EMG amplitude in maximal and submaximal exercise is dependent on signal capture rate

This study analysed the effect of different electromyographic (EMG) capture rates during maximal voluntary contraction, sub-maximal and maximal dynamic cycling activity on EMG amplitude and signal characteristics. Ten healthy subjects participated in this study. Peak power output (PPO) and maximal isometric force output (MVC) were measured, followed by a progressive cycle ride on a cycle ergometer. Electromyographic (EMG) data were simultaneously captured during the MVC and cycling activities at frequencies of 32, 64, 128, 256, 512, 1024 and 1984 Hz. Significant differences in amplitude were found (p < 0.01) between MVC, submaximal (SUB) and maximal cycling activities (PWATT) for all capture rates. Asymptote values for IEMG amplitude occurred at EMG capture rates of 1604 +/- 235.6 Hz during MVC, 503.1 +/- 236.2 Hz during PWATT and 326.2 +/- 105.4 Hz during SUB cycling activity and were significantly different (p < 0.01). No significant differences were found for force/EMG ratios between PWATT and MVC at 1984 Hz capture rates (3.8 +/- 1.7 N/V vs 2.5 +/- 0.9 N/V) while significant differences occurred at 32 Hz capture rate (6.2 +/- 3.8 vs 16.0 +/- 8.0; p < 0.01). Low correlations were found between EMG activity captured at 1984 Hz during PWATT and lean thigh volume (r = 0.36) and MVC (r = 0.32). Asymptote values found on this study suggest that data captured below 326 Hz for SUB, 503 Hz for PWATT and 1604 Hz for MVC are not reliable. Therefore apparatus capturing EMG data at low frequencies from these values cannot be used for quantitative data analyses.