Characteristics of muscle co-contraction during isometric tracking

TR Number
Journal Title
Journal ISSN
Volume Title
Virginia Tech

The purpose of this research was to study the relationship between muscle coordination and the performance of a simple manual tracking task. The study employed an isometric, zero order, pursuit tracking task with a laterally translating, periodic sine wave forcing function. The speed of the target was varied by altering the Frequency (3 levels) of the simple sine wave. The control/response ratio for each trial was manipulated by requiring a percentage of each subject’s flexion and extension maximum voluntary contraction effort (MVC, 5 levels) to track the target. Multiple electromyograms (EMGs) of the biceps and triceps muscle groups were taken to observe flexor and extensor activity during the tracking task. Muscle modeling techniques were used to quantify the force contributions from the biceps and triceps to the observed tracking force.

It was hypothesized that significant levels of co-active muscle effort would be present during the tracking task and that this co-contraction would have a unique characteristic function about the tracking conditions which yielded optimal tracking performance. The dependent measures investigated were the absolute tracking error as a proportion of the required tracking force (proportional error, PE), the absolute antagonist muscle force (AAF), and the ratio of antagonist to agonist force (co-contraction ratio, CR). Each muscle group’s maximum muscle force (MMF) required to track each condition was also determined. The experimental design was a 3 by 5 by 2 mixed factor, repeated measures ANOVA with Gender (5 male, 5 female) as the blocking variable.

The ANOVA results revealed that both target Frequency and tracking Force level had significant effects on tracking error (PE). Orthogonal polynomial contrasts showed that the Frequency effect was characteristically linear while the Force effect was quadratic in nature. A polynomial regression function was used to predict PE from the Force and Frequency conditions. This model accounted for over 96% of the variance in the PE cell means. Further analysis revealed the optimal Force level for isometric tracking to be approximately 61% MVC.

Analysis of the force contributions from each muscle group revealed quadratic relationships for the actual muscle force (%MMF) of the biceps during flexion and of the triceps during extension. These results show that optimal tracking performance during flexion occurs at approximately 66% of the biceps MMF and 65% of the triceps MMF during extension. Actual MMF values were consistently larger than net force values indicating that due to the presence of co-contraction, the measured force output at the wrist underestimated the actual muscle forces involved in tracking.

Neither Force nor Frequency had significant effects on absolute co-activity (AAF) showing that antagonist activity remained largely constant over the tracking conditions. However, co-activity was higher for the extension phase than for the flexion phase of the task.

Both Force and Frequency had significant effects on the co-contraction ratio (CR). However, no characteristic function of co-activity was found to explain the optimal tracking performance at median levels of flexion and extension force. CR increased with increasing target speed (Frequency) while it decreased with higher tracking Force levels. Since antagonist activity (AAF) remained almost constant, these results for CR must be due to changes in the level of agonist activity needed to perform the tracking task.

Higher co-contraction was also found during decreasing force production (release) than for increasing force production (exertion). Since there was no significant difference in tracking error for these parts of the task, co-activity may serve to facilitate tracking performance by controlling the rate of force release.