| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Research Reports |
EW Miller, PT, PhD, is Professor of Physical Therapy and Executive Director, Center for Aging and Community, University of Indianapolis, 901 S Shelby, Indianapolis, IN 46203 (USA)
SA Combs, PT, MS, NCS, is Assistant Professor, Krannert School of Physical Therapy, University of Indianapolis, Indianapolis, Ind
C Fish, PT, DPT, is a staff physical therapist, Rehabilitation Services, Elkhart General Hospital, Elkhart, Ind
B Bense, PT, DPT, is a staff physical therapist, Rehabilitation Services, Columbus Regional Hospital, Columbus, Ind
A Owens, PT, DPT, is a staff physical therapist, Rehabilitation and Sports Medicine, Community Health Network, Indianapolis, Ind
A Burch, PT, MS, is a staff physical therapist, Rehabilitation Services, Chandler Regional Medical Center, Chandler, Ariz
Address all correspondence to Dr Miller at: emiller{at}uindy.edu
Submitted August 2, 2005;
Accepted December 11, 2007
 | Abstract |
|---|
Subject: The subject was a 38-year-old man who had a stroke 2.5 years previously.
Methods: A single-subject design with baseline, intervention, immediate postintervention, and 6-month postintervention phases was conducted. Dependent variables included 25-m sprint time, single-leg balance, running step width, running step length ratio, Stroke Impact Scale, 6-minute walk test, and lower-extremity strength (force-generating capacity).
Results: At the 6-month postintervention phase, sprint speed, left single-leg balance, and step width changed significantly from the baseline phase. Step length ratio trended toward less symmetry but more consistency, and muscle strength improved more than 20% in 6 of 8 muscle groups in the involved lower extremity and 4 of 8 muscle groups in the uninvolved lower extremity.
Discussion and Conclusion: Intensive task-specific training was feasible and effective for retraining running ability in the study subject. He returned to recreational running, which provided him with a greatly improved outlook and a better quality of life.
| Introduction |
|---|
 TopAbstract Introduction MethodResultsDiscussionConclusionReferences |
|---|
Walking and running can be distinguished by a replacement of 2 periods of double-limb support during the stance phase of walking with 2 periods of double-limb float during the swing phase of running.3,4 As speed increases from a walk to a run, swing time increases and stance time decreases.3,4 Because of the reduction in support time, greater balance becomes a critical element throughout the running cycle.5,6 Although walking demands muscular effort among proximal lower-extremity (LE) muscle groups, more muscular power must be generated during running, and structures must be able to absorb more energy to control increased ground reaction forces.3–5
Literature pertaining to the rehabilitation of running ability in people with neurological dysfunction is limited. Williams and Goldie6 investigated motor task predictors of running ability in participants with traumatic brain injury (TBI). Four tasks, including bounding onto a leg, walking on toes, stepping backward up a step, and single-limb balance, were found to be strong predictors of running ability after TBI.6 A single-subject experiment involving a participant with a high level of function and an incomplete C5-C6 spinal cord injury investigated outcomes following a speed-progressive gait intervention in a body-weight–support/treadmill (BWST) environment.7 The participant in that study7 showed significantly increased over-ground walking and running speeds following training. The authors suggested that speed-progressive gait training in a BWST environment may be a useful technique for improving functional outcomes in people with other neurological impairments.7 Additionally, animal models indicated that running after the induction of cortical lesions in rats was associated with a neuroprotective effect and a promotion of functional recovery.8,9
Recently, much has been written about locomotor training to maximize walking outcomes in people with neurological dysfunction.10–13 Appropriate sensory inputs, including adequate loading of the LEs during stance and sufficient hip extension during terminal stance to ensure the facilitation of swing, are critical.10,12 Greater intensity of training through gait speeds closer to a normal walking pace has also been cited as a key factor in improving outcomes.12,14–16 Repetitive, task-specific practice of a meaningful activity and attention to safety during training enhance practice of a complex gait task and reduce fear of falling, respectively.10,13,17
Body-weight–support/treadmill training has been suggested as a means of locomotor training that incorporates these motor learning principles.10,11,13,18 Body-weight–support/treadmill training has been shown to improve over-ground walking speed,14,18–22 endurance,19,22 stride length,20 cadence,20 balance,18,19,22 and stance symmetry18 in people with stroke. Body-weight–support/treadmill training allows prolonged training by decreasing the amount of oxygen consumption necessary during gait.23 The BWST environment also increases safety and an individual's confidence through the support offered by the harness system.10,18,24
Despite these documented benefits, a recent Cochrane Review showed no statistically significant differences in walking outcomes between treadmill training with and treadmill training without body-weight support (BWS) or other physical therapy interventions in people with stroke.25 The Cochrane Review reported, however, that training in a BWST environment resulted in a trend toward improved walking speed among people after stroke who could walk independently prior to the initiation of the intervention.25 No research, however, has investigated modifying these interventions as a means to improve running ability in people after stroke.
The purpose of this study, therefore, was to investigate the feasibility and effectiveness of the use of an intensive task-specific intervention to improve the running ability of a 38-year-old man 2.5 years after stroke. The participant could walk independently, but his running pattern was inefficient, slow, and unsafe. We hypothesized that the participant would improve his running ability, as evidenced by improved running speed and symmetry, endurance, balance, strength (force-generating capacity), and self-reported quality of life, following the completion of an 8-week program and would maintain these improvements for at least 6 months.
| Method |
|---|
|
TopAbstractIntroductionMethodResultsDiscussionConclusionReferences |
|---|
The participant's past medical history was unremarkable. Prior to the motor vehicle crash, he was healthy and active, enjoying recreational running and sports, including running half-marathons, water skiing, and playing soccer and basketball. Since the motor vehicle crash and subsequent stroke, he had achieved independence in community-based walking and was able to return to his work as a marketing manager. His 6-minute walk test and LE strength measures are shown in Tables 1 and 2. He had minimal gait deviations during normal walking but had occasional left foot slap during long-distance walking. He used his upper extremities (UEs) independently for all gross and fine motor activities. He could also drive independently and played an active role in parenting. Despite his effort and desire, however, he reported that he was too slow and too uncoordinated to run. He demonstrated an asymmetric running pattern, reduced left LE extension during stance, and minimal time in double-limb float. He also displayed a wide base of support, excessive left foot slap, and reduced left arm swing during running. Repeated attempts at over-ground running were unsuccessful.
|
|
Design
An A-B single-subject design with retention phases was used. Repeat measurements of dependent variables for single-subject analysis (SS variables) were obtained during all phases. A baseline phase with 10 repeat measurements across 3 weeks was established for performance comparisons throughout the remaining phases of the study. Following the baseline phase, the intervention was initiated. Dependent variables were measured once a week for a total of 8 measures during the intervention phase. Two retention phases with 9 measures across 3 weeks allowed comparisons of the immediate postintervention and 6-month postintervention phases with the baseline phase. There are no formal rules about the number of measures in each phase or the length of each phase.26 In an attempt to provide more confidence in our conclusions, we were very conservative in selecting 8 to 10 measurement sessions per phase. Phase length was determined by the participant's schedule. In addition to the SS variables, 3 dependent variables were measured once each during the baseline and retention phases.
Equipment
For this study, a LiteGait BWS system* and a TrimUp 3350 treadmill
were used. The LiteGait BWS system contains a Bi-Sym scale, which measures the amount of BWS provided to the participant. The TrimUp 3350 treadmill is an electric treadmill with speed measured in miles per hour and increasing in 0.1-mile-per-hour increments.
Measurement Tools
Dependent variables were selected to address the levels of the International Classification of Functioning, Disability and Health27 model, including body functions, activity limitations, and participation restrictions. Four SS variables measured activity limitations: 25-m sprint, single-leg balance, running step length ratio (calculated from right and left step length measures), and running step width. Three other dependent variables were also included to provide insight into intervention effectiveness for body functions and participation restrictions: LE strength and 6-minute walk test (body functions) and the Stroke Impact Scale (SIS) (participation restrictions).
The timed 25-m sprint was performed to assess running speed. The participant was instructed to run as fast as he could and not to slow down until after he had passed the finish line. Timing began when the researcher said "go." Sprint times were recorded in seconds, and scores were converted to meters per second.
Although not validated for people with stroke, single-limb balance was chosen because it has been shown to be a good predictor of running ability in people with brain injury.6 The participant was instructed to stand on one leg with his hands on his hips for as long as he could. Timing began as soon as the nonstance foot was lifted from the ground and ended when the nonstance foot, trunk, or UE made contact with the ground or spotter or when 60 seconds was reached. The average time in 3 trials was calculated and recorded for each side.28
Running step length, for calculating the running step length ratio, and running step width were measured by instructing the participant to run at a comfortable pace across a secure 15-m (50-ft) piece of newsprint paper with inked moleskin attached to his shoes.29 The average right and left step lengths, step widths, and stride of the middle 3 strides were measured and recorded in centimeters. The step length ratio was calculated by dividing the right by the left average step length.
A hand-held dynamometer (Microfet 2 HHD)
was used by an experienced clinician to assess LE strength through a make test.30,31 Hand-held dynamometry has been found to have high intrarater reliability (intraclass correlation coefficient [ICC]=.96–.98) and concurrent validity (ICC=.94–.97) in people with stroke.30–32 The average strength in 3 trials was measured and recorded in newtons.
Cardiovascular endurance was assessed through the use of the 6-minute walk test in accordance with guidelines from the American Thoracic Society.33 Flansbjer and colleagues34 found the 6-minute walk test to have high test-retest reliability (ICC=.99) and good sensitivity to change (SEM%=4.8%). The smallest real difference for indicating a clinical change has been found to be 13%.34
The SIS (version 3.0) was completed by participant interview and was used to evaluate the impact of stroke on participation in life activities.35,36 The test-retest reliability of the SIS has been established (ICC=.57–.92).35 Discriminant validity across the 8 domains has been determined, and criterion validity has been found to be moderate to high (.44–.84).35 The minimal clinically important difference in individual SIS domain scores is considered to be a 10- to 15-point change over time.35 The raw data were analyzed with specialized software from the Rehabilitation Outcomes Research Center
, and percent change was calculated for each area.
Procedure
Evaluator training for dependent variable measurement techniques was conducted by an experienced therapist with neurological clinical specialist certification. Training included practicing measurements on people with and people without stroke until clinical competence was attained. A familiarization session for the purposes of introducing the dependent variables and evaluation procedures was conducted with the participant. Following the familiarization session, SS variables were measured 10 times over a period of 3 weeks. Repeat measurements were obtained in order to establish stable baseline measures necessary for analysis. Measurement sessions typically took place in the late afternoon, and variables were tested in the same order. Although rest breaks were encouraged, the participant usually completed all sessions without a break. Each of the 3 other dependent variables—strength testing, the 6-minute walk test, and the SIS—was added to one of the measurement sessions, making 3 of the 10 sessions longer than the others.
During the intervention, the participant was seen 3 times per week for 8 weeks, for a total of 23 treatment sessions. Sessions were scheduled at the participant's convenience, typically in the late afternoon. One session was missed during week 6 because of a work conflict for the participant. Each of the SS variables was measured once per week at the start of an intervention session. A brief rest period was allowed before the intervention began. During the rest period, the participant and researchers often engaged in casual, unstructured conversation regarding the participant's activities and lifestyle.
Intervention sessions included 3 bouts of running, each lasting to participant tolerance or to a maximum of 10 minutes. Bouts typically consisted of a 1.5-minute warm-up at a self-selected fast walking speed, 7.5 minutes of running, and a 1-minute cool-down. All walking and running speeds were self-selected by the participant, but he was encouraged to go as fast as he could. During each intervention session, therapist feedback consisted of visual and verbal cues to facilitate optimal symmetry and alignment. Cues typically centered around instructing the participant to increase left step length, control the right ankle at initial contact, and equalize arm swing. Manual cues were not provided because the treadmill speed was too high for effective cueing.
At the beginning of the intervention phase, the participant was observed running on the treadmill at a comfortable, self-selected speed with 20%, 10%, and 0% BWS. It was determined that 20% BWS facilitated the best running technique, as defined by an upright and slightly forward trunk, maximum hip extension at the end of stance, effective swing through, and optimal step length and heel contact. This was also the amount of support preferred by the participant. Thus, the intervention began with 20% of the participant's body weight supported.
Our technique for adjusting BWS at the beginning of each bout involved fitting the participant with the harness, which was attached to the LiteGait BWS system. The participant stood on the treadmill in stationary double stance, and BWS was increased until the participant's full body weight was supported through the harness. The BWS was then decreased to the appropriate percentage of support. This technique prevented the harness from slipping and also allowed for consistency of BWS across intervention sessions.
The intervention progressed through decreases in BWS and increases in treadmill speed and running time. The participant determined running speed for each bout, but the researchers manually controlled any change in speed. The participant was unaware of the actual treadmill speed so as to allow him to focus more on his running pattern and perceived exertion, but he requested that the speed be increased or decreased. The participant was encouraged to run at a speed that he could maintain for the duration of the 10-minute bout. The participant was informed that the intervention could be terminated at any time per his request. Blood pressure and heart rate were monitored prior to and following each bout. The ACSM Guidelines for Exercise Testing and Prescription37 was to be used to guide decisions about the termination of a session; however, intervention was never stopped on the basis of these guidelines. The specific intervention progression is shown in Table 3.
|
During the retention phases, performance on all dependent variables was measured. The SS variables were measured 9 times during the retention phases, and the strength and 6-minute walk test variables were measured once during each retention phase. The SIS variable was measured only at 6 months after the intervention. The immediate postintervention phase and the 6-month postintervention phase lasted for 3 weeks to allow us time to gather the repeat measurements necessary for analysis. These phases were conducted to measure the short-term effects of the training and to determine whether the effects were maintained across time.
Data Analysis
The SS variables were checked for serial dependence by use of autocorrelation coefficients calculated with the SingWin|| program. A significant autocorrelation, indicating serial dependence of successive observations, interferes with the interpretation of statistical results. No autocorrelations were present.26 The 2-standard-deviation-band method was selected for analysis because it is appropriate for baselines with greater variability: there is a larger band width outside which a subject's performance must occur for change to be considered statistically significant.26 Two consecutive points outside the 2-standard-deviation band indicated a statistically significant difference (alpha value of <.05) across phases.
The baseline and retention phase variables from the baseline, immediate postintervention, and 6-month postintervention phases were compared for demonstration of changes in the participant's status. The changes are described here.
| Results |
|---|
|
TopAbstractIntroductionMethodResultsDiscussionConclusionReferences |
|---|
|
|
|
|
| Discussion |
|---|
|
TopAbstractIntroductionMethodResultsDiscussionConclusionReferences |
|---|
Our results also supported our hypothesis that the participant would improve his running ability and would maintain these improvements for at least 6 months. It is interesting to note that our participant not only maintained but also furthered his improvements in 25-m sprint time, single-limb balance, running step width, running step length ratio and, in 11 of 16 muscles, strength measures at 6 months after the intervention. We did not anticipate this outcome but, upon learning that during the 6-month break the participant began to participate in independent over-ground running, we found the explanation for the continued improvements.
At the first 6-month postintervention visit, the participant reported that he and his wife were running together 3.2 km (2 miles) 3 times per week. For them, this was the most important outcome from the study. Running was an activity that they had enjoyed together prior to the stroke, for both social and fitness reasons, and they were genuinely thrilled to be able to run together again. The participant viewed this achievement as a huge step in returning to the normalcy of life prior to his stroke. His return to running resulted from his newfound ability to "start off slowly, get into a rhythm, and increase my speed the rest of the way." We believe that the task-oriented training allowed him a unique opportunity to develop an adaptive motor pattern for running that was optimal within the constraints of his "new" system of motor production.38 He was able to transfer the adaptive motor pattern to over-ground running with repeated practice during the 6-month break and continue the improvements he had begun to achieve during the intervention. It is also possible that the intervention provided the participant with the self-confidence to try running in the community, resulting in his successful return to recreational running.
In order to increase running speed, an individual must increase either cadence or stride length.3 The results indicated that the participant did not have large increases in step length; therefore, increased cadence most likely led to the improved sprint times. This finding is in contrast to the findings of Gardner and colleagues,7 who reported that increased stride length occurred following a speed-progressive gait intervention for a participant with incomplete spinal cord injury. Differences in the study findings may be attributable to slightly varied training protocols and different participant diagnoses. Despite these differences, both training protocols focused on progressively increasing training speeds and resulted in increased over-ground running speed and ability following the completion of the protocols. Furthermore, our findings support evidence pertaining to speed training in athletes, in that task-specific, speed-dependent training optimizes running performance and sprint speed.39,40
Another possible explanation for the improved running speed is that the intervention, which demanded continuous fast speed training, improved the spatial and temporal aspects of the participant's running pattern. Therefore, the improved speed could indicate motor learning, as described by the participant during one of his 6-month postintervention visits, "I've figured out how to do it."
The participant's step length ratio showed a trend toward a decrease (from 1.02 to 0.92) from the baseline phase to the 6-month postintervention phase (Fig. 4). This surprising finding refuted our hypothesis that the participant would improve running symmetry with training. It is possible, however, that symmetrical step length should not have been the goal for this individual. Winter and Sienko41 argued that there is a lack of scientific basis supporting protocols that focus solely on training symmetry in people with asymmetrical neuromuscular systems. Instead of seeking symmetry, people may develop adaptive motor patterns that work within the constraints of the systems.38 This notion was demonstrated by our participant. His sprint speed improved even though his running step length became slightly less symmetrical. It is possible that his self-reported improvements in establishing a rhythmic running pattern (timing) were the key to his improved running ability. Evidence obtained from healthy people who run, however, indicates that future musculoskeletal complications may arise because of an asymmetrical running pattern.4
The participant showed significant improvement in left single-leg balance from the baseline phase to the immediate and 6-month postintervention phases (Fig. 2). Although performance demonstrated statistically significant improvement, performance remained inconsistent and variable, as indicated by the increasing standard deviations across phases. Despite improved strength in his ankle musculature (Tab. 1), he still may have been in an early phase of motor learning in which performance improves but more variability exists. The results of the present study support the findings of Williams and Goldie6 that single-leg balance is predictive of running ability. Our participant's improvement in single-leg balance may have allowed better control during the single-limb support phase and thereby may have led to better running ability, as evidenced by significantly decreased step width and increased running speed.
Although we did not directly train strength, we anticipated that facilitating appropriate and repetitive activation of muscle groups required for running would result in increases in strength. This assumption was supported. At the 6-month postintervention phase, right dorsiflexion, right hip flexion and abduction, and left hip flexion were all within 1 standard deviation of reported strength norms.42 There were generally greater gains in strength in the involved LE (left) than in the uninvolved side. The most substantial increase in strength from the baseline phase to the 6-month postintervention phase involved the left hip flexor muscles, which showed a 167% improvement in strength. This result is notable because the hip flexor muscles are primarily responsible for increased power production during running.4 The intervention may have allowed the participant to learn to activate the hip flexor muscles from a stretched position and, through repetitive practice, make impressive gains in strength.
The participant demonstrated an increase in the 6-minute walk test of 241 m (42%) from the baseline phase to the immediate postintervention phase, indicating a clinically significant real difference for this parameter.34 From the baseline phase to the 6-month postintervention phase, an increase of only 63 m (11%) occurred. Although this finding seems to indicate that the participant lost some endurance from the immediate postintervention phase to the 6-month postintervention phase, the apparent decrease was likely attributable to a difference in exertion during testing by the participant. During the immediate postintervention phase, he reported that he had worked somewhat hard during the 6-minute walk test; during the same walk test at the 6-month postintervention measurement, he reported that the walk required no exertion at all. Although the 6-minute walk test is considered a very reliable tool34 and the same instructions given for immediate postintervention testing and 6-month postintervention testing should have provided similar effort on the part of the participant, such may not have been the case in this particular situation.
Clinically important differences in the SIS were achieved in the domains of emotion, handicap, and stroke recovery (changes of 
10 points), demonstrating that training and improved running performance positively changed the participant's perception of his disability.35 He reported that he felt less "handicapped" and had increased his ability to participate in meaningful recreational activities.
There are important clinical implications to be derived from the present study. Running may be overlooked as a clinical goal for people with a high level of function and stroke, TBI, cerebral palsy, or other neurological conditions, despite the potential to regain running ability. Typically, therapists must discharge patients as soon as they have the minimal skills needed to be safe in the home and community, limiting the opportunity to work on high-level skills that may have personal importance to patients. This was the case for the participant in our study. The clinical research setting provided an opportunity for this participant to receive an intervention that allowed him to achieve his goals.
Besides the clinical research setting, running training could be delivered in an outpatient physical therapy setting, pending the availability of necessary equipment and personnel. It is unlikely that most third-party payers would provide coverage for this sort of intervention, but it is possible that some people would be able to pay privately. Running training could be incorporated into supervised postrehabilitation community-based fitness programs that have been shown to be effective in improving and maintaining mobility in people after stroke.43–45 In this setting, people might participate in supervised over-ground or BWST running programs. People should be screened for cardiac and orthopedic contraindications by a physical therapist prior to participation in this sort of community-based program, and some settings may require that a physician's release statement be obtained. The physical therapist would likely set up training parameters and provide continuing consultation during the training; however, the intervention could be delivered by a physical therapist assistant, exercise physiologist, or fitness professional who has received appropriate training in how to supervise exercise for special populations, specifically, how to provide running training for people with neurological impairments in either the over-ground or the BWST running environment. Because of concerns regarding the reduced cardiovascular fitness levels commonly found in people with neurological impairments,45,46 a running intervention program should be encouraged only if the professional overseeing the program is also trained in monitoring cardiovascular responses to exercise in accordance with American College of Sports Medicine guidelines.37
The limitations of a single-subject research design include the limited generalizability of study conclusions. The A-B design is quasi-experimental, and when a change in a dependent measure occurs at the onset of the intervention, a relationship between independent and dependent variables is established. Without the withdrawal phase (as in the A-B-A design), however, there is no control for the historical threat to internal validity, and it becomes possible that an influence other than the independent variable (in this case, training) may have caused the changes in the dependent variables (functional measures). Unfortunately, a withdrawal phase for rehabilitation interventions is not appropriate—because it is hoped that people will not return to the baseline after the interventions. In the present study, we used immediate postintervention and 6-month postintervention retention phases to help strengthen the relationship between dependent and independent variables. More studies are needed to investigate the effects of intensive, task-specific training on running in people with neurological conditions. Future studies could involve other diagnoses, larger sample sizes, or different protocols for training.
| Conclusion |
|---|
|
TopAbstractIntroductionMethodResultsDiscussionConclusionReferences |
|---|
| Footnotes |
|---|
This study was approved by the Institutional Review Board at the University of Indianapolis.
A poster presentation of this research was given at the Combined Sections Meeting of the American Physical Therapy Association; February 1–5, 2006; San Diego, Calif.
* Mobility Research, PO Box 3141, Tempe, AZ 85280. 
Nautilus, 12032 Hwy 155 North, Tyler, TX 75708.
Hogan Health Industries, Medical Products Division, 111 East 12300 South, Draper, UT 84020-0957.
Rehabilitation Outcomes Research Center, 6003 Wescoe Pavilion, Mail Stop 1039, 3901 Rainbow Blvd, Kansas City, KS 66160.
|| SingWin (to accompany Bloom, Fisher, and Orme's Evaluating Practice: Guidelines for the Accountable Professional, 4th edition), copyright 2003 by Auerbach, Schnall, and LaPorte; licensed to Allyn and Bacon, a Pearson Education Company.
| References |
|---|
|
TopAbstractIntroductionMethodResultsDiscussionConclusionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
