In this longitudinal study, researchers from Wake Forest School of Medicine and the University of Texas Southwestern in Dallas, Texas, examined the frequency and severity of head impacts experienced by youth football players and how exposure to head impacts changes from one year to the next in returning players. The researchers then compared the resulting data with findings on neuroimaging studies obtained over consecutive years in the same athletes. The comparison demonstrated a significant positive association between changes in head impact exposure (HIE) metrics and changes in abnormal findings on brain imaging studies. Full details of this study can be found in the article “Analysis of longitudinal head impact exposure and white matter integrity in returning youth football players” by Mireille E. Kelley, PhD, and colleagues, published today (June 15, 2021) in the Journal of Neurosurgery: Pediatrics.
Illustrating the considerable variability in head impacts among football players, Kelley et al. inform us that in one football season, youth football players (10-13 years of age) can sustain between 26 and 1003 head impacts, and high school players can sustain between 129 and 1258 head impacts. Fortunately, most of these impacts do not result in concussions. In fact, most of these subconcussive impacts do not produce any acute signs or symptoms attributed to concussions.
Nevertheless, there is concern that, over time, repetitive subconcussive impacts may cause damage to contact-sport athletes’ brains. This is particularly of concern when we are speaking of young athletes, who have many years ahead of them in their sport and in their lives.
To examine changes in head impact exposure (HIE) from year to year, the researchers focused on a group of 47 athletes who participated in youth football for two or more consecutive years sometime between 2012 and 2017. The athletes played on a variety of teams. All wore football helmets outfitted with the Riddell Head Impact Telemetry System™, which measures linear and rotational head accelerations that occur during a head impact. The outfitted helmets were worn during all football sessions–both practices and games. Biomechanical data on head impacts were transmitted in real time via radio waves to a sideline data collection field unit for later analysis. Altogether, the data covered 109 football athlete-seasons with 41,148 head impacts. Despite the large number of head impacts, none of the 47 youth athletes sustained a clinically diagnosed concussion during the study period.
Kelley et al. examined a variety of HIE metrics: number of head impacts, 50th percentile of impacts per football session (game, practice, and both); 95th percentile peak linear and rotational accelerations measured by the Head Impact Telemetry System System™; and risk-weighted cumulative exposure, a metric summarizing the frequency and magnitude of head impacts experienced by the athlete over a single season. The researchers found that variations in HIE metrics differed from year to year and between athletes. For example, in an examination of data from three consecutive seasons, some youths experienced more impacts in their second year of play than in their first, while other youths experienced fewer impacts in later years of play.
Although trends in increasing mean number of game impacts, mean 50th percentile of impacts per football session, and mean 50th percentile of impacts per game session were identified over three seasons, the differences between the means were not statistically significant. The researchers did note “significant variability in several HIE metrics among teams,” but this was not the focus of this study.
In 19 of the 47 youth football athletes, brain images were obtained pre- and post-season for two consecutive football seasons. This was done using diffusion tensor imaging (DTI), a type of magnetic resonance imaging that can be used to assess the integrity of the brain’s white matter, indicating possible sites of injury. Using this technique, the researchers evaluated changes in the following scalar metrics: fractional anisotropy; mean diffusivity; and linear, planar, and spherical anisotropy coefficients.
A group of 16 youth athletes who participated in non-contact sports (for example, swimming, tennis, track) underwent DTI two times (baseline and follow-up studies, four months apart) and served as a control group. Abnormal white matter voxels (small three-dimensional areas) on youth football players’ brain images were defined as voxels in which DTI scalar values increased or decreased throughout the football season significantly (two standard deviations above or below those of mean values in the control group).
Similar to the pattern of HIE, the researchers found “both increases and decreases in the number of abnormal voxels between season 1 and season 2” in the youth football athletes.
Linear regression analyses were performed to assess relationships between changes in HIE metrics and changes in DTI scalar metrics from one year to the next. There was a significant positive correlation between changes in the number of head impacts per practice session and every DTI scalar metric. In addition, significant positive correlations were determined between changes in both the 50th percentile impacts per practice sessions and the 50th percentile impacts per football sessions (including both practice and game sessions) and the various DTI scalar metrics.
In summary, the major findings of this longitudinal study include the following:
- HIE varied among individual athletes from one season to the next. Increases and decreases in HIE across three consecutive seasons in individual youth football players were observed; the changes were not significant from one season to the next.
- Trends in DTI imaging changes varied among individual athletes from one season to the next. Increases and decreases in the number of abnormal voxels on DTI from the first to second season were identified in individual players.
- The amount of HIE an athlete experienced in football, particularly in practice, was associated with the amount of change in neuroimaging metrics. Positive associations between changes in abnormal voxels on DTI and the number of head impacts per practice session, 50th percentile impacts per practice session, and 50th percentile impacts per session overall (including both practice and game sessions) between consecutive seasons (seasons 1 and 2) were found.
Because of the significant positive correlations between changes in HIE metrics and changes in the number of abnormal voxels on DTI between consecutive seasons, the authors support efforts to reduce the number and frequency of head impacts, particularly those occurring during practice sessions, when the majority of head impacts occur. They believe this action may reduce the number of abnormal imaging findings in youth football athletes from one football season to the next.
When asked about the findings of the study, Dr. Jillian Urban, Assistant Professor at Wake Forest School of Medicine, responded, “Our findings further support ongoing efforts to reduce the number of head impacts in football practices. In an upcoming study, we plan to engage stakeholders in the youth football community to develop and test practical solutions informed by the biomechanical data we collect on field to reduce head impacts in practice.”
Reference: “Analysis of longitudinal head impact exposure and white matter integrity in returning youth football players” by Kelley, ME, Urban JE, Jones DA, Davenport EM, Miller LE, Snively BM, Powers AK, Whitlow CT, Maldjian JA, Stitzel JD, 15 June 2021, Journal of Neurosurgery: Pediatrics.
Drs. Kelley, Urban, Jones, Miller, and Stitzel are affiliated with both Wake Forest School of Medicine and Virginia Tech-Wake Forest School of Biomedical Engineering and Sciences in Winston-Salem, North Carolina. Drs. Snively, Powers, and Whitlow are affiliated with Wake Forest School of Medicine. Drs. Davenport and Maldjian are affiliated with the University of Texas Southwestern in Dallas, Texas.
Disclosure: The authors report no conflict of interest concerning the materials or methods used in this study or the findings specified in this paper.
Grant Support: Three grants from the National Institutes of Health, namely National Institute of Neurological Disorders and Stroke grants R01NS094410 and R01NS082453, and National Center for Advancing Translational Sciences grant KL2TR001421.