Shifting of activation center in the brain during muscle fatigue: an explanation of minimal central fatigue?

Abstract

Accumulating evidence suggests that the overall level of cortical activation controlling a voluntary motor task that leads to significant muscle fatigue does not decrease as much as the activation level of the motoneuron pool projecting to the muscle. One possible explanation for this "muscle fatigue>cortical fatigue" phenomenon is that the brain is an organ with built-in redundancies: it has multiple motor centers and parallel pathways, and the center of activation may shift from one location to another when neurons in the previous location become fatigued. This hypothesis was tested by estimating the changes of source locations of high-density (64 channels) scalp electroencephalographic (EEG) signals collected during both fatigue and non-fatigue motor tasks. A current dipole model was used to estimate the EEG sources. The fatigue motor task induced significant muscle fatigue, and the non-fatigue task did not. The EEG signal source that indicated the center of brain activation showed substantial location shifts during the fatigue motor task. The shifts could not be explained by variations of source locations caused by error estimated from the non-fatigue task EEG and simulated data. Compared to the non-fatigue condition, the weighted-center of the source locations for all the participants shifted toward the right hemisphere (ipsilateral to the muscle activation), anterior, and inferior cortical regions under the fatigue condition. Fatigue did not alter dipole (source-signal) strength or the overall level of brain activation. The brain may avoid fatigue by shifting neuron populations that participate in a fatiguing motor task.

Fig. 1

Illustration of EEG-derived MRCP (upper panel), EMG (middle panel) and force (lower panel) time courses. The MRCP profile shows that following a baseline, a slow- to fast-rising negative potential (NP) occurs before initiation of the muscle contraction. The MRCP was derived from EEG signals of the C3 electrode located roughly above the left sensoimotor area. The EMG was recorded from the flexor digitorum superficialis muscle. The dipole fitting was performed using data from −50 ms (indicated by the first vertical dashed line) to the peak of the NP (second vertical dashed line). Because this 50-ms period immediately precedes the muscle activation, it was thought that the NP within this time represented cortical activities associated with execution of the motor action.

Fig. 2

Force and EMG results under the fatigue (a and b) and non-fatigue (control, c and d) conditions. For the fatigue condition, each data point represents an average of 40 consecutive 2-s handgrip contractions with a 5-s inter-trial interval or rest. The time taken for each 40-trial block was 280 s. For the control condition, each data point represents an average of 40 consecutive 2-s handgrip contractions with a 30-s inter-trial interval or rest. The time taken for each 40-trial block was 1280 s. The force and EMG were normalized to their respective MVC values acquired at the beginning of the experiments. FDS, flexor digitorum superficialis; FDP, flexor digitorum profundus; ED, extensor digitorum; R, right arm. The statistical analysis results are shown at the right of the force (a) or each muscle EMG (b) symbol in the format of (k:m-n), where k indicates the data point to be compared and m-n indicates the data points (m^th to n^th) that showed significant changes relative to the k^th data point (P ≤ 0.05). For example, in plot (a), (1:2-5) means that all data points from the 2^nd to 5^th are significantly different from the 1^st data point. No significant changes were found in control data (c and d).

Fig. 3

Single dipole plots showing locations of the center of brain activation. The locations of the moving dipoles are plotted in the X-Y (top row, top view), X-Z (middle row, back view), and Y-Z (bottom row, side view) planes. X: left to right; Y: front to back; Z: bottom to top. The triangles in the top and middle rows indicate the position of the nose, and positions of the ears are marked on two sides of each circle in the top and middle rows. In each plot, each subject is represented by a symbol, and each contraction block is represented by a color (block 1: dark blue; block 2: red; block 3: green; block 4: magenta; block 5: light blue). The ellipse in each plot represents the 95% confidence region that contains 95% of the data points in the data set. The dipole locations show much greater spatial dispersions for the fatigue (left column) than non-fatigue (right column) conditions. The centers (indicated by the cross points of the dashed straight lines) of the ellipses in the fatigue condition are located more toward the right hemisphere, anterior and inferior regions of the brain.

Fig. 4

Standard deviations of dipole locations along time/block. The plots a, b, and c show standard deviations of the dipole locations at the X, Y, and Z axes respectively. The solid and open circles represent experimental data of the fatigue and control conditions, respectively. The dashed straight lines are linear regressions. R² is the linearity of the fit and the slope of the fit indicates the trend of the data.

Fig. 5

Dipole strengths as a function of time. (a) Fatigue condition, (b) control condition. No significant changes in dipole strength were found in both conditions, indicating that the amplitude of brain activation during the period of muscle-contraction execution was not affected by fatigue (a) or time (b).