•  
  •  
 

THE EFFECT OF FATIGUE ON THE LONG-TERM CORRELATION OF ACCELERATION DURING AN UNSTABLE SERIAL TASK

Abstract

Dave P. Heller,1 Samantha Bostick,1 & Gracyn Carmean1

1Rockhurst University, Kansas City, Missouri

Fatigue affects movement kinematics. Previous research has used detrended fluctuation analysis (DFA) to investigate the long-term correlation of kinematic variables during movement. PURPOSE: The purpose of this study was to investigate the effect of fatigue on long-term correlation of the net acceleration (a-net) of the center of mass (CoM) during an unstable, serial task. METHODS: 10 participants (6F, 4M; 26 + 9y; 168.8 + 11.0 cm; 67.0 + 12.0 kg) completed a fatiguing protocol of bodyweight squats in time to a metronome (46 bpm) on the soft end of a Both Sides Up ball. We sampled three-dimensional CoM acceleration at 100 Hz using an inertial measurement unit. We used DFA to calculate the scaling exponent (a) of the net acceleration vector of the CoM time series (aa-net,CoM) using box sizes ranging from N = 10 to N = 300. To determine the effects of fatigue, we used dependent t-tests to compare the slope of the DFA log-log graph (i.e., aa-net,CoM) for the 1st and last 30s prior to and following a transition point in the slope previously described in research. RESULTS: The difference in the transition point of the slope of the log-log graph determining aa-net,CoM between the 1st and last 30s approached statistical significance (1st 30s: 76 + 15 vs. last 30s: 85 + 13; p = .079). There was a statistically significant difference between aa-net,CoM before and after the transition point (Before: a = 1.32 + 0.11 vs. After: a = 0.54 + 0.14; p <0.001). There was no statistically significant difference in aa-net,CoM between the 1st and last 30s prior to the transition point (1st 30s: a = 1.33 + 0.13 vs. last 30s: a = 1.30 + 0.09; p = .296). Following the transition point, there was a statistically significant difference in aa-net,CoM between the 1st and last 30s (1st 30s: a = 0.63 + 0.13 vs. last 30s: a = 0.46 + 0.11; p < .001). CONCLUSION: We found differences in the long-term correlation of the net acceleration vector of the CoM following the transition point of the log-log graph used to determine aa-net,CoM. This difference was not only numerical but also categorical exhibiting a change from persistent (a > 0.5) to anti-persistent (a < 0.5) acceleration. This difference could indicate a change in kinematics in a serial task on an unstable surface due to fatigue. Further research will investigate if these effects are present in continuous tasks such as gait.

This document is currently not available here.

Share

COinS