Sage Journals: Discover world-class research

Abstract

Background

Intracortical inhibitory/faciliatory measures are affected after stroke; however, the evidence is conflicting.

Objective

This meta-analysis aimed to investigate the changes in motor threshold (MT), motor evoked potential (MEP), short-interval intracortical inhibition (SICI), and intracortical facilitation (ICF), and identify sources of study variability using a machine learning approach.

Methods

We identified studies that objectively evaluated corticospinal excitability and intracortical inhibition/facilitation after stroke using transcranial magnetic stimulation. Pooled within- (ie, affected hemisphere [AH] vs unaffected hemisphere [UH]) and between-subjects (ie, AH and UH vs Control) standardized mean differences were computed. Decision trees determined which factors accurately predicted studies that showed alterations in corticospinal excitability and intracortical inhibition/facilitation.

Results

A total of 35 studies (625 stroke patients and 328 healthy controls) were included. MT was significantly increased and MEP was significantly decreased (ie, reduced excitability) in the AH when compared with the UH and Control (P < .01). SICI was increased (ie, reduced inhibition) for the AH when compared with the UH, and for the AH and UH when compared with Control (P < .001). ICF was significantly increased (ie, increased facilitation) in the AH when compared with UH (P = .016) and decreased in UH when compared with Control (P < 0.001). Decision trees indicated that demographic and methodological factors accurately predicted (73%-86%) studies that showed alterations in corticospinal and intracortical excitability measures.

Conclusions

The findings indicate that stroke alters corticospinal and intracortical excitability measures. Alterations in SICI and ICF may reflect disinhibition of the motor cortex after stroke, which is contrary to the notion that stroke increases inhibition of the affected side.

Keywords

hemiparesis cortical excitability brain plasticity rehabilitation TMS regression tree

Introduction

More than 795, 000 people suffer a stroke each year in the United States.¹ A stroke can lead to significant impairments that affect cognition,² speech,³ and sensorimotor systems.^4,5 Consequently, stroke is the leading cause of long-term disability in the United States.¹ Alterations in corticospinal excitability have been attributed as one of the underlying mechanisms for motor impairment after stroke.^6-9 Therefore, understanding how stroke affects corticospinal excitability could assist in the development of targeted interventions to improve motor function and aid stroke recovery.

The measurement of motor cortical and corticospinal excitability can be achieved using noninvasive brain stimulation techniques such as transcranial magnetic stimulation (TMS).^10,11 TMS uses an electromagnet (a coil) to deliver a brief magnetic pulse that stimulates the neurons in a specific area of the brain.¹⁰ The TMS intensity and the amplitude of the resulting motor-evoked responses are then commonly used to quantify the excitability of the motor cortex. Specifically, 4 prominent measures of corticospinal and intracortical excitability measured with TMS include the motor threshold (MT), motor evoked potential (MEP) amplitude, short-interval intracortical inhibition (SICI), and intracortical facilitation (ICF).¹² The MT and MEP amplitude are measures of corticospinal excitability. MT indicates the minimum amount of magnetic stimulation required to activate the motor cortex and elicit a motor response, while MEP measures the elicited motor response at a given stimulation intensity. In contrast, SICI and ICF are measures of intracortical excitability, where SICI measures the amount of intracortical inhibition and ICF measures the amount of ICF in the brain.^13,14 MT and MEP are straightforward metrics to quantify, and these procedures are largely standardized between studies.^15,16 Studies have repeatedly shown that MT increases and MEP amplitude decreases after stroke.^8,17-19 On the other hand, quantifying SICI and ICF requires more involved procedures: a stimulus of a weaker intensity than the MT (ie, a conditioning stimulus) is followed by a suprathreshold stimulus (ie, the test stimulus) delivered a short time later (ie, an interstimulus interval [ISI]).^12,20 The conditioning stimulus activates inhibitory or facilitatory interneurons in the cortex, which suppress or facilitate the response to the test stimulus. The evidence related to how stroke affects these measures in the motor cortex is somewhat conflicting, with some studies showing reduced intracortical excitability^21-25 and others showing increased intracortical excitability or no significant change after a stroke.^26-31 Developing a more concrete understanding of how stroke alters SICI and ICF will enable us to better understand how these measures contribute to motor function and recovery after stroke.

Therefore, the primary purpose of this study was to systematically review the current research related to alterations in corticospinal and intracortical excitability following a stroke and perform a meta-analysis to gain a better understanding of the magnitude (ie, effect size) and direction of changes (ie, increased or decreased) in corticospinal (MT: primary; MEP: secondary), and intracortical (SICI: primary; ICF: secondary) excitability after a unilateral stroke. We chose MT and SICI as our primary measures as these were more consistently reported. It is important to note that the measurements of MT, MEP, SICI, and ICF after stroke could be affected by the (1) use of different parameters for measuring MT (active vs resting) and SICI/ICF (e.g., conditioning stimulus intensity [CSI], ISI, and test stimulus intensity [TSI])³²; (2) small sample sizes, which could lead to sampling bias and type-II error; and (3) differences in the characteristics of the sample population (eg, timing, location, and type of stroke). Hence, a secondary purpose of this study was to identify potential sources of variability in these outcome measures using decision tree analyses. We hypothesized that stroke survivors would exhibit increased MT and SICI and reduced MEP and ICF in their affected hemisphere when compared with the unaffected hemisphere and uninjured controls and that the parameters used in testing and timing of stroke would predict differences in corticospinal and intracortical excitability outcomes.

Methods

Identification and Selection of Studies

Using guidelines of the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement,³³ a comprehensive search of published literature was done on the following electronic databases: (1) PubMed, (2) EBSCO Host, (3) Web of Science Core Collection, (4) Embase, (5) the Cochrane Library, and (6) Scopus. Keyword combinations for the search included: “transcranial magnetic stimulation,” “TMS,” “short interval intracortical inhibition,” “SICI,” “intracortical inhibition,” “stroke,” “cerebrovascular accident,” “CVA,” “hemiparesis,” and “hemiplegia.” References from included articles were manually inspected for other potential studies that could be included. All databases were searched for relevant publications up until March 1st, 2022.

Inclusion and Exclusion Criteria

Two independent reviewers used Rayyan to determine the eligibility of studies. For the studies to be included in the meta-analysis, they had to be (1) written in English, (2) original investigations related to individuals who have suffered a stroke, and (3) cross-sectional and prospective case-controlled studies that measured MT and SICI using TMS. Case studies, conference abstracts, studies that focused on other diseases or conditions and variables, or studies that did not meet the inclusion criteria were excluded. Intervention studies were included only if the study contained pre-intervention data. Initial screening was done using the title and abstract and if that was insufficient, then the reviewers read through entire articles to determine eligibility. To check for publication bias in outcome variables, funnel plots and Egger’s tests were used.³⁴

Data Extraction and Analysis

A data extraction sheet was used to summarize the following variables for stroke survivors and control participants: (1) sample size, (2) age, (3) upper or lower extremity muscles, (4) time since stroke, (5) type of measurement, and (6) neuromuscular parameters (MT, SICI, active or resting TMS, CSI, TSI, and ISI). The type of measurement referred to whether SICI was quantified by the change in the motor-evoked potential when the TSI was held constant (ie, conventional), or by the change in TSI required to elicit a similar motor-evoked potential (ie, unconventional). When there was insufficient data for these variables, efforts were taken to obtain these data by contacting the study authors. When numerical values for the outcome variables were not available, we used an online data extraction tool to estimate the means and standard deviations.^35,36

An additional secondary analysis was performed using this dataset by identifying the articles that evaluated MEP and ICF. The following variables for stroke survivors and control participants were collected from each study: (1) sample size, (2) age, (3) upper or lower extremity muscles, (4) time since stroke, (5) type of measurement, and (6) neuromuscular parameters (MEP, ICF, active or resting TMS, CSI, TSI, and ISI). When numerical values for the outcome variables were not available, we used an online data extraction tool to estimate the means and standard deviations.^35,36

Statistical Analysis

Descriptive statistics for all demographic and neuromuscular variables were calculated and tabulated for both the stroke and uninjured control groups (ie, the affected hemisphere [AH], unaffected hemisphere [UH], and uninjured participants [Control]). The following equations were used to calculate the pooled standard deviation (equation (1)) and effect size (equation (2)) between the AH and UH of the stroke survivors (ie, within-subjects analysis), assuming a modest correlation of .5^37-40:

s_{pooled} = \sqrt{s_{AH}^{2} + s_{UH}^{2} - 2 \times 0.5 \times s_{AH} \times s_{UH}}

(1)

ES (cohen's d) = \frac{(μ_{AH} - μ_{UH})}{s_{pooled}} .

(2)

The standard error of the effect size for the within-subjects analysis. (equation (3)) was calculated using the following equation^39,40:

S E_{d} = \sqrt{(\frac{1}{n} + \frac{d^{2}}{2 n})}

(3)

For the standard error, since there were occasions where the affected hemisphere was unable to output an MEP response, the smallest population was used for n.

The following equations were used to calculate the pooled standard deviation (equation (4)) and effect size (equation (5)) for between group comparisons (AH vs Control and UH vs Control)^37-40:

s_{pooled} \sqrt{\frac{s_{Stroke}^{2} (n_{Stroke} - 1) + s_{Control}^{2} (n_{Control} - 1)}{n_{Stroke} + n_{Control} - 2}}

(4)

ES (cohen’s d) = \frac{(μ_{S t r o k e} - μ_{C o n t r o l})}{s_{pooled}} .

(5)

The standard error of the effect size for the stroke and uninjured control comparison (equation (6)) was calculated using the following equation^39,40:

{SE}_{d} = \sqrt{(\frac{n_{Stroke} + n_{Control}}{n_{Stroke} \times n_{Control}} + \frac{d^{2}}{2 (n_{Stroke} + n_{Control})})}

(6)

Standardized mean difference (SMD) meta-analyses were performed for MT, SICI, MEP, and ICF using the JASP software program (Version 0.16.4; JASP-Stats, Amsterdam, The Netherlands). SMD meta-analyses were done to quantify (1) AH and UH differences, (2) AH and Control differences, and (3) UH and Control differences.^38,39 As eligible studies differed in their participant demographics and variables collected, a random-effects model was chosen to increase generalizability.³⁷ A Wald’s Test was used to determine within- (AH vs UH) and between-group differences (AH vs Control and UH vs Control) in MT, SICI, MEP, and ICF. Forest plots were used to determine the SMDs and 95% confidence intervals (95% CIs) for within- and between-group differences in MT, SICI, MEP, and ICF. $I^{2}$ statistics were used to assess heterogeneity between the eligible studies. For this, $I^{2}$ values of <40% were considered to be not important, and $I^{2}$ values of 50%— 90% were representative of substantial heterogeneity.⁴¹

Decision Tree Analysis

SICI and MT (Primary Analysis)

When there was a significant SMD, decision tree analyses were conducted to determine the factors that predicted alterations in MT and SICI. For this analysis, we computed a binary variable that indicated if the MT or SICI effect size for a study was lower than the 95% confidence interval (CI) of the pooled estimate, which served as the dependent variable. For MT, if the study effect size was less than the 95% CI of the pooled estimate, this indicated that there was no reduced excitability found in the study. For SICI, if the study effect size was less than the 95% CI of the pooled estimate, this indicated that there was no reduced inhibition. The binary variable was predicted using several independent variables including sample size, time after stroke, age, extremity (upper or lower), type of contraction (active or resting), the CSI, the TSI, ISI, and the type of measurement (ie, conventional or unconventional). Nodes identified by the decision tree that represented a relative gain of over 10% were considered to be meaningful.^42,43 Decision trees were validated using cross-validation with 5 sample folds to determine accuracy when applying these models to new data. Analyses were performed using Statistical Package for the Social Sciences (SPSS; Version 29, IBM Corporation).

MEP and ICF (Secondary Analysis)

A similar decision tree analysis was performed for MEP and ICF when there was a significant SMD. For this analysis, we computed a binary variable that indicated if the MEP or ICF effect size for a study was lower than the 95% confidence interval (CI) of the pooled estimate, which served as the dependent variable. For MEP, if the study effect size was greater than the 95% CI of the pooled estimate, this indicated there was no reduced excitability. For ICF, when comparing the AH and UH, if the study effect size was less than the 95% CI of the pooled estimate, this indicated there was no increase in facilitation; when comparing the UH and C, if the study effect size was greater than 95% CI of the pooled estimate, this indicated there was no reduction in facilitation. The binary variable was predicted using several independent variables including sample size, time after stroke, age, extremity (upper or lower), type of contraction (active or resting), the CSI, the TSI, ISI, and the type of measurement (ie, conventional or unconventional). Nodes identified by the decision tree that represented a relative gain of over 10% were considered to be meaningful.^42,43 Decision trees were validated using cross-validation with 5 sample folds to determine the accuracy when applying these trees to new data. Analyses were performed using Statistical Package for the Social Sciences (SPSS; Version 29, IBM Corporation).

Results

Literature Search, Study Selection, and Study Characteristics

Our comprehensive search identified 1395 articles. There were 592 articles after the removal of duplicate studies. During the initial screening process, a total of 515 articles were removed and 77 articles were then further evaluated. A total of 35 studies met our eligibility criteria.^{7,21-31,44-66} 29 studies reported MT, 35 reported SICI, 25 studies reported MEP, and 20 reported ICF. Subject characteristics and descriptive data for each study is outlined in Supplemental Table 1. A PRISMA flowchart for the literature search and screening process is shown in Figure 1. We note that some studies included multiple subgroups of stroke patients or performed analyses with multiple stimulation parameters, which were all included in our analyses. Hence, Forest plots may include multiple effect sizes for a single study.

Figure 1.

A schematic of the PRISMA flow diagram of the identification and selection of the studies included in this meta-analysis.

Motor Threshold

A total of 22 studies were included that compared AH versus UH with sample sizes of 344 and 397, respectively ( $I^{2}$ = 87.060%, P < .001). A total of 19 studies were included that compared the AH versus Control with sample sizes of 302 and 282, respectively ( $I^{2}$ = 88.545%, P < .001). A total of 12 studies were included that compared UH versus Control with sample sizes of 261 and 187, respectively ( $I^{2}$ = 89.667%, P < .001). Our meta-analysis indicated a significant increase in MT of AH when compared with UH (Figure 2; SMD = 0.87, 95% CI = 0.58-1.17, P < .001) and Control (Figure 3; SMD = 0.96, 95% CI = 0.52-1.40, P < .001). There were no significant differences in MT of UH when compared with Control (Supplemental Figure 1; SMD = 0.05, 95% CI = −0.49 to 0.58, P = .867).

Figure 2.

Forest plot depicting the pooled standard mean difference (95% confidence intervals (lower limit to upper limit) of the motor threshold (MT) between the affected hemisphere versus unaffected hemisphere.

Figure 3.

Forest plot depicting the pooled standard mean difference (95% confidence intervals (lower limit to upper limit) of the motor threshold (MT) between the affected hemisphere versus uninjured controls.

Short-Interval Intracortical Inhibition

A total of 28 studies were included that compared AH versus UH with sample sizes of 435 and 519, respectively ( $I^{2}$ = 89.750%, P < .001). A total of 21 studies were included that compared the AH versus Control with sample sizes of 307 and 306, respectively ( $I^{2}$ = 91.124%, P < .001). A total of 14 studies were included that compared UH versus Control with sample sizes of 288 and 213, respectively ( $I^{2}$ = 91.782%, P < .001). Our meta-analysis indicated a significant reduction in intracortical inhibition (ie, an increase in SICI) of AH when compared with UH (Figure 4; SMD = 0.70, 95% CI = 0.45-0.96, P < .001) and Control (Figure 5; SMD = 1.52. About 95% CI = 1.09-1.95, P < .001). There was also a significant reduction in intracortical inhibition of UH when compared with Control (Supplemental Figure 2; SMD = 1.13, 95% CI = 0.66-1.59, P < 0.001).

Figure 4.

Forest plot depicting the pooled standard mean difference (95% confidence intervals (lower limit to upper limit) of the short-interval intracortical inhibition (SICI) between the affected hemisphere versus unaffected hemisphere.

Figure 5.

MEP Amplitude

A total of 17 studies were included that compared AH versus UH with sample sizes of 237 and 267, respectively ( $I^{2}$ = 56.876%, P < .001). A total of 16 studies were included that compared the AH versus Control with sample sizes of 245 and 249, respectively ( $I^{2}$ = 79.391%, P < .001). A total of 10 studies were included that compared UH versus Control with sample sizes of 193 and 164, respectively ( $I^{2}$ = 74.620%, P < .001). Our meta-analysis indicated a significant decrease in MEP of AH when compared with UH (Supplemental Figure 3; SMD = −0.45, 95% CI = −0.63 to −0.27, P < .001) and Control (Supplemental Figure 4; SMD = −0.49, 95% CI = −0.83 to −0.15, P = .005). There were no significant differences in MEP of UH when compared with Control (Supplemental Figure 5; SMD = 0.23, 95% CI = −0.14 to 0.60, P = .225).

Intracortical Facilitation

A total of 18 studies were included that compared AH versus UH with sample sizes of 343 and 377, respectively ( $I^{2}$ = 88.480%, P < .001). A total of 12 studies were included that compared the AH versus Control with sample sizes of 178 and 198, respectively ( $I^{2}$ = 83.782%, P < .001). A total of 10 studies were included that compared UH versus Control with sample sizes of 174 and 164, respectively ( $I^{2}$ = 78.909%, P < .001). Our meta-analysis indicated there was a significant increase in ICF of AH when compared with UH (Supplemental Figure 6; SMD = 0.32, 95% CI = 0.06-0.57, P = .016). There was no significant change in ICF of the AH when compared with Control (Supplemental Figure 7; SMD = −0.17, 95% CI = −0.47 to 0.14, P < .288). There was a significant decrease in ICF of UH when compared with Control (Supplemental Figure 8; SMD = −0.58, 95% CI = −0.85 to −0.30, P < .001).

Risk of Publication Bias

Visual inspection of funnel plots indicated some possibilities of bias for the variables included in the analysis. The Egger’s test indicated significant bias for both MT (P < .05), SICI (P < .01), and ICF (P < .05). For MEP, Egger’s test was only significant when comparing AH and UH (P < .01).

Decision Trees

Decision tree analyses were performed to visualize how demographic and methodological variability within studies contributed to alterations in corticospinal and intracortical excitability. Overall, these models were able to categorize these studies with high accuracy (73%-86%). Specific factors and their relative contributions in constructing these models (ie, relative gain) are described below, and the results of the five-fold cross-validation are provided in the supplemental materials (Supplemental Figure 9).

Motor Threshold

The decision tree comparing MT of AH with UH (Figure 6) highlighted the time after stroke as the most important factor for predicting if patients had reduced corticospinal excitability, where studies with reduced time after stroke (ie, ≤1327.5 days) were more likely to have reduced corticospinal excitability (Gain = 100%). This factor was able to predict studies that found reduced corticospinal excitability with 73.0% accuracy. Cross-validation affected this accuracy by 18.9%. When comparing MT of AH with Control, no factors explained whether or not the patients would have reduced corticospinal excitability.

Figure 6.

Decision tree for motor threshold (MT) between the affected hemisphere and unaffected hemisphere. In each node, n refers to the number of studies and not the number of subjects. Decision tree conditions: maximum depth of tree of 3; minimum value of parent node of 8; and a minimum value of child node of 4. Nodes are considered to significantly contribute to predictions of altered intracortical inhibition if they represent a gain >10%.

Short-Interval Intracortical Inhibition

The decision tree comparing SICI of AH with UH (Figure 7) indicated that the sample size was the most important factor in predicting intracortical inhibition, where studies with lower sample sizes (ie, ≤12.5 subjects) often found reduced intracortical inhibition. Further, in studies that had lower sample sizes, the type of measurement that was performed was the next most important predictor of inhibition, where studies using the conventional method often showed reduced intracortical inhibition (Gain = 52.6%). In studies with larger sample sizes (ie, >12.5 subjects), the time after stroke was the next most important factor for predicting inhibition, where studies of more chronic patients (ie, >628.5 days after stroke) often showed no reduction in intracortical inhibition. Finally, in studies with large sample sizes (ie, >12.5 subjects) and reduced time after stroke (ie, ≤628.5 days), the average age of the subjects was the next most important factor for predicting reduced intracortical inhibition (≤63.8 years). Typically, younger patients in these studies exhibited reductions in intracortical inhibition after stroke (Gain = 28.9%). These factors were able to predict studies that found reduced intracortical inhibition with 82.8% accuracy. Cross-validation affected this accuracy by 17.2%.

Figure 7.

Decision tree for short-interval intracortical inhibition (SICI) between the affected hemisphere and unaffected hemisphere. In each node, n refers to the number of studies and not the number of subjects. Decision tree conditions: maximum depth of tree of 3; minimum value of parent node of 8; and a minimum value of child node of 4. Nodes are considered to significantly contribute to predictions of altered intracortical inhibition if they represent a gain >10%.

When comparing SICI of AH with Control (Figure 8), the time after stroke was the most important factor for predicting reduced intracortical inhibition, where studies with reduced time after stroke (ie, ≤863.4 days) were more likely to have reduced intracortical inhibition. In studies with reduced times after stroke, the next most important factor for predicting inhibition was the TSI, where studies using lower TSIs (ie, ≤127.98% MT) showed reduced intracortical inhibition. Finally, in studies with lower test stimulus intensities, the CSI was the next most important factor for predicting reduced intracortical inhibition, where studies that used higher CSIs (ie, >60% MT) were more likely to show reduced intracortical inhibition (Gain = 96.2%). These factors were able to predict studies that found reduced intracortical inhibition with 79.2% accuracy. Cross-validation affected this accuracy by 20.9%.

Figure 8.

Decision tree for short-interval intracortical inhibition (SICI) between the affected hemisphere and uninjured controls. In each node, n refers to the number of studies and not the number of subjects. Decision tree conditions: maximum depth of tree of 3; minimum value of parent node of 8; and a minimum value of child node of 4. Nodes are considered to significantly contribute to predictions of altered intracortical inhibition if they represent a gain >10%.

When comparing SICI of UH with Control (Supplemental Figure 10), the average age of the participants was the most important factor for predicting inhibition, where studies with younger participants (ie, ≤64.95 years) were more likely to show reduced intracortical inhibition. In studies with younger participants, the time after stroke was the next most important factor for predicting reduced intracortical inhibition, where studies that examined lower time after stroke (ie, ≤279.3 days) were more likely to show reduced intracortical inhibition. Further, in the studies that examined lower time after stroke, most of the studies that showed reduced inhibition examined participants that were even more acute (ie, ≤101.8 days; Gain = 55.0%). Finally, in studies with older participants (ie, >64.95 years), the study sample size was the next most important factor for predicting reduced intracortical inhibition, where studies with larger sample sizes (ie, >12 subjects) were more likely to see reductions in intracortical inhibition (Gain = 25.0). These factors were able to predict studies that found reduced intracortical inhibition with 85.7% accuracy. Cross-validation affected this accuracy by 28.6%.

MEP Amplitude

When using decision trees to compare MEP amplitude of AH with UH and AH with Control, no factors explained whether or not the patients would have reduced corticospinal excitability.

Intracortical Facilitation

The decision tree comparing ICF of AH with UH (Supplemental Figure 11) highlighted the TSI as the most important factor for predicting if patients had increased facilitation (ie, reduced inhibition), where studies with higher TSIs (ie, >122.5%MT) were more likely to have increased facilitation (Gain = 27.6%). Further, in studies that used lower TSIs, the CSI was the next most important factor for predicting if participants had increased facilitation, with studies that used higher CSIs (ie, >77.50%MT) were more likely to have increased facilitation (Gain = 37.9%). Finally, in studies that used lower CSIs, the sample size was the next most important factor for predicting if there was an increase in facilitation, where studies with smaller sample sizes (ie, ≤9 subjects) were more likely to see increased facilitation (Gain = 13.8%). These factors were able to predict studies that found increased facilitation with 76.5% accuracy. Cross-validation affected this accuracy by 23.6%.

The decision tree comparing ICF of UH with Control (Supplemental Figure 12) highlighted the average age of the subjects as the most important factor for predicting if patients had reduced facilitation (ie, increased inhibition), where studies with younger participants (ie, ≤66.35 years) were more likely see reduced facilitation. Further, in studies that tested on younger participants, the study's sample size was the next most important factor for predicting if participants had reduced facilitation, with studies that tested on fewer subjects (ie, ≤14.5 subjects) being more likely to see reduced facilitation (Gain = 61.9%). Finally, in the studies that tested on larger samples, the average age of the participants was again the next most important factor for predicting if there was reduced facilitation, where studies that tested on younger subjects (ie, ≤63.8 years) were more likely to see reduced facilitation (Gain = 19.0%). These factors were able to predict studies that found reduced inhibition with 83.8% accuracy. Cross-validation affected this accuracy by 21.6%.

Discussion

The purposes of this meta-analysis were to (1) systematically evaluate the current literature on alterations in corticospinal (MT and MEP) and intracortical excitability (SICI and ICF)following stroke and (2) use a novel decision tree analysis to identify the methodological and demographic factors that contribute to variability in study outcomes for corticospinal and intracortical excitability. The key findings of this meta-analysis were (1) corticospinal excitability of the AH was significantly reduced when compared with the UH and Control; (2) SICI of the AH was increased (ie, intracortical inhibition was reduced) when compared with the UH and Control; (3) intracortical inhibition was bilaterally reduced when compared with Control after stroke; (4) ICF was increased in the AH when compared with the UH (ie, increased facilitation) and decreased in the UH when compared with Control; (5) corticospinal excitability in the AH was found to recover in comparison with the UH with time after stroke; and (6) demographic and methodological factors accurately predicted (73%-86%) studies that showed alterations in corticospinal and intracortical excitability measures. These findings collectively improve our understanding of neurophysiological alterations in corticospinal and intracortical excitability after stroke and emphasize the importance of restoring functional connectivity and normal intracortical excitability after stroke.

A key finding of this meta-analysis was that corticospinal excitability of the AH was significantly reduced when compared with the UH and Control; however, there were no differences when comparing the UH with Control. This indicates that stroke-induced changes in corticospinal excitability are hemisphere-specific, and largely confined to the affected area of the brain. The reduced corticospinal excitability in the AH is most likely indicative of a reduction of functional connectivity and is of cortical origin, as previous studies have demonstrated that spinal excitability increases after stroke.⁶⁷ Interestingly, the results of the decision tree analysis on MT indicated that studies that included participants that were greater than 3.5 years after stroke did not show reduced corticospinal excitability, indicating that some measures of excitability were generally restored with time after stroke. We note that, although the decision tree analysis did not predict the chronicity of stroke improves MEP amplitude, our Forest plot indicated that many studies evaluating chronic stroke survivors generally showed recovery of MEP (Supplemental Figure 3). While MEP amplitudes can be confounded by the MT used for the stimulation, as MEP amplitudes are typically reported at a pre-defined %MT, the observation that MT recovered with time indicates that this was not the case.

We also found that the SICI was reduced (ie, increased excitability) in the AH when compared with both the UH and Control, and that ICF was increased in the AH when compared with the UH. These observations are interesting, as they are contrary to the general notion that stroke results in inhibition of the affected side. The reorganization of the brain after a stroke, such as the reinforcement of the ipsilateral projections of the non-lesioned hemisphere,⁶⁸ naturally results in the facilitation of inhibitory circuits (eg, transcallosal inhibitory pathway) of the lesioned hemisphere. Following this logic, many interventions, such as transcranial direct current stimulation (tDCS) and repetitive TMS (rTMS), have been applied in an attempt to increase cortical excitability and reduce inhibition of the lesioned hemisphere.^69-71 While we found that corticospinal excitability of the AH was reduced in stroke patients, our results also dispel the idea that there is increased intracortical inhibition of the AH. Instead, intracortical inhibition is reduced (ie, disinhibition) and ICF is increased (increased facilitation) in the AH following the injury.

It is possible that the observed disinhibition and facilitation could be a compensatory mechanism to account for the reduced corticospinal excitability after stroke.⁷² Hence, increasing the intracortical excitability may work to facilitate the recovery of MT and MEP. Our decision tree analysis partially supports this reasoning, as we found that MT tends to recover with time, while SICI also begins to normalize with time, especially in younger patients. However, we note that SICI was reduced in the UH while the MT was unaffected. This finding suggests that the reduction in SICI after stroke is mediated by a more generalized central mechanism (eg, molecular and pharmacological factors),^73-75 rather than lesion-specific reorganization. It is important to note that, unlike SICI, we observed a significant reduction in ICF of the UH when compared with the Control, which does not support a more generalized central mechanism. However, upon examining the Forest plot, this finding appeared to be primarily driven by studies that involved individuals with central post-stroke pain.²⁷

It is to be recognized that our findings do not completely dispel the notion other types of inhibition may be occurring following an injury. Notably, SICI is a measure of intracortical inhibition; hence it measures the activity of inhibitory circuits that act within the hemisphere. However, other inhibitory circuits may exist that act to mediate impairment. Specifically, interhemispheric inhibition—inhibition that occurs through circuits connecting the 2 hemispheres (eg, transcallosal pathways)⁷⁶—could still be working to reduce the activity of the affected hemisphere. While testing on this line of research is ongoing, current evidence on the presence of interhemispheric inhibition after stroke is also conflicting.⁷⁷ The limited research available suggests that, like with intracortical inhibition, inter-hemispheric inhibition after stroke is also dependent on the methods used to measure inhibition and the stage of recovery.⁷⁷ It is generally found that corticospinal excitability appears to be less suppressed by interhemispheric inhibition during the sub-acute phase of stroke when compared with controls, but this pattern reverses in chronic stroke.^7,78,79 However, our findings are somewhat contrary to the interhemispheric inhibition model during the chronic phase of stroke, as our decision tree analysis indicated that excitability of the affected hemisphere, as determined by a reduction in MT, also increases in this chronic group. This observation suggests that interhemispheric inhibition gradually decreases, rather than increases, with time. Indeed, a longitudinal study by Takechi et al⁷ found transcallosal inhibition reduces gradually with chronicity after stroke. Unfortunately, there are not many longitudinal studies that have evaluated both intracortical and interhemispheric inhibition after stroke, which are critically needed to improve our understanding of post-stroke neuronal reorganization.

Our decision tree analysis uncovered several important factors influencing SICI and ICF. We found that both demographical (ie, age and sample size) and methodological factors (ie, TSI, CSI, and the type of measurement) contributed to the observed variability in study outcomes for intracortical excitability. Particularly, higher CSIs and conventional methods of measuring SICI were generally found to be significant predictors of increased intracortical excitability. Hence, we feel that applying standardized methods should be considered when designing future studies that use SICI to measure neural excitability. Similarly, studies with lower sample sizes (<10-15) were generally showing reduced inhibition or increased intracortical excitability. Thus, future studies should consider using a reasonable sample size to ensure robustness and minimize spurious findings.

TMS-based measures, such as MT, MEP amplitude, SICI, and ICF, are commonly used to assess changes in cortical and corticospinal excitability after neurological injury^10,12,72,80; however, they only offer an indirect measurement of cortical excitability and can be affected by spinal and peripheral changes in excitability. For example, MT values will be increased if a subject has greater skull thickness (although this is typically not a problem when performing within-subjects analysis or comparing across an entire sample with similar demographics), and MEP values can be altered due to inhibition or potentiation at the spinal, peripheral, or muscular levels.⁸¹ Although SICI and ICF measures are normalized to remove non-cortical factors, these measures are still collected from the muscle; hence, they are not completely independent of the descending pathways. These potential confounding factors were reduced by performing a cross-sectional analysis, where many of the analyses were within subjects (ie, comparing AH to UH). Regardless, this meta-analysis reveals important information about how these commonly used measures are affected after stroke and also about potential factors that could contribute to the variability in study results.

While this study contributes significantly to our understanding of stroke-induced changes in intracortical and corticospinal excitability, several limitations must be considered. First, the presence of publication bias would typically raise concerns about the validity of our findings, as studies with non-significant results may have been unpublished, skewing the overall analysis. However, we note that we performed a cross-sectional analysis and were not looking to determine the effects of an intervention. As such we scrutinized the results from other cross-sectional studies as well as the baseline measures from studies that performed interventions. Hence, while there may be studies that examined interventions that have gone unpublished and may have contributed data to our analyses, we do not feel that the exclusion of these studies would have altered our findings. Second, our analyses were performed on individual studies using their average findings; hence, it is difficult to ascertain the heterogeneity of the stroke populations (ie, variability in stroke type, location, severity, and the time elapsed since the stroke) included in these studies and analysis. Controlling for these factors would require intimate knowledge of the datasets used in each study, which adds a level of complexity to the analysis and reduces the feasibility of performing this research. However, the ideal method for answering these research questions would contain the individual patient data from all of the included studies, which may be possible in the future given the drive for data sharing policies established by several journals and funding agencies. Third, the sample size (ie, the number of studies) in some of the nodes in the decision tree analyses contained a small number of observations (ie, n < 5). Therefore, these results need to be interpreted with caution as the accuracy of the model can become unreliable. However, we note that cross-validation on untrained datasets reduced this accuracy only by 17% to 25%, indicating that the model was generally robust in prediction. Although we used the commonly recommended node sizes based on our overall sample size, there is no clear consensus on the choice of these parameters, which makes it difficult for the reader to interpret the meaningfulness of the decision tree model. While we addressed this issue by adding criteria for determining which nodes are meaningful based on the gain of individual nodes (ie, >10% is considered to be meaningful), it is to be noted that these gains can only be used to predict the target variable (eg, increased intracortical inhibition in the AH when compared with the UH) and cannot be used to interpret the non-target variable (eg, no change in intracortical inhibition in the AH when compared with UH). Last, the quality of systematic reviews and meta-analyses is highly dependent on the quality of the studies included in the analysis. While we had strict inclusion criteria to exclude studies that involved case studies and case series, most studies were of small sample size (<30 participants) and were cross-sectional in nature (~75%). Further, most studies used TSIs ≤120%MT, whereas there is some evidence to suggest that SICI is best captured at TSIs >120%MT.⁴⁰ Hence, the limitations of these individual studies that were included in this meta-analysis need to be considered while interpreting the study results.

Conclusions

In conclusion, this meta-analysis indicates the presence of altered neural excitability in individuals post-stroke. Specifically, the results indicate that intracortical inhibition is reduced after stroke (ie, increased excitability), which is contrary to the notion of generalized inhibition of the lesioned hemisphere after stroke. The results also indicate that some of these alterations in neural excitability tend to recover with time, albeit not to a complete extent. Our novel decision tree analysis also indicates that demographic and methodological factors can accurately predict studies that show alterations in corticospinal excitability and intracortical excitability. This finding underlines the need for methodological consistency when measuring SICI/ICF and emphasizes the dynamic nature of these changes during the recovery timeline. Future studies should aim for larger, more representative sample sizes and longitudinal designs to map the trajectory of excitability changes following a stroke. To advance the development of targeted interventions, further research is needed to better understand the functional implications of these alterations in corticospinal and intracortical excitability.

Supplemental Material

sj-docx-1-nnr-10.1177_15459683241281299 – Supplemental material for Altered Corticospinal and Intracortical Excitability After Stroke: A Systematic Review With Meta-Analysis

Supplemental material, sj-docx-1-nnr-10.1177_15459683241281299 for Altered Corticospinal and Intracortical Excitability After Stroke: A Systematic Review With Meta-Analysis by Edward P. Washabaugh, Sierra A. Foley, Emily G. Czopek and Chandramouli Krishnan in Neurorehabilitation and Neural Repair

Footnotes

Acknowledgements

None.

Author Contributions

Edward P. Washabaugh: Conceptualization; Data curation; Formal analysis; Methodology; Project administration; Supervision; Visualization; Writing—original draft; Writing—review & editing. Sierra A. Foley: Data curation; Formal analysis; Visualization; Writing—review & editing. Emily G. Czopek: Data curation; Formal analysis; Visualization; Writing—original draft. Chandramouli Krishnan: Conceptualization; Formal analysis; Investigation; Supervision; Writing—original draft; Writing—review & editing.

Declaration of Conflicting Interests

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

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iDs

Edward P. Washabaugh

Chandramouli Krishnan

Supplementary material for this article is available on the Neurorehabilitation & Neural Repair website along with the online version of this article.

References

Tsao

Aday

Almarzooq

, et al. Heart disease and stroke statistics-2022 update: a report from the American Heart Association. Circulation. 2022;145(8):e153-e639. doi:10.1161/CIR.0000000000001052

Haring

HP.

Cognitive impairment after stroke. Curr Opin Neurol. 2002;15(1):79-84. doi:10.1097/00019052-200202000-00012

Borthwick

Communication impairment in patients following stroke. Nurs Stand. 2012;26(19):35-41. doi:10.7748/ns2012.01.26.19.35.c8879

Perry

Garrett

Gronley

Mulroy

SJ.

Classification of walking handicap in the stroke population. Stroke. 1995;26(6):982-989. doi:10.1161/01.str.26.6.982

Raghavan

Upper limb motor impairment after stroke. Phys Med Rehabil Clin N Am. 2015;26(4):599-610. doi:10.1016/j.pmr.2015.06.008

Berenguer-Rocha

Baltar

Rocha

Shirahige

Brito

Monte-Silva

Interhemispheric asymmetry of the motor cortex excitability in stroke: relationship with sensory-motor impairment and injury chronicity. Neurol Sci. 2020;41(9):2591-2598. doi:10.1007/s10072-020-04350-4

Takechi

Matsunaga

Nakanishi

, et al. Longitudinal changes of motor cortical excitability and transcallosal inhibition after subcortical stroke. Clin Neurophysiol. 2014;125(10):2055-2069. doi:10.1016/j.clinph.2014.01.034

Veldema

Nowak

Gharabaghi

Resting motor threshold in the course of hand motor recovery after stroke: a systematic review. J Neuroeng Rehabil. 2021;18(1):158. doi:10.1186/s12984-021-00947-8

Volz

Sarfeld

Diekhoff

, et al. Motor cortex excitability and connectivity in chronic stroke: a multimodal model of functional reorganization. Brain Struct Funct. 2015;220(2):1093-1107. doi:10.1007/s00429-013-0702-8

10.

Kobayashi

Pascual-Leone

Transcranial magnetic stimulation in neurology. Lancet Neurol. 2003;2(3):145-156. doi:10.1016/s1474-4422(03)00321-1

11.

Ridding

Rothwell

JC.

Stimulus/response curves as a method of measuring motor cortical excitability in man. Electroencephalogr Clin Neurophysiol. 1997;105(5):340-344. doi:10.1016/s0924-980x(97)00041-6

12.

Di Pino

Pellegrino

Assenza

, et al. Modulation of brain plasticity in stroke: a novel model for neurorehabilitation. Nat Rev Neurol. 2014;10(10):597-608. doi:10.1038/nrneurol.2014.162

13.

Di Lazzaro

Pilato

Dileone

, et al. GABAA receptor subtype specific enhancement of inhibition in human motor cortex. J Physiol. 2006;575(Pt 3):721-726. doi:10.1113/jphysiol.2006.114694

14.

Florian

Muller-Dahlhaus

Liu

Ziemann

Inhibitory circuits and the nature of their interactions in the human motor cortex a pharmacological TMS study. J Physiol. 2008;586(2):495-514. doi:10.1113/jphysiol.2007.142059

15.

Caramia

Cicinelli

Paradiso

, et al. Excitability changes of muscular responses to magnetic brain stimulation in patients with central motor disorders. Electroencephalogr Clin Neurophysiol. 1991;81(4):243-250. doi:10.1016/0168-5597(91)90009-m

16.

Kammer

Beck

Thielscher

Laubis-Herrmann

Topka

Motor thresholds in humans: a transcranial magnetic stimulation study comparing different pulse waveforms, current directions and stimulator types. Clin Neurophysiol. 2001;112(2):250-258. doi:10.1016/s1388-2457(00)00513-7

17.

Bai

Zhang

Fong

KNK

. Intracortical and intercortical networks in patients after stroke: a concurrent TMS-EEG study. J Neuroeng Rehabil. 2023;20(1):100. doi:10.1186/s12984-023-01223-7

18.

Pennisi

Rapisarda

Bella

Calabrese

Maertens

Noordhout

Delwaide

PJ.

Absence of response to early transcranial magnetic stimulation in ischemic stroke patients: prognostic value for hand motor recovery. Stroke. 1999;30(12):2666-2670. doi:10.1161/01.str.30.12.2666

19.

Rapisarda

Bastings

de Noordhout

Pennisi

Delwaide

PJ.

Can motor recovery in stroke patients be predicted by early transcranial magnetic stimulation?

Stroke. 1996;27(12):2191-2196. doi:10.1161/01.str.27.12.2191

20.

Kujirai

Caramia

Rothwell

, et al. Corticocortical inhibition in human motor cortex. J Physiol. 1993;471:501-519. doi:10.1113/jphysiol.1993.sp019912

21.

Kuo

Tang

Tsai

, et al. Neurophysiological signatures of hand motor response to dual-transcranial direct current stimulation in subacute stroke: a TMS and MEG study. J Neuroeng Rehabil. 2020;17(1):72. doi:10.1186/s12984-020-00706-1

22.

Mooney

Ackerley

Rajeswaran

, et al. The influence of primary motor cortex inhibition on upper limb impairment and function in chronic stroke: a multimodal study. Neurorehabil Neural Repair. 2019;33(2):130-140. doi:10.1177/1545968319826052

23.

Liepert

Hamzei

Weiller

Lesion-induced and training-induced brain reorganization. Restor Neurol Neurosci. 2004;22(3-5):269-277.

24.

Malcolm

Vaughn

Greene

DP.

Inhibitory and excitatory motor cortex dysfunction persists in the chronic poststroke recovery phase. J Clin Neurophysiol. 2015;32(3):251-256. doi:10.1097/WNP.0000000000000143

25.

B-M

Kim

D-Y

Paik

N-J.

The relationship between the severity of motor impairment and motor cortical excitability in the unaffected hemisphere after unilateral stroke. Int Congress Ser. 2005;1278:291-294.

26.

Huynh

Vucic

Krishnan

Lin

Hornberger

Kiernan

MC.

Longitudinal plasticity across the neural axis in acute stroke. Neurorehabil Neural Repair. 2013;27(3):219-229. doi:10.1177/1545968312462071

27.

Tang

Lee

Jeng

, et al. Pathophysiology of central poststroke pain: motor cortex disinhibition and its clinical and sensory correlates. Stroke. 2019;50(10):2851-2857. doi:10.1161/STROKEAHA.119.025692

28.

Huynh

Vucic

Krishnan

Lin

Kiernan

MC.

Exploring the evolution of cortical excitability following acute stroke. Neurorehabil Neural Repair. 2016;30(3):244-257. doi:10.1177/1545968315593804

29.

Liepert

Motor cortex excitability in stroke before and after constraint-induced movement therapy. Cogn Behav Neurol. 2006;19(1):41-47. doi:10.1097/00146965-200603000-00005

30.

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(6):936-943. doi:10.1016/s1388-2457(02)00062-7

31.

Seo

Kim

Won

Park

Seo

MH.

Changes in intracortical excitability of affected and unaffected hemispheres after stroke evaluated by paired-pulse transcranial magnetic stimulation. Ann Rehabil Med. 2018;42(4):495-501. doi:10.5535/arm.2018.42.4.495

32.

Krishnan

Effect of paired-pulse stimulus parameters on the two phases of short interval intracortical inhibition in the quadriceps muscle group. Restor Neurol Neurosci. 2019;37(4):363-374. doi:10.3233/RNN-180894

33.

Moher

Liberati

Tetzlaff

Altman

Group

Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. PLoS Med. 2009;6(7):e1000097. doi:10.1371/journal.pmed.1000097

34.

Egger

Davey Smith

Schneider

Minder

Bias in meta-analysis detected by a simple, graphical test. BMJ. 1997;315(7109):629-634. doi:10.1136/bmj.315.7109.629

35.

Moeyaert

Maggin

Verkuilen

Reliability, validity, and usability of data extraction programs for single-case research designs. Behav Modif. 2016;40(6):874-900. doi:10.1177/0145445516645763

36.

Rohatgi

WebPlotDigitizer (Version 4.5). 2021. Accessed March 01, 2022. https://automeris.io/WebPlotDigitizer

37.

Borenstein

Hedges

Higgins

Rothstein

HR.

A basic introduction to fixed-effect and random-effects models for meta-analysis. Res Synth Methods. 2010;1(2):97-111. doi:10.1002/jrsm.12

38.

Dunlap

Cortina

Vaslow

Burke

MJ.

Meta-analysis of experiments with matched groups or repeated measures designs. Psychol Methods. 1996;1(2):170.

39.

Lakens

Calculating and reporting effect sizes to facilitate cumulative science: a practical primer for t-tests and ANOVAs. Front Psychol. 2013;4:863. doi:10.3389/fpsyg.2013.00863

40.

Rodriguez

Palmieri-Smith

Krishnan

How does anterior cruciate ligament reconstruction affect the functioning of the brain and spinal cord? A systematic review with meta-analysis. J Sport Health Sci. 2021;10(2):172-181. doi:10.1016/j.jshs.2020.07.005

41.

Higgins

Thomas

Chandler

, et al. Cochrane Handbook for Systematic Review of Interventions version 6.4. 2023. John Wiley & Sons Ltd. https://www.training.cochrane.org/handbook

42.

Sanjay K, Danti

Detection of fake opinions on online products using decision tree and information gain. Paper presented at: 2019 3rd International Conference on Computing Methodologies and Communication (ICCMC); March 27-29, 2019; Erode, India, IEEE: 372-375.

43.

Kwon

Karim

Topaz

Currie

LM.

Nurses “seeing forest for the trees” in the age of machine learning: using nursing knowledge to improve relevance and performance. Comput Inform Nurs. 2019;37(4):203-212. doi:10.1097/CIN.0000000000000508

44.

Acler

Robol

Fiaschi

Manganotti

A double blind placebo RCT to investigate the effects of serotonergic modulation on brain excitability and motor recovery in stroke patients. J Neurol. 2009;256(7):1152-1158. doi:10.1007/s00415-009-5093-7

45.

Beaulieu

Masse-Alarie

Brouwer

Schneider

Brain control of volitional ankle tasks in people with chronic stroke and in healthy individuals. J Neurol Sci. 2014;338(1-2):148-155. doi:10.1016/j.jns.2013.12.038

46.

Beaulieu

Masse-Alarie

Brouwer

Schneider

Noninvasive neurostimulation in chronic stroke: a double-blind randomized sham-controlled testing of clinical and corticomotor effects. Top Stroke Rehabil. 2015;22(1):8-17. doi:10.1179/1074935714Z.0000000032

47.

Beaulieu

Masse-Alarie

Ribot-Ciscar

Schneider

Reliability of lower limb transcranial magnetic stimulation outcomes in the ipsi- and contralesional hemispheres of adults with chronic stroke. Clin Neurophysiol. 2017;128(7):1290-1298. doi:10.1016/j.clinph.2017.04.021

48.

Cicinelli

Pasqualetti

Zaccagnini

Traversa

Oliveri

Rossini

PM.

Interhemispheric asymmetries of motor cortex excitability in the postacute stroke stage: a paired-pulse transcranial magnetic stimulation study. Stroke. 2003;34(11):2653-2658. doi:10.1161/01.STR.0000092122.96722.72

49.

Di Lazzaro

Profice

Pilato

, et al. The level of cortical afferent inhibition in acute stroke correlates with long-term functional recovery in humans. Stroke. 2012;43(1):250-252. doi:10.1161/STROKEAHA.111.631085

50.

Di Lazzaro

Profice

Pilato

, et al. Motor cortex plasticity predicts recovery in acute stroke. Cereb Cortex. 2010;20(7):1523-1528. doi:10.1093/cercor/bhp216

51.

Dimyan

Perez

Auh

Tarula

Wilson

Cohen

LG.

Nonparetic arm force does not overinhibit the paretic arm in chronic poststroke hemiparesis. Arch Phys Med Rehabil. 2014;95(5):849-856. doi:10.1016/j.apmr.2013.12.023

52.

Ferreiro

Andrade

Conforto

AB.

Decreased short-interval intracortical inhibition correlates with better pinch strength in patients with stroke and good motor recovery. Brain Stimul. 2018;11(4):772-774. doi:10.1016/j.brs.2018.01.030

53.

Goodwill

Teo

Morgan

Daly

Kidgell

DJ.

Bihemispheric-tDCS and upper limb rehabilitation improves retention of motor function in chronic stroke: a pilot study. Front Hum Neurosci. 2016;10:258. doi:10.3389/fnhum.2016.00258

54.

Honaga

Fujiwara

Tsuji

Hase

Ushiba

Liu

State of intracortical inhibitory interneuron activity in patients with chronic stroke. Clin Neurophysiol. 2013;124(2):364-370. doi:10.1016/j.clinph.2012.08.005

55.

Hummel

Steven

Hoppe

, et al. Deficient intracortical inhibition (SICI) during movement preparation after chronic stroke. Neurology. 2009;72(20):1766-1772. doi:10.1212/WNL.0b013e3181a609c5

56.

Kim

Kang

YJ.

Virtual reality-guided motor imagery increases corticomotor excitability in healthy volunteers and stroke patients. Ann Rehabil Med. 2016;40(3):420-431. doi:10.5535/arm.2016.40.3.420

57.

Julkunen

Maatta

Saisanen

, et al. Functional and structural cortical characteristics after restricted focal motor cortical infarction evaluated at chronic stage: indications from a preliminary study. Clin Neurophysiol. 2016;127(8):2775-2784. doi:10.1016/j.clinph.2016.05.013

58.

Kim

Lee

Cohen

Park

Choi

KG.

Motor cortical excitability in patients with poststroke epilepsy. Epilepsia. 2008;49(1):117-124. doi:10.1111/j.1528-1167.2007.01231.x

59.

Liepert

Restemeyer

Kucinski

Zittel

Weiller

Motor strokes: the lesion location determines motor excitability changes. Stroke. 2005;36(12):2648-2653. doi:10.1161/01.STR.0000189629.10603.02

60.

Liepert

Storch

Fritsch

Weiller

Motor cortex disinhibition in acute stroke. Clin Neurophysiol. 2000;111(4):671-676. doi:10.1016/s1388-2457(99)00312-0

61.

Lioumis

Mustanoja

Bikmullina

, et al. Probing modifications of cortical excitability during stroke recovery with navigated transcranial magnetic stimulation. Top Stroke Rehabil. 2012;19(2):182-192. doi:10.1310/tsr1902-182

62.

Liuzzi

Horniss

Lechner

, et al. Development of movement-related intracortical inhibition in acute to chronic subcortical stroke. Neurology. 2014;82(3):198-205. doi:10.1212/WNL.0000000000000028

63.

Manganotti

Acler

Zanette

Smania

Fiaschi

Motor cortical disinhibition during early and late recovery after stroke. Neurorehabil Neural Repair. 2008;22(4):396-403. doi:10.1177/1545968307313505

64.

Nardone

Tezzon

Inhibitory and excitatory circuits of cerebral cortex after ischaemic stroke: prognostic value of the transcranial magnetic stimulation. Electromyogr Clin Neurophysiol. 2002;42(3):131-136.

65.

Swayne

Rothwell

Ward

Greenwood

RJ.

Stages of motor output reorganization after hemispheric stroke suggested by longitudinal studies of cortical physiology. Cereb Cortex. 2008;18(8):1909-1922. doi:10.1093/cercor/bhm218

66.

Takeuchi

Tada

Toshima

Ikoma

Correlation of motor function with transcallosal and intracortical inhibition after stroke. J Rehabil Med. 2010;42(10):962-966. doi:10.2340/16501977-0628

67.

Phadke

Robertson

Condliffe

Patten

Upper-extremity H-reflex measurement post-stroke: reliability and inter-limb differences. Clin Neurophysiol. 2012;123(8):1606-1615. doi:10.1016/j.clinph.2011.12.012

68.

McPherson

Chen

Ellis

Yao

Heckman

Dewald

JPA

. Progressive recruitment of contralesional cortico-reticulospinal pathways drives motor impairment post stroke. J Physiol. 2018;596(7):1211-1225. doi:10.1113/JP274968

69.

Hummel

Celnik

Giraux

, et al. Effects of non-invasive cortical stimulation on skilled motor function in chronic stroke. Brain. 2005;128(Pt 3):490-499. doi:10.1093/brain/awh369

70.

Khedr

Ahmed

Fathy

Rothwell

JC.

Therapeutic trial of repetitive transcranial magnetic stimulation after acute ischemic stroke. Neurology. 2005;65(3):466-468. doi:10.1212/01.wnl.0000173067.84247.36

71.

Madhavan

Weber

2nd Stinear

JW.

Non-invasive brain stimulation enhances fine motor control of the hemiparetic ankle: implications for rehabilitation. Exp Brain Res. 2011;209(1):9-17. doi:10.1007/s00221-010-2511-0

72.

Ding

Triggs

Kamath

Patten

Short intracortical inhibition during voluntary movement reveals persistent impairment post-stroke. Front Neurol. 2018;9:1105. doi:10.3389/fneur.2018.01105

73.

Green

Hainsworth

Jackson

DM.

GABA potentiation: a logical pharmacological approach for the treatment of acute ischaemic stroke. Neuropharmacology. 2000;39(9):1483-1494. doi:10.1016/s0028-3908(99)00233-6

74.

Lindsay

Kouzouna

Simcox

Pandyan

AD.

Pharmacological interventions other than botulinum toxin for spasticity after stroke. Cochrane Database Syst Rev. 2016;10(10):CD010362. doi:10.1002/14651858.CD010362.pub2

75.

Paparella

Vanderwalle

Stagg

Maquet

An integrated measure of GABA to characterize post-stroke plasticity. Neuroimage Clin. 2023;39:103463. doi:10.1016/j.nicl.2023.103463

76.

Caleo

Plasticity of transcallosal pathways after stroke and their role in recovery. J Physiol. 2018;596(10):1789-1790. doi:10.1113/JP276081

77.

Gerges

ANH

Hordacre

Pietro

Moseley

Berryman

. Do adults with stroke have altered interhemispheric inhibition? A systematic review with meta-analysis. J Stroke Cerebrovasc Dis. 2022;31(7):106494. doi:10.1016/j.jstrokecerebrovasdis.2022.106494

78.

Butefisch

Wessling

Netz

Seitz

Homberg

Relationship between interhemispheric inhibition and motor cortex excitability in subacute stroke patients. Neurorehabil Neural Repair. 2008;22(1):4-21. doi:10.1177/1545968307301769

79.

Branscheidt

Schambra

, et al. Rethinking interhemispheric imbalance as a target for stroke neurorehabilitation. Ann Neurol. 2019;85(4):502-513. doi:10.1002/ana.25452

80.

Edwards

Meehan

Linsdell

, et al. Changes in thresholds for intracortical excitability in chronic stroke: more than just altered intracortical inhibition. Restor Neurol Neurosci. 2013;31(6):693-705. doi:10.3233/RNN-120300

81.

Spampinato

Ibanez

Rocchi

Rothwell

Motor potentials evoked by transcranial magnetic stimulation: interpreting a simple measure of a complex system. J Physiol. 2023;601(14):2827-2851. doi:10.1113/JP281885

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

4.59 MB