Evidence that the long-latency stretch responses of the human wrist extensor muscle involve a transcerebral pathway.

Ballistic wrist extensor movements were interrupted by a sudden perturbing force in two conditions, and subjects were asked either to compensate for ('oppose') the perturbation or to cease the extensor movement ('let-go') when the interruption occurred. In the first condition, the interruption occurred unexpectedly in some of the trials, whereas in the second it occurred on every trial. Responses were recorded from the flexor and extensor carpi radialis muscles, and from the scalp and cervical spine. Unexpected perturbing forces were associated with an early extensor EMG response (M1) and longer latency EMG responses in both flexor and extensor muscles. When the interruptions occurred in every trial, both the late EMG activity and the cerebral responses were markedly attenuated. The relation between expectation of perturbation and the occurrence or size of the late EMG activity indicates that this response is under supraspinal control. The potentials recorded over the cervical spine and scalp were, like the late EMG activity, influenced by the expectancy of interruption. These potentials persisted even when the potentials related to the mechanically elicited SEP were removed by subtraction of responses to anticipated interruptions from those to unanticipated interruptions. The earliest of these potentials preceded the late EMG activity by 30-50 ms. They were present when the late responses were present, absent when the late responses were absent, and became progressively smaller when the late responses became smaller, suggesting a tight coupling between these events and the late EMG events. These results indicate that cerebral activity, unrelated merely to mechanical stimulation, consistently precedes the late EMG activity, and thereby directly suggests that the long-latency response has a cerebral (possible cortical) pathway.