Weight-supported treadmill versus over-ground training after spinal cord injury: from a physical therapist's point of view.

To the Editor:

In this letter, I would like to address the significance for physical therapy of a recent study by Dobkin et al. (1) This randomized clinical trial investigated the effectiveness of a combination of body-weight-supported treadmill training (BWSTT) and over-ground mobility training versus solely over-ground mobility therapy in subjects with an incomplete spinal cord injury (iSCI). A large number of the subjects with iSCI regained walking ability independent of the group to which they were assigned. Compared with historic data, an unexpectedly high percentage (92%) of patients with an American Spinal Injury Association Impairment Scale (ASIA) classification C regained an independent walking ability. In addition, the walking speed of the subjects classified as ASIA C and D did not differ between the groups. The authors concluded that BWSTF was not superior in restoring walking ability compared with over-ground mobility therapy in people with iSCI early after trauma.

From a physical therapist's point of view, the results of the study by Dobkin et al (1) could be considered very encouraging. As even a sensitive outcome measure such as walking speed (2) did not show any difference, Dobkin and colleagues had evidence that their intervention, which consisted of task-oriented over-ground training, was equally effective when compared with BWSTT. Body-weight-supported treadmill training can be considered an intervention with a thorough theoretical framework based on animal experiments and an impressive amount of animal and (mainly uncontrolled) human evidence indicating effectiveness (for a review, see Dietz (3)). Because physical therapy interventions often are criticized for their lack of scientific arguments based on theoretical background or treatment effectiveness, the results of Dobkin and colleagues' study are positive findings.

Several ideas might explain the lack of differences found. (1,4,5) In my opinion, the 3 prerequisites for improving locomotor abilities, as identified by Grillner and Wallen, (6) are sometimes forgotten, but they might partly explain the findings by Dobkin et al (1):

1. "The basic or stereotyped movement synergy to achieve propulsion" (6) can be regarded as the rudimentary stepping that is generated by the central pattern generator (CPG), and the CPG can be well trained by applying BWSTT. (3)

2. The "maintenance of equilibrium during propulsive movement" (6) is not trained to a large extent by BWSTT due to the body-weight unloading.

3. "Adaptation of the locomotor pattern to the behavioural goals of the person and the constraints of the environment" (6) is, in general, not trained by BWSTT, perhaps with the small exceptions of implementing changes in walking speed or treadmill inclination.

Body-weight-supported treadmill training can be considered a safe and practical intervention that can be applied at an early stage during rehabilitation where body-weight support and physical (or robotic) assistance enable patients with iSCI with partial paresis to perform leg movements. It can also be considered task-specific because it retrains reciprocal leg movements with appropriate sensory inputs. (3) Training the CPG improves leg muscle activation, but this cannot be sufficiently transferred into function, because patients who remained classified as ASIA B (motor complete) showed no gain in locomotor function. (1) Therefore, supraspinal input appears to be inevitable to restore locomotion, and BWSTT can complement but not replace task-oriented over-ground adaptive locomotor training.

In conclusion, physical therapy, in general, might profit from the results of the study by Dobkin et al (1) as these results increase the body of evidence of the effectiveness of task-oriented training and might get physical therapy out of its "not-better-than-placebo" status. The present multidisciplinary approach to treating patients after an iSCI appears to be successful, as a high percentage of patients with iSCI regained walking ability. More specifically for physical therapists, task-oriented over-ground training with or without BWSTT can result in a promising functional outcome for the patient, which is, from the therapist's and patient's point of view, all that matters.

Hubertus JA van Hedel, PT, PhD

Spinal Cord Injury Center

Balgrist University Hospital

Forchstrasse 340

CH-8008 Zurich, Switzerland

[e-mail]

References

(1) Dobkin B, Apple D, Barbeau H, et al; Spinal Cord Injury Locomotor Trial Group. Weight-supported treadmill vs over-ground training for walking after acute incomplete SCI. Neurology. 2006;66:484-493.

(2) van Hedel HJ, Wirz M, Curt A. Improving walking assessment in subjects with an incomplete spinal cord injury: responsiveness. Spinal Cord. 2006;44:352-356.

(3) Dietz V. Spinal cord pattern generators for locomotion. Clin Neurophysiol. 2003;114:13 79-1389.

(4) Wolpaw JR. Treadmill training after spinal cord injury: good but not better. Neurology. 2006;66:466-467.

(5) Dietz V. Good clinical practice in neurorehabilitation. Lancet Neurol. 2006;5:377-378.

(6) Grillner S, Wallen P. Central pattern generators for locomotion, with special reference to vertebrates. Annu Rev Neurosci. 1985;8:233-261.

[DOI:10.2522/ptj.2006.86.10.1444]

Editorial Board Response:

Thanks to Dr van Hedel for bringing this recently completed study and its findings to the attention of the Physical Therapy readership. Van Hedel highlights the importance of a recently completed multisite randomized clinical trial (RCT) in acute spinal cord injury (SCI). (1,2) The study compared 2 approaches to walking rehabilitation for individuals acutely status post incomplete spinal cord injury (iSCI). Regardless of the approach used, the outcome measures from both groups were much better than historically reported. (1) Specifically, a high percentage of the participants, classified with American Spinal Injury Association Impairment Scale (ASIA) C impairment, regained "independent walking ability" and attained walking speeds of ~l.0 m/s and nearing normal values. The trial was conducted in response to a National Institutes of Health request for applications (RFA) to initiate a clinical efficacy trial of body-weight-supported ambulation training following incomplete spinal cord injury (RFA HD-98-005, National Institute of Child Health and Human Development). (3)

Several questions can be raised from a clinician's viewpoint when considering translating this knowledge to practice, including: Who was in the study? What were the interventions? What was the delivery intensity? When after injury was the intervention delivered, and for how long (duration)? In other words, what decisions do we make as clinicians from the outcome of this trial, and what questions does it generate as a scientist? This response serves as a follow-up to the dialogue begun by van Hedel and offers only an introduction to a dialogue that certainly deserves a more in-depth analysis.

The recent RCT in SCI (2) studied people within 8 weeks of traumatic SCI and admitted for rehabilitation at 6 partnering clinical sites: Magee Rehabilitation Hospital (Philadelphia, Pa), Shepherd Center (Atlanta, Ga), Ohio State University (Columbus, Ohio), Rancho Los Amigos Rehabilitation Hospital (Los Angeles, Calif), McGill University/ Institut de Readaptation de Montreal (Montreal, Quebec, Canada), and University of Ottawa Rehabilitation Hospital (Ottawa, Ontario, Canada). (1,2) All participants received usual rehabilitation care plus 1 of 2 locomotor interventions 5 times per week for 12 weeks. As van Hedel comments, 1 of the interventions--locomotor training (LT)--is a physiologically based intervention supported by a vast literature in basic science of the neurobiology of walking and activity-based plasticity in animal models and some human evidence providing the theoretical background. The other comparison intervention comprised the conventional arm of the study, with subjects receiving mobility training for standing and walking over ground using supplemental equipment to accomplish the activity (eg, assistive devices, standing frames, parallel bars), comparable to current practice.

Although "over-ground training" may appear to be common to both interventions, the over-ground training provided in each study arm differed substantially in their theoretical underpinnings and subsequent execution and process of delivery. The LT protocol provided for intense practice of stepping in the body-weight support and treadmill environment followed by an immediate translation of skills to an over-ground environment. The intent of LT was to provide a task-specific sensorimotor experience to generate a stepping response. The goal of step training on the treadmill was primarily development of the capacity to step. Dally over-ground training immediately followed step training on the treadmill and included an assessment of transfer of the skills acquired on the treadmill to an over-ground environment (eg, step initiation, stepping, stopping, balance). The over-ground training also prepared the individual for home and community ambulation (eg, introduction of assistive devices, negotiating uneven terrain, varying walking speed). Locomotor training on the treadmill and then over ground was conducted following specific guidelines for maximizing weight bearing of the lower extremities, minimizing loading of the arms, mirroring the kinematics of walking (eg, upright posture, joint positions, arm swing, speed), and minimizing use of compensatory movements to walk. (1,2,4)

In comparison, training for the control group occurred only over ground and was consistent with usual care for gait rehabilitation following SCI. Thus, load bearing was initiated using a variety of equipment: tilt table, standing frame, or parallel bars with or without bracing, if needed. If gait training continued over ground, bracing and/or an assistive device were provided initially to afford support and compensate for weakness, balance deficits, and uncoordinated movement patterns. Training targeted weight bearing, improving the gait pattern, and balance. Changes in assistive device or bracing were part of progression. The control group matched the experimental intervention for a minimum of a half hour of weight bearing per day; this element was required to equate the intensity of training across groups. The control intervention did not include any treadmill-based training.

Van Hedel refers to the over-ground control training as "task-oriented" training and contrasts it to the LT intervention using body-weight support and the treadmill. Both interventions may be "task-oriented" in that both addressed walking, but each was executed quite differently. Certainly, the approaches were not impairment-based strategies (eg, strength training, endurance training, or spasticity-reducing methods (5)), but both afforded weight-bearing and walking practice.

Van Hedel points to the surprising outcomes of this study for people with an ASIA C classification of impairment and the high percentage of people who benefited (92%) relative to past reports of 40% achieving independent ambulation, defined as a Functional Independence Measure (FIM) Locomotor score of [greater than or equal to] 5. (1) Subjects with ASIA C classification walked an average of 1.0 m/s at 6 months following injury regardless of the intervention and achieved an FIM Locomotor score of [greater than or equal to] 5. This is an overwhelmingly positive and remarkable outcome.

One interpretation and clinical response may be that early, intense intervention targeting weight-bearing and walking practice is beneficial after SCI for recovery of walking, in particular for the ASIA C population. Comparisons of walking outcomes from recent databases from National Institute of Disability and Rehabilitation Research (NIDRR) Model SCI Centers, the VA SCI Medical Center Network, or the NeuroRecovery Network (Reeve Foundation) may provide a useful comparison of a comparable cohort for walking recovery and the current standard of care. A more in-depth analysis and description of the current rehabilitation practice for gait training after SCI would be helpful in identifying the critical elements of the conventional over-ground mobility training and progression that may have contributed to the RCT outcomes. The clinical response to the interventions as measured by the FIM and gait speed provides valuable information for efficacy and functional gains. Knowledge of the process of recovery and mechanisms of response to the interventions, however, would be advantageous from a scientific and clinical decision-making perspective.

In this trial, the safety and feasibility of providing this intervention in the acute SCI population was established with no differences reported in adverse reactions across study groups. (2) This result may be advantageous in designing and executing future studies advancing multiple or combined therapeutic strategies. Furthermore, a greater-than-expected attrition rate from randomization to 6-month follow-up accounted for a substantial reduction in the upper motor neuron study population from ASIA C (n=52) and D(n=7) to ASIA C (n=38) and D (n=7) (see Fig. 1 in Dobkin et al (2)), although withdrawal rates were reportedly similar across the 2 groups.

Two populations that did not benefit stand out from these findings. First, people with an SCI impairment classification of ASIA B did not demonstrate improved walking outcomes from LT or conventional over-ground training. This population may require combined or augmented therapies (ie, pharmacology, electrical stimulation) to improve outcomes for walking recovery. Second, 8% of the ASIA C population did not achieve independent walking and community walking speeds. The reasons contributing to the differential outcomes would be of interest in categorizing "responders" and "nonresponders," in modifying the intervention, or understanding other factors. People who did not show significant improvement at 6 months may be candidates for alternative approaches or, again, combined therapies.

This RCT was a first for examining an alternative physical intervention, a physiologically based intervention, targeting walking recovery in the acute SCI population. From that perspective, the trial is historical for SCI rehabilitation and an important event for clinicians and people with SCI, with strong evidence for early mobility and walking practice for people with ASIA C impairments. Evidence-based practice is a dynamic process mirroring both the evolution of science and the requisite translation to clinical practice. Findings of clinical trials likely provide the science and yet stimulate controversy and generate many more questions as translation into practice is considered. An example is an RCT that examined the use of methylprednisolone acutely after SCI to enhance motor recovery. Outcomes from that RCT continue to drive ongoing debate as to the clinical use and benefit of and scientific inquiry related to methlyprednisolone. (6-12)

Planning and conducting a clinical trial in SCI is a daunting process that continues to spawn discussion in scientific and rehabilitation communities delineating guidelines and criteria for executing a successful RCT and translation of findings to practice. (13-15) With this backdrop, dialogue examining this RCT in SCI for the recovery of walking, the outcomes, and its relevance within a continuum of scientific findings is necessary and productive in advancing the science of rehabilitation for people after SCI.

Andrea L Behrman, PT, PhD

Associate Professor

University of Florida 

Research Scientist

VA Brain Rehabilitation Research Center-Gainesville, Fla

References

(1) Dobkin B, Apple D, Barbeau H, et al; Spinal Cord Injury Locomotor Trial Group. Methods for a randomised trial of weight-supported treadmill training versus conventional training for walking during inpatient rehabilitation after incomplete traumatic spinal cord injury. Neurorehabil Neural Repair 2003;17:153-167.

(2) Dobkin B, Apple D, Barbeau H, et al; Spinal Cord Injury Locomotor Trial Group. Weight-supported treadmill vs over-ground training for walking after acute incomplete SCI. Neurology. 2006;66:484-493.

(3) Body weight supported ambulation training after spinal cord injury (RFA HD-98-005, National Institute of Child Health and Human Development). Available at: http:// grants.nih.gov/grants/guide/rfa-files/RFAHD-98-005.html. Accessed August 2006.

(4) Behrman AL, Harkema SJ. Locomotor training after human spinal cord injury: a series of case studies. Phys Ther. 2000;80:688-700.

(5) Barbeau H. Locomotor training in neurorehabilitation: emerging rehabilitation concepts. Neurorehabil Neural Repair. 2003; 17:3-11.

(6) Sipski ML, Pearse DD. Methylprednisolone and other confounders to spinal cord injury clinical trials. Nat Clin Pract Neurol. 2006;2:402-403.

(7) Eck JC, Natchigall D, Humphreys SC, Hodges SD. Questionnaire survey of spine surgeons on the use of methylprednisolone for acute spinal cord injury. Spine. 2006;31: E250-E253.

(8) McCutcheon EP, Selassie AW, Gu JK, Picklesimer EE. Acute traumatic spinal cord injury, 1993-2000: a population-based assessment of methylprednisolone administration and hospitalization. J Trauma. 2004;56:1076-1083.

(9) Hurlbert RJ, Moulton R. Why do you prescribe methylprednisolone for acute spinal cord injury? A Canadian perspective and a position statement. Can J Neurol Sci. 2002;29:236-239.

(10) Short D. Use of steroids for acute spinal cord injury must be reassessed. BMJ. 2000;321 (7270):1224.

(11) Bracken MB. Methylprednisolone and acute spinal cord injury: an update of the randomized evidence. Spine. 2001;26(24 suppl): S47-S54.

(12) Bracken MB, Shepard MJ, Holford TR, et al. Administration of methylprednisolone for 24 or 48 hours or tirilazad mesylate for 48 hours in the treatment of acute spinal cord injury: results of the Third National Acute Spinal Cord Injury Randomized Controlled Trial. National Acute Spinal Cord Injury Study. JAMA. 1997;277:1597-1604.

(13) Kleitman N. Keeping promises: translating basic research into new spinal cord injury therapies. J Spinal Cord Med. 2004;27:311-318.

(14) Anderson DK, Beattie M, Blesch A, et al. Recommended guidelines for studies of human subjects with spinal cord injury. Spinal Cord. 2005;43:453-458.

(15) Curt A, Schwab ME, Dietz V. Providing the clinical basis for new interventional therapies: refined diagnosis and assessment of recovery after spinal cord injury. Spinal Cord. 2004;42:1-6.

[DOI: 10.2522/pti.2006.86.10.1445]