I think it depends on the type of muscle contraction. If sEMG was used to detect the fatigue during isometric or constant-force fatiguing tasks the RMS would be high. However, many studies noted that most human activities are dynamic activities (Time-varying changes in forces), so that during sustained maximum voluntary contraction (sMVC), the RMS decreased by 70% when the muscle reaches 50% MVC estimating that EMG response to sMVC also diminished.

thank you Dr. Waleed

In regards to your first question, I think it is because a change in EMG parameters provides very little precision in identifying the physiological changes induced by fatigue. Supraspinal, spinal, and even muscle membrane properties are known to be altered by fatigue and each would be expected to influence EMG. For your second question, I believe research using EMG during fatigue has shown great variance in its response. In addition, the intensity (submaximal or maximal) of the fatiguing isometric contraction would be influential on the expected response. If the goal is simply to detect fatigue, while there are several proposed definitions, I believe an exercise-induced decrease in muscle force or power is most widely accepted at least with regard to "performance" fatigue. Hope this helps.

Thank you Dr. Garrett for your explanation

EMG is a measure of the electrical activity within the muscle and not necessarily indicative of what is going on mechanistically in the muscle. Fatigue can happen in several different areas during muscle contraction not just in changes in neuromuscular excitation (electrical activity). If fatigue is caused by a build up of metabolic waste in the sarcomere you could potentially see a lack of calcium reuptake (SERCA pump), which is not affected by excitation. I would strongly agree with Garrett when he says muscle fatigue is best determined by a decrease in muscle force output or power.

Thanks Dr. Micah for the explanation