Reliability and efficacy of the long-latency stretch reflex in the human thumb.

1. The amount of positional compensation afforded by the long-latency reflex in the flexor pollicis longus has been investigated in ten normal human subjects. 2. The interphalangeal joint of the thumb was extended by between 2 and 40 degrees at up to 900 deg/s by suddenly increasing the standing force applied to the lever against which the subject was pressing with the pad of the thumb. 3. Electromyographic (e.m.g.) responses at spinal-latency were very small or absent for stretches of this magnitude. The long-latency stretch reflex produced an average positional correction of about 50% for disturbances in the range of 5-25 degrees. The response began to saturate for disturbances of greater than 25 degrees. 4. The e.m.g. response was pulsatile, lasting only some 50 ms, even during continuously increasing disturbances; frequently it terminated despite a remaining positional error. 5. There was a large variation from subject to subject in the average amount of positional correction provided by the stretch reflex. Examination of single responses to the same stretch in individual subjects showed an even greater variation from trial to trial. 6. Variation in the compensation produced by the long-latency stretch reflex from trial to trial could not be explained by the slight variation in size or maximum velocity of the individual stretches.