Dynamic response of human muscle spindle afferents to stretch.

1. One hundred and twenty-four muscle afferents from the finger extensor muscles were recorded from the radial nerve in human subjects. 2. The afferents were provisionally classified as muscle spindle primary (78/124) and secondary afferents (25/124), and Golgi tendon organ afferents (21/124), on the basis of their response to 1) maximal twitch contractions, 2) 20- and 50-Hz sinusoids superimposed on ramp-and-hold stretches, 3) stretch sensitization, and 4) isometric contractions and sudden relaxations. 3. Ramp-and-hold stretches at two velocities, 10 and 50 degrees/s, were applied to the appropriate metacarpophalangeal (MCP) joint while the parent muscle remained relaxed. For each unit three discrete parameters were assessed: the presence or absence of 1) an initial burst at the commencement of the ramp stretch, 2) a deceleration response at the beginning of the hold phase, and 3) a prompt silencing at muscle shortening. In addition, two kinds of dynamic indexes were calculated for 79 of the muscle spindle afferents. 4. Most spindle afferents responded readily to stretch, whereas the Golgi tendon organ afferents produced very poor stretch responses. All of them lacked a static response, whereas the dynamic response, when present at all, consisted of only a few impulses. 5. The dynamic index was higher for spindle primaries than for secondaries, and this difference was statistically significant although the distribution was unimodal for spindle afferents as a group. Hence, this parameter was a poor discriminator. 6. Initial bursts, deceleration responses, and silences during imposed shortening were more common in spindle primaries than in secondaries. The differences were significant in all these respects. 7. The three discrete parameters were statistically pairwise independent for the spindle afferents, justifying the combination of the three into a useful battery for discrimination between primary and secondary spindle afferents and the use of this battery as a partial data base for a probability approach towards a solid classification of human muscle afferents.