Locomotor treadmill training with body-weight support prior to over-ground gait: promoting symmetrical gait in a subject with acute stroke.(A clinical report)

McCain, Karen J., and Patricia S. Smith. "Locomotor treadmill training with body-weight support prior to over-ground gait: promoting symmetrical gait in a subject with acute stroke.(Clinical report). ." Topics in Stroke Rehabilitation.  14.5 (Sept-Oct 2007): 18(10). Expanded Academic ASAP. Gale. King County Library System. 5 Feb. 2008 

Full Text:  COPYRIGHT 2007 Aspen Publishers, Inc.

Background and Purpose:

An asymmetrical gait pattern is frequently observed in persons recovering from stroke. Locomotor training with partial body-weight support (BWS) has been demonstrated to be effective for restoring ambulation abilities in persons poststroke. However, the optimal treatment parameters for this intervention have not been defined. The purpose of this case study was to report outcomes for a person in the acute period poststroke who experienced locomotor treadmill training with BWS prior to walking over ground. Method: The subject of this report was a 60-year-old male with a large right-sided infarct extending into the basal ganglia. Locomotor training with BWS began on day 10 following the infarct. The subject had five sessions, totaling 40 minutes of walking practice, prior to starting gait training over ground. Results: The subject walked with a single-point cane (contact guard to standby assistance) for a total of 1,000 feet at a speed of 0.94 m/s (185 ft/min) when discharged on day 25 of rehabilitation. He demonstrated comparable stance time and step length bilaterally. Discussion/Conclusion: The subject of this case report experienced a good outcome using a combination of early locomotor treadmill training and traditional therapeutic activities. Further investigation of early treadmill training may be warranted in subjects with acute stroke. Key words: body weight support, gait, stroke, treadmill

**********

Gait impairment after stroke is common, with many stroke survivors living with residual gait problems, despite extensive rehabilitation. (1) Gait after stroke is perhaps best distinguished by asymmetry and lack of smoothness in forward progression. (2-8) Some of the characteristics of gait after stroke that have been identified include reduced stride and step length, wide base of support, as well as increased stance periods and altered swing phase periods. (9,10) In addition to changes on the affected side, abnormal movements of the upper extremity, trunk, and pelvis, as well as the less affected lower extremity, are common. (5) Wade and colleagues found as many as 30% of subjects after stroke were nonambulatory at 3 months postonset, (11) yet some subjects regain the ability to walk at normal speeds. (12) However, average velocity of gait in subjects after stroke varies between 0.18 and 1.03 m/s, (3,11,13,14) while healthy adults typically walk at speeds of approximately 1.4 m/s. (15)

Dissatisfaction with gait recovery is common after stroke and often persists into the chronic stage of recovery (16) with patients eager to improve their gait quality and stability. (17) However, there is increasing evidence that the typical gait abnormalities seen after stroke are highly resistant to change in the chronic phase of recovery. Buzzelli and colleagues (17) followed 42 subjects (who were at least 1 year poststroke) during a 3-month daily physical therapy program designed specifically to improve gait quality. Kinematic gait analysis was unable to detect any measurable change after the intensive therapy, despite the patients' motivation and expectation of improvement. (17)

There is evidence that locomotor treadmill training with partial body-weight support (BWS) may be an effective method of improving gait quality, speed, and trunk stability following stroke. (18,19) It has been suggested that it may be an easier way to encourage symmetrical gait early in the rehabilitation process (20,21) as well as a more appropriate method to facilitate sensory input= and maximize much needed repetition during the critical recovery period after a CNS insult. (19,23)

The application of locomotor training with BWS in subjects with stroke was first proposed in 1985 (24) with early clinical studies done with subjects after stroke and spinal cord injury (SCI). (24,25) Since those initial studies, small clinical trials have been done to establish the feasibility and safety of the intervention. (26-29) More recently, da Cunha and colleagues (30) determined that locomotor training with BWS was a safe intervention to use with subjects in the acute phase of recovery. There is increasing evidence that treadmill training with BWS is superior to conventional approaches for locomotor training after stroke. (18,27) Gait training on the treadmill allows the manipulation of postural instability and balance through a weight-bearing progression while facilitating stepping. (31) It also eliminates the fear of falling, (32,33) making gait training possible with subjects who cannot safely be guarded during over-ground gait training (21) and allowing the intervention to be initiated earlier than conventional methods. (34) Treadmill training, with or without BWS, after stroke has consistently been shown to produce a more symmetrical gait pattern (35-37) while concurrent training of the contra-lateral lower extremity (LE) may improve gait quality, as some of the variation in gait speed after stroke has been attributed to the less affected LE. (14)

The most effective parameters to be used in the application of treadmill training with BWS in individuals poststroke have yet to be defined. Animal studies have indicated that early treadmill training may improve long-term neurological outcome after ischemic stroke, (38) and there is a general consensus that early therapeutic intervention is optimum to encourage motor recovery after stroke. (39-41) However, many of the studies with individuals poststroke have failed to address the timing and sequencing of the treadmill and over-ground components of the gait progression. The progression from treadmill to over-ground gait has been described with subjects with SCI. (42,43) In fact, Behrman and colleagues (43) described a detailed decision-making algorithm for such a progression in a subject with incomplete SCI. Studies involving persons poststroke have reported early application of locomotor treadmill training (6 weeks to 12 month postonset), (19,29,30) but no studies have reported delaying the initiation of over-ground gait until the basic components of normal gait have been established. Canning and colleagues (44) presented evidence that walking after stroke does eventually become automatic (i.e., well learned) and suggested that automatic control of gait should be a goal of gait retraining after stroke. Early treadmill training after stroke onset may be a method to promote automatic gait that is closer to normal gait.

Therefore, the purpose of this case study was to describe the application of locomotor treadmill training with partial BWS with a subject prior to the initiation of over-ground gait training.

Case Description

The subject of this case study was a 60-year-old male who had a stroke and did not receive medical care until 3 days later when he was admitted to the hospital and placed on the acute neurology service. Initial computed axial tomography studies revealed a prominent infarct involving the right middle cerebral artery. Magnetic resonance imaging revealed a moderately large infarct of the right frontal and temporal lobes, including the insular cortex and right basal ganglia. The subject had been completely independent prior to the stroke, living alone and working in an office setting. He had no risk factors for stroke except hyperlipidemia. At the time of admission, the subject was alert and oriented to name, place, and time. He stated his goal was to walk and to return to his previous independent lifestyle. While on the acute service, the subject was given an explanation of locomotor treadmill training with partial BWS and questioned about his interest in participating in this intervention as a part of his rehabilitation program. He expressed his interest and agreement with the plan. He had no contraindications to early treadmill training, such as recent myocardial infarction, uncontrolled hypertension, or LE degenerative disease. (30) His physician agreed with and authorized the treatment plan. As locomotor treadmill training with BWS is considered standard of care treatment, no institutional review board approval was sought for this case report.

While on the acute floor, physical therapy (PT) and occupational therapy (OT) were initiated. PT and OT focused primarily on passive/active-assistive range of motion and bedside mobility activities. Gait was not initiated on the acute service. The subject was transferred to the rehabilitation unit on day 4 after hospital admission where a typical rehabilitation schedule (i.e., 3 hours of therapy per day) was initiated.

Examination

Upon admission to the rehabilitation unit, an initial physical therapy evaluation was done, with results of tests and measures outlined in Table 1. All assessments had proven validity and reliability for persons poststroke. (45) All assessments, initial and discharge, were done by one of the authors (K.M.).

Intervention

Treadmill training

The subject was admitted to the rehabilitation unit on a Friday, and the initial therapy assessments were completed on that day. During the next 2 days, the subject had once-daily therapy consisting of beside activities, including supine and sitting activities. Gait was not initiated on the weekend. The following Monday, day 4 on the rehabilitation unit, a full therapy schedule was started. In the morning, the subject participated in basic mobility activities, including standing balance, transfers, standing weight shifts, and weight-bearing activities with the upper and lower extremities. In the afternoon of day 4, gait training was initiated on the treadmill with partial BWS. The system used was the single Pneu-Weight (Pneumex, Inc., 3115 North Boyer Ave., Sandpoint, ID 83864). The unit was comprised of an unweighting device that was positioned over a treadmill and powered by a small pneumatic compressor. A specially designed suspension vest was applied to the subject in supine, and he was transferred to the treadmill via wheelchair. Once on the treadmill, the vest was secured to the Pneu-Weight via a cable connected to the crossbar of the frame, and a percentage of the subject's body weight was removed by adjusting dials located on the frame of the Pneu-Weight. The initial treadmill training effort consisted of a total of 3.5 minutes of gait at a speed of 0.5 mph with approximately 65 pounds of BWS, which represented 40% of the subject's weight of 165 pounds. The subject required moderate-to-maximum assistance for left LE advancement and minimum assistance for weight shift over his affected leg with stance phase. He was allowed to hold onto the treadmill bar with his right upper extremity (UE) and his left hand was secured to the bar, as he was not able to hold it independently. As was the case for all treadmill sessions except the last session, two therapists assisted the subject during treadmill training. One therapist was seated next to the treadmill on a low stool and assisted with advancement of the left LE and the second was positioned behind the subject to assist with weight shift. The goal of the training was to replicate as closely as possible normal gait parameters, including heel contact at initial contact, followed by plantarflexion and knee flexion. In addition, every effort was made to prevent knee hyperextension at midstance and to promote a trailing limb at terminal stance.

The complete treadmill data, as well as overground data, are outlined in Table 2. BWS was initially set at 40%, based on previous work suggesting that to be an appropriate staring point. (19) BWS was reduced approximately 6% after the first two training sessions. A reduction of 3% was made for session 5, while 6% reductions were made for sessions 6 and 7. Reductions of approximately 9% were made for sessions 8 and 9, with full weight bearing achieved by session 9. The reduction schedule was intended to maximize muscle activation, as previous studies have suggested that excessive or prolonged BWS results in less activation and stimulation of key muscles, specifically the quadriceps and plantarflexors. (46) The amount of time on the treadmill was increased from 3.5 minutes for the first session to a maximum of 12 minutes for the sixth session. The final treadmill session lasted 6 minutes. The speed of the treadmill increased from a starting value of 0.5 mph to a final speed of 0.8 mph.

Over-ground gait training

Over-ground gait did not commence until rehabilitation day 8. At that time, the subject had completed five treadmill training sessions, totaling 40 minutes of gait. Over-ground gait began when the subject was able to advance his left LE with minimal assistance and consistently activate his quadriceps with loading on the left LE during treadmill training. The decision to initiate over-ground gait at this point in the training was based not only on the muscle activity in the subject's LE, but also on the fact that he had experienced multiple gait cycles with symmetrical step length and stance time during the treadmill training. Gait over ground was initiated with the use of a grocery cart, which encouraged symmetry by allowing bilateral UE support. An OT assisted with the gait by facilitating left hand/arm control on the handle of the grocery cart, as well as cueing the subject for proper trunk alignment. The PT provided minimal cues to hip and pelvis during swing and stance phases on the left. A cue for a posterior pelvic tilt at midstance on the left effectively controlled the tendency for the knee to hyperextend in midstance, secondary to weakness in the plantarflexors. (47) An ace wrap was used to hold the subject's ankle in neutral dorsiflexion until the custom ankle-foot orthosis (AFO) was delivered. The decision to fit the subject with an AFO was made during the second week of the subject's rehabilitation stay and was based on the subject's in ability to generate adequate strength to control his ankle and knee during gait. The orthosis of choice for this subject was an AFO with a double adjustable ankle joint, with metal uprights and plastic calf and foot components. The orthosis was ready for delivery on the day of his discharge.

Gait training over ground was conducted with and without assistive devices. After initial over-ground training with the grocery cart, no assistive device was used on day 4 of over-ground training. The rationale for attempting gait without an assistive device was to encourage equal LE weight bearing and equal stance time. On day 14 of his rehabilitation stay, it was determined in the family conference that the subject was going to go to South Carolina after discharge to stay with his daughter, since he was not ready to return to independent living. Gait training was instituted with a single-tip cane (STC) on rehabilitation day 21 in preparation for discharge. Some authors have suggested that there is not a natural sequence of progression from parallel bars to quad cane to STC as is commonly employed in the clinical setting. (48) Instead, these authors suggested that the movement patterns required for use of the various devices are quite different. Changing devices may prolong the progression to independence and gait training may be more efficient if it is done only with the device that is anticipated to be used at home, as was done with this subject. The subject discharged from the rehabilitation unit 25 days after admission using a STC with contact guard to standby assistance of his daughter, He continued with outpatient therapy in South Carolina.

Results

At the time of discharge from inpatient rehabilitation, the initial assessments were repeated and are detailed in Table 1. Muscle strength improved primarily in the larger, more proximal muscle groups of the lower extremity. Sensation and reflexes were unchanged at the final assessment. The greatest changes were observed in FIM[TM] * and balance scores.

Gait speed at time of discharge, on level surfaces, was 0.94 m/s (185 ft/min), which represents 69% of comfortable gait speed for a man his age, (49) and gait tolerance was 1,000 ft (304.8 m) with an STC and AFO. The subject's gait was analyzed using the observational gait analysis system. (47) (Although this system does not have published data concerning its reliability and validity, (50) observational methods are known to vary based on the skill of the rater. (50) In this case report, the rater [K.M.] was very experienced in the use of the system and also videotaped the subject throughout his stay so gait observations could be further reviewed for accuracy.) He demonstrated good step length and stance time symmetry. He demonstrated an occasional past retract (a visible backward movement of the thigh during terminal swing) on the left, infrequent foot drag on the left with swing, occasional left knee hyperextension from mid to terminal stance on the left, and occasional slight left lateral trunk lean with stance on the left. He also frequently contacted the ground with a flat foot and increased knee flexion.

Discussion

Locomotor treadmill training with partial BWS has been demonstrated to be an effective intervention to apply in the acute phase of recovery after stroke. (19,29,30) However, no published studies to date have reported delaying over-ground gait practice until the subject has the opportunity to experience a normal gait pattern using the treadmill and partial BWS. This report presented such a case.

At the time of discharge, based on observational gait analysis, the subject walked with a consistent step-through pattern, as well as comparable step length and stance time bilaterally. His score on the Berg Balance Scale improved from 15/56 to 44/56. He also walked at a speed of 0.94 m/s for a total of 304.8 m (1,000 ft) with a STC with contact guard to standby assistance.

There are many variables concerning the application of locomotor treadmill training with BWS after stroke that have not been optimally defined. The introduction and integration of the over-ground gait training component has received little attention in the studies to date. Studies have reported increased over-ground gait speeds after treadmill training with BWS, (18,19) as well as improved over-ground endurance compared to treadmill gait without BWS, (19,30) but the reports have provided little information to guide clinicians in the timing of over-ground walking. In this case report, the subject began over-ground gait training on day 8 of inpatient rehabilitation, only after the initial gait training had been done on the treadmill. During this initial skill acquisition period, the subject was placed in an environment designed to eliminate the features of gait after stroke that result in asymmetry. For example, the treadmill provided the opportunity for a trailing limb on the affected side, while concurrently training balance without the fear of falling. It is interesting that the subject remarked after the first over-ground gait training session that he "knew what to do" during that first attempt.

In this case report, treadmill training began very early in the subject's rehabilitation. It has been demonstrated that earlier gait recovery after stroke is associated with future gait independence (51) and that task-specific interventions that are applied early and intensively may be the most effective. (52) To date, the largest published report of application of locomotor treadmill training with and without BWS with subjects with acute stroke applied the intervention from day 27 to day 148 after onset of stroke, with a mean of 68.2 days for the BWS group and 78.4 days for the no BWS group. (19) The subjects were classified at the start of the trial as low or high level ambulatory status, so gait over ground had begun before the initiation of treadmill training. Yet these same authors conclude that treadmill training with partial BWS should have a "substantial impact" when it is applied during the acute phase of recovery. However, there are various definitions of the term acute as it relates to recovery after stroke. Some authors have defined the acute phase as being less than 6 weeks after stroke, (53) while others have concluded that the acute phase continues for up to 1 year. (54) In this case report, the subject began gait training on the treadmill on day 10 after stroke onset, clearly in the acute phase of recovery.

There were undoubtedly factors other than the application of locomotor treadmill training with partial BWS that influenced this subject's outcome, including traditional therapeutic interventions (balance, mobility, transfer training), the subject's motivation, premorbid lifestyle and health status, and family support. Spontaneous recovery was also a probable contributor to his outcome. However, given the size of his stroke and the involvement of the basal ganglia, (55) spontaneous recovery to the degree that he experienced in the 25 days that he was on the rehabilitation unit would seem unlikely. The severity of his sensory loss would also suggest that his gait recovery was better than what might have been anticipated. Perry and colleagues (56) have suggested that intact proprioception is a basic requirement for independent gait. Despite impaired proprioception, the subject progressed to gait on level surfaces with contact guard to standby assistance during his brief rehabilitation stay. In addition, it is known that AFOs of different types can have various effects on gait parameters (10); although the benefits of the adjustable AFO have been identified, (57,58) the degree to which the AFO impacted the subject's outcome is not known. Likewise, it is not known what impact the specific training parameters had on the subject's outcome. For example, the fastest speed the subject walked on the treadmill was 0.8 mph. There is good evidence in the literature suggesting that faster speeds on the treadmill might produce even better gait outcomes in subjects after stroke. (59,60)

Clearly, there is much that remains to be determined concerning the most effective way to apply locomotor treadmill training with BWS in the rehabilitation of subjects after stroke. Recent studies have begun to explore some of the variables in an effort to define optimum training parameters. For example, Chen and colleagues (61) examined speed, support stiffness, and handrail use. Dobkin (62) has summarized the evidence on locomotor treadmill training with BWS and identified several of the parameters that need clarification, including the timing and length of the intervention, the amount of BWS used, the progression to over-ground gait, the impact of upper extremity support, and the speed of the treadmill. This case report provides additional information concerning the potential impact of the intervention when applied in the very early days of recovery, when motor learning potential may be the greatest. Many persons in the early days after stroke onset do not possess the necessary strength and motor control to reproduce normal gait over ground initially, (5,63) as this subject did not. If the application of locomotor treadmill training with partial BWS begins before over-ground gait training has been initiated, it may provide the opportunity for subjects with stroke to practice initial gait efforts with a pattern that is close to normal gait. As a result, the gait learned from the beginning may not be marked by asymmetry as well as by other commonly seen gait deviation such as knee hyperextension in stance, (3) poorly controlled pelvic and trunk motion, (64) and decreased step length on the less affected extremity. (2) Buzzelli and colleagues (17) provide compelling evidence that extensive therapy efforts 1 year poststroke may not be able to produce changes in gait, despite patients' expectations of improvement. Given the evidence of asymmetry in gait after stroke and the apparent difficulty in changing that asymmetry, (17,65) identification of an intervention that has potential to rehabilitate gait without the asymmetries would represent significant progress in the rehabilitation of individuals after stroke.

As a case report, there are inherent limitations that must be considered. Obviously, there was only one subject, with no control for comparison, and there was not a period of no treatment. In addition, it is not possible to assess the influence of the training parameters that were selected. As noted earlier, observational gait analysis was used to assess the subject's gait quality, but the rater was experienced and had the advantage of videotape for further review. Likewise, with one subject, the results cannot be generalized to the stroke population at large.

Conclusion

This case report describes a novel application of locomotor treadmill training with partial BWS in a subject in the acute phase of recovery after stroke. The subject described in this case report experienced a good outcome after 25 days of inpatient rehabilitation, using a combination of early gait training on the treadmill, over-ground gait training, and balance and mobility training.

Further investigation of this novel application of locomotor treadmill training in a larger cohort of subjects may be warranted.