Voluntary suppression of neck reflexes during passive head-on-trunk rotations: reflex gain control versus proprioceptive feedback.

Normal subjects can completely eliminate resistance upon imposed head-on-trunk rotations when they are asked to relax. It is not, however, clear how neck reflexes to stretch can be voluntarily suppressed. Reflexive responses might be modified by adjusting the gain of the reflex loop through descending control. Theoretically, necessary corrections upon interfering disturbances during coordinated motor performance requiring the interplay of relaxation/activation may be missing if muscle relaxation is taking place exclusively by this mechanism. It has been alternatively proposed that sensory information from the periphery may be allowed to "neutralize" neck reflexes if it is fed back with opposite sign to the structures driving the reflexes. Six healthy subjects were asked to relax while subjected to head-on-trunk rotations generated by a head motor. After any initial resistance had completely subsided, the head was unexpectedly exposed to "ramp-and-hold" perturbations of up to 2° amplitude and 0.7 s duration. Resistance to stretch consistently reappeared thereupon, suggesting that stretch reflex gain had not been set to zero during the previously achieved complete relaxation. Resistance to perturbations under these circumstances was compared with the forces generated when the same ramp-and-hold displacements were delivered unpredictably to the head held stationary. A quantitative model of neck proprioceptive reflexes suppression has been thus constructed. Gain scheduling or "motor set" cannot sufficiently account for the voluntary reflex suppression during slow passive head rotations. Instead, we propose as underlying mechanism, the "neutralization" of the controlling servo by means of continuous feedback tracking displacement and force signals from the periphery.