Abstract
Studies using blood‐oxygenation‐level‐dependent functional magnetic resonance imaging (BOLD fMRI) have characterized how the resting brain is affected by concussion. The literature to date, however, has largely focused on measuring changes in the spatial organization of functional brain networks. In the present study, changes in the temporal dynamics of BOLD signals are examined throughout concussion recovery using scaling (or fractal) analysis. Imaging data were collected for 228 university‐level athletes, 61 with concussion and 167 athletic controls. Concussed athletes were scanned at the acute phase of injury (1–7 days postinjury), the subacute phase (8–14 days postinjury), medical clearance to return to sport (RTS), 1 month post‐RTS and 1 year post‐RTS. The wavelet leader multifractal approach was used to assess scaling () and multifractal () behavior. Significant longitudinal changes were identified for , which was lowest at acute injury, became significantly elevated at RTS, and returned near control levels by 1 year post‐RTS. No longitudinal changes were identified for . Secondary analyses showed that clinical measures of acute symptom severity and time to RTP were related to longitudinal changes in . Athletes with both higher symptoms and prolonged recovery had elevated values at RTS, while athletes with higher symptoms but rapid recovery had reduced at acute injury. This study provides the first evidence for long‐term recovery of BOLD scale‐free brain dynamics after a concussion.
Highlights
To date, most of the BOLD fMRI research in this domain has focused on healthy adults and has examined how scaling is modulated by cognitive tasks and other sources of effort (Barnes et al, 2009; Churchill et al, 2016; Ciuciu et al, 2012; He, 2011)
The present study is the first to examine the relationship between scaling behavior in resting-state fMRI and sport-related concussion, with reference to a large control group
This study is the first to show that functional brain dynamics are significantly altered in the course of concussion recovery, with effects that are modified by clinical variables of acute symptom severity and time to return to sport (RTS)
Summary
Sixty-one concussed athletes were recruited from university-level sport teams at a single institution (including volleyball, hockey, soccer, football, rugby, basketball, lacrosse, and water polo; see Supplementary Table S1 for athlete numbers by sport) through the academic sport medicine clinic, following a concussion diagnosis. Log-cumulant values were averaged over all ROIs showing significant longitudinal effects, and at each imaging session (ACU, SUB, RTS, 1MO, 1YR) concussed and control group means were compared. Given the demographic effects seen in the control cohort (see Section 2.5), a GLM was used to evaluate the effects of concussion with covariates adjusting for age, sex and history of concussion This was done in a bootstrap resampling framework (1,000 iterations) to obtain coefficient b with a 95%CI, BSR, and p-value for each imaging session, with significant imaging sessions identified at an FDR threshold of 0.05. A secondary set of analyses, performed within the concussed cohort, tested for effects of clinical covariates on log-cumulant values over the course of recovery This included (a) total symptom severity at acute injury and (b) days to RTS. Afterwards, logcumulant values were averaged across significant ROIs and the values reanalyzed with a bootstrapped LMM to obtain summary statistics, including a fixed-effect coefficient b with a 95%CI, BSR, and p-value
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
