Comparison of automatic and voluntary chewing patterns and performance.

Chewing, like respiration, is ordinarily performed as an automatic motor act, yet both can be voluntarily controlled. No in-depth analyses of voluntary chewing exist. Therefore, we have analyzed on a cycle-by-cycle basis voluntarily controlled chewing, and compared it with automatic chewing. We assessed the performance during voluntarily controlled chewing by obtaining constant error and variable error scores. Nine healthy adults with full dentition were subjects (Ss). Their three-dimensional jaw movements and movement times were derived from Kinesiograph recordings obtained while chewing a standard piece of gum on the right side of the mouth. Burst durations and onset latencies of masseter activity were obtained from surface-recorded EMGs. Frequency during automatic chewing was obtained from data recorded while the subject viewed a film. Next, the subject chewed in time with a metronome set at this "automatic" rate. Intrasubject variability among 30 consecutive chewing cycles during voluntary was less than during automatic chewing. In every S gape and ipsilateral jaw excursions and the variability of burst durations of masseter activity were less during voluntary than during automatic chewing, showing that both the spatial and temporal aspects of the two types of chewing differ significantly. Ss varied in ability to follow the metronome. A S's constant error might be small, yet his variable error might be large, as if feed-back-based corrections influenced cycle-to-cycle variability. Fast chewers had smaller constant and variable error scores than did slow chewers, suggesting a speed-accuracy relationship. In summary, both temporal and spatial aspects of voluntary chewing were modified compared with those of automatic chewing. During voluntarily controlled chewing, cycle-to-cycle variability was less compared with automatic chewing due to reductions in variability of occlusal phase and the masseter's burst durations, and total jaw excursions were less because gapes and ipsilateral deviations during closing were reduced.