Stimulus control of EMG burst duration in human locomotion

To better understand constraints on EMG activity, burst duration was restricted for the rectus femoris (RF) muscle as subjects walked on a motor-driven treadmill. Once in each step cycle (touchdown to touchdown of the left foot) a computer-activated light discriminated a short burst (if green) and silence (if red). All four subjects rapidly met the criterion of 90% success in 50 cycles for a burst 100–300 msec long within a performance duration, PDUR, of 700 msec. In contrast, two of the four failed to reduce the burst to conform to a 100–200 msec requirement. Several variables were ruled out as bearing on these between-subject performance differences: (1) complete failure to produce a short burst within the PDUR, since many were successful; (2) a differential effect upon subjects of two “reinforcer” types (a high tone after success and a low tone after failure, or else audible EMG biofeedback before the high or low tone); (3) appreciable differences in cycle duration or swing/stance ratios; (4) the presence or absence of concomitant activity in other leg muscles whose EMG was recorded. The frequent failure to accurately terminate the burst was suggested to be due to a greater susceptibility, whether reflexive or conditioned, to simulation arising in the contracting RF muscle itself.