NIH Public Access Author Manuscript J Neurol Phys Ther. Author manuscript; available in PMC 2012 July 11. NIH-PA Author Manuscript Published in final edited form as: J Neurol Phys Ther. 2010 December ; 34(4): 202–207. doi:10.1097/NPT.0b013e3181fd5eab. SPLIT-BELT TREADMILL TRAINING POST-STROKE: A CASE STUDY Darcy S. Reisman1,2, Heather McLean2, and Amy J. Bastian2,3,4 1Department of Physical Therapy, University of Delaware, Newark, Delaware 19716 2Kennedy Krieger Institute, Johns Hopkins University School of Medicine, Baltimore, Maryland 21205 3Department of Neurology, Johns Hopkins University School of Medicine, Baltimore, Maryland 21205 4Department of Physical Medicine and Rehabilitation, Johns Hopkins University School of Medicine, Baltimore, Maryland 21205 NIH-PA Author Manuscript Abstract Background/Purpose—Even after rehabilitation, many individuals with stroke have residual gait deviations and limitations in functional walking. Applying the principles of motor adaptation through a split-belt treadmill walking paradigm can lead to short-term improvements in step length asymmetry after stroke. The focus of this case study was to determine whether it is possible to capitalize on these improvements for long-term gain. Case Description—The participant was a 36-year-old female who was 1.6 years post-stroke. She had a slow walking speed and multiple specific gait deviations, including step length asymmetry. Intervention—The participant walked on the split-belt treadmill 3 days/week for four weeks, with the paretic leg on slow belt. The goal was 30 minutes of split-belt treadmill walking each day, followed by overground walking practice to reinforce improvements in step length symmetry. NIH-PA Author Manuscript Outcome—With training, step length asymmetry decreased from 21% to 9%, and decreased further to 7% asymmetry 1 month after training. Self-selected walking speed increased from 0.71 m/s to 0.81 m/s after training and 0.86 m/s 1 month later. Percent recovery, measured by the Stroke Impact Scale, increased from 40% to 50% post-training and to 60% 1 month later. Discussion—Improvements in step length symmetry were observed following training and these improvements were maintained 1 month later. Concomitant changes in clinical measures were also observed, although these were relatively modest. The outcomes for this participant are encouraging given the relatively small dose of training. They suggest that short-term adaptation can be capitalized on through repetitive practice and can lead to longer-term improvements after stroke. BACKGROUND AND PURPOSE According to the World Health Organization, 15 million people worldwide suffer a stroke each year. One of the primary concerns for individuals who experience stroke is the ability to regain walking function.1 Consequently, significant effort is focused on gait retraining Corresponding author: Darcy S. Reisman, Ph.D., P.T., University of Delaware, Dept of Physical Therapy, 322 McKinly Laboratory, Newark, DE 19716, PH: 302-831-0508, FAX: 302-831-4234, dreisman@udel.edu. Reisman et al. Page 2 NIH-PA Author Manuscript during rehabilitation following a stroke2 and efforts to develop and improve locomotor retraining programs are a major focus of rehabilitation research. Despite these efforts, many individuals with stroke have residual gait deficits3–8 and never achieve community ambulation status.9, 10 Spatiotemporal gait asymmetries are prevalent after stroke and are related to slow walking speed.6, 11, 12 Step length asymmetry also has consequences relative to other gait deviations. Specifically, a shorter non-paretic step length is related to a decreased propulsive force on the paretic leg,3 which limits forward propulsion of the body, is related to slow gait speed,3 and may be related to reduced efficiency.13 Thus, recent poststroke gait retraining interventions have targeted step length asymmetry to promote greater recovery of walking.14 NIH-PA Author Manuscript Many novel rehabilitation interventions for those with stroke have emerged in the past several years.14–16 Recently, a group of studies utilizing the principles of motor adaptation applied to walking have gained interest in the rehabilitation community because of their unique ability to target specific gait deviations.17–20 In these studies, motor adaptation has a specific definition: it is the process of modifying or adjusting a movement from trial-to-trial based on error feedback21 such that a new movement pattern is temporarily learned to respond to new task demands. Once the adaptation is complete, if the new demand is removed, movements are once again erroneous, this time in the opposite manner, because the adapted pattern remains. These initial, oppositely directed errors are termed after-effects. With continued practice with the demands removed, the movement pattern returns to baseline, again by modifying or adjusting the movement from trial-to-trial based on error feedback. Using this definition, then, motor adaptation is one specific component of true motor skill learning. To fully learn a novel motor skill (i.e., retain permanently) requires much longer time periods; the process is influenced not only by adaptive mechanisms, but also by offline learning, consolidation, and long-term storage.22, 23 NIH-PA Author Manuscript Recent studies in persons post-stroke have shown that applying the principles of motor adaptation through a split-belt treadmill walking paradigm can lead to short-term improvements in step length asymmetry.19, 20 In order to obtain these improvements, a person’s step length asymmetry must be exaggerated while walking on the split-belt treadmill. Specifically, a person with stroke who takes a longer step with the paretic leg and a shorter step with the non-paretic leg during regular walking would need to walk on the split-belt treadmill with the paretic leg on the slow belt. Initially this will cause the person to take an even longer step with the paretic leg and an even shorter step with the non-paretic leg. Over time (less than 15 minutes), the person will correct the exaggerated asymmetry by lengthening the non-paretic leg step and shortening the paretic leg step. When the belts are returned to the same speed, the person will continue this adjusted pattern, resulting in improved step length symmetry.19, 20 The improvements in asymmetry are short-lived, and the person with stroke returns to the baseline asymmetry after several minutes of overground or treadmill walking. However, these results demonstrate that persons post-stroke retain the capacity to produce a more symmetric walking pattern, which could be capitalized on during rehabilitation. A critical component of the split-belt treadmill studies described above is the strategy employed to achieve an after-effect of improved symmetry. Specifically, the participants step length asymmetry is exaggerated with the split-belt treadmill, thereby augmenting their “error” (asymmetry) during split-belt walking. This error augmentation is critical because it provides the nervous system with a cue to correct the asymmetry.19, 20 This suggests an interesting and important concept for rehabilitation -- movement error augmentation may prompt the nervous system to make a movement correction. This could be especially useful for persons with chronic movement deviations, where the deviation may no longer be perceived by the nervous system as a movement error that requires correction.18 This would J Neurol Phys Ther. Author manuscript; available in PMC 2012 July 11. Reisman et al. Page 3 suggest that rather than providing interventions that correct movement errors, it may be useful to enhance error and allow the individual to correct the movement. NIH-PA Author Manuscript The temporary improvements in step length and double support asymmetry observed in persons with stroke following split-belt treadmill adaptation occur on the treadmill and also transfer to overground walking.19, 20 Thus, repetitive practice of the improved walking pattern following split-belt treadmill adaptation can be undertaken over ground, providing for optimal locomotor task-specific training. This makes the utility of this type of locomotor adaptation training particularly appealing for rehabilitation. However, in order to determine the potential value of split-belt locomotor adaptation in rehabilitation, it is important to know whether the short-term improvements observed can be translated into longer-term improvements through repetitive training. Thus, the purpose of this case study was to provide repetitive split-belt treadmill training in a person with stroke to determine whether longer-term changes in step length asymmetry and gait function could be achieved. PARTICIPANT NIH-PA Author Manuscript The participant in this case study was a 36-year-old female who sustained a single hemorrhagic stroke, of undetermined etiology, involving the right insular region one year and seven months prior to training. Her other past medical history was unremarkable. She gave informed consent to participate in training based on procedures approved by the Human Subjects Review Board at the John Hopkins School of Medicine. Following surgery and acute management, the participant received inpatient and outpatient rehabilitation, which ended several months prior to the start of this training intervention. Prior to her stroke, the participant was a full-time doctoral student. At the time of the training, she was continuing part-time with her studies and was not currently undergoing formal therapy. Her exercise activity was inconsistent and included walking on the treadmill and over ground outside when the weather permitted. She was independent in all activities of daily living and reported no history of falls in the past year. Observational analysis revealed that the participant walked without orthotic or assistive device, but with a slow walking speed for someone of her age. One the paretic side, stance time was noticeably shortened and step length was longer. She had difficulty clearing the ground with her paretic foot, and used compensatory patterns such as hip elevation and circumduction. She reported that fatigue limited her walking and the consistency of her exercise routine. The participant’s goal from this intervention was to improve her walking quality, speed, and endurance. NIH-PA Author Manuscript BASELINE EXAMINATION Clinical Evaluation A detailed baseline clinical assessment was performed before the start of training. The participant had no major cognitive deficits or symptoms of hemispatial neglect as determined by the star cancellation test.24 Self-selected walking speed was measured as the average of 3 trials along a 6-meter walkway, and fast walking speed was measured as the fastest of the 3 trials. The Timed Up and Go (TUG) test, lower extremity Fugl-Meyer, and Stroke Impact Scale (SIS) were administered using customary procedures.25–27 Self-selected and fast walking speed, TUG test, lower extremity Fugl-Meyer, and SIS were repeated immediately following the completion of training and then again one month later. The participant’s scores on these clinical tests are shown in Table 1. J Neurol Phys Ther. Author manuscript; available in PMC 2012 July 11. Reisman et al. Page 4 Gait Analysis NIH-PA Author Manuscript NIH-PA Author Manuscript Baseline overground walking characteristics of the participant were measured with computerized gait analysis using OPTOTRAK (Northern Digital, Waterloo ON) sensors that recorded 3-dimensional position data at 100 Hz from both sides of the body. Infrared emitting diodes (IREDs) were placed bilaterally on the foot (5th metatarsal head), ankle (lateral malleolus), knee (lateral joint space), hip (greater trochanter), pelvis (iliac crest) and shoulder (acromion process). The participant walked over ground down a 9-meter walkway at her self-selected gait speed. Data from 7–8 trials were recorded and the middle two strides from each trial were used for analysis. Step length was calculated as the anterior-posterior distance between the ankle markers at the time when each foot contacted the ground. Step length was labeled paretic or non-paretic based on the leading leg. The percent time in double limb support was the time that both feet were in contact with the floor, expressed as a percentage of the stride time for each leg. The stance time (the time from foot contact to liftoff) was also expressed as a percentage of the stride time. Data from instrumented gait analysis in persons with chronic stroke has shown high day-to day-reliability (ICC’s of 0.97–0.99) for spatiotemporal measurements during overground walking.28 Furthermore, step length data collected from 3 persons post-stroke during overground walking with the motion capture system used in this study found day-to-day changes of no greater than 0.7 cm. As shown in Figure 1, the participant demonstrated a 10 cm step length asymmetry at baseline, with the paretic leg taking a longer step compared to the non-paretic leg. Given her substantial step length asymmetry, we configured the split-belt treadmill belt speeds during training to target step length asymmetry (see training protocol below). INTERVENTION The participant trained 3 days/week for four weeks. Each session lasted approximately 1 hour, including a warm-up on the treadmill and rest periods between bouts. The training was designed so the participant would walk on the split-belt treadmill in six 5-minute bouts for a total of 30 minutes of split-belt treadmill walking each day. However, for the first session of training, the participant completed only 4 bouts and for the second and third sessions, she completed only 5 bouts due to fatigue. For all subsequent sessions, she was able to complete all 6 bouts. During each training block, heart rate and the Rating of Perceived exertion (RPE)29 were recorded at 2 and 4 minutes. If RPE exceeded 15 or heart rate exceeded 80% of age-predicted heart rate maximum, the participant was given a rest break. During the rest period heart rate and blood pressure were monitored. NIH-PA Author Manuscript During split-belt treadmill training, the participant walked with the paretic leg on the slow belt and the non-paretic leg on the fast belt. This leg-belt speed configuration was chosen so that the participant’s baseline step length asymmetry would be exaggerated by the split-belts (initially, the leg on the slow belt takes a longer step than the leg on the fast belt, thus exaggerating the participant’s asymmetry). The speed of the fast belt was close to the participant’s fast overground walking speed (1.0 m/s) and the slow belt speed was half of the fast belt speed (0.5 m/s). During all treadmill walking the participant wore a ceilingmounted harness for safety. The harness did not support body weight or interfere with participant’s walking. The participant was allowed to use the front handrail of the treadmill for safety, but was encouraged not to use this support. By session 6, the participant could consistently walk on the treadmill without using the handrail. Each day, following completion of all treadmill walking bouts, the participant practiced walking over ground for approximately 5 minutes with verbal cueing from the physical therapist to reinforce the improved step length symmetry. J Neurol Phys Ther. Author manuscript; available in PMC 2012 July 11. Reisman et al. Page 5 NIH-PA Author Manuscript All evaluations were completed by a physical therapist with over 17 years of clinical experience (DSR) and all intervention sessions were completed by a different physical therapist with 19 years of experience (HM). OUTCOMES The clinical evaluation and gait analysis performed at baseline were repeated within one week of the end of training, and then again 1 month following the end of training. Clinical Assessment The outcomes from the clinical assessment are presented in Table 1. Self-selected walking speed increased from 0.71 to 0.81 m/s after-training and continued to improve slightly to 0.86 m/s 1 month later. Fast walking speed increased from 1.13 m/s to 1.2 m/s following training and this was maintained at the 1 month follow up. The time taken to complete the TUG test decreased from 11.62 to 10.21 seconds after training, and decreased further to 8.85 seconds 1 month later. On the participation domain of the SIS, scores declined slightly posttraining, but then increased at 1 month post-training. Percent recovery as measured by the SIS increased by 10% at post-training, and increased an additional 10% 1 month later. The lower extremity portion of the Fugl-Meyer scale increased by two points at post-training and this increase was maintained 1 month later. NIH-PA Author Manuscript Gait Analysis The outcomes for step length are illustrated in Figure 1. The step length on both legs increased from baseline to post-training and this gain was maintained at one month followup. Most importantly, however, step length asymmetry, which was the gait deficit targeted by the intervention, improved from 21% asymmetry at baseline to 9% asymmetry posttraining, and improved further to 7% asymmetry by 1 month after training. The actual step length difference decreased from a 9.3 cm difference at baseline to a 4.5 cm difference posttraining, and then a 3.7 cm difference one month later. Stance and double support time asymmetries, which were not targeted by the intervention, remained essentially unchanged (Table 2). DISCUSSION NIH-PA Author Manuscript This case study examined whether short-term improvements in step length asymmetry following split-belt treadmill walking can be translated into longer-term improvements through repetitive training in a person with stroke. Improvements in step length asymmetry were observed following training and these improvements were maintained 1 month later. Concomitant changes in clinical measures were also observed, although in some cases these changes were relatively small. Nonetheless, the outcomes for this participant are encouraging, especially given the relatively small dose of training utilized. Step length asymmetry is a pervasive gait deviation observed after stroke.8, 11, 12 It is related to other gait deviations, such as decreased paretic leg propulsive force11 and to slow selfselected walking speed.11 The participant in this case study demonstrated an improvement of more than 50% in her step length asymmetry following split-belt treadmill training. Her actual step length difference decreased from almost 10 cm to less than 5 cm following training. This means that her step length asymmetry as a percentage of her paretic step length decreased from 21% to only 9%. Data from our lab has consistently shown that the average ± one standard deviation step length asymmetry in neurologically intact persons is around 5 cm,19, 20, 30 and that step length asymmetry as a percentage of step length is approximately 2%. Thus, this participant with stroke now demonstrates a step length asymmetry that is closer to the normal range observed for neurologically intact persons. In J Neurol Phys Ther. Author manuscript; available in PMC 2012 July 11. Reisman et al. Page 6 NIH-PA Author Manuscript addition, the improvement in step length asymmetry was achieved by an increase in step length on both legs, but with a relatively greater increase on the non-paretic leg. This is important because an improvement in asymmetry could also be achieved by a decrease in paretic leg step length. That type of change in asymmetry would not be as functionally desirable given that increased step length has been associated with increased walking speed following a treadmill and overground walking program in persons post-stroke.31 Step length asymmetry is relatively resistant to change with gait interventions after stroke15, 32 (although see Kahn and Hornby, 2009).14 One of the differences between the intervention in this case study and previous studies is the incorporation of error augmentation. During split-belt treadmill walking, the enhancement of the participant’s asymmetry lead to trial-and-error adjustment of step length and to the symmetric after-effect observed. Trial-and-error practice has been shown to be important for human motor learning33 and thus, may have been the mechanism underlying the improvements observed here. Small improvements in the participant’s walking speed were also observed following splitbelt treadmill training. However, this change did not exceed the minimal detectable change (MDC) in self-selected walking speed for persons with chronic stroke.34 NIH-PA Author Manuscript The time required by the participant to complete the TUG improved immediately following split-belt treadmill training, and showed continued improvement when tested 1 month later. The change of 2.77 seconds on the TUG exceeds the average change in TUG scores following overground gait training in persons post-stroke (1.81 seconds (95% CI=−2.29 to −1.33)),36 but does not exceed the MDC previously reported for persons with chronic stroke.34 The participant also reported an improvement in her recovery from stroke and an increase in participation (both measured by the SIS). The 10-point change in participation from baseline to 1 month follow-up for our participant is much larger than the approximately 2-point change observed in a previous study.38 A ten- to fifteen-point change in a single domain of the SIS has been suggested to represent real and meaningful change.37 Also, important for the participant, is that her perception of her recovery from the stroke increased from 40% at baseline to 50% following the intervention, and then to 60% 1 month later. NIH-PA Author Manuscript It is noteworthy that the participant’s perception of her recovery, her self-selected walking speed, and her time on the TUG all improved steadily from baseline to post-training to one month follow-up. Why did these measures continue to improve 1 month following the end of training? It is possible that the improvements in step length symmetry and walking speed observed following training allowed for greater ease of walking and greater walking practice, ultimately resulting in the continued improvement observed 1 month after training. Previous studies have shown that better step length symmetry is associated with faster walking speed,11 and that increases in walking speed are accompanied by decreased energy expenditure and increased daily step activity in persons post-stroke.39, 40 Unfortunately, we did not measure energy expenditure or daily step activity, which would have been useful information. LIMITATIONS There are several limitations to this case study. First, the participant experienced the testing protocol on multiple occasions, and therefore changes could have been due to experience and/or practice. Second, the changes observed could have been due to natural recovery, though this seems unlikely given that she was more than 1.5 years post-stroke at the time of testing. Third, the physical therapist performing the evaluations was not blinded to the J Neurol Phys Ther. Author manuscript; available in PMC 2012 July 11. Reisman et al. Page 7 NIH-PA Author Manuscript intervention. Finally, it is possible that the improvement in step length asymmetry was due solely to the overground walking training and had nothing to do with the split-belt treadmill training. However, 33% of community-dwelling persons living with chronic stroke (including this participant) continue to demonstrate persistent asymmetries following participation in conventional physical therapy, of which over ground locomotor training is a common component.8 Furthermore, other more conventional treadmill training studies that have also included an overground walking component have not found changes in step length asymmetry in persons post-stroke41. Thus, it seems unlikely that the overground walking component of the intervention alone (without the split-belt training prior to overground walking practice) accounted for the changes observed. However, a case study design does not allow us to rule out this possibility. CONCLUSIONS NIH-PA Author Manuscript Previous single-session studies have suggested that rehabilitation interventions that utilize motor adaptation may be effective at targeting specific movement deficits after stroke.19, 20, 42 This case study is the first report demonstrating that, through repetitive practice, it is possible to capitalize on the short-term adaptations observed in previous studies and obtain longer-term improvements. Specifically, improvements were observed in the participant’s step length asymmetry (the gait deviation targeted by the intervention), and these improvements were maintained one month later. Changes in walking speed, time on the Timed Up and Go test, social participation, and her perception of recovery from the stroke all showed more modest improvements; the impact of the intervention on these variables is not clear. Current studies are underway to determine whether similar improvements are observed across a larger group of individuals with stroke, and to determine baseline characteristics that influence a person’s response to this intervention after stroke. REFERENCES NIH-PA Author Manuscript 1. Mayo NE, Wood-Dauphinee S, Ahmed S, et al. Disablement following stroke. Disabil Rehabil. 1999 May-Jun;21(5–6):258–268. [PubMed: 10381238] 2. Jette DU, Latham NK, Smout RJ, Gassaway J, Slavin MD, Horn SD. Physical therapy interventions for patients with stroke in inpatient rehabilitation facilities. Phys Ther. 2005 Mar; 85(3):238–248. [PubMed: 15733048] 3. Bowden MG, Balasubramanian CK, Neptune RR, Kautz SA. Anterior-posterior ground reaction forces as a measure of paretic leg contribution in hemiparetic walking. Stroke. 2006 Mar; 37(3): 872–876. [PubMed: 16456121] 4. Chen G, Patten C, Kothari DH, Zajac FE. Gait differences between individuals with post-stroke hemiparesis and non-disabled controls at matched speeds. Gait Posture. 2005 Aug; 22(1):51–56. [PubMed: 15996592] 5. Nadeau S, Gravel D, Arsenault AB, Bourbonnais D. Plantarflexor weakness as a limiting factor of gait speed in stroke subjects and the compensating role of hip flexors. Clin Biomech (Bristol, Avon). 1999 Feb; 14(2):125–135. 6. Olney SJ, Griffin MP, McBride ID. Temporal, kinematic, and kinetic variables related to gait speed in subjects with hemiplegia: a regression approach. Phys Ther. 1994 Sep; 74(9):872–885. [PubMed: 8066114] 7. Olney SJ, Richards C. Hemiparetic gait following stroke. Part I: Characteristics. Gait Posture. 1996; 4:136–148. 8. Patterson KK, Parafianowicz I, Danells CJ, et al. Gait asymmetry in community-ambulating stroke survivors. Arch Phys Med Rehabil. 2008 Feb; 89(2):304–310. [PubMed: 18226655] 9. Lord SE, McPherson K, McNaughton HK, Rochester L, Weatherall M. Community ambulation after stroke: how important and obtainable is it and what measures appear predictive? Arch Phys Med Rehabil. 2004 Feb; 85(2):234–239. [PubMed: 14966707] J Neurol Phys Ther. Author manuscript; available in PMC 2012 July 11. Reisman et al. Page 8 NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript 10. Lord SE, Rochester L. Measurement of community ambulation after stroke: current status and future developments. Stroke. 2005 Jul; 36(7):1457–1461. [PubMed: 15947265] 11. Balasubramanian CK, Bowden MG, Neptune RR, Kautz SA. Relationship between step length asymmetry and walking performance in subjects with chronic hemiparesis. Arch Phys Med Rehabil. 2007 Jan; 88(1):43–49. [PubMed: 17207674] 12. Kim CM, Eng JJ. Symmetry in vertical ground reaction force is accompanied by symmetry in temporal but not distance variables of gait in persons with stroke. Gait Posture. 2003 Aug; 18(1): 23–28. [PubMed: 12855297] 13. Brouwer B, Parvataneni K, Olney SJ. A comparison of gait biomechanics and metabolic requirements of overground and treadmill walking in people with stroke. Clin Biomech (Bristol, Avon). 2009 Nov; 24(9):729–734. 14. Kahn JH, Hornby TG. Rapid and long-term adaptations in gait symmetry following unilateral step training in people with hemiparesis. Phys Ther. 2009 May; 89(5):474–483. [PubMed: 19282361] 15. Hornby TG, Campbell DD, Kahn JH, Demott T, Moore JL, Roth HR. Enhanced gait-related improvements after therapist- versus robotic-assisted locomotor training in subjects with chronic stroke: a randomized controlled study. Stroke. 2008 Jun; 39(6):1786–1792. [PubMed: 18467648] 16. Pohl M, Mehrholz J, Ritschel C, Ruckriem S. Speed-dependent treadmill training in ambulatory hemiparetic stroke patients: a randomized controlled trial. Stroke. 2002 Feb; 33(2):553–558. [PubMed: 11823669] 17. Blanchette A, Bouyer LJ. Timing-specific transfer of adapted muscle activity after walking in an elastic force field. J Neurophysiol. 2009 Jul; 102(1):568–577. [PubMed: 19420121] 18. Reisman DS, Bastian AJ, Morton SM. Neurophysiologic and Rehabilitation Insights From the Split-Belt and Other Locomotor Adaptation Paradigms. Phys Ther. 2009 Dec 18. 19. Reisman DS, Wityk R, Silver K, Bastian AJ. Locomotor adaptation on a split-belt treadmill can improve walking symmetry post-stroke. Brain. 2007 Jul; 130(Pt 7):1861–1872. [PubMed: 17405765] 20. Reisman DS, Wityk R, Silver K, Bastian AJ. Split-belt treadmill adaptation transfers to overground walking in persons poststroke. Neurorehabil Neural Repair. 2009 Sep; 23(7):735–744. [PubMed: 19307434] 21. Martin TA, Keating JG, Goodkin HP, Bastian AJ, Thach WT. Throwing while looking through prisms. II Specificity and storage of multiple gaze-throw calibrations. Brain. 1996 Aug; 119(Pt 4): 1199–1211. [PubMed: 8813283] 22. Ghilardi MF, Moisello C, Silvestri G, Ghez C, Krakauer JW. Learning of a sequential motor skill comprises explicit and implicit components that consolidate differently. J Neurophysiol. 2009 May; 101(5):2218–2229. [PubMed: 19073794] 23. Krakauer JW, Shadmehr R. Consolidation of motor memory. Trends Neurosci. 2006 Jan; 29(1): 58–64. [PubMed: 16290273] 24. Azouvi P, Samuel C, Louis-Dreyfus A, et al. Sensitivity of clinical and behavioural tests of spatial neglect after right hemisphere stroke. J Neurol Neurosurg Psychiatry. 2002 Aug; 73(2):160–166. [PubMed: 12122175] 25. Duncan PW, Bode RK, Min Lai S, Perera S. Rasch analysis of a new stroke-specific outcome scale: the Stroke Impact Scale. Arch Phys Med Rehabil. 2003 Jul; 84(7):950–963. [PubMed: 12881816] 26. Fugl-Meyer AR, Jaasko L, Leyman I, Olsson S, Steglind S. The post-stroke hemiplegic patient. 1. a method for evaluation of physical performance. Scand J Rehabil Med. 1975; 7(1):13–31. [PubMed: 1135616] 27. Podsiadlo D, Richardson S. The timed "Up & Go": a test of basic functional mobility for frail elderly persons. J Am Geriatr Soc. 1991 Feb; 39(2):142–148. [PubMed: 1991946] 28. Ng SS, Hui-Chan CW. The timed up & go test: its reliability and association with lower-limb impairments and locomotor capacities in people with chronic stroke. Arch Phys Med Rehabil. 2005 Aug; 86(8):1641–1647. [PubMed: 16084820] 29. Borg G. Ratings of perceived exertion and heart rates during short-term cycle exercise and their use in a new cycling strength test. Int J Sports Med. 1982 Aug; 3(3):153–158. [PubMed: 7129724] J Neurol Phys Ther. Author manuscript; available in PMC 2012 July 11. Reisman et al. Page 9 NIH-PA Author Manuscript NIH-PA Author Manuscript 30. Reisman DS, Block H, Bastian AJ. Inter-limb coordination during locomotion: What can be adapted and stored? J Neurophysiol. 2005; 94:2403–2415. [PubMed: 15958603] 31. Ada L, Dean CM, Hall JM, Bampton J, Crompton S. A treadmill and overground walking program improves walking in persons residing in the community after stroke: a placebo-controlled, randomized trial. Arch Phys Med Rehabil. 2003 Oct; 84(10):1486–1491. [PubMed: 14586916] 32. Patterson SL, Forrester LW, Rodgers MM, et al. Determinants of walking function after stroke: differences by deficit severity. Arch Phys Med Rehabil. 2007 Jan; 88(1):115–119. [PubMed: 17207686] 33. Criscimagna-Hemminger SE, Shadmehr R. Consolidation patterns of human motor memory. J Neurosci. 2008 Sep 24; 28(39):9610–9618. [PubMed: 18815247] 34. Flansbjer UB, Holmback AM, Downham D, Patten C, Lexell J. Reliability of gait performance tests in men and women with hemiparesis after stroke. J Rehabil Med. 2005 Mar; 37(2):75–82. [PubMed: 15788341] 35. Vereeck L, Wuyts F, Truijen S, Van de Heyning P. Clinical assessment of balance: normative data, and gender and age effects. Int J Audiol. 2008 Feb; 47(2):67–75. [PubMed: 18236239] 36. States RA, Pappas E, Salem Y. Overground Physical Therapy Gait Training for Chronic Stroke Patients With Mobility Deficits. Stroke. 2009 Sep 17. 37. Duncan PW, Wallace D, Lai SM, Johnson D, Embretson S, Laster LJ. The stroke impact scale version 2.0. Evaluation of reliability, validity, and sensitivity to change. Stroke. 1999 Oct; 30(10): 2131–2140. [PubMed: 10512918] 38. Schmid A, Duncan PW, Studenski S, et al. Improvements in speed-based gait classifications are meaningful. Stroke. 2007 Jul; 38(7):2096–2100. [PubMed: 17510461] 39. Moore JL, Roth EJ, Killian C, Hornby TG. Locomotor Training Improves Daily Stepping Activity and Gait Efficiency in Individuals Poststroke Who Have Reached a "Plateau" in Recovery. Stroke. 2009 Nov 12. 40. Reisman DS, Rudolph KS, Farquhar WB. Influence of speed on walking economy poststroke. Neurorehabil Neural Repair. 2009 Jul-Aug;23(6):529–534. [PubMed: 19126838] 41. Plummer P, Behrman AL, Duncan PW, et al. Effects of stroke severity and training duration on locomotor recovery after stroke: a pilot study. Neurorehabil Neural Repair. 2007 Mar-Apr;21(2): 137–151. [PubMed: 17312089] 42. Patton JL, Stoykov ME, Kovic M, Mussa-Ivaldi FA. Evaluation of robotic training forces that either enhance or reduce error in chronic hemiparetic stroke survivors. Exp Brain Res. 2006 Jan; 168(3):368–383. [PubMed: 16249912] NIH-PA Author Manuscript J Neurol Phys Ther. Author manuscript; available in PMC 2012 July 11. Reisman et al. Page 10 NIH-PA Author Manuscript NIH-PA Author Manuscript Figure 1. Step length (in meters) during overground walking for the paretic (black bars) and nonparetic (grey bars) leg at baseline, post-training, and at follow-up 1 month later. Each bar represents the average step length over strides and error bars represent the standard deviation of the step length over strides. NIH-PA Author Manuscript J Neurol Phys Ther. Author manuscript; available in PMC 2012 July 11. Reisman et al. Page 11 Table 1 Clinical outcome measures NIH-PA Author Manuscript Self-selected walking speed (m/s) Pre-training Post-training 1 month follow-up 0.71 0.81 0.86 Fast walking speed (m/s) 1.13 1.2 1.2 Timed Up and Go (sec) 11.62 10.21 8.85 59 56 69 Stroke Impact Scale participation % recovery Fugl-Meyer (Lower extremity) 40% 50% 60% 24/34 26/34 26/34 NIH-PA Author Manuscript NIH-PA Author Manuscript J Neurol Phys Ther. Author manuscript; available in PMC 2012 July 11. NIH-PA Author Manuscript 19.7 (1.3) % double support* 20.5 (1.5) 77 (1.4) 19.4 (1.1) 61 (1.9) 17.2 (1.5) 76 (1.4) Non-paretic Post-training Paretic Values reported are the mean (standard deviation) across strides * 62 (1.7) % stance* Non-paretic Pre-training Paretic NIH-PA Author Manuscript Temporal gait measures 19.5 (1.1) 61 (2.0) 17.3 (1.6) 76 (1.0) Non-paretic 1 month follow-up Paretic NIH-PA Author Manuscript Table 2 Reisman et al. Page 12 J Neurol Phys Ther. Author manuscript; available in PMC 2012 July 11.