| | Hemiplegic Gait After Stroke: Is Measurement of Maximum Speed Required?Accepted 1 November 2005. Abstract Kollen B, Kwakkel G, Lindeman E. Hemiplegic gait after stroke: is measurement of maximum speed required? ObjectivesTo study the relation between comfortable and maximum walking speed in stroke rehabilitation and to determine which parameters are predictive in this relation and increase the relations’ precision. DesignOne-year prospective cohort study. Longitudinal information was obtained for 10-m comfortable and maximum walking speeds, hemiplegic limb muscle strength, and balance. In addition, subjects’ ages and the type of rehabilitation they received were registered. SettingStroke service facilities. ParticipantsEighty-one acute stroke patients. InterventionsNot applicable. Main Outcome MeasureTen-meter maximum walking speed. ResultsWe found a progressive improvement in walking speed and a mean systematic difference between comfortable and maximum walking speeds. An overall mean intraclass correlation coefficient for consistency of ρ equal to .96 and a within- and between-subject regression coefficient of 1.32 were demonstrated for the relation between comfortable and maximum walking speeds. None of the covariables included were statistically significant in the final linear regression prediction model. ConclusionsIndependent of time after onset of stroke, maximum walking speed can be predicted by comfortable walking speed with considerable accuracy. The precision of this estimation is not increased by considering patients’ age, hemiplegic muscle strength, balance, or therapeutic intervention.
CLINICALLY, IT IS WIDELY accepted that walking speed is a simple but highly reliable and responsive parameter of gait.1, 2(p78,79) Reliability has been established for test-retest and between-observer measurements.3, 4 Repeated measurements with a simple timed walking test has been qualified as a responsive method to measure change in walking performance over time.5 Several randomized controlled trials (RCTs) have demonstrated favorable effects of gait training, particularly on walking speed.6, 7, 8, 9, 10 Supporting the validity of speed measurement as a useful tool for objectively monitoring progress in hemiplegic gait, statistical evaluations have found significant correlations between time-distance parameters such as cadence, cycle time, stance time, swing time, stride length, and step length and walking speed,11, 12, 13 as well as ambulatory performance.1, 3, 11, 13, 14, 15, 16, 17, 18 In addition, gait speed correlates strongly with other parameters such as balance,19 use of walking aids,3 and number of falls; it also reflects activities of daily living function in geriatric patients.20
Given this body of evidence, it is suggested that improvement in walking speed reflects a genuine improvement in mobility, even if other mostly categoric measures fail to detect it.
Obtaining sufficient information on adequate walking speed requires the recording of both comfortable speed and maximum walking speed, because the latter may be important in community-based activities such as crossing a street. Because most patients with stroke tend to walk slowly, it is necessary to determine to what safe level they are capable of increasing their walking speed.
Bohannon17 studied this relation cross sectionally in 20 subacute stroke patients and found Spearman rank-order correlation coefficients for FIM instrument scores of .826 (P<.001) and .673 (P<.01) for comfortable and fast walking speed measurements, respectively, and a linear regression coefficient of 1.40 for comfortable speed versus maximum walking speed. This latter finding denotes its clinical relevance, in that 1.4 times the comfortable speed corresponds with the maximum speed.
In this study, we elaborate on these findings by investigating the predictive relation between comfortable and maximum walking speed at regular postonset times in a larger group of stroke patients. Therefore, we addressed the following research questions: (1) Can maximum walking speed be accurately predicted by comfortable walking speed in patients with severe stroke? (2) Can the precision of this possible predictive relation be increased by considering a stroke patient’s age, hemiplegic limb muscle strength, balance, and his/her therapeutic intervention? (3) Is the relation between comfortable and maximal walking speed stable over time?
Methods  Design and Procedures This prospective cohort study was part of a 3-group RCT conducted to study the effects of augmented exercise training on stroke outcome.21 Within 14 days of stroke onset, 101 severely disabled patients with a primary middle cerebral artery stroke were randomly assigned to a basic rehabilitation program that was supplemented with additional arm or leg training, or to a control program in which the arm and leg were immobilized with an inflatable pressure splint. Each treatment regimen was applied for 30 minutes, 5 days a week during the first 20 weeks poststroke. Patients were included if they met the following criteria: (1) were between 30 to 80 years old; (2) had suffered an ischemic, first-ever, stroke involving the middle or anterior cerebral artery, as revealed by computed tomography or magnetic resonance imaging scan; (3) were unable to walk at first assessment; (4) had no complicating medical history such as cardiac, pulmonary, or orthopedic disorders; (5) had no severe deficits in communication, (6) had no severe deficits in memory and understanding; and (7) provided written or verbal informed consent and demonstrated sufficient motivation to participate. Details of design and outcome have been published elsewhere.21 Inclusion in this study subset was also contingent on one’s ability to walk 10m without the physical assistance of a therapist. Measurements Functional Ambulation Categories Walking ability was assessed with the Functional Ambulation Categories (FAC).2 The FAC is a reliable and valid assessment tool that includes 6 categories designed to provide information on the level of physical support needed by patients to ambulate safely. The first measurement of gait speed was taken as soon as the patient could walk 10m without physical assistance while under the supervision of a therapist. This criterion corresponds to FAC level 3.3, 4 Walking devices were allowed during the measurements, with the exception of rollators and walkers. Measurements began within 14 days of stroke onset, and were done weekly up to 10 weeks, fortnightly up to 20 weeks, and thereafter only at 26, 38, and 52 weeks after onset. All functional assessments were done by 1 observer (GK) who was blinded to subjects’ treatment assignment.21 Speed We studied gait speed at comfortable and safe maximum walking speeds using a standard approach to assess gait performance.3, 16 To reduce the measurement error of the timed walking test, we calculated the mean of 3 repeated measurements.22 During each session, the patient walked 10m at a comfortable and at a maximum pace. Timing with a digital stopwatch that registered time in 1/100 of a second was manually initiated at the “go” instruction and stopped when the subject crossed the 10-m mark. The patient rested for about 1 minute between each test. Registered speed was subsequently converted into meters per second by dividing the distance walked by the time required. No encouragement to facilitate performance during a walking session was permitted. Motricity Index Recovery of strength in upper extremity and lower extremity was assessed with the Motricity Index.23 This instrument reliably assesses paresis in the upper and lower extremities of stroke patients. It uses a weighted score of a maximum of 100 points for each extremity and is derived from the Medical Research Council grades. It tests 6 limb movements. Timed balance test Balance was measured with the timed balance test (TBT).24 This instrument has 5 components on an ordinal scale and measures timed balance on progressively diminishing support surfaces. In addition to a basic rehabilitation program, patients received 1 of 3 therapeutic interventions that emphasized upper- or lower-extremity training or air splint immobilization of the upper and lower limb.21 Age, hemiplegic limb muscle strength, balance, and therapy are believed to be related to walking ability. We used these parameters as predictive covariables in the linear regression model to increase the precision of the prediction of the relation between comfortable and maximum walking speed.25 Statistical Analysis We applied cross-sectional and longitudinal analysis for all included measurements and we used Cronbach α coefficients to determine internal consistency of repeated measurements. We conducted paired Student t testing to demonstrate systematic differences between both speeds. Two-way intraclass correlation coefficients for consistency (ICCconsistency) were calculated to test the relative agreement between comfortable and maximum walking speeds at each measurement. Research question 1 The predictive relation between comfortable and maximum walking speeds was investigated cross sectionally by applying a linear regression model, using maximum walking speed as the dependent variable and comfortable walking speed as the independent variable. Research question 2 Covariables were added to this model to increase precision of prediction. Entering of covariables into the linear regression model was based on the stepwise backward selection technique at α equal to .05. These variables included age, Motricity Index of arm and leg, TBT, and applied therapeutic intervention. Compliance with the assumptions of the linear regression model was confirmed for all variables. We used a 2-tailed significance level of .05 for all tests. Research question 3 Finally, we applied random coefficient analysis to obtain a single coefficient reflecting within- and between-subject regression for all measurements, again using maximum walking speed as the dependent variable and comfortable walking speed as the independent variable. This method analyzes cross-sectional and longitudinal treatment and time effects simultaneously while correcting for the correlated observations within subjects over time. It allows for regression coefficients to differ between subjects. Because time constitutes an independent covariate in such a model, this statistical method makes possible longitudinal analysis of unequally spaced time points of measurement. Finally, in random coefficient analysis, missing data are presumed to be missing at random.26 Results Patient characteristics are presented in table 1. The mean interval ± standard deviation (SD) between stroke onset and first unassisted walk (FAC ≥3) was 4.8±2.9 weeks. None of the patients could walk unassisted in the first week poststroke. Because not all 101 patients progressed to unassisted walking at some point, comfortable walking speed was measured in 85 patients and maximum walking speed in 81 patients. Therefore, our study results were based on a subset population of a maximum of 81 subjects. The mean comfortable walking speed progressively increased from the 2nd to the 52nd week, from .037 to .635m/s; mean maximum speed increased from .071 to .851m/s (fig 1). The overall mean comfortable speed and maximum speed were .472 and .638m/s, respectively. | | |  | Group | Total | |
|---|
| Number | 81 | |
| Sex (male/female) | 47/34 | |
| Age (y) | 64.3±10.8 | |
| MMSE (range, 0–30) | 26.6±2.3 | |
| Stroke hemisphere (left/right) | 33/48 | |
| Type of stroke (OCSP) | | |
|  TACI (0/1)⁎ | 43 | |
| PACI (0/1)⁎ | 31 | |
| LACI (0/1)⁎ | 7 | |
| OPS (range, 1.6–6.8) | 4.2±0.9 | |
| GCS (range, 0–15) | 14.8±0.9 | |
| Impairments (cognitive) (%) | | |
| Visual inattention (0/1)⁎ | 45.7 | |
| Hemianopia (0/1)⁎ | 25.9 | |
| Visual gaze deficit (0/1)⁎ | 21.0 | |
| No. of days between CVA and first measurement | 8.2±2.7 | |
| Motricity Index lower extremity (range, 0–100) | 33.0±28.4 | |
| TCT (0/200) | 61.84±28.3 | |
| Brunnstrom stage (range, 1–6) | 2.53±1.4 | |
| Comfortable walking speed (m/s) | 0.04±0.1 | |
| Barthel Index (range, 0–20) | 6.90±3.7 | |
| FAC (range, 0–5) | 0.89±1.0 | | | | |
Internal consistency of repeated measurements yielded high coefficients for comfortable walking speed (α=.98), maximum walking speed (α=.99), Motricity Index of the arm (α=.99), Motricity Index of the leg (α=.99), and TBT (α=.97). A paired t test revealed systematic differences between comfortable and maximum gait velocities (table 2). The distribution of these measurement differences within the 95% limits of agreement decreased as more patients entered the study (fig 2). | | | | Week | N | Mean Difference (95% CI) | SD | t | ρ⁎ (95% CI) | |
|---|
| 2 | 74 | −.033(−.063to−.003) | .129 | −2.21 | .86(.78–.91) | |
| 3 | 78 | −.060(−.100to−.020) | .178 | −2.99 | .94(.90–.96) | |
| 4 | 78 | −.083(−.128to−.038) | .199 | −3.69 | .95(.92–.97) | |
| 5 | 79 | −.102(−.148to−.057) | .202 | −4.50 | .96(.94–.97) | |
| 6 | 76 | −.090(−.128to−.052) | .168 | −4.69 | .97(.95–.98) | |
| 7 | 75 | −.110(−.160to−.060) | .218 | −4.37 | .96(.93–.97) | |
| 8 | 78 | −.147(−.204to−.089) | .255 | −5.09 | .94(.91–.96) | |
| 9 | 79 | −.159(−.206to−.112) | .211 | −6.71 | .96(.94–.98) | |
| 10 | 75 | −.150(−.208to−.091) | .253 | −5.12 | .95(.92–.97) | |
| 12 | 72 | −.175(−.231to−.120) | .237 | −6.27 | .96(.93–.97) | |
| 14 | 72 | −.196(−.253to−.139) | .242 | −6.86 | .95(.92–.97) | |
| 16 | 73 | −.212(−.270to−.153) | .253 | −7.16 | .95(.92–.97) | |
| 18 | 71 | −.199(−.251to−.147) | .220 | −7.63 | .96(.93–.97) | |
| 20 | 72 | −.190(−.241to−.139) | .219 | −7.37 | .96(.93–.97) | |
| 26 | 73 | −.187(−.254to−.121) | .285 | −5.63 | .93(.88–.95) | |
| 38 | 71 | −.191(−.234to−.148) | .181 | −8.89 | .97(.95–.98) | |
| 52 | 69 | −.213(−.274to−.152) | .253 | −6.98 | .94(.90–.96) | | | | |
ICCconsistency values were calculated for the relation between comfortable and maximum walking speeds at each measurement and the overall mean (ie, the mean of all comfortable and maximum walking speed measurements) and are presented in a scatterplot (see table 2, fig 3). There was increased measurement error toward the highest speeds, while there was decreased discrimination between measurements at lower speeds. Significance for all ICC measurements was obtained at P equal to .000. Research Question 2 Covariables added to the final regression models were not significant and therefore did not contribute to the precision of the cross-sectional relation between comfortable and maximum walking speed. Research Question 3 Random coefficient analysis produced a significant within- and between-subject regression coefficient of 1.32 (95% CI, 1.29–1.36; P=.000), with a fixed and random intercept and slope indicating that 1.32 times the comfortable walking speed generated the expected maximum walking speed. Discussion Our results show that, in stroke patients with marked hemiplegia, maximum walking speed can be reliably estimated by measuring comfortable walking speed time (in m/s). Cross sectionally and longitudinally applied regression analyses demonstrated that the relation between comfortable and maximum walking speed does not change over time after stroke. We found that maximum speed was 1.32 times that of comfortable speed. Furthermore, this relation remained constant after adding the following covariables to the linear regression model to increase its precision: patients’ age, balance, hemiplegic arm and leg muscle strength, and type of therapeutic intervention. Bohannon’s17 cross-sectional study found a regression coefficient–based relation of 1.40 between comfortable and maximum walking speed in 20 subacute stroke patients. This relation is identical to our uncorrected mean overall linear regression coefficient of 1.43, even though Bohannon measured walking speed over a 7-m distance. However, after correcting for the within-subject dependency of multiple measurements, we found an overall regression coefficient of 1.32. The validity of our findings is strengthened by the observed consistency over time and applied correction for within-subject dependency. Moreover, our findings were not affected by learning effects or measurement errors because the relation was stable over time. This relation may be unique for gait in hemiplegic subjects. Our findings may have several implications. First, in a therapeutic setting, maximum walking speed can be estimated with considerable accuracy by the converted comfortable walking speed. This in turn may provide the therapist with information as to the speed a stroke patient can generate to safely meet functional demands such as crossing streets or keeping pace with others who are walking. Second, this relation remains constant irrespective of poststroke time and does not need to be corrected for age, balance, or paresis in the arm or leg, nor is it affected by the type of therapeutic intervention administered. Third, the implications for gait training are important because this information allows therapists to safely train stroke patients at (1.3) higher walking speeds to better prepare them for meeting the demands imposed by independent community living. It enables therapists to measure progress and set targets for maximizing functional gait speed during the rehabilitation process. In stroke patients, speed intensive gait training induces marked speed-related improvements in body and limb kinematics and muscle activation patterns.27 Future research may address the nature of this apparent fixed relationship between comfortable and maximum gait speeds in age-matched healthy subjects and in subjects with other neurologic disorders. It would be of particular interest to clarify the significance of the constant we observed in our study and the biomechanic and energy transformation-related mechanisms involved. Study Limitations In our study, 3 comfortable walking speed measurements preceded 3 maximal walking speed measurements. Despite frequent rest periods, fatigue-related speed changes may have occurred in the maximum speed measurement group. Because we adhered to our restrictive inclusion criteria, our results are based on a homogenous stroke population with no independent walking ability at onset. As a consequence, our results may not extrapolate to the stroke population at large; further investigation is required. Conclusions Poststroke maximum walking speed can be estimated with considerable accuracy by multiplying the converted comfortable walking time (in m/s) by approximately 1.3. This relation is stable over time; its precision does not increase when patients’ age, hemiplegic arm or leg muscle strength, balance, or therapeutic intervention are considered and it is independent of time from onset. References 1.
1
Collen FM
, Wade DT
, Bradshaw CM
.
Mobility after stroke
(reliability of measures of impairment and disability)
.
Int Disabil Stud
. 1990;12:6–9
.
MEDLINE 2.
2
Wade DT
.
Measurement in neurological rehabilitation
.
Oxford: Oxford Univ Pr; 1992;
.
3.
3
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
.
MEDLINE 4.
4
Wade DT
, Langton Hewer R
.
Functional abilities after stroke
(measurement, natural history and prognosis)
.
J Neurol Neurosurg Psychiatry
. 1987;50:177–182
.
MEDLINE |
CrossRef
5.
5
Friedman PJ
, Richmond DE
, Baskett JJ
.
A prospective trial of serial gait speed as a measure of rehabilitation in the elderly
.
Age Ageing
. 1988;17:227–235
.
MEDLINE 6.
6
Richards CL
, Malouin F
, Wood-Dauphinee S
, Williams JI
, Bouchard JP
, Brunet D
.
Task-specific physical therapy for optimization of gait recovery in acute stroke patients
.
Arch Phys Med Rehabil
. 1993;74:612–620
.
MEDLINE |
CrossRef
7.
7
Salbach NM
, Mayo NE
, Wood-Dauphinee S
, Hanley JA
, Richards CL
, Cote R
.
A task-orientated intervention enhances walking distance and speed in the first year post stroke
(a randomized controlled trial)
.
Clin Rehabil
. 2004;18:509–519
.
MEDLINE |
CrossRef
8.
8
Richards CL
, Malouin F
, Bravo G
, Dumas F
, Wood-Dauphinee S
.
The role of technology in task-oriented training in persons with subacute stroke
(a randomized controlled trial)
.
Neurorehabil Neural Repair
. 2004;18:199–211
.
MEDLINE 9.
9
Duncan P
, Studenski S
, Richards L
, et al.
Randomized clinical trial of therapeutic exercise in subacute stroke
.
Stroke
. 2003;34:2173–2180
.
CrossRef
10.
10
Van Peppen RP
, Kwakkel G
, Wood-Dauphinee S
, Hendriks HJ
, Van der Wees PJ
, Dekker J
.
The impact of physical therapy on functional outcomes after stroke
(what’s the evidence?)
.
Clin Rehabil
. 2004;18:833–862
.
MEDLINE |
CrossRef
11.
11
Mizrahi J
, Susak Z
, Heller L
, Najenson T
.
Objective expression of gait improvement of hemiplegics during rehabilitation by time-distance parameters of the stride
.
Med Biol Eng Comput
. 1982;20:628–634
.
MEDLINE |
CrossRef
12.
12
Brandstater ME
, Bruin de H
, Gowland C
, Clarke BM
.
Hemiplegic gait
(analysis of temporal variables)
.
Arch Phys Med Rehabil
. 1983;64:583–587
.
MEDLINE 13.
13
Wall JC
, Turnbull GI
.
Gait asymmetries in residual hemiplegia
.
Arch Phys Med Rehabil
. 1986;67:550–553
.
MEDLINE 14.
14
Dettmann MA
, Londer MT
, Sepic SB
.
Relationships among walking performance, postural stability, and functional assessments of the hemiplegic patient
.
Am J Phys Med
. 1987;66:77–90
.
MEDLINE 15.
15
Friedman PJ
.
Gait recovery after hemiplegic stroke
.
Int Disabil Stud
. 1990;12:119–122
.
MEDLINE 16.
16
Wade DT
, Wood VA
, Heller A
, Maggs J
, Hewer RL
.
Walking after stroke
.
Scand J Rehabil Med
. 1987;19:25–30
.
MEDLINE 17.
17
Bohannon RW
.
Walking after stroke
(comfortable versus maximum safe speed)
.
Int J Rehabil Res
. 1992;15:246–248
.
MEDLINE 18.
18
Ozgirgin N
, Bolukbasi N
, Beyazova M
, Orkun S
.
Kinematic gait analysis in hemiplegic patients
.
Scand J Rehabil Med
. 1993;25:51–55
.
MEDLINE 19.
19
Bohannon RW
.
Selected determinants of ambulatory capacity in patients with hemiplegia
.
Clin Rehabil
. 1989;3:47–53
.
CrossRef
20.
20
Potter JM
, Evans AL
, Duncan G
.
Gait speed and activities of daily living function in geriatric patients
.
Arch Phys Med Rehabil
. 1995;76:997–999
.
Abstract |
Full-Text PDF (308 KB)
|
CrossRef
21.
21
Kwakkel G
, Wagenaar RC
, Twisk JW
, Lankhorst GJ
, Koetsier JC
.
Intensity of leg and arm training after primary middle-cerebral-artery stroke
(a randomised trial)
.
Lancet
. 1999;354:191–196
.
CrossRef
22.
22
Connelly DM
, Stevenson TJ
, Vandervoort AA
.
Between- and within-rater reliability of walking tests in a frail elderly population
.
Physiother Can
. 1996;48:47–51
.
23.
23
Collin C
, Wade D
.
Assessing motor impairment after stroke
(a pilot reliability study)
.
J Neurol Neurosurg Psychiatry
. 1990;53:576–579
.
MEDLINE |
CrossRef
24.
24
Bohannon RW
.
Correlation of lower limb strengths and other variables with standing performance in stroke patients
.
Physiother Can
. 1989;41:198–202
.
25.
25
Kwakkel G
, Wagenaar RC
.
Effect of duration of upper- and lower-extremity rehabilitation sessions and walking speed on recovery of interlimb coordination in hemiplegic gait
.
Phys Ther
. 2002;82:432–448
.
MEDLINE 26.
26
Twisk JW
.
Applied longitudinal data analysis for epidemiology
.
Cambridge: Cambridge Univ Pr; 2003;
.
27.
27
Lamontagne A
, Fung J
.
Faster is better
(Implications for speed-intensive gait training after stroke)
.
Stroke
. 2004;35:2543–2548
.
CrossRef
a Research Bureau, Isala Klinieken, Zwolle b Rehabilitation Centre De Hoogstraat, Rudolf Magnus Institute of Neuroscience, UMC Utrecht, Utrecht c Department of Rehabilitation VU Medical Center, Amsterdam, The Netherlands  Reprint requests to Boudewijn J. Kollen, Research Bureau, Isala Klinieken, PO Box 10400, 8000 GK Zwolle, the Netherlands
Supported by the Netherlands Heart Foundation (project no. 93.134) and ZONmw (grant no. 14.350004). No commercial party having a direct financial interest in the results of the research supporting this article has or will confer a benefit upon the author(s) or upon any organization with which the author(s) is/are associated. PII: S0003-9993(05)01385-7 doi:10.1016/j.apmr.2005.11.007 © 2006 American Congress of Rehabilitation Medicine and the American Academy of Physical Medicine and Rehabilitation. Published by Elsevier Inc. All rights reserved. | |
|
FULL TEXT01385-7/journalimage?img=PIIS0003999305013857.gr1.lrg.gif&fig=fig1&kwhquery=&issn=0003-9993&ishighres=false&allhighres=false&free=yes','journalimage');)
01385-7/journalimage?img=PIIS0003999305013857.gr2.lrg.gif&fig=fig2&kwhquery=&issn=0003-9993&ishighres=false&allhighres=false&free=yes','journalimage');)
01385-7/journalimage?img=PIIS0003999305013857.gr3.lrg.gif&fig=fig3&kwhquery=&issn=0003-9993&ishighres=false&allhighres=false&free=yes','journalimage');)