Neuromuscular fatigue differs with biofeedback type when performing a submaximal contraction

The aim of the study was to examine alterations in contractile and neural processes in response to an isometric fatiguing contraction performed with EMG feedback (constant-EMG task) when exerting 40% of maximal voluntary contraction (MVC) torque with the knee extensor muscles. A task with a torque feedback (constant-torque task) set at a similar intensity served as a reference task. Thirteen men (26 +/- 5 yr) attended two experimental sessions that were randomized across days. Endurance time was greater for the constant-EMG task compared with the constant-torque task (230 +/- 156 s vs. 101 +/- 32 s, P < 0.01). Average EMG activity for the knee extensor muscles increased from 33.5 +/- 4.5% to 54.7 +/- 21.7% MVC EMG during the constant-torque task (P < 0.001), whereas the torque exerted during the constant-EMG task decreased from 42.8 +/- 3.0% to 17.9 +/- 5.6% MVC torque (P < 0.001). Comparable reductions in knee extensors MVC (-15.7 +/- 8.7% for the constant-torque task vs. -17.5 +/- 9.8% for the constant-EMG task, P > 0.05) and voluntary activation level were observed at exhaustion. In contrast, excitation-contraction coupling process, assessed with an electrically evoked twitch and doublet, was altered significantly more at the end of the constant-EMG task despite the absence of M-wave changes for both tasks. Present results suggest that prolonged contractions using EMG biofeedback should be used cautiously in rehabilitation programs. (c) 2006 Elsevier Ltd, All rights reserved.