The combined effect of muscle contraction history and motor commands on human position sense.

Along with afferent information, centrally generated motor command signals may play a role in joint position sense. Isometric muscle contractions can produce a perception of joint displacement in the same direction as the joint would move if unrestrained. Contradictory findings of perceived joint displacement in the opposite direction have been reported. As this only occurs if muscle spindle discharge in the contracting muscle is initially low, it may reflect increased muscle spindle firing from fusimotor activation, rather than central motor command signals. Methodological differences including the muscle contraction task and use of muscle conditioning could underlie the opposing findings. Hence, we tested perceived joint position during two contraction tasks ('hold force' and 'hold position') at the same joint (wrist) and controlled muscle spindle discharge with thixotropic muscle conditioning. We expected that prior conditioning of the contracting muscle would eliminate any effect of increased fusimotor activation, but not of central motor commands. Muscle conditioning altered perceived wrist position as expected. Further, during muscle contractions, subjects reported wrist positions displaced ~12 degrees in the direction of contraction, despite no change in wrist position. This was similar for 'hold force' and 'hold position' tasks and occurred despite prior conditioning of the agonist muscle. However, conditioning of the antagonist muscle did reduce the effect of voluntary contraction on position sense. The errors in position sense cannot be explained by fusimotor activation. We propose that central signals combine with afferent signals to determine limb position and that multiple sources of information are weighted according to their reliability.