Hysteresis in corticospinal excitability during gradual muscle contraction and relaxation in humans

Many studies have demonstrated that the firing behavior of single motor units varies in a nonlinear manner to the exerted torque during gradual muscle contraction and relaxation. However, it is unclear whether corticospinal excitability has such a hysteresis-like feature. In this study, we examined corticospinal excitability using transcranial magnetic stimulation (TMS) during gradual muscle contraction and relaxation for torque regulation in elbow flexor muscles. Eight healthy male subjects performed two different isometric elbow flexion tasks, namely, sinusoidal and tonic torque exertion tasks. In the sinusoidal task, the subjects sinusoidally increased and decreased the isometric elbow flexion torque (range of 0-15% of maximum voluntary contraction) at three different frequencies (0.33, 0.17, and 0.08 Hz). For each ascending (contraction: CON) and descending (relaxation: REL) period of the exerted torque, a single TMS was applied at 5 phases. In the tonic task, the elbow flexion torque was tonically exerted at 7 levels in a similar range as that in the sinusoidal task. EMG activities were recorded from the agonists, the biceps brachii (BB) and brachioradialis (BRD) muscles, and an antagonist, the triceps brachii (TB) muscle. The results demonstrated that the EMG activities of both the agonists and antagonist were larger in the CON period than the REL period, even when the exerted torque was the same. However, there were no significant differences in EMG activation profiles among the different frequencies of contraction. In BB and BRD, the motor-evoked potential (MEP) elicited by the TMS was also greater in the CON period than in the REL period. This CON-REL difference of MEP amplitudes was still observed when corrections were made for the increased EMG activities; that is, the MEP amplitudes to the identical EMG activities were greater in the CON period than in the REL period, and this phenomenon was more pronounced at higher frequencies. In addition, the degree to which sinusoidal MEPs exceeded tonic MEPs in the CON period and were smaller than tonic MEPs in the REL period became more pronounced at higher frequencies. On the other hand, there were significant correlations between the BB and BRD MEP amplitudes and the rate of change of elbow flexion/extension torque. These results indicate that corticospinal excitability during muscle contraction and relaxation has a neural hysteresis to the muscle activity, i.e., spinal motoneuronal activity, according to the rate of change of the exerted torque, i.e., muscle tension. This suggests that corticospinal excitability modulation depends not only on concurrent spinal motoneuronal activity and muscle tension but also on the time-series pattern of their changes during muscle contraction and relaxation.