- Email: [email protected]
Effect of Balance Support on the Energy Cost of Walking After Stroke
Archives of Physical Medicine and Rehabilitation journal homepage: www.archives-pmr.org Archives of Physical Medicine and Rehabilitation 2013;94:2255-61
ORIGINAL ARTICLE
Effect of Balance Support on the Energy Cost of Walking After Stroke Trienke IJmker, MSc,a,b Han Houdijk, PhD,a,b Claudine J. Lamoth, PhD,c Ameerani V. Jarbandhan, MSc,a Danie¨lle Rijntjes, PT,b Peter J. Beek, PhD,a Lucas H. van der Woude, PhDc From the aMOVE Research Institute Amsterdam, Faculty of Human Movement Sciences, VU University Amsterdam, Amsterdam, bHeliomare Research and Development, Wijk aan Zee and cCenter for Human Movement Sciences and Center for Rehabilitation, University Medical Center Groningen, University of Groningen, Groningen, The Netherlands.
Abstract Objective: To examine the influence of balance support on the energy cost of treadmill and overground walking in ambulatory patients with stroke. Design: Cross-sectional. Setting: Research laboratory at a rehabilitation center. Participants: Patients with stroke depending on a walking aid in daily life (nZ12; walking aid dependent ambulators) and walking aid independent ambulators (nZ12), all able to walk for at least 5 minutes. Interventions: Not applicable. Main Outcome Measures: Energy cost (J$kg1$m1) and temporal gait parameters (walking speed, mean and coefficient of variation of stride time, and symmetry index) were obtained during 4 walking trials at preferred walking speed: overground with and without a cane and on a treadmill with and without handrail support. Results: On the treadmill, handrail support resulted in a significant decrease in energy cost of 16%, independent of the group. Although walking aid dependent ambulators had on average a larger reduction in energy cost than walking aid independent ambulators (19% vs 14%), this interaction did not reach statistical significance (PZ.11). Interestingly, overground walking with support resulted in an 8% reduction in energy cost for walking aid dependent ambulators, but a 6% increase for walking aid independent ambulators. The reduction in energy cost with support was accompanied by changes in temporal gait parameters, most notably an increase in stride time and symmetry and a decrease in stride time variability. Conclusions: Balance support can result in a significant reduction in the energy cost of walking in stroke patients, the magnitude of which depends on walking ability and the walking task. Impaired balance control should not be overlooked as a contributing factor to the increased energy cost of walking in patients with stroke, and improving or assisting balance control should be considered to reduce the energy cost of hemiplegic gait. Archives of Physical Medicine and Rehabilitation 2013;94:2255-61 ª 2013 by the American Congress of Rehabilitation Medicine
Stroke is one of the leading causes of death and disability in the Western world for people >65 years.1 It can result in various physical and neuropsychologic impairments that limit Presented to the Joint World Congress of International Society for Posture and Gait Research and Gait and Mental Function, June 25, 2012, Trondheim, Norway; and the European Society for Movement Analysis in Adults and Children, September 13e14, 2012, Stockholm, Sweden. No commercial party having a direct financial interest in the results of the research supporting this article has or will confer a benefit on the authors or on any organization with which the authors are associated.
performance of activities of daily living and participation in society. Two major physical problems that often arise in this population are poststroke fatigue and difficulties with maintaining balance.2 Poststroke fatigue is thought to be a multidimensional concept involving physical as well as neuropsychologic factors.2,3 Physiologically, fatigue can be seen as the consequence of an imbalance between physical capacity, or energy resources, and energy demands.2 In patients with stroke, physical capacity is often
0003-9993/13/$36 - see front matter ª 2013 by the American Congress of Rehabilitation Medicine http://dx.doi.org/10.1016/j.apmr.2013.04.022
2256
T. IJmker et al
decreased because of deconditioning,4,5 whereas the energy demand of activities of daily life is frequently increased. For instance, patients with stroke often have an energy cost of walking (expressed in J$kg1$m1) of up to twice the cost of healthy counterparts walking at the same speed.6,7 The nature of this increase in the energy cost of walking is not fully understood. Increased muscle work because of stroke-related impairments, such as increased muscle tone and spasticity, and compensatory strategies, such as increased displacement of the center of mass, hip hiking, and circumduction of the paretic leg, have all been related to this increase.6,8-10 However, in these analyses, the potential metabolic effort for maintaining balance and ensuring a stable gait pattern is often overlooked. Literature has shown that maintaining balance during gait involves a metabolic energy cost even in healthy subjects.11-13 Moreover, previous research suggests that balance problems, which are often seen in stroke survivors, can result in an increased metabolic effort for balance maintenance. It has been shown that balance perturbations during upright standing, for instance standing on foam, lead to an increase in energy expenditure, which is up to twice as high for patients with stroke than for healthy controls.14 If perturbing balance in patients with stroke elicits an extra metabolic demand, supporting balance, by contrast, could lead to a decrease in energy cost. This should especially be true for stroke survivors with more residual deficits affecting balance control, such as those relying on a walking aid, and for more challenging walking conditions. Therefore, the purpose of this study was to investigate the effect of balance support on the energy cost of walking in walking aid independent and walking aid dependent ambulatory patients with stroke, and whether this effect was similar for treadmill and overground walking. We hypothesized that providing support would decrease the energy cost of walking in patients with stroke, and that this decrease would be larger for patients depending on a walking aid in daily life than for independent ambulators. For patients with stroke, the energy cost of walking is generally higher on the treadmill than overground,15 and treadmill walking is claimed to be more challenging to balance control for these patients because of altered visual and proprioceptive input, the moving support surface, and the constant beltspeed. We thus hypothesized that for patients with stroke, the effect of support would be larger on the treadmill than overground.
Methods Participants Subjects were recruited from the stroke unit at the rehabilitation center Heliomare (Wijk aan Zee, the Netherlands). From the total number of stroke patients admitted to the rehabilitation center
List of abbreviations: BBS CV Egross FAC RER SI TMWT VO2
Berg Balance Scale coefficient of variation gross metabolic energy expenditure Functional Ambulatory Category respiratory exchange ratio symmetry index ten-meter walk test oxygen consumption
during the time of the measurements (nZ88 over 2.5mo), 24 patients (mean age SD, 5215.9y; 14 men, 10 women) who fulfilled the inclusion criteria and were available for measurements, agreed to participate. All patients were able to walk for at least 3 minutes without, and 5 minutes with support of an assistive device overground and on a treadmill, and achieved a Functional Ambulatory Category (FAC) score of 4.16 Prior to the experiment, all subjects had walked on the treadmill at least once during regular physical therapy sessions. Exclusion criteria were vestibular disease, cardiovascular or pulmonary comorbidities contraindicating moderate exercise, visual impairments, cognitive and communicative disorders that could interfere with the protocol, and medication use and/or nonstroke-related sensory/motor impairments that could interfere with balance control. Subjects were classified as either dependent or independent ambulators based on their dependence on walking aids in daily life. Of the 24 subjects, 12 subjects used a walking aid in daily life (dependent ambulators), whereas the other 12 did not (independent ambulators). Based on previous studies on the energy cost for balance control, we estimated the effect size in this group to be .40.12,17 With this effect size, an alpha of .05, and a beta power of .80, 12 subjects per group were deemed appropriate to detect statistically significant differences in energy cost between groups and conditions. Balance and walking ability were characterized by the FAC score, score on the Berg Balance Scale (BBS),18 and score on a timed ten-meter walk test (TMWT).19 The scores were obtained from a standard clinical assessment by a physical therapist no longer than 1 month prior to participation. Descriptive characteristics did not differ significantly between the groups except for walking speed on the TMWT and BBS, with lower scores for dependent ambulators (table 1). Prior to giving their written informed consent, all participants were fully informed about the content and purpose of the study. This study was approved by the local medical ethics committee.
Experimental protocol The experiment consisted of 2 support conditions (with and without support), which were performed overground and on a treadmill (task conditions), resulting in 4 walking trials, which all lasted 5 minutes. Trials were separated by a resting period of approximately 5 minutes. Subjects were randomly assigned to start on the treadmill or overground, and supported and unsupported trials were then performed in random sequence. Prior to the walking trials, resting metabolism was measured in a seated position for 5 minutes, after a 10-minute resting period. Support was provided by a standard height adjustable cane on the nonparetic side during overground walking, and by handrail support on the nonparetic side during treadmill walking. For some patients, treadmill walking without support for 5 minutes can be very challenging; therefore, balance support was allowed if necessary during the first 2 minutes of the unsupported treadmill trial. The last 3 minutes were performed without support to ensure steady-state oxygen consumption (VO2) during the final 2 minutes of the trial from which energy cost was calculated. To avoid excessive weight bearing on the handrail and cane, subjects were given the instruction to use the support to maintain balance without leaning on it. Post hoc analysis on force platform data from the instrumented treadmill revealed that weight bearing on the handrail was on average no more than 5% of the subject’s body weight. www.archives-pmr.org
Balance control and energy cost after stroke Table 1
2257
Descriptive characteristics of study population
Characteristics
All
Independent
Dependent
Age (y) Sex (male/female) Time since stroke onset (d) Body mass index FAC (score of 4/ score of 5) BBS TMWT (km/h)
51.815.9 15/9 79.050.4
46.917.3 8/4 64.848.4
57.113.0 7/5 94.548.4
25.53.0 6/17
26.82.6 2/10
24.12.9 4/7
52.54.5 3.51.3
54.91.3 4.31.1
49.95.5* 2.71.0*
NOTE. Values are mean SD or as otherwise indicated. * Significant difference from the independent group (P<.05).
All walking trials were performed at the preferred speed. Prior to the first treadmill trial, preferred walking speed with and without support was determined by having subjects walk on the treadmil at a low beltspeed (0.4km$h1), and increasing beltspeed gradually with 0.2km$h1 every 10 seconds until the subject subjectively identified their preferred walking speed. This speed was maintained for approximately 30 seconds after which the subjects were asked to reevaluate, if necessary, the beltspeed was adjusted. This was repeated until the subjects preferred walking speed was established. This process concurrently served as a habituation period for treadmill walking. If necessary, this habituation period was extended until the subject reported feeling comfortable walking on the treadmill. During overground trials, subjects walked at a comfortable speed back and forth on a 2-m indoor path set between 2 cones. During each lap, the duration to walk the middle 10m of the path was timed with a stopwatch. During turns, subjects were supported at the waist by one of the investigators to minimize the possible additional metabolic cost of turning.
Instrumentation Breath by breath VO2 was obtained from a pulmonary gas exchange system.a Oxygen consumpation data from the final 2 minutes of each trial were analyzed to ensure steady-state energy expenditure. During all walking trials, trunk accelerations were measured using a triaxial ambulant accelerometerb fixed with an elastic belt near the level of the third lumbar spine segment (sampling frequency 100Hz). Treadmill trials were executed on an instrumented treadmillc equipped with a forceplate (1.51m, sampling frequency 300Hz), from which the vertical ground reaction force was derived to calculate weight bearing on the handrail.
Data analysis 1
1
Steady-state energy expenditure (J$kg $min ) was calculated from VO2 (mL$kg1$min1) and the respiratory exchange ratio (RER) obtained during the final 2 minutes of each trial. Gross metabolic energy expenditure (Egross) was calculated using the following equation20: Egross J$kg1 $min1 Zð4:960 RER þ 16:040Þ Vo2 Net energy expenditure was calculated by subtracting resting metabolism from Egross. Net metabolic cost (J$kg1$m1) was obtained by normalizing for body weight and walking speed. www.archives-pmr.org
Oxygen data for 1 subject (dependent ambulator) did not show prolonged periods of steady-state VO2; therefore, this subject was removed from all analyses. Walking speed was determined from the beltspeed during treadmill walking, and for overground walking from the average time to walk the middle 10m of the path of all laps walked during the final 2 minutes of a trial. The accelerometer data were used to calculate the following gait parameters: mean and variability of stride time and symmetry index (SI). The accelerometer data were first low-pass filtered with a third-order zero-lag Butterworth filter with a cutoff frequency of 20Hz. Subsequently, moments of foot contact were determined from anterior-posterior trunk accelerations based on the method by Zijlstra et al.21 These moments of foot contact were used to calculate step time. Outliers in the step time data during overground trials caused by turning were removed from the data using a median filter, which removed data points more than 3 SDs away from the median. Step time was quantified as the time interval (s) between foot contact of the contralateral limb and the next foot contact of the ipsilateral limb. Stride time was calculated by adding left and right consecutive step times. Variability of stride times was quantified by the coefficient of variation (CV) of the stride times of individual participants within a trial. The SI was calculated as the ratio between the shorter and longer step time, with perfect symmetry resulting in an SI of 1 and smaller values corresponding to larger asymmetry.
Statistical analysis The level of significance for all statistical analyses was set at P<.05. Data were tested for normality using the KolmogorovSmirnov and Shapiro-Wilk tests, which revealed that most of the outcome measures were not normally distributed. Therefore, nonparametric tests were used. Main effects of the support condition (unsupported vs supported) and task condition (overground vs treadmill) were tested using the Wilcoxon signed-rank test. Main group effects (independent vs dependent) were investigated using the Mann-Whitney U test. Group support condition interaction was tested using the Mann-Whitney test on difference scores (unsupported vs supported) between groups. Task condition support condition interaction was investigated using a Wilcoxon signed-rank test on difference scores (unsupported vs supported) for overground and treadmill walking.
Results Effect of support on the treadmill A significant main effect of support on the energy cost was found for treadmill walking. Supported walking resulted in a significant decrease in energy cost of 16.2% (0.99J$kg1$m1) on average (P<.001). Although the reduction in the energy cost with support was on average larger for dependent (19%) than independent ambulators (14%), no significant group support condition interaction effect was found (PZ.107). A significant main effect of group showed that independent of the support condition, dependent ambulators had a significantly higher energy cost of walking than independent ambulators (P<.01) (fig 1, table 2). The effect of support on gait parameters for treadmill walking are shown in figure 2A through 2D and table 2. No significant
2258
T. IJmker et al
Overground Energy cost (J ·kg-1 · m-1)
Energy cost (J ·kg-1 · m-1)
Treadmill 10 8 6 4 2 0
Independent
10
Unsupported
8
Supported
6 4 2 0
Independent
Dependent
Dependent
Fig 1 Effect of support on the energy cost for the independent and dependent ambulators seperately for treadmill (A) and overground walking (B). Error bars represent SDs.
for walking speed, but a significant group support condition interaction for speed (P<.05) revealed that independent ambulators decreased their walking speed significantly with support (P<.05), whereas no change in speed was found for dependent ambulators. Stride time increased significantly during supported walking (P<.01). No significant main effect of support on the CV of stride time was found, but a significant group support condition interaction effect on the CV of stride time showed that supported walking resulted in a significantly reduced CV of stride time for dependent ambulators only (P<.01). No significant effect on step time symmetry was found.
main effect of support was found for walking speed, but results did show a significant group support condition interaction (P<.05). Further analysis showed that for independent ambulators walking speed did not change because of support, whereas dependent ambulators showed a trend toward an increased speed with support (PZ.058). A significant main effect of support for mean stride time showed that during walking with support, stride time increased significantly (P<.01). No significant effect of support on the CV of stride time was found. A significant group support condition interaction on step time symmetry was found (P<.05); further analysis showed that step time symmetry improved (became more symmetrical) for dependent ambulators only (P<.05).
Treadmill versus overground walking Effect of support overground
Treadmill walking resulted in a significantly higher energy cost than overground walking both with (15%) and without (38%) support (P<.01). Furthermore, the effect of support on the energy cost was significantly larger on the treadmill than during overground walking (P<.05). Regarding gait parameters, walking speed was significantly lower on the treadmill compared with overground (P<.01). No significant support condition task condition interaction was found for any of the gait parameters.
No significant main effect of the support condition was found for overground walking. A main effect of the group revealed that independent of the support condition, dependent ambulators had a significantly higher energy cost than independent ambulators (P<.01) (see fig 1, table 2). Furthermore, a significant group support condition interaction was found (P<.05); dependent ambulators showed a significant decrease in the energy cost due to support (8.4%, P<.05), whereas independent ambulators showed a significant increase in the energy cost due to support (6.1%, P<.05). The effect of support on gait parameters for overground walking is shown in figure 3A through 3D and table 2. Similar to treadmill walking, no significant main effect of support was found Table 2
Discussion In this study we hypothesized that walking with support would decrease the metabolic effort to maintain balance and would
Energy cost and gait parameters during unsupported and supported walking for independent and dependent ambulators Treadmill Independent
Overground Dependent
Independent
Dependent
Outcome parameters
US
S
US
S
US
S
US
S
Energy cost (J$kg1$m1) Walking speed (m/s) Stride time (s) CV of stride times (%) SI
3.191.12 0.910.30 1.230.13 4.501.90 0.930.06
2.660.69* 0.870.22 1.320.14* 4.102.20 0.890.10y
6.452.91 0.500.30 1.420.35 10.809.40 0.750.13
4.951.70* 0.590.23 1.500.39* 5.603.30 0.810.11*y
2.580.57 1.130.26 1.220.12 4.401.90 0.880.12
2.950.56*y 1.050.22*y 1.290.15* 4.801.90y 0.890.13
4.542.13 0.700.28 1.420.39 9.005.80 0.740.09
4.001.47*y 0.740.27y 1.500.43* 6.402.40*y 0.760.10
NOTE. Values are mean SD. Abbreviations: S, supported; US, unsupported. * Significantly different from unsupported walking. y Significant difference in effect of support between independent and dependent ambulators (group support interaction).
www.archives-pmr.org
Balance control and energy cost after stroke
2259
Stride Time 2
1.5
Stride time (s)
Walking speed (m/s)
Walking speed
1 0.5 0
Supported
1.6 1.4 1.2 1
Independent
Dependent
Independent
Symmetry index
20 15 10 5
Dependent
Symmetry index
CV of stride time CV of stride time (%)
Unsupported
1.8
1.1
0.9 0.7 0.5
0
Independent
Dependent
Independent
Dependent
Fig 2 Effect of support on walking speed (A), stride time (B), stride time variability (C), and symmetry index (D) during treadmill walking for independent and dependent ambulators separately. Error bars represent SDs.
therefore decrease the energy cost of walking. Moreover, we expected that this effect would be stronger in dependent ambulators than in independent ambulators, and higher for treadmill walking compared with overground walking. Support significantly reduced the energy cost of walking on the treadmill, with a reduction in the energy cost of 16% on average, corresponding to a decrease of 0.99J$kg1$m1. For overground walking, the effect of support was smaller and beneficial only for dependent ambulators (8.4%). These results demonstrate that providing support can result in a substantial and clinically relevant reduction of the energy cost of walking in patients with stroke.22 The effect of support appears to be mediated by the individual’s walking ability; dependent ambulators showed a higher energy cost of walking in general and a larger effect of support than independent ambulators. In addition, the effect of support appears to be mediated by the walking task. We expected larger effects of support on the treadmill, because we assumed treadmill walking to be more challenging for balance control. Indeed, the energy cost was higher, and the reduction in cost because of support was considerably larger on the treadmill than overground. During overground walking, independent ambulators even experienced a disadvantage of support. However, this difference in effect of support on the treadmill and overground might also be affected by the type of support on the treadmill (handrail) and overground (cane). While cane-assisted walking can provide a biomechanical and somatosensory advantage for balance control, it also represents a dual task and may lead to an increase in the energy cost because of handling and carrying the cane.23,24 In healthy subjects, an increase in cost of 30% in cane-assisted walking has been reported.23,25 Similarly, these metabolic demands of cane-assisted walking would be encountered by www.archives-pmr.org
patients with stroke, negating positive effects of cane support because of facilitation of balance control. Nonetheless, the extra metabolic cost of cane-assisted walking in independent ambulators (6.1%), who were less accustomed to using a cane and showed a more normative gait pattern, was far less than what has been reported for healthy subjects. Moreover, dependent ambulators even showed a reduction in the energy cost with support, implying that the benefit of facilitating balance control outweighs the disadvantages of cane-assisted walking in this group. Although many factors could influence the energy cost of walking after stroke, the manipulation that was used in this study was aimed to primarily alter balance control. Furthermore, the reduction in the energy cost with support was accompanied by changes in temporal gait parameters, most notably an increase in stride time, a decrease in stride time variability, and an increase in gait symmetry. Previous studies have found that people reduce stride time when gait stability is challenged.26,27 Also, lower stride time variability and improved symmetry is often seen as an indicator of a more stable gait pattern.28-30 Together, these changes can be regarded as evidence that providing support during walking, as imposed in this study, indeed facilitates balance control, thereby reducing the strain on the balance control system and reducing metabolic cost. Note, however, that providing support does not necessarily reduce the energy cost to normative values (especially not in the dependent ambulators). With balance support, the energy cost of walking in stroke patients might still be higher compared with able-bodied subjects, because of other stroke-related problems, such as impairments in leg swing and propulsion. The mechanisms underlying the energetic requirement of balance control are currently not well understood, but might be
2260
T. IJmker et al
Walking speed 1.5
2
1 0.5 0
1.8 1.6 1.4 1.2 1
Independent
Dependent
Independent
CV of stride time
Dependent
Symmetry index 1.1
20
Symmetry index
CV of stride time (%)
Supported
Stride time Stride time (s)
Walking speed (m/s)
Unsupported
15
10 5
0.9 0.7 0.5
0
Independent
Independent
Dependent
Dependent
Fig 3 Effect of support on walking speed (A), stride time (B), stride time variability (C), and symmetry index (D) during overground walking for independent and dependent ambulators separately. Error bars represent SDs.
related to the selection of energetically less optimal gait parameters, such as step frequency or walking speed, or increased muscle cocontraction, to improve stability despite a penalty on energetic cost. In addition, part of the changes in the energy cost when walking with support could be related to changes in walking speed, because energy cost is known to depend on walking speed following a U-shaped function.31,32 For instance, on the treadmill, dependent ambulators showed a trend toward a small increase in speed accompanied by a reduction in the energy cost with support. Possibly, walking with support enables these patients to walk at an energetically more optimal walking speed. In future studies, a more elaborate investigation of gait parameters related to gait stability might shed more light on the underlying changes related to the decrease in the energy cost because of support.
the group of dependent ambulators, because of the inclusion criteria necessary to be able to complete the protocol (FAC>3 and able to walk without support on a treadmill). Therefore, this study may actually underestimate the effect of support in patients with more limited ambulatory capacity. Finally, the group sizes in this study were relatively small (nZ12 for both groups), especially given the large variation in individual impairments in patients with stroke, which may have reduced the statistical power of this study. However, despite the small sample sizes and the heterogenous population, we were able to detect statistically significant results, suggesting that the obtained results reflect more generic principles regarding the metabolic effort for balance control.
Conclusions Study limitations It can be questioned whether treadmill walking, as used in this study to demonstate the effect of support in a challenging situation, is relevant in the context of daily functioning. However, the same could be said for the overground walking task in a perturbation-free environment, which can be seen as a simplification of real-life walking in which perturbations continuously occur. Although both conditions can thus be considered somewhat artificial, they do provide an indication of the magnitude of the effect of support that could occur in daily life. Another limitation of the current study is the relatively young age and good ambulatory ability of the study population, even in
Balance support can result in a significantly reduced energy cost of walking in patients with stroke, providing further evidence for the notion that balance control exacts a metabolic demand during walking, particularly in balance-impaired populations. The effect of balance support on the energy cost of walking is dependent on both ambulatory capacity and the walking task itself, with larger reductions for subjects depending on a walking aid in daily life and during challenging tasks, such as walking on a treadmill. Based on these results, it can be argued that, besides other stroke-related problems, the increased cost of walking in patients with stroke partly originates from an increased metabolic effort for balance control. Improving or assisting balance control during rehabilitation could lead to www.archives-pmr.org
Balance control and energy cost after stroke a decreased metabolic effort for balance control and a reduced cost of walking in patients with stroke.
Suppliers a. Oxycon Delta; CareFusion, De Molen 8/10, 3994 DB, Houten, the Netherlands. b. Dynaport Hybrid; McRoberts, Raamweg 43, 2596 HN, The Hague, the Netherlands. c. C-Mill; ForceLink BV, Industrieweg 6G, 4104 AR, Culemborg, the Netherlands.
Keywords Energy metabolism; Gait; Postural balance; Rehabilitation; Stroke
Corresponding author Trienke IJmker, MSc, MOVE Research Institute Amsterdam, Faculty of Human Movement Sciences, VU University Amsterdam, Van der Boechorststraat 9, 1081 BT, Amsterdam, the Netherlands. E-mail address: [email protected].
References 1. Roger VL, Go AS, Lloyd-Jones DM, et al. Heart disease and stroke statisticse2012 update: a report from the American Heart Association. Circulation 2012;125:e2-220. 2. Colle F, Bonan I, Gellez Leman MC, Bradai N, Yelnik A. Fatigue after stroke. Ann Readapt Med Phys 2006;49:272-6, 361-4. 3. Ingles JL, Eskes GA, Phillips SJ. Fatigue after stroke. Arch Phys Med Rehabil 1999;80:173-8. 4. Mackay-Lyons MJ, Makrides L. Exercise capacity early after stroke. Arch Phys Med Rehabil 2002;83:1697-702. 5. Mackay-Lyons MJ, Makrides L. Longitudinal changes in exercise capacity after stroke. Arch Phys Med Rehabil 2004;85:1608-12. 6. Platts MM, Rafferty D, Paul L. Metabolic cost of over ground gait in younger stroke patients and healthy controls. Med Sci Sports Exerc 2006;38:1041-6. 7. Detrembleur C, Dierick F, Stoquart G, Chantraine F, Lejeune T. Energy cost, mechanical work, and efficiency of hemiparetic walking. Gait Posture 2003;18:47-55. 8. Wagenaar RC, Beek WJ. Hemiplegic gait: a kinematic analysis using walking speed as a basis. J Biomech 1992;25:1007-15. 9. Waters RL, Mulroy S. The energy expenditure of normal and pathologic gait. Gait Posture 1999;9:207-31. 10. Maeda N, Kato J, Azuma Y, et al. Energy expenditure and walking ability in stroke patients: their improvement by ankle-foot orthoses. Isokinet Exerc Sci 2009;17:57-62. 11. Dean JC, Alexander NB, Kuo AD. The effect of lateral stabilization on walking in young and old adults. IEEE Trans Biomed Eng 2007;54: 1919-26. 12. Donelan JM, Shipman DW, Kram R, Kuo AD. Mechanical and metabolic requirements for active lateral stabilization in human walking. J Biomech 2004;37:827-35.
www.archives-pmr.org
2261 13. Ortega JD, Fehlman LA, Farley CT. Effects of aging and arm swing on the metabolic cost of stability in human walking. J Biomech 2008;41: 3303-8. 14. Houdijk H, Ter Hoeve N, Nooijen C, Rijntjes D, Tolsma M, Lamoth C. Energy expenditure of stroke patients during postural control tasks. Gait Posture 2010;32:321-6. 15. Brouwer B, Parvataneni K, Olney SJ. A comparison of gait biomechanics and metabolic requirements of overground and treadmill walking in people with stroke. Clin Biomech (Bristol, Avon) 2009;24: 729-34. 16. Holden MK, Gill KM, Magliozzi MR, Nathan J, Piehl-Baker L. Clinical gait assessment in the neurologically impaired. Reliability and meaningfulness. Phys Ther 1984;64:35-40. 17. Wezenberg D, de Haan A, van Bennekom CA, Houdijk H. Mind your step: metabolic energy cost while walking an enforced gait pattern. Gait Posture 2011;33:544-9. 18. Berg KO, Wood-Dauphinee SL, Williams JI, Maki B. Measuring balance in the elderly: validation of an instrument. Can J Public Health 1992;83(Suppl 2):S7-11. 19. Collen FM, Wade DT, Bradshaw CM. Mobility after stroke: reliability of measures of impairment and disability. Int Disabil Stud 1990;12:6-9. 20. Garby L, Astrup A. The relationship between the respiratory quotient and the energy equivalent of oxygen during simultaneous glucose and lipid oxidation and lipogenesis. Acta Physiol Scand 1987;129:443-4. 21. Zijlstra W, Hof AL. Assessment of spatio-temporal gait parameters from trunk accelerations during human walking. Gait Posture 2003;18: 1-10. 22. Brehm MA, Knol DL, Harlaar J. Methodological considerations for improving the reproducibility of walking efficiency outcomes in clinical gait studies. Gait Posture 2008;27:196-201. 23. Bateni H, Maki BE. Assistive devices for balance and mobility: benefits, demands, and adverse consequences. Arch Phys Med Rehabil 2005;86:134-45. 24. Woollacott M, Shumway-Cook A. Attention and the control of posture and gait: a review of an emerging area of research. Gait Posture 2002; 16:1-14. 25. McBeath AA, Bahrke M, Balke B. Efficiency of assisted ambulation determined by oxygen consumption measurement. J Bone Joint Surg Am 1974;56:994-1000. 26. Hak L, Houdijk H, Steenbrink F, et al. Speeding up or slowing down?: Gait adaptations to preserve gait stability in response to balance perturbations. Gait Posture 2012;36:260-4. 27. Monsch ED, Franz CO, Dean JC. The effects of gait strategy on metabolic rate and indicators of stability during downhill walking. J Biomech 2012;45:1928-33. 28. Hausdorff JM, Rios DA, Edelberg HK. Gait variability and fall risk in community-living older adults: a 1-year prospective study. Arch Phys Med Rehabil 2001;82:1050-6. 29. Hausdorff JM. Gait dynamics, fractals and falls: finding meaning in the stride-to-stride fluctuations of human walking. Hum Mov Sci 2007;26:555-89. 30. Bradley CE, Wutzke CJ, Zinder S, Lewek MD. Spatiotemporal gait asymmetry is related to balance/fall risk in individuals with chronic stroke. Gainesville: American Society of Biomechanics; 2012. 31. Mian OS, Thom JM, Ardigo LP, Narici MV, Minetti AE. Metabolic cost, mechanical work, and efficiency during walking in young and older men. Acta Physiol (Oxf) 2006;186:127-39. 32. Malatesta D, Simar D, Dauvilliers Y, et al. Energy cost of walking and gait instability in healthy 65- and 80-yr-olds. J Appl Physiol 2003;95: 2248-56.
ORIGINAL ARTICLE
Effect of Balance Support on the Energy Cost of Walking After Stroke Trienke IJmker, MSc,a,b Han Houdijk, PhD,a,b Claudine J. Lamoth, PhD,c Ameerani V. Jarbandhan, MSc,a Danie¨lle Rijntjes, PT,b Peter J. Beek, PhD,a Lucas H. van der Woude, PhDc From the aMOVE Research Institute Amsterdam, Faculty of Human Movement Sciences, VU University Amsterdam, Amsterdam, bHeliomare Research and Development, Wijk aan Zee and cCenter for Human Movement Sciences and Center for Rehabilitation, University Medical Center Groningen, University of Groningen, Groningen, The Netherlands.
Abstract Objective: To examine the influence of balance support on the energy cost of treadmill and overground walking in ambulatory patients with stroke. Design: Cross-sectional. Setting: Research laboratory at a rehabilitation center. Participants: Patients with stroke depending on a walking aid in daily life (nZ12; walking aid dependent ambulators) and walking aid independent ambulators (nZ12), all able to walk for at least 5 minutes. Interventions: Not applicable. Main Outcome Measures: Energy cost (J$kg1$m1) and temporal gait parameters (walking speed, mean and coefficient of variation of stride time, and symmetry index) were obtained during 4 walking trials at preferred walking speed: overground with and without a cane and on a treadmill with and without handrail support. Results: On the treadmill, handrail support resulted in a significant decrease in energy cost of 16%, independent of the group. Although walking aid dependent ambulators had on average a larger reduction in energy cost than walking aid independent ambulators (19% vs 14%), this interaction did not reach statistical significance (PZ.11). Interestingly, overground walking with support resulted in an 8% reduction in energy cost for walking aid dependent ambulators, but a 6% increase for walking aid independent ambulators. The reduction in energy cost with support was accompanied by changes in temporal gait parameters, most notably an increase in stride time and symmetry and a decrease in stride time variability. Conclusions: Balance support can result in a significant reduction in the energy cost of walking in stroke patients, the magnitude of which depends on walking ability and the walking task. Impaired balance control should not be overlooked as a contributing factor to the increased energy cost of walking in patients with stroke, and improving or assisting balance control should be considered to reduce the energy cost of hemiplegic gait. Archives of Physical Medicine and Rehabilitation 2013;94:2255-61 ª 2013 by the American Congress of Rehabilitation Medicine
Stroke is one of the leading causes of death and disability in the Western world for people >65 years.1 It can result in various physical and neuropsychologic impairments that limit Presented to the Joint World Congress of International Society for Posture and Gait Research and Gait and Mental Function, June 25, 2012, Trondheim, Norway; and the European Society for Movement Analysis in Adults and Children, September 13e14, 2012, Stockholm, Sweden. No commercial party having a direct financial interest in the results of the research supporting this article has or will confer a benefit on the authors or on any organization with which the authors are associated.
performance of activities of daily living and participation in society. Two major physical problems that often arise in this population are poststroke fatigue and difficulties with maintaining balance.2 Poststroke fatigue is thought to be a multidimensional concept involving physical as well as neuropsychologic factors.2,3 Physiologically, fatigue can be seen as the consequence of an imbalance between physical capacity, or energy resources, and energy demands.2 In patients with stroke, physical capacity is often
0003-9993/13/$36 - see front matter ª 2013 by the American Congress of Rehabilitation Medicine http://dx.doi.org/10.1016/j.apmr.2013.04.022
2256
T. IJmker et al
decreased because of deconditioning,4,5 whereas the energy demand of activities of daily life is frequently increased. For instance, patients with stroke often have an energy cost of walking (expressed in J$kg1$m1) of up to twice the cost of healthy counterparts walking at the same speed.6,7 The nature of this increase in the energy cost of walking is not fully understood. Increased muscle work because of stroke-related impairments, such as increased muscle tone and spasticity, and compensatory strategies, such as increased displacement of the center of mass, hip hiking, and circumduction of the paretic leg, have all been related to this increase.6,8-10 However, in these analyses, the potential metabolic effort for maintaining balance and ensuring a stable gait pattern is often overlooked. Literature has shown that maintaining balance during gait involves a metabolic energy cost even in healthy subjects.11-13 Moreover, previous research suggests that balance problems, which are often seen in stroke survivors, can result in an increased metabolic effort for balance maintenance. It has been shown that balance perturbations during upright standing, for instance standing on foam, lead to an increase in energy expenditure, which is up to twice as high for patients with stroke than for healthy controls.14 If perturbing balance in patients with stroke elicits an extra metabolic demand, supporting balance, by contrast, could lead to a decrease in energy cost. This should especially be true for stroke survivors with more residual deficits affecting balance control, such as those relying on a walking aid, and for more challenging walking conditions. Therefore, the purpose of this study was to investigate the effect of balance support on the energy cost of walking in walking aid independent and walking aid dependent ambulatory patients with stroke, and whether this effect was similar for treadmill and overground walking. We hypothesized that providing support would decrease the energy cost of walking in patients with stroke, and that this decrease would be larger for patients depending on a walking aid in daily life than for independent ambulators. For patients with stroke, the energy cost of walking is generally higher on the treadmill than overground,15 and treadmill walking is claimed to be more challenging to balance control for these patients because of altered visual and proprioceptive input, the moving support surface, and the constant beltspeed. We thus hypothesized that for patients with stroke, the effect of support would be larger on the treadmill than overground.
Methods Participants Subjects were recruited from the stroke unit at the rehabilitation center Heliomare (Wijk aan Zee, the Netherlands). From the total number of stroke patients admitted to the rehabilitation center
List of abbreviations: BBS CV Egross FAC RER SI TMWT VO2
Berg Balance Scale coefficient of variation gross metabolic energy expenditure Functional Ambulatory Category respiratory exchange ratio symmetry index ten-meter walk test oxygen consumption
during the time of the measurements (nZ88 over 2.5mo), 24 patients (mean age SD, 5215.9y; 14 men, 10 women) who fulfilled the inclusion criteria and were available for measurements, agreed to participate. All patients were able to walk for at least 3 minutes without, and 5 minutes with support of an assistive device overground and on a treadmill, and achieved a Functional Ambulatory Category (FAC) score of 4.16 Prior to the experiment, all subjects had walked on the treadmill at least once during regular physical therapy sessions. Exclusion criteria were vestibular disease, cardiovascular or pulmonary comorbidities contraindicating moderate exercise, visual impairments, cognitive and communicative disorders that could interfere with the protocol, and medication use and/or nonstroke-related sensory/motor impairments that could interfere with balance control. Subjects were classified as either dependent or independent ambulators based on their dependence on walking aids in daily life. Of the 24 subjects, 12 subjects used a walking aid in daily life (dependent ambulators), whereas the other 12 did not (independent ambulators). Based on previous studies on the energy cost for balance control, we estimated the effect size in this group to be .40.12,17 With this effect size, an alpha of .05, and a beta power of .80, 12 subjects per group were deemed appropriate to detect statistically significant differences in energy cost between groups and conditions. Balance and walking ability were characterized by the FAC score, score on the Berg Balance Scale (BBS),18 and score on a timed ten-meter walk test (TMWT).19 The scores were obtained from a standard clinical assessment by a physical therapist no longer than 1 month prior to participation. Descriptive characteristics did not differ significantly between the groups except for walking speed on the TMWT and BBS, with lower scores for dependent ambulators (table 1). Prior to giving their written informed consent, all participants were fully informed about the content and purpose of the study. This study was approved by the local medical ethics committee.
Experimental protocol The experiment consisted of 2 support conditions (with and without support), which were performed overground and on a treadmill (task conditions), resulting in 4 walking trials, which all lasted 5 minutes. Trials were separated by a resting period of approximately 5 minutes. Subjects were randomly assigned to start on the treadmill or overground, and supported and unsupported trials were then performed in random sequence. Prior to the walking trials, resting metabolism was measured in a seated position for 5 minutes, after a 10-minute resting period. Support was provided by a standard height adjustable cane on the nonparetic side during overground walking, and by handrail support on the nonparetic side during treadmill walking. For some patients, treadmill walking without support for 5 minutes can be very challenging; therefore, balance support was allowed if necessary during the first 2 minutes of the unsupported treadmill trial. The last 3 minutes were performed without support to ensure steady-state oxygen consumption (VO2) during the final 2 minutes of the trial from which energy cost was calculated. To avoid excessive weight bearing on the handrail and cane, subjects were given the instruction to use the support to maintain balance without leaning on it. Post hoc analysis on force platform data from the instrumented treadmill revealed that weight bearing on the handrail was on average no more than 5% of the subject’s body weight. www.archives-pmr.org
Balance control and energy cost after stroke Table 1
2257
Descriptive characteristics of study population
Characteristics
All
Independent
Dependent
Age (y) Sex (male/female) Time since stroke onset (d) Body mass index FAC (score of 4/ score of 5) BBS TMWT (km/h)
51.815.9 15/9 79.050.4
46.917.3 8/4 64.848.4
57.113.0 7/5 94.548.4
25.53.0 6/17
26.82.6 2/10
24.12.9 4/7
52.54.5 3.51.3
54.91.3 4.31.1
49.95.5* 2.71.0*
NOTE. Values are mean SD or as otherwise indicated. * Significant difference from the independent group (P<.05).
All walking trials were performed at the preferred speed. Prior to the first treadmill trial, preferred walking speed with and without support was determined by having subjects walk on the treadmil at a low beltspeed (0.4km$h1), and increasing beltspeed gradually with 0.2km$h1 every 10 seconds until the subject subjectively identified their preferred walking speed. This speed was maintained for approximately 30 seconds after which the subjects were asked to reevaluate, if necessary, the beltspeed was adjusted. This was repeated until the subjects preferred walking speed was established. This process concurrently served as a habituation period for treadmill walking. If necessary, this habituation period was extended until the subject reported feeling comfortable walking on the treadmill. During overground trials, subjects walked at a comfortable speed back and forth on a 2-m indoor path set between 2 cones. During each lap, the duration to walk the middle 10m of the path was timed with a stopwatch. During turns, subjects were supported at the waist by one of the investigators to minimize the possible additional metabolic cost of turning.
Instrumentation Breath by breath VO2 was obtained from a pulmonary gas exchange system.a Oxygen consumpation data from the final 2 minutes of each trial were analyzed to ensure steady-state energy expenditure. During all walking trials, trunk accelerations were measured using a triaxial ambulant accelerometerb fixed with an elastic belt near the level of the third lumbar spine segment (sampling frequency 100Hz). Treadmill trials were executed on an instrumented treadmillc equipped with a forceplate (1.51m, sampling frequency 300Hz), from which the vertical ground reaction force was derived to calculate weight bearing on the handrail.
Data analysis 1
1
Steady-state energy expenditure (J$kg $min ) was calculated from VO2 (mL$kg1$min1) and the respiratory exchange ratio (RER) obtained during the final 2 minutes of each trial. Gross metabolic energy expenditure (Egross) was calculated using the following equation20: Egross J$kg1 $min1 Zð4:960 RER þ 16:040Þ Vo2 Net energy expenditure was calculated by subtracting resting metabolism from Egross. Net metabolic cost (J$kg1$m1) was obtained by normalizing for body weight and walking speed. www.archives-pmr.org
Oxygen data for 1 subject (dependent ambulator) did not show prolonged periods of steady-state VO2; therefore, this subject was removed from all analyses. Walking speed was determined from the beltspeed during treadmill walking, and for overground walking from the average time to walk the middle 10m of the path of all laps walked during the final 2 minutes of a trial. The accelerometer data were used to calculate the following gait parameters: mean and variability of stride time and symmetry index (SI). The accelerometer data were first low-pass filtered with a third-order zero-lag Butterworth filter with a cutoff frequency of 20Hz. Subsequently, moments of foot contact were determined from anterior-posterior trunk accelerations based on the method by Zijlstra et al.21 These moments of foot contact were used to calculate step time. Outliers in the step time data during overground trials caused by turning were removed from the data using a median filter, which removed data points more than 3 SDs away from the median. Step time was quantified as the time interval (s) between foot contact of the contralateral limb and the next foot contact of the ipsilateral limb. Stride time was calculated by adding left and right consecutive step times. Variability of stride times was quantified by the coefficient of variation (CV) of the stride times of individual participants within a trial. The SI was calculated as the ratio between the shorter and longer step time, with perfect symmetry resulting in an SI of 1 and smaller values corresponding to larger asymmetry.
Statistical analysis The level of significance for all statistical analyses was set at P<.05. Data were tested for normality using the KolmogorovSmirnov and Shapiro-Wilk tests, which revealed that most of the outcome measures were not normally distributed. Therefore, nonparametric tests were used. Main effects of the support condition (unsupported vs supported) and task condition (overground vs treadmill) were tested using the Wilcoxon signed-rank test. Main group effects (independent vs dependent) were investigated using the Mann-Whitney U test. Group support condition interaction was tested using the Mann-Whitney test on difference scores (unsupported vs supported) between groups. Task condition support condition interaction was investigated using a Wilcoxon signed-rank test on difference scores (unsupported vs supported) for overground and treadmill walking.
Results Effect of support on the treadmill A significant main effect of support on the energy cost was found for treadmill walking. Supported walking resulted in a significant decrease in energy cost of 16.2% (0.99J$kg1$m1) on average (P<.001). Although the reduction in the energy cost with support was on average larger for dependent (19%) than independent ambulators (14%), no significant group support condition interaction effect was found (PZ.107). A significant main effect of group showed that independent of the support condition, dependent ambulators had a significantly higher energy cost of walking than independent ambulators (P<.01) (fig 1, table 2). The effect of support on gait parameters for treadmill walking are shown in figure 2A through 2D and table 2. No significant
2258
T. IJmker et al
Overground Energy cost (J ·kg-1 · m-1)
Energy cost (J ·kg-1 · m-1)
Treadmill 10 8 6 4 2 0
Independent
10
Unsupported
8
Supported
6 4 2 0
Independent
Dependent
Dependent
Fig 1 Effect of support on the energy cost for the independent and dependent ambulators seperately for treadmill (A) and overground walking (B). Error bars represent SDs.
for walking speed, but a significant group support condition interaction for speed (P<.05) revealed that independent ambulators decreased their walking speed significantly with support (P<.05), whereas no change in speed was found for dependent ambulators. Stride time increased significantly during supported walking (P<.01). No significant main effect of support on the CV of stride time was found, but a significant group support condition interaction effect on the CV of stride time showed that supported walking resulted in a significantly reduced CV of stride time for dependent ambulators only (P<.01). No significant effect on step time symmetry was found.
main effect of support was found for walking speed, but results did show a significant group support condition interaction (P<.05). Further analysis showed that for independent ambulators walking speed did not change because of support, whereas dependent ambulators showed a trend toward an increased speed with support (PZ.058). A significant main effect of support for mean stride time showed that during walking with support, stride time increased significantly (P<.01). No significant effect of support on the CV of stride time was found. A significant group support condition interaction on step time symmetry was found (P<.05); further analysis showed that step time symmetry improved (became more symmetrical) for dependent ambulators only (P<.05).
Treadmill versus overground walking Effect of support overground
Treadmill walking resulted in a significantly higher energy cost than overground walking both with (15%) and without (38%) support (P<.01). Furthermore, the effect of support on the energy cost was significantly larger on the treadmill than during overground walking (P<.05). Regarding gait parameters, walking speed was significantly lower on the treadmill compared with overground (P<.01). No significant support condition task condition interaction was found for any of the gait parameters.
No significant main effect of the support condition was found for overground walking. A main effect of the group revealed that independent of the support condition, dependent ambulators had a significantly higher energy cost than independent ambulators (P<.01) (see fig 1, table 2). Furthermore, a significant group support condition interaction was found (P<.05); dependent ambulators showed a significant decrease in the energy cost due to support (8.4%, P<.05), whereas independent ambulators showed a significant increase in the energy cost due to support (6.1%, P<.05). The effect of support on gait parameters for overground walking is shown in figure 3A through 3D and table 2. Similar to treadmill walking, no significant main effect of support was found Table 2
Discussion In this study we hypothesized that walking with support would decrease the metabolic effort to maintain balance and would
Energy cost and gait parameters during unsupported and supported walking for independent and dependent ambulators Treadmill Independent
Overground Dependent
Independent
Dependent
Outcome parameters
US
S
US
S
US
S
US
S
Energy cost (J$kg1$m1) Walking speed (m/s) Stride time (s) CV of stride times (%) SI
3.191.12 0.910.30 1.230.13 4.501.90 0.930.06
2.660.69* 0.870.22 1.320.14* 4.102.20 0.890.10y
6.452.91 0.500.30 1.420.35 10.809.40 0.750.13
4.951.70* 0.590.23 1.500.39* 5.603.30 0.810.11*y
2.580.57 1.130.26 1.220.12 4.401.90 0.880.12
2.950.56*y 1.050.22*y 1.290.15* 4.801.90y 0.890.13
4.542.13 0.700.28 1.420.39 9.005.80 0.740.09
4.001.47*y 0.740.27y 1.500.43* 6.402.40*y 0.760.10
NOTE. Values are mean SD. Abbreviations: S, supported; US, unsupported. * Significantly different from unsupported walking. y Significant difference in effect of support between independent and dependent ambulators (group support interaction).
www.archives-pmr.org
Balance control and energy cost after stroke
2259
Stride Time 2
1.5
Stride time (s)
Walking speed (m/s)
Walking speed
1 0.5 0
Supported
1.6 1.4 1.2 1
Independent
Dependent
Independent
Symmetry index
20 15 10 5
Dependent
Symmetry index
CV of stride time CV of stride time (%)
Unsupported
1.8
1.1
0.9 0.7 0.5
0
Independent
Dependent
Independent
Dependent
Fig 2 Effect of support on walking speed (A), stride time (B), stride time variability (C), and symmetry index (D) during treadmill walking for independent and dependent ambulators separately. Error bars represent SDs.
therefore decrease the energy cost of walking. Moreover, we expected that this effect would be stronger in dependent ambulators than in independent ambulators, and higher for treadmill walking compared with overground walking. Support significantly reduced the energy cost of walking on the treadmill, with a reduction in the energy cost of 16% on average, corresponding to a decrease of 0.99J$kg1$m1. For overground walking, the effect of support was smaller and beneficial only for dependent ambulators (8.4%). These results demonstrate that providing support can result in a substantial and clinically relevant reduction of the energy cost of walking in patients with stroke.22 The effect of support appears to be mediated by the individual’s walking ability; dependent ambulators showed a higher energy cost of walking in general and a larger effect of support than independent ambulators. In addition, the effect of support appears to be mediated by the walking task. We expected larger effects of support on the treadmill, because we assumed treadmill walking to be more challenging for balance control. Indeed, the energy cost was higher, and the reduction in cost because of support was considerably larger on the treadmill than overground. During overground walking, independent ambulators even experienced a disadvantage of support. However, this difference in effect of support on the treadmill and overground might also be affected by the type of support on the treadmill (handrail) and overground (cane). While cane-assisted walking can provide a biomechanical and somatosensory advantage for balance control, it also represents a dual task and may lead to an increase in the energy cost because of handling and carrying the cane.23,24 In healthy subjects, an increase in cost of 30% in cane-assisted walking has been reported.23,25 Similarly, these metabolic demands of cane-assisted walking would be encountered by www.archives-pmr.org
patients with stroke, negating positive effects of cane support because of facilitation of balance control. Nonetheless, the extra metabolic cost of cane-assisted walking in independent ambulators (6.1%), who were less accustomed to using a cane and showed a more normative gait pattern, was far less than what has been reported for healthy subjects. Moreover, dependent ambulators even showed a reduction in the energy cost with support, implying that the benefit of facilitating balance control outweighs the disadvantages of cane-assisted walking in this group. Although many factors could influence the energy cost of walking after stroke, the manipulation that was used in this study was aimed to primarily alter balance control. Furthermore, the reduction in the energy cost with support was accompanied by changes in temporal gait parameters, most notably an increase in stride time, a decrease in stride time variability, and an increase in gait symmetry. Previous studies have found that people reduce stride time when gait stability is challenged.26,27 Also, lower stride time variability and improved symmetry is often seen as an indicator of a more stable gait pattern.28-30 Together, these changes can be regarded as evidence that providing support during walking, as imposed in this study, indeed facilitates balance control, thereby reducing the strain on the balance control system and reducing metabolic cost. Note, however, that providing support does not necessarily reduce the energy cost to normative values (especially not in the dependent ambulators). With balance support, the energy cost of walking in stroke patients might still be higher compared with able-bodied subjects, because of other stroke-related problems, such as impairments in leg swing and propulsion. The mechanisms underlying the energetic requirement of balance control are currently not well understood, but might be
2260
T. IJmker et al
Walking speed 1.5
2
1 0.5 0
1.8 1.6 1.4 1.2 1
Independent
Dependent
Independent
CV of stride time
Dependent
Symmetry index 1.1
20
Symmetry index
CV of stride time (%)
Supported
Stride time Stride time (s)
Walking speed (m/s)
Unsupported
15
10 5
0.9 0.7 0.5
0
Independent
Independent
Dependent
Dependent
Fig 3 Effect of support on walking speed (A), stride time (B), stride time variability (C), and symmetry index (D) during overground walking for independent and dependent ambulators separately. Error bars represent SDs.
related to the selection of energetically less optimal gait parameters, such as step frequency or walking speed, or increased muscle cocontraction, to improve stability despite a penalty on energetic cost. In addition, part of the changes in the energy cost when walking with support could be related to changes in walking speed, because energy cost is known to depend on walking speed following a U-shaped function.31,32 For instance, on the treadmill, dependent ambulators showed a trend toward a small increase in speed accompanied by a reduction in the energy cost with support. Possibly, walking with support enables these patients to walk at an energetically more optimal walking speed. In future studies, a more elaborate investigation of gait parameters related to gait stability might shed more light on the underlying changes related to the decrease in the energy cost because of support.
the group of dependent ambulators, because of the inclusion criteria necessary to be able to complete the protocol (FAC>3 and able to walk without support on a treadmill). Therefore, this study may actually underestimate the effect of support in patients with more limited ambulatory capacity. Finally, the group sizes in this study were relatively small (nZ12 for both groups), especially given the large variation in individual impairments in patients with stroke, which may have reduced the statistical power of this study. However, despite the small sample sizes and the heterogenous population, we were able to detect statistically significant results, suggesting that the obtained results reflect more generic principles regarding the metabolic effort for balance control.
Conclusions Study limitations It can be questioned whether treadmill walking, as used in this study to demonstate the effect of support in a challenging situation, is relevant in the context of daily functioning. However, the same could be said for the overground walking task in a perturbation-free environment, which can be seen as a simplification of real-life walking in which perturbations continuously occur. Although both conditions can thus be considered somewhat artificial, they do provide an indication of the magnitude of the effect of support that could occur in daily life. Another limitation of the current study is the relatively young age and good ambulatory ability of the study population, even in
Balance support can result in a significantly reduced energy cost of walking in patients with stroke, providing further evidence for the notion that balance control exacts a metabolic demand during walking, particularly in balance-impaired populations. The effect of balance support on the energy cost of walking is dependent on both ambulatory capacity and the walking task itself, with larger reductions for subjects depending on a walking aid in daily life and during challenging tasks, such as walking on a treadmill. Based on these results, it can be argued that, besides other stroke-related problems, the increased cost of walking in patients with stroke partly originates from an increased metabolic effort for balance control. Improving or assisting balance control during rehabilitation could lead to www.archives-pmr.org
Balance control and energy cost after stroke a decreased metabolic effort for balance control and a reduced cost of walking in patients with stroke.
Suppliers a. Oxycon Delta; CareFusion, De Molen 8/10, 3994 DB, Houten, the Netherlands. b. Dynaport Hybrid; McRoberts, Raamweg 43, 2596 HN, The Hague, the Netherlands. c. C-Mill; ForceLink BV, Industrieweg 6G, 4104 AR, Culemborg, the Netherlands.
Keywords Energy metabolism; Gait; Postural balance; Rehabilitation; Stroke
Corresponding author Trienke IJmker, MSc, MOVE Research Institute Amsterdam, Faculty of Human Movement Sciences, VU University Amsterdam, Van der Boechorststraat 9, 1081 BT, Amsterdam, the Netherlands. E-mail address: [email protected].
References 1. Roger VL, Go AS, Lloyd-Jones DM, et al. Heart disease and stroke statisticse2012 update: a report from the American Heart Association. Circulation 2012;125:e2-220. 2. Colle F, Bonan I, Gellez Leman MC, Bradai N, Yelnik A. Fatigue after stroke. Ann Readapt Med Phys 2006;49:272-6, 361-4. 3. Ingles JL, Eskes GA, Phillips SJ. Fatigue after stroke. Arch Phys Med Rehabil 1999;80:173-8. 4. Mackay-Lyons MJ, Makrides L. Exercise capacity early after stroke. Arch Phys Med Rehabil 2002;83:1697-702. 5. Mackay-Lyons MJ, Makrides L. Longitudinal changes in exercise capacity after stroke. Arch Phys Med Rehabil 2004;85:1608-12. 6. Platts MM, Rafferty D, Paul L. Metabolic cost of over ground gait in younger stroke patients and healthy controls. Med Sci Sports Exerc 2006;38:1041-6. 7. Detrembleur C, Dierick F, Stoquart G, Chantraine F, Lejeune T. Energy cost, mechanical work, and efficiency of hemiparetic walking. Gait Posture 2003;18:47-55. 8. Wagenaar RC, Beek WJ. Hemiplegic gait: a kinematic analysis using walking speed as a basis. J Biomech 1992;25:1007-15. 9. Waters RL, Mulroy S. The energy expenditure of normal and pathologic gait. Gait Posture 1999;9:207-31. 10. Maeda N, Kato J, Azuma Y, et al. Energy expenditure and walking ability in stroke patients: their improvement by ankle-foot orthoses. Isokinet Exerc Sci 2009;17:57-62. 11. Dean JC, Alexander NB, Kuo AD. The effect of lateral stabilization on walking in young and old adults. IEEE Trans Biomed Eng 2007;54: 1919-26. 12. Donelan JM, Shipman DW, Kram R, Kuo AD. Mechanical and metabolic requirements for active lateral stabilization in human walking. J Biomech 2004;37:827-35.
www.archives-pmr.org
2261 13. Ortega JD, Fehlman LA, Farley CT. Effects of aging and arm swing on the metabolic cost of stability in human walking. J Biomech 2008;41: 3303-8. 14. Houdijk H, Ter Hoeve N, Nooijen C, Rijntjes D, Tolsma M, Lamoth C. Energy expenditure of stroke patients during postural control tasks. Gait Posture 2010;32:321-6. 15. Brouwer B, Parvataneni K, Olney SJ. A comparison of gait biomechanics and metabolic requirements of overground and treadmill walking in people with stroke. Clin Biomech (Bristol, Avon) 2009;24: 729-34. 16. Holden MK, Gill KM, Magliozzi MR, Nathan J, Piehl-Baker L. Clinical gait assessment in the neurologically impaired. Reliability and meaningfulness. Phys Ther 1984;64:35-40. 17. Wezenberg D, de Haan A, van Bennekom CA, Houdijk H. Mind your step: metabolic energy cost while walking an enforced gait pattern. Gait Posture 2011;33:544-9. 18. Berg KO, Wood-Dauphinee SL, Williams JI, Maki B. Measuring balance in the elderly: validation of an instrument. Can J Public Health 1992;83(Suppl 2):S7-11. 19. Collen FM, Wade DT, Bradshaw CM. Mobility after stroke: reliability of measures of impairment and disability. Int Disabil Stud 1990;12:6-9. 20. Garby L, Astrup A. The relationship between the respiratory quotient and the energy equivalent of oxygen during simultaneous glucose and lipid oxidation and lipogenesis. Acta Physiol Scand 1987;129:443-4. 21. Zijlstra W, Hof AL. Assessment of spatio-temporal gait parameters from trunk accelerations during human walking. Gait Posture 2003;18: 1-10. 22. Brehm MA, Knol DL, Harlaar J. Methodological considerations for improving the reproducibility of walking efficiency outcomes in clinical gait studies. Gait Posture 2008;27:196-201. 23. Bateni H, Maki BE. Assistive devices for balance and mobility: benefits, demands, and adverse consequences. Arch Phys Med Rehabil 2005;86:134-45. 24. Woollacott M, Shumway-Cook A. Attention and the control of posture and gait: a review of an emerging area of research. Gait Posture 2002; 16:1-14. 25. McBeath AA, Bahrke M, Balke B. Efficiency of assisted ambulation determined by oxygen consumption measurement. J Bone Joint Surg Am 1974;56:994-1000. 26. Hak L, Houdijk H, Steenbrink F, et al. Speeding up or slowing down?: Gait adaptations to preserve gait stability in response to balance perturbations. Gait Posture 2012;36:260-4. 27. Monsch ED, Franz CO, Dean JC. The effects of gait strategy on metabolic rate and indicators of stability during downhill walking. J Biomech 2012;45:1928-33. 28. Hausdorff JM, Rios DA, Edelberg HK. Gait variability and fall risk in community-living older adults: a 1-year prospective study. Arch Phys Med Rehabil 2001;82:1050-6. 29. Hausdorff JM. Gait dynamics, fractals and falls: finding meaning in the stride-to-stride fluctuations of human walking. Hum Mov Sci 2007;26:555-89. 30. Bradley CE, Wutzke CJ, Zinder S, Lewek MD. Spatiotemporal gait asymmetry is related to balance/fall risk in individuals with chronic stroke. Gainesville: American Society of Biomechanics; 2012. 31. Mian OS, Thom JM, Ardigo LP, Narici MV, Minetti AE. Metabolic cost, mechanical work, and efficiency during walking in young and older men. Acta Physiol (Oxf) 2006;186:127-39. 32. Malatesta D, Simar D, Dauvilliers Y, et al. Energy cost of walking and gait instability in healthy 65- and 80-yr-olds. J Appl Physiol 2003;95: 2248-56.