Effects of Treadmill Exercise on Transcranial Magnetic Stimulation−Induced Excitability to Quadriceps After Stroke

Effects of Treadmill Exercise on Transcranial Magnetic Stimulation−Induced Excitability to Quadriceps After Stroke

229 ORIGINAL ARTICLE Effects of Treadmill Exercise on Transcranial Magnetic StimulationⴚInduced Excitability to Quadriceps After Stroke Larry W. For...

229KB Sizes 7 Downloads 17 Views

229

ORIGINAL ARTICLE

Effects of Treadmill Exercise on Transcranial Magnetic StimulationⴚInduced Excitability to Quadriceps After Stroke Larry W. Forrester, PhD, Daniel F. Hanley, MD, Richard F. Macko, MD ABSTRACT. Forrester LW, Hanley DF, Macko RF. Effects of treadmill exercise on transcranial magnetic stimulation⫺induced excitability to quadriceps after stroke. Arch Phys Med Rehabil 2006;87:229-34. Objective: To determine characteristics of transcranial magnetic stimulation (TMS)–induced measures of central motor excitability to the paretic and nonparetic quadriceps muscles of chronic hemiparetic stroke patients in the context of a shortterm, submaximal bout treadmill exercise. Design: Cross-sectional. Setting: Motor control and gait biomechanics laboratory. Participants: Convenience sample of 11 patients including cohorts of treadmill untrained (n⫽8) and trained (n⫽3) stroke patients with chronic hemiparetic gait. Intervention: Short-term submaximal treadmill exercise. Main Outcome Measures: Thresholds, amplitudes and latencies of TMS-induced motor evoked potentials at vastus medialis in paretic and nonparetic lower extremities. Results: Baseline characteristics of the motor evoked potentials (MEPs) show significantly higher motor thresholds, longer latencies, and reduced amplitudes on the paretic side. In crosssectional comparisons a group of treadmill-trained patients had greater paretic MEP amplitude changes after treadmill exercise versus paretic MEP responses from a group of untrained patients. Conclusions: These results indicate that treadmill training for 3 months or more may alter responsiveness of the lower-extremity central motor pathways to a short-term treadmill stimulus. Key Words: Evoked potentials, motor; Magnetics; Neuronal plasticity; Rehabilitation; Stroke; Treadmill test. © 2006 by the American Congress of Rehabilitation Medicine and the American Academy of Physical Medicine and Rehabilitation

From the Baltimore Veterans Affairs Medical Center Rehabilitation Research Service, Baltimore, MD (Forrester); Departments of Physical Therapy & Rehabilitation Science (Forrester), Neurology (Forrester, Hanley, Macko), and Medicine, Divisions of Gerontology and Rehabilitation Medicine (Macko), University of Maryland School of Medicine, Baltimore, MD; Geriatric Research, Education and Clinical Center, Baltimore, MD (Macko); and Division of Brain Injury Outcome, Department of Neurology, Johns Hopkins University, Baltimore, MD (Hanley). Supported by the Veterans Affairs (VA) Rehabilitation Research and Development (career development award no. B2375V), National Stroke Association, National Institutes of Health (grant no. R29 AG14487), National Institute on Aging, Claude D. Pepper Older Americans Independence Center (grant no. P60AG 12583), and the Baltimore VA Geriatrics Research, Education & Clinical Center. No commercial party having a direct financial interest in the results of the research supporting this article has or will confer a benefit upon the authors or upon any organization with which the authors are associated. Reprint requests to Larry W. Forrester, PhD, Dept of Physical Therapy & Rehabilitation Science, University of Maryland School of Medicine, 100 Penn St, Baltimore, MD 21201-1082, e-mail: [email protected]. 0003-9993/06/8702-10178$32.00/0 doi:10.1016/j.apmr.2005.10.016

HERE IS EMERGING EVIDENCE that chronic neuroT logic deficits due to stroke can be improved through intensive repetitive task-oriented motor training. The basis for 1-5

these improvements is thought to involve mechanisms of central neuroplasticity that are responsive to fundamental principles of motor learning, including intensive practice coupled with task-oriented problem solving. However, the precise temporal profile and mechanisms of long-term reorganization in neuromotor control (ie, motor re-learning) in stroke patients for functional activities such as walking are not well understood. One perspective is that durable long-term adaptations are a consequence of repeated exposures to rapid short-term plasticity associated with a given “dose” of training. This model of adaptation has been studied for normal motor learning situations using transcranial magnetic stimulation (TMS) to assess the short-term effects of specific training activities on corticospinal excitability to task-specific muscles. Using TMS in healthy human subjects, Classen et al6 showed rapid plasticity of kinematic encoding to single bouts of motor training following as few as 20 minutes of intensive exercise. Caramia et al7 reported that brief periods of moderate intensity exercise increased TMS-induced excitability of central motor hand pathways, peaking 15 minutes postexercise. Other studies show transient reductions in TMS motor evoked potentials (MEPs) after higher intensity, longer duration exercise.8-11 A better understanding of short-term central nervous system (CNS) plasticity responses in stroke patients may provide insight into the mechanisms underlying the more durable functional adaptations attained through motor re-learning. Although there is an upsurge of interest in using exercise to advance the recovery of gait function poststroke, there have been few studies that investigate exercise-induced lowerextremity (LE) neural plasticity.12,13 Most attention has been given to the use of treadmills to improve gait patterning during recovery (for review, see Dobkin14). We15 and others4,16,17 have shown that compared with overground walking, the treadmill walking provides an immediate stimulus for improved symmetry in loading forces, interlimb phasing, and stanceswing ratios. Furthermore, baseline comparisons of velocitymatched treadmill versus overground walking have demonstrated significant shifts in timing of electromyographic activation in both paretic and nonparetic quadriceps muscles.18 Other preliminary studies provide evidence that a progressive program of aerobic treadmill training improves strength in the quadriceps muscles of chronic stroke patients,3 as well as mediating improved overground gait patterns.19 Hence short-term facilitation effects of the treadmill on biomechanic, muscle activation, and interlimb coordination variables may contribute to long-term treadmill training benefits that improve gait patterning. One suggestion would be that long-term task-specific exercise training may lead to altered excitability of the central motor pathways. If so, the time course of how these changes emerge over the course of training is unknown. In this study, we use TMS-induced MEPs to quantify CNS excitability to the quadriceps muscles of chronic stroke patients to characterize neurophysiologic responses between limbs before and after a single session of submaximal treadmill exerArch Phys Med Rehabil Vol 87, February 2006

230

QUADRICEPS TMS AND TREADMILL IN STROKE, Forrester Table 1: Patient Characteristics, Assistive Devices, and Manual Muscle Test Patient/Sex

Age/Time Poststroke (y)

MMT Knee Flexion/Extension

Assistive Device

30ft (9m) Walk (m/s)

Lesion Location

U-1/M U-2/F U-3/F U-4/M U-5/F U-6/M U-7/M U-8/M T-1/M T-2/M T-3/M

58.4/3.0 69.1/10.0 48.2/3.0 63.3/1.0 68.2/1.7 62.3/1.7 59.3/0.8 68.4/0.8 60.9/4.6 62.5/1.8 72.5/1.5

4/4 4–/4⫹ 4/4 2/2 4⫹/4 4/4 4⫹/4⫹ 3/3 4–/4⫹ 5/5 4⫹/5

SPC W SPC, AFO SPC, AFO SPC, AFO SPC, AFO None SPC, AFO SPC None SPC

0.94 0.06 0.50 0.52 0.57 0.59 1.21 0.56 0.60 1.02 0.79

L subcortical R cortical R cortical R brainstem L subcortical L brainstem R subcortical L subcortical R subcortical L cortical R unavailable

Abbreviations: AFO, ankle-foot orthosis; F, female; L, left; M, male; MMT, manual muscle test; R, right; SPC, single point cane; T, trained; U, untrained; W, walker.

cise. We hypothesized that nonparetic quadriceps would exhibit greater excitability than paretic quadriceps, and that a single session of treadmill exercise would alter the short-term excitability to these muscles, seen as changed MEP amplitudes and/or latencies. We also compare the MEP responses of a small treadmill-trained cohort versus an untrained cohort to assess the potential for long-term neuroplasticity through treadmill training. METHODS We recruited 11 patients with persistent LE hemiparesis due to remote stroke (⬎6mo), and who had completed available rehabilitation therapies, from the University of Maryland, Baltimore (UMB) and the Veterans Affairs Maryland Health Care System (VAMHCS) clinics. All patients were screened for contraindications to TMS (ie, the presence of metal implants in the head, neck, and trunk regions, cardiac pacemakers, or history of seizures). Additional inclusion and exclusion criteria were oriented toward medical clearance necessary for aerobic treadmill training and are described in detail elsewhere.20 A group of 8 untrained patients was enrolled consecutively prior to group assignment in a randomized, controlled, 6-month progressive treadmill-training study. Prior experience with treadmill walking was limited in these patients to baseline cardiovascular screening tests for aerobic capacity and economy of gait. The group of 3 trained patients had already been randomized completed at least 3 months of progressive treadmill exercise training. Informed consent was obtained in accordance with joint UMB and VAMHCS Institutional Review Board policies in accordance with the Helsinki Declarations. Patients in the untrained group included 5 men and 3 women (mean age, 62.2⫾7.0y; mean time poststroke, 2.8⫾3.1y). The trained group consisted of 3 men (mean age, 65.3⫾6.3y; mean time poststroke, 2.6⫾1.7y). A summary of patient characteristics is provided in table 1. We affixed bipolar surface electromyography electrodes (Ag-AgCl; diameter, 1cm2) to the vastus medialis muscles of both legs. Patients were seated on a dynamometer,a stabilized by lap and shoulder belts, and were provided continuous visual feedback for maintaining prescribed knee isometric extensor force (fig 1). The vertex (Cz) was located according to the 10-20 International System and marked as the anatomic reference point for stimulator coil placements, aligned to the anteroposterior plane. Foam earplugs were inserted into the patient’s external ear canals to attenuate the stimulator discharge sound. Arch Phys Med Rehabil Vol 87, February 2006

We performed stimulations with a Magstim 200 unitb and a custom double cone coil (diameter, 110mm; maximum intensity, 1.4T) designed for deeper stimulus penetration of the cortices to elicit MEPs at the quadriceps of each leg. Electromyographic signals were amplified (1000⫻) and sampled at 2kHz. Stimulation onset times were concurrently recorded from the stimulator unit’s synchronization output signal. The TMS protocol involved determination of the optimal quadriceps stimulation site for each leg referenced to Cz, with motor thresholds defined as minimum stimulator power (% maximum) needed to elicit MEP responses of 50␮V or more in at least 5 of 10 consecutive stimulations under passive (rest) condition. This was followed by a series of 10 suprathreshold stimulations (10% above threshold) delivered at approximately 0.2Hz with active quadriceps facilitation defined by force feedback (approximately twice the resting weight of the lower leg). After baseline TMS measures of each leg, patients performed submaximal effort treadmill walking that mimicked the

Fig 1. Patient positioned for isometric testing of knee extension. Surface electromyography electrodes placed over the vastus medialis muscles bilaterally for recording TMS MEP responses. Visual feedback of knee extensor force output presented on computer monitor. Abbreviations: A-to-D, analog to digital; EMG, electromyography; PC, personal computer; TTL synch, transistor-transistor logic synchronization signal.

231

QUADRICEPS TMS AND TREADMILL IN STROKE, Forrester

RESULTS

50 L a te n c y ( m s)

Amplitude (mV)

2.0

1.0

0.0

25

0

NP

P

NP

P

Fig 2. Means ⴞ standard error of pretreadmill exercise comparison of MEPs between nonparetic (NP) and paretic (P) vastus medialis showing lower amplitudes and longer latencies on the paretic side. Data pooled for all subjects (nⴝ11). *P<.01.

routine treadmill aerobic training session as previously described.21,22 Typically, such sessions are conducted 3 times weekly for 6 months, with effort targeted to 60% heart rate reserve at initial training durations of 15 to 20 minutes per session. Here we sought to assess CNS responses to the treadmill stimulus without the confound of fatigue; thus, each trial was conducted at the patient’s preferred velocity on a level grade, with continual monitoring of comfort level. Sessions were terminated either when the effort exceeded a mild-tomoderate level or if the time reached 20 minutes. The postexercise TMS series were administered at the same stimulation levels to the same sites for each hemisphere. The retesting was completed within a 10- to 15-minute window following the single bout of treadmill exercise. Data Reduction and Analysis Amplified electromyography data were passively demeaned and digitally filtered with a 10 to 1000Hz bandpass. MEP onsets were defined by the occurrence of an electromyographic amplitude spike exceeding the mean ⫾ 3 standard deviations of electromyographic level during the 100ms prior to stimulation. Latency was defined as the elapsed time from stimulation to MEP onset averaged over 10 successive stimulations. MEP amplitudes were quantified as peak-to-peak differences poststimulation and averaged over the 10 stimulations. Because the data were not normally distributed in 14 of 18 variable samples between the 2 groups (Lilliefors test for normality), nonparametric Wilcoxon signed-rank tests were used for all comparisons, with ␣ set to .05 for 2-tailed probability.

Pre-Exercise MEPs Differences between the nonparetic and paretic MEP-derived measures at the quadriceps were robust. Thresholds for nonparetic quadriceps averaged 55.5%⫾11.2% maximum stimulator power compared with 63.5%⫾10.8% on the paretic side (P⫽.01). With the pre-exercise data pooled across all subjects (fig 2), the mean of nonparetic amplitudes was 3.5 times greater than on the paretic side (1.498mV vs 0.428mV, P⫽.01). The mean of nonparetic latencies was significantly lower than on the paretic side, indicating faster conduction times in the former (30.22ms vs 34.17ms, P⫽.01). Effects of Submaximal Treadmill Exercise on MEPs The pre-post treadmill exercise data for each group are summarized in table 2, and figure 3 depicts examples of MEP waveforms pre-post treadmill exercise. After the bout of treadmill exercise, the MEP amplitudes for the untrained group remained significantly different between limbs (nonparetic, 1.489mV vs paretic, 0.495mV; P⫽.01). However, their MEP amplitudes did not change significantly within either limb after the exercise session. Also in the untrained group, the betweenlimb differences in the latencies persisted after the bout of treadmill exercise (nonparetic, 27.82ms vs paretic, 31.56ms; P⫽.02). This between-limb difference was maintained despite small but significant reductions in latencies to both limbs (nonparetic, P⫽.03; paretic, P⫽.05). In the trained group, postexercise increases in paretic-side MEP amplitudes approached significance (.133mV, P⫽.11), while those on the nonparetic side remained unchanged. It should be noted that the relative change in the paretic-side MEP amplitude was nearly 10-fold greater than that of any of the other exercise-induced changes in MEP parameters. In this group the effect of exercise on latency differences between sides approached significance after exercise (P⫽.11), with nonparetic times decreasing 1.67ms versus a 0.35-ms increase on the paretic side. However, latency measures within-limbs did not differ significantly after the exercise session. Comparisons of Trained Versus Untrained Groups In this analysis, we compared the trained and untrained groups on the extent of MEP amplitude and latency changes

Table 2: MEP Amplitudes and Latencies for Both Nonparetic and Paretic Vastus Medialis in Each Group Amplitudes and Latencies

Untrained Amplitudes (mV) Nonparetic Paretic Latencies (ms) Nonparetic Paretic Trained Amplitudes (mV) Nonparetic Paretic Latencies (ms) Nonparetic Paretic

Pre

Post

⌬ (post–pre)

%⌬ (post–pre)

1.470⫾0.325 0.515⫾0.038

1.489⫾0.304 0.495⫾0.046

0.019 ⫺0.020

1.3 ⫺3.9

29.93⫾1.02 33.86⫾1.91

27.82⫾0.90 31.56⫾1.94

⫺2.11 ⫺2.30

⫺7.0 ⫺6.8

1.574⫾0.558 0.196⫾0.046

1.677⫾0.845 0.329⫾0.027

30.98⫾0.75 34.98⫾2.78

29.32⫾0.94 35.33⫾1.42

0.103 0.133 ⫺1.66 0.35

6.5 67.9 ⫺5.4 1.0

NOTE. Values are means ⫾ standard error.

Arch Phys Med Rehabil Vol 87, February 2006

232

QUADRICEPS TMS AND TREADMILL IN STROKE, Forrester

TM-Untrained (S50)

TM-Trained (S12)

NP pre-TM

NP post-TM

P pre-TM

P post-TM

NP pre-TM

NP post-TM

P pre-TM

P post-TM

0.1 mV 10 ms

Fig 3. Examples of 10 averaged TMS-induced MEPs at vastus medialis before and after a single session of treadmill walking exercise. Top 4 panels show trained patient’s nonparetic and paretic responses. Lower 4 panels show responses of an untrained patient. Arrows denote stimulus onset. Abbreviation: TM, treadmill.

due to the submaximal bout of treadmill exercise. The changes in amplitude (see table 2) did differ between groups for the paretic side (P⫽.05), but not the nonparetic side (P⫽.78). Notably the paretic MEP amplitudes increased for all 3 trained patients, whereas 6 of 8 decreased among the untrained group. The changes in latency did not differ between the groups for either the nonparetic (P⫽.59) or paretic sides (P⫽.16). DISCUSSION The present study was undertaken to make initial characterizations of the baseline corticospinal excitability in the quadriceps muscles of hemiparetic stroke survivors and to examine the short-term effects produced by a single submaximal effort treadmill walking session. We report significantly decreased MEP amplitudes and increased thresholds and latencies in paretic versus nonparetic responses to TMS in quadriceps of chronic hemiparetic stroke patients. Our findings further suggest that paretic quadriceps MEP responses are differentially affected by a single bout of submaximal effort treadmill exercise in trained versus untrained patients with chronic stroke. To our knowledge, this is the first report to demonstrate reduced excitability of central motor pathways to the paretic quadriceps utilizing TMS and provide evidence of short-term neuroplasticity with treadmill training in chronic hemiparetic patients. Numerous studies have shown baseline evidence of decreased motor control to the paretic upper extremity (UE) after stroke. Few TMS studies have examined the LEs in these patients, and none have reported on CNS motor control to the Arch Phys Med Rehabil Vol 87, February 2006

proximal leg muscles essential to gait. Previously it was shown that chronic stroke patients have reduced and altered functional magnetic resonance imaging (fMRI) cortical activation maps for paretic knee control compared to nonparetic legs and to healthy subjects.2 We now report that nonparetic quadriceps MEP profiles have more than 3 times the amplitude, 13% lower motor thresholds, and 12% faster latencies than paretic-side MEPs. Our results are in agreement with fMRI studies, documenting substantial decrements in all paretic leg TMS MEP parameters evaluated in this study. These findings are also in agreement with studies that show similar responses in reduced paretic UE excitability.23-27 Seitz et al25 reported MEP amplitudes to paretic hand muscles at 64% of the nonparetic amplitudes. In subacute phase stroke patients, nonparetic hand muscles had MEP amplitudes that were 4 times greater and motor thresholds that were 41% lower than on the paretic side.28 Koski et al23 also showed differences in hand muscle MEP thresholds (paretic 20% greater), amplitudes (paretic 22% of nonparetic), and latencies (paretic were 120% of nonparetic). Reports of UE proximal muscle excitability also indicate diminished paretic MEP amplitudes at the deltoid muscles.27 While further studies are needed to examine the relationship between abnormal MEP profiles and deficit severity across the spectrum of stroke recovery, it is clear that TMS MEPs provide discriminative measures between paretic and nonparetic central motor pathways. TMS may also be useful to study whether neuroplasticity underlies locomotor relearning after stroke. Studies of UE movement-based interventions provide a strong precedent that noninvasive TMS measures may index elements of adaptations in CNS motor control that are associated with functional gains. For example, TMS maps before and after therapy utilizing 2 weeks of intensive, task-oriented practice with paretic UE activities have shown increased brain areas eliciting MEP responses to the paretic hand muscles of stroke patients.1 More recently, Koski23 demonstrated that UE training cannot only affect long-term plasticity associated with improved function, but also that CNS excitability can be enhanced in both limbs after single therapy sessions. This latter finding has parallels in earlier work in healthy subjects showing that 20 to 30 minutes of task-repetitive practice produces short-term cortical motor plasticity.6 Moreover, recent studies in stroke patients reveal that application of transcranial direct current stimulation to the motor cortex for 20 minutes produces immediate adaptations in CNS motor excitability to hand muscles.29 This response was linked to improved motor function scores that persisted at least 25 minutes. The possible connections between short-term CNS changes and durable long-term changes in motor control with improved function are not proven definitively from these results. More longitudinal studies with higher density of noninvasive assessments of CNS motor function may eventually converge on this process. While an increasing body of literature demonstrates intervention-induced neuroplasticity in the UE after stroke, little is known regarding the potential for neuroplasticity in LE motor control. No prior studies, to our knowledge, have applied TMS to study training-related neuroplasticity to LE after stroke. However, other noninvasive imaging studies demonstrate the potential for promoting changes in CNS motor control after stroke. Near infrared spectroscopy was used during actual treadmill walking with stroke patients to show differences in cortical activation during 3 different therapy approaches.13 Greater cortical activation was found with pelvic manipulations to induce paretic leg swing compared with physically moving the thigh and foot forward. Another recent study used fMRI to study brain areas that control ankle dorsiflexion, an action that

QUADRICEPS TMS AND TREADMILL IN STROKE, Forrester

contributes significantly to the performance of walking.12 While this study reported on only 4 stroke patients in the chronic phase, it did demonstrate changes in several cortical areas over the course of body weight-supported treadmill training. We now report that in the paretic leg after treadmill exercise changes in MEP amplitudes were positive and significantly greater in the treadmill-trained versus untrained group. This cross-sectional comparison between trained and untrained subjects suggests that even a brief submaximal effort treadmill training session can produce different effects on LE central motor responses, depending on whether there has been significant exposure to progressive treadmill training. That this effect was seen in all the trained subjects raises the prospect that a regular program of treadmill training may promote long-term adaptive neuroplasticity in leg motor control. Notably, we have reported that treadmill exercise training improves paretic leg quadriceps strength and gait function in chronic hemiparetic stroke patients.3,19 Further studies are needed to establish whether TMS indices of neuroplasticity are related to the exercise-mediated gains in LE motor function after stroke. Results of this small cross-sectional investigation must be interpreted with caution. We do not know if the responses we report are treadmill specific, or whether similar effects could be elicited with comparable amounts and effort of overground walking. Future studies are needed to discern the potential for progressive aerobic overground training with stroke survivors. Also, the mechanisms and site(s) of neuroplasticity that may be invoked through treadmill training cannot be established in this study. The use of single-pulse TMS limits interpretation of where changes are occurring within the CNS. These data reflect the full neuromotor ensemble, including motor cortex, spinal pathways, and peripheral segments, any of which could be affected by the exercise stimulus. Moreover, our findings are limited by the small sample size and cross-sectional experimental design. Prospective randomized studies utilizing both TMS and neuroimaging to evaluate the profile of neuroplasticity in knee motor control with exercise are now underway. CONCLUSIONS TMS demonstrates a distinct neurophysiologic profile of higher motor threshold, reduced amplitude and prolonged latency to paretic quadriceps in chronic hemiparetic stroke patients. Cross-sectional analyses provide evidence that a single treadmill exercise training session produces short-term adaptations consisting of increased paretic leg TMS MEP amplitude in trained but not untrained chronic hemiparetic patients. Further studies are needed to develop temporal profiles of the training-induced responses in stroke patients at multiple levels spanning functional motor performance, neurophysiologic integrity, and imagery of cortical activation during relevant motor tasks. References 1. Liepert J, Bauder H, Miltner WH, Taub E, Weiller C. Treatmentinduced cortical reorganization after stroke in humans. Stroke 2000;31:1210-6. 2. Luft AR, Forrester LW, Macko RF, et al. Brain activation of lower extremity movement in chronically impaired stroke survivors. Neuroimage 2005;26:184-94. 3. Smith GV, Macko RF, Silver KH, Goldberg AP. Treadmill aerobic exercise improves quadriceps strength in patients with chronic hemiparesis following stroke: a preliminary report. J Neurol Rehabil 1998;12:111-7. 4. Visintin M, Barbeau H. The effects of body weight support on the locomotor pattern of spastic paretic patients. Can J Neurol Sci 1989;16:315-25.

233

5. Whitall J, McCombe-Waller S, Silver KH, Macko RF. Repetitive bilateral arm training with rhythmic auditory cueing improves motor function in chronic hemiparetic stroke. Stroke 2000;31: 2390-5. 6. Classen J, Liepert J, Wise SP, Hallett M, Cohen LG. Rapid plasticity of human cortical movement representation induced by practice. J Neurophysiol 1998;79:1117-23. 7. Caramia MD, Scalise A, Gordon R, Michalewski HJ, Starr A. Delayed facilitation of motor cortical excitability following repetitive finger movements. Clin Neurophysiol 2000;111:1654-60. 8. Gandevia SC, Petersen N, Butler JE, Taylor JL. Impaired response of human motoneurones to corticospinal stimulation after voluntary exercise. J Physiol 1999;521.3:749-59. 9. Rollnik JD, Schubert M, Albrecht J, Wohlfarth K, Dengler R. Effects of somatosensory input on central fatigue: a pilot study. Clin Neurophysiol 2000;111:1843-6. 10. Samii A, Wassermann EM, Hallett M. Post-exercise depression of motor evoked potentials as a function of exercise duration. Electroencephalogr Clin Neurophysiol 1997;105:352-6. 11. Zanette G, Bonato C, Polo A, Tinazzi M, Manganotti P, Fiaschi A. Long-lasting depression of motor-evoked potentials to transcranial magnetic stimulation following exercise. Exp Brain Res 1995; 107:80-6. 12. Dobkin BH, Firestine A, West M, Saremi K, Woods R. Ankle dorsiflexion as an fMRI paradigm to assay motor control for walking during rehabilitation. Neuroimage 2004;23:370-81. 13. Miyai I, Yagura H, Oda I, et al. Premotor cortex is involved in restoration of gait in stroke. Ann Neurol 2002;52:188-94. 14. Dobkin BH. An overview of treadmill locomotor training with partial body weight support: a neurophysiologically sound approach whose time has come for randomized clinical trials. Neurorehabil Neural Repair 1999;13:157-65. 15. Harris-Love ML, Forrester LW, Macko RF, Silver KH, Smith GV. Hemiparetic gait parameters in treadmill versus overground walking. Neurorehabil Neural Repair 2001;15:105-12. 16. Barbeau H, Visintin M. Optimal outcomes obtained with bodyweight support combined with treadmill training in stroke subjects. Arch Phys Med Rehabil 2003;84:1458-65. 17. Hesse S, Bertelt C, Schaffrin A, Malezic M, Mauritz KH. Restoration of gait in nonambulatory hemiparetic patients by treadmill training with partial body-weight support. Arch Phys Med Rehabil 1994;75:1087-93. 18. Harris-Love ML, Macko RF, Whitall J, Forrester LW. Improved hemiparetic muscle activation in treadmill versus overground walking. Neurorehabil Neural Repair 2004;18:154-60. 19. Silver KH, Macko RF, Forrester L, Goldberg AP, Smith GV. Effects of aerobic treadmill training on gait velocity, cadence and gait symmetry in chronic hemiparetic stroke: a preliminary report. Neurorehabil Neural Repair 2000;14:65-71. 20. Macko RF, Katzel LI, Yataco A, et al. Low-velocity graded treadmill stress testing in hemiparetic stroke patients. Stroke 1997; 28:988-92. 21. Macko RF, DeSouza CA, Tretter L, et al. Treadmill aerobic exercise training reduces energy expenditure and cardiovascular demands of hemiparetic gait in chronic stroke patients. Stroke 1997;28:326-30. 22. Macko RF, Smith GV, Dobrovolny CL, Sorkin J, Goldberg AP, Silver KH. Treadmill training improves fitness reserve in chronic hemiparetic stroke patients. Arch Phys Med Rehabil 2001;82: 879-84. 23. Koski L, Mernar TJ, Dobkin BH. Immediate and long-term changes in corticomotor output in response to rehabilitation: correlation with functional improvements in chronic stroke. Neurorehabil Neural Repair 2004;18:230-49. Arch Phys Med Rehabil Vol 87, February 2006

234

QUADRICEPS TMS AND TREADMILL IN STROKE, Forrester

24. Liepert J, Miltner WH, Bauder H, et al. Motor cortex plasticity during constraint-induced movement therapy in stroke patients. Neurosci Lett 1998;250:5-8. 25. Seitz RJ, Hoflich P, Binkofski F, Tellmann L, Herzog H, Freund HJ. Role of the premotor cortex in recovery from middle cerebral artery infarction. Arch Neurol 1998;55:1081-8. 26. Thickbroom GW, Byrnes ML, Archer SA, Mastaglia FL. Motor outcome after subcortical stroke: MEPs correlate with hand strength but not dexterity. Clin Neurophysiol 2002;113:2025-9. 27. Turton A, Lemon RN. The contribution of fast corticospinal input to the voluntary activation of proximal muscles in normal subjects and in stroke patients. Exp Brain Res 1999;129:559-72.

Arch Phys Med Rehabil Vol 87, February 2006

28. Cicinelli P, Traversa R, Rossini PM. Post-stroke reorganization of brain motor output to the hand: a 2-4 month follow-up with focal magnetic transcranial stimulation. Elctroencephalogr Clin Neurophysiol 1997;105:438-50. 29. Hummel F, Celnik P, Giraux P, et al. Effects of non-invasive cortical stimulation on skilled motor function in chronic stroke. Brain 2005;128(Pt 3):490-9. Suppliers a. Kin-Com 125 AP; Chattecx Corp, Chattanooga Group, 4717 Adams Rd, Hixson, TN 37343. b. Magstim Co Ltd, Spring Gardens, Whitland, Dyfed, SA34 0HR, UK.