Sage Journals: Discover world-class research

Abstract

Background. Stroke is a major cause of disability worldwide, and effective rehabilitation is crucial to regain skills for independent living. Recently, novel therapeutic approaches manipulating the excitatory-inhibitory balance of the motor cortex have been introduced to boost recovery after stroke. However, stroke-induced neurophysiological changes of the motor cortex may vary despite of similar clinical symptoms. Therefore, better understanding of excitability changes after stroke is essential when developing and targeting novel therapeutic approaches. Objective and Methods. We identified recovery-related alterations in motor cortex excitability after stroke using magnetoencephalography. Dynamics (suppression and rebound) of the ~20-Hz motor cortex rhythm were monitored during passive movement of the index finger in 23 stroke patients with upper limb paresis at acute phase, 1 month, and 1 year after stroke. Results. After stroke, the strength of the ~20-Hz rebound to stimulation of both impaired and healthy hand was decreased with respect to the controls in the affected (AH) and unaffected (UH) hemispheres, and increased during recovery. Importantly, the rebound strength was lower than that of the controls in the AH and UH also to healthy-hand stimulation despite of intact afferent input. In the AH, the rebound strength to impaired-hand stimulation correlated with hand motor recovery. Conclusions. Motor cortex excitability is increased bilaterally after stroke and decreases concomitantly with recovery. Motor cortex excitability changes are related to both alterations in local excitatory-inhibitory circuits and changes in afferent input. Fluent sensorimotor integration, which is closely coupled with excitability changes, seems to be a key factor for motor recovery.

Keywords

stroke recovery motor cortex excitability proprioception passive movement beta rhythm

Introduction

Approximately 75% of stroke survivors suffer from permanent disability; thus, stroke causes significant human suffering and poses a major economic burden on the society.¹ Recovery from stroke is based on brain’s plasticity. Studies in both animals and humans have shown that a period of enhanced plasticity occurs 1-4 weeks after stroke.^2-5 After this sensitive period, the effectiveness of poststroke rehabilitation diminishes dramatically. Recently, there have been promising attempts to prolong or enhance the sensitive period with pharmacological manipulations^6-8 or with noninvasive brain stimulation,^9,10 both aiming at changing the cortical excitation-inhibition balance. However, patients with initially similar clinical symptoms may recover differently, possibly because the underlying neurophysiological changes vary between these patients. Thus, understanding and monitoring recovery-related neurophysiological mechanisms and their temporal evolution is crucial for developing efficient, personalized rehabilitation.

Fluent upper limb motor function is important for independency in daily life. Integration of proprioceptive and tactile input with motor plans forms the basis of smooth and precise movements.¹¹ Afferent input mediates its effect on motor functions by modulating the motor cortex excitability.¹² Accordingly, our previous study in healthy subjects indicated that proprioceptive input strongly modulates the ~20-Hz motor cortex rhythm, causing an initial suppression followed by a strong and robust rebound.¹³ Prior studies have suggested that the ~20-Hz rebound reflects deactivation or inhibition of the motor cortex.^14
-17 Moreover, a combined magnetiencephalography (MEG) and magnetic resonance spectroscopy study showed that the ~20-Hz rebound strength is associated with the concentration of the inhibitory neurotransmitter GABA (γ-aminobutyric acid).¹⁸

To study alterations in motor cortex excitability after stroke and its association with motor recovery, we measured the dynamics of ~20-Hz motor cortex oscillations during passive movement of the index fingers in 23 stroke patients at the acute phase and during 1-year recovery. The motivation of this study was to understand the neurophysiological mechanisms underlying stroke recovery, which is instrumental for developing novel therapeutic interventions.

Methods

Subjects

We recruited 30 patients with first-ever stroke in the middle cerebral artery territory causing unilateral upper limb paresis (12 females, age 45-78 years, mean 67 ± 2 years) from the Department of Neurology, Helsinki University Hospital. The paresis of the upper limb was determined on the basis of the clinical examination by a neurologist at the time of the recruitment. The inclusion criteria required at least a sensory deficit and hand weakness or clumsiness. Twenty-two healthy control subjects (11 females, age 42-72 years, mean 59 ± 2.0 years; data presented in detail in our previous study¹³) were included. Exclusion criteria were earlier neurological diseases, mental disorders, neurosurgical operations, and unstable cardiovascular condition. After recruitment, 7 patients were excluded; 2 died, 4 declined the upcoming MEG sessions, and the data of 1 patient were contaminated with artifacts preventing reliable analyses. Twenty-three patients participated in all 3 follow-up measurements (10 females, age 45-78 years, mean 65 ± 2 years). The local ethics committee of the Helsinki and Uusimaa Hospital District approved the study protocol, and all subjects gave written informed consent prior to the measurements.

Clinical Evaluation

The patients underwent clinical examination in conjunction with the MEG recordings 1 to 7 days (T₀), 1 month (T₁), and 12 months (T₂) after stroke. Impairment caused by stroke was evaluated according to clinical neurological examination and the National Institutes of Health Stroke Scale (NIHSS; 0-42). Independency in daily life was scored with Barthel Index (BI; 0-100). Jamar Hydraulic Hand Dynamometer was used to measure the grip strength. Fine and gross manual dexterity was tested with Nine-Hole-Peg Board (NHPT; time to remove and replace 9 pegs into 9 holes, maximum 180 seconds) and Box and Block (BB; number of cubes moved from one compartment to another in 60 seconds) tests, respectively. Proprioception was tested qualitatively; the impaired hand was placed to a certain position and the ability of the patient to blindly mimic this position with the healthy hand was evaluated to be normal or abnormal.

Furthermore, we aimed to study whether the patients with different clinical outcomes (well-recovered vs poorly recovered patients) would show distinct changes in the modulation of the ~20-Hz rhythm. We attempted to split our patient group according to the “proportional recovery rule”¹⁹; however, only 2 patients reached the required 70% of the initial recovery potential. We also tried to compare the patients according to the lesion site (cortical vs cortico-subcortical vs subcortical lesions). Again, these subgroups were too small to draw reliable conclusions.

Neuroradiological Evaluation

Anatomical magnetic resonance images (MRIs) were acquired at T₀ and T₁ with a 3-T MRI scanner (Philips Achieva 3T, Philips Medical Systems, Best, the Netherlands). A diffusion-weighted sequence was used to detect ischemic lesions, and a T2-weighted 3D volume (1 × 1 × 1 mm³ voxels) was acquired for lesion size measurement at T₀ with the MRIcron software (McCausland Center for Brain Imaging, Columbia, SC, USA). The lesions were classified to be cortical when confined to the cortex and the immediate subcortical white matter without involvement of basal ganglia and/or capsula interna; subcortical if nucleus caudatus, putamen, globus pallidus, capsula interna and/or thalamus were involved; and cortico-subcortical if both cortical and subcortical areas were involved.²⁰

MEG Recordings

A 306-channel (204 planar gradiometers, 102 magnetometers) whole-scalp MEG system (Vectorview; Elekta Oy, Helsinki, Finland) was used for recordings. The data of 18 control subjects were recorded in Aalto University and four controls and all the patients with similar measurement and stimulus devices in the BioMag Laboratory (Helsinki University Hospital, Finland). During the MEG measurements, the subjects were either sitting or in supine position (4 patients at T₀) and instructed to relax, not to pay attention to the finger lift and to avoid excessive blinking.¹³

Four indicator coils, 3 anatomical landmarks (right and left preauricular points and nasion) and 50 to 100 additional points on the head surface were used for co-registration. The MEG and vertical electro-oculogram signals were pass-band filtered to 0.03 to 330 Hz and digitized at 1000 Hz. Online averaging was performed to monitor the number of accepted trials (~60 for each hand) while acquiring continuous data for analysis. Resting state data with eyes open and closed (3 minutes each) were recorded.

Passive Movement

The index finger was extended briskly once every 3 seconds by a trained nurse; in the patients first of the healthy and then of the impaired hand. To ensure minimal cutaneous tactile stimulation the middle phalanx was covered with a surgical tape to which a rigid aluminum stick was attached with a Velcro strap. The finger was not allowed to touch the device during the movement. A 3-axis accelerometer (ADXL335 iMEMS accelerometer Analog Devices Inc, Norwood, MA, USA) was attached on the nail of the index finger and its signals, acquired with the MEG system, were used to define the finger kinematics. The accelerometer signal was available only for 17 controls and 16 patients. Since the lag (time from actual movement to recorded movement onset) distributions were narrow (335 ± 8, 317 ± 13, and 326 ± 4 ms at T₀, T₁, and T₂, respectively), the average lag was used for the subjects with remaining accelerometer signals.

Timing and amplitude of the movements were kept constant by monitoring them with 2 optical gates separated by 30 mm along the direction of the movement (Figure 1); only movements passing through both gates within 500 ms were accepted as valid trials for online averaging and later for off-line analysis (~60 for each hand). The kinematics of the movements are reported in detail in the Results section.

Figure 1.

(A) Setup for passive movement. (B) Representative signals of 1 patient at T₀ (1-7 days), T₁ (1 month), and T₂ (12 months) after stroke. Two upper rows: Magnetoencephalography signals from a single gradiometer channel (raw and filtered to 15-25 Hz over the primary sensorimotor cortex. The ~20-Hz modulation is observable even to a single movement. Third row: Magnitude of acceleration. Total duration of the movement highlighted in gray.

Data Analysis

The raw data were preprocessed with the temporal signal-space separation (tSSS) method²¹ to suppress magnetic interference of nearby sources and compensated for head movement^22,23 using the MaxFilter software (version 2.2.11; Elekta Oy). The 204 planar gradiometer channels were used in subsequent analysis due to their focal sensitivity patterns.

Amplitude spectra were estimated from the resting-state data (eyes open) with the Welch method using 2048-sample Hanning-windowed segments, to define the peak amplitudes and frequencies of spontaneous brain activity. As the strongest peaks were observed in the 15- to 25-Hz range in both the controls and the patients, this band was used for further analysis in all subjects. Time-frequency representations (TFR) of passive-movement responses were calculated over all channels for the 3- to 40-Hz range with 7-cycle Morlet wavelets, to ensure the frequency range of the strongest modulation.

The temporal spectral evolution (TSE) method¹⁴ was used to quantify the modulation of rhythmic activity; the continuous data were filtered to 15 to 25 Hz, rectified, and averaged (−100 to 3000 ms) time-locked to stimulus onset.

Peak amplitudes of suppression and rebound over the rolandic area were quantified in both ipsi- and contralateral hemispheres with respect to the stimulated hand from 4 channels (2 from each hemisphere) showing the strongest suppression and rebound of 15- to 25-Hz activity. The relative peak amplitudes were calculated as amplitude changes with respect to the individual baseline (−100 to 0 ms) and defined as percentage values. Peak latencies were determined at the point of maximal suppression/rebound. To characterize the movement kinematics, the peak acceleration magnitude was computed for each successful movement, and averaged across the movements for each subject separately.

Statistical Analysis

The normality of the data was tested by using the Kolmogorov-Smirnov test; with 4 variables, the null hypothesis of a normal distribution could be rejected at P < .05. To make all variables normally distributed, we transformed the original values x into new values y = ln(x + 1) where ln(·) is the natural logarithm. After this transformation, the Kolmogorov-Smirnov test indicated normal distribution of all variables. The statistical analyses were performed on these transformed variables.

For the patients, the kinematics of passive movements and clinical test results between the impaired and healthy hands were compared using a 2-way (hands: impaired and healthy; times: T₀, T₁, T₂) repeated-measures analysis of variance (ANOVA). The kinematics of passive movements were compared between the patients and the controls (right and left hands pooled) with 1-way, 6-level (2 × hand; 3 × time) ANOVA. The TSE results from all sessions (T₀, T₁, and T₂) were studied in both hemispheres (AH; UH) to both healthy- and impaired-hand passive movement. The variance within factors time, hemispheres, and side of stimulation (contralateral/ipsilateral) was studied with a 2-way within-subject ANOVA. Significant (P < .05) main effects (F) were compared with paired-samples t tests. Independent-samples t tests were used when comparing statistically significant effects between the controls and the patients. Pearson’s parametric test was used for correlation analyses (P < .05).

Results

Neuroradiological Findings and Clinical Outcome

MRIs revealed cortical strokes in 2, subcortical strokes in 6, and cortico-subcortical strokes in 15 patients. The lesion volumes varied from 0.24 to 220 cm³ (mean 40 ± 12 cm³; Supplementary material, Table S1).

Table 1 shows the mean (±SEM) scores of the clinical tests of all the patients. At T₀, all clinical scores of the impaired hand were significantly worse (P < .001) than those of the healthy hand. During the follow-up, the impaired hand function improved significantly; BB improved from T₀ to T₁ (P < .001) and from T₁ to T₂ (P < .01). NHPT improved from T₀ to T₁ (P < .05), but not significantly from T₁ to T₂. The impaired hand function did not reach the level of the healthy hand by T₂ (P < .001). No significant improvement of Jamar scores was detected.

Table 1.

Clinical Scores of the Patients.^a

Time	Jamar		NHPT		Box and Block		NIHSS	BI
Time	Impaired Hand	Healthy Hand	Impaired Hand	Healthy Hand	Impaired Hand	Healthy Hand	NIHSS	BI
T₀	16 ± 4	31 ± 3	104 ± 15	27 ± 2	22 ± 5	45 ± 3	6 ± 1	59 ± 8
T₁	18 ± 3	31 ± 2	92 ± 16	24 ± 1	32 ± 5	54 ± 2	4 ± 1	80 ± 6
T₂	19 ± 3	30 ± 2	82 ± 15	24 ± 1	36 ± 5	56 ± 2	2 ± 0	92 ± 3
Reference values	25 ± 7.8 (females, age >50 y)		21 ± 9 (females, age 65-69 y)		72 ± 22 (females, age 65-69 y)		0	100
	38 ± 7.5 (males, age >50 y)		22 ± 14 (males, age 65-69 y)		71 ± 32 (males, age 65-69 y)

Abbreviations: Jamar, hydraulic handgrip dynamometer (kg); NHPT, 9-hole-peg time (seconds); Box and Block, number of blocks; NIHSS, National Institutes of Health Stroke Scale (0-42); BI, Barthel Index (0-100).

T₀ = 1-7 days; T₁ = 1 month; T₂ = 12 months after stroke. Reference values in healthy people. Scores of the impaired hand that differ significantly (P < .001) from those of the healthy hand are in boldface.

Also, the healthy hand function improved during follow-up (Table 1). However, BB scores of the healthy hand were clearly lower than the reference values at all time points. Both BB and NHPT scores improved significantly from T₀ to T₁ (P < .001 and P < .01, respectively). The Jamar scores of the healthy hand did not differ from the reference values at any time point.

The overall clinical scores (NIHSS, BI) improved from T₀ to T₁ (P < .001 and P < .001, respectively) and from T₁ to T₂ (P < .001 and P < .01, respectively).

The proprioceptive sense of the impaired hand was normal in 5 of 23 patients at T₀, in 8 of 23 patients at T₁, and in 11 of 23 patients at T₂.

Kinematics of Passive Movements

In the patients, the peak acceleration of the index finger did not differ significantly between the healthy and the impaired hand at any time point: T₀ (healthy 0.39 ± 0.12 g, impaired 0.38 ± 0.13 g, P = .52), T₁ (healthy 0.45 ± 0.12 g, impaired 0.39 ± 0.12 g, P = .07), or T₂ (healthy 0.57 ± 0.09 g, impaired 0.55 ± 0.09 g, P = .25). Neither were differences within one hand observed between T₀ and T₁. However, the passive movements of both the healthy and impaired hand were brisker at T₂ than at T₀ (P < .001) and at T₁ (P < .01).

In the controls, the peak acceleration of the passive movement did not differ between the right and left index fingers (right: 0.40 ± 0.10 g and left: 0.40 ± 0.06 g, N = 17; P = .96).

In the patients versus controls, no significant differences in the peak acceleration were found at T₀ or T₁. However, at T₂, the peak acceleration in the patients for both the healthy and impaired hand was brisker than in the controls (P < .001).

The movement durations were significantly shorter (P < .01) in the patients versus controls in all measurement sessions. However, the movement durations of the healthy and affected hands of the patients did not differ between T₀ and T₁.

Spontaneous Brain Activity

In the controls, the amplitude spectra during rest revealed strongest peaks around 20-Hz (beta) at 15.7 ± 0.4 Hz (β₁) and 21.8 ± 0.5 Hz (β₂) over the right rolandic area and at 15.3 ± 0.4 Hz (β₁) and 21.5 ± 0.5 Hz (β₂) over the left rolandic area.

In the patients, strongest beta peaks at T₀ were observed in AH at 16.0 ± 0.4 Hz (β₁) and at 20.9 ± 0.7 Hz (β₂) and in the UH at 15.3 ± 0.2 Hz (β₁) and at 19.3 ± 0.6 Hz (β₂). The strength of β₁- or β₂-peaks did not differ significantly between the hemispheres, the time points, or between the patients and controls (Table 2A).

Table 2.

Characteristics of the ~20-Hz Rhythm.^a

A: Amplitudes of strongest peaks (fT/cm) of spontaneous activity at ~20 Hz
	Patients (N = 23)
Hemisphere	AH						UH						Controls (N = 22)
Time	T₀		T₁		T₂		T₀		T₁		T₂		Right	Left
β₁	11 ± 1.4		10 ± 1.1		12 ± 0.4		12 ± 1.4		12 ± 1.3		12 ± 0.9		12 ± 1.9	13 ± 2.2
β₂	8.5 ± 0.9		8.4 ± 0.9		8.2 ± 0.7		9.7 ± 1.2		10 ± 1.0		9.7 ± 0.8		8.7 ± 0.9	8.9 ± 0.8
B: Relative suppression and rebound amplitudes (%)
	Patients
	AH-impaired			UH-impaired			AH-healthy			UH-healthy			Controls
Time	T₀	T₁	T₂	T₀	T₁	T₂	T₀	T₁	T₂	T₀	T₁	T₂	Contra H	Ipsi H
Suppression	27 ± 2.5	27 ± 2.4	33 ± 2.9	24 ± 2.1	25 ± 1.7	28 ± 2.1	26 ± 1.9	26 ± 1.9	29 ± 2.0	39 ± 2.0	38 ± 2.4	42 ± 2.3	34 ± 1.6	31 ± 1.5
Rebound	24 ± 4.3	39 ± 6.0	44 ± 5.0	18 ± 3.8	26 ± 2.7	36 ± 4.0	33 ± 4.3	42 ± 4.5	42 ± 4.3	48 ± 4.4	59 ± 6.6	66 ± 6.5	97 ± 9.3	54 ± 5.2

Abbreviations: AH, affected hemisphere; UH, unaffected hemisphere; AH-impaired, affected hemisphere, impaired-hand stimulation; UH-impaired, unaffected hemisphere, impaired-hand stimulation; AH-healthy, affected hemisphere, healthy-hand stimulation; UH-healthy, unaffected hemisphere, healthy-hand stimulation; Contra H, hemisphere contralateral to stimulation; Ipsi H, hemisphere ipsilateral to stimulation.

T₀ = 1-7 days; T₁ = 1 month; T₂ = 12 months after stroke. Values that differ significantly from the controls are in boldface.

Modulation of the ~20-Hz Rhythm

Congruently with earlier studies,^15,24 the ~20-Hz rhythm was bilaterally modulated to unilateral stimulation. The maximum ~20-Hz rebounds were observed over the same rolandic channels as the strongest beta peaks in the amplitude spectra and appeared anterior to the maximal suppression on the planar gradiometers.

In the controls, no differences between suppression and rebound strength were observed between the left and right hemispheres, in line with our earlier studies,^5,13 hence, the responses of the left and right hemispheres were pooled together for further analyses.

In the following, we use the nomenclature AH-impaired (affected hemisphere, contralateral impaired-hand stimulation), AH-healthy (affected hemisphere, ipsilateral healthy-hand stimulation), UH-healthy (unaffected hemisphere, contralateral healthy-hand stimulation) and UH-impaired (unaffected hemisphere, ipsilateral impaired-hand stimulation).

Figure 2 shows the grand-average TSE of the ~20-Hz rhythm (from the maximum channel of each subject) to passive movement in the patients and the controls. After the stimulus, the rhythm is briefly suppressed, maximally at 530 ± 10 ms (patients) and 540 ± 20 ms (controls), and subsequently increased, peaking at 1370 ± 30 ms (patients) and 1450 ± 30 ms (controls). The peak latencies of suppression and rebound did not differ between the hemispheres, the time points or between the patients and controls

Figure 2.

Modulation of the ~20-Hz oscillation. Grand average of the alterations of the ~20-Hz rhythm to passive movement of (A) impaired hand of the stroke patients (N = 23) in the contralateral affected (AH-impaired) and ipsilateral unaffected (UH-impaired) hemispheres compared to the responses of the controls (N = 22, hemispheres pooled). (B) Healthy hand in the ipsilateral affected (AH-healthy) and unaffected contralateral (UH-healthy) hemispheres, both at T₀ (1-7 days), T₁ (1 month), and T₂ (12 months) after stroke.

The mean amplitudes of suppression and rebound of the patients and the controls are presented in Table 2B. No significant differences in the strength of the suppression were detected between the hemispheres of the patients, between different time points or between the patients and the controls. In contrast, Figure 2 shows that in the patients, the rebound strength increases from the acute phase during recovery in both hemispheres.

ANOVA showed significant main effects for the factors time, F(2, 44) = 13.9, P < .000; hemisphere, F(1, 22) = 2.2, P < .000, partial η² = 0.090; and side of stimulation, F(1, 22) = 29.0, P < .000, partial η² = 0.569; for the rebound strength. A significant interaction was found between the factors hemisphere and side of stimulation, F(1, 22) = 23.9, P < .001.

Impaired-Hand Stimulation

At T₀, no rebound was detected in AH in 6 patients, and at T₁, in 2 patients. Figure 3 shows that the AH-impaired rebound increased significantly from T₀ to T₁ and to T₂ (P < .01 and P < .001, respectively). No significant increase from T₁ to T₂ was detected. The AH-impaired rebounds of the patients were significantly weaker than those of the controls at all time points (P < .001 for T₀, T₁, and T₂).

Figure 3.

Rebound strength. Rebound amplitudes (%) in the controls (N = 22, hemispheres pooled) and the patients (N = 23) in the affected (AH) and unaffected hemispheres (UH) to the stimulation of the impaired hand and healthy hands. AH-impaired, affected hemisphere, impaired-hand stimulation; UH-impaired, unaffected hemisphere, impaired-hand stimulation; AH-healthy, affected hemisphere, healthy-hand stimulation; UH-healthy, unaffected hemisphere, healthy-hand stimulation; CTRL contra, controls, contralateral stimulation; CTRL ipsi, controls, ipsilateral stimulation. T₀ (1-7 days), T₁ (1 month), and T₂ (12 moths) after stroke. *P < .05; **P < .01; ***P < .001.

At T₀, no rebound was found in the UH in 7 patients. The UH-impaired rebound increased significantly from T₀ to T₁ and T₂ (P < .01 and P < .001, respectively) and from T₁ to T₂ (P < .01). The UH-impaired rebound was significantly weaker than that of the controls at all time points (P < .001 for T₀, P < .01 for T₁, and P < .05 for T₂).

The UH-impaired rebound was significantly weaker than the UH-healthy rebound (P < .001 for T₀, P < .001 for T₁, and P < .001 for T₂).

Healthy-Hand Stimulation

For healthy-hand stimulation, no AH-healthy rebound was detected in 2 patients at T₀. Figure 3 shows that the AH-healthy rebound approached, but did not achieve significance from T₀ to T₁ and T₂ (P = .05); yet it was significantly weaker than that of the controls at T₀ (P < .05).

The AH-healthy rebound was significantly stronger than the AH-impaired rebound at T₀ (P < .05). At T₁ and T₂, such difference was not observed.

The UH-healthy rebound increased significantly from T₀ to T₂ (P < .01) but not from T₀ to T₁ or from T₁ to T₂. The UH-healthy rebound was significantly weaker than that of the controls in all sessions (P < .001 for T₀, P < .01 for T₁, and P < .05 for T₂).

To summarize, the AH rebound was decreased compared with that of the controls at T₀ and increased during follow-up, not only when the impaired (AH-impaired) but also when the healthy hand was stimulated (AH-healthy). The AH-healthy rebound was decreased only at the acute phase, whereas the AH-impaired rebound did not reach the level of the controls during follow-up. The UH rebound to both healthy and impaired hand stimulation was also diminished at the acute phase and increased during follow-up but did not reach the level of the controls.

Correlation of Rebound Amplitude With Clinical Outcome

Impaired-Hand Stimulation

Since only 15 of 23 patients were able to perform NHPT while 19 of 23 patients obtained above-zero BB scores at T₂, we chose BB for correlating the rebound amplitude with clinical outcome. Figure 4 shows that the AH-impaired rebound amplitude correlated with results of BB at all time points (r = 0.6, P < .01 at T₀; r = 0.7, P < .001 at T₁; and r = 0.6, P < .01 at T₂), the stronger the rebound the better the BB score. The rebound strength correlated also with NHPT results of the impaired hand; the stronger the rebound the faster the patient performed the test at T₀ (r = −0.5, P < .05), at T₁ (r = −0.7, P < .001) and at T₂ (r = −0.5, P < .05). In addition, the rebound amplitude at T₀ correlated with the BB scores achieved at T₁ (r = 0.6, P < .001) and at T₂ (r = 0.7, P < .001).

Figure 4.

Rebound strength and clinical correlation. Linear correlation of the rebound amplitude (%) in the affected hemisphere to the contralateral stimulation of the impaired hand, and result of Box and Block test of the impaired hand. T₀ (1-7 days), T₁ (1 month), and T₂ (12 months) after stroke.

The AH-healthy rebound strength did not correlate with any of the hand function tests.

Healthy-Hand Stimulation

The UH-healthy rebound strength correlated with BB scores of the healthy hand at T₀ (r = 0.04, P < .05) and at T₁ (r = 0.5, P < .05) and with NHPT scores of the healthy hand at T₁ (r = −0.4, P < .05) and at T₂ (r = −0.6, P < .01).

Correlation With the Lesion Volume

The rebound strength did not correlate with the lesion volume. However, the overall clinical scores correlated with the lesion volume: NIHSS (r = 0.5, P < .05 at T₀; r = 0.4, P < .05 at T₁; and r = 0.6, P < .001 at T₂) and BI (r = −0.4, p < .05; r = −0.4, P < .05 and r = 0.5, P < .05, at T₀, T₁, and T₂, respectively). NHPT and BB of the impaired hand correlated with the lesion volume only at T₁ (r = 0.4, P < .05 and r = −0.4, P < .05, respectively).

Discussion

The present longitudinal study shows that the rebound amplitudes of the ~20-Hz motor cortex rhythm to passive movement are decreased compared with the controls in both hemispheres at the acute phase after stroke. The rebound amplitudes increase to some extent during follow-up but only the rebound strength in the AH to healthy-hand stimulation reached the level of the controls, indicating long-lasting excitability changes in both hemispheres after stroke.

Rebound and Motor-Cortex Excitability

Most animal studies have suggested stroke-triggered hyperexcitability in the AH that promotes experience-dependent plasticity, essential for functional recovery after stroke.^25-28 A recent study in mice showed that after subcortical infarction, induced by an occlusion of the middle cerebral artery, tonic GABAergic inhibition was significantly decreased in the peri-infarct zone, likely via glutamate-driven excitoxicity.²⁹ In contrast, another recent study indicated increased tonic GABAergic inhibition in the peri-infarct zone in mice.⁸ Counteracting this overinhibition by pharmacological means led to better motor recovery, underlining the close linkage between excitability changes and motor recovery after stroke. Despite these controversial observations, both studies indicate that an initially increased excitability is essential for motor recovery. In our stroke patients, we observed decreased inhibition (hyperexcitability) at the acute phase after stroke, congruently with most of the earlier studies.^29-32 During recovery, the clinical parameters of hand motor function improved in conjunction with increased inhibition (increase in the rebound strength) of the motor cortex. This is in line with earlier studies indicating that although disinhibitory changes may be necessary for functional recovery,^33,34 a normalization of excitability is associated with good recovery after stroke.^5,35,36

In the present study, excitability changes were studied indirectly by recording the modulation of the ~20-Hz rhythm, which has been shown to be associated with changes in the cortical excitation-inhibition balance.^5,13 The ~20-Hz oscillations are suppressed during movement or after somatosensory stimulation, and increase transiently (rebound) after cessation of the movement or stimulation.^14,15,17

Earlier findings have shown that the rebound and suppression have different generator areas,^37,38 frequencies,³⁹ and functional roles.^13,40 Several studies have indicated that the ~20-Hz rebound reflects active inhibition or removal of excitation of the motor cortex,^18,40,41 whereas the functional significance of the suppression has remained unclear. Accordingly, pharmacological studies have proposed that although both suppression and rebound seem to be GABA-dependent processes, their effects are mediated via different receptor subtypes; the suppression via fast ionotropic GABA_A and the rebound via both GABA_A and slower metabotropic GABA_B receptor.^42-44

In the present study, the ~20-Hz rebound in the patients was altered throughout the follow-up period but no changes were detected in the suppression strength. Although we cannot evaluate the exact nature of the inhibitory processes underlying the ~20-Hz rebound, our result advocates the notion of distinct neuronal populations underlying suppression and rebound.

Rebound in the AH

The strength of the AH rebound to stimulation of the impaired hand (AH-impaired) correlated significantly with functional recovery, but it did not reach the level of the controls during follow-up. Earlier studies have shown that afferent input modulates motor cortex excitability,^12,45 and it is likely that altered afferent input strongly affects the AH-impaired rebound.

A prior study in monkeys indicated that cortico-cortical projections from the somatosensory cortex to the motor cortex are crucial in learning new motor skills but not in the execution of existing skills.¹¹ Accordingly, a study using repetitive transcranial magnetic stimulation (rTMS) showed that disturbing somatosensation during task training impaired motor learning.⁴⁶ Our finding, the correlation of the AH-impaired rebound strength with motor recovery, suggests that fluent integration of afferent input with motor functions is a key element in relearning motor skills. Modulatory afferent input may even be on its own an important driver of cortical remapping and motor recovery.⁴⁷

Afferent input may reach the motor cortex via different pathways. Our earlier study in stroke patients showed that the AH-impaired rebound strength did not correlate with the strength of the primary somatosensory cortex activation, but did correlate with the activation strength of the secondary somatosensory (SII) information with motor functions.⁵ This is in line with earlier animal and human studies implicating the importance of SII areas in the sensorimotor integration.^4,48,49

The strength of the AH rebound to healthy-hand stimulation (AH-healthy) was decreased compared with the controls at the acute phase after stroke but reached the level of the controls by 1 month. As the afferent input is assumed to be intact in this case, the decrease of rebound to healthy-hand stimulation likely reflects stroke-induced changes in local excitatory-inhibitory circuits rather than changes in afferent input. In contrast to the AH-impaired rebound, the AH-healthy rebound did not correlate with any of the clinical parameters.

As stroke-induced changes in the local excitatory-inhibitory circuits affect both hemispheres, it is likely that the strength of AH-impaired rebound comprises the combined effects of deficient afferent input and sensorimotor integration and changes in local cortical excitability and effects. The exact role and functional significance of all these distinct influences on motor cortex excitability still remain unclear. Nevertheless, as only the AH-impaired rebound correlated with motor recovery, the results indicate the important role of modulatory afferent input during motor recovery.

Rebound in the UH

At the acute phase, the UH-healthy rebound was diminished compared to the controls despite of intact afferent input, indicating increased cortical excitability in the UH. During the follow-up, the amplitude of the UH-healthy rebound increased but it did not reach the level of the controls by 1 year, implicating long-lasting excitability changes of the UH in accordance with earlier findings.^50,51 Earlier TMS studies in stroke patients have suggested that hyperexcitability of the UH may hamper functional recovery.^33,50 Inhibition of the hyperexcited UH with rTMS may release the AH from excessive interhemispheric inhibition and could thus improve functional recovery.¹⁰

In our patients, the impaired hand function did not reach the level of the healthy hand by 1 year. The poor recovery of hand function in conjunction with long-lasting excitability changes of the UH might imply a harmful effect of the hyperexcited UH congruently with earlier studies in stroke patients.^33,50,51 However, it is possible that hyperexcitation of the UH is not only harmful, but may provide at least some degree of regained function in poorly recovering patients. However, our data are not sufficient to address this important issue, which needs to be investigated in future studies.

Effect of Stimulus Kinematics on Rebound

In the current study, passive movements were performed manually; therefore, the stimuli may have varied slightly in terms of peak acceleration and movement duration. The movements were all brisk, but significantly more so in the last session (T₂). However, the acceleration did not differ between the first 2 sessions (T₀ and T₁) where the strongest recovery of rebound amplitude occurred. In addition, at 1 year (T₂), the movements were brisker and longer for the patients than for the control subjects but still the rebound amplitude remained weaker in the patients; thus, it is unlikely that the observed changes in the rebound could be explained by changes in stimulus kinematics. Yet, we suggest that the repeatability of passive-movement responses should be addressed in future studies. Recently introduced robust MEG-compatible movement actuators, using computer-controlled pneumatic artificial muscles with millisecond accuracy,⁵² are thus recommended for clinical studies in the future.

Perspectives to Development of Stroke Rehabilitation

Our findings of altered motor cortex rebound in both hemispheres agree with earlier studies indicating that hyperexcitability occurs predominantly 1 to 4 weeks after stroke.^4,5,53,54 Recently, there have been attempts to enhance plasticity after stroke using pharmacological manipulations with antidepressive agents such as serotonin-reuptake inhibitors (SSRIs).^6,7 In humans, a randomized clinical study indicated less severe residual motor deficits in patients treated with SSRI during the first 3 months after stroke.⁶ Animal studies have suggested that also GABAergic drugs^8,29 could reopen or enhance the sensitive period through reducing GABAergic inhibition.¹¹ In addition to pharmacological manipulations, noninvasive brain stimulation, such as rTMS, has been applied to boost plasticity but the success has been variable.⁵⁵

As these therapeutic approaches are aiming at changing cortical excitability, it is of utmost important to understand and monitor changes in excitability. The present results suggest that neurophysiological methods such as MEG and EEG could be useful in detecting alterations in cortical excitability to monitor the effect of interventions at different stages (acute, subacute, and chronic) of recovery. Earlier TMS studies have shown that motor cortex excitability depends on the lesion site.^56,57 Because of the small number of patients, we could not reliably compare patients having cortical, cortico-subcortical, or subcortical lesions. However, despite this heterogeneous sample, our results clearly show that the rebound strength increased throughout follow-up in conjunction with clinical outcome.

The AH-impaired rebound amplitude at the acute phase correlated with motor skills (BB scores) achieved 1 month and 12 months after stroke. Although our data are not sufficient to make conclusions about recovery potential, this relationship suggests that a diminished rebound at the acute phase may predict poor motor recovery. Further studies, both in healthy subjects and stroke patients, aiming at monitoring changes in the ~20-Hz rebound during interventions, are needed to address the usability and sensitivity of the ~20-Hz rebound as a neurophysiological marker of motor cortex recovery after stroke.

Conclusions

The present results indicate that the strength of the rebound in the AH to impaired-hand stimulation is associated with recovery of hand motor function after stroke. The close coupling of afferent input with excitability changes seems to be a key factor for motor recovery, and needs to be paid attention to when planning therapeutic interventions aiming at changing cortical excitability. Since the strength of the rebound is likely associated with stroke-induced excitability changes, it could be used to guide therapeutic interventions for optimal recovery. Future studies in a larger patient group and in patients matched for their impairment caused by stroke should aim at monitoring the proposed neurophysiological markers to assess their value in stroke rehabilitation.

Footnotes

Acknowledgements

We thank Mia Illman for help with all the MEG recordings, Jyrki Mäkelä for supporting our MEG measurements in Biomag Laboratory (Helsinki University Hospital), and Suvi Heikkilä for the assistance in those recordings. We also thank the Helsinki University Hospital occupational therapists for performing the clinical hand motor tests.

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The study was financially supported by Helsinki University Hospital Research Fund, SalWe Research Program for Mind and Body (Tekes; Finnish Funding Agency for technology and Innovation, Grants No. 1104/10 and 1988/31/2015), Finnish Medical Foundation, and Academy of Finland Postdoctoral Research Grant No. 266133.

Supplementary material

Supplementary material for this article is available on the Neuro-rehabilitation & Neural Repair website at .

References

Mozaffarian

Benjamin

Heart disease and stroke statistics—2015 update: a report from the American Heart Association. Circulation. 2015;131:e29-e322.

Nudo

Wise

SiFuentes

Milliken

GW.

Neural substrates for the effects of rehabilitative training on motor recovery after ischemic infarct. Science. 1996;272:1791-1794.

Barbay

Plautz

Friel

Behavioral and neurophysiological effects of delayed training following a small ischemic infarct in primary motor cortex of squirrel monkeys. Exp Brain Res. 2006;169:106-116.

Forss

Mustanoja

Roiha

Activation in parietal operculum parallels motor recovery in stroke. Hum Brain Mapp. 2012;33:534-541.

Laaksonen

Kirveskari

Mäkelä

Effect of afferent input on motor cortex excitability during stroke recovery. Clin Neurophysiol. 2012;123:2429-2436.

Chollet

Tardy

Albucher

J-F

Fluoxetine for motor recovery after acute ischaemic stroke (FLAME): a randomised placebo-controlled trial. Lancet Neurol. 2011;10:123-130. Erratum in: Lancet Neurol. 2011;10:205.

Maya Vetencourt

Sale

Viegi

The antidepressant fluoxetine restores plasticity in the adult visual cortex. Science. 2008;320:385-388.

Clarkson

Huang

Macisaac

Mody

Carmichael

ST.

Reducing excessive GABA-mediated tonic inhibition promotes functional recovery after stroke. Nature. 2010;468:305-309.

Nitsche

Paulus

Sustained excitability elevations induced by transcranial DC motor cortex stimulation in humans. Neurology. 2001;57:1899-1901.

10.

Fregni

Boggio

Mansur

Transcranial direct current stimulation of the unaffected hemisphere in stroke patients. Neuroreport. 2005;16:1551-1555.

11.

Pavlides

Miyashita

Asanuma

Projection from the sensory to the motor cortex is important in learning motor skills in the monkey. J Neurophysiol. 1993;70:733-741.

12.

Favorov

Sakamoto

Asanuma

Functional role of corticoperipheral loop circuits during voluntary movements in the monkey: a preferential bias theory. J Neurosci. 1988;8:3266-3277.

13.

Parkkonen

Laaksonen

Piitulainen

Parkkonen

Forss

Modulation of the ~20-Hz motor-cortex rhythm to passive movement and tactile stimulation. Brain Behav. 2015;5:e00328.

14.

Salmelin

Hari

Spatiotemporal characteristics of sensorimotor neuromagnetic rhythms related to thumb movement. Neuroscience. 1994;60:537-550.

15.

Salenius

Schnitzler

Salmelin

Jousmäki

Hari

Modulation of human cortical rolandic rhythms during natural sensorimotor tasks. Neuroimage. 1997;5:221-228.

16.

Pfurtscheller

Central beta rhythm during sensorimotor activities in man. Electroencephalogr Clin Neurophysiol. 1981;51:253-264.

17.

Hari

Salmelin

Mäkelä

Salenius

Helle

Magnetoencephalographic cortical rhythms. Int J Psychophysiol. 1997;26:51-62.

18.

Gaetz

Edgar

Wang

Roberts

TPL

. Relating MEG measured motor cortical oscillations to resting γ-aminobutyric acid (GABA) concentration. Neuroimage. 2011;55:616-621.

19.

Prabhakaran

Zarahn

Riley

Inter-individual variability in the capacity for motor recovery after ischemic stroke. Neurorehabil Neural Repair. 2008;22:64-71.

20.

Dromerick

Reding

MJ.

Functional outcome for patients with hemiparesis, hemihypesthesia, and hemianopsia. Does lesion location matter?

Stroke. 1995;26:2023-2026.

21.

Taulu

Simola

Spatiotemporal signal space separation method for rejecting nearby interference in MEG measurements. Phys Med Biol. 2006;51:1759-1768.

22.

Nenonen

Nurminen

Kičić

Validation of head movement correction and spatiotemporal signal space separation in magnetoencephalography. Clin Neurophysiol. 2012;123:2180-2191.

23.

Taulu

Kajola

Presentation of electromagnetic multichannel data: the signal space separation method. J Appl Phys. 2005;97:124905.

24.

Forss

Jousmäki

Hari

Interaction between afferent input from fingers in human somatosensory cortex. Brain Res. 1995;685:68-76.

25.

Domann

Hagemann

Kraemer

Freund

Witte

OW.

Electrophysiological changes in the surrounding brain tissue of photochemically induced cortical infarcts in the rat. Neurosci Lett. 1993;155:69-72.

26.

Buchkremer-Ratzmann

Witte

OW.

Extended brain disinhibition following small photothrombotic lesions in rat frontal cortex. Neuroreport. 1997;8:519-522.

27.

Ward

Brown

Thompson

Frackowiak

RS.

Neural correlates of motor recovery after stroke: a longitudinal fMRI study. Brain J. Neurol. 2003;126:2476-2496.

28.

Moon

S-K

Alaverdashvili

Cross

Whishaw

IQ.

Both compensation and recovery of skilled reaching following small photothrombotic stroke to motor cortex in the rat. Exp Neurol. 2009;218:145-153.

29.

Jaenisch

Liebmann

Guenther

Hübner

Frahm

Witte

OW.

Reduced tonic inhibition after stroke promotes motor performance and epileptic seizures. Sci Rep. 2016;6:26173.

30.

Witte

OW.

Lesion-induced plasticity as a potential mechanism for recovery and rehabilitative training. Curr Opin Neurol. 1998;11:655-662.

31.

Witte

Bidmon

Schiene

Redecker

Hagemann

Functional differentiation of multiple perilesional zones after focal cerebral ischemia. J Cereb Blood Flow Metab. 2000;20:1149-1165.

32.

Sieber

Guenther

Jaenisch

Age-specific transcriptional response to stroke. Neurobiol Aging. 2014;35:1744-1754.

33.

Liepert

Hamzei

Weiller

Motor cortex disinhibition of the unaffected hemisphere after acute stroke. Muscle Nerve. 2000;23:1761-1763.

34.

Buetefisch

CM.

Role of the contralesional hemisphere in post-stroke recovery of upper extremity motor function. Front. Neurol. 2015;6:214. http://www.ncbi.nlm.nih.gov/pmc/articles/PMC4607877/. Accessed June 28, 2016.

35.

Calautti

Leroy

Guincestre

Marié

Baron

JC.

Sequential activation brain mapping after subcortical stroke: changes in hemispheric balance and recovery. Neuroreport. 2001;12:3883-3886.

36.

Swayne

Rothwell

Ward

Greenwood

RJ.

Stages of motor output reorganization after hemispheric stroke suggested by longitudinal studies of cortical physiology. Cereb Cortex. 2008;18:1909-1922.

37.

Salmelin

Forss

Knuutila

Hari

Bilateral activation of the human somatomotor cortex by distal hand movements. Electroencephalogr Clin Neurophysiol. 1995;95:444-452.

38.

Jurkiewicz

Gaetz

Bostan

Cheyne

Post-movement beta rebound is generated in motor cortex: evidence from neuromagnetic recordings. Neuroimage. 2006;32:1281-1289.

39.

Pfurtscheller

Stancák

Jr Edlinger

On the existence of different types of central beta rhythms below 30 Hz. Electroencephalogr Clin Neurophysiol. 1997;102:316-325.

40.

Salmelin

Hämäläinen

Kajola

Hari

Functional segregation of movement-related rhythmic activity in the human brain. Neuroimage. 1995;2:237-243.

41.

Pfurtscheller

Event-related synchronization (ERS): an electrophysiological correlate of cortical areas at rest. Electroencephalogr Clin Neurophysiol. 1992;83:62-69.

42.

Hall

Stanford

Yamawaki

The role of GABAergic modulation in motor function related neuronal network activity. Neuroimage. 2011;56:1506-1510.

43.

Muthukumaraswamy

Myers

Wilson

The effects of elevated endogenous GABA levels on movement-related network oscillations. Neuroimage. 2013;66:36-41.

44.

Rossiter

Davis

Clark

Boudrias

M-H

Ward

NS.

Beta oscillations reflect changes in motor cortex inhibition in healthy ageing. Neuroimage. 2014;91:360-365.

45.

Asanuma

Larsen

Zarzecki

Peripheral input pathways projecting to the motor cortex in the cat. Brain Res. 1979;172:197-208.

46.

Vidoni

Acerra

Dao

Meehan

Boyd

LA.

Role of the primary somatosensory cortex in motor learning: an rTMS study. Neurobiol Learn Mem. 2010;93:532-539.

47.

Winship

Murphy

TH.

Remapping the somatosensory cortex after stroke: insight from imaging the synapse to network. Neuroscientist. 2009;15:507-524.

48.

Forss

Jousmäki

Sensorimotor integration in human primary and secondary somatosensory cortices. Brain Res. 1998;781:259-267.

49.

Hoffer

Hoover

Alloway

KD.

Sensorimotor corticocortical projections from rat barrel cortex have an anisotropic organization that facilitates integration of inputs from whiskers in the same row. J Comp Neurol. 2003;466:525-544.

50.

Manganotti

Patuzzo

Cortese

Palermo

Smania

Fiaschi

Motor disinhibition in affected and unaffected hemisphere in the early period of recovery after stroke. Clin Neurophysiol. 2002;113:936-943.

51.

Liepert

Graef

Uhde

Leidner

Weiller

Training-induced changes of motor cortex representations in stroke patients. Acta Neurol Scand. 2000;101:321-326.

52.

Piitulainen

Bourguignon

Hari

Jousmäki

MEG-compatible pneumatic stimulator to elicit passive finger and toe movements. Neuroimage. 2015;112:310-317.

53.

Nudo

Milliken

GW.

Reorganization of movement representations in primary motor cortex following focal ischemic infarcts in adult squirrel monkeys. J Neurophysiol. 1996;75:2144-2149.

54.

Rossini

Caltagirone

Castriota-Scanderbeg

Hand motor cortical area reorganization in stroke: a study with fMRI, MEG and TCS maps. Neuroreport. 1998;9:2141-2146.

55.

Raffin

Siebner

HR.

Transcranial brain stimulation to promote functional recovery after stroke. Curr Opin Neurol. 2014;27:54-60.

56.

Liepert

Restemeyer

Kucinski

Zittel

Weiller

Motor strokes: the lesion location determines motor excitability changes. Stroke. 2005;36:2648-2653.

57.

Liepert

Restemeyer

Münchau

Weiller

Motor cortex excitability after thalamic infarction. Clin Neurophysiol. 2005;116:1621-1627.

Strength of ~20-Hz Rebound and Motor Recovery After Stroke

Abstract

Keywords

Introduction

Methods

Subjects

Clinical Evaluation

Neuroradiological Evaluation

MEG Recordings

Passive Movement

Data Analysis

Statistical Analysis

Results

Neuroradiological Findings and Clinical Outcome

Kinematics of Passive Movements

Spontaneous Brain Activity

Modulation of the ~20-Hz Rhythm

Impaired-Hand Stimulation

Healthy-Hand Stimulation

Correlation of Rebound Amplitude With Clinical Outcome

Impaired-Hand Stimulation

Healthy-Hand Stimulation

Correlation With the Lesion Volume

Discussion

Rebound and Motor-Cortex Excitability

Rebound in the AH

Rebound in the UH

Effect of Stimulus Kinematics on Rebound

Perspectives to Development of Stroke Rehabilitation

Conclusions

Footnotes

Acknowledgements

Declaration of Conflicting Interests

Funding

Supplementary material

References