Improvements in Speed-Based Gait Classifications Are Meaningful
Background and Purpose— Gait velocity is a powerful indicator of function and prognosis after stroke. Gait velocity can be stratified into clinically meaningful functional ambulation classes, such as household ambulation (<0.4 m/s), limited community ambulation (0.4 to 0.8 m/s), and full community ambulation (>0.8 m/s). The purpose of the current study was to determine whether changes in velocity-based community ambulation classification were related to clinically meaningful changes in stroke-related function and quality of life.
Methods— In subacute stroke survivors with mild to moderate deficits who participated in a randomized clinical trial of stroke rehabilitation and had a baseline gait velocity of 0.8 m/s or less, we assessed the effect of success versus failure to achieve a transition to the next class on function and quality of life according to domains of the Stroke Impact Scale (SIS).
Results— Of 64 eligible participants, 19 were initially household ambulators, and 12 of them (68%) transitioned to limited community ambulation, whereas of 45 initially limited community ambulators, 17 (38%) became full community ambulators. Function and quality-of-life SIS scores after treatment were significantly higher among survivors who achieved a favorable transition compared with those who did not. Among household ambulators, those who transitioned to limited or full community ambulation had significantly better SIS scores in mobility (P=0.0299) and participation (P=0.0277). Among limited community ambulators, those who achieved the transition to full community ambulatory status had significantly better scores in SIS participation (P=0.0085).
Conclusions— A gait velocity gain that results in a transition to a higher class of ambulation results in better function and quality of life, especially for household ambulators. Household ambulators possibly had more severe stroke deficits, reducing the risk of “ceiling” effects in SIS-measured activities of daily living and instrumental activities of daily living. Outcome assessment based on transitions within a mobility classification scheme that is rooted in gait velocity yields potentially meaningful indicators of clinical benefit. Outcomes should be selected that are clinically meaningful for all levels of severity.
Stroke, the primary cause of adult disability, leads to impairments in mobility and walking.1,2 Decreased gait velocity reflects impaired mobility; is a reliable, valid, and sensitive measure of recovery of poststroke mobility3–5 that discriminates the effects of stroke; and is related to the potential for rehabilitation recovery.6 Gait velocity is also a widely used indicator in other health states and among older adults, in whom it predicts future health status and health care utilization.7
The relations between gait velocity and community ambulation have been reported in individuals after stroke. Previous work regarding ambulation classification was completed by Perry et al.5 That group used a questionnaire to assess the current mobility of 147 stroke patients. Additional assessments included muscle strength, proprioception, walking velocity, and functional walking ability at home and in the community. Gait velocity was determined to be the most efficient in predicting ambulation classification. The authors condensed 6 original walking classifications into the 3 used in the current analysis: household, limited community, and full community ambulation. Perry et al5 determined that household ambulation was equal to severe gait impairment and a velocity of <0.4 m/s; limited community ambulation was equivalent to moderate gait impairments and a walking speed of between 0.4 and 0.8 m/s; and full community ambulation indicated mild impairment and a walking speed of >0.8 m/s; furthermore, increases in gait velocity were related to improved home and community ambulation. Similarly, Lord and colleagues8 used 4 categories to define community ambulation. They found that gait velocity differed between the categories and ranged from an average of 0.52 m/s for those unable to leave home to an average of 1.14 m/s for those able to ambulate in a shopping center.
Gait velocity has been described as the “almost perfect” measure.9 However, in a recent review of measurements of community ambulation after stroke, Lord and Rochester10 concluded that there is “no guarantee that increases in gait velocity will denote a meaningful improvement in performance.” It is therefore important to determine changes in poststroke gait velocity that are “clinically meaningful,” in that they are associated with important improvements in function and quality of life. The objective of this study was to determine whether a change in gait velocity based on a classification scheme related to community ambulation was clinically meaningful in terms of changes in stroke-related function and quality of life as measured by the Stroke Impact Scale (SIS). In this classification scheme, gait velocity was stratified into 3 clinically meaningful functional ambulation classes: household ambulation (<0.4 m/s), limited community ambulation (0.4 to 0.8 m/s), and full community ambulation (>0.8 m/s). We wished to determine whether favorable transitions across the functional mobility classes were associated with differences in self-reported SIS scores of activities of daily living (ADLs), instrumental activities of daily living (IADLs), mobility, and social participation after stroke.
Data in this study were derived from a parent study: a prospective, randomized, controlled, single-blind, clinical intervention trial.11 The trial was designed to determine the effects of a therapist-supervised 12- to 14-week home-based, structured, and progressive exercise intervention in participants after stroke. This secondary analysis was approved by the University of Florida Health Science Center institutional review board.
This analysis did not examine differences between intervention and usual-care groups, which have been published elsewhere11; rather, we examined the differences in ADLs, IADLs, mobility, and social participation between those who did and did not successfully advance to the next ambulation classification as defined by Perry et al.5 For example, the participants considered successful in walking recovery transitioned from an initial gait velocity of <0.4 m/s to 0.4 to 0.8 m/s or faster, or they advanced from an initial gait velocity of between 0.4 and 0.8 m/s to >0.8 m/s. We realize that there are other classifications available. However, agreed-on exact cutoffs are unavailable, as in classifications from Perry et al5 and Lord et al.8 However, it is likely that there are transition points when a change in function can be expected. Therefore, we chose to use the classifications described by Perry et al5 to assess whether changes assessed in classifications are reflected in meaningful changes in function.
All participants in this study were enrolled in the Kansas City Stroke Registry. All those registered gave informed consent and permission to be screened for eligibility for future research studies. Those eligible for the registry had a confirmed diagnosis of stroke within 3 to 28 days, were >50 years old, and lived within a 50-mile radius. Registry exclusion criteria included the following: subarachnoid hemorrhage, being lethargic, obtunded, or comatose; having uncontrolled blood pressure, hepatic or renal failure, New York Heart Association class III/IV heart failure, a known limited life expectancy, or prestroke disability in self care; or having previously lived in a nursing home before the stroke.
Those in the registry were eligible to be screened for the randomized clinical trial. Inclusion for the parent clinical trial included the following: (1) stroke within 30 to 150 days; (2) ability to ambulate 25 feet independently; (3) mild to moderate stroke deficits defined by a Fugl-Meyer score of 29 to 90 for upper and lower extremities, an Orpington Prognostic Scale score of 2.0 to 5.2, and palpable wrist extension on the involved side; and (4) a Folstein Mini-Mental Status examination score >16. Exclusion criteria included the following: (1) serious cardiac conditions (hospitalization for heart disease within 3 months, active angina, serious cardiac arrhythmias, hypertrophic cardiomyopathy, severe aortic stenosis, pulmonary embolus, or infarction); (2) dependence on supplemental oxygen; (3) severe weight-bearing pain; (4) other serious organ system disease; and (5) a life expectancy of <1 year as reported in the medical chart and defined by the medical team at the time of stroke onset. All participants signed an informed consent form to participate in the trial, and all had approval from a primary care physician.11 Ninety-two individuals completed the clinical trial. For this secondary analysis, individuals were excluded when they could walk >0.8 m/s at baseline. Sixty-four participants walked <0.8m/s at the baseline assessment. This analysis did not compare the intervention with the usual-care groups but combined all 64 participants walking <0.8 m/s at baseline to include those who changed and those who did not change gait velocity over time.
Gait Velocity and Walking Recovery
Gait velocity was assessed at baseline and was repeated at 3 months. It was measured as the average of 2 trials of the 10-m walk.12 Participants were asked to walk at their normal comfortable pace and used a walking aid when needed. All participants were able to accelerate before the assessment began and started the assessment from a stopped position. Successful walking recovery was defined when participants shifted from <0.4 m/s to 0.4 to 0.8 m/s (household to limited community ambulation) or from 0.4 to 0.8 m/s to >0.8 m/s (limited to full community ambulation).5 A dichotomous variable of “success” or “failure” was used for the analysis.
Function and Quality of Life
Function and quality of life were assessed according to the activity and participation constructs of the SIS, a comprehensive and psychometrically robust stroke-specific outcome measure.13,14 The SIS was developed from the perspective of patients, caregivers, and health professionals with stroke expertise. The SIS is reliable, valid, and sensitive to changes in stroke-related recovery. There are 8 domains and 59 items in version 3.0. For this study, we use the ADL/IADL (all SIS scores were converted to a scoring scale of 0 to 100), mobility, and participation domains. The SIS-16 is a shorter version of the SIS composite physical function domain and is limited to assessment of physical functioning.14–16 We also included the SIS-16 in this analysis.
All analyses were completed with SAS statistical software, version 8.2.17 Simple descriptive statistics were used for demographics and outcome scores. Independent-sample t tests were used to determine significant differences in SIS outcomes between those who succeeded and those who failed to transition to a higher level of walking. χ2 tests were used as appropriate to compare results from categorical variables. Because age and baseline scores for the SIS-16 and the SIS ADL/IADL were at least marginally significantly different between groups, analyses were adjusted for all variables in an ANCOVA, with age and baseline score as the covariates.
Baseline demographics for the 64 participants are presented in Table 1. These participants had an average age of 71 years, were diverse in sex and ethnicity, and had mild to moderate stroke deficits. Baseline characteristics of participants who succeeded or failed in transitioning to a higher ambulatory class are presented in Table 1. In the sample as a whole, those who succeeded were younger than those who failed (P=0.0023), but initial gait velocity did not differ: 0.51 m/s for those who succeeded and 0.49 m/s for those who failed (P=0.689; Table 2). In addition, those who successfully transitioned to a higher ambulation class had a significantly higher SIS-16 score than those who failed to transition (P=0.0323). The difference between SIS ADL/IADL scores verged on significance at the 0.05 level (P=0.0570). Those with an initial gait velocity of 0.4 to 0.8 m/s demonstrated increased baseline scores on all SIS measures when compared with those with an initial gait velocity of <0.4 m/s (Table 3). However, all those successful in reaching the next ambulation class demonstrated increased SIS scores compared with those who failed to transition.
The 3-month gait velocity for the entire group was 0.68 m/s (Table 2). As anticipated, 3-month gait velocity scores were higher among those who succeeded in transitioning to a higher ambulation class (0.82 m/s) than those who failed to do so (0.56 m/s) (P<0.0001; Table 2). In all, 45% of the sample was successful and 55% failed to increase to a higher ambulation class at 3 months. Of 19 initial household ambulators, 12 (63%) succeeded in transitioning to limited community ambulation or higher, whereas of 45 limited community ambulators, 17 (38%) transitioned to full community ambulators (Table 4).
For the group as a whole, success in transition was associated with improved stroke-related quality of life as assessed with the SIS, where such transition led to higher SIS participation scores (P=0.0009), and SIS mobility scores neared a significant difference (P=0.0576; Table 5). Among household ambulators, those who succeeded in transitioning to limited community ambulation or higher had significantly better SIS mobility and social participation scores than those who failed (Table 6). Among initially limited community ambulators, those who succeeded had significantly higher SIS participation scores (Table 6).
Gait velocity is a valid and reliable measure of walking recovery after stroke3 and has been shown to be a valuable indicator of future health and function.7,18 Changes in gait velocity have been reported in observational studies and clinical trials,7,11,19–21 but the amount of change that is considered clinically meaningful and reflective of the level of community ambulation has not been established. Lord and Rochester10 recently reported that “continued reliance on the 10-meter walk as a proxy measure for community ambulation is misplaced.” They suggested that self-report of community ambulation is the most useful outcome measure.
However, the results of the current study indicate that when 10-m gait velocity measures are stratified into clinically meaningful functional ambulation classes such as household ambulation (<.4 m/s), limited community ambulation (0.4 to 0.8 m/s), and full community ambulation (>0.8 m/s), changes in 10-m gait velocity are clinically meaningful. Transitioning to a higher class of ambulation is associated with substantially better function and quality of life, especially with regard to mobility and community participation, in initial household ambulators.
The results of this study also demonstrate that it is necessary to consider the severity of stroke deficits when examining relations between changes in gait velocity and changes in function and quality of life. The household ambulators who successfully transitioned to limited community ambulators demonstrated significant changes in SIS-scored mobility and participation. In contrast, those who transitioned from limited community ambulation to full community ambulation only exhibited significant changes in participation. Those classified as household ambulators at baseline possibly had the greatest deficits, making “ceiling” effects in mobility and participation measures less likely. Additionally, the SIS mobility assessment includes items regarding climbing stairs and walking long distances; if these items are not in the participants’ environments, they do not know whether they are able to perform such activities, and therefore, they receive a lower score, decreasing their chance to demonstrate improvement. Although both groups were likely unable to complete all activities, it is more likely that those walking at an increased gait speed and who were demonstrating increased function and quality of life were more likely able to rely on these more difficult items to demonstrate an increased score. Because these individuals are more likely to leave the home environment and engage in the community, there are more items available to be answered, and thus, they are able to gain a higher score.
A limitation of the study is the small sample size. Future studies with larger samples are necessary to fully understand the effects of stroke severity on the probability of transition to different levels of mobility and the associated effects on quality of life. However, the study results have several important implications for clinical research design of poststroke recovery trials. When gait velocity is used to stratify subjects into clinically meaningful functional ambulation classes such as those used herein, then gait velocity is a clinically meaningful outcome measure. Second, it is important to make sure that chosen outcomes are clinically meaningful for all levels of severity. This can be achieved by not assuming that similar changes in gait velocity are clinically meaningful for all levels of severity. In this study, successful walking recovery was defined by a sliding dichotomy.22 The idea of a sliding dichotomy is based on a definition that is tailored for each participant’s baseline function or prognosis, rather than a single definition of good outcome for all participants.
Using this sliding dichotomy, we observed (as expected) that those who walked the fastest at baseline were already functioning well with respect to ADL/IADL and that changes in gait velocity may not be associated with changes in ADL/IADL but rather that changes in gait velocity are associated with recovery of more complex community participation. In contrast, in those with the slowest gait velocity, we observed changes in mobility and participation.
Duncan et al23 reported wide variation and inconsistent use of outcome measures in stroke clinical trials and little agreement about what constitutes a favorable or unfavorable outcome. Our systematic assessment of gait velocity as an outcome measure needs to be replicated with other measures to ensure that the outcomes we observed are clinically meaningful for all levels of severity and that we appropriately selected measures that did not suffer from floor or ceiling effects.
In summary, outcome assessment based on transitions within a mobility classification scheme that is rooted in gait velocity yields potentially meaningful indicators of clinical benefit. However, it is necessary to ensure that chosen outcomes are clinically meaningful for all levels of stroke severity.
Sources of Funding
This project was funded by a predoctoral dissertation award through the Department of Veteran Affairs, Rehabilitation Outcomes Research Center of Excellence, North Florida/South Georgia Malcolm Randal VA Hospital. The original research study was funded by the National Institute on Aging, Claude D. Pepper Older Americans Center grant 5P60AG14635, and was completed at the University of Kansas Theo and Alfred M. Landon Center on Aging and the University of Kansas Medical Center, Kansas City.
- Received October 18, 2006.
- Revision received December 13, 2006.
- Accepted January 22, 2007.
AHCPR. Clinical Practice Guidelines: Post-Stroke Rehabilitation. Washington, DC: US Department of Health and Human Services; 1995.
O’Sullivan SB. Stroke. In: O’Sullivan SB, Schmitz TJ, eds. Physical Rehabilitation: Assessment and Treatment. Philadelphia: FA Davis Co; 2001: 519–582.
Richards C, Malouin F, Wood-Dauphinee S. Gait velocity as an outcome measure of locomotor recovery after stroke. In: Craik RA, Oatis C, eds. Gait analysis: Theory and Applications. St Louis, Mo: Mosby; 1995: 355–364.
Perry J, Garrett M, Gronley JK, Mulroy SJ. Classification of walking handicap in the stroke population. Stroke. 1995; 26: 982–989.
Wade D. Measurement in Neurological Rehabilitation. Oxford, UK: Oxford University Press; 1992.
Lord SE, Rochester L. Measurement of community ambulation after stroke: current status and future developments. Stroke. 2005; 36: 1457–1461.
Duncan P, Studenski S, Richards L, Gollub S, Lai SM, Reker D, Perera S, Yates J, Koch V, Rigler S, Johnson D. Randomized clinical trial of therapeutic exercise in subacute stroke. Stroke. 2003; 34: 2173–2180.
Eng JJ, Chu KS, Dawson AS, Kim CM, Hepburn KE. Functional walk tests in individuals with stroke: relation to perceived exertion and myocardial exertion. Stroke. 2002; 33: 756–761.
Lai SM, Studenski S, Duncan PW, Perera S. Persisting consequences of stroke measured by the Stroke Impact Scale. Stroke. 2002; 33: 1840–1844.
Duncan PW, Lai SM, Bode RK, Perera S, DeRosa J. Stroke Impact Scale-16: a brief assessment of physical function. Neurology. 2003; 60: 291–296.
Green J, Forster A, Young J. Reliability of gait speed measured by a timed walking test in patients one year after stroke. Clin Rehabil. 2002; 16: 306–314.
Duncan PW, Jorgensen HS, Wade DT. Outcome measures in acute stroke trials: a systematic review and some recommendations to improve practice. Stroke. 2000; 31: 1429–1438.