Fatigue-induced changes of impedance and performance in target tracking

Abstract

Kinematic variability is caused, in part, by force fluctuations. It has been shown empirically and numerically that the effects of force fluctuations on kinematics can be suppressed by increasing joint impedance. Given that force variability increases with muscular fatigue, we hypothesized that joint impedance would increase with fatigue to retain a prescribed accuracy level. To test this hypothesis, subjects tracked a target by elbow flexion and extension both with fatigued and unfatigued elbow flexor and extensor muscles. Joint impedance was estimated from controlled perturbations to the elbow. Contrary to the hypothesis, elbow impedance decreased, whereas performance, expressed as the time-on-target, was unaffected by fatigue. Further analysis of the data revealed that subjects changed their control strategy with increasing fatigue. Although their overall kinematic variability increased, task performance was retained by staying closer to the center of the target when fatigued. In conclusion, the present study reveals a limitation of impedance modulation in the control of movement variability.

Fig. 1

Schematic description of the experiment. a Section of a tracking trial. The gray area indicates the target area and the black line the projection of the pointer position. b Section of the fatigue protocol. Time varying torque is indicated in gray and the elbow angle in black. c Time line of the experiment. Unfatigued tracking (UFT) and post fatigue tracking (PFT) trials were all followed by 1-min rest. The fatigue protocol (FP) was followed immediately by fatigued tracking (FT). After the last UFT trial and after the last FT trial, subjects had a 5-min rest period. Ratings of perceived exertion were assessed immediately after UFT, FP, FT and PFT and are indicated by the vertical arrows

Fig. 2

Mean ratings of perceived exertion (RPE) over the entire experimental protocol. The gray area indicates ±1 SD. Ratings were obtained immediately after un-fatigued tracking (UFT), the fatigue protocol (FP), fatigued tracking (FT) and post-fatigue tracking (PFT). RPE scores after FT were used for statistical comparisons between conditions

Fig. 3

Mean and SD over subjects of tracking performance measures, itemized for the four repeated trials (dark to light bars) and the three experimental conditions [un-fatigued tracking (UFT), fatigued tracking (FT), post fatigue tracking (PFT)]. Note that for UFT and FT the results are based on ten subjects and for PFT on five subjects. This is indicated by the vertical dashed line. See main text for explanation of the performance variables

Fig. 4

Post fatigue tracking (PFT) performance. Five subjects performed an additional PFT task. Repeated measures ANOVAs were executed for the mean of the performance variables in the last two trials of each condition. Stiffness values were averaged over movement directions. Overlines indicate P-values smaller than 0.05

Fig. 5

Control variability. Left column shows the mean performance measures from minute to minute un-fatigued (UFT), fatigued (FT) and post fatigue tracking (PFT) for a typical subject. Right column shows the mean and SD for all subjects of the standard deviation over 8-min for UFT and FT to show that the depicted effect generalized across subjects. Statistical significance is also indicated. First row: percentage time-on-target (%TT), second row: percentage of time that the pointer lagged the center of the target (%LAG), third row: mean distance to the center of the target (MDT), fourth row: SD of the distance to the center of the target (SDDT), fifth row: RMS distance to the center of the target (RMSDT)

Fig. 6

Mean and SD of stiffness (K, left) and damping (B, right) for un-fatigued (UFT), fatigued (FT) and post fatigue (PFT) tracking. Left and right bars indicate, respectively, estimates during elbow extension and elbow flexion

Fig. 7

Mean and SD over subjects of speed pulse characteristics, optimized for the four repeated trials and the three experimental conditions. Note that for un-fatigued tracking (UFT) and fatigued tracking (FT) the results are based on ten subjects and for post fatigue tracking (PFT) on five subjects. This is indicated by the vertical dashed line. See main text for explanation of the speedpulse characteristics