Background: Boxing has been an event in the Canada Winter Games since 1971, providing a platform for young athletes to showcase their skills. However, due to decreasing popularity of the sport as well as concerns over safety, it was removed from the 2015 Games. Youth participation in boxing has been controversial, as it requires balancing the benefits of participation in boxing programs with the potential risks posed by a sport involving repetitive head impacts. Understanding injury risks is an important first step for informing future decisions about boxing’s role in the Games and ensuring the safety of all youth athletes involved.

Purpose: To examine concussion incidence, rates, and mechanisms for athletes participating in boxing at the Canada Games.

Methods: This retrospective cohort study examined registration data and medical reports from the 2011 and 2019 Canada Winter Games. New and subsequent injuries, directly and indirectly related to boxing participation, were coded in alignment with the 2020 International Olympic Committee consensus statement on methods for recording and reporting epidemiological data on injury and illness in sport. Incidence rates (injuries per 1000 athlete-days) and clinical incidence (injuries per 100 athletes) with 95% confidence intervals (CI) were calculated for the two Games.

Resutls: In total, 108 males (67 in 2011, 41 in 2019) [mean age (range): 17.0 (15.2-18.1)] participated in 3-day or 4-day events, with a total of 387 athlete-days across the two Games. Of these athletes, 47 sustained at least one injury, with a total of 51 injuries directly related to participation. There were eight head injuries directly related to participation across the two Games; five of which were concussions. In 2011, three concussions were reported [13.1 concussions per 1000 athlete‐days (95% CI 3.5‐48.6)] followed by two in 2019 [12.6 concussions per 1000 athlete‐days (95% CI 3.8‐42.1)], all resulting from direct contact with another athlete. The clinical incidence of concussion in both 2011 and 2019 were similar [4.5 concussions (95% CI 0.9-13.1); 4.9 concussions (95% CI 0.6-17.6) per 100 athletes, respectively]. Of athletes who sustained a concussion, one athlete reported having a previous concussion within 1-year of participating in the Games.

Conclusions: Approximately 1 in every 20 youth boxers were estimated to sustain a concussion at the Canada Games. The clinical incidence of concussion in boxing at the 2011 (4.5%) and 2019 (4.9%) Canada Games is higher than previous reports of concussion in Olympic boxing events which ranged between 0 to 3.2%. Despite increased concussion awareness and improved concussion recognition tools, concussion rates remained consistent between both Games. These findings highlight an important starting point for future studies examining concussion prevention strategies. Continuing to explore injury rates and mechanisms at the Games are an important step to help protect young athletes.

Introduction: Emerging evidence suggests that direct or indirect impact on the head alters white matter (WM) tract integrity even without clinical symptoms. Long-term exposure to repetitive sub-concussive impact has been associated with neurocognitive impairments, with or without a defined impact injury. Diffusion Tensor Imaging (DTI), an advanced Magnetic Resonance Imaging, has been used to study structural changes in the brain by quantifying the direction and magnitude of water diffusion in the WM tracts and the brain. This study aims to use DTI to characterise the effect of playing a season of contact sports on brain microstructure among contact sport athletes (1) relative to baseline scans and (2) non-contact sport athletes.

Methods: Thirty-three male contact sports players underwent brain scans (3.0 T GE MRI scanner, multi-shell diffusion MRI, and 3D-T1w sequences) at three-time points: (1) pre-season, (2) mid-season, and (3) post-season. Twenty age-matched non-contact sports male athlete controls underwent a single timepoint scan at baseline. Diffusion images were preprocessed using FMRIB's Diffusion Toolbox, and diffusion tensors were reconstructed using Mrtrix3. Tract-based Spatial Statistics (TBSS) was used to compare diffusion metrics - fractional anisotropy (FA), mean diffusivity (MD), axial diffusivity (AD), and radial diffusivity (RD) - between groups. Independent t-tests compared controls and contact sport players across three-timepoints, while paired t-tests assessed within-group changes in contact sport players between two-timepoints. Cohen's d (d) was used to quantify the effect size of differences between groups in the identified significant cluster, while its size is expressed as a percentage of the whole WM skeleton, which TBSS compared across subjects.

Results: TBSS identified significant clusters, where compared to the control group, RD was higher for rugby players in postseason (p < 0.05, d = 1.182, 4.24%), and FA was lower in midseason (p < 0.05, d = 1.986, 1.55%), and postseason (p < 0.05, d = 1.733, 13.85%). Prominent WM tracts identified to be significantly different were the left internal and external capsule and left cerebellar peduncle. Significant clusters were identified more in the left hemisphere than in the right hemisphere of the brain. No significant difference was found for other comparisons.

Conclusions: Our results reveal statistically significant differences in white matter microstructure between rugby players and non-contact sport athletes, particularly within the left internal and external capsule and left cerebellar peduncle. Elevated RD and reduced FA values indicate potential disruption or damage to white matter integrity, likely influenced by repetitive head impacts. These findings emphasize the importance of continuous monitoring of brain health in contact with sport athletes to better understand the long-term effects of repeated sub-concussive events. Further study is warranted to characterise the effect of contact sports on brain microstructure in association with head impact measures and brain development in adolescents.

Biomarkers that support the diagnosis and prognosis of mild traumatic brain injury (mTBI) remain elusive. Consequently, management decisions for mTBI patients rely heavily on subjective symptom reports and clinical assessments. Magnetic resonance imaging (MRI) offers a promising, non-invasive method for studying changes in brain structure and function following mTBI, with the potential to identify markers that could aid in diagnostic and prognostic decisions. However, studies have primarily focused on identifying group differences between participants with mTBI and controls, yielding mixed results. There is limited literature on how MRI data can be utilized to classify or cluster acute mTBI patients versus uninjured individuals, and fewer studies have explored the integration of multimodal MRI data for this purpose. In this work, we leverage similarity-driven multi-view linear reconstruction (SiMLR) to explore how multimodal MRI datasets can advance our understanding of acute mTBI. Fifteen adolescent males (15-18 years) with acute sport-related mTBI (< 2 weeks post-injury) and 27 age- and sex-matched healthy collision sport athletes served as a comparison group. Participants underwent 3T T1-weighted, diffusion-weighted, and resting-state functional MRI scans, as each of these modalities show promise in identifying subtle features of mTBI. Diagnosis of mTBI was made by an experienced physician. Imaging-derived quantitative metrics were computed with ANTsPyMM v1.4.0, utilizing standard atlas-based anatomical and functional imaging postprocessing pipelines within the ANTsX ecosystem. This produced a matrix of over 17,000 quantitative neuroimaging metrics, including anatomically constrained cortical measurements, white matter microstructure, and functional connectivity for each participant. SiMLR was then applied to these matrices using ANTsR v0.6.1 with default settings to identify low-dimensional embeddings that optimize the joint predictive power across the three MRI modalities while controlling for age. Generalized linear models were used to explore the importance of each low-dimensional embedding in relation to group membership (acute mTBI versus control) using 1000 permutations and p < 0.05. Effect sizes were calculated for embeddings that were significant predictors of group membership. This process identified three low-dimensional embeddings that were moderate-strong predictors (d’s 0.64-0.83) of group membership. Two embeddings were derived from diffusion-weighted metrics with the third based on contributions from resting-state measures. An advantage of SiMLR is the preservation of interpretability enabling identification of features that contribute most to the joint embedding space. Using this feature, asymmetry in mean fractional anisotropy within the cingulum and anterior corona radiata, along with connectivity between the temporoparietal and executive control networks, were identified as the strongest contributors to the joint space for each significant embedding. Future directions include further optimization of SiMLR on a larger dataset and integration of additional MRI modalities such as FLAIR and perfusion imaging. Additionally, future work will explore how SiMLR features may be used to understand longitudinal symptom burden and recovery outcomes.

Background: Concussion management and return-to-play/sport medical decisions have evolved and now include exertional testing. We developed a multimodal exertional test (MET) requiring minimal resources and are assessing its validity.

Purpose: To examine physiological and subjective responses, and failure rates between healthy athletes and athletes with concussion on the MET.

Methods: In this prospective observational study, 116 healthy intercollegiate athletes (females=49, males=67) underwent the MET at a baseline assessment. Meanwhile, 49 athletes with concussion (females=36, males=13) completed the MET at various intervals post-injury (1, 2, 3, 4, 6, and 8 weeks). The study evaluated average and peak heart rates, changes in symptoms, and pass/fail outcomes of the MET. Advanced Bayesian hierarchical modelling techniques were employed to differentiate group responses.

Results: Overall, average and peak heart rates were identified to increase from pre-MET to Stage 1 (average heart rate difference = 25.4 beats per minute [bpm], 90% Credible Interval [90% CrI] = 17.0 – 33.3 bpm; peak heart rate difference = 28.6 bpm, 90% CrI = 20.3 – 36.4 bpm), maintain increased between Stages 1-3, and increase again from Stages 3-4 (average heart rate difference = 10.9 bpm, 90% CrI = 1.0 – 20.7 bpm; peak heart rate difference = 15.7 bpm, 90% CrI = 25.3 – 35.0 bpm) for healthy athletes. A similar pattern was identified in athletes with concussion who were able to successfully pass the MET. Additionally, female participants had higher average and peak heart rates compared to male participants (average heart rate difference = 4.0 bpm, 90% CrI = 1.1 – 6.7 bpm; peak heart rate difference = 4.4 bpm, 90% CrI = 1.1 – 7.7 bpm). Minimal changes in symptom provocation were identified in healthy athletes following each task, however symptom exacerbation was found in athletes with concussion. At 1-, 2-, 3-, and 4-weeks post-injury, 53%, 48%, 24%, and 16% of athletes with concussion failed the MET, respectively, although no healthy athletes failed.

Conclusions: While physiological responses to the MET were consistent across groups, athletes with concussion exhibited symptom provocation and higher failure rates. These findings underscore the potential of the MET in enhancing concussion management strategies by providing a nuanced understanding of recovery trajectories and readiness for safe return to play.

Background: Concussion is a common youth sports injury with significant implications if poorly managed. Current return-to-play assessments include symptom checks and single domain assessment of cognitive and physical recovery. However, these assessments do not reflect the speed, complexity, or skill integration required in sport and may fail to elicit symptoms associated with combined physical, cognitive and sensory demands. We developed R2Play, a multidomain concussion assessment to better simulate the demands of sport. R2Play consists of 6 tablets in a standard configuration that participants run between and tap in an alphanumerical sequence (1-A-2-B…6-F). Increasing physical (exercise task) and cognitive (inhibition task) loads are introduced in subsequent levels and each level is performed with two conditions (auditory interference; multidomain switching) for added challenge. This study explored clinical feasibility, face and content validity of R2Play.

Methods: Clinician-youth dyads completed the R2Play assessment. Heart rate (HR) and level completion times were logged. Multidomain cost scores were calculated to describe performance changes between levels and conditions. Semi-structured interviews with clinicians and youth collected feedback on their experience of R2Play. Clinicians completed the System Usability Scale (SUS). Quantitative data were analyzed via descriptive statistics. Interview transcripts were analyzed via content analysis to generate categories organized by domains of feasibility, face and content validity.

Results: Five clinicians and 11 youth (10-22 years) participated. Assessment times ranged from 30-40 minutes with only minor technical challenges (e.g., momentarily unresponsive equipment). Usability was good-to-excellent (SUS=81±8.4). Youth reported that instructions were clear and the combination of running, agility (e.g., lunging, direction changes), and complex thinking made R2Play feel sport-like. Clinicians reported that R2Play enabled observation of cognitive-motor skill integration within a sport-like task. Moderate-to-high intensity exertion (peak HR=80±11% age-predicted max) was reached. Completion times were 21.8±21.1% slower and heart rate 15.0±12.2% higher for the inhibition level relative to baseline. For the exercise level, HR was 12.7±6.5% higher relative to the baseline while completion times were equivalent. Multidomain switching and auditory interference conditions increased the completion time relative to the standard/baseline level by 23.1±10.1% and 3.8±8.3%, respectively.

Conclusions: This work demonstrates initial proof-of-concept and evidence supporting feasibility, face and content validity of R2Play. Validating R2Play’s level structure and cognitive, exercise, and sensory loading is an important first step towards using the tool in active rehabilitation and return-to-play decision-making. Next steps include establishing cross-site feasibility and psychometric properties of R2Play.

Some contact and collision sport athletes notice a decline in their memory functioning as they age. Memory impairment is a core diagnostic criterion for traumatic encephalopathy syndrome, and former professional collision sport athletes are believed to be at risk for this syndrome. The purpose of this study was to examine predictors and correlates of perceived decline in memory functioning among former elite and professional rugby league players in Australia. Participants were 130 male former elite/professional rugby league players who completed a health survey (age: mean = 51.7, SD = 12.5; 50% of the sample was 43-64 years old). Their first calendar year of elite play was before 1970 for 18.2% of the sample and before 1990 for 60.1%. Their mean number of seasons played at the elite or professional level was 9.9 (SD = 4.3). They were asked ‘Compared to how your memory was 5 years ago, how do you think that your memory is now?’ with response options of: ‘Much Better, Better, The Same, Worse, or Much Worse.’ There were 64 (49.2%) former players who endorsed having worse or much worse memory compared to 5 years ago. The two groups (i.e., those with vs. without perceived decline in memory) did not significantly differ in body mass index, number of seasons of elite or professional participation in sports, number of past concussions, or on a composite measure of signs and symptoms of concussions during their playing years (all p-values were greater than .10). The proportions of the sample with perceived memory decline versus no perceived memory decline who had certain health conditions were as follows: high blood pressure = 20.3% vs. 19.7% (p = .930), diabetes = 12.7% vs. 3.1% (p = .043), pain medication use = 46.9% vs. 33.9% (p = .131), headache medication use = 19.1% vs. 9.2% (p = .110), sleep apnea = 24.2% vs. 6.2% [p < .001, odds ratio (OR) = 4.87, 95% confidence interval (CI) = 1.52-15.63], and medications for anxiety or depression = 35.9% vs. 9.2% (p < .001, OR = 5.52, 95% CI = 2.06-14.74). In a multivariate logistic regression model that included age and those six health variables [χ2(7)=22.64, p=.002; Nagelkerke R2=.22], perceived decline in memory was marginally significantly associated with sleep apnea (OR=4.07, 95% CI=.99-16.65) and being prescribed medications for anxiety or depression (OR=3.64, 95% CI=1.22-10.87). In former professional rugby league players years of exposure to collision sport and concussion history were not associated with perceived decline in memory functioning, whereas sleep apnea and mental health problems were associated with perceived memory decline.

There is considerable interest in understanding later in life brain health in former professional football players. Cognitive functioning is a measurable marker of brain health. The purpose of this study is to examine associations between football exposures, such as years of participation and concussion signs and symptoms during playing years, and neuropsychological functioning in former professional American-style football players. Participants were 98 former professional football players who underwent neuropsychological testing for research purposes. Their median age was 48 (interquartile range=41-55). There were 51 (52%) who identified as Black or African American and 47 (48%) who identified as White. The battery included (i) six memory test scores from the Neuropsychological Assessment Battery (NAB, derived from List Learning, Story Learning, and Daily Living Memory), (ii) the NAB Naming Test, (iii) the raw copy score from the Rey Complex Figure, (iv) five test scores from the Delis-Kaplan Executive Function System (derived from Verbal Fluency and Color Word Interference), (v) the Test of Premorbid Functioning Reading Standard Score, and (vi) the Reynolds Intellectual Screening Test. Specific football exposure variables included total years of participation, years of professional play, age of first exposure to football, and concussion signs and symptoms during playing years. There were 21 individual test scores that were correlated with four football exposure variables. We defined a negligible correlation, having no practical or clinical significance, as a correlation of r=0.20 or lower, meaning that the two variables had 4% or less overlapping variance. More than 95% of the correlations were negligible. There were a few very small negative correlations between concussion signs and symptoms during playing years, memory test scores, and the category verbal fluency score. There were no small meaningful correlations between any neuropsychological test score and age of first exposure to football, years of professional play, or total years of exposure to football. In this sample of former professional football players, age of first exposure to football and years of exposure to football were not associated with any neuropsychological test score. Retrospective ratings of concussion signs and symptoms during players years had a negligible to small association with neuropsychological test scores.

Background: Knowledge translation, including updated recommendations after an international consensus, can be challenging. One way to share knowledge could be using a Massive Open Online Course (MOOC), making it available to more people, promoting global engagement.

Objectives: The objective of this preliminary study was to evaluate the feasibility and effectiveness of a MOOC as a global knowledge translation (KT) strategy in improving knowledge after the 6th International Conference on Concussion in Sport.

Design: A pre-experimental design was used to evaluate de-identified learning analytics including participation, and knowledge gains. Participants were self-registered in the MOOC on concussion spring 2024.

Intervention: The MOOC included six modules over an 8-week period: 1) Introduction to Concussion, 2) Prevention, 3) Detection, 4) Initial management, 5) Access to Care (individuals with persisting symptoms), and 6) Adaptation, Implementation and Revision of Protocols. Each module included interactive learning materials and evaluations.

Main Outcome Measurements: The main outcomes were the number of registrants; pre- and post-course exam scores. Optional survey questions documented participant type [e.g. parent, teacher or school administrator, health care professionals (HCPs), athlete, coach or sport organization stakeholder, other (non-HCP) and intent to develop/update a protocol.

Results: A total of 2,162 participants registered and 1022 (47.3%) accessed the course in the initial six weeks. The majority [(n=612/829), 73.8%] indicated being HCPs and 315/807 (39.0%) indicated the intent to develop/update a protocol. The pre-course test was completed by 859 participants (84.0%), with a mean score of 54.7% (95% CI: 34.6 - 74.8). The mean post-test score was 79.0% (95% CI: 68.0 – 90.1), with a significant difference between test results (t=-4.50, p=0.001). For new content questions from the Amsterdam consensus, the average pre-test score was 43.9% (95%CI: 21.4 – 66.3) post-test was 75.9% (95%CI: 61.2 – 90.6) (t=-6.3, p< 0.001).

Conclusions: The results suggest that the MOOC could be an effective approach to knowledge translation and effectively translate new recommendations for concussion following the Amsterdam International Consensus Statement publication.