These results showed that VR wrist exercise following tDCS had greater immediate and sustained post-exercise corticospinal facilitation effects than exercise without tDCS or than tDCS without exercise. Furthermore, post-exercise corticospinal facilitation was sustained for 20 min after exercise in the VR and tDCS conditions compared to the other conditions.
VR training-induced increase in cortical excitability
In the present study, VR motor training facilitated post-exercise corticospinal excitability more than simple active exercise in healthy volunteers. Although not consistent, most previous studies revealed that repeated motor training increased motor cortex excitability immediately and reduced intra-cortical inhibition [37–40]. However, the type of motor training is crucial for the post-exercise facilitatory effects on the motor cortex [41, 42]. In this study, we compared the post-exercise facilitatory effects between VR wrist exercise paradigm and active exercise. Short-term changes in corticospinal excitability after visuomotor adaptation using a VR program were assessed as a marker for learning-related processes; these facilitatory effects might accelerate motor recovery in stroke patients . These changes in M1 excitability may lead to sustained, cumulative changes, and are associated with motor learning and better post-stroke clinical outcomes [40, 43–45].
The detection rate of wrist movements was about 30 Hz in our VR system, a rate that was sufficient for examining wrist movements that were performed one cycle about every 6 s. This was much slower than 30 Hz, so the wrist movements were fully detectable, although conventional movement measuring systems have much higher recording rates [46, 47].
To the best of our knowledge, this is the first report to show that VR wrist exercise facilitates post-exercise corticospinal excitability more than a paced- active wrist exercise (Figure 3; A, B). Comparing VR to the paced- active wrist exercise, there was no significant difference in duration of exercise (15 min), performance metrics (movement speed, total distance), or MEP facilitation during exercise. This suggests that the superior facilitation by VR exercise resulted from factors other than differences in muscle activation during exercise. One possible reason for the enhanced post-exercise facilitation of corticospinal activity by the VR exercise could be found in the characteristics of VR exercise, which is task-oriented (catching coins and successful jumping), more interactive, and more interesting. Thus, it draws subject attention that might activate the ipsilesional extended motor network, including a putative mirror neuron system .
The present study and many previous findings support our interpretation. We found that VR wrist exercise produced a higher level of attention and a lower fatigue scale score than active wrist exercise (Figure 4). It is known that the type of motor activation is important for the occurrence of the post-training facilitatory effects of MEP [41, 42]. Perez et al. showed that repetitive motor skill exercises but not non-skill motor passive exercises, increased post-exercise corticospinal excitability evoked by TMS . They also observed a decrease in intra-cortical inhibition after motor skill exercises and explained the increase in MEPs as due to possible modulation by local intra-cortical circuits . These findings indicate that observed changes in activities at the cortical level may be related to the type of motor activity, degree of attention and fatigue, and goal-directed behavior in the motor tasks [37, 41, 49, 50].
Another possible explanation also supports the idea that coordination between visual input and motor performance is the decisive factor for the VR exercise-induced changes in cortical excitability. Visuo-motor training similar to that used in the present study increased activity in cortical neurons in monkeys and can improve motor performance in humans [27, 49, 51].
Combined tDCS and VR wrist exercise-induced increases in cortical excitability
Nitsche and Paulus noted that up to 40% of MEP excitability changes appeared and lasted for several minutes after the end of anodal tDCS in healthy volunteers . Additionally, anodal tDCS increased the MEPs of affected muscle in patients with stroke and healthy subjects . These facilitatory effects are believed to accelerate motor recovery in stroke patients . The strength and duration of these after-effects could be controlled by varying the intensity and duration of anodal tDCS . In this study, after 20 min of tDCS, the MEP returned to baseline after 20 min of no exercise. However, if VR exercise was performed immediately after tDCS, the corticospinal excitability was sustained for another 20 min.
A possible mechanism for the increased duration is the cortical excitability effect of anodal tDCS, VR exercise induced corticospinal facilitation, and the VR exercise-induced decrease in cortico-cortical inhibition may act synergistically. Although we did not assess intracortical inhibition, many previous studies have demonstrated a decrease in intracortical inhibition during and after skill-acquisitive voluntary motor training or tDCS [40, 54–56]. Reduced intracortical inhibition is important for inducing neural plasticity after injury [40, 55, 57]. However, in present study, VR exercise combined with tDCS did not increase or decrease the subjects’ rated attention or fatigue scores versus those in VR exercise alone.
Few studies have assessed the synergistic effect of tDCS and motor skill training. Two studies reported a beneficial effect of combined peripheral nerve stimulation and tDCS on motor sequence performance in chronic stroke patients , and increased corticomotor excitability of the motor cortex that persisted after anodal tDCS during robotic wrist training . One recent randomized multicenter trial noted no significant functional improvement in robot arm training during tDCS in 96 stroke patients . Considering that this study enrolled mostly patients whose upper extremities were severely impaired (FMS < 10) and had cortical lesions, unlike our study, further studies are needed to address its effects.
In this study, a single session of VR exercise following tDCS did not produce a higher score for the rate of coin acquisitions than VR exercise alone. One reason might be that tDCS was performed before the task in the present study. While our study did not show that a single session of combined therapy (tDCS and VR) would improve motor performance more than single therapy (VR alone), the rationale for this combined therapy is that cortical facilitation across multiple practice sessions may translate into enhanced and sustained neuroplastic changes in the affected hemisphere . Several studies have demonstrated long-lasting learning effects or functional improvement of skilled motor training following multiple sessions of tDCS in healthy volunteers and stroke patients [21, 33, 34, 59].
There were several limitations to our study. First, this study was conducted using a small sample of mildly impaired stroke patients. All subacute stroke patients were in a period of spontaneous recovery. We believe the impact of this factor was minimized because we tested three tasks in randomized order and the test was completed over 3 or 4 days. Second, heterogeneity in type of lesion (cortical or subcortical) and side of stimulation could be other limitations. Subcortical stroke patients with intact cortical connectivity may profit more from tDCS than patients with disrupted neural pathways . Furthermore, use-dependent cortical plasticity may differ according to the stimulated hemisphere, according to previous studies using TMS . Thus, we analyzed hemisphere-specific facilitation in stroke patients, which showed comparable results. Third, we did not compare the full performance or behavioral measurement of the upper extremity (except rate of coin acquisition) among VR conditions with or without tDCS. We focused on the change in post-exercise corticospinal facilitation according to the various exercise conditions, especially the synergistic effect of tDCS and VR conditions. Fourth, a lack of sham stimulation or multiple mode simulation of tDCS (e.g., dual hemisphere stimulation ) was another limitation. The mode, repetition, and duration of stimulation of tDCS, anatomical location of lesions, grade of severity of impairment, and type of training could affect the cortical facilitation of tDCS after stroke. Future studies should investigate whether the simultaneous application of tDCS and exercise may induce greater behavioral changes, and these results should be compared to those from the present study.