Limb-loaded cycling program for locomotor intervention following stroke.

Intervention to improve locomotor capability is problematic during the early stages of rehabilitation following stroke. Functional training of people who have severely limited ambulatory skills (eg, those with a Functional Independence Measure [FIM] score of 1) is often a time-consuming task, which requires extensive assistance because these individuals may be unsafe in gait activities. Effective interventions for gait training during the early stages following stroke are scarce. For example, although body-weight-supported treadmill training has been shown to improve the walking function of people with hemiplegia, (1-8) few adjunctive exercises exist that can complement the force-generating capability training and lower-limb patterning that is practiced during body-weight-supported treadmill training.

For many people in the early stages following stroke, cyclical leg exercise is a safe, task-oriented locomotor intervention that is used to supplement functional ambulation training. (9-12) The kinematic patterns of both locomotor tasks are similar. Both locomotor tasks are cyclical; require reciprocal flexion and extension movements of hip, knee, and ankle; and have alternating muscle activation of antagonists (13,14) in a well-timed and coordinated manner. Kautz and Brown (9) reported that successful cyclical leg exercise requires that the legs, whether impaired or not, follow essentially the same trajectory. Because many characteristics of cyclical leg exercise are similar to those of walking, cyclical leg exercise can potentially play a valuable role as an adjunctive form of locomotor intervention for people with ambulatory dysfunction in the acute rehabilitation setting. Although cyclical leg exercise is not walking, the task of maintaining a simplified locomotor pattern and simultaneously supporting partial body weight and learning to weight shift is task-oriented rather than muscle-specific (eg, exercises that target specific muscle groups in planar movements or nonfunctional patterns of movement).

Ergometer cycling offers training characteristics that differ from those of walking, some of which are advantageous to the study of mobility and muscle coordination. (9-11) Because balance is not a factor in seated pedaling, individuals are not subjected to risky postural disturbances associated with the initial stages of upright locomotor training. By excluding balance from the demands of the locomotor task, individuals may be more independent and capable of focusing their attention on resolving lower-limb locomotor impairments, such as muscle weakness, that contribute to walking dysfunction. In locomotor activities that are free from balance demands, existing muscle coordination patterns applicable to propulsion will dominate over patterns associated with balance reactions. (12)

A limitation to traditional cyclical leg exercise is the lack of weight bearing and weight shifting required. In contrast, walking involves the task of shifting weight while performing a cyclical and repetitive task, such that the ability to bear weight dynamically (ie, periodically adjusting to task demands) on the lower extremities underlies ambulation capability. (15) Walking requires that people have adequate lower-extremity force-generating capacity for weight bearing and the ability to time lower-extremity musculature appropriately to load, unload, flex, and extend during specific points in the gait cycle. Weight shift during walking requires bilateral coordination for controlled weight transfer while continuing on a cyclic trajectory. Therefore, we sought to develop an intervention that could include weight-bearing demands as an added dimension to the task of cyclical leg exercise.

Limb-loaded cycling (LLC), as described in this report, offers a task-oriented locomotor training intervention to people who are unable to support their entire body weight. Because only 27% of individuals are able to walk independently within 1 week following stroke, (16,17) cycling provides an alternative means of assessment of weight-bearing capability and intervention during the acute stages of rehabilitation.

Limb-loaded cycling requires coordinated transfer of weight between paretic and nonparetic limbs and demands well-timed lower-extremity force production. Our purpose was to determine whether LLC could provide a safe and feasible adjunct to physical therapy in the training of functional ambulation. Thus, the purpose for this case report is to describe the implementation of an intervention for patients with severely limited ambulation skills following a stroke using LLC training as an adjunct to other methods of gait training in physical therapy. Two individuals who were within their first 2 months following stroke and whose ambulation skills initially were severely limited underwent an LLC intervention and are described in this report.

Case Description

Patients

Patients were recruited from an inpatient rehabilitation facility. Charts were reviewed upon admission to the hospital to determine suitability. We sought individuals who had a first-time stroke that resulted in unilateral hemiparesis, had severely limited ambulation skills, and had the ability to follow simple verbal commands. Consent from the patients' physicians was obtained prior to contact. Patients were given an explanation of the protocol and signed an informed consent form prior to the initial session, which the Northwestern University Institutional Review Board approved.

Patient 1. Patient 1 (P1) was a 77-year-old, right-hand-dominant, man, 10 days following a right parietal ischemic stroke with dysphagia, left facial droop, and left-sided hemiplegia with the upper extremity more involved than the lower extremity. He had a history of borderline hypertension and untreated prostate cancer for more than 3 years. Upon admission to the inpatient rehabilitation facility, the patient's medications included atenolol, lisinopril, aspirin, and clopidogrel bisulfate.

He was in a stable state of health when he developed slurred speech, numbness, and facial droop on the left side. All diagnostic tests for stroke at that time were negative, and symptoms resolved within 45 minutes of onset. The next morning, the symptoms returned, with the addition of left-sided weakness and unsteady gait. A magnetic resonance imaging (MRI) scan confirmed the parietal stroke, and he was admitted to an inpatient rehabilitation facility following an acute care stay of 9 days, which was complicated by persistent hypertension.

Initial medical evaluation showed left-sided perceptual neglect, with mild increases in resistance to passive stretch on the left side. Sensation testing revealed decreased sensitivity to light touch, pain and temperature, position sense, and 2-point discrimination in the left upper and lower extremities. His emotional state was labile and included sudden episodes of crying. The physician referred him for physical therapy to improve his muscle force, balance, and endurance and to increase his independence in activities of daily living.

The initial physical therapist evaluation described P1 as having severely limited ambulation skills. He demonstrated anxiety and fear of falling, which made him an appropriate candidate for the LLC intervention. Anxiety and fear were evidenced by statements about past falling incidents and a strong reluctance to get out of the wheelchair and to practice walking unless at least 2 therapists were available to ensure his safety. His therapist assigned an initial FIM score of 1 (FIM score of 1: total assistance--individual performs less than 25% of effort, requires assistance of 2 people, or does not walk a minimum of 15.2 m [50 ft]) because P1 needed complete assistance to maintain upright support during walking. For this reason, standing balance could not be tested, although P1 was able to sit unsupported for more than 5 minutes. Transfers from bedside to chair required full assistance. Results of range-of-motion testing of the lower extremities were within normal limits, and muscle force in the left lower extremity was poor to fair, indicating that diminished muscle force may have contributed to his severely limited ambulation status. He had slightly increased resistance to passive stretch in the left lower extremity, with increased effort and with passive range of motion (PROM) applied at rapid speeds. Observation of his standing posture indicated decreased hip, knee, and trunk extension on the left side, decreased weight bearing on the left lower extremity, impaired midline orientation, and a right gaze preference. The physical therapy plan of care was to see P1 five times per week, for 1 hour per session, focusing on therapeutic exercise, gait, and balance training to improve mobility. The therapeutic goal was to discharge P1 to a home environment where he could be ambulatory with minimal assistance from a family member.

Patient 2. Patient 2 (P2) was a 68-year-old, right-hand--dominant, Albanian-born woman who had a left frontal lobe cerebrovascular accident. She was sitting at a bus stop in Switzerland when she experienced sudden mild weakness on the right side of her body that progressed quickly, until she was unable to walk. She was admitted to a hospital in Macedonia, where she stayed for 2 weeks. No medical records of her stay were available. Upon return to the United States where she currently was living, she was admitted to an acute care hospital and was diagnosed as having had a left frontal lobe infarct with periventricular demyelination. After 5 weeks at the acute care hospital, she was transferred to the inpatient rehabilitation hospital, where our training began approximately 7.5 weeks after the date of infarct.

The initial medical evaluation note said that the patient had right lower-extremity paresis, hypertension, and slight resistance to passive stretch in the right upper and lower extremity. She had slight nonfluent aphasia, but had the ability to follow multistep commands. Sensation testing showed intact sensitivity to pain, light touch, and position in all extremities. The physician prescribed physical therapy to improve muscle force, balance, and endurance; heparin to prevent deep vein thrombosis; and aspirin as a stroke prophylactic. She was to take acetaminophen and temazepam at the discretion of the registered nurse practitioner.

The physical therapist examination showed no limitations in PROM. However, P2 had little active movement in the right lower extremity. Manual muscle testing scores for the right lower extremity were 0/5 to 2/5 for all tested planes. She showed slight resistance to right hip flexion. P2 stood with decreased weight bearing on the right lower extremity and with hyperextension of the right knee. She was able to maintain standing balance without upper-extremity support, but required contact guard assistance. Minimal to moderate assistance (ie, amount of effort produced by the therapist to prevent her from collapsing or falling over) was required to maintain standing balance when perturbations were introduced.

Initial mobility assessment showed that sit-to-stand transfers required bilateral upper-extremity support. P2 was able to ambulate 1.5 m (5 ft) in the parallel bars with assistance to advance the right lower extremity, and she severely hyperextended her right knee with weight bearing. Because of the limited distance she could walk and maximal assistance that was required, her initial FIM score was documented as 1 for walking capabilities.

The physical therapist's long-term ambulatory goal was that P2 would be able to walk 45.7 m (150 ft) with minimal assistance using an assistive device. The physical therapy plan of care was to see P2 five to seven times per week, for 30 minutes to 1 hour per session, focusing on therapeutic exercise, gait, and transfer training to improve mobility.

Throughout the period that P1 and P2 were undergoing the LLC intervention, they were engaged in twice-daily physical therapy sessions consisting of 1 hour of individualized stretching to all of the major joints of the upper and lower extremities; muscle force training; pre-gait activities, such as bridging and rolling exercises to encourage weight bearing and sitting/standing balance training; and 1 hour of group muscle force training classes consisting of supine and side-lying mat exercises. P1 and P2 also received 1 hour of occupational therapy.

Equipment Used: Limb-Loaded Cycle Ergometer

The patients were evaluated and trained using a modified Biodex semirecumbent isokinetic ergometer. * This device included a waist belt, upper-body seat belt, and straps on the pedals to provide support and safety. The apparatus included a releasable seat that was allowed to slide along a linear track similar to a leg press machine. To prevent the seat from sliding forward, the patients were required to maintain a loading force on the sliding seat mechanism, set at a specific chosen percentage of their body weight, throughout the pedaling cycle. To successfully complete the task, the patients were required to maintain their support lower extremity in an extended and loaded position. If they were unable to maintain the load, the seat slid forward until stopped by a metal safety block on the track. The block prevented excessive forward displacement of the seat and provided auditory feedback so that when the patients heard the seat collide with the block, they knew they were in an unloaded position. The position of the block was determined so that, when the seat was resting on it, the crank of the pedal of the nonparetic lower extremity was in a position perpendicular to the ground, with the nonparetic knee in 90 degrees of flexion.

Progressive levels of loading were accomplished by an adjustable gas-spring mechanism that was attached to the seat by a cable. Two 6-axis ATI Delta triaxial force transducers ([dagger]) were mounted on the pedals to measure load settings by detecting normally directed pedal reaction forces when the patient's nonparetic lower extremity was in a fully extended, loaded position. Data from the force transducers were collected at a sampling rate of 200 samples per second for up to 30 seconds using LabView data acquisition software ([double dagger]) and a National Instruments A/D board. ([double dagger])

Initial Limb-Loaded Cycling Examination

The initial bicycle examination began with a resting heart rate and blood pressure reading. After the patient was transferred onto the bicycle, adjustments were made for individual leg length dimensions. A 2-minute warm-up was performed at a comfortable pedaling speed with the seat in a fixed position prior to LLC examination.

The LLC required the patients to load their limbs with an imposed load equivalent to a percentage of their body weight by pushing a sliding recumbent seat posteriorly using force generated at the legs and received at the pedals. The patient was to cycle while remaining in this loaded position, with the seat held posteriorly. A successful cycle was achieved when the paretic lower extremity completed the extension (loading) phase of the pedaling cycle while preventing the seat from colliding with the metal block placed on its track.

Patients began the test with a load equivalent to 20% of their body weight. They were instructed to begin cycling at a comfortable speed for 2 to 3 revolutions. They were then asked to push the seat back as they continued to pedal and to remain off of the metal block. A total of 10 limb-loaded revolutions were attempted, and the number of successful revolutions on the paretic lower extremity was recorded. Five or more successful revolutions were considered a successful set and allowed for testing at a higher load, and fewer than 5 revolutions would warrant decreasing the load or stopping the test. Load changes were made in 5% increments until the patient's initial LLC capability was determined.

Initial pedaling examination for P1 showed an inability to perform LLC at 15% of his body weight with his paretic lower extremity. P1 was able, however, to pedal a standard cycle ergometer with a fixed seat and no imposed load. P2 also showed no LLC capabilities at 15% of body weight. She had repeated, involuntarily interrupted and backward cycling, and the paretic lower extremity resisted upstroke in its pedaling cycle and needed assistance to complete a full cycle. She also experienced difficulty completing the transition from upstroke to downstroke on the paretic lower extremity and required assistance in making the transition.

These initial pedaling results are of interest because P1 showed greater lower-extremity control and muscle force during the standard cycling task despite the differences in the patients' physical therapy ambulatory examinations. P2 was ambulatory over short distances (less than 10 m), yet was unable to carry out the standard cycling task. P1 was more capable of performing the locomotor task of LLC, yet was initially unable to stand without assistance from 2 therapists.

P1's success at the pedaling task, when compared with walking, made him a good candidate for the LLC intervention. Ambulation training was expected to be difficult because he had to overcome the deficits associated with unilateral neglect and would have difficulty loading his entire body weight on the paretic limb. In contrast to gait training, the LLC training was expected to provide him with a safe setting in which he could focus his attention on weight bearing and weight shifting onto his paretic lower extremity, combating his unilateral neglect, while learning how to coordinate the limbs during a cyclic locomotor task. In contrast to P1, P2 was able to walk over short distances. The rationale for LLC training for P2 was based on the expectation that consistent and repetitive practice of a locomotor weight-shifting task would enhance the effectiveness of daily gait training sessions.

Limb-Loaded Cycling Intervention

The LLC training began at the load deemed successful during the initial examination or at the minimum load of 15% of body weight. The intervention consisted of 10 to 14 training sessions, with 10 sets of repetitions in each session. A 2-minute break was provided between trials, during which heart rate and blood pressure were evaluated. In contrast to the initial load examination when 10 repetitions were attempted, 20 repetitions were attempted in each set and the number of successful repetitions completed by the paretic limb determined the load setting for the following set.

A specific protocol was developed that models a weight-training program emphasizing motor control. Muscle force training protocols are common in the rehabilitation literature; however, our emphasis on motor control is unique. The standard set for this control-focused muscle force training program required that an individual be able to complete 12 to 15 repetitions at a particular load to be deemed successful at that load. Patients were required to perform more than 15 successful repetitions at a particular load before attempting a higher load. If fewer than 12 successful repetitions were completed, the load would be decreased.

P1 underwent 13 training sessions of independent LLC training. At no point in the intervention were we required to interrupt training as a result of the patient's response to exercise. Both patients' heart rates never exceeded 110 beats per minute, and blood pressure readings never exceeded 180/100 mm Hg. The intervention did not cause either patient to display signs or symptoms of overexertion, as defined by our limits on blood pressure and heart rate values.

Based on both the physical therapist examination and the initial pedaling examination of P2, it was evident that she had severe right lower-extremity weakness and lack of lower-extremity coordination. She was unable to bear weight when loading was required or to adequately flex during unloading tasks with her paretic lower extremity during both walking and pedaling. Although the initial examination revealed that she was unable to complete the pedaling task while sustaining 15% of body weight on either the paretic or the nonparetic lower extremity without assistance, we believed that she would benefit from our proposed intervention. Her frequent display of interrupted pedaling and backward pedaling suggested that she required some basic pedaling training to improve her lower-extremity coordination and muscle control, and therefore we found it necessary to implement a preliminary LLC training program.

Hybrid LLC Program for P2

The decision to provide P2 with a hybrid training program instead of decreasing the load to less than 15% of body weight was made to preserve the nature of the LLC task. Excessively low loads alter the demands of the activity so that generation of flexor support moments are required to stabilize the seat in place of extensor support moments during the support phase of the task. Maintenance of the demands of the LLC task is necessary for maintaining its representation of a locomotor task.

A hybrid training program was implemented and comprised 2 activities: assisted LLC and assisted single-leg LLC. Once P2 had the ability to perform the pedaling task independently, she began LLC training similar to P1 and continued for the final 5 sessions before discharge. In assisted LLC, the therapist held the seat back, supporting the portion of load that P2 was unable to maintain. Similar to an active-assisted range-of-motion exercise, as the patient's ability to support higher loads improved, the amount of assistance was decreased. We anticipated that the assisted LLC training would help her become familiar with the LLC principles, while decreasing the demands of the activity. Her coordination was challenged with a load that was graded to her abilities. We also expected that this graded limb-loaded exercise would allow her to experience the sensation of loading on her paretic lower extremity and that she soon would be able to understand how to resist and maintain appropriate amounts of load while performing a locomotor task.

Assisted single-leg LLC involved removing the nonparetic lower extremity from the pedal and holding it so that it could not contribute to the generation of force in the paretic lower extremity. Again, the seat was held back to decrease the load imposed on the paretic lower extremity and to support the load during upstroke. Introduction of this activity allowed for the isolation of muscle force and coordination training to the affected lower extremity. It also allowed P2 to focus on the paretic side during the activity, and we anticipated that the skills gained from this intervention would carry over to 2-leg cycling. P2 was intermittently allowed to experience the entire load (15% of body weight) under both conditions to decrease her dependence on the assistance and to remind her that independent pedaling at 15% of body weight was the eventual goal of the intervention.

The patient's muscle force and coordination improved, which allowed her to work on 3 preparatory subtasks of the independent LLC protocol: (1) generation of adequate extensor force from both lower extremities during downstroke; (2) reduction of inappropriate extensor resistance in the paretic lower extremity during upstroke, thus allowing for flexion during upstroke and successful extension during the downstroke of the nonparetic lower extremity; and (3) achievement of appropriate timing for the activation and deactivation of flexors and extensors during the transition phases of the pedaling cycle for the paretic and nonparetic lower extremities.

Physical therapy intervention during this 5-day period included gait training with the use of a rolling walker, transfer training, and strengthening exercises for the paretic lower extremity. The patient progressed in her walking function over the course of the assisted LLC intervention. Initially, she hyperextended her paretic knee during the stance phase of walking in approximately 75% of her steps. The frequency of hyperextension decreased to 40%, and weight shifting onto the paretic side improved. The patient's ambulation with a rolling walker improved from walking 15.2 m (50 ft) with moderate assistance to walking 30.5 m (100 ft) with minimal assistance. Her FIM score improved from 1 to 2 (FIM score of 2: maximal assistance--patient performs 25% to 49% of locomotion effort to go a minimum of 15.2 m; requires assistance of one person only).

Primary Outcomes

Tables 1 and 2 depict the patients' improvement in LLC and ambulatory function over the course of the intervention. A unit described as the dynamic load index was used to calculate the product of load tolerated (expressed as percentage of body weight) and the number of times that the paretic limb successfully completed a limb-loaded cycle. This index approximates the amount of mechanical work done during each session. Both patients progressed during the intervention in both measures of interest: percentage of body weight load tolerance and dynamic load index score (eg, product of percentage of body weight x number of repetitions).

Ambulation Outcomes

A 6-minute walk test was included in the protocol for the walking evaluation. This test has been demonstrated to yield reliable and valid measurements of functional walking ability for people following stroke. (18) To do the 6-minute walk test, the patient needed to be able to walk with contact guard assistance or less, according to the recommendation of the patient's physical therapist. The patients were to ambulate with any assistive device or orthosis that they normally used during physical therapy. Testing was conducted on an indoor 10-m track. The floor surface was wood, and the patients wore shoes suitable for walking. A physical therapist was present during the walk test to ensure safety, but did not provide any verbal or tactile cueing during the walking assessment. The first 10 m of the walk test was recorded using a digital video camera, and the videotape was later analyzed to calculate speed and cadence. This test is equivalent to the standard 10-m walk test, which has been shown to yield reliable and valid measurements of gait speed following stroke.18.19 A stopwatch was used to time the 6-minute walk test, and taped indicators on the floor were set every 2 m to calculate distance traveled. The verbal instruction prior to commencing the walk test was, "Walk as quickly as you can and as safely as you can." Rather than label the test as fast or habitual, we have stated the instructions because the combination of speed and safety results in fast walking for some people and "safe" walking for others.

Progression in ambulatory function coincided with LLC function. On his first day of LLC training, P1 walked and performed pre-gait activities poorly during physical therapy. He needed maximal verbal and tactile cueing to maintain his paretic hip and knee extended, and he had considerable difficulty with weight shifting. These problems led the physical therapist to lower expectations and to believe that P1 would always need maximum assistance to walk safely. By the end of the first week of training (day 5), P1 was able to walk 3 m using a rolling walker, with moderate assistance (ie, moderate levels of effort by the therapist to hold the patient upright and prevent the patient from falling). At the end of the second week (day 10), P1 was able to walk 26 m, 2 times, using a wide-based quad cane with minimal assistance. He also was able to ascend and descend 3 steps twice with moderate assistance. Although he continued to require tactile cues (eg, tapping and brushing) and maximal verbal cues (eg, "remember to move your leg forward") to perform ambulatory activities, he had exceeded the physical therapist's expectations, and his FIM score improved from 1 to 2.

P2 progressed from ambulation with a rolling walker for a distance of 33 m with minimal assistance to 58 m with supervision over the 5 independent LLC sessions (sessions 6-10). Her FIM score improved from 2 to 4 for walking function over this period, and her FIM score was 5 at the time of discharge (FIM score 5: supervision--individual requires standby supervision, cueing, or coaxing to go a minimum of 45.7 m [150 ft]). At the conclusion of the training period, P1's gait speed was 0.12 m/s, his cadence was 0.39 steps per second, and he walked 42 m on the 6-minute walk test. P2 walked with a speed of 0.2 m/s, her cadence was 0.82 steps per second, and she walked a distance of 58 m during the 6-minute walk test using a rolling walker with close supervision.

Discussion

The patients' outcomes indicated that the LLC intervention was safe and that an individual with unilateral neglect and an individual with initially poor limb-loading capability could participate in the intervention. Although locomotor training often is difficult during the early phases of locomotor acquisition following stroke, our patients' experiences indicate that the device and protocol described in this report may give pre-locomotor training that can be achieved at a high intensity and with a focus on weight bearing and load tolerance.

Both patients illustrated rapid progression over the first week of training (Tabs. 1 and 2). In addition to the possible role of exercise-induced muscle force gains, this outcome may be attributed to familiarity gained in the new task that most often occurs during the early stages of an intervention and to the natural neurological recovery that is associated with acute stages of rehabilitation following stroke.

The intervention was safe for both of these individuals because neither had interruptions of training due to exceeding safety parameters. The parameters were defined by the heart rate and blood pressure limits set by the physician, the patients' complaints of pain or shortness of breath, and the physical therapists' observations of overexertion signs and symptoms. Thus, we deemed the intervention safe because the patients had no changes in heart rate and blood pressure that exceeded the physician's set limits during the intervention. They also had no signs of abnormal response to exercise such as pallor, cold sweat, or unusual heart rate response (ie, too rapid an increase, failure to increase, or a decrease with exercise) and had no symptoms of overexertion such as anginal response, no dyspnea, no excessive fatigue, no mental confusion or dizziness, and no claudication of the lower extremity. Neither individual reported any incident of musculoskeletal strain.

P1's anxiety and fear of falling made him an appropriate candidate for the LLC intervention. As training progressed, P1 enthusiastically participated in walking sessions with his physical therapist and no longer demanded extra therapist assistance to provide assurance of safety. In our experience, people who have a fear of falling most often establish a dependency on therapists or assistive devices that may limit their function. In such cases, the use of an LLC ergometer can offer a safer environment in which they can gain comfort and confidence in using the lower limb during locomotion. The LLC ergometer also can offer the benefits of reducing individual anxiety and promoting independence.

Variability of load tolerance during sessions was attributed to the patients' discovery of appropriate load tolerance for 20 repetitions. These fluctuations were due, in part, to the nature of the protocol, where the achievement of the predetermined standard of proficiency (ie, at least 16 successful cycles) at a particular load resulted in a 5% load increase. Initially, the patients were rarely able to perform the task at the new load setting, and thus alternated between the levels until either their locomotor function improved or they became fatigued.

P1's left-sided neglect seemed to be the main barrier to progress in his therapy. In an effort to address this impairment, he was encouraged to focus on his affected side, and performance of the paretic lower extremity was emphasized. In addition, the strenuous daily schedule often caused generalized fatigue and may have caused variability in performance between trials. Despite fatigue, he continued to tolerate increases in weight-bearing demands without exceeding the cardiovascular parameters.

P1's emotional status was also an important factor in his pedaling performance. His performance improved with verbal encouragement and goal setting. He enjoyed keeping track of his best paretic lower-extremity scores and attempted to surpass them. He appeared to show less effort when the load was increased, perhaps because he feared advancing in level of difficulty, and he felt more comfortable training at lower loads. When he was told that the load had been decreased, he seemed more willing to provide his maximal effort. When he was told that there were few trials remaining, he was most willing to expend energy.

In some cases, our instructions caused undue anxiety during the exercise session. Initially, we standardized the instructions and feedback given to the patients during the intervention. We found that performance improved, however, when we modified our cueing to adapt to the patients' learning style and cognitive deficits. For example, P2 required maximal cueing during the intervention to remind her to maintain the load with her paretic lower extremity while pedaling. She responded to the attention of coaching with increased force output. P1 required instructions to keep the seat away from the metal block and to focus on his paretic lower extremity during each trial. He also required goals of number of successes to be set before each trial and performed best when told how many successes he had achieved. Informing him of load increases often caused him to express apprehension for the next trial, and his performance declined more than expected. When told that the exercise was expected to be more tolerable for him, he often exceeded expectations. Thus, although cueing and instructions were not standardized, they were intended to optimize individual performance.

P2 was able to achieve her highest value on the dynamic load index because she was consistently successful at a moderate load. P2 appeared to be highly motivated by verbal encouragement and performed best with praise and by setting personal goals. On numerous occasions, she said that her paretic lower extremity felt stronger than before and that the intervention was helping her walk more independently. She also indicated that she intended to use a stationary bicycle for muscle force training purposes following discharge.

Limitations of Limb-Loaded Cycling

One possible limitation to the exercise protocol was the fatigue caused by the activity. We found that both patients performed at lower loads on the fifth consecutive day of training, when compared with those of the previous session. The training schedule can be modified to allow for adequate rest between training sessions. For example, a "day off' may be provided every third day to reduce or eliminate the observed fatigue effects.

Although cycle ergometers are commonly used in the clinic to address cardiovascular training for people following stroke, they are not often applied as a task-oriented locomotor training tool. The limitation to specificity of training lies in the lack of challenge to balance and postural systems in an upright position. For example, P1 was unable to stand and balance in an upright position while showing an ability to pedal the limb-loaded cycle ergometer. In contrast, P2 was able to stand and balance in an upright position, but was unable to pedal the limb-loaded cycle ergometer. Therefore, balance and posture issues do not appear to be addressed well with LLC.

Given these limitations, it would seem appropriate to pair cycle ergometry, in particular LLC, with other physical therapy interventions that target balance and postural control in an upright position. Examples include balance board exercises, functional reaching exercises, and more standard gait training techniques, including those carried out in body-weight-supported environments (eg, overhead harnesses and aquatic therapy). At least one current phase III clinical trial is examining the combined effects of LLC and bodyweight-supported treadmill training for people following stroke. (20)

Conclusions

Limb-loaded cycling, used with an adaptive training protocol, was a safe and feasible intervention for the 2 patients with acute stroke. When LLC is used as an adjunct to usual and customary physical therapy, it may provide an early intervention for accelerating the return of greater muscle performance in people following stroke, and further studies should examine this question. These cases form some basis for a larger-scale clinical trial that utilizes LLC as an adjunct to physical therapy gait intervention for patients following stroke.

References

(1) Hesse S, Bertelt C, Jahnke MT, et al. Treadmill training with partial body weight support compared with physiotherapy in nonambulatory hemiparetic patients. Stroke. 1995;26:976-981.

(2) Hesse S, Malezic M, Schaffrin A, Mauritz KH. Restoration of gait by combined treadmill training and multichannel electrical stimulation in non-ambulatory hemiparetic patients. Scand J Rehabil Med. 1995;27: 199-204.

(3) Hesse S, Uhlenbrock D, Sarkodie-Gyan T. Gait pattern of severely disabled hemiparetic subjects on a new controlled gait trainer as compared to assisted treadmill walking with partial body weight support. Clin Rehabil. 1999;13:401-410.

(4) Hesse S, Konrad M, Uhlenbrock D. Treadmill walking with partial body weight support versus floor walking in hemiparetic subjects. Arch Phys Med Rehabil. 1999;80:421-427.

(5) Kosak MC, Reding MJ. Comparison of partial body weight-supported treadmill gait training versus aggressive bracing assisted walking post stroke. Neurorehabil Neural Repair. 2000;14:13-19.

(6) Nilsson L, Carlsson J, Danielsson A, et al. Walking training of patients with hemiparesis at an early stage after stroke: a comparison of walking training on a treadmill with body weight support and walking training on the ground. Clin Rehabil. 2001;15:515-527.

(7) Teixeira da Cunha Filho I, Lira PA, Qureshy H, et al. A comparison of regular rehabilitation and regular rehabilitation with supported treadmill ambulation training for acute stroke patients. J Rehabil Res Dev. 2001;38:245-255.

(8) Visintin M, Barbeau H, Korner-Bitensky N, Mayo NE. A new approach to retrain gait in stroke patients through body weight support and treadmill stimulation. Stroke. 1998;29:1122-1128.

(9) Kautz SA, Brown DA. Relationships between timing of muscle excitation and impaired motor performance during cyclical lower extremity movement in post-stroke hemiplegia. Brain. 1998;121: 515-526.

(10) Brown DA, Kautz SA. Increased workload enhances force output during pedaling exercise in persons with poststroke hemiplegia. Stroke. 1998;29:598-606.

(11) Brown DA, Kautz SA. Speed-dependent reductions of force output in people with poststroke hemiparesis. Phys Ther. 1999;79:919-930.

12 Brown DA, Kautz SA, Dairaghi CA. Muscle activity adapts to anti-gravity posture during pedalling in persons with post-stroke hemiplegia. Brain. 1997;120:825-837.

(13) Raasch C, Zajac F. Locomotor strategy for pedaling: muscle groups and biomechanical functions. J Neurophysiol 1999;81:515-525.

(14) Ting LH, Kautz SA, Brown DA, Zajac FE. Phase reversal of biomechanical functions and muscle activity in backward pedaling. J Neurophysiol. 1999;81:544-551.

(15) Winter D. The Biomechanics and Motor Control of Human Gait. Waterloo, Ontario, Canada: University of Waterloo Press; 1991.

(16) Jorgensen HS, Nakayama H, Raaschou HO, Olsen TS. Recovery of walking function in stroke patients: the Copenhagen Stroke Study. Arch Phys Med Rehabil. 1995;76:27-32.

(17) Jorgensen HS, Nakayama H, Raaschou HO, et al. Outcome and time course of recovery in stroke, part II: time course of recovery. The Copenhagen Stroke Study. Arch Phys Med Rehabil. 1995;76:406-412.

(18) da Cunha-Filho IT, Henson H, Wankadia S, Protas EJ. Reliability of measures of gait performance and oxygen consumption with stroke survivors. J Rehabil Res Dev. 2003;40:19-25.

(19) 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.

(20) Physical Therapy Clinical Research Network Web site. Available at: http://pt.usc.edu/clinresnet/index.html. Accessed October 26, 2004.

* Biodex Medical Systems Inc, 20 Ramsay Rd, PO Box 702, Shirley, NY 11967.

([dagger]) ATI Industrial Automation, 1031 Goodworth Dr, Apex, NC 27502.

([double dagger]) National Instruments Corp, 11500 N Mopac Expressway, Austin, TX 78759-3504.

DA Brown, PT, PhD, is Assistant Professor, Department of Physical Therapy and Human Movement Sciences, The Feinberg School of Medicine, Northwestern University, 645 N Michigan Ave, Suite 1100, Chicago, IL 60601 (USA) ([e-mail]). Address all correspondence to Dr Brown.

S Nagpal, PT, MS, was a graduate student in the Department of Physical Therapy, School of Health and Human Services, D'Youville College, Buffalo, NY, when this work was conducted.

S Chi, PT, MSPT, was Research Physical Therapist, Department of Physical Therapy and Human Movement Sciences, The Feinberg School of Medicine, Northwestern University, when this work was conducted.

All authors provided concept/idea/project design, writing, data collection and analysis, project management, and patients. Dr Brown provided fund procurement and facilities/equipment.

Institutional review board approval for this work was provided by Northwestern University.

This work was supported by National Institute of Disability and Rehabilitation Research, US Department of Education, NIDRR grant H133B980021. The limb-loaded cycle ergometer was donated by Biodex Medical Systems Inc.

This article was received February 27, 2004, and was accepted July 19, 2004.

Table 1.

Progression of Training Sessions With Patient 1 (Body Weight=82.6 kg
[182 lb]) (a)

           Load (% BW (b))
                                          No. of Load
Session    Minimum    Maximum    Mode     Changes

 1         15         15         15       0
 2         15         15         15       0
 3         20         25         20       2
 4         20         25         20       7
 5 (c)     20         25         25       4
 6 (c)     15         20         20       6
 7         15         25         20       9
 8         20         25         20       7
 9         20         25         20/25    8
10         20         30         25       7
11         20         30         25       9
12         20         30         25/30    8
13         25         35         25/30    9

           Dynamic Load Index              Ambulatory Status
           (Product of Load x     FIM      (From Physical
Session    No. of Repetitions)    Score    Therapist's Notes)

 1          6.2                   1        Nonambulatory
 2         12.3
 3         16.3
 4         26.3
 5 (c)     21.4                   1        3 m (10 ft), rolling
                                             walker, moderate
                                             assistance
 6 (c)     22.3
 7         24.7
 8         28.1
 9         27.9
10         31.8                   2        24.4 m (80 ft), wide-
                                             based quad cane,
                                             minimum assistance;
                                             up and down 3
                                             steps, 2 times,
                                             moderate assistance
11         33.4
12         33.3
13         31.1

(a) Final walk assessment: Functional Independence Measure (FIM)
score=2, 6-minute walk test measurement=42 m, gait speed=0.12 m/s,
cadence=0.39 step per second.

(b) BW=percentage of body weight.

(c) Session interrupted, includes fewer than 10 trials.

Table 2.

Progression of Training Sessions With Patient 2 (Body Weight=70.8 kg
(156 Ibl) (a)

           Load (%BW (b))
                                         No. of Load
Session    Minimum    Maximum    Mode    Changes

 1 (c)     15         15         15      0
 2 (c)     15         15         15      0
 3 (c)     15         15         15      0
 4 (c)     15         15         15      0
 5 (c)     15         15         15      0
 6         15         15         15      0
 7         15         15         15      0
 8         15         20         15      9
 9         15         20         20      9
10         20         20         20      0

           Dynamic Load Index              Ambulatory Status (From
           (Product of Load x     FIM      Physical Therapist's
Session    No. of Repetitions)    Score    Notes)

 1 (c)     N/A d                  1        15.2 m (50 ft), rolling
                                             walker, moderate
                                             assistance, knee hyper-
                                             extension during stance
                                             in 75% of steps
 2 (c)     N/A
 3 (c)     N/A
 4 (c)     N/A
 5 (c)     N/A                    2        30.5 m (100 ft), rolling
                                             walker, minimum assis-
                                             tance, knee hyperextension
                                             during stance in 40% of
                                             steps
 6         15.0
 7         16.2
 8         20.9
 9         22.7
10         25.8                   4        53.3 m (175 ft), rolling
                                             walker, standby
                                             supervision

(a) Final walk assessment: Functional Independence Measure (FIM)
score=5, 6-minute walk test measurement=58 m, gait speed=0.2 m/s,
cadence=0.83 step per second.

(b) %BW=percentage of body weight.

(c) Session consisted of "assisted" limb-loaded cycling training.

(d) N/A=not applicable.