A comparison of motor control characteristics of the dominant and non-dominant arms in response to assistive force during unilateral task

Abstract

BACKGROUND:

In recent years, power assist suits have been used in nursing and rehabilitation scenarios. The external assistive force might disturb the user’s motor control system during the process of assisted training using power assist suit, thus affect the progress of rehabilitation. With the consideration of the non-negligible physiological differences between upper limbs, this study focused on the physiological responses of the two arms against the external assistive force.

OBJECTIVE:

To investigate and contrast the impact of assistive force on dominant and non-dominant arms during unilateral isometric elbow flexion.

METHODS:

Participants were instructed to adjust force exertion to a target value based on the visual feedback. Task performances including muscle activity of agonist and antagonist muscles, force steadiness and rated perceived exertion were evaluated at multiple workload and assistive load conditions.

RESULTS:

No significant differences in muscle activity of agonist and antagonist muscles between the two arms. In contrast, the dominant arm showed a higher assist efficiency at a low assistive level, whereas the non-dominant arm had a lower level of force fluctuation during a unilateral force matching task.

CONCLUSIONS:

Both arms could utilize the assistive force to reduce muscle activity. However, the two arms showed different abilities in response to external assistive force. This indicates distinct motor control strategies for each arm and implicates the necessity of side-individualized rehabilitative approach for achieving a better training effect.

Keywords

Unilateral isometric elbow flexion dominant/non-dominant arm muscle activity force steadiness

1. Background

The power assist suit (PAS), a wearable machine that augments human physical capabilities, has been applied in nursing and rehabilitation scenarios [1]. As a complementary tool to therapist, PAS can be integrated to drive the impaired limb in rehabilitation training [2] or be used as an assistive device in daily life [3]. This kind of motion assistive device will be more widely acceptable in the near future. Recently, research on PAS has focused not only on mechanical development but also on its clinical efficacy. Studies that compared the long-term effects of robotic therapies with conventional occupational therapy indicated similar levels of improvement in motor control capabilities of the two training methods [4], thus confirming the effectiveness and applicability of PAS-based training. Nevertheless, most of these studies mainly focused on the training effect at the final stage, and research on human-PAS interactions during operating progress is still insufficient.

Humans are capable of modulating physical output based on the feedback of the muscle tension and their own experiences. However, when an external assistive force is applied, the physical output could be different from what was expected. This gap might disturb the motor control system [5], influence the training effect and prolong the training progress. Accordingly, such human physiological interactions with robotic assistance devices must be considerate during the development of PAS.

This research focuses on the PAS for the upper limbs, since high functioning upper limbs are crucial for quality of life. When performing daily behaviors, humans preferentially use one arm over the other for tasks such as opening a drawer or using a broom. Daily preferential use was shown to alter physiological and mechanical properties of skeletal muscle. It has been reported that during submaximal isometric flexion of the first dorsal interosseous muscle, motor units in the dominant hand show lower average firing rates and recruitment thresholds compared to the non-dominant hand, which suggest higher mechanical effectiveness of motor units in the dominant hand [6]. Additionally, the dominant limb shows stronger isometric muscular strength and better skill performance and learning rate due to preferred use over time [7, 8]. Moreover, from the perspective of the motor control mechanism, movements of the dominant hand are controlled mainly by the brain hemisphere contralateral to that hand, whereas the non-dominant hand is controlled by both hemispheres [9, 10]. Because of interhemispheric transmission delays, the non-dominant arm has disadvantage in performance, such as movement speed and accuracy, compared to the dominant arm [11].

However, some studies indicated that dominant and non-dominant arms have different specialties during unilateral movements due to their different control strategies. Functional magnetic resonance imaging (fMRI) showed that the right arm is controlled by the left motor cortex, which specializes in dynamic control and predictable situations, while the left arm is controlled by the right motor cortex, which specializes in impedance control and unpredictable situations [9, 12]. Correspondingly, studies comparing the performance of the two arms during target reaching movements found lower average muscle activity in the dominant arm, which suggests that the dominant arm more effectively anticipates task dynamics; the non-dominant arm more accurately compensated for an unexpected load, indicating that the non-dominant arm specializes in the control of the steady-state position [13, 14].

It is still difficult to determine whether the dominant arm is superior, or if the two arms have advantage in specific aspects of control, since the underlying neural mechanisms of handedness remain incompletely understood. Notwithstanding this lack of data, task capability and motor control strategies of the two arms are distinct to some degree, which might cause a difference in physiological responses to external assistive force between the two arms. On the other hand, isometric exercise, which movements are simple and applicable to all age groups, has been proved to be effective to improve ability in muscular-neuron disease sufferers [15, 16]. Therefore, the first objective was to investigate the physiological adaptability during unilateral isometric elbow flexion of each dominant and non-dominant arm against assistive force. The second objective was to evaluate bilateral arm performance through analysis of electromyography of agonist and antagonist muscles and force exertion steadiness. The comparison between the dominant and non-dominant arms may give insight into the training strategy of PAS-assisted rehabilitation.

2. Materials and methods

2.1 Experimental participants

Thirteen healthy male university students were recruited (Table 1). Handedness of each participant was assessed using the Edinburgh Handedness Inventory [17]. All participants gave written informed consent and the experiment was approved by the Ethics Committee of the Faculty of Design, Kyushu University, Japan.

Table 1
Characteristics of participants ( $n=$ 13)

	The number of participants
Left-handed	2
Right-handed	11
	Mean $\pm$ SD ${}^{\ast}$
Age (years)	22.6 $\pm$ 1.1
Height (cm)	171.9 $\pm$ 6.7
Weight (kg)	65.2 $\pm$ 9.8
Dominant arm (cm)
Forearm length	31.9 $\pm$ 1.5
Upper arm length	23.5 $\pm$ 1.2
Non-dominant arm (cm)
Forearm length	32.0 $\pm$ 1.4
Upper arm length	23.6 $\pm$ 1.3

${}^{*}$ SD: Standard deviation.

Figure 1.

Experimental set-up for unilateral isometric elbow flexion with assistance.

2.2 Experimental setup

Figure 1 illustrates the experimental setup. Participants sat upright in an armless chair facing a monitor. The value of force exertion was shown on the monitor in real time, and each participant was instructed to adjust it to a target value. The dominant or the non-dominant arm was positioned at 90 ${}^{\circ}$ elbow flexion with the palm horizontally in supination.

A strap was fastened on the wrist to connect the forearm to a tension sensor that acted as workload and force measurement. Another strap was fixed in the middle of the length of the radius to support the forearm by connecting the forearm to assistive load through a fixed pulley system [5]. The height of the monitor was adjusted to eye level, and each participant affirmed that the value of force displayed on the monitor was clearly visible. A barrier was placed between participant and assistive load to prevent participants from anticipating the degree of assistance.

2.3 Experimental conditions

First, the maximum force that the arm can generate during isometric elbow flexion was identified. Workload conditions were determined based on the maximum force (20% and 40% maximum force), and four levels of assistive load were determined based on the workload conditions equal to 0%, 33%, 67%, and 100% workload. With the combination of workload and assistive load conditions, participants performed 8 trials in counterbalanced order.

2.4 Experimental protocol

Participants were asked to avoid high-intensity exercise that could cause muscle fatigue and influence task performance before the experiment day. Before the experiment, participants changed clothes into short-sleeved shirts and short pants that had been prepared. The experimental session consisted of a maximal isometric voluntary contraction (MVC) task and submaximal force matching task. The MVC task included the measurement of the maximal isometric voluntary force (MVF) and the maximal amplitude of the surface electromyography (sEMG) of biceps brachii (BB) and triceps brachii (TB). The submaximal force matching task with assistance was performed immediately after MVC. The dominant and non-dominant arms performed the experimental tasks in random sequence.

2.4.1 MVC task

Participants were instructed to maintain their arm positioned at 90 ${}^{\circ}$ elbow flexion with the hand in supination. To measure the MVC of BB, a strap that connected to a ground-fixed tension sensor was set on the wrist. Each participant was asked to pull against the tension sensor with their maximum effort for 5 seconds. As for the TB, the participant’s forearm was positioned on the armrest, and they were then asked to push down against the armrest at maximum effort for 5 seconds. Three MVC trials were performed for each muscle; 60 seconds rest was interspersed between each trial to avoid fatigue. The maximal tension and sEMG amplitude were quantified by averaging the data over a 3-second interval. The mean tension of three trials was used to calculate the workload and assistive load, and the mean sEMG amplitude was used to normalize sEMG during the submaximal force matching task.

2.4.2 Force matching task with assistance

During the force matching task, participants were asked to maintain a workload that equals 20% and 40% MVF for 30 seconds with their elbow flexion maintained at 90 ${}^{\circ}$ based on the visual feedback displayed on the monitor. The force matching task contained two phases: (1) from 0 to 10 seconds, no assistive force was applied; (2) from 11 to 30 seconds, an assistive force was applied at four levels equal to 0%, 33%, 67%, and 100% workload (Fig. 2). Participants were required to perform a set of 8 trials (20%MVF workload with 0%, 33%, 67%, 100% assistive level and 40%MVF workload with 0%, 33%, 67%, 100% assistive level) in random order. The rest time between each trial was at least 60 seconds.

Figure 2.

Two phases of the force matching task with assistance.

2.5 Measurements

2.5.1 sEMG recordings

Before attaching the surface electrodes, the skin was gently scrubbed with an abrasive gel and cleaned with alcohol; an electrolytic gel was placed between the skin and the metallic part of the electrode to ensure signal quality. sEMG was recorded from BB and TB muscles. Disposable electrodes were placed above the belly of each muscle and the distance between the two electrodes was 2–3 cm. sEMG signals were amplified using a bio-amplifier (Bio Amp ML 132, AD Instruments Co., Australia).

2.5.2 Tension recordings

Tension during MVC and force matching tasks were recorded using a tension sensor (Attachment for tensile force, 30 kg, T.K.K. 1269f, Takei Scientific Co., Japan). The tension signal was amplified by a strain amp (strain amp TSA-210, T.K.K. 1268b, Takei Scientific Co., Japan).

Amplified analog signals of sEMG and tension were synchronized and converted into digital signals using PowerLab 16/30 (ML880, ADInstruments Co., Australia) and sampled at a frequency of 1 kHz with a band-pass filter of 10–500 Hz. These digital signals were imported to a personal computer and analyzed using LabChart v7.1.1 (AD Instruments Co., Australia) for further analysis. The sEMG signals were full-wave rectified and the amplitudes were normalized with respect to the MVC.

2.5.3 Perceived exertion

The rate of perceived exertion (RPE) during the force matching task was evaluated using Borg’s CR-10 Scale, which has a range from 0 to 10, with 0 being no muscle effort and 10 being maximum muscle effort. Participants rated the exertion directly after each experiment trial [18].

2.6 Data analysis

2.6.1 Muscle activity of BB and TB (%MVC)

The sEMG amplitude of BB and TB measured during the force matching trial was divided by the maximal sEMG of MVC for normalization. Muscle activity was evaluated by the normalized sEMG (%MVC).

2.6.2 Force steadiness (CV)

Fluctuation (force steadiness) in isometric force exertion was quantified by the coefficient of variation (CV).

2.6.3 Assist efficiency

Assist efficiency reflects the reduction rate of %MVC ${}_{\text{BB}}$ at each assistive condition compared with the unassisted condition (0% assistance). It was calculated by the following equation:

Assist efficiency $\displaystyle=\frac{\begin{array}[]{c}\text{\%MVC}_{\text{BB}}^{\text{Measured% }}\text{ at x\% assist}-\\ \text{\%MVC}_{\text{BB}}^{\text{Measured}}\text{ without assist}\\ \end{array}}{\begin{array}[]{c}\text{\%MVC}_{\text{BB}}^{\text{Theoretical}}% \text{ at x\% assist}-\\ \text{\%MVC}_{\text{BB}}^{\text{Theoretical}}\text{ without assist}\\ \end{array}}\times 100(\text{\%})$ $\displaystyle\text{x}=33,67,100$

Figure 3.

The changes of muscle activity of BB (biceps brachii) at 20%MVF workload: (a) for the dominant arm; (b) for the non-dominant arm.

For the theoretical value of %MVC ${}_{\text{BB}}$ : (1) the assistive force is equivalent to the workload at 100% assist, therefore the theoretical value of %MVC ${}_{\text{BB}}$ should be zero; (2) the theoretical %MVC ${}_{\text{BB}}$ should be equal to the measured %MVC ${}_{\text{BB}}$ when no assistive force is applied (0% assist); (3) for the 33% and 67% assistive levels, theoretical %MVC ${}_{\text{BB}}$ was determined with respect to the assistive level, according to the following equation:

$\displaystyle\text{\%MVC}_{\text{BB}}^{\text{Theoretical}}\text{ at x\% assist}$ $\displaystyle=({\text{\%MVC}_{\text{BB}}^{\text{Theoretical}}\text{without % assist}-0})$ $\displaystyle\times\left({1-\frac{\text{x}}{100}}\right)$ $\displaystyle\text{x}=33,67$

Where, “0” refers to the theoretical %MVC ${}_{\text{BB}}$ at 100% assistance.

2.6.4 Analytic period

The analytic period of this experiment was defined as a 10-second interval from the 20th second when the force exertion began to stabilize to the end of the force matching task (emphasized by the shaded region in Fig. 2). This interval was also referred to as the steady period.

Figure 4.

The changes of muscle activity of BB (biceps brachii) at 40%MVF workload: (a) for the dominant arm; (b) for the non-dominant arm.

Figure 5.

Mean muscle activity of BB (biceps brachii) of the dominant and non-dominant arms during the steady period: (a) for 20%MVF workload; (b) for 40%MVF workload.

Figure 6.

Assist efficiency of the dominant and non-dominant arms at different assistive levels: (a) for 20%MVF workload; (b) for 40%MVF workload.

Figure 7.

Mean muscle activity of TB (triceps brachii) of the dominant and non-dominant arms during the steady period: (a) for 20%MVF workload; (b) for 40%MVF workload.

Figure 8.

Perceived exertion of unilateral task performed by the dominant and non-dominant arms at different assistive levels: (a) for 20%MVF workload; (b) for 40%MVF workload.

Figure 9.

Steadiness (CV) of the dominant and non-dominant arms at different assistive levels: (a) for 20%MVF workload; (b) for 40%MVF workload.

2.7 Statistical analysis

All statistical analyses were performed using IBM SPSS (version 24.0 software, Cary, NC, USA). Three-way ANOVA (arm conditions $\times$ workload levels $\times$ assistive levels) with repeated measures on three factors was used to compare %MVC ${}_{\text{BB}}$ , %MVC ${}_{\text{TB}}$ , RPE, assist efficiency, and CV. When the main effect and interaction were significant, Bonferroni-corrected pairwise comparison was used as a post hoc test to examine the mean differences in the factors. The significance level of all statistical tests was set at $p<$ 0.05. All data were reported as means $\pm$ standard deviations.

3. Results

3.1 %MVC ${}_{\text{BB}}$ of dominant and non-dominant arms

The changes of %MVC ${}_{\text{BB}}$ at different assistive levels are shown in Fig. 3 (20%MVF workload) and Fig. 4 (40%MVF workload). %MVC ${}_{\text{BB}}$ of the dominant and non-dominant arms changed with similar tendencies that decreased with the intensity of the assistive level at both workload conditions.

The mean %MVC ${}_{\text{BB}}$ during the steady period of the dominant and non-dominant arms are shown in Fig. 5. The result showed that workload ( $F$ [1, 12] $=$ 57.16, $p<$ 0.001) and assistive level ( $F$ [3, 36] $=$ 163.52, $p<$ 0.001) had a significant effect and a significant interaction ( $F$ [3, 36] $=$ 83.62, $p<$ 0.001). In contrast, the main effect of arm condition, as well as the interaction that concerned arm condition, were not significant.

Post hoc comparisons indicated significant difference among all assistive levels at both workloads. In general, at 20% and 40%MVF workloads, the %MVC ${}_{\text{BB}}$ of both arms decreased with increases in the assistive level (Fig. 5). At the same assistive level, the %MVC ${}_{\text{BB}}$ at 40%MVF workload was larger than that at 20%MVF workload.

3.2 The assist efficiency of dominant and non-dominant arms during the steady period

The results of assist efficiency of the two arms at two workload conditions are showed in Fig. 6. Significant main effects of workload ( $F$ [1, 12] $=$ 74.47, $p<$ 0.001), assistive level ( $F$ [2, 24] $=$ 27.86, $p<$ 0.001), and the workload $\times$ assistive level interaction ( $F$ [2, 24] $=$ 1.78, $p<$ 0.001) were detected. Assist efficiency of both dominant and non-dominant arms showed similar tendencies in response to increasing assistive levels, nevertheless, the tendencies were variable at the two workload conditions. At 20%MVF workload, assist efficiency of each arm was approximately the same at each assistive condition (Fig. 6a), whereas at 40%MVF workload, the assist efficiency of both arms decreased with increases in assistive levels. Most importantly, a significant assistive level $\times$ arm interaction ( $F$ [2, 24] $=$ 4.68, $p<$ 0.05) indicated a higher assist efficiency of the dominant arm compared with the non-dominant arm, especially at the 33% assistive level (Fig. 6).

3.3 %MVC ${}_{\text{TB}}$ of dominant and non-dominant arms

The mean %MVC ${}_{\text{TB}}$ of two arms during steady state are shown in Fig. 7. Similar to the result of %MVC ${}_{\text{BB}}$ , %MVC ${}_{\text{TB}}$ was only significantly affected by workload level ( $F$ [1, 12] $=$ 47.89, $p<$ 0.001), assistive level ( $F$ [3, 36] $=$ 80.25, $p<$ 0.001), and their interaction ( $F$ [3, 36] $=$ 40.88, $p<$ 0.001). The main effect of arm ( $F$ [1, 12] $=$ 0.02, $p=$ 0.881) and the interaction related to arm condition were not significant. %MVC ${}_{\text{TB}}$ of both arms decreased with the intensity of assistive level at the two workload conditions, which was similar to the results observed for %MVC ${}_{\text{BB}}$ (Fig. 7).

3.4 Perceived exertion

RPE of both dominant and non-dominant arms did not differ significantly ( $F$ [1, 12] $=$ 0.23, $p=$ 0.638). In contrast, RPE was significantly affected by workload ( $F$ [1, 12] $=$ 67.41, $p<$ 0.001) and assistive level ( $F$ [3, 36] $=$ 82.65, $p<$ 0.001). The results show that at both 20% and 40%MVF workloads, perceived exertion of unilateral task was alleviated when assistive level increased, whether it was performed by the dominant or non-dominant arm (Fig. 8). Participants perceived greater force exertion at 40%MVF workload compared with 20%MVF workload.

3.5 Force steadiness of dominant and non-dominant arms

Force steadiness was significantly affected by workload level ( $F$ [1, 12] $=$ 5.44, $p<$ 0.05), assistive level ( $F$ [3, 36] $=$ 24.94, $p<$ 0.001), and arm ( $F$ [1, 12] $=$ 5.77, $p<$ 0.05). In general, CV of both arms showed similar tendencies in response to increasing assistive level at both workloads. Nevertheless, the tendency was variable at each workload: at 20%MVF workload, CV of both arms was maintained at a similar level despite the increased assistive level (Fig. 9a). In contrast, at 40%MVF workload, CV of both arms decreased with increasing assistive level (Fig. 9b). Additionally, CV at 40%MVF workload was higher than that at 20%MVF workload, particularly at the 0% assistive level. It is noteworthy that CV of the non-dominant arm was generally smaller than that of the dominant arm.

4. Discussion

The purpose of this research was to contrast the task performance of dominant and non-dominant arms at different workload levels (20% and 40%MVF) with different assistive levels (0%, 33%, 67% and 100%) by analyzing the muscle activity of the biceps, triceps, and force variation of the two arms. According to dominant arm advantages in task performance [7, 8], we hypothesized that the dominant arm would have a better performance to accommodate the assistive force compared with the non-dominant arm.

4.1 A comparison of muscle activity of dominant and non-dominant arms

4.1.1 Muscle activity of biceps brachii

The agonist muscle of both arms showed similar responses when adapting to the assistive force: decreasing gradually after the assistive force was applied and stabilizing after several seconds. By visual inspection, there was no drastic changes in the %MVC ${}_{\text{BB}}$ of both arms at each assistive level during the steady period (Figs 3 and 4). Additionally, %MVC ${}_{\text{BB}}$ of the two arms decreased with the increase of assistive force at both workloads (Fig. 5). These results indicate that both dominant and non-dominant arms are capable of utilizing the assistive force to reduce the exertion of BB.

4.1.2 The assist efficiency

Assist efficiency, which reflects the reduction rate of %MVC ${}_{\text{BB}}$ at each assistive level compared with the unassisted condition, was expected to be approximately 100% at the 100% assistive level at both workloads since the assistive force counterbalanced the workload. Nevertheless, at both workloads, both arms retained muscle activity of 6%MVC ${}_{\text{BB}}$ at 100% assist, and the assist efficiency at 20%MVF and 40%MVF workloads were 71% and 87%, respectively (Fig. 6). These lower assist efficiencies could be due to co-contraction of the antagonist muscle. As reported in previous research [19], muscle activity associated with co-contraction of the BB and TB contribute to elbow stiffness during isometric elbow flexion. At assisted conditions, in order to prevent injury caused by sudden external force, participants would try to increase elbow joint stiffness by increasing muscular co-contraction, which required further contraction of BB, thus reducing the assist efficiency [5]. This result also suggests that it might be necessary for both dominant and non-dominant arms to maintain muscle activity of the primary working muscle at a certain level to prevent injury at a higher assistive level.

Most importantly, the results in this experiment showed that the assist efficiency of the dominant arm is significantly higher than that of the non-dominant arm at 33% assistance (Fig. 6). This lower efficiency of the non-dominant arm, especially at the low assistive level, could be due to motor control differences between the two arms. The explanation can be discussed as follows:

(1) Why was the laterality of assist efficiency only found at a low assistive level?

As reported in previous study, forceful voluntary isometric muscle contraction that last at least 5 seconds can bring about rapid changes in excitability in the motor cortex and spinal reflexes that last long after the cessation of the contraction, thus resulting in a prolonged involuntary contraction [20]. In the experiment trials, the application of intensive assistance could be regarded as cessation of the muscle contraction. Before the assistance can be applied, participants maintained 20% or 40%MVF isometric elbow flexion for 10 seconds, which might be intensive enough to induce an involuntary contraction of BB. Additionally, voluntary contraction of BB occurred during assistance to prevent injury caused by sudden assistive force. Consequently, after an intensive assistive force is applied, especially at 100% assistance, voluntary and involuntary muscle contractions might exist simultaneously. In contrast, for low assistance (33%), the assistive force might not be powerful enough to induce cessation of muscle contraction, therefore, there would primarily be voluntary muscle contraction throughout the trial. Moreover, the control of voluntary contraction would be affected by motor control differences between the arms (explained in detail subsequently). Thus, the laterality of assist efficiency was demonstrated at a low assistive level.

(2) Why was assist efficiency of the non-dominant arm lower than that of the dominant arm?

It has been suggested that voluntary movement of the dominant hand is controlled mainly by the brain hemisphere contralateral to that hand, whereas the non-dominant hand is controlled by both hemispheres [9, 10]. Due to interhemispheric transmission delays, the non-dominant arm has disadvantages in performance, such as motor control and coordination [6, 11]. Correspondingly, a previous study also reported greater muscle activity in the non-dominant arm compared with the dominant arm during dynamic movement [21]. Similarly, during isometric elbow flexion, the motor control of the non-dominant arm might not be as efficient as the dominant arm, thus resulting in extra contractions of BB when accommodating external assistive force.

Overall, both dominant and non-dominant arms showed the ability to utilize the external assistive force to reduce the muscle activity of BB. However, due to the performance disadvantages of motor control, the non-dominant arm reduced the muscle activity of BB less efficiently compared with the dominant arm, especially at a low assistive level.

4.1.3 Muscle activity of triceps brachii

During isometric contraction, both the agonist and antagonist muscles contract at the same time to maintain joint stability [22, 23]. Similarly, in this experiment, the antagonist (TB) was co-activated with the agonist (BB) at an unassisted stage, and the %MVC ${}_{\text{TB}}$ was decreased with the intensity of assistive force, which was a similar tendency as that being observed in %MVC ${}_{\text{BB}}$ (Fig. 7). These results suggest that the muscle activity of TB changes corresponding to BB, which also reinforces the idea that TB is essential to maintain the stability of the elbow joint during isometric contraction. Moreover, even though the assistance in this experiment was aimed to assist the agonist, it also showed the same inhibitory effect on the muscle activities of the antagonist.

4.1.4 Rated perceived exertion

RPE also showed a similar tendency in response to increased assistive force as that observed in %MVC ${}_{\text{BB}}$ (Fig. 8), suggesting that the relief of perceived exertion correlates with the reduction of main working muscle activity. In addition, no difference in RPE between the two arms indicated that a unilateral task performed with either the dominant or non-dominant arm perceived the same level of exertion.

4.2 A comparison of force steadiness (CV) of the dominant and non-dominant arm

It is known that muscle force steadiness could be influenced by the muscle group performing the task, the type and the intensity of the muscle contraction, and the physical activity status of the individual [24]. In the present study, CV of both dominant and non-dominant arms were greater at 40%MVF workload compared with that at 20%MVF workload. This result is in line with a previous study that indicated that CV of force increases as the level of the target force increases [25].

Bilateral difference was also found in the CV result. The non-dominant arm appears to be more capable of controlling force exertion with higher accuracy compared with the dominant arm (Fig. 9). This laterality of force control ability between arms might be due to two reasons: different intensity of the target force and a bilateral difference in motor control strategies.

In the present experiment, participants performed the isometric elbow contraction with each arm. Although the workload conditions (20% and 40%MVF) were the same for both arms, the target force intensity was different. The mean MVFs of the dominant and non-dominant arms were 161.8 $\pm$ 27.5 N and 149.1 $\pm$ 27.5 N, respectively, representing a significantly higher value for the dominant arm. Therefore, during the force matching tasks at 20%MVF and 40%MVF workloads, the dominant arm has a greater target force, which is often associated with a higher CV [25]. Consequently, the dominant arm showed relatively poor force control performance compared with the non-dominant arm.

Additionally, from the perspective of lateralization of control strategies of the two arms, Bagesteiro and Sainburg (2003) proposed a dynamic dominance model in which the dominant arm system more effectively anticipated task dynamics due to feedforward control mechanisms, whereas the non-dominant arm specialized in the control of the steady-state position due to sensory feedback mediated error correction mechanisms [13, 26, 27]. Similarly, in this experiment, from the result of assist efficiency, one can observe the specialization of the dominant arm for controlling muscle contractions in response to the external assistive force. From the CV results, one can observe the specialization of the non-dominant arm with greater force exertion accuracy at the steady state. Accordingly, control strategies in the dynamic dominance model might also be adaptable to this experiment: performances of the dominant arm might be characterized by a feedforward control strategy that enables it to take advantage of external assistive force. In contrast, performances of the non-dominant arm are probably related to impedance control that selectively modifies arm impedance independently of force [28].

4.3 Implication on rehabilitation

In the PAS-assisted rehabilitation scenario, pre-training session is required. The PAS is often directly mounted on the patients’ body, which makes it necessary for the patients to practice interacting with the dynamic operation as well as to coordinate with the external force provided by the PAS. Our findings regarding muscle activity indicated a comparatively lower ability of the non-dominant arm to utilize the external assistive force, which indicates that the non-dominant arm may require longer pre-training in order to adapt to the assistive force and make more effective use of it.

Moreover, our findings also suggested different control strategies for the two arms of healthy people when performing PAS-assisted movement. Further, a previous clinical study has demonstrated distinct control mechanisms involved in the production and coordination of forces for post-stroke [29]. These findings give insight into the control methods of rehabilitative PAS. For instance, PAS might need to control each arm in accordance with its control characteristics to achieve a better training effect, such as, feedforward control for the dominant arm and impedance/feedback control for the non-dominant arm, or distinct control mechanisms for the less-impaired limb and the impaired limb.

4.4 Limitations and future study

The present study found different specialties of the two arms in response to external assistive force. However, there are several limitations of our study, including movement form (static, isometric contraction, unilateral elbow flexion), workload intensity (20% and 40%MVF) and duration (30 seconds). If these factors are changed, physiological responses of the dominant and non-dominant arms might be different, and different specialties of the two arms might be observed. Further research is necessary to examine the specific responses of the two arms against various intensities of external assistive force during bilateral elbow flexion and dynamic movement.

5. Conclusion

Comparison of the task performance of dominant and non-dominant arms at various workloads and assistive levels showed that during unilateral elbow isometric contraction, the muscle activity of BB and TB, perceived exertion, and CV of both arms decreased with the intensity of assistive force, which indicates that both arms are capable of utilizing external assistive force. However, the two arms showed different abilities in response to external assistive force. The dominant arm utilized the assistive force more efficiently, especially at a lower assistive level, which suggests the necessity for the non-dominant arm to take more practices to accommodate to the PAS at pre-training session. On the other hand, the non-dominant arm could maintain a force with higher accuracy. The different advantages of the two arms in performance indicate distinct motor control strategies of each arm and implicate the necessity of the corresponding control method of rehabilitative PAS of each arm to achieve a better training effect.

Footnotes

Acknowledgments

This work was supported by KAKENHI JP15K 14619 and JP17H01454.

Conflict of interest

None to report.

References

Young

Ferris

. State of the art and future directions for lower limb robotic exoskeletons. IEEE Trans Neural Syst Rehabil Eng. 2017; 25(2): 171–82.

Veneman

Kruidhof

Hekman

EEG

Ekkelenkamp

Van Asseldonk

EHF

Van Der Kooij

. Design and evaluation of the LOPES exoskeleton robot for interactive gait rehabilitation. IEEE Trans Neural Syst Rehabil Eng. 2007; 15(3): 379–86.

Perry

Rosen

Burns

. Upper-limb powered exoskeleton design. In: IEEE/ASME Transactions on Mechatronics. 2007. pp. 408–17.

Stephenson

Lockwood

. Effectiveness of robotic assisted rehabilitation for mobility and functional ability in adult stroke patients: a systematic review. Vol. 15, JBI database of systematic reviews and implementation reports. 2017; 3049–91.

Nasir

Hayashi

Loh

Muraki

. The effect of assistive force on the agonist and antagonist muscles in elbow flexion. Movement, Heal Exerc. 2017; 6(2): 35–52.

Adam

De Luca

Erim

. Hand dominance and motor unit firing behavior. J Neurophysiol [Internet]. 1998; 80(3): 1373–82. Available from: http://www.physiology.org/doi/10.1152/jn.1998.80.3.1373.

Armstrong

Oldham

. A comparison of dominant and non-dominant hand strengths. J Hand Surg Am. 1999; 24B(4): 421–5.

Demura

Miyaguchi

Aoki

. The difference in output properties between dominant and nondominant limbs as measured by various muscle function tests. J Strength Cond Res. 2010; 24(10): 2816–20.

Grabowska

Gut

Binder

Forsberg

Rymarczyk

Urbanik

. Switching handedness: fMRI study of hand motor control in right-handers, left-handers and converted left-handers. Acta Neurobiol Exp (Wars). 2012; 72(4): 439–51.

10.

Hlustík

Solodkin

Gullapalli

Noll

Small

. Functional lateralization of the human premotor cortex during sequential movements. Brain Cogn. 2002; 49(1): 54–62.

11.

Derakhshan

. Laterality of the command center in relation to handedness and simple reaction time: A clinical perspective. J Neurophysiol. 2006; 96(6): 3556.

12.

Mutha

Haaland

Sainburg

. The effects of brain lateralization on motor control and adaptation. J Mot Behav. 2012; 44(6): 455–69.

13.

Bagesteiro

Sainburg

. Nondominant arm advantages in load compensation during rapid elbow joint movements. J Neurophysiol [Internet]. 2003; 90(3): 1503–13. Available from: http://jn.physiology.org/cgi/doi/10.1152/jn.00189.2003.

14.

Duff

Sainburg

. Lateralization of motor adaptation reveals independence in control of trajectory and steady-state position. Exp Brain Res. 2007; 179(4): 551–61.

15.

Zahmatkeshan

Delaviz

. Effect of isometric exercises on ability and balance of patients with multiple sclerosis. Iran Rehabil J. 2017; 15(4): 415–20.

16.

Khosrojerdi

Tajabadi

Amadani

Akrami

Tadayonfar

. The effect of isometric exercise on pain severity and muscle strength of patients with lower limb fractures: a randomized clinical trial study. Med – Surg Nurs J. 2018; 7(1): e68104.

17.

Oldfield

. The assessment and analysis of handedness: the Edinburgh inventory. Neuropsychologia. 1971; 9(1): 97–113.

18.

Borg

GAV

. Borg – MSSE 1982 – Perceived_Exertion. Med Sci Sports Exerc. 1982; 14(5): 377–81.

19.

McGuire

Green

Gabriel

. Task complexity and maximal isometric strength gains through motor learning. Physiol Rep. 2014; 2(11): e12218.

20.

Suzuki

Kaiya

Watanabe

Hutton

. Contraction-induced potentiation of human motor unit discharge and surface EMG activity. Med Sci Sports Exerc. 1988; 20(4): 391–5.

21.

Diederichsen

Nørregaard

Dyhre-Poulsen

Winther

Tufekovic

Bandholm

, et al. The effect of handedness on electromyographic activity of human shoulder muscles during movement. J Electromyogr Kinesiol. 2007; 17(4): 410–9.

22.

Petrofsky

Batt

Hye

Jones

Ushak

Tucker

, et al. Muscle use during isometric cocontraction of agonist-antagonist muscle pairs in the upper and lower body compared to abdominal crunches and a commercial multi gym exerciser. J Appl Res. 2006; 6(4): 300–28.

23.

Frey-Law

Avin

. Muscle coactivation: a generalized or localized motor control strategy? Muscle and Nerve. 2013; 48(4): 578–85.

24.

Enoka

Christou

Hunter

Kornatz

Semmler

Taylor

, et al. Mechanisms that contribute to differences in motor performance between young and old adults. J Electromyogr Kinesiol. 2003; 13(1): 1–12.

25.

Sherwood

Schmidt

. The relationship between force and force variability in minimal and near-maximal static and dynamic contractions. J Mot Behav. 1980; 12(1): 75–89.

26.

Schaefer

Haaland

Sainburg

. Hemispheric specialization and functional impact of ipsilesional deficits in movement coordination and accuracy. Neuropsychologia. 2009; 47(13): 2953–66.

27.

Wang

Sainburg

. The dominant and nondominant arms are specialized for stabilizing different features of task performance. Exp Brain Res. 2007; 178(4): 565–70.

28.

Franklin

. Impedance control balances stability with metabolically costly muscle activation. J Neurophysiol [Internet]. 2004; 92(5): 3097–105. Available from: http://jn.physiology.org/cgi/doi/10.1152/jn.00364.2004.

29.

Lodha

Coombes

Cauraugh

. Bimanual isometric force control: asymmetry and coordination evidence post stroke. Clin Neurophysiol. 2012; 123(4): 787–95.

A comparison of motor control characteristics of the dominant and non-dominant arms in response to assistive force during unilateral task

Abstract

BACKGROUND:

OBJECTIVE:

METHODS:

RESULTS:

CONCLUSIONS:

Keywords

1. Background

2. Materials and methods

2.1 Experimental participants

Table 1 Characteristics of participants ( n = 13)

2.3 Experimental conditions

2.4 Experimental protocol

2.4.1 MVC task

2.4.2 Force matching task with assistance

2.5.1 sEMG recordings

2.5.2 Tension recordings

2.5.3 Perceived exertion

2.6 Data analysis

2.6.1 Muscle activity of BB and TB (%MVC)

2.6.2 Force steadiness (CV)

2.6.3 Assist efficiency

3. Results

3.1 %MVC BB of dominant and non-dominant arms

3.2 The assist efficiency of dominant and non-dominant arms during the steady period

3.3 %MVC TB of dominant and non-dominant arms

3.4 Perceived exertion

3.5 Force steadiness of dominant and non-dominant arms

4. Discussion

4.1 A comparison of muscle activity of dominant and non-dominant arms

4.1.1 Muscle activity of biceps brachii

4.1.2 The assist efficiency

(1) Why was the laterality of assist efficiency only found at a low assistive level?

(2) Why was assist efficiency of the non-dominant arm lower than that of the dominant arm?

4.1.3 Muscle activity of triceps brachii

4.1.4 Rated perceived exertion

4.2 A comparison of force steadiness (CV) of the dominant and non-dominant arm

4.3 Implication on rehabilitation

4.4 Limitations and future study

5. Conclusion

Footnotes

Acknowledgments

Conflict of interest

References

Table 1
Characteristics of participants ( $n=$ 13)

3.1 %MVC ${}_{\text{BB}}$ of dominant and non-dominant arms

3.3 %MVC ${}_{\text{TB}}$ of dominant and non-dominant arms