Aftereffects in the responses of cat muscle spindles and errors of limb position sense in man

This is a report of two series of experiments, one, observations on muscle spindles in the anesthetized cat, the other, based on the animal observations, concerned with limb-position sense in human subjects. When the cat soleus muscle is contracted at a short length and then after a brief period is brought to an intermediate length, the integrated afferent activity recorded in dorsal roots is maintained at a high level. If the contraction is applied while the muscle is held at a long length, the subsequent level of afferent activity back at the intermediate length is much lower. A similar result is obtained if large amplitude vibration of the muscle is used instead of contraction. Recordings from single afferents confirm that these alterations in discharge could be attributed to the primary and secondary endings of muscle spindles. Following the proposal that the resting discharge of muscle spindles contributes to the sense of limb position in man, we have used a sequence of contractions or periods of vibration of biceps and triceps brachii followed by an arm-matching task to look for errors in limb position consistent with the patterns of muscle afferent discharge observed in the animal experiments. Blindfolded subjects were told to contract biceps of one, the test arm, with the arm flexed. After the arm had relaxed, it was placed by the experimenter at an intermediate angle that the subject was asked to match with the other, indicator arm. Then the subject was asked to contract triceps with the arm extended, and the matching task was repeated. The hypothesis was that contracting a muscle while it was short raised the level of afferent discharge coming from it when the arm was moved to an intermediate position so that the muscle would be perceived as more stretched and the arm felt to be displaced further than it really was. Thus, when biceps was contracted with the arm flexed, this would lead the arm in the intermediate position to be perceived as more extended than it really was. When triceps was contracted with the arm extended, this would lead the arm in the intermediate position to be perceived as more flexed than it really was. For a total of 106 pairs of matches carried out by 18 subjects the error in the position adopted by the indicator arm was often not in the expected direction, and it became apparent that subjects were not able to match accurately a passively positioned arm. However, when the difference between errors after a conditioning contraction of the extended arm and then of the flexed arm was calculated, it was highly significant. Nearly all subjects perceived the arm's position after a contraction in the flexed position as more extended than in a temporally adjacent match where the arm was contracted while held extended. These matching errors are consistent with the changes in resting discharge of muscle spindles seen in the animal experiements and, therefore, provide further support for the proposal that spindles are concerned with signaled limb position.