strategies to train obstacle crossing in people with stroke, therefore it is unlikely that results are due to a specific training regime. This study did not explore which factors were associated with improvement in obstacle crossing. Future studies should measure changes in limb control, limb strength and balance so that factors associated with change in obstacle crossing can be explored. 4. Conclusions Obstacle crossing improved over a one month time period when leading with the unaffected limb in people with a subacute stroke receiving rehabilitation. Less improvement was observed when leading with the affected limb. Change in obstacle crossing did not correlate with increased level over-ground gait speed. This suggests there is a need to specifically train obstacle crossing in people with stroke. Further studies are needed to evaluate the effectiveness of specific training programmes. Acknowledgements Dr Said received salary support from NHMRC Health Professional Training Fellowship Grant No. 310612 and NHMRC project grant 385002 and a Career Interruption Fellowship from The University of Melbourne. Conflict of interest statement None of the authors have any conflicts of interest associated with this work.
217
References [1] Said CM, Goldie PA, Patla AE, Sparrow WA, Martin KE. Obstacle crossing in subjects with stroke. Archives of Physical Medicine and Rehabilitation 1999;80:1054–9. [2] Den Otter AR, Geurts AC, de Haart M, Mulder T, Duysens J. Step characteristics during obstacle avoidance in hemiplegic stroke. Experimental Brain Research 2005;161(2):180–92. [3] Said CM, Goldie PA, Culham E, Sparrow WA, Patla AE, Morris ME. Control of lead and trail limb clearance during obstacle crossing following stroke. Physical Therapy 2005;85:413–27. [4] Said CM, Goldie PA, Patla AE, Culham E, Sparrow WA, Morris ME. Balance during obstacle crossing following stroke. Gait Posture 2008;27(1):23–30. [5] Park M, Lee S, Lee M, Shim J, Goo B. Plantar foot pressure distribution of stroke patients while crossing obstacles of different heights. Journal of Physical Therapy Science 2010;22:135–9. [6] Chou LS, Kaufman KR, Brey RH, Draganich LF. Motion of the whole body’s center of mass when stepping over obstacles of different heights. Gait Posture 2001;13(1):17–26. [7] Eng JJ, Winter DA, Patla AE. Strategies for recovery from a trip in early and late swing during human walking. Experimental Brain Research 1994;102: 339–49. [8] Patla AE, Rietdyk S. Visual control of limb trajectory over obstacles during locomotion: effect of obstacle height and width. Gait Posture 1993;1: 45–60. [9] Schmid A, Duncan PW, Studenski S, Lai SM, Richards L, Perera S, et al. Improvements in speed-based gait classifications are meaningful. Stroke 2007;38(7):2096–100. [10] Said CM, Galea M, Lythgo N. People with stroke who fail an obstacle crossing task have a higher incidence of falls and utilise different gait patterns compared with people who pass. Physical Therapy 2013;93(3):334–44. [11] Said CM, Galea M, Lythgo N. Obstacle crossing performance does not differ between the first and subsequent attempts in people with stroke. Gait Posture 2009;30:455–8. [12] Davis RB, Ounpuu S, Tyburski D, Gage JR. A gait analysis data collection and reduction technique. Human Movement Science 1991;10:575–87. [13] Kadaba MP, Ramakrishnan HK, Wooten ME. Measurement of lower extremity kinematics during level walking. Journal of Orthopaedic Research 1990;8:383–92. [14] Portney LG, Watkins MP. Foundations of clinical research: applications to practice. 3rd ed. New Jersey: Pearson Education Inc.; 2009. [15] Field AP. Discovering statistics using SPSS. London: Sage; 2005. [16] Lu T-W, Yen H-C, Chen H-L, Hsu W-C, Chen S-C, Hong S-W, et al. Symmetrical kinematic changes in highly functioning older patients post-stroke during obstacle-crossing. Gait Posture 2010;31(4):511–6. [17] Carr J, Shepherd R. Training motor control, increasing strength and fitness and promoting skill acquisition. In: Carr J, Shepherd R, editors. Neurological rehabilitation: optimizing motor performance. 2rd ed., Elsevier: Churchill Livingstone; 2010. [18] Jaffe DL, Brown DA, Pierson-Carey CD, Buckley EL, Lew HL. Stepping over obstacles to improve walking in individuals with poststroke hemiplegia. Journal of Rehabilitation Research & Development 2004;41(3A):283–92. [19] Deprey SM. Random practice to enhance safe compensation of neurological deficits: a case report. Neurology Report 1999;23(2):52–6. [20] Mirelman A, Maidan I, Herman T, Deutsch JE, Giladi N, Hausdorff JM. Virtual reality for gait training: can it induce motor learning to enhance complex walking and reduce fall risk in patients with Parkinson’s disease? Journals of Gerontology Series A Biological Sciences and Medical Sciences 2011;66(2):234–40. [21] Lamoureux E, Sparrow WA, Murphy A, Newton RU. The effects of improved strength on obstacle negotiation in community-living older adults. Gait Posture 2003;17(3):273–83. [22] Lamoureux EL, Murphy A, Sparrow A, Newton RU. The effects of progressive resistance training on obstructed-gait tasks in community-living older adults. Journal of Aging and Physical Activity 2003;11(1):98–110. [23] Kim H. Effects of Tai Chi exercise on the center of pressure trace during obstacle crossing in older adults who are at a risk of falling. Journal of Physical Therapy Science 2009;21(1):49–54. [24] Zhang C, Mao D, Riskowski JL, Song Q. Strategies of stepping over obstacles: the effects of long-term exercise in older adults. Gait Posture 2011;34(2):191–6. [25] Weerdesteyn V, Nienhuis B, Duysens J. Exercise training can improve spatial characteristics of time-critical obstacle avoidance in elderly people. Human Movement Science 2008;27(5):738–48. [26] Weerdesteyn V, Rijken H, Geurts ACH, Smits-Engelsman BCM, Mulder T, Duysens J. A five-week exercise program can reduce falls and improve obstacle avoidance in the elderly. Gerontology 2006;52(3):131. [27] Lam T, Dietz V. Transfer of motor performance in an obstacle avoidance task to different walking conditions. Journal of Neurophysiology 2004;92:2010–6.
Contents lists available at SciVerse ScienceDirect
Gait & Posture journal homepage: www.elsevier.com/locate/gaitpost
Obstacle crossing following stroke improves over one month when the unaffected limb leads, but not when the affected limb leads§ Catherine M. Said a,b,*, Mary Galea c, Noel Lythgo d a
Physiotherapy, Melbourne School of Health Sciences, University of Melbourne, Parkville, Victoria, Australia Physiotherapy Department, Austin Health, Heidelberg West, Victoria, Australia c Department of Medicine, University of Melbourne, c/- Royal Melbourne Hospital, Parkville, Victoria, Australia d School of Medical Sciences, RMIT University, Victoria, Australia b
A R T I C L E I N F O
A B S T R A C T
Article history: Received 15 January 2013 Received in revised form 24 May 2013 Accepted 8 July 2013
While it is well established that obstacle crossing is impaired following stroke, it is not known whether obstacle crossing improves as gait improves following stroke. The purpose of this study was to determine whether obstacle crossing changed over a one month time period in people with a recent stroke. Twenty participants receiving rehabilitation following a recent stroke were tested on two occasions one month apart. Participants received usual care rehabilitation, including physiotherapy, between the tests. The main outcome measure was obstacle crossing speed as participants stepped over a 4-cm high obstacle. Secondary measures were spatiotemporal variables. Data were collected via a three dimensional motion analysis system. When leading with the affected limb no changes in obstacle crossing speed or spatiotemporal variables were observed over the one month period. When leading with the unaffected limb, crossing speed significantly increased (p = .002), and affected trail limb swing time (p = .03) and crossing step double support time reduced (p = .016). While not significant, the lead and trail limb preobstacle distance increased (p = .08), and lead swing time (p = .052) reduced. Change in obstacle crossing speed did not correlate with change in level gait speed. Obstacle crossing does not necessarily improve over a one month time period in people receiving rehabilitation following stroke. These findings suggest that there may be a need for more targeted training of obstacle crossing, particularly when leading with the affected limb. ß 2013 Elsevier B.V. All rights reserved.
Keywords: Stroke Gait disorder Neurologic Rehabilitation Obstacle crossing
Obstacle crossing is impaired for many people following a stroke [1,2]. Said et al. found over 50% of people able to walk and receiving rehabilitation following stroke either contacted the obstacle or lost balance when attempting to clear a small obstacle [1]. People with stroke also utilised different movement patterns to clear an obstacle compared to unimpaired participants [3,4]. They had significantly slower gait speeds as they crossed the obstacle, which accounted for some differences in the gait pattern such as lead limb placement before the obstacle [3]. However, speed did not account for all gait adjustments. Compared with healthy participants walking at matched speed, people with stroke placed the unaffected lead limb and affected and unaffected trail limb closer to the obstacle after crossing [3]. They also positioned their
§ This data was presented in part at the Australian Physiotherapy Conference Week, Brisbane, 2011. * Corresponding author at: Melbourne School of Health Sciences, The University of Melbourne, Parkville 3010, Australia. Tel.: +61 3 9496 2055. E-mail addresses: [email protected], [email protected] (C.M. Said).
0966-6362/$ – see front matter ß 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.gaitpost.2013.07.008
centre of mass closer to the base of support when leading with the affected limb [4]. Many of these studies assessed people undergoing rehabilitation within six months of stroke, thus some people may not have attained optimal walking recovery [3–5]. While obstacle crossing is more challenging than level ground walking [6–8], it is fundamentally a locomotor task. It is reasonable to expect that if walking ability improves, obstacle crossing may also improve. Gait speed is an important clinical marker of gait improvement [9], and there are valid rationales for anticipating that increased level over-ground gait speed would lead to improvements in obstacle crossing. Some movement deficits during obstacle crossing following stroke were partly speed associated [3,4] and people who fail an obstacle crossing task cross the obstacle more slowly compared with people who pass [10]. Thus if gait speed improves following stroke, aspects of obstacle crossing performance may also normalise. In addition, for some participants, the obstacle crossing task may have been ‘novel’ following their stroke, as their exposure to complex walking tasks may have been limited at this stage of recovery. With repeated exposure to complex locomotor
214
C.M. Said et al. / Gait & Posture 39 (2014) 213–217
tasks, including obstacle crossing, people with subacute stroke may change their obstacle crossing strategy. We therefore wanted to investigate whether obstacle crossing changes over time in people with subacute stroke. While many variables have been examined during obstacle crossing, the primary outcome of interest for this study was gait speed during obstacle crossing, as this differentiates between people with stroke and controls [3]. It is also associated with failure on an obstacle crossing task [10]. To explore mechanisms behind observed changes, temporal and sagittal plane spatial variables were examined. The purpose of this study was to determine whether obstacle crossing gait speed improves over time in people recovering from a recent stroke. It was hypothesised that if level over-ground (unobstructed) walking speed increased following stroke, it would be associated with an increase in obstacle crossing gait speed. It was also anticipated that spatial and temporal variables would change and approach values achieved by unimpaired adults. 1. Method This was an observational study. Ethics approval was obtained and participants provided informed consent. Obstacle crossing performance and falls, obstacle crossing performance and spatiotemporal characteristics, and variability in spatiotemporal characteristics in these participants have been previously reported [10,11]. 1.1. Participants Data were collected from 20 participants recruited from two hospital physiotherapy departments. A sample of 20 was sufficient to detect a large difference between the two tests (d = 0.8), with a power of 70% and a two tailed alpha of 0.05. Participants were receiving ongoing physiotherapy for a gait or balance disorder following a recent stroke, and capable of walking 10 m without a gait aid or physical assistance. Participants were excluded if they had other medical, musculoskeletal or neurological conditions that may have impacted walking. Mean age was 61.1 years (SD = 15), mean height 167.9 cm (SD = 9.23) and participants were first tested a median of 60.7 days (SD = 47.2) post stroke. Participant characteristics are provided in Table 1. 1.2. Apparatus Data were recorded by a six or eight camera VICON 612 3D motion system1 and AMTI forceplate.2 During the obstructed trials a red coloured balsa wood obstacle measuring 40 mm high 1.5 mm thick 600 mm long was positioned after the forceplate, approximately 5 m from the start. Data processing utilised Vicon BodyBuilder1 Version 3.55 (build 136).1 1.3. Procedure Participants wore loose fitting shorts, walking shoes and any prescription eyewear. Anthropometric measures were obtained (Vicon Plug-in Gait Product Guide1) and twenty-one 14 mm passive reflective markers were placed on the lower limbs, [12,13] acromions and obstacle as previously described [3]. Participants performed four unobstructed walking trials at selfselected speed, followed by eight trials with the obstacle. They were instructed to walk at self-selected speed and step over the 1 Oxford Metrics Ltd., 14 Minns Estate, West Way, Oxford, OX2OJB, United.ingdom. 2 Advanced Mechanical Technology, Inc.: Watertown, MA.
obstacle without contacting it or losing balance. Participants were accompanied by a therapist, who walked behind and to the side of the subject lightly holding a safety belt. Assistance was only provided if required. Participants repeated the test procedure one month later (mean 29.5 days, SD = 5.4). During this time they received their usual physiotherapy. No attempt was made to standardise or prescribe the type of treatment received; however, as ‘receiving physiotherapy for a gait or balance disorder’ was a criterion for inclusion, it was assumed that a portion of therapy was directed towards these issues. 1.4. Gait variables: data processing For each participant, one trial leading with the affected limb and one trial leading with the unaffected limb were analysed for each test. The first trial with adequate data (minimal marker occlusion and clean forceplate strike, if available) was selected. The gait pattern utilised by people with stroke to cross an obstacle was fairly consistent over three attempts within a single session [11], so the selection of one trial for analysis was justified. Trials were excluded from motion analysis if the participant required assistance to maintain balance. Data were filtered using Woltring filtering routine with a predicted Mean-Squared-Error value of 20. BodyBuilder11 was used to create ‘virtual markers’ on the shoe at the most distal point of the toe and heel. Data were exported to Microsoft Excel to calculate obstacle crossing gait speed (trail heel contact preobstacle to trail heel contact post-obstacle), lead and trail limb preobstacle horizontal distance (lead or trail heel position before the obstacle to the obstacle), vertical toe clearance (top of the obstacle to the lead or trail toe) and post-obstacle horizontal distance (lead or trail heel position after the obstacle to the obstacle). Foot contact and toe off events were obtained from BodyBuilder1 by visual inspection of the position of the virtual markers on the heel and toe. Lead limb swing time was from lead limb toe off pre-obstacle to lead limb foot contact post-obstacle. Trail limb swing time was from trail limb toe off pre-obstacle to foot contact post-obstacle. Crossing double support time was from lead foot contact postobstacle to trail toe off pre-obstacle. More details on data processing have been previously reported [3]. 1.5. Statistical analysis Statistical analyses were undertaken using IBM SPSS Version 20 for Windows.3 Paired t tests were used to determine whether level over-ground walking speed had changed between the two test sessions. To determine whether obstacle crossing gait speed or spatiotemporal variables changed over the one month period, data from trials leading with the affected limb and unaffected limb were analysed separately. While most spatial data were normally distributed, temporal data were skewed. Therefore paired t tests were used for spatial data and Wilcoxon Signed Rank tests were used for temporal data. Effect sizes were calculated for all data. For parametric tests, effect sizes of 0.8, 0.5 and 0.2 were interpreted as large, medium and small respectively [14]. For nonparametric tests, effect sizes of 0.5, 0.3 and 0.1 were interpreted as large, medium and small respectively [15]. To determine whether changes in obstacle crossing speed were associated with changes in level over-ground walking speed, correlations between the difference scores for the two tests were calculated using Pearson’s r. As this was an exploratory study and risks associated with a Type I error were low, no corrections for multiple tests were applied. 3
IBM SPSS Incß IBM Corporation.
C.M. Said et al. / Gait & Posture 39 (2014) 213–217
215
Table 1 Characteristics of participants. Participant no.
Age (y)
Lesion site
Days post-stroke
Gait aida
Gait speed (m/s)
1 2 3 4 5 6 7 8b 9c,d 10 11 12 13 14 15 16b 17d 18 19 20
75 33 57 65 60 26 85 49 74 70 52 71 73 70 62 43 68 76 55 58
R subcortical infarct L frontal infarct R basal ganglia haemorrhage R thalamic haemorrhage L corona radiata infarct R MCA infarct R MCA infarct R subcortical stroke L post limb internal capsule infarct Brainstem stroke R pontine, L corona radiata L MCA infarct Multiple infarcts both hemispheres and cerebellum L MCA infarct L basal ganglia haemorrhage L putamen/corona radiata lacunar infarct L pontine infarct R MCA/ACA infarct R occipitoparietal haemorrhage L medullary infarct L MCA infarct
10 42 34 28 12 155 48 58 97 27 62 55 50 50 77 207 72 35 66 28
Nil Nil Nil Nil Nil Nil SPS AFO, SPS AFO, SPS Nil Nil 4WF 2WF 4WF SPS SAB, SPS Nil Nil Nil Nil
1.1 1.1 0.8 0.9 0.8 0.9 0.4 0.2 0.2 0.8 1.1 0.6 0.6 0.1 0.4 0.8 0.6 1.2 1.3 1.1
Adapted from Said et al. [10], with permission of the American Physical Therapy Association. ß2013 American Physical Therapy Association. Abbreviations: R, right; L, left; MCA, middle cerebral artery; ACA, anterior cerebral artery; SPS, single point stick; 2WF, 2 wheeled frame; 4WF, 4 wheeled frame; AFO, ankle foot orthosis; SAB, Swedo ankle brace. a Gait Aid is the aid the participant reported using for walking at home/in the community. No SPS/frames were used during testing. For safety participants were allowed to use AFO or SAB. b Did not lead with the unaffected limb for any trials on Test 1 or Test 2. c Unable to complete eight trials on Test 1 due to fatigue. d Did not lead with unaffected limb for any trials on Day 2.
2. Results All 20 participants attended testing on both occasions. All participants led with their affected limb for at least one trial on both tests. Two participants did not lead with their unaffected limb on either test and two participants did not lead with their unaffected limb on Test 2, therefore data for this condition are only available for 16 participants. Six of the 20 participants failed at least one obstacle crossing trial at Test 1 and five failed at least one trial at Test 2; five participants contacted the obstacle with their affected lead limb, three participants contacted the obstacle with their unaffected lead limb and three participants required therapist assistance to maintain balance. Unobstructed gait speed increased from 0.76 m/s (SD = 0.36) at Test 1 to 0.85 m/s (SD = 0.36) at Test 2 (t(19) = 2.373, p = .028). Table 2 illustrates that when leading with the affected limb no significant changes in obstacle crossing speed were observed, although the effect size was medium. In contrast, Table 3 illustrates that when leading with the unaffected limb crossing speed
significantly increased (t(15) = 3.790, p = .002), with a large effect size. No correlations between change in unobstructed gait speed and change in obstacle crossing speed were detected when leading with either the affected (r = .092) or unaffected limb (r = .227). Inspection of scatterplots revealed one outlier when leading with the affected limb (participant 15) and two outliers when leading with the unaffected limb (participants 10 and 15). While correlations improved with outliers removed, they were not significant (affected r = .406, p = .084); unaffected r = .498, p = .070). Further inspection of the spatial and temporal variables confirmed no significant changes to the gait pattern during obstacle crossing when the affected limb led (Table 2). Inspection of effect sizes revealed a medium reduction in trail (unaffected) limb clearance. There were small to medium increases in lead and trail limb pre-obstacle distance and trail post-obstacle distance and a small to medium reduction in lead post-obstacle distance. Remaining variables showed small effect sizes. Data were
Table 2 Gait speed and spatiotemporal variables during obstacle crossing for Test 1 and Test 2 for the affected lead limb (n = 20). Variable
Crossing speed (m/s) Lead pre-obstacle distance (cm) Trail pre-obstacle distance (cm) Lead toe clearance (cm) Trail toe clearance (cm) Lead post-obstacle distance (cm) Trail post-obstacle distance (cm) Lead swing time (s) Trail swing time (s) Double support time (s)
Unimpaireda (n = 12) [3,4]
Stroke
Mean self-selected speed (SD)
Mean matched speed (SD)
Mean Test 1 (SD)
Mean Test 2 (SD)
ES
1.19 (0.21) 116.8 (15.9) 49.9 (4.7) 11.6 (4.8) 9.0 (3.9) 28.4 (8.6) 97.7 (18.0) 0.51 (0.02) 0.46 (0.02) 0.13 (0.03)
0.83 (0.25) 105.3 (13.4) 47.5 (6.3) 11.4 (5.3) 8.4 (5.3) 20.3 (10.7) 77.3 (21.4) 0.63 (0.10) 0.56 (0.07) 0.19 (0.07)
0.72 (0.32) 93.5 (26.7) 47.2 (7.8) 11.3 (4.2) 10.1 (5.8) 13.6 (7.3) 60.5 (24.7) 0.57b (0.23) 0.47b (0.19) 0.15b (0.10)
0.76 (0.34) 96.2 (29.3) 48.8 (9.2) 11.4 (2.9) 9.1 (4.4) 12.5 (8.0) 62.7 (24.6) 0.57b (0.20) 0.47b (0.13) 0.15b (0.15)
0.63 0.25 0.25 0.05 0.42 0.32 0.38 0.04 0.12 0.00
Abbreviations: ES, effect size. a Unimpaired data from previous publications [3,4]. b Median (Interquartile range).
Mean Change (SD) 0.04 (0.09) 2.7 (15.0) 1.5 (9.1) 0.1 (3.0) 1.0 (3.4) 1.1 (4.9) 2.1 (7.7) 0.01b (0.12) 0.00b (0.07) 0.00b (0.05)
p Value .074 .427 .457 .828 .204 .337 .231 .778 .452 1.000
C.M. Said et al. / Gait & Posture 39 (2014) 213–217
216
Table 3 Gait speed and spatiotemporal variables during obstacle crossing for Test 1 and Test 2 for the unaffected lead limb (n = 16). Variable
Crossing speed (m/s) Lead pre-obstacle distance (cm) Trail pre-obstacle distance (cm) Lead toe clearance (cm) Trail toe clearance (cm) Lead post-obstacle distance (cm) Trail post-obstacle distance (cm) Lead swing time (s) Trail swing time (s) Double support time (s)
Unimpaireda (n = 12) [3,4]
Stroke
Mean self-selected speed (SD)
Mean slow speed (SD)
Mean Test 1 (SD)
Mean Test 2 (SD)
ES
1.18 (0.27) 109.1 (19.8) 45.7 (8.8) 11.2 (4.0) 9.9 (4.4) 31.4 (9.2) 101.4 (18.8) 0.50 (0.05) 0.46 (0.04) 0.13 (0.04)
0.86 (0.25) 109.8 (14.0) 47.9 (8.0) 10.6 (4.3) 11.3 (9.7) 20.0 (8.5) 80.4 (19.1) 0.63 (0.12) 0.56 (0.07) 0.19 (.06)
0.78 (0.27) 98.5 (23.7) 48.6 (8.3) 12.6 (4.7) 8.6 (6.0) 13.9 (8.1) 64.0 (24.7) 0.54b (0.17) 0.53b (0.10) 0.16b (0.08)
0.88 (0.29) 105.8 (29.4) 52.5 (9.6) 12.7 (4.4) 8.5 (3.5) 14.3 (6.3) 67.8 (21.6) 0.53b (0.12) 0.49b (0.13) 0.15b (0.06)
1.29 0.66 0.67 0.06 0.04 0.11 0.63 0.34 0.38 0.42
Mean Change (SD) 0.10 (0.11) 7.3 (15.8) 3.9 (8.5) 0.1 (2.6) 0.1 (4.5) 0.4 (5.4) 3.8 (9.4) 0.02b (0.11) 0.02b (0.09) 0.02b (0.04)
p Value .002 .083 .082 .836 .897 .759 .129 .052 .033 .016
Abbreviation: ES, effect size. a Unimpaired data from previous publications [3,4]. b Median (Interquartile range).
re-analysed excluding participants who did not lead with the unaffected limb (i.e. only including participants able to lead with both limbs) with similar results. When leading with the unaffected limb (Table 3), affected trail swing time (Z = 2.135, p = .03) and crossing step double support time (Z = 2.401, p = .016) were significantly reduced, with medium effect sizes. Lead swing time was also reduced, however this did not reach significance (Z = 1.942, p = .052). There were nonsignificant increases of medium effect size for lead and trail limb pre-obstacle distance, and trail post-obstacle distance. 3. Discussion Results demonstrated that while obstacle crossing when leading with the unaffected limb improves over one month with usual rehabilitation, there was no significant improvement when leading with the affected limb. Furthermore, changes in obstacle crossing speed did not correlate with changes in level over-ground gait speed; even with outliers removed the correlations were only of moderate strength and were non-significant. Thus improvements observed in obstacle crossing speed when leading with the unaffected limb cannot be attributed to improvements in level over-ground walking speed. The lack of association between improvement in level ground walking and obstructed walking probably reflect differences between the tasks. Obstacle crossing places additional demands on both limb control and balance compared with unobstructed over-ground walking [6–8]. While this study did not measure changes in limb control or balance, future studies should consider whether improvement in obstacle crossing is associated with changes in limb control or balance. While no significant spatial or temporal differences in obstacle crossing performance were observed when leading with the affected limb, the increase in crossing speed had a medium effect size. Thus more participants or a longer follow-up period may have resulted in a significant finding. Increases of small effect size were also noted for foot placement before and after the obstacle, suggesting that increases in speed were accomplished by modifying step length. In contrast differences in temporal variables had a very small effect size, suggesting that as a group there was little change on these variables. When leading with the unaffected limb there was a significant increase in obstacle crossing speed, accompanied by reductions in affected trail limb swing time and double support. While there were not significant changes in foot placement before or after the obstacle, effect size calculations showed a moderate increase in lead and trail pre-obstacle distance and trail post-obstacle distance. These changes ‘normalise’ deficits previously observed following stroke (Table 3), and contribute to the increase in
obstacle crossing speed. Interestingly, there was no change in lead post-obstacle distance. It has been observed that people with stroke contact the obstacle with the heel of the lead limb [3,10]. Placement of the lead limb closer to the obstacle increases the risk of lead limb contact with the obstacle on landing [3,10], although once lead limb post-obstacle distance is normalised for gait speed, it did not differ between people who fail and those who pass an obstacle crossing task [10]. It is therefore of concern that while other features of the obstacle crossing task improved when leading with the unaffected limb, the potentially risky foot placement did not improve. It is not surprising that lead limb clearance did not change when leading with either the affected or unaffected limb. While some studies found lead limb clearance is increased following stroke [1,16], a previous study using a similar methodology showed no difference between people with stroke and controls [3] and lead clearances observed in the current study are similar to previously reported values. While no significant changes in trail limb clearance were observed, two points should be considered. First, previous research has shown that affected trail limb clearance is reduced in people with stroke [3], which may place them at risk of falls. It would have been desirable to see an increase in affected trail toe clearance, however inspection of data (Table 3) showed no change. Second, there was a moderate reduction in unaffected trail limb clearance at the second test (Table 2). While it is not clear whether this reduction is clinically significant, as reduced toe clearance may be associated with a trip, this finding should be monitored in future studies. The findings suggest that obstacle crossing may require taskrelated training following stroke to address specific identified deficits [17] and to train leading with both the affected and unaffected limb. While obstacle crossing did improve when leading with the unaffected limb, two participants did not lead at all with the unaffected limb and two participants did not lead with the unaffected limb on the second test. While it is not clear why these participants preferred to lead with the affected limb, reluctance to cross an obstacle with the unaffected limb first may reflect a more limited ability to modify the gait pattern. In addition, as previously noted, while there was a trend for most spatial parameters to increase in line with changes in gait speed, unaffected lead post obstacle distance did not increase. There may be a role for training this parameter, particularly in people with very low gait speeds after stroke. There is preliminary evidence that obstacle crossing can be trained in people with stroke and other neurological disorders. Repeated practice of obstacle crossing may be beneficial in people with stroke [18,19] and Parkinson’s Disease [20], although these studies did not include control groups that underwent conventional walking training. Strength training
C.M. Said et al. / Gait & Posture 39 (2014) 213–217
[21,22], Tai Chi [23,24] and exercise that includes training on an obstacle course [25,26] can improve obstacle crossing performance in older people without stroke. Auditory feedback about spatial parameters has been used to train young (mean age 26 years) healthy participants to clear an obstacle on a treadmill with minimal toe clearance [27]. While more research into effective programmes to train obstacle crossing following stroke is required, these studies suggest that obstacle crossing can be trained and provide treatment paradigms that could be explored. The use of safety harnesses and body weight support systems are also being more widely used in clinical practice and could be used to enable people with stroke to practice these more challenging tasks in a safe manner. 3.1. Study limitations The main limitation of this study is the small sample size. This is particularly an issue when leading with the unaffected limb, as four subjects did not lead with the unaffected limb on one or both tests. In addition, a longer time period between test sessions may have resulted in greater improvements in obstacle crossing between the two tests. The one month time period was chosen to reduce the risk of attrition during the study and to ensure participants continued with ‘usual care’ rehabilitation during the study period. Participants were also tested at varying times poststroke, however all were still receiving physiotherapy and thus the treating physiotherapist had determined they were still making improvements in their walking. It is also acknowledged that as rehabilitation during the one month time period was not regulated, it is not known what specific training each participant received. There are currently no recommended strategies to train obstacle crossing in people with stroke, therefore it is unlikely that results are due to a specific training regime. This study did not explore which factors were associated with improvement in obstacle crossing. Future studies should measure changes in limb control, limb strength and balance so that factors associated with change in obstacle crossing can be explored. 4. Conclusions Obstacle crossing improved over a one month time period when leading with the unaffected limb in people with a subacute stroke receiving rehabilitation. Less improvement was observed when leading with the affected limb. Change in obstacle crossing did not correlate with increased level over-ground gait speed. This suggests there is a need to specifically train obstacle crossing in people with stroke. Further studies are needed to evaluate the effectiveness of specific training programmes. Acknowledgements Dr Said received salary support from NHMRC Health Professional Training Fellowship Grant No. 310612 and NHMRC project grant 385002 and a Career Interruption Fellowship from The University of Melbourne. Conflict of interest statement None of the authors have any conflicts of interest associated with this work.
217
References [1] Said CM, Goldie PA, Patla AE, Sparrow WA, Martin KE. Obstacle crossing in subjects with stroke. Archives of Physical Medicine and Rehabilitation 1999;80:1054–9. [2] Den Otter AR, Geurts AC, de Haart M, Mulder T, Duysens J. Step characteristics during obstacle avoidance in hemiplegic stroke. Experimental Brain Research 2005;161(2):180–92. [3] Said CM, Goldie PA, Culham E, Sparrow WA, Patla AE, Morris ME. Control of lead and trail limb clearance during obstacle crossing following stroke. Physical Therapy 2005;85:413–27. [4] Said CM, Goldie PA, Patla AE, Culham E, Sparrow WA, Morris ME. Balance during obstacle crossing following stroke. Gait Posture 2008;27(1):23–30. [5] Park M, Lee S, Lee M, Shim J, Goo B. Plantar foot pressure distribution of stroke patients while crossing obstacles of different heights. Journal of Physical Therapy Science 2010;22:135–9. [6] Chou LS, Kaufman KR, Brey RH, Draganich LF. Motion of the whole body’s center of mass when stepping over obstacles of different heights. Gait Posture 2001;13(1):17–26. [7] Eng JJ, Winter DA, Patla AE. Strategies for recovery from a trip in early and late swing during human walking. Experimental Brain Research 1994;102: 339–49. [8] Patla AE, Rietdyk S. Visual control of limb trajectory over obstacles during locomotion: effect of obstacle height and width. Gait Posture 1993;1: 45–60. [9] Schmid A, Duncan PW, Studenski S, Lai SM, Richards L, Perera S, et al. Improvements in speed-based gait classifications are meaningful. Stroke 2007;38(7):2096–100. [10] Said CM, Galea M, Lythgo N. People with stroke who fail an obstacle crossing task have a higher incidence of falls and utilise different gait patterns compared with people who pass. Physical Therapy 2013;93(3):334–44. [11] Said CM, Galea M, Lythgo N. Obstacle crossing performance does not differ between the first and subsequent attempts in people with stroke. Gait Posture 2009;30:455–8. [12] Davis RB, Ounpuu S, Tyburski D, Gage JR. A gait analysis data collection and reduction technique. Human Movement Science 1991;10:575–87. [13] Kadaba MP, Ramakrishnan HK, Wooten ME. Measurement of lower extremity kinematics during level walking. Journal of Orthopaedic Research 1990;8:383–92. [14] Portney LG, Watkins MP. Foundations of clinical research: applications to practice. 3rd ed. New Jersey: Pearson Education Inc.; 2009. [15] Field AP. Discovering statistics using SPSS. London: Sage; 2005. [16] Lu T-W, Yen H-C, Chen H-L, Hsu W-C, Chen S-C, Hong S-W, et al. Symmetrical kinematic changes in highly functioning older patients post-stroke during obstacle-crossing. Gait Posture 2010;31(4):511–6. [17] Carr J, Shepherd R. Training motor control, increasing strength and fitness and promoting skill acquisition. In: Carr J, Shepherd R, editors. Neurological rehabilitation: optimizing motor performance. 2rd ed., Elsevier: Churchill Livingstone; 2010. [18] Jaffe DL, Brown DA, Pierson-Carey CD, Buckley EL, Lew HL. Stepping over obstacles to improve walking in individuals with poststroke hemiplegia. Journal of Rehabilitation Research & Development 2004;41(3A):283–92. [19] Deprey SM. Random practice to enhance safe compensation of neurological deficits: a case report. Neurology Report 1999;23(2):52–6. [20] Mirelman A, Maidan I, Herman T, Deutsch JE, Giladi N, Hausdorff JM. Virtual reality for gait training: can it induce motor learning to enhance complex walking and reduce fall risk in patients with Parkinson’s disease? Journals of Gerontology Series A Biological Sciences and Medical Sciences 2011;66(2):234–40. [21] Lamoureux E, Sparrow WA, Murphy A, Newton RU. The effects of improved strength on obstacle negotiation in community-living older adults. Gait Posture 2003;17(3):273–83. [22] Lamoureux EL, Murphy A, Sparrow A, Newton RU. The effects of progressive resistance training on obstructed-gait tasks in community-living older adults. Journal of Aging and Physical Activity 2003;11(1):98–110. [23] Kim H. Effects of Tai Chi exercise on the center of pressure trace during obstacle crossing in older adults who are at a risk of falling. Journal of Physical Therapy Science 2009;21(1):49–54. [24] Zhang C, Mao D, Riskowski JL, Song Q. Strategies of stepping over obstacles: the effects of long-term exercise in older adults. Gait Posture 2011;34(2):191–6. [25] Weerdesteyn V, Nienhuis B, Duysens J. Exercise training can improve spatial characteristics of time-critical obstacle avoidance in elderly people. Human Movement Science 2008;27(5):738–48. [26] Weerdesteyn V, Rijken H, Geurts ACH, Smits-Engelsman BCM, Mulder T, Duysens J. A five-week exercise program can reduce falls and improve obstacle avoidance in the elderly. Gerontology 2006;52(3):131. [27] Lam T, Dietz V. Transfer of motor performance in an obstacle avoidance task to different walking conditions. Journal of Neurophysiology 2004;92:2010–6.