Body+Weight+Suspension+Devices+as+a+Gait+Training+Intervention

Clinical Question:

 * The purpose of this Clinical Research Project was to evaluate the effectiveness of using automatic suspension devices with partial body-weight support during gait training to rehabilitate various patient conditions.

Search Terms:

 * automatic suspension
 * body-weight support and gait training
 * partial body-weight support
 * body-weight gait
 * body weight treadmill

Inclusion Criteria:

 * Any device that provides partial body weight support during overground or treadmill gait training

Exclusion Criteria:

 * The use of any exoskeleton or mechanical device in conjunction with the partial body weight support during gait training.
 * Complete Spinal Cord Injury

Background Information:

 * Description of the BWS system: “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.”
 * Most of these devices have a microprocessor that maintains a set suspension force and has emergent locking that detect when user has a sudden change in load or height.

= Evidence = REHABOT:

“Automatic Suspension Device for Gait Training” (6)

 * Use of REHABOT, which is an automatic suspension arm that swings in a circle and supports body weight while patients walk in a circle with a handrail and no propulsion.
 * 23 participants - 16 Orthopedic (bilateral joint replacements, unilateral joint replacements, juvenile arthritis and bilateral trans-femoral amputee) and 7 neurologic (hemiparesis, incomplete quadriplegia) patients who had trouble ambulating in parallel bars.
 * Walked 10-100m, once a day, 5x a week.
 * Average use was for 5 weeks (3 weeks orthopedic and 9 weeks neuro).
 * At the end of the program participants functional level: CNS: 1 only Rehabot, 2 parallel bars, 2 walker, 2 cane. Orthopedic: 2 walker, 13 cane, 1 independent.
 * 3 specific cases where they couldn’t initiate gait in a therapeutic pool 1 had bilateral tibial osteotomy ended with cane, 1 transfemoral amputee with hip replacement on opposite leg ended with prostheses and crutch, and 1 incomplete quadriplegia who ended with using a walker in home.

FLOAT:

Background

 * Previous studies suggested that active overground walking is better at improving skilled walking compared to treadmill gait training because it enables walking on natural surfaces, turning, and obstacle negotiation.
 * FLOAT is an overhead cable system that controls vertical forces and two horizontal planes and permits a variety of training scenarios including level surface ambulation, stair training and transfer training.
 * Purpose: to understand the gait patterns of healthy individuals in order to create the best program for patients that require a similar device for gait training.

Methods

 * 19 healthy volunteers (9 female, 10 male) participated in the study.
 * Force sensors on the FLOAT controlled the unloading force; unloading force was strictly vertical.
 * Gait kinematics and EMG activity in the rectus femoris, biceps femoris, tibialis anterior, and gastrocnemius medialis were recorded.
 * Subjects walked barefoot along 8m walkway at .56 m/s to match walking speeds observed in neurological patients. Only the middle 6 cm were used for analysis to eliminate acceleration and deceleration.
 * At least 20 gait cycles were recorded per individual.
 * There were six different loading conditions assessed - baseline, 10%, 20%, 30%, 40%, and 50% of bodyweight unloading - order was pseudorandomized.
 * Data analyzed - step length, cadence, gait phases, upper body inclination angle for forward and backward tilt, lower limb joint angles.

Results and Discussion

 * BWS system does have effects on time-distance parameters and requires compensatory muscle activation to maintain normal gait pattern - step length increased, cadence decreased, stance and double support phases decreased, swing and single support phases increased - increased biceps femoris activity and decreased gastrocnemius medialis activity during stance phase with 50% bodyweight support.
 * FLOAT did not distort gait kinematics but the subjects adjusted gait-phase timing and muscle activation to maintain lower limb kinematics while keeping predetermined speed.
 * Body-weight support can help with balance training for patients with stroke or Parkinson’s, forcing them to spend more time in swing phase and single support phase; body-weight support leads to a more upright position during gait, which is good for patients with a SCI that typically crouch over (* could be from harness or reduced trunk loading) (study found more extended knee and hip angles and foot flat with unloading).

Conclusion

 * With unloading devices, specific muscle activity adjustments ensure near normal gait patterns - but can impaired patients adapt to gait modulation in the same way as the healthy subjects?

“Treadmill training with partial body-weight support after anterior cruciate ligament reconstruction” (8)

 * ====**Study Design**: randomized controlled trial====
 * **Purpose**: to compare the effectiveness of conventional physical therapy with treadmill training with partial weight support on rehabilitation after ACL reconstruction
 * **Why it Matters**: ACL injuries are very common and can lead to future injuries. Early ACL reconstruction recovery is limited by weight bearing restrictions after surgery. The first demand of subjects undergoing any LE surgery is restoring walking condition.
 * ** Subjects: ** 40 participants were randomly assigned to groups. Inclusion criteria: unilateral knee injury, preoperative MRI showing rupture of ACL, positive drawer test, Lachman test and pivot shift test confirming knee instability, no other injuries, ACL reconstruction performed by the same group of surgeons.
 * **Methods**:
 * Both the control and treatment groups performed therapeutic exercises 3x / week
 * The treatment group also performed treadmill training with partial body weight support with a brace limiting knee flexion to 60 deg starting at 1 week post-op.
 * Initial weight bearing was 30% with a 20% increase each week, aiming for double LE full WB and ability to shift gravity stably at 6 weeks
 * Initial speed of treadmill was 0.2 m/s and increased by 0.2 m/s each week
 * ** Outcome Measures: **
 * Measured at 12 and 24 weeks post-op:
 * Circumference of LE’s (10 cm above patella)
 * Holden classifications (walking capacity)
 * 10 meter walking times (using brace)
 * International Knee Documentation Committee (IKDC) score
 * Knee joint stability was assessed 24 weeks post-operation using a KT-1000
 * **Results**:
 * No significant difference between the groups at 24 weeks post-op in most of the measures. At 24 weeks, no one had post-op knee swelling, effusion or infection.
 * No significant difference in the treatment group at 12 weeks and the control group at 24 weeks in circumference of LE’s, Holden classification, and 10 meter walking time.
 * Significant differences were found at 12 weeks between the groups in the Holden classification, 10 meter walking time, circumference of the LE’s and IKDC scores.
 * **Conclusion**:
 * “The improvement in function in a subject’s LEs was clearly accelerated by the intervention of treadmill training with partial body weight support without compromising the stability of the knee in a given follow-up period.”
 * Partial body weight support during treadmill training provides a safer environment for recovering walking ability.
 * It can also help add closed chain exercises earlier on in rehab programs which place less stress on the implant.
 * ** Limitations ** : sample size and short-term follow-up period

Stroke:

“Treadmill walking with body weight support in subacute non-ambulatory stroke improves walking capacity more than overground walking” (3)

 * **Study Design**: double blinded, randomized control trial
 * ** Why it Matters: ** many patients have residual walking disabilities after a stroke and being able to walk independently is a major determinant of whether a patient will be able return home. It also has lasting implications for their quality of life and ability to complete ADL’s.
 * **Subjects**: 126 participants within 4 weeks of a stroke who were undergoing inpatient rehabilitation that were unable to walk.
 * **Intervention**: the experimental group did 30 min/day of treadmill walking with body weight support using an overhead harness, the control group did 30 min of overground walking per day.
 * **Outcome Measures**: walking quality and capacity, walking perception, community participation and falls
 * **Results**: at 6 months there was no difference between the groups in terms of speed or stride. However, on average, the experimental group walked 57 m further in the 6 min Walk Test compared to the control group. The experimental group also rated their walking 1 point, out of 10, higher than the control group. No difference between groups in community participation or number of falls.
 * **Conclusion**: treadmill training with body weight support results in better walking capacity and perception of walking compared to overground walking without harmful effects on walking quality.

"Optimal outcomes obtained with body-weight support combined with treadmill training in stroke subjects" (2)

 * ** Study Design ** : Randomized control trial
 * ** Subjects ** : 100 subjects past the acute stage of recovery after a stroke
 * Intervention: partial body weight support up to 40% (enough to support proper posture during gait) that was decreased with progression of training. Training was 4x per week for 6 wks, each session being 3 trials of no more than 20 minutes.
 * ** Outcome measures ** : balance, motor recovery, overground walking speed, and overground walking endurance
 * ** Results ** : significantly greater effect of locomotor training with BWS on more functionally impaired stroke subjects, as characterized by lower pre-training scores of overground walking speed and endurance, functional balance, and lower-limb motor recovery. Also, 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.
 * ** Conclusion ** : Using partial body weight support can be effective in improving aspects of gait such as walking speed and endurance, functional balance and lower limb motor recovery in older and more impaired populations.
 * ** Why it matters ** **:** This is an applicable training method that can help to provide a dynamic way of training where task-specific methods can be implemented in training. This method of rehabilitation also encourages the use of the affected limb(s) in a directly applicable manner to gait training after stroke.

“Body weight-supported treadmill training after stroke” (5)

 * The authors of this study reviewed other articles which focused on various forms of stroke rehab, including task-specific repetitive approaches and the Bobath program. Task-specific repetitive approaches prevailed in improving general motor functions, so the authors of this study based treadmill training after stroke on the principle of “whoever wants to relearn walking has to walk.”
 * One of the big advantages of using partial BWS in gait training after stroke is it allows hemiparetic patients to actually practice the movement of walking, instead of focusing on tone-inhibiting maneuvers in which sitting and standing are common, which is the conventional method of treatment
 * To participate in treadmill rehabilitation with partial BWS a patient should be able to sit at the edge of the bed independently, also cardiac risk factors, a history of recent DVT of LE, LE joint contractures, and arthosis can be limiting factors
 * Review of early studies in the utilization of partial BWS for post-stroke training revealed that locomotor training (with or without FES) had a greater effect on improving gait ability compared to conventional physiotherapy.
 * [[image:https://lh6.googleusercontent.com/e8L-u1JktjdW1fUDm8SQtBS4NQIL5dRz4tlqVSaX8jaiyHZRkCk1A0mZTZH-M_OZ3_50Gr7lOyR-a_QHavfGFs-_FszZwgzyWnOIGMmSb1HIoMtNgO86j4YYur4_6GAPeVwjZeMc width="289" height="202"]]
 * All patients in this study who were wheelchair bound before became ambulatory (with verbal cueing) by the end of the study.
 * The next study reviewed involved a RCT of 100 patients in which 50 were trained with partial BWS and 50 were trained full weight-bearing on LE. AFter 6 weeks of training, the partial BWS group scored significantly higher on “functional balance (P=0.001), motor recovery (P=0.001), overground-walking speed (P=0.029), and overground walking endurance (P=0.018). Even when a follow up was done 3 months later, the partial BWS group “continued to have significantly higher scores for overground walking speed (P=0.006) and motor recovery (P=0.039).”
 * A third study the authors reviewed was from the Berlin group, in which results favored a more intense approach to rehab following a stroke involving conventional physiotherapy in combination with treadmill training with partial BWS.
 * These authors also reviewed a recent study by Kosak and Reding involving 56 acute stroke patients in which physiotherapy was directly compared to treadmill training. The therapy given to these patients involved “aggressive, early, therapy-assisted ambulation using knee-ankle combination bracing and hemi-bar if needed”. Sessions for both groups lasted up to 45 mins/day x 5 days a week. Interestingly, overall scores between groups did not differ, but the treadmill training group had higher endurance and speed scores. This study concluded that both therapies were effective, but treadmill training may be more beneficial in patients who are difficult to mobilize using physiotherapy alone.
 * [[image:https://lh3.googleusercontent.com/FsHMttrTIX3590sf9u7OlfbQl3F8ppVnCLaDmoJkYAhtEOKz2XJniCHgnb4km6wiYcbKWBk7W9w_DPRWwoSjYlJtmsYx2-F_rKf7SRmXiKBRug6HBquoE4QDTfXK3PJpQ7JN7OjD width="287" height="277"]]
 * Daniellson and Sunnerhagen revealed that treadmill training with 30% BWS required less O2 consumption when compared to no BWS. This would be significant when working with patients who have comorbidities including cardiovascular problems.
 * Take-aways:
 * When treadmill training with partial BWS, support should not exceed 40% as to optimize weight-bearing training post-stroke
 * 15% BWS allowed hemiparetic patients to walk more symmetrically
 * During a 30-minute session of treadmill training with BWS, patients have the ability to practice up to 1000 gait cycles, a significantly greater number than 50 gait cycles during a typical therapy session
 * Recent studies have revealed that gait is more efficient with a higher belt speed, indicating that gait speed should be increased with treadmill training as soon as possible during stroke rehab
 * Treadmill training with partial BWS is a promising new approach to gait training after stroke, “patients walk more symmetrically, less spastically, and more efficiently on the treadmill as compared with floor walking.”

Cerebral Palsy: ==== “ Effect of body-weight suspension training versus treadmill training on gross motor abilities of children with spastic diplegic cerebral palsy” (4) ====
 * **Background**: Children with cerebral palsy have limited neuromotor capabilities, poor postural control and adaption, and low stance stability with various environments. Body-weight suspension training and treadmill training have both been used in the past to improve functional gross motor skills in children with spastic diplegic cerebral palsy (SDCP).
 * ** Study Design ** : Assessor-blinded randomized controlled trial
 * **Purpose**: To compare the effects that body-weight suspension training and treadmill training have on the gross motor functional skills of standing, walking speed, and sit-to-stand transitional movements in children with SDCP.
 * ** Subjects ** : 20 children with SDCP (7 boys and 13 girls) ages 6-8 were recruited from an outpatient clinic
 * ** Methods ** : Subjects were split into 2 groups - Group 1 received treadmill training and Group 2 received suspension training using dynamic spider cage. Subjects participated in 36 30-minute sessions (over a span of 12 weeks) of either the body-weight suspension or treadmill training. The suspension group used a dynamic spider cage that unweighted 30% of their body weight. Suspended subjects were instructed to walk forward, backward, sideways and to step over obstacles. For the treadmill group, walking speed started as self-paced, and it increased from 1.1 km/hr in the first 18 sessions to 3.3 km/hr in the last 18 sessions due to motor progress. In addition to the gait training, both groups received the same 40-minute therapeutic exercise program (3x/wk) for balance and postural control.
 * ** Outcome measures ** : The three outcome measures the researchers used were the dimensions “D” (standing) and “E” (walking) of the standardized observational instrument gross motor function measure (GMFM), timed 10-m Walking Test, and 5 times sit-to-stand. The timed 10-m Walking Test assessed their speed of walking, and the 5 times sit-to-stand assessed how well they can perform transitional movements. All of these measurements were taken at baseline, after 6 weeks (18 sessions), and after 12 weeks (36 sessions).
 * ** Statistical analysis: ** 2X3 mixed ANOVA was used to look at the effects of therapy and time
 * ** Results ** : After 6 weeks, both groups showed significant improvements in all outcome measures from baseline, but there was no significant difference in the outcome measures between the groups (showing that the effect of time of therapy was significant). After 12 weeks, the suspension group reported higher scores in both dimensions D and E on the GMFM assessment than the treadmill group. As far as the 10-m Walking Test, the suspension group had a higher mean walking speed than the treadmill group after 12 weeks. For the 5 times sit to stand, the suspension group decreased their time more than the treadmill group at 12 weeks.
 * ** Conclusion ** : The results show that both treatments have positive effects on walking abilities after 18 sessions, but that there isn’t a significant difference between the interventions until 36 sessions of treatment. Therefore, the body-weight suspension intervention is superior to the treadmill intervention only when done for an extended period of time. The researchers conclude that using body-weight support is favorable for children with SDCP because it allows for more gradual functional training to improve walking abilities, but that 12 weeks of the intervention is necessary to see significant progress.

“Treadmill training with partial body weight support in nonambulatory patients with cerebral palsy” (9)

 * Participants **
 * 10 children 4 boys 6 girls. Mean age 11.5. Range 6-18. 3 with spastic diplegia, 4 with spastic tetraparesis, and 3 with spastic tetraparesis with additional ataxia.
 * “(1) ability to stand holding on with both arms for at least 3 seconds, (2) no fixed contractures of the lower limb joints, (3) no other orthopedic or neurological diseases impairing mobility, (4) no previous operations (eg, tendon transfer or lengthening) or neurolytic treatment 6 months before study onset, and (5) no severe cognitive or communicative impairment, ie, all children could follow instructions.”
 * Methods **
 * Lasted 3 months. 3x a week for 36 times. Each session lasted 30 minutes, 5 minutes on and off walking.
 * FAC measurement = “Level 0 can not walk or requires the help of two or more people. Level 1 needs continuous support from one person with carrying weight and balance, and level 2 dependent on continuous or intermittent support of one person to help with balance or coordination. Level 3 needs only verbal supervision, level 4 help on stairs and uneven surfaces, and level 5 walk independently anywhere. It was tested on a 20-meter walkway.”
 * GM function measure of standing and walking
 *  Results **
 * FAC grade increased from a mean of 1.1 to 1.9 after the treatment
 * The standing score from the GM Function Measure increased from a mean of 10.9 to 15.9 after the treatment

“Efficacy of partial body weight-supported treadmill training compared with overground walking practice for children with cerebral palsy: A randomized controlled trial” (10)

 * **Purpose:** “(1) determine the safety and feasibility of a PBWSTT program held in a special school environment and (2) investigate if PBWSTT can increase the walking endurance, walking speed, and walking function at school of children and adolescents with CP and moderate to severe walking difficulty (GMFCS III and IV) compared with overground walking practice.”
 * **Subjects**: Randomized into support treadmill group or overground walking group.
 * 2x a week for 9 weeks. Walked maximum 30 minutes. “(1) systematically reduce body weight support, (2) progressively increase treadmill speed, and (3) emphasize upright standing posture and facilitate the normal kinematic components of the gait cycle.”
 *  **Outcome Measures**: Measures of study included criteria such as preferred self-selected walking speed (10-meter walk test - high reliability), walking endurance (10MWT), [|5] and walking function in the school environment (School Function Assessment - correlation of .8 and .9)
 * **Results**: No statistical difference between groups for any measures. Could be because other studies were not randomized/had a small sample size or because this training protocol had less training time (2 instead of 3x a week) and no before-study training on treadmill.

Developmental Delays:

“Body Weight Support Treadmill Training for Children With Developmental Delay Who Are Ambulatory” (7)

 * **Background**: PT is important for children with developmental delay because it can help manage their underlying deficits to improve their task performance and independence. It is especially beneficial to start gait training at an early age because at age 7 most children have already developed the gait pattern that they will use the rest of their life. Body weight supported treadmill training (BWSTT) is a unique method that makes it easier for the PT to provide feedback, and also allows the child to use feedback to vary their performance in a safe environment.
 * ** Study Design ** **:** randomized controlled trial
 * **Purpose**: To measure the effectiveness that BWSTT has on gross motor skill development and gait abilities in ambulatory children ages 2-5 with developmental delays.
 * ** Subjects ** : 24 children ages 2-5 who are diagnosed with developmental delay and are independently ambulatory. 12 were in the BWSTT intervention group and 12 were in the control group.
 * ** Methods ** : Subjects were randomly split into two groups. All 24 subjects did usual PT sessions with therapeutic exercises, but the 12 subjects in the BWSTT group received 15 min BWSTT sessions 3x/wk for 6 wks . The BWSTT intervention included being hooked into the LiteGait gait training device and walking at a self-selected pace at a set incline. The treadmill speed and incline increased and the amount of body-weight support decreased each session, as tolerated, so far as there weren’t increased gait deviations. While walking, the subjects received tactile feedback to correct deviations.
 * ** Outcome measures ** : They chose to measure self-selected gait velocity (10-m Walking Test) and GMFM dimensions D and E to objectively assess functional ambulation tasks. Outcome measures were tested at baseline, 4 weeks, 6 weeks, and 6 weeks post-intervention.
 * ** Results ** : The walking velocity outcome measure revealed a significant relationship between group and time (the BWSTT group had higher scores in the 10-m Walking Test than the control group at every time interval). The treatment group also showed higher GMFM scores than the control group at every time interval. The GMFM scores significantly increased for the control group over time, but not as much as the treatment group. Not only did the BWSTT increase gait velocities, but it also helped the toe-walkers have increased heel-toe gait and reciprocal arm swing.
 * ** Conclusion ** : The results show that the combination of traditional PT treatment and a high-intensity protocol of BWSTT is extremely efficient in increasing gross motor skills and gait abilities in young children with developmental delays. This study shows that BWSTT doesn’t only affect the outcomes temporarily, but that improvements are still maintained 6 weeks later when no longer doing the treatment.

=Videos:= media type="youtube" key="4suWpwMErPk" width="560" height="315" media type="youtube" key="8JbgbzB2pM4" width="560" height="315" media type="youtube" key="0R5SfHt2Nws" width="560" height="315" media type="youtube" key="bNOQyA5FhLY" width="560" height="315" media type="youtube" key="um8Gd1bUijA" width="560" height="315"

**Clinical Recommendations:**
 * We would NOT recommend REHABOT because the suspension track is a circle and is not practical for walking in everyday life
 * We would recommend FLOAT as long as they have enough muscle activity to ensure near normal gait patterns.
 * Partial body-weight support is recommended for ACL reconstruction rehabilitation to jumpstart functional gait while they have post-surgery weight-bearing restrictions.
 * The majority of evidence says it is recommended with patient who have had a stroke but the evidence is mixed. After a stroke, it can help with walking capacity, confidence, normalizing gait, LL recovery, endurance, and balance.
 * For patients with Cerebral Palsy, body weight suspension is recommended to help with postural stability, situational walking, and increases toward independent walking. However, some research says it must be done for an extended amount of time (12 weeks) to see the benefit of doing it instead of regular treadmill training.
 * For children with developmental delays, using body-weight suspension can assist with self-selected gait velocity and gross motor skills. In combination with traditional PT treatment, it improves these factors within 6 weeks and the results are lasting.

References:
1. Awai, L., Franz, M., Easthope, C. S., Vallery, H., Curt, A., & Bolliger, M. (2017). Preserved gait kinematics during controlled body unloading. Journal of NeuroEngineering and Rehabilitation,14 (1). doi:10.1186/s12984-017-0239-9, from <span style="background-color: #ffffff; color: #1155cc; font-family: Arial; font-size: 10pt; vertical-align: baseline;">[]

2. Barbeau, H., & Visintin, M. (2003, October). Optimal outcomes obtained with body-weight support combined with treadmill training in stroke subjects. Retrieved April 14, 2017, from []

3. Dean, C. M., Ada, L., Bampton, J., Morris, M. E., Katrak, P. H., & Potts, S. (n.d.). Treadmill walking with body weight support in subacute non-ambulatory stroke improves walking capacity more than overground walking: a randomised trial. Retrieved April 14, 2017, from []

4. Emara, H., El-Gohary, T., & Al-Johany, A. (2016, February 4). Effect of body-weight suspension training versus treadmill training on gross motor abilities of children with spastic diplegic cerebral palsy. Retrieved April 14, 2017, from []

5. Hesse, S., Werner, C., Bardeleben, A., & Barbeau, H. (2001, July). Body weight-supported treadmill training after stroke. Retrieved April 14, 2017, from []

6. Kawamura, J. Ide, T. Hayashi, S. Ono, H. and Honda, T. (1993). Automatic suspension device for gait training. Prosthetics and Orthotics International. Retrieved April 14, 2017, from []

7. Lowe, L., McMillan, A. G., & Yates, C. (2016, January 1). Body Weight Support Treadmill Training for Children With Developmental Delay Who Are Ambulatory. Retrieved April 14, 2017, from []

8. Luo, Y., Shen, W., Jiang, Z., & Sha, J. (2016, December). Treadmill training with partial body-weight support after anterior cruciate ligament reconstruction: a randomized controlled trial. Retrieved April 14, 2017, from []

<span style="background-color: #ffffff; color: #333333; font-family: Arial; font-size: 10pt; vertical-align: baseline;">9. Schindl, M. R., Forstner, C., Kern, H., & Hesse, S. (2000). Treadmill training with partial body weight support in nonambulatory patients with cerebral palsy. Archives of Physical Medicine and Rehabilitation,81(3), 301-306. doi:10.1053/apmr.2000.0810301, from <span style="background-color: #ffffff; color: #1155cc; font-family: Arial; font-size: 10pt; vertical-align: baseline;">[]

<span style="background-color: #ffffff; color: #333333; font-family: Arial; font-size: 10pt; vertical-align: baseline;">10. Willoughby, K. L., Dodd, K. J., Shields, N., & Foley, S. (2010). Efficacy of Partial Body Weight–Supported Treadmill Training Compared With Overground Walking Practice for Children With Cerebral Palsy: A Randomized Controlled Trial. Archives of Physical Medicine and Rehabilitation,91(3), 333-339. doi:10.1016/j.apmr.2009.10.029, from <span style="background-color: #ffffff; color: #1155cc; font-family: Arial; font-size: 10pt; vertical-align: baseline;">[]