Linking corticospinal tract activation and upper-limb motor control in adults with cerebral palsy

Abstract

Aim:

To quantify the cervicomedullary motor evoked potentials (CMEPs) at the cervical spinal level in adults with cerebral palsy (CP) and determine if altered CMEPs are linked with upper-extremity motor function in this population.

Method:

This cross-sectional study consisted of a cohort of adults with CP (n = 15; mean age = 33 years 5 months [SD = 11 years 8 months]); Manual Ability Classification System levels I–IV) and neurotypical controls (n = 18; mean age = 30 years 10 months [SD = 10 years 4 months]), who were recruited to participate at an academic medical center. Adults with CP and typical adults (controls) were stimulated at the cervicomedullary junction to assess CMEPs at the cervical spinal cord level. Upper-extremity motor function was quantified using the Box and Blocks and Purdue Pegboard tests, self-reported upper-extremity function (UEF), and assessments of selective motor control.

Results:

At higher stimulation levels, the contralateral CMEP responses of adults with CP were different from typical adults (p = 0.032). Reduced CMEP was correlated with reduced upper-limb function, including worse performance on the Box and Blocks (rho = 0.625, p = 0.025) and Purdue Pegboard tests (rho = 0.701, p = 0.010), lower self-reported UEF (rho = 0.761, p = 0.009), and overall selective motor control (rho = 0.731, p = 0.007).

Interpretation:

Changes in the activation of spinal motoneurons through corticospinal pathways may have an important role in the altered upper-extremity motor function of individuals with CP.

Graphical Abstract

Cerebral palsy (CP) is the most common motor disability affecting children across the world.^1,2 However, changes in motor function in individuals with CP can be observed over the life span.³ Muscle weakness has been implicated in partially driving the motor impairment observed in individuals with CP;^4,5 however, the precise physiological mechanisms that govern weakness are currently unclear. Changes to the motor unit (consisting of the spinal motoneurons and the muscle fibers they innervate) are at least partly responsible.⁶ Previous transcranial magnetic stimulation research suggested that the strength of the descending motor pathways that activate the spinal motoneurons is reduced, leading to muscle weakness.⁷ Of these pathways, corticospinal tracts (CSTs) are structurally impacted, probably early during development.^8–13 However, it is not clear whether damage to these pathways is limited solely to the brain or whether these effects extend to the spinal cord where the CSTs terminate.

Previous neurophysiological studies used cervicomedullary motor evoked potentials (CMEPs) to isolate the influence of spinal cord interneurons on motor evoked potentials (MEPs). This measure is obtained using noninvasive electrical or magnetic stimulation of the CSTs that can evoke large, short-latency responses in the muscles of the upper and lower limbs.^14,15 Importantly, CMEP size is an indicator of the responsiveness of spinal motoneurons to corticospinal inputs and overall spinal motoneuron excitability.¹⁶ The purpose of this study was to evaluate and compare the reorganization of corticospinal pathways at the cervical level in adults with CP. In these experiments, we investigated changes in CMEP size from the flexor carpi radialis (FCR) muscle in a cohort of adults with CP and typical adults (controls). These responses allowed us to isolate the spinal contribution to MEPs traditionally obtained from cortical stimulation. We then examined the potential link between CMEP changes and measures of overall motor impairment (manual dexterity, self-reported upper-extremity function [UEF], and selective motor control) in this population.

METHOD

Participants

The Boys Town National Research Hospital’s institutional review board reviewed and approved the protocol for this study; the experimental work conformed to the standards set out by the Declaration of Helsinki, except for registration in a database. Participants or their guardians provided written informed consent for the experimental procedures and all participants assented to participate in the study. We based our sample size on the effect size observed in a previous study conducted in adults with a spinal cord injury, which used a similar CMEP protocol.¹⁶ Using an effect size of 1.79, we estimated that eight participants in each group would provide greater than 80% power to detect a similar difference at a 0.01 alpha level. As such, 15 were adults with unilateral or bilateral spastic CP (mean age = 33 years 5 months [SD = 11 years 8 months]; seven females; Manual Ability Classification System [MACS] levels I–IV) and 18 were typical adults (mean age = 30 years 10 months [SD = 10 years 4 months]; 10 females). Participants were recruited from local clinics, institutional registries, and the local Omaha community to participate in this study. Individuals with CP were excluded from the study if they had a dorsal root rhizotomy at the cervical level. Individuals taking anti-spasticity medications (e.g. baclofen and tizanidine) were also excluded. One participant had been given botulinum neurotoxin A in the tested FCR muscle in the past year. Further details on the participants with CP are provided in Table 1. Controls (typical adults) had no known neurological or musculoskeletal impairments that affected their hand movements at the time of the study. The control group was selected from a convenience sample of healthy adults in the local community. Controls were not matched according to age but were recruited within an age range. The ages of the respective groups were not significantly different (p > 0.05).

Table 1:

Demographics of the participants with cerebral palsy included in this study

Age (years)	Sex	Ethnicity	MACS level	GMFCS level	Clinical presentation	Brain insults	Maximum stimulation used (mA)
23	F	White	II	II	Spastic bilateral	Bilateral polymicrogyria most prominent in the occipital and parietal lobes, cerebellar hypoplasia and dysplasia, callosal dysgenesis	70
23	M	White	I	I	Spastic unilateral (left-sided)	Right periatrial trigone focus of remote ischemia	65
22	M	White	III	III	Spastic bilateral	Corpus callosum agenesis, bilateral posterior white matter atrophy, parietal cortical atrophy	95
34	F	Black	II	II	Spastic bilateral	No abnormal findings noted	95
29	M	White	III	II	Spastic unilateral (right-sided)	Left frontal atrophy and polymicrogyria, corpus callosum aplasia with parietal and cerebellar atrophy	95
38	M	Asian American	IV	IV	Spastic bilateral (quadriplegia)	Bilateral posterior parietal periventricular volume loss and diffuse abnormal white matter signal, atrophy of the posterior corpus callosum	95
26	M	White	III	II	Spastic unilateral (left-sided)	Right parietal polymicrogyria, mild asymmetric enlargement of right lateral ventricle, corpus callosum enlargement	95
24	M	White	II	II	Spastic bilateral	Partial agenesis of the corpus callosum	95
20	F	White	I	I	Spastic unilateral (right-sided)	Unable to complete MRI due to metal	95
33	F	White	IV	IV	Spastic bilateral (quadriplegia)	Bilateral posterior superior ventricle periventricular volume loss, empty sella	95
40	F	White	I	I	Spastic bilateral	No abnormal findings noted	95
60	M	White	III	I	Spastic bilateral	No abnormal findings noted	95
26	M	White	II	I	Spastic unilateral (right-sided)	Unable to complete MRI due to metal	95
50	F	Black	I	I	Spastic bilateral	No abnormal findings noted	95
46a	F	White	III	III	Spastic bilateral	No abnormal findings noted	95

^aOne participant received botulinum neurotoxin A in the FCR muscle in the past year. The information provided on each participant’s brain insults represent the radiologist’s read of a T1-weighted brain MRI.

Abbreviations: F, female; FCR, flexor carpi radialis; GMFCS, Gross Motor Function Classification System; M, male; MACS, Manual Ability Classification Scale; MRI, magnetic resonance imaging.

Experimental design

All participants were comfortably seated in a custom, upright neck-supporting chair. The examined arm was positioned with the shoulder in a neutral position and elbow flexed at 90 degrees. The forearm of the tested arm was pronated and restrained by straps to limit compensatory movements and to isolate wrist flexion. The experimental tasks were performed with the right arm. However, for the two individuals with left-sided spastic presentation, the left arm was tested. Participants received median nerve stimulation to quantify the FCR maximum compound muscle action potential (Mmax), which was used to normalize the stimulation responses. Participants subsequently underwent a stimulus–response curve to examine changes in CMEPs with increasing stimulus intensity. After these neurophysiological assessments, participants completed clinical assessments of upper-limb function. For all testing, table height was adjusted for bench versus wheelchair sitting. A neck-supporting chair was provided for neurophysiological testing.

Median nerve stimulation

Initially, the Mmax for the FCR muscle was calculated to normalize the CMEP outcomes in the final analyses. To evoke Mmax, single electrical stimuli were delivered to the median nerve of the tested limb with the unipolar surface electrodes positioned along the medial bicipital groove in the antecubital fossa, with the cathode positioned proximal to the anode at a distance of approximately 2cm. Single rectangular electrical pulses of 1ms duration were delivered using a constant current stimulator (DS8R, Digitimer North America, Fort Lauderdale, FL, USA); the FCR compound muscle action potentials were collected from surface electromyography (EMG). Stimulation intensity was started below the motor threshold and was sequentially increased by 1mA increments until a maximum M-wave was elicited. Mmax was defined through visual inspection as the point beyond which M-wave amplitude no longer increased with increased current intensity.

EMG recordings

EMG recordings of the FCR were performed during the median nerve and CMEP stimulation experimental paradigms. The EMG was recorded bilaterally from the FCR muscle through surface electrodes taped to the skin over the muscle (Trigno, Delsys, Natick, MA, USA). Placement of the EMG sensors (at one-third of the distance between the medial epicondyle and radial styloid) was done with respect to anatomical landmarks and photographed to ensure similar electrode positions between individuals. To optimize the quality of the EMG signal, the skin over the FCR muscle was cleaned with 70% isopropyl alcohol prep pads before electrode placement to reduce impedance at the skin–electrode interface. The collected signals were amplified, bandpass-filtered (20–500Hz), and sampled at 1926Hz for online and a posteriori analysis with a custom MATLAB program (R2018b, MathWorks, Natick, MA, USA).

Cervicomedullary stimulation

Cervicomedullary stimulation was used to activate corticospinal axons subcortically at the cervicomedullary junction as shown in Figure 1. The CST was stimulated by passing an electrical current (200ms duration, 5–95mA amplitude, 400V maximum output voltage) between a cup electrode (cathode) placed behind the contralateral mastoid process at the cervicomedullary level and the ipsilateral deltoid (anode) on the shoulder (Figure 1a). This experimental setup is similar to the methods used in recent MEP investigations.^17,18 With this electrode arrangement, CMEPs were recorded in the bilateral FCR muscles but were largely more prominent in the contralateral FCR muscle at lower stimulus intensities. The presence and onset of ipsilateral CMEP responses were monitored in all participants. To minimize discomfort, all CMEP recordings were performed while participants performed a low-level isometric wrist flexion contraction. Placement of the cathode at the contralateral mastoid process allowed us to minimize ipsilateral peripheral nerve depolarization, which occurs closer to the cathode.¹⁵ In addition, cathode placement allowed for larger CMEP responses at lower stimulus intensities, which increased the tolerability of the stimulation protocol.

Figure 1:

(a) Assessment of corticospinal tract activation of spinal motoneurons in adults with cerebral palsy and typical adults using electrical medullary stimulation. The location of stimulating electrodes at the contralateral mastoid process (cathode) and ipsilateral deltoid (anode) is shown. The location of the electromyography (EMG) on the flexor carpi radialis of the contralateral arm is also shown. (b) The putative activated corticospinal pathway originating in the contralateral sensorimotor cortices and terminating at the ipsilateral cervical spinal motoneurons to produce the flexor carpi radialis cervicomedullary motor evoked potential response is shown.

CMEP stimulus recruitment curves were obtained by adjusting stimulus intensities in 5mA steps and toward a 95mA maximum. FCR CMEP onset latencies of approximately 8ms to 10ms were considered acceptable. To ensure that responses were elicited by direct corticospinal activation, CMEP onset latency was monitored throughout the study because a reduction in onset latency of approximately 2ms with increasing stimulus intensity represents direct, postsynaptic activation of motoneurons at the cervical roots. Lastly, CMEP amplitude was expressed as a percentage of Mmax for the outcome variable.

Clinical outcome assessments

All participants completed the standardized Purdue Pegboard test of fine motor dexterity¹⁹ (Lafayette Instrument Company, Lafayette, IN, USA). This was measured from the tested limb as the total number of pegs picked up and placed into the holes of a board within a 30-second period. All participants also completed the standardized Box and Blocks test of hand dexterity.^20,21 Participants were instructed to move as many blocks as possible from one compartment across to another within a 60-second period using the tested limb. One trial was attempted with the nondominant hand (tested limb) for these assessments.

The test of arm selective control²² was also performed in all individuals with CP to measure the selective voluntary motor control of the shoulder, elbow, and fingers. These assessments were performed with the participant sitting on a bench or, if necessary, in their wheelchair. The examiner demonstrated the test requirements for the starting positions and joint movements. For each joint movement, the examiner passively moved the tested upper-extremity joint to assess the full range of motion and demonstrate the desired movement. Then, participants were instructed to actively move the joint in a similar manner using a 3-second verbal count. Each position was video-recorded and scored by an independent assessor (KS) as 0 (unable), 1 (impaired), or 2 (intact selective voluntary motor control), with a total maximum score for each tested limb of 16.

Lastly, the Neuro-QoL UEF Scale²³ was completed by all participants with CP. This scale is a part of the Patient-Reported Outcomes Measurement Information System (PROMIS) self-reported measures and examines ability across fine motor activities and activities of daily living involving digital, manual, and reach-related functions.²³ We administered the eight-item Neuro-QoL UEF Scale to adults with CP in this study.

Statistical analyses

For the respective outcome measures, equality of variance was initially checked using Levene’s equality of variance testing. A two-way mixed-factor analysis of variance was used to examine the effect of CMEP amplitude at the respective stimulation intensities and group (typical adults or adults with CP) differences. Sixteen points of stimulation intensity were used in this analysis. Separate independent t-tests were used to assess group differences in the Purdue Pegboard and Box and Box tests. A Shapiro–Wilk test revealed that CMEP size was not normally distributed, so nonparametric Spearman’s rank correlation coefficients were calculated to examine the relationship between clinical assessments and CMEP size in adults with CP. In all tests, statistical significance was assumed if p < 0.05. All statistical analyses were performed in JASP 0.14 (https://jasp-stats.org/) and Prism 9 (GraphPad Software, Boston, MA, USA); the results are presented as the mean plus or minus the standard error of the mean.

RESULTS

CMEP responses

Participants with CP and controls tolerated cervical stimulation at the levels used in this study with no reported adverse events. Exemplary CMEP responses at a representative low (60mA) and high (95mA) intensity stimulation level for representative controls and adults with CP are shown in Figure 2a,b. The figure shows that responses were larger with a higher stimulus intensity in both controls and adults with CP, suggesting that CMEP size was modulated by stimulus intensity in both groups.

Figure 2:

Cervicomedullary motor evoked potential (CMEP) sizes are reduced at higher stimulus intensities in adults with cerebral palsy (CP) compared to typical adults. (a) Raw contralateral CMEP responses from an exemplary typical adult. (b) Adult with CP. Exemplary raw traces at low (60mA) and high (95mA) stimulus intensities are shown. Electromyography (EMG) amplitude from the flexor carpi radialis (FCR) muscle is shown on the y-axis and time (ms) is denoted on the x-axis, with 0ms defined as the onset of the stimulus artifact. (c) The maximum compound action potential (Mmax) plus or minus the standard error of the mean of the contralateral EMG responses for each stimulus intensity between 20mA and 95mA at 5mA intervals. The data are normalized to a percentage of Mmax. As shown, CMEP size was significantly larger at stimulus intensities of 90mA and 95mA in typical adults (controls) than in adults with CP. *p < 0.05.

Statistical analysis of the stimulus–response curve results revealed an effect for stimulus intensity (p < 0.001) but not for group (p = 0.184). There was also a stimulus intensity × group interaction (p < 0.001). As shown in Figure 2c, the post hoc analysis using the Holm–Šidák multiple-comparisons correction revealed that adults with CP had a lower CMEP amplitude than controls at stimulus intensities of 90mA (p = 0.032, with a mean difference of 0.117) and 95mA (p = 0.001, with a mean difference of 0.160). The 95% confidence interval (CI) of the mean difference was −0.020 to 0.010. Subgroup analysis was performed with only adults with bilateral CP (n = 10) and controls, with a difference between means of 0.054 (SEM = 0.031). The 95% CI of the mean difference was −0.011 to 0.118. Subgroup post hoc analysis showed similar effects at 90mA (p = 0.013, with a mean difference of 0.140) and 95mA (p < 0.001, with a mean difference of 0.179). These results suggest that adults with CP (both bilateral and unilateral presentation) were unable to increase CMEP size at higher stimulus intensities.

Relationships with manual dexterity

Adults with CP moved fewer blocks on the Box and Blocks test (controls: 70.38 [SEM = 0.45] blocks; individuals with CP: 37.07 [SEM = 1.37] blocks; p < 0.001, 95% CI = 21.5–45.2, mean difference = 33.3), and placed fewer pegs on the Purdue Pegboard test (controls: 15.39 [SEM = 0.08] pegs; individuals with CP: 6.27 [SEM = 0.37] pegs; p < 0.001, 95% CI = 5.95–12.3, and mean difference = 9.12) compared to controls. As seen in Figure 3a, there was a significantly positive rank-order relationship between maximum CMEP amplitude and the number of blocks moved for adults with CP (rho = 0.625, p = 0.025). There was also a significantly positive rank-order relationship between the number of pegs placed and maximum CMEP size for adults with CP (Figure 3b; rho = 0.701, p = 0.010). These results suggest that adults with CP who had a larger maximum CMEP amplitude tended to have better manual dexterity.

Figure 3:

The size of the cervicomedullary motor evoked potentials (CMEPs) is related to motor function in adults with cerebral palsy. The relationship between flexor carpi radialis CMEP size and clinical outcomes in adults with cerebral palsy (CP) is shown. (a) Rank-order correlation between the number of blocks moved on the Box and Blocks test and the average CMEP size for adults with CP. The rank-order of CMEP size is shown on the x-axis, while the rank-order of the number of blocks moved is shown on the y-axis. (b) Rank-order correlation between the number of pegs placed on the Purdue Pegboard test and the average CMEP size for adults with CP. The rank-order of CMEP size is shown on the x-axis, while the rank-order of the number of pegs placed is shown on the y-axis. As shown, there were significantly positive correlations between CMEP size and each participant’s Box and Blocks (rho = 0.625; p = 0.025) and Pegboard (rho = 0.701; p = 0.010) performance. (c) Rank-order correlation between the upper extremity function (UEF) survey score and the average CMEP size for adults with CP. The rank-order of CMEP size is shown on the x-axis, while the rank-order of the UEF score is shown on the y-axis. As shown, there were significantly positive correlations between CMEP size and the participant’s UEF score (rho = 0.761; p = 0.009).

Relationships with selective motor control

On the test of arm selective control, adults with CP scored an average of 9.87 (SEM = 0.17) out of a maximum score of 16. We observed a significantly positive rank-order relationship between maximum CMEP amplitude and arm selective control scores (rho = 0.731, p = 0.007). On the Neuro-QoL UEF Scale, adults with CP scored an average T-score of 41.69 (SEM = 0.62). For comparison, the normative value²³ for this scale was previously reported to be 47.9 (SEM = 0.27). As shown in Figure 3c, there was a significantly positive rank-order relationship between maximum CMEP amplitude and the Neuro-QoL UEF Scale scores (rho = 0.761, p = 0.009). These correlations suggest that adults with CP who had larger maximum CMEP amplitudes tended to have better voluntary motor control and less difficulty with fine motor tasks and activities of daily living, although no causal relationship can be established with these data.

DISCUSSION

The purpose of this study was to evaluate and compare the reorganization of corticospinal pathways at the cervical level in adults with CP. One of our primary findings was that adults with CP had smaller CMEP sizes at the highest stimulus intensities (90mA and 95mA) relative to controls. This finding suggests that there may be a lack of corticospinal activation of highest-threshold wrist flexor motor neurons in adults with CP. The literature has shown that CST physiology is strongly linked to deficits in fine motor function and selective motor control in adults with CP. An earlier study showed that facilitation of MEPs with voluntary contraction is impaired in adults with CP.⁷ In that study, stimulation at the motor cortical level was used, producing smaller responses at all levels of voluntary contraction. Our study focused on the spinal contribution to these responses and did not reveal any changes at lower stimulus intensities. This may suggest that the reduction in MEP amplitudes previously noted may be largely cortical in origin, with a spinal contribution being more important for activation of higher-threshold spinal motoneurons.

There is strong evidence that CMEPs arise from spinal motoneuron activation to the descending volleys elicited by the excitation of corticospinal axons.^14,24,25 These responses are largely monosynaptic in the upper limbs of humans.²⁶ The integrity of the CSTs is altered in individuals with CP^10,11,27,28 and this probably extends to the spinal cord level. Thus, lack of facilitation of the CMEP response at higher stimulus intensities may be the result of axonal or synaptic changes in the higher-threshold CSTs and their connections to cervical motoneurons in adults with CP.

The diminished CMEP responses in adults with CP at higher stimulus intensities could also reflect changes in the excitability of the cervical spinal motor pools. Intriguingly, recent neuroimaging work indicated that the cervical gray matter cross-sectional area is smaller in adults with CP²⁹ and the number of motor units in the thenar muscles of adults with CP is lower than that seen in controls.³⁰ A reduction in the excitability of cervical motoneurons through CST stimulation could be linked to or contribute to changes in the number of motoneurons available to enable the voluntary contraction of upper-extremity muscles in adults with CP. Given the variability of the responses in both individuals with CP and controls, the strength of the CMEP response to stimulation may be linked to the presence or absence of upper-limb involvement or to handedness. Future studies may be needed to examine this further.

Relating CMEP amplitude to motor impairment in adults with CP

For adults with CP, the diminished CMEP amplitude at the higher stimulus intensities was linked to a reduced performance on the Box and Blocks and Purdue Pegboard tests. Previous studies indicated that individuals with CP are unable to fully activate their motor units and that higher levels of voluntary effort may be necessary to adequately increase the recruitment of motor units.^7,31 Our CMEP stimulus–response curves for adults with CP indicate that a reduced proportion of the spinal motor pools is activated by the CSTs, further fueling the premise that they are less likely to be able to fully activate wrist flexor motor units. This is partially linked with manual ability; diminished activation of the spinal cord motor pools with cervicomedullary stimulation was tightly connected with poor manual dexterity. Altogether, these findings suggest that reduced recruitment of upper-extremity motor units probably has a prominent role in the deficits in manual dexterity seen in adults with CP.

Diminished CMEP amplitude was also linked with greater impairments in upper-extremity selective motor control seen in participants with CP. Furthermore, our results show that those with either a unilateral, bilateral, or bilateral (quadriplegia) presentation exhibited deficits in upper-extremity selective motor control. These observations suggest that selective motor control impairments are probably ubiquitous in those with CP independent of presentation. Intriguingly, individuals with a lower CMEP amplitude also tended to have less upper-extremity selective motor control. This may reflect the significance of the CSTs in selective motor control in people with CP and suggests that activation of the spinal motoneurons by these tracts is possibly related to the noted deficits.

We also identified that diminished CMEP amplitude at higher stimulus intensities was linked with reductions in the self-reported ability to perform upper-extremity motor tasks. Importantly, individuals with a smaller CMEP amplitude also self-reported greater difficulties with fine motor tasks and activities of daily living. These results further support the overall picture that CST deficits at the spinal level may have a significant role in shaping daily manual function and overall quality of life in adults with CP.

Limitations

While stimulation of the CSTs at the cervicomedullary junction facilitates the evaluation of the corticospinal projections on the spinal motor pools, it is possible to have contaminated responses, particularly at higher stimulation levels.^14,26,32 This could include the indirect stimulation of the spinal motor pools through spinal interneurons or the direct stimulation of motoneurons or ventral roots.^14,26,32 However, the stimulation approach we used with the electrode positioning only on the contralateral cervicomedullary junction, along with lower levels of stimulation, were designed to limit this possibility.

Also, in this study, one of the individuals with CP had received botulinum neurotoxin A injection in the FCR muscle in the year leading up to the study. However, the vast majority (14 of 15) of our participants with CP did not have any major treatments that could alter their FCR muscle spinal circuitry (botulinum neurotoxin A, anti-spasticity medications, or dorsal root rhizotomies). We also tested only the more affected hand with the CMEP, Pegboard, and Box and Blocks tests in our individuals with CP. Furthermore, our participants attempted only one trial with the Pegboard and Box and Blocks tests. Finally, the results in this study were obtained from a small sample of individuals, which may potentially limit generalizability. Future studies using a larger sample with more extensive testing may be useful to study the role of the less affected hand in adults with CP and improve the strength of the relationships uncovered in the present study.

Conclusions

Adults with CP have smaller CMEPs at higher simulation levels compared with controls. This suggests that adults with CP are less capable of activating the higher-threshold spinal motoneurons that activate the type II motor units governing the production of larger muscular forces. The reduction in the excitation of the higher-threshold spinal motoneurons is connected with the extent of the altered upper-extremity motor performance. Our findings are largely consistent with the idea that upper motor neuron syndromes like CP are characterized by a recruitment disorder, a plateauing of recruitment, and the inability to use it correctly in activities of fine motor control, such as manual dexterity.

It is unclear whether alterations to the CSTs themselves or typical development of higher-threshold spinal motoneurons^33,34 are responsible for the deficits in motor function in adults with CP from the results obtained in this study. Comparisons with both adults with acquired stroke and young children with CP could shed some light on this question. Furthermore, simultaneous CMEP and transcranial magnetic stimulation of the motor cortex^7,35 should be used in future studies to unravel the spinal motoneuron contribution to the changes in CST activation in these populations.

Given the implication of lowered activation of higher-threshold motoneurons in adults with CP, it is feasible that neuromodulation targeted toward the spinal cord could be used to increase overall levels of upper-limb function in this population. Recent promising advances in spinal neuromodulation^36,37 in people with CP may be directly applied and tested using CMEP responses. Furthermore, we suggest that physical therapy approaches that target activities necessitating the activation of higher-threshold motoneurons, such as strength or power training, might ignite beneficial neuroplasticity at the spinal cord level.³⁸ Future studies in adults with CP using these approaches in combination with spinal neurophysiological assessments may be fruitful.

Supplementary Material

Table S1

Table S1: Response at different stimulus levels for individuals with cerebral palsy who were included in this study.

What this paper adds.

• Spinal cervicomedullary motor evoked potentials responses were diminished at higher stimulation levels.

• Changes were linked to differences in upper-limb function.

• Overall, the findings point toward spinal neuromodulation as a therapeutic target in this population.

ABBREVIATIONS

CMEP

cervicomedullary motor evoked potential

CST

corticospinal tract

FCR

flexor carpi radialis

MEP

motor evoked potential

Mmax

maximum compound muscle action potential

UEF

upper-extremity function