Subconcussive head impact exposure differences between drill intensities in U.S. high school football

Purpose: USA Football established five levels of contact (LOC) to guide the intensity of high school football practices. However, it remains unclear whether head impact exposure differs by LOC. The purpose of this study was to examine head impact frequency and magnitude by LOC in the overall sample and three position groups. Methods: This longitudinal observational study included 24 high school football players during all practices and games in the 2019 season. Players wore a sensor-installed mouthguard that monitored head impact frequency, peak linear acceleration (PLA), and rotational head acceleration (PRA). Practice/game drills were filmed and categorized into 5 LOCs (air, bags, control, thud, live), and head impact data were assigned into 5 LOCs. Player position was categorized into linemen, hybrid, and skill. Results: A total of 6016 head impacts were recorded during 5 LOCs throughout the season. In the overall sample, total number of impacts, sum of PLA, and PRA per player increased in an incremental manner (air < bags < control < thud < live), with the most head impacts in live (113.7-17.8 hits/player) and the least head impacts in air (7.7-1.9 hits/player). The linemen and hybrid groups had consistently higher impact exposure than the skill group. Average head impact magnitudes by position group were higher during live drills (PLA (41.0-45.9g) and PRA (3.3-4.6 krad/s^2) per head impact), whereas other LOCs had lower magnitudes (PLA (18.2-23.2g) and PRA (1.6-2.3krad/s^2) per impact). Conclusion: Our data suggest that LOC may influence cumulative head impact exposure in high school football, with players incurring frequent head impacts during live, thud, and control. The data indicate the importance of considering LOCs to refine practice guidelines and policies to minimize head impact burden in high school football athletes.


INTRODUCTION
The long-term consequence of sport-related head injury is a complex public health issue 54 with no concrete solution (1, 2). Despite inherent risk of head injury in contact sports (e.g., 55 American football, hockey, soccer), participating in these team sports, especially during 56 developmental age, provides well-documented benefits, including higher levels of physical 57 activity, improved mental health, and lower likelihood of smoking cigarettes and using illegal 58 substances (3). In an attempt to promote a safer environment, youth football organizations have 59 implemented the teaching of proper tackling techniques and reduced the size of the field and 60 number of players on the field (4). These modifications are further substantiated in high school, 61 college, and professional football, where kickoffs and touchbacks were adjusted to reduce 62 injuries (5), well-designed protective helmets are worn, and tackling of defenseless players and 63 blindside blocking are prohibited (6, 7). Consistent with these adjustments, concussion and 64 catastrophic injury rates have been reduced (8, 9); however, subconcussive head impact exposure 65 has proven more complex. 66 Subconcussive head impact is defined as a hit to the head that does not induce overt 67 concussion symptoms (10). These head impacts are most common in American football, where 68 athletes can experience several hundred impacts with some exceeding 1,000 head impacts in a 69 single season (11). Evidence has emerged to indicate that both high school and college football 70 players with frequent subconcussive head impact exposure exhibit neuronal microstructural 71 damage (12, 13), abnormal brain activation (14, 15), ocular-motor impairment (16), and 72 elevation in brain-injury blood biomarkers (17). One line of research suggests that long-term 73 exposure to these hits is a key factor in developing neurodegenerative disorders later in life (18,74 19). Although USA Football (for high school) and the NCAA (for college) have eliminated 75 All rights reserved. No reuse allowed without permission. author/funder, who has granted medRxiv a license to display the preprint in perpetuity. Therefore, we conducted a longitudinal observational study to examine cumulative head 91 impact exposure across different drill intensities in high school football players over the course 92 of a single season. Our primary hypothesis was that there would be an incremental increase in 93 cumulative head impact kinematics between LOCs in the overall sample with live recording the 94 greatest head impact exposure and air recording the lowest: air < bags < control < thud < live. 95 Since the proximity to opponents and nature of contact during nearly every play for linemen 96 (23), we also hypothesized that linemen would have greater head impact exposure in most LOCs, 97 compared to the hybrid and skill positions. Our secondary aim was to identify the average head 98 All rights reserved. No reuse allowed without permission. author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint (which was not peer-reviewed) is the . https://doi.org/10.1101/2020.03.29.20047167 doi: medRxiv preprint impact magnitudes by LOC and position group. Lastly, we aimed to explore frequency of head 99 impacts that were within 25-60 g, 60-100 g, or >100 g in each LOC. author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint (which was not peer-reviewed) is the . https://doi.org/10.1101/2020.03.29.20047167 doi: medRxiv preprint varsity games, and 9 junior varsity games over a 12-week season. Video data were collected 122 using Hudl (Agile Sports Technologies, Inc.) during the same timeframe as subconcussive head  author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint (which was not peer-reviewed) is the . https://doi.org/10.1101/2020.03.29.20047167 doi: medRxiv preprint number of hits, peak linear acceleration, and peak rotational acceleration were used for 145 descriptive analyses. Timestamps of all impacts were used to correspond with timeframes of 146 each LOC.  author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint (which was not peer-reviewed) is the . https://doi.org/10.1101/2020.03.29.20047167 doi: medRxiv preprint A series of descriptive analyses was conducted to examine all aims. Our outcome 168 variables for primary and secondary aims were head impact count, sum of peak linear 169 acceleration, and sum of peak rotational acceleration. These outcomes were expressed as mean   Table 1. 182 183 LOC-dependent cumulative head impact exposure 184 LOCs displayed an influence on head impact exposure in high school football players, as 185 illustrated by the increasing cumulative exposure corresponding to increased levels of contact. 186 Specifically, in the overall sample, total number of impacts per player as well as sum of PLA and 187 PRA increased in an incremental manner (air < bags < control < thud < live), with the most head 188 impacts being observed in live (113.7  17.8 hits, 2,657.6  432.0 g, 233.9  40.1 krad/s 2 ) and 189 the least head impact exposure in air (7.7  1.9 hits, 176.9  42.5 g, 16.7  4.2 krad/s 2 ) ( Figure   190 All rights reserved. No reuse allowed without permission. author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint (which was not peer-reviewed) is the . https://doi.org/10.1101/2020.03.29.20047167 doi: medRxiv preprint 1A-C). Of note, live impacts were primarily during games (N=2,512, 92.08%) as opposed to 191 during practices (N=216, 7.92%). 192 All 3 position groups exhibited similar incremental patterns in head impact kinematics 193 (air < bags < control < thud < live). However, the linemen and hybrid groups had consistently 194 higher exposure in both frequency and magnitudes than the skill group across all LOCs ( Figure   195 2A-C). Head impact kinematics for each LOC are detailed in Supplemental Table 1. Subconcussion research is still at its infancy, but it is a rapidly growing area of concern in 210 sport injury prevention. To contribute to this emerging field, this study examined whether 211 subconcussive head impact kinematics differed across LOCs and between player position groups. 212 There were four key findings in this study. First, cumulative head impact exposure increased as 213 All rights reserved. No reuse allowed without permission. author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint (which was not peer-reviewed) is the . https://doi.org/10.1101/2020.03.29.20047167 doi: medRxiv preprint the LOC increased, with frequent head impacts observed during live, thud, and control drills. 214 Second, in line with literature (11), linemen and hybrid groups had consistently higher head 215 impact exposure compared to the skill position group across all LOCs. Third, the mean head 216 impact magnitude was similar (18 to 23 g) across air, bags, control, and thud, but head impacts 217 during live showed greater magnitudes (40 to 45 g). Lastly, very high impact magnitudes (>100 218 g) were small in number overall but were more frequent in thud and live than other LOCs. Taken 219 together, our data, for the first time, empirically support the USA football's categorization of 220 LOC while calling for a need to dissect football practice guidelines to make more specific 221 recommendations for practice and games to minimize cumulative burden on adolescents' brain 222 health. 223 Owing to the sensor-installed helmets, mouthguards, headbands, and skin patches, our 224 knowledge of head impact exposure in American football has drastically improved in the past 15 225 years. These technological advancements allowed researchers to evaluate head impact frequency 226 and magnitude in various position groups, practice types (e.g., shell-only, full-gear), play types 227 (e.g., running, passing, special teams), and time-based hit rates (29-34). Previous research has 228 suggested that overall head impact exposure elevates in relation to increased practice duration, 229 contact intensity, and time spent in high risk drills (21, 32). However, these variables differ 230 greatly between players and position groups. For example, previous literature demonstrated that 231 the differences between individual players accounted for 48% of the variance in head impact 232 exposure during practice (35). Additionally, recent research suggests different types of plays 233 (e.g., running, passing, special teams) also have different average head impact magnitudes which 234 will influence cumulative head impact exposure (34). Despite this high degree of variance 235 between players and play types, the linemen and hybrid position groups consistently have higher 236 All rights reserved. No reuse allowed without permission. author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint (which was not peer-reviewed) is the . https://doi.org/10.1101/2020.03.29.20047167 doi: medRxiv preprint head impact exposure compared to skill positions such as receivers, defensive backs, and 237 quarterbacks (30, 31). We were able to corroborate these observations in our study in the context 238 of five different LOC, where frequent head impacts were notable especially during high intensity 239 drills (live, thud, control), with linemen and hybrid players experiencing more head impacts than 240 skill players. 241 Another major finding from this study was that live had the highest average PLA and 242 PRA compared to air, bags, control, and thud, which shared similar impact magnitudes. author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint (which was not peer-reviewed) is the . https://doi.org/10.1101/2020.03.29.20047167 doi: medRxiv preprint substantiate the importance of regulating the duration of specific drill types, rather than 260 restricting practice type and frequency. 261 There are several limitations to this study. Our examination of head impacts in high 262 school football is limited in that it was conducted on a single high school football team in the 263 Midwest composed of primarily white males; therefore, the results are not generalizable to the 264 broader U.S. population of high school football teams. A second limitation of the study is that  author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint (which was not peer-reviewed) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.