Edited by: Ruoli Wang, KTH Royal Institute of Technology, Sweden
Reviewed by: Francesco Cenni, University of Jyväskylä, Finland; Hu Xiaoling, Hong Kong Polytechnic University, Hong Kong
This article was submitted to Biomechanics, a section of the journal Frontiers in Bioengineering and Biotechnology
This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
There is a significant influence of muscle fatigue on the coupling of antagonistic muscles while patients with post-stroke spasticity are characterized by abnormal antagonistic muscle coactivation activities. This study was designed to verify whether the coupling of antagonistic muscles in patients with post-stroke spasticity is influenced by muscle fatigue. Ten patients with chronic hemipare and spasticity and 12 healthy adults were recruited to participate in this study. Each participant performed a fatiguing isometric elbow flexion of the paretic side or right limb at 30% maximal voluntary contraction (MVC) level until exhaustion while surface electromyographic (sEMG) signals were collected from the biceps brachii (BB) and triceps brachii (TB) muscles during the sustained contraction. sEMG signals were divided into the first (minimal fatigue) and second halves (severe fatigue) of the contraction. The power and coherence between the sEMG signals of the BB and TB in the alpha (8–12 Hz), beta (15–35 Hz), and gamma (35–60 Hz) frequency bands associated with minimal fatigue and severe fatigue were calculated. The coactivation ratio of the antagonistic TB muscle was also determined during the sustained fatiguing contraction. The results demonstrated that there was a significant decrease in maximal torque during the post-fatigue contraction compared to that during the pre-fatigue contraction in both stroke and healthy group. In the stroke group, EMG-EMG coherence between the BB and TB in the alpha and beta frequency bands was significantly increased in severe fatigue compared to minimal fatigue, while coactivation of antagonistic muscle increased progressively during the sustained fatiguing contraction. In the healthy group, coactivation of the antagonistic muscle showed no significant changes during the fatiguing contraction and no significant coherence was found in the alpha, beta and gamma frequency bands between the first and second halves of the contraction. Therefore, the muscle fatigue significantly increases the coupling of antagonistic muscles in patients with post-stroke spasticity, which may be related to the increased common corticospinal drive from motor cortex to the antagonistic muscles. The increase in antagonistic muscle coupling induced by muscle fatigue may provide suggestions for the design of training program for patients with post-stroke spasticity.
The ability of the central nervous system to appropriately control agonistic and antagonistic muscles is a key mechanism to maintain body coordination during human voluntary movements and postural adjustments, which can be damaged by post-stroke spasticity (
Neuromuscular fatigue has the tendency to arise when performing physical activities after stroke, which has received little attention in clinical rehabilitation research (
Frequency-based electromyography (EMG) assessment has been used as an effective tool to explore the motor control strategy of movement. In particular, EMG–EMG coherence analysis is a method of reflecting the similarity of two signals in frequency domain and has been applied to evaluate the common synaptic input to co-contracting muscles or to two parts of the same muscle from branches of last order neurons to spinal motoneurons (
Based on these studies, we wondered whether the coupling and controlling strategies of antagonistic muscles in patients with post-stroke spasticity can be influenced by muscle fatigue. The current study was designed to verify this hypothesis. EMG activities of the biceps brachii (BB) and triceps brachii (TB) were recorded from both patients with post-stroke spasticity and healthy subjects during fatiguing isometric elbow flexion contraction. The amount of antagonist activation and EMG-EMG coherence between the BB and TB were compared between the first (stage 1 with minimal fatigue) and second halves (stage 2 with severe fatigue) of the contraction.
Twelve healthy adults (12 men; age: 21.1 ± 3.4 years; height: 177.6 ± 7.7 cm; weight: 66.25 ± 9.5 kg) and 10 patients with chronic hemiparesis (7 men and 3 women; age: 60.5 ± 9.1 years; height: 169.9 ± 8.8 cm; weight: 69.9 ± 11.3 kg) participated in this study. All healthy subjects reported no known neuromusculoskeletal impairments and were right-handed. The inclusion criteria for the patients with stroke were as follows: (1) hemiplegia secondary to an ischemic or hemorrhagic stroke, (2) at least 6 months post-stroke, (3) residual voluntary elbow flexion force, (4) spastic hypertonia in elbow flexors of the impaired side, rated as Modified Ashworth Scale (MAS) less than 3, and (5) able to understand and follow instructions related to the experiment. All subjects were fully informed of all procedures and risks associated with the experimental tests before providing their informed written consent to participate. The experiment was approved by the Ethics Committee of Tongji University.
The experiment consisted of three motor tasks. First, participants performed isometric maximal elbow flexion contraction test for three times with 5 min rest after each test, in order to acquire the maximal isometric elbow flexion torque of each subject without muscle fatigue. Second, after 5-min rest, each participant performed a sustained elbow flexion fatiguing contraction at 30% maximal elbow flexion force for as long time as possible. Lastly, as soon as the fatiguing elbow flexion contraction task was complete, the subjects were instructed to perform another isometric maximal elbow flexion contraction test. All motor tasks were finished with the affected limb.
In the experiment, the participants sat with the upper arm vertically placed, and the elbow angle was maintained at 90°. The right forearm was positioned parallel to the ground and supinated. The participants were instructed to maintain their posture by flexing the elbow with the elbow joint maintained as close as possible to 90° until they experienced exhaustion and were no longer able to continue the contraction. The participants were verbally vigorously encouraged to continue the sustained contraction for as long as possible. During the sustained contraction, surface EMG signals of the BB and TB muscles of the affected limb were recorded. The experimental setup has been depicted in
Experimental setup. Participants sat with the upper arm vertically placed and the elbow angle keeping in 90°. The right forearm was positioned parallel to the ground and supinated. During the sustained elbow flexion contraction, a weight was suspended from the distal part of the right forearm so as to produce a target force of 30% maximal elbow flexion force
The elbow flexion torque was measured using a torque sensor (TRS-500, Transducer Techniques, Temecula, CA). The sensor was located in line with the center of the rotation of the active elbow joint. Surface electromyographic signals were recorded by bipolar surface electrodes of NeuroScan system (NeuroScan Inc., El Paso, TX). Two pairs of electrodes were placed over the belly of the right BB and TB muscles with a center-to-center inter-electrode distance of 2 cm. A common reference electrode was placed to the left of the processus mastoideus. The skin was shaved and cleaned with alcohol wipes before filling the electrodes with conducting gel and fixing. Medical adhesive tape and plastic casts were applied to fix the electrodes. The EMG signals were amplified, band-pass filtered (3–1000 Hz), digitized (2000 samples/s), and acquired by using the NeuroScan system.
Data were analyzed off-line by custom-written programs using MATLAB R2016a software (Mathworks Inc., Natick, MA).
The raw EMG signals recorded from both the BB and TB muscles were band-pass filtered offline at 5–500 Hz using a fourth order zero-phase-shift Butterworth filter. Then, the filtered surface EMG (sEMG) signals of both the BB and TB muscles were further full-wave rectified using the Hilbert transform according to the research of
The EMG median frequency and power spectrum based on Fourier transform of filtered-only EMG signals were calculated for non-overlapping 2.048-s epochs during the sustained fatiguing contraction and were averaged for each epoch during the first and second halves of the sustained elbow flexion contraction. The EMG power in the alpha (8–12 Hz), beta (15–35 Hz), and gamma (35–60 Hz) frequency bands were acquired during the first and second halves of the sustained elbow flexion contraction.
The coactivation ratio of the antagonistic TB muscle, described as the quotient of the TB EMG amplitude divided by the sum of the BB and TB EMG amplitudes, was also calculated for non-overlapping 2.048-s epochs during the sustained fatiguing contraction and time normalized to 20 points for each subject and then averaged for all subjects.
The filtered and rectified sEMG signals of each participant were equally divided into two segments: the first and second halves of the contraction. Coherence analysis was conducted on each half segment between EMG signals recorded from the BB and TB muscles. The coherence was calculated for a segment length of 2048 samples with 50% overlap, using a Hanning window, as suggested by previous research.
The magnitude squared coherence, Cxy(f) between the two EMG signals (rectified EMG signals) recorded from the BB and TB muscles, x(t) and y(t), for a given frequency f was calculated as follows:
where Sxy(f) is the cross spectrum and Sxx(f) and Syy(f) are the auto spectra of x(t) and y(t), respectively.
The significance level of the coherence spectrum was calculated based on the methods described by
Statistical analysis was performed using IBM SPSS 20.0 (SPSS Inc., Chicago, IL). Normality was tested using the Kolmogorov–Smirnov test. A repeated-measures general linear model was used to determine the difference between the maximal torque and median frequency. Pearson’s cross-correlation analysis was employed to determine the relationship between the coactivation ratio of antagonistic muscles and the contraction duration time. The non-parametric Wilcoxon test was adopted to test the difference in EMG power and coherence between the first and second halves of the contraction. All significance thresholds were set at α = 0.05.
The fatiguing contraction times of the stroke group and healthy group subjects were averages of 220.66 ± 109.19 s and 258.58 ± 51.13 s, respectively. There was no significant difference between the fatiguing contraction times of the stroke and healthy group subjects (
Comparison of maximal torque between the pre- and post-fatigue
Changes in the average coactivation ratio among stroke and healthy subjects for each 5% duration times are shown in
Changes of average coactivation ratio plotted as a percentage of contraction time during the sustained fatiguing contraction. The coactivation ratio was calculated as quotient of the antagonistic muscle TB EMG amplitude divided by the sum of agonistic muscle BB EMG amplitude and TB EMG amplitude.
The average EMG power of the BB and TB muscles in the alpha, beta and gamma frequency bands during the first and second halves of the contraction in the stroke and healthy group were presented in
Comparisons of the average EMG power of BB
In order to present the characteristics of the coherence spectra and provide clues to verify the impact level of cross-talk on the results, representative coherence spectra of two subjects as well as pooled coherence spectra of all the subjects in the stroke and healthy group were presented in
Representative (subject 4, 7, and 10) and pooled coherence spectra of the BB and TB sEMG signals during the first (thick line) and second (thin line) halves of the fatiguing contraction.
Comparisons of significant coherence integral in the alpha, beta, and gamma frequency bands during the first and second halves of the fatiguing contraction were presented in
Comparisons of significant coherence integral for the stroke
In the present study, we examined the influence of fatigue on the antagonistic muscle coupling of limbs with spasticity induced by stroke. The sEMG activation and coherence between antagonist muscles were compared between the first and second halves of the 30% maximal-level sustained fatiguing isometric elbow flexion contraction between patients with post-stroke spasticity and healthy adults. We found that the intermuscular coherences between the antagonistic BB and TB muscles in the alpha and beta frequency bands were significantly increased during the second half of the contraction compared to those during the first half of the contraction in the stroke group. As far as we know, this is the first study conducted to examine the effect of fatigue on antagonistic muscle coupling of limbs with stroke-induced spasticity.
It may be argued that the reliability of the results may have been reduced by cross-talk contamination, which has been widely studied in relevant studies of antagonist muscle coactivation activities (
In this study, the maximal torque of elbow flexion decreased significantly during the post-fatigue contraction compared to that during the pre-fatigue contraction, while the EMG median frequency of the BB muscle during the second half of the contraction showed a significant reduction compared to that during the first half of the contraction. This demonstrated that muscle fatigue of the agonist muscle BB occurred during the sustained elbow flexion contraction. In previous researches, it has been found that post-stroke spasticity increased force inaccuracy and variability and thus increase the movement instability (
EMG-EMG coherence in the beta band has been suggested to be closely related to the common corticospinal drive from the motor cortex to the muscles (
However, no significant coherence was found between the first and second halves of the fatiguing contraction in healthy adults in the current study, which is inconsistent with the results acquired in sustained isometric fatiguing contraction with 20% MVC force level in previous research (
In this study, we have mainly focused on the fatigue-related EMG coupling properties of antagonistic muscles in the population of patients with post-stroke spasticity. As a limitation of this study, the stroke survivors without spasticity have not been acquired and compared. This makes it hard to explicitly demonstrate the direct correlation between spasticity and muscle coupling. Actually, the results in the current study may be close related to cortically originated muscular discoordination, or motor recruitment, rather than spasticity in passive motion. Therefore, whether spasticity may significant influence EMG coupling of antagonistic muscles in stroke patients still need further research.
EMG synchrony within the Piper frequency band is considered a biomarker of functional corticomotor integrity (
A significant increase in intermuscular coherence in the beta frequency band between the antagonistic elbow muscles was observed during the second half of the 30% maximal-level sustained isometric contraction compared to that during the first half of the contraction, indicating that muscle fatigue significantly increases the coupling of antagonistic muscles in patients with post-stroke spasticity, which may be related to the increased common corticospinal drive from the motor cortex to the antagonistic muscles. The increase in antagonistic muscle coupling induced by muscle fatigue may provide suggestions for the design of training programs for patients with post-stroke spasticity.
The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.
The studies involving human participants were reviewed and approved by the Ethics Committee of the Tongji University. The patients/participants provided their written informed consent to participate in this study.
L-JW, X-MY, and W-XN conceived and designed the study. S-JH and X-MY recruited the subjects and collected the basic characteristics of the subjects. X-MY performed the clinical assessment. L-JW, Q-NS, CW, and HY performed the experiments. L-JW, S-JH, X-MY, and W-XN made a contribution to the data analysis. L-JW wrote this manuscript. All authors contributed to the article and approved the submitted version.
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
We acknowledge Yan Lu and Honglin Wang for the recruitment of study participants; Fanhui Zheng for assistance with EEG experiments. We would like to thank all the participants of this study.