Measurement of the head impacts in a sub-elite Australian Rules football team with an instrumented patch: An exploratory analysis

Abstract

An investigation was made of the frequency, magnitude, and distribution of head impacts in Australian Football League players over a season of matches. In a prospective cohort analysis of impact magnitude, frequency, and distribution on data collected with a wireless head impact sensor worn behind the ear of 23 players, a total of 4903 impacts were recorded. Players experienced on average 407 ± 143 impacts over the duration of the study resulting in 30 ± 38 impacts per-player per-match. Linear accelerations ranged from 10 g to 153 g with a mean, median, and 95th percentile value of 17 g, 13 g, and 40 g, respectively. Rotational accelerations ranged from 130 rad/s² to 21,890 rad/s² with a mean, median, and 95th percentile value of 2426 rad/s², 1556 rad/s², and 7571 rad/s², respectively. This study obtained initial measurements on the frequency, magnitude, distribution, and risk weighted exposure of head impacts in Australia Rules Football in order to better inform medical personnel in the identification and evaluation of at-risk players for concussion. The location of impacts varied considerably with the back of the head recording more total impacts than the front, side, and top. Midfielders sustained more impacts per-player, per-match, and had higher median resultant linear accelerations than forwards and defenders. The results of this study, in which most impacts were within the low severity limit for linear, rotational, HIT_SP, and RWE_CP, indicate that Australian Rules football needs to include more encompassing methods of examination of player exposure.

Keywords

Biomechanics concussion head impact biomechanics instrumented patch linear acceleration rotational acceleration

Introduction

As the most popular football code in Australia,¹ Australian Rules Football (ARF) is a full-contact, invasion game² that generally involves less collisions and tackling than rugby union or rugby league. Similar to football (soccer) in the objectives of the game, ARF is a dynamic running game combining athleticism with speed, necessitating skillful foot and hand passing.^1,3 Unlike soccer, ARF involves aggressive tackling with sudden collisions between players and the ground.^2,4 There are 18 players on field, per team, during match activities, with up to four players utilized as interchange players.⁵ The interchange players can be rotated onto the field as frequently as the coach requires. Players partake in ARF in one of three positional groups, forwards, midfielders, and defenders.¹ Forwards typically mark the ball (catch the ball on the full from a kick greater than 15 m) in the best position possible then take a free kick at the goals; midfielders typically work alongside forwards and defenders picking up loose balls from the ground or regaining possession and clearing the ball from their goal area; and defenders typically attempt to “spoil” the opposition forwards marking the ball.¹

Playing ARF involves activities such as running, kicking, frequent jumping/landing actions, and sudden changes in direction.^1,6 As a result of the fast paced nature of ARF, there is a high risk of injury.⁷ The most common match injuries that occur to ARF players are muscle strains,^6,8–11 joint or ligament sprains,^6,8–11 contusions or haematomas,^6,8,9,11 and concussions.¹¹ Ranked as the fourth most common injury (4.3%) in community (under 18) ARF players,^2,11 the incidence of concussion is reported to have increased.¹² The incidence of concussion at the professional level of participation, the Australian Football League (AFL), is reported to be 5 to 6 per 1000 player hours making this one of the most common injuries that occur during match participation.¹¹ As a result of the increase in sports-related concussions occurring, these injuries have become an increasingly serious concern.^13–15

As a subset of mild traumatic brain injuries¹⁶ information relating to the cause, course, and sequelae of concussion has increased. As a result of this information, the knowledge surrounding the recognition and management of concussion has also improved.¹⁷ This is important in sports such as ARF where collisions can result in concussion.¹⁷ Furthermore, there is a concern for the immediate, and long-term, effects of players involved in sports such as ARF and rugby that are subjected to repeated impacts to the head resulting in sub-concussive impacts occurring,^18,19 and how these may adversely affect cerebral functions.^18–20 As a result research into head impacts,²¹ and prevention strategies such as body-contact/tackling skills¹⁷ have increased over the years, leading to greater insight into the likely causes and the longitudinal effects of these injuries.

Research has sought to better define the resultant linear and rotational accelerations that are involved in concussion injuries to the head through the use of telemetry or impact sensors.^21–23 These impact sensors (gyroscopes/accelerometers) have been utilized in helmeted, and non-helmeted, sports to understand the link between the biomechanics of the head impact and the clinical outcomes of concussion in athletes.^21–28 Of the non-helmeted sports, soccer²⁹ and rugby union²¹ have started to accumulate impact sensor data. Using this emerging technology to gain real-time data^26,30–33 on sports collisions, along with the mandatory use of helmets in American football, has afforded the systematic analysis of injury biomechanics that occur in match and training activities. This new technology has led to a more extensive knowledge of the magnitudes, distribution, and frequency of accelerations involved in head injuries, that can be applied to football or any other circumstances where repetitive head injury can occur.

However, this technology has not been incorporated in ARF concussion research and management. The aim of this exploratory study was to investigate the head impact acceleration characteristics with the use of wireless head impact sensors during 12 matches in adult sub-elite level ARF players. This is the first study to embark on this emerging area of research within ARF. The generation of new knowledge may result in comparisons being drawn between player positions to identify injury risk and incidence within the sport.

Methods

A prospective observational cohort study was conducted on a single West Australian Football League (WAFL) sub-elite (reserves) team competing in the senior WAFL competition during the 2015 competition. All members of the team were invited to participate in the study. A total of 23 players (21.0 ± 2.4 years, 182.0 ± 8.3 cm, and 77.6 ± 6.5 kg) agreed to participate and were enrolled in the study. These players consisted of seven forwards; 11 midfielders and 5 defenders. The consent was obtained from the players before enrolling in the study. The researchers’ university ethics committee approved all procedures (MUHREC 2015/061). The approval was provided by the Western Australian Football Commission, participating team, and players prior to commencing the study.

Study participants wore the XPatch impact-sensing skin patch (X2Biosystems Ltd, Seattle, WA; www.x2biosystems.com) on the skin covering their mastoid process (right side) during each match. The XPatch sensor sampling at 1024 Hz was placed behind the player’s right ear just before they participated in match activities and was removed immediately after the match was completed. The positioning of the XPatch over the mastoid process is important to ensure that the sensor was not activated by enhanced soft-tissue effects when impacts occur.³⁴

The XPatch contained a low-power, high-g triaxial accelerometer with 200 g maximum per axis and a triaxial angular rate gyroscope to capture six degrees of freedom for linear and rotational time history accelerations of the heads center of gravity for all impacts that occurred during match participation. The time history incorporated three axes (x, y, z) of acceleration and three axes of velocity. Standing in an upright position, these planes describe the medial-lateral, anterior-posterior, and vertical acceleration and deceleration.

If an accelerometer exceeded the predetermined 10 g linear acceleration threshold, 100 milliseconds (ms) of data (10 ms pre-trigger and 90 ms post-trigger) from each accelerometer and gyroscope were recorded to the on-board memory for later downloading. This data acquisition limit was based on a review of data acquisition limits utilized in previously published studies.²¹

Following the match, the XPatch were removed from the player and the data downloaded to the Injury Management Software (IMS) (X2Biosystems). The IMS enabled the raw accelerometer data to be transformed to the head center of gravity by using a rigid-body transformation for linear acceleration and a 5-point stencil for rotational acceleration.^21,34 The biomechanical measures of head impact severity consisted of impact duration (ms), linear acceleration (g), and rotational head acceleration (rad/s²). Resultant linear acceleration is the rate of change in velocity of the estimated center of gravity of the head attributable to an impact and the associated direction of motion of the head.³⁵ Resultant rotational acceleration is the rate of change in rotational velocity of the head attributable to an impact, and its direction in a coordinate system with the origin at the estimated center of gravity of the head.³⁵ False impacts were removed by the X2Biosystems proprietary “de-clacking” algorithm²¹ by comparing the waveform of each impact to a “Gaussian-like” reference waveform using cross-correlation.²¹ Impacts with a resultant linear acceleration of <10 g were removed. The remaining impacts were downloaded to an Excel spreadsheet and time-filtered to include only those impacts that occurred during match participation.

Head impact exposure including frequency, magnitude, and location of impacts were quantified using previously established methods.^26,36 Three measures of impact frequency were computed for each player: player impacts, the total, median 25th–75th interquartile range [IQR], and the 95th percentile of head impacts recorded for a player during all the matches observed, player group impacts, the total, median [IQR], and the 95th percentile of impacts recorded for each of the player groups (forwards, midfielders, and backs) during all matches observed, and impacts per match, the total, median [IQR], and the 95th percentile of head impacts recorded for a player during all the matches observed.

All filtered data on the Microsoft Excel spreadsheet was analyzed with SPSS V.22.0.0. The impact variables were not normally distributed (Kolmogorov-Smirnov; p < 0.001). Therefore, data were expressed as median [IQR], and as severity measures (95th percentile linear acceleration, 95th percentile rotational acceleration).^37,38 The impact location variables were computed as azimuth and elevation angles relative to the center of gravity (CG) of the head centered on the mid-sagittal plane.³⁹ These were categorized as front (Left: θ = 180° to −135°; Right: θ = 180° to 135°), side (Left: θ = –135° to −45°; Right: θ = 135° to 45°), back (Left: θ = −45° to 0°: Right: θ = 45° to 0°), and top (Left: θ = 180° through negative θ to 0°; Right: θ = 180° through positive θ to 0°). Impacts to the top of the head were defined as all impacts above an α of 65° from a horizontal plane through the CG of the head.⁴⁰ Impact locations were analyzed by front, back, side, and top impacts using a Friedman repeated measures ANOVA on ranks.

Head impacts were assessed by injury tolerance level for a concussion using previously published injury tolerance levels^31,32,41 for linear (>95 g) and rotational acceleration (>5500 rad/s²). Head impacts were assessed for impact severity using previously published levels for linear acceleration (mild < 66 g, moderate 66–106 g, severe > 106 g) and rotational acceleration (mild < 4600 rad/s², moderate 4600–7900 rad/s², severe > 7900 rad/s²).^42–44

Two additional risk equations were included in the analysis of the impact data to identify players at risk of a concussion. The Head Impact Telemetry Severity Profile (HIT_SP)⁴⁰ is weighted composite score including linear and rotational accelerations, impact duration, as well as impact location. The Risk Weighted Exposure Combined Probability (RWE_CP)⁴⁵ is a logistic regression equation and regression coefficient of injury risk prediction of an injury occurring based on previously published analytical risk functions. The RWE_CP combines the resultant linear and rotational accelerations to elucidate individual player and team-based exposure to head impacts. As a value of 63 is a 75% indicator for a concussive injury,^40,46 the HIT_SP values were evaluated by limits of less than 25% risk (<21), 25% to 75% risk (21–63), and >75% risk (>63). The RWE_CP values were also evaluated by the same values of 25% risk (<0.2500) 25% to 75% risk (0.2500–0.7500) and >75% risk (>0.7500).

The impact data were further assessed by grouping the resultant impacts by player age, height, and body mass. As there were only 23 players enrolled in the study, this was undertaken by dividing the players into three groups based on player age, height, and body mass distribution. These were termed lower (age: <20 years; height < 1.76 m; body mass: < 74 kg), middle (age: 20–21 years; height: 1.76–1.87 m; body mass: 74–78 kg), and higher (age: > 21 years; height > 1.87 m; body mass: > 78 kg). The resultant linear and rotational accelerations, HIT_SP and RWE_CP were analyzed using a Friedman repeated measures ANOVA on ranks. Post hoc analysis with Wilcoxon-signed-rank tests was conducted with a Bonferroni correction applied.

The HIT_SP and RWE_CP were analyzed by total, forwards, backs, and midfielders recorded head impacts using a Friedman repeated measures ANOVA on ranks. Post hoc analysis with Wilcoxon-signed-rank tests was conducted with a Bonferroni correction applied. A one-sample chi-squared (χ²) test and risk ratio (RR) with 95% confidence intervals (CI) were used to determine head impact risk. Statistical significance was set at p < 0.05.

Results

Throughout the course of the project, 12 matches and 23 players were monitored. For analysis, players were separated into three positional groups: forwards (n = 7), midfielders (n = 11), and defenders (n = 5). Over the duration of the study, 4903 impacts were recorded. Players experienced on average 213 ± 315 impacts per-player over the 12 matches resulting in 29 ± 37 impacts per-player per-match. Linear accelerations ranged from 10 g to 153 g with a mean, median, and 95th percentile value of 17 g, 13 g, and 40 g, respectively. Rotational accelerations ranged from 130 rad/s² to 21,890 rad/s² with a mean, median, and 95th percentile value of 2426 rad/s², 1556 rad/s², and 7571 rad/s², respectively (see Table 1).

Table 1.

Impacts to the head greater than 10 g for total impacts recorded, impacts by forwards midfielders and defenders in an amateur Australian Football League team over a season of matches.

Total impacts	Per-player per-match season	Per-player per-match	Impact duration (ms)	PLA (g)			PRA(rad/s²)			HIT_SP			RWE_CP
n =	Mean ± SD	Mean ± SD	Mean ± SD	Mean ± SD	Median [IQR]	95%	Mean ± SD	Median [IQR]	95%	Mean ± SD	Median [IQR]	95%	Mean ± SD	Median [IQR]	95%
Total
4903	213 ± 315	29 ± 37	7.1 ± 6.3	17 ± 12	13 [11–18]	40	2426 ± 2480	1556 [1020–2794]	7571	19 ± 15	16 [14–19]	39	0.0298 ± 0.1404	0.0003 [0.0001–0.0009]	0.9315
Forwards
764	109 ± 105^bc	14 ± 8^bc	6.5 ± 5.8	17 ± 13	10 [7–17]	37	2505 ± 2427	1764 [1092–2898]	7292	19 ± 16	16 [14–19]	35	0.0287 ± 0.1392	0.0003 [0.0002–0.0010]	0.0681
Midfielder
3131	261 ± 417^a	36 ± 49^a	7.0 ± 6.3	17 ± 11	13 [11–18]	39	2351 ± 2407	1520 [1018–2655]	7155	19 ± 15	16 [14–19]	39	0.0270 ± 0.1338	0.0002 [0.0002–0.0008]	0.0721
Defender
1008	202 ± 180^a	26 ± 18^a	7.7 ± 6.7	18 ± 13	13 [11–18]	44	2623 ± 2738	1553 [978–3105]	13,327	20 ± 15	16 [14–20]	44	0.0403 ± 0.1621	0.0003 [0.0001–0.0012]	0.3024

Data are presented as mean and standard deviation (±SD), median [interquartile range] and 95th percentile for total impacts, impacts per player position group, impact duration (ms), resultant linear and rotational acceleration, Head Impact Telemetry Severity Profile and Risk Weighted Exposure Combined Probability.

IQR: interquartile (25th to 75th) percentile; 95%: 95th percentile; PLA (g): peak linear acceleration in gravitational force (g); PRA (rad/s²): peak rotational acceleration in radians/second²; HIT_SP: Head Impact Telemetry Severity Profile; RWE_CP: risk Weighted Exposure Combined Probability; significant difference (p < 0.05) than (a) forward; (b) midfielder; (c) defender.

The results for forwards, midfielders, and defenders (see Table 1) revealed midfielders experience significantly more impacts per player, per match (36 ± 49) than forwards (RR: 4.1 [95% CI: 3.8 to 4.4]; p < 0.0001) and defenders (RR: 3.1 [95% CI: 2.9 to 3.3]; p < 0.0001). Forwards recorded higher resultant median rotational accelerations (1764 [1092 to 2898] rad/s²) than defenders (p = 0.8742) and midfielders (p = 0.2235).

The number of impacts to the head varied over the season (see Table 2). The back of the head recorded more total impacts (n = 2466) than the front (RR: 3.7 [95% CI: 3.4 to 4.0]; p < 0.0001), side (RR: 1.5 [95% CI: 1.4 to 1.6] p < 0.0001), and top (RR: 26.8 [95% CI: 21.9 to 33.0]; p < 0.0001) of the head. The top of the head recorded the highest median resultant linear acceleration (18 [13 to 26] g) when compared with impacts recorded to the front (p = 0.8993), side (p < 0.0001), and back (p < 0.0001) of the head. Forwards recorded a higher median resultant linear acceleration (22 [18 to 48] g) to the top of the head than midfielders (p = 0.0280) and defenders (p = 0.1282). The front of the head recorded the highest resultant median rotational accelerations (3186 [2072 to 5408] rad/s²) when compared with the back (p < 0.0001), side (p < 0.0001), and top (p = 0.0987) of the head. Forwards recorded the highest resultant median rotational accelerations to the top of the head (7560 [2531 to 10556] rad/s²) when compared with midfielders (p = 0.0180) and defenders (p = 0.0280). The front of the head recorded the highest median HIT_SP (19 [16 to 27]) and RWE_CP (0.0013 [0.0004 to 0.0104]) over the duration of the study. Forwards recorded the highest median HIT_SP (26 [18–49]) and RWE_CP (0.5400 [0.0007–0.6458]) to the top of the head when compared with midfielders (p = 0.0180; p = 0.0277) and defenders (p = 0.0425; p = 0.0425) over the duration of the study.

Table 2.

Impacts to the head greater than 10 g by impact location for total impacts, impacts recorded by forwards midfielders and defenders in an amateur Australian Football League team over a season of matches.

Impacts				PLA (g)			PRA(rad/s²)			HITsp			RWE_(CP)
Location	No.	%	Duration (ms)	Mean ± SD	Median [IQR]	95%	Mean ± SD	Median [IQR]	95%	Mean ± SD	Median [IQR]	95%	Mean ± SD	Median [IQR]	95%
Total
Front	672	13.7	10.5 ± 8.2	23 ± 17^ef	17 [13–27]	57	4257 ± 3322^ef	3186 [2072–5408]	11,078	26 ± 23^ef	19 [16–27]	61	0.0853 ± 0.2285^ef	0.0013 [0.0004–0.0104]	0.761
Back	2466	50.3	6.3 ± 5.7	16 ± 10^dfg	12 [11–16]	34	1946 ± 2109^dfg	1299 [896–2069]	5564	18 ± 13^dfg	15 [14–17]	34	0.0195 ± 0.1160^dfg	0.0002 [0.0001–0.0004]	0.0163
Side	1673	34.1	6.8 ± 5.7	17 ± 11^deg	13 [11–17]	36	2355 ± 2222^deg	1633 [1059–2767]	6720	19 ± 12^deg	16 [14–19]	34	0.0220 ± 0.1189^deg	0.0003 [0.0002–0.0008]	0.486
Top	92	1.9	9.3 ± 7.2	22 ± 13^ef	18 [13–26]	53	3262 ± 2697^ef	2216 [1359–4757]	8599	23 ± 12^ef	19 [15–25]	46	0.0439 ± 0.1613^ef	0.0007 [0.0002–0.0061]	0.4181
Forwards
Front	119	15.6	9.9 ± 7.6	23 ± 18^befg	17 [13–27]	61	4100 ± 3085^ef	3252 [2179–4435]	10,389	25 ± 21^ef	19 [16–27]	60	0.0741 ± 0.2091^ef	0.0012 [0.0004–0.0058]	0.6602
Back	334	43.7	5.2 ± 4.8	15 ± 11^bdfg	12 [11–15]	27	1962 ± 2,13^dfg	1334 [896–2144]	6086	18 ± 14^bdfg	15 [14–17]	26	0.0198 ± 0.1215^dfg	0.0002 [0.0001–0.0004]	0.0202
Side	304	39.8	6.5 ± 5.4	16 ± 11^de	13 [11–17]	31	2370 ± 1979^deg	1800 [1162–2850]	6066	19 ± 15^deg	16 [15–19]	30	0.0152 ± 0.1006^deg	0.0003 [0.0002–0.0009]	0.0206
Top	7	0.9	11.4 ± 8.6	32 ± 21^bde	22 [18–48]	*	7189 ± 4342^bcef	7560 [2531–10,556]	*	35 ± 22^bcef	26 [18–49]	*	0.2635 ± 0.3927^bcef	0.5400 [0.0007–0.6458]	*
Midfielders
Front	393	12.6	11.0 ± 8.6	23 ± 16^aef	17 [13–27]	58	4353 ± 3418^ef	3335 [2069–5505]	11,652	27 ± 25^ef	20 [16–27]	68	0.0893 ± 0.2389^ef	0.0015 [0.0004–0.0123]	0.8422
Back	1656	52.9	6.4 ± 5.8	16 ± 9^ad	13 [11–16]	34	1889 ± 1938^dfg	1315 [926–2073]	5124	18 ± 12^adfg	15 [14–18]	34	0.0151 ± 0.0992^dfg	0.0002 [0.0001–0.0004]	0.0121
Side	1013	32.4	6.5 ± 5.3	17 ± 11^d	13 [11–17]	36	2296 ± 2228^de	1556 [1043–2629]	6867	19 ± 13^deg	15 [14–18]	34	0.0223 ± 0.1197^de	0.0003 [0.0002–0.0462]	0.0462
Top	69	2.2	9.2 ± 7.3	21 ± 12^a	16 [12–23]	54	2845 ± 2322^ace	1958 [1148–4219]	8009	22 ± 11^acef	18 [15–24]	44	0.0296 ± 0.1253^ace	0.0004 [0.0002–0.0044]	0.2408
Defenders
Front	160	15.9	9.6 ± 7.7	22 ± 16^ef	17 [12–27]	56	4120 ± 3231^ef	2947 [1954–5267]	10,802	25 ± 20^ef	19 [16–27]	59	0.0820 ± 0.2141^ef	0.0009 [0.0004–0.0096]	0.7104
Back	476	47.2	6.6 ± 5.9	16 ± 12^df	12 [11–15]	40	2154 ± 2621^dfg	1214 [820–2230]	8378	18 ± 13^df	15 [14–17]	41	0.0353 ± 0.1585^df	0.0002 [0.0001–0.0005]	0.1818
Side	356	35.3	8.2 ± 6.9	18 ± 13^de	13 [11–19]	45	2545 ± 2428^de	1714 [1053–3138]	7711	20 ± 14^de	16 [14–20]	39	0.0297 ± 0.1394^de	0.0003 [0.0002–0.0013]	0.088
Top	16	1.6	8.7 ± 6.2	21 ± 9	20 [14–30]	*	3346 ± 2031^abe	2554 [1668–5385]	*	22 ± 7^ab	19 [16–26]	*	0.0091 ± 0.0171^ab	0.0008 [0.0004–0.0146]	*

Data are presented as mean (±SD), median [IQR] and 95th percentile for total impacts, impacts per player position group, impact duration (ms), resultant linear and rotational acceleration, Head Impact Telemetry Severity Profile and Risk Weighted Exposure Combined Probability.

IQR: interquartile (25th to 75th) percentile; 95%: 95th percentile; PLA (g): peak linear acceleration in gravitational force (g); PRA (rad/s²): peak rotational acceleration in radians/second²; HIT_SP: Head Impact Telemetry Severity Profile; RWE_CP: Risk Weighted Exposure Combined Probability; significant difference (p < 0.05) than (a) forward; (b) midfielder; (c) defender; (d) front; (e) back; (f) side; (g) top.

not able to be calculated.

There were 14 (0.3%) impacts recorded above the linear injury risk limit and 436 (8.9%) impacts above the rotational injury risk limit (see Table 3). Midfielders recorded more impacts in the injury risk limit for rotational accelerations (>5500 rad/s²) than forwards (p = 0.7111) and defenders (p = 0.0278) over the duration of the study. Defenders recorded more impacts (6.7%) above the severe rotational acceleration (>7900 rad/s²) threshold than midfielders (p = 0.0009) and forwards (p = 0.0461). The majority of the impacts were in the low impact severity limit for linear (98.8%), rotational (88.3%), HIT_SP (80.7%), and RWE_CP (96.2%) data acquisition limits.

Table 3.

Impacts to the head greater than 10g by injury tolerance, injury severity for resultant linear and rotational accelerations, Head Impact Telemetry Severity Profile and Risk Weighted Exposure Combined Probability for total impacts, impacts recorded by forwards midfielders and defenders in an amateur Australian Football League team over a season of matches.

		Total	Forwards	Midfielders	Defenders
		n (%)	n (%)	n (%)	n (%)
Injury tolerance
Linear	>95 g	14 (0.3)	4 (0.5)	7 (0.2)	3 (0.3)
Rotational	>5500 rad/s²	436 (8.9)	67 (8.8)	261 (8.3)^c	108 (10.7)^b
Injury severity (linear)
Mild	<66 g	4282 (98.8)	753 (98.7)	3098 (98.9)	991 (98.3)
Moderate	66–106 g	52 (1.1)	10 (1.3)	26 (0.8)^c	16 (1.6)^b
Severe	>106 g	9 (0.2)	2 (0.3)	6 (0.2)	1 (0.1)
Injury severity (rotational)
Mild	<4600 rad/s²	4238 (88.3)	673 (88.1)	2795 (89.3)	860 (85.3)
Moderate	4600–7900 rad/s²	344 (7.0)	57 (7.5)	207 (6.6)	80 (7.9)
Severe	>7900 rad/s²	231 (4.7)	34 (4.5)^c	129 (4.1)^c	68 (6.7)^ab
HIT_SP
Mild	<21	3958 (80.7)	624 (81.7)	2543 (81.2)	791 (78.5)
Moderate	21–63	863 (17.6)	128 (16.8)	536 (17.1)	199 (19.7)
Severe	>63	82 (1.7)	12 (1.6)	52 (1.7)	18 (1.8)
RWE_CP
Mild	<0.2500	4719 (96.2)	736 (96.3)	3029 (96.7)	954 (94.6)
Moderate	0.2500–0.7500	96 (2.0)	14 (1.8)	51 (1.6)^c	31 (3.1)^b
Severe	>0.7500	88 (1.8)	14 (1.8)	51 (1.6)	23 (2.3)

Data are presented as number of impacts and percentage of impacts recorded.

rad/s²: radians/second²; HIT_SP: Head Impact Telemetry Severity Profile; RWE_CP: risk Weighted Exposure Combined Probability; significant difference (p < 0.05) than (a) forward; (b) midfielder; (c) defender.

There were statistically significant differences observed for the resultant linear (χ²₍₂₎ = 10.39; p = 0.0055), rotational (χ²₍₂₎ = 21.44; p < 0.0001), HIT_SP (χ²₍₂₎ = 24.64; p < 0.0001), and RWE_CP (χ²₍₂₎ = 25.04; p < 0.0001) when comparing impacts for the lower, middle, and higher age groups (see Table 4). There was a statistically significant difference observed on post hoc analysis that the lower age group recorded a higher median result when compared with the middle age group for resultant linear (14 g vs. 13 g; z = −2.82; p = 0.0048) and rotational (1863 rad/s² vs. 1519; z = −3.67; p = 0.0002) accelerations; RWE_CP (0.0003 vs. 0.0002; z = −3.89; p = 0.0001) and the middle and higher age group for HIT_SP (19 vs. 20; z = −2.98; p = 0.0029).

Table 4.

Impacts to the head greater than 10 g by age, height and weight groups in an amateur Australian Football League team over a season of matches.

			Impacts	PLA (g)			PRA (rad/s²)			HITsp			RWEcp
Range	No.	Age	per-player per-season	Mean SD	Median [IQR]	95%	Mean SD	Median [IQR]	95%	Mean SD	Median [IQR]	95%	Mean SD	Median [IQR]	95%
Age¹²³⁴
L: <20 years	7	18.7 ± 0.8	123 ± 79	17 ± 11^b	14 [11–18]	37	2579 ± 2279^bc	1863 [1180–3071]	7705	19 ± 13^c	16 [14–19]	35	0.0250 ± 0.1230^bc	0.0003 [0.0002–0.0011]	0.0609
M: 20–21 years	7	20.4 ± 0.5	190 ± 170	17 ± 11^a	13 [11–18]	39	2365 ± 2397^ac	1519 [1048–2642]	7157	19 ± 14	16 [14–19]	39	0.0266 ± 0.1318^ac	0.0002 [0.0002–0.0008]	0.0681
H: >21 years	9	24.0 ± 1.5	299 ± 480	17 ± 14	13 [11–17]	43	2455 ± 2758^ab	1457 [861–2894]	8901	20 ± 17^a	15 [14–19]	42	0.0394 ± 0.1654^ab	0.0002 [0.0001–0.0009]	0.2405
Height¹²³⁴
L: <1.76 m	7	172.0 ± 6.3	346 ± 534	16 ± 11^bc	13 [11–16]	34	2229 ± 2290^c	1431 [993–2497]	6770	18 ± 14^bc	15 [14–18]	34	0.0231 ± 0.1229^c	0.0002 [0.0001–0.0006]	0.0437
M: 1.76–1.87 m	8	183.0 ± 2.6	137 ± 159	18 ± 13^ac	13 [11–19]	44	2567 ± 2757^c	1588 [903–3020]	8877	20 ± 16^ac	16 [14–20]	45	0.0400 ± 0.1655^c	0.0003 [0.0001–0.0012]	0.2326
H: >1.87 m	8	189.7 ± 2.1	153 ± 124	19 ± 13^ab	14 [11–20]	45	2676 ± 2521^ab	1873 [1209–3073]	7808	21 ± 15^ab	16 [15–21]	46	0.0329 ± 0.1447^ab	0.0004 [0.0002–0.0013]	0.16
Weight¹²³⁴
L: <74 kg	9	71.9 ± 1.3	264 ± 488	16 ± 11^bc	13 [11–16]	36	2243 ± 2372^b	1427 [962–2509]	6957	19 ± 14^bc	15 [14–18]	38	0.0266 ± 0.1334^bc	0.0002 [0.0001–0.0006]	0.0639
M: 74–78 kg	6	76.5 ± 1.4	205 ± 73	18 ± 13^a	14 [11–20]	41	2569 ± 2555^ac	1765 [1114–2993	7256	21 ± 18^a	16 [14–21]	41	0.0307 ± 0.1470^a	0.0003 [0.0002–0.0012]	0.0756
H: >78 kg	8	85.3 ± 4.5	160 ± 168	18 ± 12^a	13 [11–18]	44	2633 ± 2584^b	1679 [1059–3092]	8271	20 ± 12^a	16 [14–20]	44	0.0351 ± 0.1465^a	0.0003 [0.0002–0.0012]	0.2127

Data are presented as mean (±SD), median [IQR] and 95th percentile for total impacts per player per-season, resultant linear and rotational acceleration, Head Impact Telemetry Severity Profile and Risk Weighted Exposure Combined Probability.

IQR: Interquartile (25th to 75th) percentile; 95%: 95th percentile; PLA (g): peak linear acceleration in gravitational force (g); PRA (rad/s²): peak rotational acceleration in radians/second²; HIT_SP: Head Impact Telemetry Severity Profile; RWE_CP: Risk Weighted Exposure Combined Probability; L: lower range group; M: middle range group; H: higher range group; significant difference (p < 0.05) than (a) lower range group; (b) middle range group; (c) higher range group; significant difference (p < 0.05) for Friedman repeated measures ANOVA on ranks by (1) PLA (g); (2) PRA (rad/s²); (3) HIT_SP; (4) RWE_CP.

Discussion

This study reports, for the first time, the head impact biomechanics experienced whilst participating in ARF league matches. From this data, the magnitude, distribution, frequency, and severity of impacts to the head experienced by participants during ARF match play can be characterized. The primary aim of this study was to quantify the frequency, magnitude, and location of head impacts over the course of a football season and then compare the results between player position and to similar codes such as in junior, high school, and college American football players and senior amateur rugby union rugby and American football.

This is the first study to record and report the magnitude, frequency, and distribution of impacts to the head for players participating in ARF. During the study, there was an average of 213 ± 315 impacts per-player for the season recorded for ARF players resulting in 29 ± 37 impacts per-player per-match. This is less than the impacts recorded in senior amateur rugby union²¹ (564 ± 618 per-player for the season; 95 ± 133 impacts per-player, per-match) and collegiate level American football (950³⁰ impacts per-athlete, per season to 1353⁴⁷ impacts per-athlete (practice and games combined) exposure). Documenting impacts to the head is important because of the cumulative effects on the brain.⁴⁸ However, measuring linear and rotational accelerations provides even greater understanding of the potential damage of head impacts.^31,32,41,49 Additionally, and unique to this study, was the inclusion of a risk-weighted cumulative exposure (RWE) measure.⁴⁵ By adjusting the impacts’ contribution to cumulative exposure according to its associated impact tolerance, the RWE for linear, rotational, and a combined (linear and rotational) probability measure can be established.⁴⁵ In order to accurately predict the risk of concussion, both linear and rotational accelerations should be accounted for, to determine the concussion risk.⁴⁵ The RWE combined probability (RWE_CP) enabled this concussion risk prediction to be undertaken. By recording the RWE_CP of individual players, player groups, and for the sport,⁵⁰ the resulting values may assist with the identification of players with a potential cumulative exposure resulting in concussion. Only one other study⁴⁵ has reported the RWE_CP and this was in High School level American Football. Utilizing the same logistic regression equations and regression coefficients, it can be seen that the median RWE_CP in High School level American football players was 0.497⁴⁵ and this was higher than the current study (0.0003) recorded on ARF players.

In the current study, the mean, median, and 95th percentile value of the resultant linear (17 ± 12 g, 13 g, and 40 g) and rotational accelerations (2426 ± 2480 rad/s², 1556 rad/s², and 7571 rad/s²) were established. When compared with collegiate American football participation,³⁶ the 95th percentile resultant linear (62 g) acceleration was higher than the ARF players but the resultant rotational (4378 rad/s²) acceleration were lower than ARF players. When compared with senior amateur rugby union,²¹ the mean resultant linear (22 g) and rotational (3903 rad/s²) accelerations were higher than those reported for the ARF players. Further research into the biomechanics of ARF match and training activities at all levels is warranted.

The location of impacts recorded in the current study varied considerably with the back of the head (50%) recording more total impacts than the front (14%), side (34%), and top (2%) of the head. However, the top of the head recorded the highest median resultant linear acceleration (18 g) over the duration of the study. A possible reason for the top of the head having the highest resultant linear acceleration may be related to the type of match activities undertaken in ARF. As seen in Table 2, forwards and defenders recorded median linear accelerations of 22 g and 20 g, respectively, to the top of the head while midfielders recorded higher median linear accelerations (17 g) to the front of the head. The roles of the forwards are to mark the ball and this is done by jumping or jostling against defenders and midfielders in an attempt to get the ball. In doing this, defenders will be attempting to interfere with the forwards as they try to mark the ball and this can result in contact with the top of the head, knocking the player out of the way resulting in the forward and defender recording impacts to the top of the head or may fall onto the group resulting in the higher linear accelerations being recorded. In senior amateur rugby union,²¹ impacts were more commonly recorded to the right (45%) side of the head while collegiate level American football up to 45% of impacts occurred to the front of the head.²⁷ The variations observed between the three comparative codes may stem from the differences in player formations, the protective equipment utilized and the rules of the sporting code in terms of match participation. Further research is warranted to evaluate the differences between the different sporting codes and to identify inter-code differences such as player-position groups.

Examining the results between player-positions identified that midfielders sustained more impacts (261) per-player, per-match, and recorded a higher median resultant linear acceleration (13 g) than forwards and defenders. However, forwards recorded higher resultant median rotational accelerations (1764 rad/s²) and this may be related to the different activities and roles that these players participate in through match participation. Despite these differences, defenders recorded a higher median HIT_SP (16) than midfielders and forwards and these may also be as a result of the differences in positional play during match activities. The variations observed between player-positions warrants further research to identify what match related activities result in these differences being observed. For example, while all players are able to be on any part of the field at any point in a game, they are normally assigned to a certain area for their role.¹ For example, the center, in conjunction with two wingmen, are normally positioned in the midfield area of the oval.¹ The center normally attains the ball and links the defense and attack. In the forward area, a full forward [Forward category] is the main target for the ball when attacking.¹ The full forward must be strong at one-on-one contests to wrestle off opponents.¹ Due to the free-flowing and complex nature of each position, more data are required on each player in order to determine the level of risk associated with the number of impacts to the head.

An interesting finding in analyzing the impact data by age, height, and body mass were the differences observed. Players in the lower age group (<20 years) had fewer average impacts per-player per-season but recorded higher median resultant linear and rotational accelerations but interestingly had a lower median HIT_SP and RWE_CP. When viewed by player height, players above 1.87 m recorded higher median resultant linear and rotational accelerations and as a result had a higher median HIT_SP and RWE_CP. When viewed by body mass players in the middle group recorded a higher median resultant linear and rotational accelerations and as a result had a higher median HIT_SP and RWE_CP. A possible reason for this finding is that players under the age of 20, over 1.87 m tall, and with a body mass of 74–78 kg may be exposed to more severe impacts than other players. This may be as a result of less playing experience, playing in positions, where their height would be an advantage strategically and have the body mass that enables the player to be exposed to positions on the field of play and record more impacts than other player positions. Further research is warranted exploring these aspects on a larger cohort in different sporting codes to identify if there is a similar pattern.

More knowledge on the frequency, magnitude, distribution, and risk weighted exposure of head impacts may assist in the identification of high-risk events and better inform medical personnel of the need to evaluate a player for concussion.^51,52 It is critical to understand the biomechanical basis for mild traumatic brain injuries (concussion injuries), correlating head impact exposure with the clinical variables associated with these injuries, and understanding the acute and long-term effect of repeated sub-concussive impacts.³⁶

Limitations

This study has several limitations which need to be considered when interpreting the results. The use of the X2Patch was novel in that the patch incorporates the accelerometer and gyroscope into a reusable monitor adhered to the side of the head behind the ear directly over the mastoid process of the player. The patches were applied utilizing an adhesive patch to hold the patch to the adhesive and this was applied to the side of the head behind the ear. It was important that the XPatch was applied directly over the mastoid process as the average estimated mass of the skin-patch system is greater than the mass of the sensor itself³⁴ and placement of the XPatch in a lower position may activate soft tissue in the neck during impacts providing false impacts and associated impact biomechanics. Additionally, the effects of sweating by the player, and some extreme weather conditions, resulted in some of the adhesives falling off the player during the match activities. As well, players who were grabbed around the head during the tackle sometimes had the patch pulled off. As a result, the data reported was incomplete and the number of impacts would be more than have been reported.

Although the XPatch has undergone some reported validation studies, and been compared with the Head Impact Telemetry System (HITS), the results have varied. The accelerometers utilized in this study have been reported to have a strong correlation for anterior-posterior translation (r²= 0.93),³⁴ a normalized root square error of 18% for peak linear acceleration (PLA) with an over-prediction of 15 ± 7 g and 2500 ± 1200 rad/s² for peak rotational accelerations (PRA).³⁴ It has also been reported⁵³ that the XPatch had good estimates of PLA but underestimated PRA by more than 25% but recorded more impacts than were visibly seen. Although we found this data, there are no consistent reliability studies for the XPatch and the interpretation of these results should be interpreted with some caution.

Videotape analysis to enable verification of the impacts recorded and was not undertaken as there were not enough video-recorders to enable coverage of the complete field. The size of the field where ARF players participate in match activities was 155 m wide by 185 m long. There was no one position to stand where the whole field was able to be covered with a video camera enabling verification of the impact to be undertaken post-match.

Comparison across different sports codes requires some caution. ARF is more free flowing in nature with faster ball movement on a larger playing field⁵⁴ when compared with American football, rugby union, and rugby league. Sports such as American Football, rugby union, rugby league, and ARF are participated under different contact rules and utilize different participation activities. In addition to this, the number of tackles in ARF is considerably less than other contact invasion sports such as rugby union and rugby league.⁵⁵ Comparing position results to other codes is complex. For example, when an ARF player makes a tackle, they tend to travel at slightly faster velocities compared to the opposition player being tackled. In rugby union,⁵⁶ and rugby league,^57,58 where lines of attackers and defenders face off, the ball carrier’s velocity prior to being tackled is relatively stable. The tackler, with a higher velocity entering the tackle, adjusts their relative velocity before making contact.⁵⁹ In American football, defensive and offensive linemen have the lowest head impact magnitudes of all player positions, but the greatest number of head impacts recorded.^27,32,47 Although comparisons between the different player positions of these different sporting codes may be complex, the value of this comparison is at the individual level, where the contribution of sub-concussive impacts may lead to impairment on clinical measures of neurologic function.²⁰

The current study only reports on 23 players in 12 matches and should not be suggestive of the frequency, magnitude, and location of other levels of participation of ARF. This is seen as a limitation and further studies are recommended on a larger cohort of players. Only 12 matches were recorded during the 2015 domestic competition season. As a result, the findings of this study should not be seen as reflective of the average number of impacts, and their associated impact biomechanics, of a full competition season for ARF players.

Conclusion

For the first time, the head impact biomechanics experienced with participation in Australian Football league matches was measured. Using instrumented mastoid-based impact sensors (accelerometer) over the course of a single season, 4903 impacts were recorded. Players experienced 29 ± 37 impacts per-player per-match with 14 impacts recorded above the linear injury risk limit, and 436 impacts about the rotational injury risk limit signifying that some impacts above the suggested 95 g and 5500 rad/s² are occurring. Although the majority of the impacts recorded were under the linear and rotational injury risk limits, the effects of the frequency and location of these impacts remains unknown. In comparison to senior amateur rugby union and American football players, players in this study were subjected to less overall, and high risk impacts, but the longitudinal effects of this remains unclear. However, inter-code comparisons are complex and the value of this comparison is at the individual level, where the contribution of sub-concussive impacts may lead to impairment on clinical measures of neurologic function. The key to this study was obtaining and thus gaining initial measurements on the frequency, magnitude, distribution, and risk weighted exposure of head impacts in Australia Rule Football in order to assist in the identification at-risk players, which will better inform medical personnel of the need to evaluate a player for concussion.

Footnotes

Acknowledgements

We would like to thank the West Australian Football Commission and, in particular, Troy Kirkham for their support of this research project. We also recognize the work of Myles Murphy, Head Physiotherapist, and student Mason Gittins. According to the definition given by the International Committee of Medical Journal Editors (ICMJE), the authors listed above qualify for authorship on the basis of making one or more of the substantial contributions to the intellectual content of the manuscript.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

References

Gray

Jenkins

. Match analysis and the physiological demands of Australian football. Sports Med 2010; 40: 347–360.

Orchard

Wood

Seward

. Comparison of injuries in elite senior and junior Australian football. J Sci Med Sport 1998; 1: 83–88.

Robertson

Back

Bartlett

. Explaining match outcome in elite Australian Rules football using team performance indicators. J Sports Sci 2016; 34: 637–644.

Hrysomallis

Buttifant

Buckley

. Weight training for Australian football, Melbourne: Lothian Books, 2006.

Orchard

Seward

. Epidemiology of injuries in the Australian football league, seasons 1997 - 2000. Br J Sports Med 2002; 36: 39–45.

Orchard

Seward

Orchard

. Results of 2 decades of injury surveillance and public release of data in the Australian football league. Am J Sports Med 2013; 41: 734–741.

Fortington

Berry

Buttifant

. Shorter time to first injury in first year professional football players: A cross-club comparison in the Australian Football League. J Sci Med Sport 2016; 19: 18–23.

Stevenson

Hamer

Finch

. Sport, age, and sex specific incidence of sports injuries in Western Australia. Br J Sports Med 2000; 34: 188–194.

McManus

Stevenson

Finch

. Incidence and risk factors for injury in non-elite Australian Football. J Sci Med Sport 2004; 7: 384–391.

10.

Orchard J, Seward H and Orchard J. 2014 AFL injury report. http://s.afl.com.au/staticfile/AFL%20Tenant/AFL/Files/2014-AFL-Injury-Report.pdf (2014).

11.

Hrysomallis

. Injury incidence, risk factors and prevention in Australian rugby football. Sports Med 2013; 43: 339–354.

12.

Finch

Clapperton

McCrory

. Increasing incidence of hospitalisation for sport-related concussion in Victoria, Australia. Med J Aust 2013; 198: 427–430.

13.

Covassin

Elbin

Sarmiento

. Educating coaches about concussion in sports: Evaluation of the CDC’s “Heads Up: Concussion in Youth Sports” initiative. J School Health 2012; 82: 233–238.

14.

De Beaumont

Brisson

Lassonde

. Long-term electrophysiological changes in athletes with a history of multiple concussions. Brain Inj 2007; 21: 631–644.

15.

McCrory

. Sports concussion and the risk of chronic neurological impairment. Clin J Sports Med 2011; 21: 6–12.

16.

McCrory

Meeuwisse

Aubry

. Consensus statement on concussion in sport: the 4th International Conference on Concussion in Sport held in Zurich, November 2012. Br J Sports Med 2013; 47: 250–258.

17.

Fortington

Twomey

Finch

. Concussion in community Australian football - Epidemiological monitoring of the causes and immediate impact on play. Inj Epidem 2015; 2: 20.

18.

Baugh

Stamm

Riley

. Chronic traumatic encephalopathy: Neurodegeneration following repetitive concussive and subconcussive brain trauma. Brain Imaging Behav 2012; 6: 244–254.

19.

Gavett

Stern

McKee

. Chronic traumatic encephalopathy: A potential late effect of sport-related concussive and subconcussive head trauma. Clin Sports Med 2011; 30: 179–188.

20.

Gysland

Mihalik

. The relationship between subconcussive impacts and concussion history on clinical measures of neurologic function in collegiate football players. Ann Biomed Eng 2012; 40: 14–22.

21.

King

Hume

Brughelli

. Instrumented mouthguard acceleration analyses for head impacts in amateur rugby union players over a season of matches. Am J Sports Med 2015; 43: 614–624.

22.

Hernandez

Yip

. Six degree-of-freedom measurements of human mild traumatic brain injury. Ann Biomed Eng 2015; 43: 1918–1934.

23.

Bartsch

Samorezov

Benszel

. Validation of an “Intelligent Mouthguard” single event head impact dosimeter. Stapp Car Crash J 2014; 58: 1–27.

24.

Rowson

Brolinson

Goforth

. Linear and angular head acceleration measurements in collegiate football. J Biomed Eng 2009; 131: 1–7.

25.

Brainard

Beckwith

Chu

. Gender differences in head impacts sustained by collegiate ice hockey players. Med Sci Sports Exerc 2012; 44: 297–304.

26.

Crisco

Fiore

Beckwith

. Frequency and location of head impact exposures in individual collegiate football players. J Athl Train 2010; 45: 459–559.

27.

Mihalik

Bell

Marshall

. Measurements of head impacts in collegiate football players: An investigation of positional and event type differences. Neurosurgery 2007; 61: 1229–1235.

28.

Wilcox

Beckwith

Greenwald

. Biomechanics of head impacts associated with diagnosed concussion in female collegiate ice hockey players. J Biomech 2015; 48: 2201–2204.

29.

Hanlon

Bir

. Real-time head acceleration measurements in girls youth soccer. Med Sci Sports Exerc 2012; 44: 1102–1108.

30.

Guskiewicz

Mihalik

. Biomechanics of sport concussion: Quest for the elusive injury threshold. Exerc Sport Sci Rev 2011; 39: 4–11.

31.

Guskiewicz

Mihalik

Shankar

. Measurement of head impacts in collegiate football players: relationship between head impact biomechanics and acute clinical outcome after concussion. Neurosurgery 2007; 61: 1244–1253.

32.

Broglio

Schnebel

Sosnoff

. Biomechanical properties of concussions in high school football. Med Sci Sports Exerc 2010; 42: 2064–2071.

33.

Broglio

Sosnoff

Shin

. Head impacts during high school football: A biomechanical assessment. J Athl Train 2009; 44: 342–349.

34.

Nangia

Bui

. In vivo evaluation of wearable head impact sensors. Ann Biomed Eng 2015; 44: 1234–1245.

35.

Mihalik

Blackburn

Greenwald

. Collision type and player anticipation affect head impact severity among youth ice hockey players. Pediatrics 2010; 125: e1394–e1401.

36.

Crisco

Wilcox

Beckwith

. Head impact exposure in collegiate football players. J Biomech 2011; 44: 2673–2678.

37.

Hopkins

Marshall

Batterham

. Progressive statistics for studies in sports medicine and exercise science. Med Sci Sports Exerc 2009; 41: 3–12.

38.

Hopkins

Marshall

Quarrie

. Risk factors and risk statistics for sports injuries. Clin J Sports Med 2007; 17: 208–210.

39.

Crisco

Chu

Greenwald

. An algorithm for estimating acceleration magnitude and impact location using multiple nonorthogonal single-axis accelerometers. J Biomech Eng 2004; 126: 849–854.

40.

Greenwald

Gwin

Chu

. Head impact severity measures for evaluating mild traumatic brain injury risk exposure. Neurosurgery 2008; 62: 789–798.

41.

Broglio

Eckner

Martini

. Cumulative head impact burden in high school football. J Neurotrauma 2011; 28: 2069–2078.

42.

Harpham

Mihalik

Littleton

. The effect of visual and sensory performance on head impact biomechanics in college football players. Ann Biomed Eng 2014; 42: 1–10.

43.

Ocwieja

Mihalik

Marshall

. The effect of play type and collision closing distance on head impact biomechanics. Ann Biomed Eng 2012; 40: 90–96.

44.

Zhang

Yang

King

. A proposed injury threshold for mild traumatic brain injury. J Biomed Eng 2004; 126: 226–236.

45.

Urban

Davenport

Golman

. Head impact exposure in youth football: High school ages 14 to 18 years and cumulative impact analysis. Ann Biomed Eng 2013; 41: 2474–2487.

46.

Broglio

Eckner

Surma

. Post-concussion cognitive declines and symptomatology are not related to concussion biomechanics in high school football players. J Neurotrauma 2011; 28: 2061–2068.

47.

Schnebel

Gwin

Anderson

. In vivo study of head impacts in football: A comparison of National Collegiate Athletic Association division I versus high school impacts. Neurosurgery 2007; 60: 490–496.

48.

McKee

Stein

Nowinski

. The spectrum of disease in chronic traumatic encephalopathy. Brain 2012; 136: 43–64.

49.

Pellman

Viano

Tucker

. Concussion in professional football: reconstruction of game impacts and injuries. Neurosurgery 2003; 53: 799–814.

50.

King

Hume

Gissane

. The influence of head impact threshold for reporting data in contact and collision sports: Systematic review and original data analysis. Sports Med 2016; 46: 151–169.

51.

Broglio

Eckner

Kutcher

. Field-based measures of head impacts in high school athletes. Curr Opin Pediatr 2012; 24: 702–708.

52.

Greenwald

Chu

Beckwith

. A proposed method to reduce underreporting of brain injury in sports. Clin J Sport Med 2012; 22: 83–85.

53.

Nevins

Smith

Kensrud

. Laboratory evaluation of wireless head impact sensor. Procedia Eng 2015; 112: 175–179.

54.

Gastin

McLean

Spittle

. Quantification of tackling demands in professional Australian football using integrated wearable athlete tracking technology. J Sci Med Sport 2013; 16: 589–593.

55.

Deutsch

Kearney

Rehrer

. Time-motion analysis of professional rugby union players during match-play. J Sports Sci 2007; 25: 461–472.

56.

Quarrie

Hopkins

. Tackle injuries in professional rugby union. Am J Sports Med 2008; 36: 1705–1716.

57.

King

Hume

Clark

. Nature of tackles that result in injury in professional rugby league. Res Sports Med 2012; 20: 87–105.

58.

King

Hume

Clark

. Video analysis of tackles in professional rugby league matches by player position, tackle height and tackle location. Int J Perform Anal Sport 2010; 10: 214–254.

59.

Hendricks

Karpul

Nicolls

. Velocity and acceleration before contact in the tackle during rugby union matches. J Sports Sci 2012; 30: 1215–1224.