Control of stiffness by the medium latency electromyographic response to limb perturbation.

It is hypothesized that the medium latency electromyographic (EMG) response (ML) to limb perturbation functions to preset limb stiffness to a constant initial level. Three predictions are derived from this hypothesis: Firstly, in the presence of an instruction calling for opposition to limb perturbation, a control signal, ML, will be observed and will lead to the establishment of a constant level of limb stiffness. Secondly, in the absence of an instruction to oppose, no control signal will be observed and correspondingly a constant stiffness will not be generated. Thirdly, the latency of onset of stiffness control will covary with the latency of onset of ML. These predictions were tested in experiments involving perturbation of the human forearm about the elbow joint, with surface EMG measurements and computation of the limb stiffness function. The results are in accord with these predictions, and support the hypothesis that ML functions in the feedforward control of limb stiffness.