Fatigue effect on low-frequency force fluctuations and muscular oscillations during rhythmic isometric contraction

Abstract

Continuous force output containing numerous intermittent force pulses is not completely smooth. By characterizing force fluctuation properties and force pulse metrics, this study investigated adaptive changes in trajectory control, both force-generating capacity and force fluctuations, as fatigue progresses. Sixteen healthy subjects (20-24 years old) completed rhythmic isometric gripping with the non-dominant hand to volitional failure. Before and immediately following the fatigue intervention, we measured the gripping force to couple a 0.5 Hz sinusoidal target in the range of 50-100% maximal voluntary contraction. Dynamic force output was off-line decomposed into 1) an ideal force trajectory spectrally identical to the target rate; and 2) a force pulse trace pertaining to force fluctuations and error-correction attempts. The amplitude of ideal force trajectory regarding to force-generating capacity was more suppressed than that of the force pulse trace with increasing fatigue, which also shifted the force pulse trace to lower frequency bands. Multi-scale entropy analysis revealed that the complexity of the force pulse trace at high time scales increased with fatigue, contrary to the decrease in complexity of the force pulse trace at low time scales. Statistical properties of individual force pulses in the spatial and temporal domains varied with muscular fatigue, concurrent with marked suppression of gamma muscular oscillations (40-60 Hz) in the post-fatigue test. In conclusion, this study first reveals that muscular fatigue impairs the amplitude modulation of force pattern generation more than it affects the amplitude responsiveness of fine-tuning a force trajectory. Besides, motor fatigue results disadvantageously in enhancement of motor noises, simplification of short-term force-tuning strategy, and slow responsiveness to force errors, pertaining to dimensional changes in force fluctuations, scaling properties of force pulse, and muscular oscillation.

Figure 1. Experiment procedure, system setup, and data recording.

(a) Schematic diagrams of the apparatus and fatigue protocol, (b) Typical recordings of force and target signals in the pre-fatigue and post-fatigue tests. The force signal was decomposed into a ideal force trajectory and a force pulse trace, visibly smaller after fatiguing contraction. Measures of amplitude and duration of a force pulse in the force pulse trace are displayed.

Figure 2. Contrasts of spectral features of force pulse trace between the pre-fatigue and post-fatigue tests.

Spectral profiles of force pulse trace for two typical subjects are shown in the left plots. Fatigue effect on mean frequency, mode frequency and spectral dispersion of force pulse trace is summarized in the right plots. (^***: Post-fatigue<Pre-fatigue, P<.001; ^†: Post-fatigue>Pre-fatigue, P<.05).

Figure 3. Contrasts of SampEn versus time scales and MSE area between pre-fatigue and post-fatigue tests.

Time scale for MSE plot in temporal domain is 10(^**: Post-fatigue<Pre-fatigue, P<.01; ^†††: Post-fatigue>Pre-fatigue, P<.001)(SampEn: sample entropy; MSE: multi-scale entropy).

Figure 4. Fatigue effect on pooled spectral profiles and spectral features of the rectified EMG.

Two spectral peaks in the 8–12 Hz and 40–60 Hz bands were noted, and the 40–60 Hz spectral peak was visibly suppressed after fatiguing exercise. (^***: Post-fatigue<Pre-fatigue, P<.001).