| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Research Reports |
We thank Carey for recognizing the importance and challenges of using a scientific approach to investigate practical clinical problems. Rehabilitation randomized clinical trials (RCTs) are challenging in both the design and implementation. In addition to meeting the CONSORT1 requirements for RCTs, a clinical trial that has practical relevance also incorporates features that increase its clinical applicability.2 As acknowledged by Carey, the STEPS investigators chose to: (1) design clinically relevant interventions that physical therapists commonly implement in practice, (2) use multiple clinical sites, including 2 sites that were in community-based, outpatient physical therapy settings, (3) use randomization procedures to balance participant severity across groups, and (4) apply appropriate statistical methods for sample size determination and the intention-to-treat analysis.3 Our trial specifically targeted walking performance in individuals poststroke. The literature in this area was already populated with previous phase I studies of body-weight–supported treadmill training (BWSTT),4 locomotor-based resistive cycling,5 and strength training in individuals with stroke6 that provided evidence for the apparent efficacy of the interventions we investigated in the phase II STEPS trial. In essence, we owe a great deal of gratitude to the many investigators who came before us and who laid the groundwork for the efficacy of various treatment modalities designed to improve gait performance in individuals poststroke.
As Carey points out, this study was not without its variations and problems. The logistics of implementing a multi-site rehabilitation trial are complex. The STEPS trial trained and standardized 8 intervention therapists and 10 blinded assessors across 3 clinical sites in 2 geographical regions. Despite regular oversight by investigators and research coordinators, protocol variations did arise. We worked to identify how these variations and problems might affect the interpretation of our data. We used consultation with other Physical Therapy Clinical Research Network (PTClinResNet) investigators to determine the best approach to deal with protocol violations, and we feel this helped to foster objectivity and to diminish any bias in our analysis.
Carey correctly points out that we interpreted our results to indicate that training specificity combined with high intensity was the primary factor in the superior gains in walking speed achieved by subjects in the BWSTT groups when compared with those in the combined resistive leg cycling and upper-extremity ergometry (CYCLE/UE-EX) intervention group. We favor this interpretation because the cycling intervention was not designed to focus on walking speed as a training variable. Rather, the cycling protocol emphasized muscle strength and endurance. The fact that walking endurance (as measured by the 6-minute walk distance) was improved in the resistive lower-extremity cycling (CYCLE) group supports the specificity interpretation for the CYCLE intervention. Furthermore, gains in walking speed and endurance associated with the BWSTT supports the specificity interpretation for the BWSTT intervention because the intervention focused on both progression in speed (ie, progression of treadmill speed across sessions) and endurance (ie, progression in continuous walking time). The over-ground walking training, following each BWSTT session, may be a confounder, as Carey suggests. However, we believe that this possibility is not likely and disagree that this "remains unproven." Previous studies comparing the effects of BWSTT with those of over-ground walking have demonstrated that walking on a treadmill with body-weight support is more effective than over-ground walking alone7 and that walking on a treadmill at higher speeds is more effective than walking on a treadmill at slow speeds. Thus, BWSTT provides a more intense, task-specific walking experience that is more likely the "active" ingredient associated with intervention effectiveness than walking over the short distance of 15 m. In order to increase the clinical relevance of a gait training session that includes BWST, we included the short over-ground walk to better represent what therapists do in the clinic. Typically, if a therapist uses BWSTT as a therapeutic modality, it is incorporated with some—and most likely more than 15 m—of over-ground gait training.
We agree with Carey's excellent survey of the current questions about underlying neurophysiologic mechanisms associated with training effects postexercise. In fact, functional magnetic imaging studies of ankle dorsiflexion movements before and after a minimum of 12 sessions of BWSTT that used the same training parameters used in the STEPS study suggest evidence of cortical reorganization (ie, experience-dependent neuroplasticity) as a result of this type of high-intensity, task-specific training modality in adults with acquired stroke8 and in children with hemiparetic cerebral palsy who experienced stroke at birth.9 Furthermore, we, too, were interested in the underlying mechanisms associated with the STEPS interventions. As a pilot study to investigate underlying biomechanical and neurophysiologic mechanisms associated with the interventions, we enrolled the first 20 subjects (5 from each of the 4 intervention groups) to participate in a fully instrumented gait and motion analysis study with fine-wire electromyographic (EMG) recording of lower-extremity muscle activity that occurred before and after treatment. We reported the results of an individual case study of one of our STEPS participants in the combined BWSTT/CYCLE group who demonstrated a 30% increase in composite torque production in the hemiparetic limb after intervention.10 We argue that changes in strength that occur as a result of a relatively short activity or exercise bout such as the 6-week program provided in this case would be an example of neural adaptation. In addition, this participant demonstrated improvements in hip extension motion during terminal stance and increases in hip extensor torque that were accompanied with increased EMG activation and appropriate timing during gait that could be attributed to changes in volitional drive during walking. Further studies can be designed to investigate neural mechanisms that may be responsible for these improved activation patterns.
In addition, there are many factors that also may play a role in postintervention improvements in walking speed after stroke such as increased confidence and familiarity with a person's capabilities and limits, as well as improvements in self-efficacy that are accrued as an individual takes an active role in improving his or her skill level. Furthermore, lack of responsiveness to therapy may reflect the magnitude of white matter tract damage that reflects individual variation in lesion location and volume. We strongly agree with Carey that many future studies can be designed to understand not only the mechanisms associated with intervention effectiveness but also the individual factors that influence therapy responsiveness on a case-by-case basis.
We do not question the possibility that neurotrophic factors may be stimulated by exercise. In fact, we carefully controlled for intensity in the "active" interventions (ie, BWSTT, CYCLE, lower-extremity progressive-resistive exercise [LE-EX]) to ensure that exercise intensity, perceived effort, and exercise progression were moderately high across these interventions. Carey points out that the brain-derived neurotrophic factor (BDNF) has been demonstrated to increase during treadmill, cycling, or running tasks due to increases in physical activity. If we accept our argument that physical activity was generally equal across the active interventions and generally lower in the BWSTT/UE-EX intervention, then the only logical interpretation is that differences between the BWSTT groups and CYCLE group must be attributed to the task-specific differences between these interventions. In fact, the neurotrophic argument is further diluted because the BWSTT/CYCLE and BWSTT/LE-EX groups actually received a higher activity stimulus and, thus, a greater exercise-induced neurotrophic effect. Yet, walking gains were not better than in the BWSTT/UE-EX group, and strength gains were less.
Carey questions the use of the upper-extremity ergometry protocol as a "sham" comparison with the lower-limb exercise regimens. In the past, effects due to nonspecific activity and time spent with a supportive and motivational therapist might have been described as a "placebo" effect, but, as Carey points out, often these "nonspecific effects" have a real physiologic basis underlying the behavioral changes. We can state, with confidence, that the upper-extremity ergometry intervention was not designed to increase muscle strength in the lower limbs, nor was this intervention designed to be at the same level of intensity of the other 3 interventions. Some recent research appears to indicate that there is a neural connection between the arms and legs that may account for short-term effects related to reflex connectivity; however, it is unlikely that these short-term reflex influences could substantially account for the long-term effects on muscle torque production that we observed. We do not agree that an additive effect from the "sham" intervention may be responsible for the superior outcomes in the BWSTT/UE-EX group. Moreover, the work of Kwakkel and colleagues11 provides strong support for the primary role of task specificity in motor improvement with rehabilitation intervention after stroke; they demonstrated that when patients with stroke trained in upper-extremity tasks, improvements occurred only in the upper extremities, and vice versa with lower extremities.
As did Carey, we, too, struggled to understand the overall effectiveness of the intervention in the BWSTT/UE-EX group, which was the only group to realize significant increases in walking speed, walking distance, and lower-extremity strength. After consultation with researchers in exercise science, we believe, along with our colleagues, that the most likely explanation is that we induced an "overtraining" effect similar to that observed in individuals who do not have neurologic weakness conditions such as those that occur with stroke.
We would like to thank Carey, once again, for his thought-provoking insights into the results of the STEPS trial. Carey's final statements about possible future directions related to spatial and complex learning as an adjunct to the physical aspects of exercise training are exciting and stimulating. We hope that the work of Carey and others who are exploring the effects of task complexity during physical training will stimulate improved outcomes and richer, more satisfying lives for people recovering from stroke.
 | References |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
