• P.M. Santos
  • A. Veloso


The myoelectric activity of agonist and antagonist muscles in fast human movements is characterized by a phasic pattern, normally composed by three bursts and known as the friphasicpatfern. The most important events of this pattern are the initial agonist activation (AGI) responsible for the initial limb acceleration, and the antagonist activation (AMI), coincident with the electrical silence of the agonist muscle. On last 15 years, a lot of experimental research was done to understand how the Central Nervous System programs ballistic movements performed at different speeds. The most part of these studies were done with single-joint movements, performed an laboratorial environments, which became more attractive as a paradigm for testing hypotheses but didn't allow direct applications to more complex and "natural" movements. Purposes of this study were (1) to characterize the activation pattern of agonist and antagonist muscles of elbow extension on a dart throw to a target, (2) to analyze the effects of movement velocity on EMG and cinematic parameters and (3) to compare the above mentioned variables on trained and untrained subjects. Experiments were performed on four skilled dart throwers and nine untrained subjects. The task was a dart throw to a concentric target at a distance of 3 m and 1,70 m of centre height. Each subject performed 30 trials on each of three different conditions. Subjects were instructed before each condition: (P) "try to reach the target centre", (C) "try to reach the target centre as fast as possible" and (V) '3ust try to be as fast as possible". Two electrogoniometers on shoulder and elbow joints were used to measure position and velocity movement parameters on each joint. Surface EMGs were recorded from triceps (vastus medialis) and biceps brachii with active bipolar electrodes. The EMG signals were filtered, fullwave rectified and smoothed. The angle, velocity and EMG signals were time normalized using the shortest movement time as a reference and averaged in blocks of 20 trials representatives of each condition. Then, the EMG bursts (AGI and ANT) were isolated for parameters determination. A laser system developed on our Faculty allowed the dart speed determination. The elbow extension on initial condition (P) were accomplished by a tri- or biphasic EMG pattern with a clear tendency for a reciprocal activation: the maximum peak of ANT was always coincident with the silent period between the two agonist bursts AG 1 was a short duration burst (90- 135 ms) which began before movement onset and ceased before the moment of elbow peak velocity. The antagonist pattern contained a phasic burst (ANT) which began before the end of AGl and presented its maximum peak near the end of the acceleration phase. This burst was normally preceded by a tonic activation which started near movement imitation. The results show that either the modulation on amplitude and timing of AGI can be seen together or separate when we compare different throwing conditions. The subjects adjusted the time and the size af the burst to the nature of the task with a possible individual predominance of one of these mechanisms. Nevertheless, the increase of the initial EMG slope, measured by the integrated EMG of the first 30 rns of AGI [iEMG30), revealed to be the most related variable with the elbow movement speed. This finding agrees with a "dual-strategy" model (Gottlieb et al., 1989a,b, 1990). The antagonist EMG presented identical timing structure between different throwing conditions, but revealed a general tendency to an increase on its intensity when subjects throw quickly. The similar behaviour of agonist and antagonist activation intensities suggests a common mechanism of magnitude control. The untrained subjects show longer acceleration phases and longer duration of EMG events (duration of AGl, time to agonist and antagonist EMG peak, time of antagonist onset) than the skilled throwers. Those trained subjects presented reduced variability on temporal structure among throwing conditions.