Responses to load disturbances in human shoulder muscles: the hypothesis that one component is a pulse test information signal.

Human motor control has been investigated by applying displacements acting to rotate the shoulder while the subject was endeavouring to maintain a constant position against a pre-existing force delivered by a system of finite stiffness. Four separate stages of the force response were distinguished. First, for the initial 100 msec, an increase in force which was attributed to the viscoelastic properties of activated muscle. Second, after approximately 100 msec, a "medium latency" increase in force accompanied by an increase in EMG activity. Third, a "long latency" increase in force consistent with voluntary action restored the arm to its original position. Fourth, some 500 msec thereafter and dependent upon the final steady force level, a tremor might develop. No changes of force were seen that were of sufficiently short latency to be attributed either to Ia monosynaptic action or immediately following (within 15 msec) polysynaptic action of muscle mechano-receptors. Even the "medium latency" response was too weak to make an appreciable contribution to restoring the arm to its original position; at the best it provided only 15% of the force required. This response is often considered as a "stretch reflex" responsible for maintaining posture in its own right; for example, by means of a servo-assisted transcortical loop (Marsden et al., 1972). Instead, it is now suggested that it might be a test signal designed to inform the central nervous system of the current loading on the muscle and thus permit the CNS to select an appropriate pre-programmed response from its repertoire of motor actions.