Specific facilitation of the agonist motoneuron pool (pretibial muscle; TA) was investigated under the two different reaction time conditions, i, e, simple (S) and choice (C) reaction time tasks. The motor task was bilateral ankle dorsiflexion, and two different strengths of voluntary ankle dorsiflexion (25%, (S) and 50% (L) of maximum contraction force) were controlled by visually guided tracking. The excitability of the agonist motoneuron pool was investigated by the H-reflex method.
As the differences between the EMG reaction times of TA on both sides were extremely small under both the S and C tasks, it might be reasonable to use the non-stimulated side as the onset of voluntary movement. Not only the onset of H-reflex facilitation with respect to EMG onset, but also the reaction times were significantly prolonged under the conditions of task C compared with those of task S. In addition, movement times were also prolonged and peak dF/dt values were significantly decreased under the conditions of task C. These results allow us to conclude that the modulation of pre-motor facilitation of the agonist H-reflex is controlled by descending command from the brain, and would be caused by slower EMG reaction times. In addition, we suggest that the prolonged motor times and decreased dF/dt values under the conditions of task C are attributable to slower recruitment and a lower firing rate of motor unit.

The present study was undertaken to examine the effects of acoustic stimulation on human spinal motoneuron excitability. For this purpose, we used the soleus (Sol) H-reflex as a test reflex, and three different types of acoustic stimuli as conditioning stimuli. The features of the acoustic stimuli were as follows, 1) click sound (CS), 2) tone burst composed of 11 click sounds (TBL, interstimulus interval 10 ms), 3) tone burst composed of 21 click sounds (TBH, interstimulus interval 5 ms) . The intensity and frequency of each sound was 110 dB and 0.5 kHz, respectively.
Significant facilitation of the Sol H-reflex occurred at conditioning-testing (C-T) intervals of 50 ms in all subjects when the TBL stimulus was used (mean±S.D.; 50.3±8.2 ms) . This facilitatory effect appeared early and later when TBH (43.7±3.7 ms) and CS (59.2±4.5 ms) stimuli were used, respectively. The maximum facilitatory effect appeared at a C-T interval of 100 ms (mean and S. E.; 98.0±0.8 ms) and the amount of peak facilitation at that time was 156.1±1.4% (relative to the control value) . Thereafter, the amount of facilitation decreased sharply up to a C-T interval of 200 ms. However, slight but significant facilitation was observed continuously up to a C-T interval of 500 ms. No significant inhibition of the Sol H-reflex was observed between a C-T interval of 0 and 500 ms in all subjects.
Irrespective of whether the magnetic cortical stimulation was given before or after the Sol H-reflex at a short interval (-2 to 2 ms), the acoustic facilitation of the Sol H-reflex was not changed with a C-T interval of 50 ms. Additional facilitation due to magnetic cortical stimulation was, however, obtained in one of the four subjetcts when a C-T interval of 100 ms was used.
These results suggest that the acoustic facilitation of the Sol H-reflex is composed of three different facilitatory mechanisms; 1) a pathway with a high threshold and shorter latency (about 40 ms), 2) a pathway with a lower threshold and medium latency (about 60 ms), and 3) a pathway with a relativelyhigh threshold and longer latency (longer than 200 ms) . In addition, we discussed the mechanisms that underlie the facilitatory effects of the Sol H-reflex.

Task-dependent changes in the cutaneous reflex in the upper and lower leg muscles were examined in normal human subjects (n=11) . After instruction, the subjects were asked to selectively contract agonist muscles (SC task) and to co-contract antagonistic muscles (CC task) for the ankle or knee joint while standing. The cutaneous reflex was elicited by applying non-noxious electrical stimulation to the superficial peroneal nerve at the ankle joint (200 Hz, 5 pulses) . The EMG signal was rectified, averaged (n=10), cumulatively summated up to 150ms after the end of the stimulation artifact, and then divided by the time interval for the summation (ACRE₁₅₀) . A strong inhibitory effect was determined at a latency of 50 ms and was followed by a facilitatory effect after the electrical stimulation during the SC task in all muscles. In contrast, it was found that the early inhibition and the later facilitation tended to be decreased and increased during the cc task, respectively. A linear regression analysis between the ACRE₁₅₀ and the background EMG revealed that the regression slopes were significantly decreased during CC task except for the tibialis anterior (TA) and biceps femoris. The reflex ratio (ACRE₁₅₀/background EMG) was also negative for the SC task in all muscles tested, but was significantly reduced or showed a positive value for the CC task. These results suggest that the brain may control the cutaneous reflex pathways to enhance the facilitatory effects of the thigh and ankle extensor muscles during the CC task. This reflex action during the CC task may serve to prevent an undesirable fall in the center of gravity in response to a sudden tactile sensation to the dorsal surface of the foot.

A study was conducted to investigate changes in the excitability of the ankle extensor and flexor motoneurons during voluntary contraction of the quadriceps femoris muscle (Quad) . For this purpose, we used a reaction time task and the H-reflex technique. Subjects lay in the prone position on a bed and performed isometric contraction of the Quad. The strength of the contraction force was about 30% of maximum.
In all subjects we observed strong facilitation of the soleus (Sol) H-reflex, which occurred from the EMG onset of vastus lateralis muscle (VL) . The pretibial muscle (TA) H-reflex was also facilitated at almost the same time as that seen in the Sol H-reflex in nine out of ten subjects. The peak of Sol and TA H-reflex facilitation appeared between 50 and 100 ms after the EMG onset of the VL, and then these facilitations gradually decreased. Weak but constant activities of the Sol and medial gastrocnemius muscle (MG) were observed on the full-wave rectified and averaged EMG record after 50 to 80 ms from the EMG onset of VL. No such EMG activity was observed in TA.
These results suggest that excitatory inputs including those of both descending and peripheral origin induced by voluntary contraction of Quad are responsible for facilitation of the ankle extensor and flexor motoneurons. In addition, removal of presynaptic inhibition of the Ia terminal of the motoneurons by descending motor command might explain the present results.

Premovement facilitation of spinal monosynaptic reflex was investigated under different movement modalities using both visually guided tracking movement and the H-reflex technique. Subjects performed fast step and slow ramp movements under reaction time (RT) conditions and self time determining (Self) conditions. For each of the conditions, the motor task was bilateral simultaneous dorsiflexion.
Under both RT and Self conditions, the onset of premovement H-reflex facilitation (OPHF) was significantly prolonged in the slow ramp movement (RT, 37, 3±11.8 ms ; Self, 79.4±21.8 ms) compared with the fast step movement (RT, 24.4+6.2 ms ; Self, 43.4±14.2 ms) . In addition, OPHF was significantly prolonged under Self conditions in both the step and ramp movements. Movement time (MT) did not differ significantly under RT and Self conditions except in three subjects. The peak value of torque change (PV) was larger under Self conditions than under RT conditions in four out of ten subjects.
From these results, we conclude that OPHF is modulated not only by changes in movement speed but also by changes in movement modality. It is suggested that this contextual dependency of OPHF might be controlled by the supraspinal motor center.

The purpose of the present study was to determine electrophysiological differences in muscular fatigue between the soleus (Sol) and tibialis anterior (TA) muscles in normal human subjects (n=5) . The subjects were asked to make four 20-sec maximum voluntary contractions (MVCs), each separated by 3-min intervals of rest. A 3-sec MVC (V_task) or a 3-sec supramaximum electrical stimulation (E_task) was imposed at 1-min intervals during the resting period. The plantar flexion and dorsiflexion forces were significantly decreased throughout both the V_ and E_tasks. In particular, the decrease in the dorsiflexion force during the V_task was found to be very steep. However, wide inter-subject variations were found in the time course of the decrease in the MVC force in all tasks. Changes in mean power frequency (MPF) of the electromyographic (EMG) recordings in the Sol were found to be small in both tasks. In contrast, the MPF was significantly decreased during the 20-sec MVC in the TA. The root mean square (RMS) of the EMG gradually declined during both the V_ and E_tasks. The ratio of the root mean square of EMG (RMS) and the exerted force (RMS/F) was also determined during both tasks and for both muscles. The RMS/F was markedly increased during the V_task in the TA. An increase was also found in the Sol, but the magnitude of the increase was small. A small but consistent decrease in the M-wave was found in the V_task. The Sol H-reflex was decreased until the second 20 s MVC, then reached a plateau, and further decreased at the end of the fourth 20s MVC. It was suggested that the electrophysiological differences in the Sol and TA during muscular fatigue induced by the repetitive 20-sec MVC reflected differences in the physiological properties of these muscles. The RMS/F was suggested to be a useful parameter for determining the local muscular fatigue in intact human lower leg muscles.

Eleven healthy subjects repetitively performed maximal cycling movement for 10 s with 20 s rest intervals. The load of the cycling was respectively set to 30% (high frequency task, lIF' task) and 80% (high power task, TIP task) of the optimal load for exerting maximum anaerobic power. Each task was finished when the exerted maximal power was decreased to 80% of the initial value. While performing each task, transcranial magnetic stimulation (TMS) was delivered to the motor cortex which was effectively able to evoke motor evoked potential (MEP) from the thigh muscles. Elec-tromyographic (EMG) activity of the left rectos femoris (RF), vastus lateralis (VL) and the MEP was analyzed.
The maximal power exerted was decreased to 80.6±1.58 % in the HF task, and 77.3±0.77 % in the HP task. The number of repeated sets in each task was 10.1 ± 1.45 (HF task) and 4.1±0.25 sets (HP task) . The MEP area of the RF and VL was not changed significantly in the HF task, though it was significantly increased in the latter half of the HP task. A two-way ANOVA showed that the time course of the changes in the MEP area was significant in the VL (p<0.01), but not in the RF. In both tasks, the duration of the MEP was progressively prolonged in each 10 sec pedaling, and the prolongation was evident in the latter half of the tasks. However, the magnitude of the prolongation was significantly larger during the HP task. The ratio of the integrated amplitude of the EMG and the exerted power at the initial 5 bouts of cycling (EMG/Power ratio) was significantly increased in both the RF and VL, suggesting that peripheral muscular fatigue was induced during at the latter half of each task. Furthermore, the EMG/Power ratio in the VL was significantly higher during the HP task than the HF task.
These results suggest that central fatigue plays a significant role in decreasing the maximum power output, and that it takes place in a muscle-dependent fashion. It was also suggested that during low load, but relatively higher cadence frequency, central fatigue other than that involving the motor cortex accounts for the decreased power output.

Changes in the motor evoked potential (MEP) evoked by transcranial magnetic motor cortex stimulation (TMS) of rectos femoris (RF) and vastus lateralis (VL) was examined during constant cadence cycling tasks for 60 sec. Subjects were 11 normal male volunteers aged between 19 and 25 years. Pedaling load was set at 100% and 80% of the estimated optimal value for maximum anaerobic power output. For the low load task (LL task), the pedaling rate was set at half the value of the maximum pedaling rate with the load set at 80% of the optimal for maximum anaerobic power output. For the high load task (HL task), the pedaling rate was set such that the power was equivalent to the LL task.
The route mean square of the electromyographic (EMG) activity amplitude tended to steeply increase during the latter half of the task. The magnitude of the increase in the RMS was significantly larger in the HL task than the LL task. The area of the MEP also tended to increase in both tasks, though the degree of the increase was significantly larger in the LL task than the HL task. The EMG silent period (SP) after the MEP tended to steeply increase just after the task initiation and to decrease in the latter half of the task in the HL task. However, in the LL task the facilitation of MEP was not found, but it showed a gradual decrease while performing the task. The duration of the MEP tended to increase in both tasks, though the degree of the increase in the VL was significantly larger in the LL task than the HL task. The linear regression analysis between the size of the MEP and the background EMG shows a significant positive correlation coefficient during isometric contraction, but not during the two types of cycling tasks.
These results suggest that the neural circuit responsible for the MEP was controlled differentially during isometric contraction and constant cadence pedaling. Also it is likely that the mechanism of central fatigue differed depending on the cadence and or load in a task-dependent fashion irrespective of the same power output.

The present study investigated how resistance training affects behaviors related to central and peripheral fatigue during a sustained maximal voluntary contraction (MVC) . The subjects were well-trained (TR, n=8) and sedentary untrained (UT, n=6) males. The subjects were asked to repetitively perform 3 sets of MVC (elbow flexion) for 1 min with a rest interval of 1 min. Transcranial magnetic stimulation (TMS) was delivered to the contralateral motor cortex to evoke the motor evoked potential (MEP) and electromyographic (EMG) silent period (SP) after the MEP. Ratio of root mean square (RMS) of the EMG and elbow flexion force (RMS/F) was also calculated.
The time course of the decrease in elbow flexion force that was standardized with respect to the maximal value obtained at the beginning of the first MVC was almost identical in both TR and UT. At the end of the task, the elbow flexion force decreased to around 30 % of the initial value in both groups. Decrease in voluntary activation (VA) estimated by the increment of the force after TMS was significantly larger in UT (77.3%) than in TR (88.2%) at the end of the task. Although the increase in MEP during the first set was significantly greater in UT than in TR, elongation of SP was significantly larger in UT than in TR. Increase in RMS/F, which is a manifestation of peripheral fatigue, was significantly larger in TR than in UT.
These results suggest that decrease in MVC in UT and in TR is respectively more attributable to central and peripheral fatigue, and that inhibitory inputs to motor cortex were larger in UT than in TR. It is concluded that expression of central and peripheral fatigue is affected by resistance training.