2.1

Patients

The study included patients with hemiparesis fulfilling the following:

• •
  

  first clinical episode of stroke due to an ischemic or hemorrhagic supratentorial lesion;

  
  
• •
  

  right or left hemiparesis, age 16 to 65 years, duration of hemiparesis > 3 months;

  
  
• •
  

  impaired ability to walk independently or need for one person to help with balance and coordination (Functional Ambulation Category [FAC] II–IV) .

We excluded patients with recurrent stroke, receptive aphasia with inadequate comprehension to understand and follow the training schedule, significant cognitive deficits affecting participation (Mini Mental Status Examination score ≤ 23) , significant depression interfering with active participation (score < 12 on the 24-item Hamilton Rating Scale for Depression) , movement disorders interfering with training, recent myocardial infarction (< 6 months), ischemia or angina at rest or during exercise, and orthopedic conditions such as severe arthritis and total hip or knee joint replacements.

2.2

Design

The study was a randomized, controlled, single-blind design. The sample size was computed before the study based on the primary outcome measures and gait parameters required to be recorded. Included patients were randomly allocated to 1 of 3 groups by a table of random numbers. Group I received overground task-oriented gait training by a conventional physiotherapy approach. Group II received gait training on a treadmill without bodyweight support (full weight bearing); patients wore an overhead harness for security and to ensure similar experimental conditions, but no mechanical support was provided. Group III received gait training on a treadmill with bodyweight support (40% unweighting of body weight), which remained unchanged throughout all training sessions. Horizontal movement was provided by a slow moving treadmill, the speed of which was gradually increased on the basis of the patient’s comfort with the speed in both groups II and III. The gait training was provided by the therapists giving routine physiotherapy. Patients in groups II and III were blinded by ensuring similar experimental conditions (harness ± unweighting), but patients in group I were not blinded because they received overground task-oriented gait training. The physical therapist who assessed the gait parameters before and after the gait-training protocol was blinded to patient group for patients in groups II and III.

All patients received simultaneous comprehensive rehabilitation input in the form of medical management, routine physiotherapy, occupational therapy, orthotics, speech therapy and cognitive retraining, as needed. They all received basic physiotherapy in the form of stretching and strengthening exercises, spasticity-reducing exercises, balance and weight-shifting activities and postural control activities for 40 min/day.

Gait training was given for 30 min/day in all 3 groups. For group I, receiving conventional physiotherapy, overground task-oriented gait training involved walking in front of a mirror and walking along a straight line to give visual feedback for reducing gait deviations, walking a minimum distance during each session and gradually being encouraged to walk with increased speed. Patients trained on a treadmill were allowed to walk for a maximum of 5 trials during each session, each 6 min long, with intermittent rest for a total duration of at least 30 min.

Patients were evaluated before the start of training (baseline or pre-training) and at the completion of 20 sessions of training, 5 days/week for 4 weeks (post-training). Follow-up evaluation was at 3 months after completion of the training. Patients who did not return for the scheduled follow-up were contacted by mail and telephone. A blinded evaluator performed all evaluations and was not aware of group assignment for groups II and III.

2.3

BWSTT (Biodex Gait Trainer™)

The BWSTT system used was the Biodex Gait Trainer™, which is designed specifically for rehabilitation and retraining of gait for patients with neurologic and orthopaedic gait dysfunctions. The treadmill permits walking to be initiated from 0.0 km/h and increased by increments of 0.16 km/h. The patient can also hold onto a horizontal bar attached to the front of the treadmill for stability. The bodyweight support system includes an overhead harness with a pelvic band 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 supported.

2.4

Measurement tools

To measure the effectiveness of the gait training, the primary outcome variables were the quantitative clinical gait parameters of overground walking speed and overground walking endurance. Overground walking speed was measured in meters per second as the patient walked across a 10 m walkway. Patients were allowed to use the walking aids they required, and necessary assistance to compensate for the loss of balance was provided. Walking speed is an objective yet inexpensive method to monitor gait rehabilitation . Overground walking endurance was measured in meters by asking the patients to walk back and forth along the 10 m walkway until they were unable to continue. Patients were permitted to continue up to a maximum distance of 320 m .

Secondary outcome variables were the Scandinavian Stroke Scale (SSS) and FAC. Impairment after stroke was assessed by the SSS, which consists of 8 variables ranging from consciousness, cognitive deficits, motor power, and walking ability graded from 0 to 12. The maximum score is 58 and higher scores indicate better outcome . The FAC distinguishes 6 levels of required support during gait without taking into consideration any aid used. This scale has been validated for classifying walking handicap after stroke . Patients perform the test with a cane but without orthosis. Levels are defined as:

• •
  

  0: the subject cannot walk at all or requires the help of 2 or more people;

  
  
• •
  

  1: the subject needs continuous support from one person who helps carry the subject’s weight and helps with balance;

  
  
• •
  

  2: the subject is dependent on the continuous or intermittent support of one person to help with balance and coordination;

  
  
• •
  

  3: the subject needs only verbal supervision;

  
  
• •
  

  4: the subject requires help for stairs and uneven surfaces;

  
  
• •
  

  5: the subject can walk independently anywhere.

2.5

Statistical analysis

We compared the 3 groups on baseline outcome measures by ANOVA followed by post-hoc test for continuous variables and Kruskal-Wallis and chi-square tests for ordinal and nominal variables. Within each of the 3 groups, we analyzed change in each outcome variable by paired t test for continuous variables and Wilcoxon Rank test for ordinal variables (pre-training vs post-training and pre-training vs 3-month follow-up scores). Then we compared the 3 groups on outcome variables at the completion of training (post-training) and follow-up (3 months) by ANOVA followed by post-hoc test for continuous variables and Kruskal-Wallis test for ordinal variables. P < 0.05 was considered statistically significant.

2.1

Patients

The study included patients with hemiparesis fulfilling the following:

• •
  

  first clinical episode of stroke due to an ischemic or hemorrhagic supratentorial lesion;

  
  
• •
  

  right or left hemiparesis, age 16 to 65 years, duration of hemiparesis > 3 months;

  
  
• •
  

  impaired ability to walk independently or need for one person to help with balance and coordination (Functional Ambulation Category [FAC] II–IV) .

We excluded patients with recurrent stroke, receptive aphasia with inadequate comprehension to understand and follow the training schedule, significant cognitive deficits affecting participation (Mini Mental Status Examination score ≤ 23) , significant depression interfering with active participation (score < 12 on the 24-item Hamilton Rating Scale for Depression) , movement disorders interfering with training, recent myocardial infarction (< 6 months), ischemia or angina at rest or during exercise, and orthopedic conditions such as severe arthritis and total hip or knee joint replacements.

2.2

Design

The study was a randomized, controlled, single-blind design. The sample size was computed before the study based on the primary outcome measures and gait parameters required to be recorded. Included patients were randomly allocated to 1 of 3 groups by a table of random numbers. Group I received overground task-oriented gait training by a conventional physiotherapy approach. Group II received gait training on a treadmill without bodyweight support (full weight bearing); patients wore an overhead harness for security and to ensure similar experimental conditions, but no mechanical support was provided. Group III received gait training on a treadmill with bodyweight support (40% unweighting of body weight), which remained unchanged throughout all training sessions. Horizontal movement was provided by a slow moving treadmill, the speed of which was gradually increased on the basis of the patient’s comfort with the speed in both groups II and III. The gait training was provided by the therapists giving routine physiotherapy. Patients in groups II and III were blinded by ensuring similar experimental conditions (harness ± unweighting), but patients in group I were not blinded because they received overground task-oriented gait training. The physical therapist who assessed the gait parameters before and after the gait-training protocol was blinded to patient group for patients in groups II and III.

All patients received simultaneous comprehensive rehabilitation input in the form of medical management, routine physiotherapy, occupational therapy, orthotics, speech therapy and cognitive retraining, as needed. They all received basic physiotherapy in the form of stretching and strengthening exercises, spasticity-reducing exercises, balance and weight-shifting activities and postural control activities for 40 min/day.

Gait training was given for 30 min/day in all 3 groups. For group I, receiving conventional physiotherapy, overground task-oriented gait training involved walking in front of a mirror and walking along a straight line to give visual feedback for reducing gait deviations, walking a minimum distance during each session and gradually being encouraged to walk with increased speed. Patients trained on a treadmill were allowed to walk for a maximum of 5 trials during each session, each 6 min long, with intermittent rest for a total duration of at least 30 min.

Patients were evaluated before the start of training (baseline or pre-training) and at the completion of 20 sessions of training, 5 days/week for 4 weeks (post-training). Follow-up evaluation was at 3 months after completion of the training. Patients who did not return for the scheduled follow-up were contacted by mail and telephone. A blinded evaluator performed all evaluations and was not aware of group assignment for groups II and III.

2.3

BWSTT (Biodex Gait Trainer™)

The BWSTT system used was the Biodex Gait Trainer™, which is designed specifically for rehabilitation and retraining of gait for patients with neurologic and orthopaedic gait dysfunctions. The treadmill permits walking to be initiated from 0.0 km/h and increased by increments of 0.16 km/h. The patient can also hold onto a horizontal bar attached to the front of the treadmill for stability. The bodyweight support system includes an overhead harness with a pelvic band 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 supported.

2.4

Measurement tools

To measure the effectiveness of the gait training, the primary outcome variables were the quantitative clinical gait parameters of overground walking speed and overground walking endurance. Overground walking speed was measured in meters per second as the patient walked across a 10 m walkway. Patients were allowed to use the walking aids they required, and necessary assistance to compensate for the loss of balance was provided. Walking speed is an objective yet inexpensive method to monitor gait rehabilitation . Overground walking endurance was measured in meters by asking the patients to walk back and forth along the 10 m walkway until they were unable to continue. Patients were permitted to continue up to a maximum distance of 320 m .

Secondary outcome variables were the Scandinavian Stroke Scale (SSS) and FAC. Impairment after stroke was assessed by the SSS, which consists of 8 variables ranging from consciousness, cognitive deficits, motor power, and walking ability graded from 0 to 12. The maximum score is 58 and higher scores indicate better outcome . The FAC distinguishes 6 levels of required support during gait without taking into consideration any aid used. This scale has been validated for classifying walking handicap after stroke . Patients perform the test with a cane but without orthosis. Levels are defined as:

• •
  

  0: the subject cannot walk at all or requires the help of 2 or more people;

  
  
• •
  

  1: the subject needs continuous support from one person who helps carry the subject’s weight and helps with balance;

  
  
• •
  

  2: the subject is dependent on the continuous or intermittent support of one person to help with balance and coordination;

  
  
• •
  

  3: the subject needs only verbal supervision;

  
  
• •
  

  4: the subject requires help for stairs and uneven surfaces;

  
  
• •
  

  5: the subject can walk independently anywhere.

2.5

Statistical analysis

We compared the 3 groups on baseline outcome measures by ANOVA followed by post-hoc test for continuous variables and Kruskal-Wallis and chi-square tests for ordinal and nominal variables. Within each of the 3 groups, we analyzed change in each outcome variable by paired t test for continuous variables and Wilcoxon Rank test for ordinal variables (pre-training vs post-training and pre-training vs 3-month follow-up scores). Then we compared the 3 groups on outcome variables at the completion of training (post-training) and follow-up (3 months) by ANOVA followed by post-hoc test for continuous variables and Kruskal-Wallis test for ordinal variables. P < 0.05 was considered statistically significant.