Gustavo Sandri Heidner, Caitlin O'Connell, Zachary Domire, Chris Mizelle, Patrick Rider, Nicholas Murray. East Carolina University, Greenville, NC.

BACKGROUND: Traumatic brain injuries are a major health concern in the US, accounting for approximately 224,000 hospitalizations and 61,000 deaths in 2017. Fatiguing activities create a mismatch between central nervous system control and the mechanical output, because muscles have lower neural excitation and are less responsive to balance perturbations. Evaluation of regional EEG patterns between fatigued concussed participants could shed light on this as we should see upregulated activity in the areas with higher sensory weighting and could tell us if concussion impairs shifts in multimodal sensorimotor integration. METHODS: Thirty-two (N = 32) participants, seventeen (n = 17) concussed, completed the research protocols. Non-concussed participants had their balance perturbed by a whole-body fatigue protocol, a 30-minute walk on an inclined treadmill (1.9 m/s) while wearing a weighted backpack (27 kg). Participants were assessed under a variety of conditions: eyes-closed (EC), eyes-open (EO), single leg (SL), mental distraction (SS), unstable surface (US), virtual reality-induced optical flow (VR), and virtual reality baseline (VB). EEG was measured pre- and post-fatigue. For participants with mTBI, EEG was only measured once. A four-way ANOVA was conducted to compare the groups across all conditions. The dependent variable was spectral power. RESULTS: After visual inspection, data processing, and removal of outliers, there were 2802 (92.3%) valid and 236 (7.7%) missing data points. Spectral power was normalized by taking the natural logarithm of the data points. Natural log spectral power (LnP; M = 2.60, SD = 1.57) was positively skewed (Skew/SESkew = 10.69, p < .001, two-tailed) and mesokurtotic (Kurt/SEKurt = 1.14, p = .127, two-tailed). The test of between-subjects effects revealed a significant main effect of Condition, F(6, 2606) = 8.08, p < .001, ηp2 = .018, and significant interactions between Region x Bandwidth, F(6, 2606) = 4.64, p < .001, ηp2 = .011, Region x Group, F(6, 2606) = 17.34, p < .001, ηp2 = .038, and Bandwidth x Group, F(1, 2606) = 22.79, p < .001, ηp2 = .009. Spectral power was generally lower in the concussed group. CONCLUSIONS: Fatigue appears to have a distinct neural spectral power signature when compared to acute concussion. More specifically, fatigue spectral power was greater in the fatigued group than in the concussed group.

This document is currently not available here.