ONE OF THE MAJOR disabilities after stroke is the inability

Transcription

1 1458 Optimal Outcomes Obtained With Body-Weight Support Combined With Treadmill Training in Stroke Subjects Hugues Barbeau, PhD, Martha Visintin, MSc ABSTRACT. Barbeau H, Visintin M. Optimal outcomes obtained with body-weight support combined with treadmill training in stroke subjects. Arch Phys Med Rehabil 2003;84: Objectives: To identify stroke patients who are most likely to benefit from locomotor training with body-weight support (BWS), to determine the extent of carryover from treadmill training to overground locomotion, and to determine the variables that are most likely to influence the recovery of locomotion. Design: A randomized clinical trial. Setting: Inpatient rehabilitation hospital. Participants: Of 100 stroke subjects, 50 were randomized to receive locomotor training with BWS (BWS group), and 50 were randomized to receive locomotor training with full weight bearing (no-bws group). The subjects were stratified according to their initial overground walking speed and endurance, initial treadmill speed and endurance, functional balance, motor recovery, side of the lesion, and age. Intervention: Fifty subjects were trained to walk on a treadmill with up to 40% of their body weight supported by a BWS system with an overhead harness (BWS group), and 50 subjects were trained to walk while bearing their full weight (no-bws group) Main Outcome Measures: Clinical outcome measures included overground walking speed and endurance, functional balance, and motor recovery. The effect of confounding variables such as age, comorbidity, and depression on locomotor outcome was also investigated. Results: After 6 weeks of locomotor training, the BWS group scored significantly higher in all clinical outcomes. When the subjects were stratified according to their initial overground walking speed, endurance, balance, and motor recovery, a significant statistical difference in gait and balance dysfunction of all outcomes occurred in the more severely impaired subjects. An important transfer from treadmill speed to overground walking speed was observed in subjects in the BWS group. Finally, a significantly greater effect was observed in older subjects (65 85y) in the BWS group. Conclusions: Retraining gait in severely impaired stroke subjects with a percentage of their body weight supported resulted in better walking and postural abilities than did gait training in patients bearing their full weight. It appears that subjects with greater gait impairments benefited the most from From the School of Physical and Occupational Therapy, McGill University, Montreal, QC; and Research Center, Jewish Rehabilitation Hospital, Laval, QC, Canada. Supported by the Heart & Stroke Foundation of Canada, the National Health Research & Development Program, and the JRH Foundation. No commercial party having a direct financial interest in the results of the research supporting this article has or will confer a benefit upon the author(s) or upon any organization with which the author(s) is/are associated. Reprint requests to Hugues Barbeau, PT, PhD, Sch of Physical and Occupational Therapy, McGill University, 3630 Promenade-Sir-William-Osler, Montreal, QC H3G 1Y5, Canada, hugues.barbeau@mcgill.ca /03/ $30.00/0 doi: /s (03) training with BWS, as did the older patients with stroke. There is evidence of transfer from treadmill training to overground locomotion. Key Words: Cerebrovascular accident; Hemiplegia; Patient care management; Rehabilitation; Treatment outcome; Walking by the American Congress of Rehabilitation Medicine and the American Academy of Physical Medicine and Rehabilitation ONE OF THE MAJOR disabilities after stroke is the inability to walk. More than 50% of patients with stroke who survive the acute phase are unable to walk 1-3 and require a period of rehabilitation to achieve a functional level of ambulation. Research with both animals and humans has shown that the strategy adopted to retrain walking in patients with neurologic injury can significantly influence the degree of locomotor recovery. 4-6 A newly developed gait-training strategy for neurologic patients uses a harness system that supports a percentage of a patient s body weight, thereby unloading the lower extremities while the patient is being trained to walk on a treadmill Studies suggest that retraining gait with bodyweight support (BWS) leads to a more successful recovery of ambulation with respect to overground walking speed and endurance, functional balance, and lower-limb motor recovery. It also reduces the amount of physical assistance required to walk. 5,8,12,13 The strategy of retraining gait with BWS during treadmill walking was investigated in a large randomized clinical trial (RCT) involving 100 subjects with stroke. 14 The experimental group that was trained with BWS (BWS group) scored significantly higher in overground walking speed and endurance, functional balance, and lower-limb motor recovery than did a control group that trained while bearing full weight on the lower extremities (no-bws group). During the trial, the experimental group was provided with up to 40% BWS, and the amount of weight supported was progressively decreased as gait patterns improved. After a 6-week training period, 80% of the subjects in the BWS group could train at full weight bearing, which implies that such training is feasible in a clinical setting. Follow-up evaluation at 3 months showed that those trained with BWS continued to have significantly higher scores for overground walking speed and lower-limb motor recovery. It can be concluded that retraining gait in patients with stroke by supporting a percentage of their body weight resulted in their having better walking abilities than patients who trained while bearing their full weight. This article presents a further analysis of the data collected on the 100 subjects in the RCT described above. 14 The objectives were (1) to identify subjects who would most likely benefit from locomotor training with or without BWS, as determined by clinical outcome measures such as treadmill and overground walking speed and endurance, functional balance, and motor recovery; (2) to determine the extent to which locomotor training on the treadmill with and without BWS could be transferred to overground walking; and (3) to deter-

2 BODY-WEIGHT SUPPORT WITH TREADMILL TRAINING, Barbeau 1459 mine the variables that are most likely to influence the recovery of locomotion. METHODS Participants A total of 375 patient admissions to the Jewish Rehabilitation Hospital, Laval, QC, between October 1992 and January 1995 for physical rehabilitation after stroke, were reviewed (fig 1). The average age of the group was 69.2 years (range, 27 93y); 45.6% of the admissions were women. Two hundred thirty-seven patients did not meet the study s inclusion criteria for the following reasons: walked with a normal gait pattern (n 73); had severe cardiac problems (n 39); treadmill training was contraindicated because of comorbid conditions (n 29); had cerebellar, bilateral, or brainstem cerebral vascular disease (n 28); were unable to understand simple commands because of language, cognitive, behavioral, or psychiatric disorders (n 19); anticipated length of stay was less than 4 weeks (n 15); onset of cerebrovascular accident was more than 6 months ago (n 10); were readmitted during the study period (n 9); did not ambulate before their stroke (n 4); and were excluded for other reasons (n 11). Fourteen additional subjects were not recruited because the treadmill was overbooked (n 6) and because at one point high functional walkers were not sought (n 8). One hundred twenty-four subjects with Fig 1. Clinical trial profile of this study. right or left cortical stroke were eligible; 24 refused to participate, whereas 100 subjects provided informed consent to participate in the study, which was approved by the hospital s ethics committee. Patients who refused to participate were slightly older (mean, y) than those who did (mean, y). Experimental and Control Groups The 100 subjects were randomized into the experimental group (BWS, n 50) and the control group (no-bws, n 50) by block randomization within strata identified according to the initial level of ambulatory status (low, high). Low ambulatory status was defined as being nonambulatory or requiring maximal assistance to walk. High ambulatory status was defined as needing moderate or minimal assistance or walking independently with or without supervision, but with residual gait deviations. The experimental group trained on a treadmill while a percentage of their body weight was supported by an overhead harness. The control group trained on a treadmill but without BWS. The BWS system and the overhead harness have been described in detail Briefly, the harness consists of a pelvic belt that attaches around the hips and 2 thigh straps with anterior and posterior attachments to the pelvic band. The harness vertically supports the patient over the treadmill and is attached to a suspension system with a force transducer that signals the amount of body weight being supported. Individuals in the BWS group were provided up to 40% BWS at the beginning of training, and that percentage was progressively decreased as subject s gait pattern and ability to walk improved. Subjects in the control group wore the harness for security and to ensure similar experimental conditions, but no BWS was provided. Both groups trained 4 times a week for 6 weeks under the supervision of a physical therapist. During each session, patients walked a maximum of 3 trials for no more than 20 minutes. Subjects pulse and heart rates were monitored before each session began and again after each trial to ensure that they did not exceed a baseline that was established by their physician. Walking on the treadmill (Burdick T500 model a ) was initiated at 0.0km/h and increased by increments of.15km/h. Subjects could grip a horizontal bar attached to the front of the treadmill to provide stability. In addition to gait training, all subjects regardless of their group allocation received regular, weekday physical therapy. Training Strategy The training strategy focuses on a straight trunk and limb alignment with proper weight shift and weight bearing onto the hemiplegic limb during the loading phases of gait, as well as stepping to advance the limb. At the start of training, subjects walked at 10%, 20%, 30%, and 40% of BWS. The therapist then selected the percentage that facilitated proper trunk and limb alignment and transfer of weight onto the hemiplegic limb. Subjects in both groups trained with the assistance of 1 or 2 therapists, as needed. The more impaired subjects were assisted by 2 therapists. One stood behind the subject while straddling the treadmill to provide assistance for proper trunk alignment and weight shifting as the subject walked. The second therapist stood beside the hemiplegic lower limb and helped with stepping and limb control during the stance and swing phases. Other variables were manipulated during training, including treadmill speed and use of the horizontal bar to increase stability. When the subject could walk with an upright posture and shift weight from limb to limb during loading, the treadmill

3 1460 BODY-WEIGHT SUPPORT WITH TREADMILL TRAINING, Barbeau Table 1: Baseline Demographic Characteristics and Pretraining Scores on Clinical Outcome Measures for the BWS and No-BWS Groups Variable BWS Group (n 50) No-BWS Group (n 50) Age (y) (27 87) (44 84) Sex: F/M (%) 19/31 (38%/62%) 22/28 (44%/56%) Side of lesion: R/L (%) 20/30 (40%/60%) 29/21 (58%/42%) Total comorbidity (1 6) (1 7) Depression (Zung scale) (range, ) ( ) (25 75) Cognitive status (Pfeiffer scale) (range, 0 10) (2 10) (3 10) Delay onset of stroke to study (d) (27 138) (33 148) Balance (Berg scale) (range, 0 56) (3 55) (3 54) Motor recovery (STREAM) (range, 0 55) (5 51) (3 51) Overground walking speed (m/s) (range, ) ( ) ( ) Overground walking endurance (m) (range, 0 320) (2 320) (0 320) NOTE. Values are mean SD (range), unless otherwise indicated. Abbreviations: F, female; L, left; M, male; R, right; SD, standard deviation. speed was increased. In the first sessions after the speed was increased, it was sometimes necessary to augment BWS to facilitate walking at the higher speed. Once the subject was accustomed to that speed, the percentage of BWS was again decreased. For subjects in the no-bws group, treadmill speed was also increased as their gait improved and they were able to walk faster. Subjects in both groups who progressed to walking with good trunk alignment and stepping with good weight shift from side to side were trained to walk without using the treadmill horizontal support bar, to challenge their balance and postural responses. For subjects in the BWS group, support was initially increased to facilitate walking without holding onto the bar and was decreased as they were able to accomplish this with more ease. Measurement Tools Two types of variables were measured: outcome variables, with which the effectiveness of the BWS system was judged, and confounding variables that are associated with recovery of ambulation and function. All subjects were evaluated before training commencement and again at completion of the 6-week program and at a 3-month follow-up. All evaluations were performed by an evaluator who was blinded to group assignments. Outcome variables. The 2 groups were compared on balance, motor recovery, overground walking speed, and over- Fig 2. Pretraining, posttraining, and follow-up overground walking speed in the BWS and no-bws groups for subjects with an initial overground walking speed (A) between 0.0 and 0.2m/s and (B) 0.2m/s and more. *P<.01. Fig 3. Pretraining, posttraining, and follow-up overground walking endurance in the BWS and no-bws groups for subjects with an initial overground walking endurance (A) between 0 and 20m and (B) 20m and more. *P<.05.

4 BODY-WEIGHT SUPPORT WITH TREADMILL TRAINING, Barbeau 1461 Cognitive status was measured with the 10-item Pfeiffer Short Portable Mental Status Questionnaire. 21 The score was calculated out of a possible 10, with higher scores indicating better functioning. Cognitive scores were not available for subjects who had communication difficulties that were associated with aphasia. Mood was assessed with the short 10-item version of the Zung Self-Rating Depression Scale. 22 Scores range from 25 to 100, with scores more than 50 indicating depression. Statistical Analyses Descriptive statistics were used to compare the baseline characteristics and the pretraining gait scores of the 2 groups. A 2-way analysis of covariance (ANCOVA) with a repeated measure on 1 factor was used to determine differences in the 6 clinical outcome measures across the 2 groups (treadmill and overground walking speed, treadmill and overground endurance, balance, and motor recovery). The independent variable was group (BWS, non-bws), the second factor was time (after the program, follow-up), and, finally, the covariate was the pretraining score. Fig 4. Pretraining, posttraining, and follow-up balance scores in the BWS and no-bws groups for subjects with an initial balance score (A) between 0 and 15 and (B) between 15 and 56. *P<.05. ground walking endurance. Balance was assessed with the Berg Balance Scale, which evaluates 14 sitting and standing activities, each on a 5-point scale. 17 The maximum score is 56, with higher scores indicating better balance. The scale has been tested on patients with stroke and has good inter- and intrarater reliability (.98,.99, respectively). 18 Motor recovery was assessed by using the lower-extremity portion of an early version of the STREAM (STroke REhabilitation Assessment of Movement), a 25-item scale evaluated on a 4-point scale for some items and on a 2-point scale for other items. 19 More specifically, the STREAM evaluates voluntary movement of the limbs and basic mobility. The maximum score is 55, with high scores signaling better function. Overground walking speed was measured in meters per second as the subject walked across a 10-m walkway using the walking aids he/she required. Walking speed was recorded over the middle 3m of the walkway with a stopwatch. Comfortable treadmill walking speed was measured by covering the treadmill speed indicator and slowly increasing the speed until the subject indicated that a comfortable walking speed had been reached. When the subject had sufficient endurance, he/she was requested to complete the 10-m walk 3 times, and the average of the 3 trials was recorded as the speed. Overground endurance was measured by having the subject walk back and forth along the 10-m walkway until he/she was unable to continue or to a maximum distance of 320m. Treadmill walking endurance was measured by recording the total time the subject walked on the treadmill during a session up to a maximum of 20 minutes. When overground walking speed and endurance were measured, the subjects used the walking aids, if required, and were given the assistance necessary to compensate for lack of balance. Confounding variables. Information on age, gender, side of lesion, time since stroke, previous strokes, and other comorbidities, with the weighted classification scheme developed by Charlson et al, 20 was abstracted from the medical dossier. RESULTS Of the 100 subjects, 79 completed the entire study protocol (they completed all 24 training sessions). Forty-three (86%) subjects were in the BWS group, and 36 (72%) were in the no-bws group. An analysis of these patients profiles emerged when we compared subjects who completed the training with those who did not. The patients who dropped out were the elderly women with multiple comorbid conditions, a fact that was not surprising. Therapists would be hesitant to place an elderly, more sickly patient in such an apparatus. The subjects did not differ with respect to disability, as measured by balance, motor recovery, walking speed, and walking endurance. Fig 5. Pretraining, posttraining, and follow-up motor recovery scores in the BWS and no-bws groups for subjects with an initial motor recovery score (A) between 0 and 20 and (B) between 20 and 55. *P<.05.

5 1462 BODY-WEIGHT SUPPORT WITH TREADMILL TRAINING, Barbeau Variable Table 2: Statistical Analysis Using ANCOVA on Clinical Outcomes Group Time Interaction BWS/No-BWS (n) BWS/No-BWS (post/follow-up) (group/time) Overground walking speed (m/s) /23.008* / Treadmill walking speed (m/s) / NA NA / NA NA Overground walking endurance (m) / * / *.210 Treadmill walking endurance (m) / NA NA 6 17/ NA NA Balance score (max 56) / * / *.611 Motor recovery score (max 55) / * / *.951 Walking speed and side of CVA BWS 26/ FWB 16/ *.248 Walking speed (m/s), stratified by age 20 65y 18/ * y 25/ Walking endurance (m), stratified by age 20 65y 18/ * y 25/ Balance score (max 56), stratified by age 20 65y 17/ * y 25/17.003*.006*.485 Motor recovery (max 55), stratified by age 20 65y 18/ y 25/ NOTE. Analysis includes the group effect, the time effect, and group/time interaction. Abbreviations: CVA, cerebrovascular accident; FWB, full weight bearing; max, maximum; NA, not available. *P.01; P.001; P.02; The sample size is for left vs right side; P.05. Table 1 outlines the subjects characteristics and the pretraining scores on the primary gait parameters. The pretraining scores for all subjects who completed the training protocol were also similar (mean score standard deviation) for balance ( vs ), motor recovery ( vs ), overground walking speed (.18.16m/s vs.17.18m/s), and overground walking endurance ( m vs m). In addition, their pretraining scores were similar for depression ( vs ) and for cognitive status ( vs ). Comparison of Locomotor Training With and Without BWS When subjects were stratified according to the pretraining scores for overground speed (figs 2A, 2B), overground endurance (figs 3A, 3B), balance (figs 4A, 4B), and motor recovery score (figs 5A, 5B), a significant change could be observed for all outcome variables with time (table 2). However, an important group effect was also observed. Subjects with a pretraining walking speed less than.20m/s (fig 2A) showed changes in walking speed that were significantly more marked (P.005) than the changes seen in subjects whose pretraining overground walking speed was more than.20m/s (P.567) (fig 2B). A similar trend toward the more severely impaired subjects was also observed for overground endurance (figs 3A, 3B), balance (figs 4A, 4B), and motor recovery (figs 5A, 5B). However, the treadmill walking speed improved significantly in both groups over time (table 2), and no significant differences were seen between the 2 groups either after training or at follow-up. Changes in Treadmill and Overground Walking Speed Figure 6 shows that both pre- and posttraining walking speeds were greater on the treadmill than they were in overground walking (figs 6A, 6B) in both groups. Furthermore, the transfer of training from treadmill walking speed to overground walking speed was greater in the BWS group than in the no-bws group (fig 6C). Comparison of Locomotor Outcomes by Age and Side of Lesion Figure 7 shows the stratification of locomotor outcomes by age (20 65y and 65 85y). A clearly significant difference was observed between the groups for subjects in the older group (P.03, fig 7C) compared with the younger group (P.457, fig 7B).

6 BODY-WEIGHT SUPPORT WITH TREADMILL TRAINING, Barbeau 1463 DISCUSSION This analysis shows a significantly greater effect of locomotor training with BWS on more functionally impaired stroke subjects, as characterized by lower pretraining scores of overground walking speed and endurance, functional balance, and lower-limb motor recovery. Training effects were greater (but not significantly different) in the BWS group than in the no-bws group for subjects who were at a higher functional level of ambulation, as defined by higher pretraining scores for the clinical outcome measures. An important, although incomplete, transfer from treadmill to overground speed was observed in the BWS group and, to a much lesser extent, in the no-bws group. Finally, a significantly greater training effect in older patients was seen in the BWS group compared with the no-bws group. Effectiveness of Locomotor Training With BWS Both groups received task-specific gait training on the treadmill, with use of BWS being the only difference between the groups. Ultimately, the BWS group had significantly better gait outcome, which supports the hypothesis that partially unloading the lower limbs during training and progressively increasing the load as the gait pattern improves will enhance the recovery of posture and locomotion. Better walking abilities in the BWS group cannot be attributed to more gait-specific training because the 2 groups did not differ in terms of the amount of time spent in gait training. 14 Thus, the benefits of retraining gait with BWS appear to derive from its effects. Unloading the lower extremities appears to be an important factor in training balance while walking. The use of this strategy has also been supported by several studies with spinal cord injury (SCI) subjects 8-11,13,15,23,24 and stroke patients. 5,25 The mechanism underlying this effect is unclear, but recent animal studies strongly support the role of modulation of the extensor load receptors, probably arising from the Golgi tendon organ. 25 This is a newly discovered function of these receptors in the regulation of stance and walking. The benefit of locomotor training with BWS may depend partly on the degree of body unloading during walking, but this needs further investigation. A previous study 10 showed that BWS training facilitated forced use of the paretic limb; patients could not adopt a compensatory gait pattern by excessive use of the less affected limb or/and the upper extremity. This represents a very important component of this locomotor training strategy. Fig 6. Scatterplots showing the treadmill speed versus the overground speed during (A) pretraining, (B) posttraining, and (C) the change (posttraining minus pretraining speed). Squares versus stars on the diagonal line denote the difference between post- and pretraining speed. Statistical analysis showed that there was no effect related to side of lesion with respect to overground walking speed in the groups individually or combined (table 2). Fig 7. Overground walking speed stratified by age for subjects (A) between 20 and 65 years and (B) between 65 and 85 years. *P<.05.

7 1464 BODY-WEIGHT SUPPORT WITH TREADMILL TRAINING, Barbeau This analysis strongly suggests that a subgroup of stroke patients with major walking deficits showed a significantly greater improvement in overground walking speed and endurance, more motor recovery, and a greater ability to transfer from treadmill to overground walking after training with BWS. These subjects represent the target group that would most benefit from locomotor training with BWS. Kosac and Reding 26 also observed this in a subgroup of severely impaired stroke subjects. However, for stroke subjects with a walking speed more than 0.2m/s, the effects of training with BWS were greater in the magnitude of change in locomotor and balance outcome. As the stroke subjects become more functional, they need to be challenged both in terms of posture and walking to further enhance their recovery. Pohl et al 27 and Sullivan et al 28 showed that challenging walking speed for 2 weeks on the treadmill significantly improved overground walking speed. Transfer to Overground These results strongly suggest that improvements in walking achieved during supported locomotion can be sustained and transferred to overground walking in stroke patients with severe gait impairments. It has been suggested that the higher walking speed on the treadmill versus overground could be confounded by the use of a horizontal bar attached to the front of the treadmill for stability, but this hypothesis needs further investigation. We actually tested the hypothesis that locomotor recovery could also be achieved in stroke subjects by using a BWS system during overground gait training. Preliminary results have shown that this new approach is feasible. 29 Another major point of interest was the significantly greater effect of locomotor training with BWS in older stroke subjects. Danielsson and Sunnerhagen 30 reported that walking with 30% of BWS resulted in less oxygen consumption than full weightbearing walking in both stroke- and age-matched healthy subjects. Thus, the older stroke subjects (65 85y) could benefit from using locomotor training with BWS, which can be tolerated by subjects with comorbidities such as cardiovascular problems. Gait training with BWS is less demanding in terms of energy consumption, which may explain why stroke and SCI patients can begin their locomotor training with BWS on the treadmill very early after their injury. Clinical Relevance Subjects recruited for this study had significant gait disabilities, as profiled by recorded clinical measures of balance and mobility. In general, they presented attributes that are typical of subacute patients with stroke who are undergoing rehabilitation. In stroke rehabilitation, use of the treadmill is increasingly mentioned as an alternative method of gait training, although it has yet to be widely used in clinical settings. A relevant finding from this study is that 79% of the stroke subjects were able to complete the 6-week training regimen on the treadmill for both paradigms: BWS or no-bws. This suggests that treadmill gait training is well tolerated by patients with stroke. CONCLUSION This study shows that gait training on a treadmill with BWS is an effective approach because it results in better locomotor and postural abilities. This type of training is well tolerated by patients with stroke and is a training strategy that is compatible with rehabilitation practices in a clinical setting. 31 Indeed, considering that a patient s regular physical therapist can supervise the training (as was done in this study), the results can be generalized to other rehabilitation settings. Gait training with BWS could be used in combination with other rehabilitation strategies, such as functional electric stimulation (FES), to assist walking and pharmacologic approaches 32 that may enhance locomotor function in patients with neurologic conditions. Further research is needed to develop rehabilitation strategies that can further enhance recovery of both locomotion and postural abilities. It is important to investigate whether recovery of gait would be further enhanced during overground gait training with BWS. Preliminary results 29 show very positive effects. Identifying the optimal period after the lesion during which the initiation of this type of training maximizes gait and posture function is also important. Recently, this new locomotor strategy has been used with patients with different neurologic conditions, such as SCI, 11,13 Parkinson s disease, 33 and cerebral palsy, 34 as well as in the elderly and elderly stroke population. 30 Other neurologic populations, such as persons with head injury and multiple sclerosis, as well as the orthopedic population, such as amputees and persons with hip prostheses, may also benefit from such approaches. This novel training strategy appears to be effective in enhancing locomotor and postural recovery and provides a dynamic task-specific and forced-use approach for the treatment of gait dysfunction after stroke. This innovative approach, originally developed from animal studies, 35 could be combined with other rehabilitation strategies, such as pharmacologic approaches and FES, to enhance the recovery of posture and locomotion in the neurologic disease population. 4,32,35 Acknowledgments: Many thanks for the precious collaboration of the JRH clinicians and G. Chilingaryan for his statistical expertise. References 1. Clifford J. Managing disability from stroke. Can Fam Physician 1986;32: Gillum R, Gommez-Martin O, Kottke T, Jacobs D, Prineas R, Folsom A. Acute stroke in a metropolitan area, 1970 and J Chronic Dis 1985;38: Nicholls ES, Jung J, Davies JW. Cardiovascular disease mortality in Canada. Can Med Assoc J 1981;125: Barbeau H, Fung J. Recovery of locomotion following spinal cord injury: new concepts and approaches in rehabilitation. In: Good DC, Copuch JR, editors. Handbook of neurorehabilitation. New York: Marcel Dekker; p Hesse SA, Bertelt C, Jahnke MT, et al. Treadmill training with partial body weight support compared with physiotherapy in nonambulatory hemiparetic patients. Stroke 1995;26: Richards CL, Malouin F, Wood-Dauphinee S, Williams JI, Bouchard JP, Brunet D. Task-specific physical therapy for optimization of gait recovery in acute stroke patients. Arch Phys Med Rehabil 1993;74: Finch L, Barbeau H. Hemiplegic gait: new treatment strategies. Physiother Can 1985;38: Visintin M, Barbeau H, Finch L. Progressive weight bearing and treadmill stimulation during gait retraining of hemiplegics. A case study [abstract]. Phys Ther 1987;68: Visintin M, Barbeau H. The effects of body weight support on the locomotor pattern of spastic paretic patients. Can J Neurol Sci 1989;16: Visintin M, Barbeau H. The effects of parallel bars, body weight support and speed on the modulation of the locomotor pattern of spastic paretic gait: preliminary communication. Paraplegia 1994; 32: Barbeau H, Danakas M, Arsenault B. The effects of locomotor training in spinal cord injured subjects: a preliminary study. Restor Neurol Neurosci 1993;5: Hesse SA, Bertlet CM, Schaffrin A, Malezic M, Mauritz KH. Restoration of gait in non-ambulatory hemiparetic patients by treadmill training with partial body weight support. Arch Phys Med Rehabil 1994;75:

8 BODY-WEIGHT SUPPORT WITH TREADMILL TRAINING, Barbeau Wernig A, Müller S, Nanassy A, Cagol E. Short communication: Laufband therapy based on rules of spinal locomotion is effective in spinal cord injured persons. Eur J Neurosci 1995;7: Visintin M, Bitensky N, Mayo N, Barbeau H. A new approach to retrain gait in stroke patients through body weight support and treadmill training. Stroke 1998;20: Wainberg M, Barbeau H, Finch L. Description and application of a system for a locomotor rehabilitation. Med Biol Eng Comput 1987;25: Norman KE, Pépin A, Ladouceur M, Barbeau H. A treadmill apparatus and harness support for evaluation and rehabilitation of gait. Arch Phys Med Rehabil 1995;76: Berg K, Wood-Dauphinee S, William JI, Gayton D. Measuring balance in the elderly: preliminary development of an instrument. Physiother Can 1989;41: Berg K, Maki BE, Williams JI, Holliday J, Wood-Dauphinee SL. Clinical and laboratory measures of postural balance in an elderly population. Arch Phys Med Rehabil 1992;73: Daley K, Mayo N, Danys I, Cabot R, Wood-Dauphinee S. The Stroke Rehabilitation Assessment of Movement (STREAM): refining and validating the content. Physiother Can 1997;49: Charlson ME, Pompie P, Ales KL, MacKenzie CR. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis 1987;40: Pfeiffer E. A short portable mental status questionnaire for the assessment of organic brain deficit in elderly patients. J Am Geriatr Soc 1975;23: Tucker MA, Ogle SJ, Davison JG, Eilenberg MD. Validation of a brief screening test for depression in the elderly. Age Ageing 1987;16: Wernig A, Muller S. Laufband locomotion with body weight support improved walking in persons with severe spinal cord injuries. Paraplegia 1992;30: Behrman AL, Harkema SJ. Locomotor training after human spiral cord injury: a series of case studies. Phys Ther 2000;80: Dietz V, Duysens J. Significance of load receptor input during locomotion: a review. Gait Posture 1999;11: Kosac MC, Reding MJ. Comparison of partial body weight supported treadmill gait training versus aggressive bracing assisted walking post stroke. Neurorehabil Neural Repair 2000;14: Pohl M, Mehrholz J, Ritschel C, Rückriem S. Speed-dependent treadmill training in ambulatory hemiparetic stroke patients. A randomized controlled trial. Stroke 2002;33: Sullivan KJ, Knowlton BJ, Dobkin BH. Step training with body weight support: effect of treadmill speed and practice paradigms on post-stroke locomotor recovery. Arch Phys Med Rehabil 2002; 83: Fung J, Barbeau H, Roopchand S. Partial weight support improves force generation and postural alignment during overground locomotion following stroke [abstract]. Neurosci Abstr 1999;25: Danielsson A, Sunnerhagen KS. Oxygen consumption during treadmill walking with and without body weight support in patients with hemiparesis after stroke and in healthy subjects. Arch Phys Med Rehabil 2000;81: Barbeau H. Locomotor training in neurorehabilitation: emerging rehabilitation concepts. Neurorehabil Neural Repair 2003;17: Barbeau H, Norman K, Fung J, Visintin M, Ladouceur M. Does neurorehabilitation play a role in the recovery of walking in neurological populations? Ann N Y Acad Sci 1998;860: Miyai I, Fujimoto Y, Veda Y, et al. Treadmill training with body weight support: its effect on Parkinson s disease. Arch Phys Med Rehabil 2000;81: Schindl MR, Forstner C, Kern H, Hesse S. Treadmill training with partial body weight support in nonambulatory patients with cerebral palsy. Arch Phys Med Rehabil 2000;81: Rossignol S. Locomotion and its recovery after spiral injury. Curr Opin Neurobiol 2000;10: Supplier a. Spacelabs Medical, Div of Instrumentarium, PO Box 7018, Issaquah, WA