STRETCH REFLEX RESPONSES IN THE HUMAN ELBOW JOINT DURING A VOLUNTARY MOVEMENT

1. The responsiveness of the stretch reflex is modulated during human voluntary limb movements. The influence of this modulation on the limb mechanical properties (stiffness) was investigated. 2. Subjects were taught to replicate accurately a rapid (4.0 rad s(-1)) targeted elbow flexion movement of 1. rad. From the onset of 12 % of the trials a sinusoidal position disturbance (0.05 rad) was superimposed on the normal (trained) movement trajectory. The net joint torque (muscle torque) resisting these stretches was computed from measurements of applied torque, acceleration and limb inertia. Electromyographic (EMG) responses in the triceps brachii (TB), brachialis (Br) and biceps brachii (BB) were monitored. 3. The EMG responses to sinusoidal stretches applied early in the movement were less than those responses to perturbations applied when the arm neared the target (especially in the antagonist muscle TB). These EMG responses caused fluctuations in the resistance to the perturbation (stiffness), as described below. 4. When the perturbation frequency was low (< 4 Hz) the resistance of the elbow muscles to the stretch increased as the arm approached the target (48% increase). In contrast, when the stretch frequency was 7 Hz the resistance decreased by 63%. This decrease can be explained by the increased reflex response, since at 7 Hz the reflex response is probably timed so that it assists, rather than resists, the stretching as a result of loop delays. This reflex timing was confirmed by observing that, after abruptly stopping the sinusoidal stretch, the reflex response persisted for 100 ms and was indeed in a direction that would have reduced the resistance, had the perturbation continued. 5. The time course of the net muscle stiffness was estimated for frequencies ranging from 4 to 8 Hz and for each 40 ms interval a Nyquist plot was constructed, forming a C-shaped curve as frequency was varied. The size of this curve gave a measure of the stiffness resulting from reflex activity. When the arm neared the target this reflexive stiffness reached a maximum, and was probably comparable in size to the intrinsic (non-reflexive) muscle stiffness. Also, in four of the five subjects the viscous component of stiffness at 7 Hz dropped significantly below zero when the arm neared the target, again indicating that at this frequency the reflex was large and acted inappropriately. 6. In summary, the contribution of the stretch reflex to the mechanical response of the arm is relatively small during a movement but increases as the target is approached, to a maximum level where it contributes substantially to the total resistance to low frequency perturbations. This increase is, however, insufficient to cause limb instability resulting from reflex delays since: (i) the reflexive component of stiffness is only transiently high and (ii) at the frequencies where the delayed reflex responses act inappropriately (6-7 Hz) the limb inertia dominates the response to disturbances.