Natural goal-directed movements and the triphasic EMG

Triphasic electromyographic (EMG) patterns have been described as characteristic of rapid, discrete, uniplanar, goal-directed movements. This experiment examined the effects of Response Type (experimenter- vs. subject-determined), Hand (preferred vs. nonpreferred), and Practice (early vs. late) on performance accuracy, and specific temporal EMG and kinematic measures during a dart throw. EMG was recorded from triceps (main agonist), brachioradialis, and biceps (main antagonist). The number of trials in which a triphasic EMG occurred varied systematically across conditions. The experimenter-determined, early practice condition resulted in greatest frequency (92%) of trials displaying a triphasic EMG and least accurate performance. In contrast, the lowest frequency (79%) of triphasic EMG and most accurate performance occurred in the subject-determined, late practice condition. The association among 14 temporal EMG, and kinematic measures for each trial of the dart throw was analyzed with multivariate factorial ANOVA. Four clusters of variables emerged: initial phase, braking phase, terminal phase, and movement speed and duration. Variables contributing to the initial-phase cluster were most strongly associated within the experimenter-determined, early practice condition, and the strength of association was directly related to diminished performance accuracy. Paradoxically, best performance accuracy (subject-determined, late practice) was identified with a weaker association among variables representing the initial phase.