Subtle impairments of perceptual-motor function and well-being are detectable among military cadets and college athletes with self-reported history of concussion

Abstract

Introduction: A lack of obvious long-term effects of concussion on standard clinical measures of behavioral performance capabilities does not preclude the existence of subtle neural processing impairments that appear to be linked to elevated risk for subsequent concussion occurrence, and which may be associated with greater susceptibility to progressive neurodegenerative processes. The purpose of this observational cohort study was to assess virtual reality motor response variability and survey responses as possible indicators of suboptimal brain function among military cadets and college athletes with self-reported history of concussion (HxC).

Methods: The cohort comprised 75 college students (20.7 ± 2.1 years): 39 Reserve Officer Training Corp (ROTC) military cadets (10 female), 16 football players, and 20 wrestlers; HxC self-reported by 20 (29.2 ± 27.1 months prior, range: 3-96). A virtual reality (VR) test involving 40 lunging/reaching responses to horizontally moving dots (filled/congruent: same direction; open/incongruent: opposite direction) was administered, along with the Sport Fitness and Wellness Index (SFWI) survey. VR Dispersion (standard deviation of 12 T-scores for neck, upper extremity, and lower extremity responses to congruent vs. incongruent stimuli originating from central vs. peripheral locations) and SFWI response patterns were the primary outcomes of interest.

Results: Logistic regression modeling of VR Dispersion (range: 1.5-21.8), SFWI (range: 44-100), and an interaction between them provided 81% HxC classification accuracy (Model χ ²[2] = 26.03, p < .001; Hosmer & Lemeshow χ ²[8] = 1.86, p = .967; Nagelkerke R ² = .427; Area Under Curve = .841, 95% CI: .734, .948). Binary modeling that included VR Dispersion ≥3.2 and SFWI ≤86 demonstrated 75% sensitivity and 86% specificity with both factors positive (Odds Ratio = 17.6, 95% CI: 5.0, 62.1).

Discussion/conclusion: Detection of subtle indicators of altered brain processes that might otherwise remain unrecognized is clearly important for both short-term and long-term clinical management of concussion. Inconsistency among neck, upper extremity, and lower extremity responses to different types of moving visual stimuli, along with survey responses suggesting suboptimal well-being, merit further investigation as possible clinical indicators of persisting effects of concussion that might prove to be modifiable.

Keywords: executive function; functional connectivity; intra-Individual variability; mild traumatic brain injury; reaction time; virtual reality.

Figure 1

Virtual reality head-mounted display view depicting green response target spheres prior to T-pose calibration procedure. Throughout 40-trial test, response targets located beyond peripheral field of view with head in neutral position (i.e., neck rotation to left or right required to locate correct response target).

Figure 2

Depictions of head-mounted display views of horizontally moving circles, with arrows representing circle movement directions. Each panel depicts a unique combination of initial moving circle location (center versus periphery) and type (open/incongruent versus solid/congruent). Determination of stimulus type and movement direction required to locate correct response target.

Figure 3

Reaching/lunging to contact virtual reality response target with hand controller.

Figure 4

Most common complaints among self-reported history of concussion (HxC) cases, with corresponding frequency of each complaint for no concussion (NoC) cases.

Figure 5

Prevalence of self-reported concussion within subgroups defined by binary categorizations of 12-metric VR dispersion and SFWI.

Figure 6

Inverse correlation between 12-metric VR dispersion and rate correct per second.

Figure 7

Graphic depiction of within-subject variability in response times for 40-trial virtual reality tests of 2 college football players: low variability (A) and high variability (B).