Frequency and location of head impact exposures in individual collegiate football players.
ABSTRACT Measuring head impact exposure is a critical step toward understanding the mechanism and prevention of sport-related mild traumatic brain (concussion) injury, as well as the possible effects of repeated subconcussive impacts.
To quantify the frequency and location of head impacts that individual players received in 1 season among 3 collegiate teams, between practice and game sessions, and among player positions.
Collegiate football field.
One hundred eighty-eight players from 3 National Collegiate Athletic Association football teams.
Participants wore football helmets instrumented with an accelerometer-based system during the 2007 fall season.
The number of head impacts greater than 10 g and location of the impacts on the player's helmet were recorded and analyzed for trends and interactions among teams (A, B, or C), session types, and player positions using Kaplan-Meier survival curves.
The total number of impacts players received was nonnormally distributed and varied by team, session type, and player position. The maximum number of head impacts for a single player on each team was 1022 (team A), 1412 (team B), and 1444 (team C). The median number of head impacts on each team was 4.8 (team A), 7.5 (team B), and 6.6 (team C) impacts per practice and 12.1 (team A), 14.6 (team B), and 16.3 (team C) impacts per game. Linemen and linebackers had the largest number of impacts per practice and per game. Offensive linemen had a higher percentage of impacts to the front than to the back of the helmet, whereas quarterbacks had a higher percentage to the back than to the front of the helmet.
The frequency of head impacts and the location on the helmet where the impacts occur are functions of player position and session type. These data provide a basis for quantifying specific head impact exposure for studies related to understanding the biomechanics and clinical aspects of concussion injury, as well as the possible effects of repeated subconcussive impacts in football.
- Cellular and Molecular Bioengineering 12/2014; 7(4):521-531. · 1.23 Impact Factor
- [Show abstract] [Hide abstract]
ABSTRACT: Abstract Objective: Chronic Traumatic Encephalopathy (CTE) is a neurodegenerative disease associated with repetitive brain trauma (RBT). Initially described in boxers, CTE has now been found in other contact sport athletes with a history of RBT. In recent years, there has been tremendous media attention regarding CTE, primarily because of the deaths of high profile American football players who were found to have CTE upon neuropathological examination. However, the study of CTE remains in its infancy. This review focuses on research from the Centre for the Study of Traumatic Encephalopathy (CSTE) at Boston University. This study reviews the formation of the CSTE, major CSTE publications and current ongoing research projects at the CSTE. The neuropathology of CTE has been well-described. Current research focuses on: methods of diagnosing the disease during life (including the development of biomarkers), examination of CTE risk factors (including genetic susceptibility and head impact exposure variables); description of the clinical presentation of CTE; development of research diagnostic criteria for Traumatic Encephalopathy Syndrome; and assessment of mechanism and pathogenesis. Current research at the BU CSTE is aimed at increasing understanding of the long-term consequences of repetitive head impacts and attempting to begin to answer several of the unanswered questions regarding CTE.Brain Injury 01/2015; 29(2):154-63. · 1.51 Impact Factor
- [Show abstract] [Hide abstract]
ABSTRACT: Current American football helmet design has a rigid exterior with a padded interior. Softening the hard external layer of the helmet may reduce the impact potential of the helmet, providing extra head protection and reducing its use as an offensive device. The objective of this study is to measure the impact reduction potential provided by external foam. We obtained a football helmet with built-in accelerometer-based sensors, placed it on a boxing mannequin and struck it with a weighted swinging pendulum helmet to mimic the forces sustained during a helmet-to-helmet strike. We then applied layers of 1.3 cm thick polyolefin foam to the exterior surface of the helmets and repeated the process. All impact severity measures were significantly reduced with the application of the external foam. These results support the hypothesis that adding a soft exterior layer reduces the force of impact which may be applicable to the football field. Redesigning football helmets could reduce the injury potential of the sport.Hawai'i journal of medicine & public health : a journal of Asia Pacific Medicine & Public Health. 08/2014; 73(8):256-61.
Frequency and Location of Head Impact Exposures in
Individual Collegiate Football Players
Joseph J. Crisco, PhD*; Russell Fiore, MEd, ATC?; Jonathan G. Beckwith,
MS`; Jeffrey J. Chu, MS`; Per Gunnar Brolinson, DO‰; Stefan Duma, PhDI;
Thomas W. McAllister, MD"; Ann-Christine Duhaime, MD#;
Richard M. Greenwald, PhD`**
*Department of Orthopaedics, The Warren Alpert Medical School of Brown University and Rhode Island Hospital,
Providence, RI;3Department of Athletics and Physical Education, Brown University, Providence, RI;4Simbex,
Lebanon, NH;1Edward Via Virginia College of Osteopathic Medicine, Blacksburg; ||Center for Injury Biomechanics,
Virginia Tech–Wake Forest, Blacksburg; "Department of Psychiatry and Neurology, Dartmouth Hitchcock Medical
School, Lebanon, NH; #Pediatric Neurosurgery, Dartmouth Hitchcock Medical Center, Hanover, NH; **Thayer School
of Engineering, Dartmouth College, Hanover, NH
Context: Measuring head impact exposure is a critical step
toward understanding the mechanism and prevention of sport-
related mild traumatic brain (concussion) injury, as well as the
possible effects of repeated subconcussive impacts.
Objective: To quantify the frequency and location of head
impacts that individual players received in 1 season among 3
collegiate teams, between practice and game sessions, and
among player positions.
Design: Cohort study.
Setting: Collegiate football field.
Patients or Other Participants: One hundred eighty-eight
players from 3 National Collegiate Athletic Association football
Intervention(s): Participants wore football helmets instru-
mented with an accelerometer-based system during the 2007
Main Outcome Measure(s): The number of head impacts
greater than 10g and location of the impacts on the player’s
helmet were recorded and analyzed for trends and interactions
among teams (A, B, or C), session types, and player positions
using Kaplan-Meier survival curves.
Results: The total number of impacts players received was
nonnormally distributed and varied by team, session type, and
player position. The maximum number of head impacts for a
single player on each team was 1022 (team A), 1412 (team
B), and 1444 (team C). The median number of head impacts
on each team was 4.8 (team A), 7.5 (team B), and 6.6 (team
C) impacts per practice and 12.1 (team A), 14.6 (team B), and
16.3 (team C) impacts per game. Linemen and linebackers
had the largest number of impacts per practice and per game.
Offensive linemen had a higher percentage of impacts to the
front than to the back of the helmet, whereas quarterbacks
had a higher percentage to the back than to the front of the
Conclusions: The frequency of head impacts and the
location on the helmet where the impacts occur are functions
of player position and session type. These data provide a basis
for quantifying specific head impact exposure for studies related
to understanding the biomechanics and clinical aspects of
concussion injury, as well as the possible effects of repeated
subconcussive impacts in football.
Key Words: biomechanics, concussions, accelerometers
Players received up to 1444 head impacts in 1 season, with an average of 6.3 impacts per practice and 14.3 impacts per
Impact frequency and location differed among player positions, with linemen and linebackers having the largest numbers
of impacts per practice and per game.
The offensive linemen had the largest percentage of impacts to the front of the helmet.
Most impacts occurred to the front of the helmet for all player positions except for quarterbacks, who received the most
impacts to the back of the helmet.
particular impact or a series of impacts to the clinical signs
and symptoms of concussion injury, second-impact syn-
drome, or delayed cognitive sequelae have not been
established. Using animal models, several researchers have
suggested that repeated impacts7,8and direction of impact
oncussion injuries, which are most often due to
head impacts, are a growing concern in sports.1–6
However, the biomechanical factors that relate a
or head rotation9–12influence clinical and pathophysio-
logic consequences of injury. To investigate these relation-
ships and the many factors potentially involved in acute
and chronic effects of head impact, numerous researchers
have used sports fields as laboratories. Quantifying events
occurring in the experimental environment of the sport
field is important for understanding the factors relevant to
impacts that result in acute symptoms and for under-
Journal of Athletic Training
gby the National Athletic Trainers’ Association, Inc
Journal of Athletic Training549
standing whether repeated impacts might have subacute or
long-term effects when no immediate symptoms are
In a study of collegiate sports injuries by the National
Collegiate Athletic Association (NCAA),13exposure to the
risk of injury was measured as an athlete-exposure (A-E),
which was defined as ‘‘1 student-athlete participating in 1
practice or competition in which he or she was exposed to
the possibility of athletic injury, regardless of the time
associated with that participation.’’ The measure of A-E
does not account for the magnitude or frequency of head
impacts to individual players. For example, 2 athletes who
participate in the same game, both of whom would have
experienced 1 A-E, might experience a different number of
head impacts with very different magnitudes and at
different locations. Because traumatic brain injuries are
likely to occur along a broad continuum, to be cumulative,
and to involve pathophysiologic events that might occur
without evidence of acute injury symptoms, the concept of
exposure needs to incorporate these elements to best
understand individual player risk and potential prevention.
We propose to define head impact exposure as a broad
term that incorporates multiple variables. Multiple mea-
sures of head impact exposure are critical because the
specific variable or combination of variables that correlates
with the risk of head injury has not been determined. The
first important variable of head impact exposure is A-E,13
which is well suited for understanding the overall risk of
head injury per session of participation. A more complete
understanding of the risk of head injury requires additional
quantifiable variables that we propose include the magni-
tude of the head impact, the number or frequency of head
impacts, the location of the head impact, and cumulative
measures of head impacts.
Obtaining detailed information on magnitude and
frequency of head impacts to individual players has been
challenging. Videotaping of athletic events can provide
some insight into injury mechanisms, but it has limited
practicality because it has limited ability to continuously
track all players, cannot accurately identify all impacts to
the helmet, and cannot provide a direct measure of impact
magnitude. The challenges associated with using video-
tapes to study head impacts in football are well illustrated
in the studies of professional football players by Pellman
et al.14,15Despite the prevalence of head impacts and
the amount of video coverage, only 31 of 182 known
concussive events were available for their analyses because
the impacts were required to be in the open field and to be
recorded from at least 2 unobstructed views. The magni-
tudes of the head impacts, which could not be recorded or
quantified directly from the videos, were computed by
reconstructing the impact scenario in the laboratory using
Hybrid III anthropomorphic test devices (General Motors,
To measure the specific details of head impact exposure
in sport participants, investigators have developed and
implemented a variety of systems.16–18With early proto-
types, players wore obtrusive data-acquisition hardware
that required manual data downloading after each activity;
consequently, these studies were limited in the number of
athletes and session data that were collected. The Head
Impact Telemetry (HIT) System technology19–21(devel-
oped by Simbex, Lebanon, NH, and marketed commer-
cially as the Sideline Response System by Riddell Inc,
Elyria, OH) was designed specifically to address these
limitations by measuring the biomechanical factors asso-
ciated with head impacts for a large number of players
without interfering with the play of the game. The HIT
System is an accelerometer-based system that is mounted
inside a football helmet and is able to directly measure
head acceleration and location of head impacts.
Investigators have used the HIT System to study head
impacts to football players.22–26 They provided new
insights into the biomechanics of head impacts in football
by examining the number of impacts and the magnitude of
the resulting head accelerations across teams and groups of
players. Although the authors of each study reported the
total number of impacts per team that were recorded, they
did not provide detailed analyses of the head impact
exposure for individual players. To increase our under-
standing of the biomechanics of concussion injuries and the
potential cognitive effects related to single or repeated head
impacts, we sought to analyze head impact exposures for
individual players by focusing on 2 specific measures of
exposure. The purpose of our study was to quantify head
impact exposure by recording the frequency and location
of head impacts that individual players received in 1
season. We tested the hypotheses that head impact
frequency and helmet impact locations would differ among
3 collegiate teams, between practice and game sessions, and
among player positions.
Players from 3 NCAA football programs (Brown
University, Dartmouth College, and Virginia Tech) volun-
teered to participate in this observational study. Partici-
pants gave written informed consent, and the study was
approved by the institutional review boards of each
institution. Two teams participated in the Ivy League,
which is an NCAA Football Championship Subdivision
conference that does not allow postseason play, and 1 team
participated in the NCAA Football Bowl Subdivision.
During this study, all 3 teams participated in approximate-
ly the same number of games, but 1 school had almost
twice as many practices as the other 2 schools. During the
2007 fall football season, 188 players from the 3 teams,
which were denoted arbitrarily as team A (n 5 65 players),
team B (n 5 60 players), and team C (n 5 63 players),
participated in our study. Each player was assigned a
unique identification number and categorized in 1 of 8
position units that were defined by the team staff as the
player’s primary position: defensive line (DL, n 5 29),
linebacker (LB, n 5 29), defensive back (DB, n 5 34),
offensive line (OL, n 5 46), offensive back (n 5 23), wide
receiver (WR, n 5 16), quarterback (QB, n 5 8), or special
teams (n 5 3).
All players wore Riddell (Riddell Inc) football helmets
instrumented with the HIT System (Figure 1A and B),
which is a device capable of recording the acceleration-time
history of an impact from 6 linear accelerometers at
1000 Hz. The HIT System continuously samples all 6
550 Volume 45 N Number 6 N December 2010
accelerometers during play. When a preset threshold for a
single accelerometer channel exceeds 14.4g, 40 milliseconds
of data (8-millisecond pretrigger and 32-millisecond post-
trigger) are transmitted to a sideline receiver connected to a
laptop computer. From the acceleration-time histories, the
severity (magnitude of linear and rotational acceleration)
and duration of the head acceleration and the location of
the impact on the helmet are computed and stored for
future analysis.19,20Head impact data from all participat-
ing institutions were uploaded to a secure central server
with a consolidated database and subsequently exported
for statistical analysis (SAS Institute Inc, Cary, NC). Data
were reduced in postprocessing to exclude any impact event
with a peak resultant linear acceleration less than 10g to
eliminate events that had been determined during initial
system development to be inconsequential, nonimpact
events (eg, running, jumping).24Any impact event in which
the acceleration-time history pattern of the 6 linear
accelerometers did not match the theoretical pattern for
rigid body-head acceleration,19such as a single accelerom-
eter spike that can occur during throwing or kicking a
helmet, also was excluded. Finally, all impacts exceeding
125g were reviewed visually to verify quality of acceleration
data. These methods have been verified by comparing
measured impacts with video footage.23,25
Definitions and Protocol
A team session was defined as a formal practice (players
wore protective equipment and had the potential of head
contact) or a game (competitions and scrimmages). A
player session was defined as occurring when at least 1 head
impact was recorded during 1 team session because this
provided confirmation that the given participant was
present and was exposed to impact. Impacts that were
recorded outside of the time of the team session, as defined
by the team staff, were excluded from the analysis. Head
impact data were recorded during 215 team sessions (172
practices and 43 games during the 2007 fall season). The
number of sessions that were analyzed for each player
ranged from 1 to the maximum number of possible team
sessions for his school (Table).
Head impact frequency and the location of the impacts
on the helmet were analyzed for each player by team,
session, and position. Head impact frequency was quanti-
fied using 5 measures: season impacts, which indicated the
total number of head impacts recorded for a player during
all sessions; practice impacts, the total number of head
impacts recorded for a player during all practices; game
impacts, the total number of head impacts recorded for a
player during all games; impacts per practice, the average
number of head impacts for a player during practices; and
impacts per game, the average of the number of head
impacts for a player during games. We plotted these data
using cumulative histograms and ordinary histograms of
impact events sustained by members of each team. Plotting
the data as cumulative histograms enables the reporting of
values for individual players normalized for the total
number of players on each team. For example, if every
player on the team received exactly 200 impacts during a
season, then the curve would simply be a vertical line at the
200 value on the x-axis.
Figure 1. A, Football players wore instrumented helmets during practices and competitions to record the frequency, magnitude, and
location of head impacts. These helmets were instrumented with 6 accelerometers (a), telemetry electronics (e), and a battery (b). B, The
HIT System (developed by Simbex, Lebanon, NH, and marketed commercially as the Sideline Response System by Riddell Inc, Elyria, OH)
comprises an instrumented helmet, a sideline receiver, and a laptop computer. C, The regions that defined the front, right, back, and top
impact locations on the helmet and face mask are shown.
Table.Team and Player Practice and Game Sessions
TeamNo. of Players (N 5 188)
Team SessionsPlayer Sessionsa
Practices (n 5 172) Games (n 5 43) Practices, Maximum (Median) Games, Maximum (Median)
aA player session was defined as 1 session (practice or game) in which a player received at least 1 head impact. The maximum number of sessions
for an individual player ranged from 1 to the number of team sessions (practices plus games).
Journal of Athletic Training551
Impact locations to the helmet and face mask were
computed as azimuth and elevation angles in an anatomic
coordinate system relative to the estimated center of
gravity of the head19and categorized as front, left, right,
back, and top (Figure 1C). Four equally spaced regions
centered on the anatomic midsagittal and coronal planes
defined the front, left, right, and back impact locations. All
impacts occurring above an elevation angle of 656 where 06
elevation was defined as a horizontal plane through the
center of gravity of the head were considered impacts to the
top of the helmet.26
To determine if season impacts, practice impacts, game
impacts, impacts per practice, and impacts per game for
individual players were different among teams, we used the
Wilcoxon test for comparing Kaplan-Meier survival
curves. Comparing the survival curves provided both a
compelling visualization and a valid (nonparametric)
method for comparisons of the positively skewed data of
season impacts, practice impacts, and game impacts. For
consistency, we used similar analyses for the rate variables
(impacts per practice and impacts per game), but these
measures were not skewed. Because only 3 possible post
hoc comparisons existed among the 3 teams, a was
adjusted using the Bonferroni method to , .0167. We
used SAS (version 9.2; SAS Institute Inc) for analyses. The
relationship between the number of season head impacts
that a player received and the number of NCAA-defined
A-Es was examined using linear regression (SigmaPlot,
Systat Software Inc, San Jose, CA), and the resulting x2
and P values are reported.
To determine if impacts per game and practice were
different among players of the various positions, we used
activity and random effects for position and activity within
each player. The percentage of impacts at various locations
within position (eg, left versus right in WRs) and location
differentials between positions (eg, left versus right in WRs
compared with QBs) were compared using mixed linear
random effects for position and location within each player.
The Holm test was used to adjust P values for multiple
comparisons because of the large number of comparisons.
We used SAS (version 9.2; SAS Institute Inc) for analyses,
and the resulting t and P values are reported.
The number of season impacts for players on team C was
higher than those for players on team A (x215 13.106,
P , .001) and tended to be higher than for those on team B
(x215 5.0830, P 5.02; Figure 2). The number of season
impacts for players did not differ between teams A and B
(x215 0.7778, P 5 .38). The maximum (median) number of
season impacts was 1022 (257), 1412 (294), and 1444 (438)
among players on teams A, B, and C, respectively. The
percentages of players receiving any given number of
season impacts are plotted in the cumulative histogram of
Figure 2A, whereas the percentages of players receiving
season impacts in bins of 200 impacts are plotted in the
ordinary histogram of Figure 2B.
Across all players in the study, the number of season
impacts increased with A-E (R25 0.415, P , .001).13
However, the variability in the number of season impacts
for a given A-E increased substantially as the number of A-
Es increased (Figure 3). For example, the number of
season impacts for players with an A-E value of 50 ranged
from 175 to 1405.
The number of practice impacts was less for players on
team A than those for players on team C (x215 9.405,
Figure 2. The total number of head impacts for individual players during the season differed with team. A, The complete distribution of
the number of season head impacts is plotted as a cumulative histogram for all players of each team. The x-axis shows the number of
season impacts, and the y-axis gives the number of players, as a percentage of the team, with the given number of season impacts or
greater. B, An ordinary histogram of the same data.
552Volume 45 N Number 6 N December 2010
P 5 .002; Figure 4A and B). We found no differences in
the number of practice impacts between players on teams B
and C (x215 2.0731, P 5 .15) and between players on
teams A and B (x215 2.1666, P 5 .14; Figure 4A). The
maximum (median) numbers of practice impacts were 761
(160), 811 (207), and 910 (210) among players on teams A,
B, and C, respectively.
The number of game impacts was higher for players on
team C than for players on team A (x215 24.508, P ,
.001) and for players on team B (x215 10.491, P 5 .001;
Figure 4 C and D). Game impacts for individual players on
teams A and B did not differ (x215 1.9185, P 5 .17). The
maximum (median) numbers of game impacts were 351
(79), 601 (102), and 775 (173) among players on teams A,
B, and C, respectively.
The number of head impacts per practice was lower for
players on team A than for players on team B (x21 5
18.9576, P , .001) or C (x21 5 11.9123, P , .001;
Figure 5A and B). Impacts per practice were not different
between teams B and C (x215 0.4935, P 5 .48). The
maximum (median) values for the number of impacts per
practice were 15.6 (4.8), 18.9 (7.5), and 24 (6.6) among
players on teams A, B, and C, respectively. In contrast to
practices, the number of head impacts per game did not
differ by team (x215 0.4921, P 5 .78; Figure 5C and D).
The maximum (median) values for head impacts per game
were 58.5 (12.1), 66.8 (14.6), and 86.1 (16.3) among players
on teams A, B, and C, respectively.
The number of impacts per practice ranged among
positions from 3.2 (QBs) to 11.5 (DLs; Figure 6A). The
number of impacts per game ranged from 7.3 (WRs) to
29.8 (DLs; Figure 6B). This increase in the number of head
impacts per game compared with head impacts per practice
was relatively constant for all positions, with a coefficient
of 2.4 times (r25 0.92, P , .001). In general, DLs, LBs,
and OLs had a greater number of impacts per practice and
impacts per game than the players at the other positions.
Across all players, the highest percentage of impacts
occurred to the front of the helmet (Figure 7). The back of
the helmet received the second highest percentage of
impacts. When examined by positions, DBs, DLs, LBs,
and OLs had higher numbers of impacts to the front than
to the back of the helmet (t804range 5 8.92–15.23; P ,
.001). The OLs had the highest percentage of impacts to the
front of the helmet compared with players at the other
positions. Conversely, QBs had a higher percentage of
impacts to the back than to the front of the helmet (t8045
3.19, P 5 .04). The percentage of impacts to the front and
the back was not different for WRs (t8045 1.89, P . .99).
We found no difference between impacts to the left and to
the right side at any position (t804range 5 0.13–0.97; P .
.99). Impacts to the top of the helmet occurred more often
than impacts to either side but with a difference found only
for OLs (t7995 4.68, P , .001).
We sought to quantify head impact frequency and
location for individual players on 3 collegiate football
teams during a single season. We focused on these 2
measures of head impact exposure because of the lack of
data on individual exposures and on concussion injury
mechanisms. To date, the only reported exposure measure
for individual players is the risk of injury through
participation defined using A-E.13 Although A-E is a
useful factor for comparing the risk of injury across sports,
sex, and other environmental factors, it has limited
applicability to the study of injury mechanisms.
The ability to directly measure head impacts of
individual players is critical to establishing the relationship
between head impacts and concussion injury and to
examining the potential effects of cumulative subconcus-
sive impacts. Only a few of the 188 players enrolled in our
study received an impact in all team practices and games.
More typically, head impacts occurred in approximately
one-half to two-thirds of the team sessions. Head impact
frequency recorded over the entire season for practices and
games varied by team. This was not unexpected given that
players on 1 team had substantially more practice sessions,
but this team also had the lowest median number of
average impacts per practice. Although the number of team
sessions certainly influenced the number of individual head
impacts, the structure of the practice plan and the
philosophies of the coaching staff also were likely factors
that were difficult to quantify. Interestingly, the number of
impacts a player received per game did not vary by team.
We presume this is because of the controlled and timed
nature of football games, which are less dependent on a
team’s style of play or specific practice tendencies. The
total numbers of impacts players received during all
practices and all games were comparable (Figure 4);
however, after accounting for the number of sessions of
each, the number of impacts per game was 2 to 3 times
greater than the number of impacts per practice, which was
consistent with the reported findings that injury rates are
higher in games than in practices.27The number of impacts
recorded per practice and per game for an individual player
reached maximums of 24 and 86.1, respectively. The
median values for these players were 6.3 impacts per
practice and 14.3 impacts per game. For some individual
Figure 3. The number of season head impacts increased with
athlete-exposures. An athlete-exposure was defined as 1 player in
1 session in which he or she is exposed to the possibility of
athletic injury.13However, athlete-exposure was a poor predictor of
the number of season head impacts for any given player.
Journal of Athletic Training553
players, the values that we recorded might be underesti-
mates of the actual impacts because a player might have
started a practice or game but might not have completed
the session and because we were not able to instrument all
players on each team.
Researchers have not reported head impact measures for
individual players, so direct comparison with our data is
limited but instructive. Using an earlier version of the
instrumented helmet technology, Duma et al25reported
2114 impacts in 35 practices while monitoring 38 different
players (up to 8 players per session), giving a value of
approximately 7.6 impacts per player per practice. In 10
games, they recorded 1198 impacts, for an estimated 15.0
impacts per player per game. Brolinson et al23recorded
11604 impacts over 84 sessions of games and practices.
During each session, they monitored up to 18 players, with
52 different players wearing the instrumented helmets over
the 2-season period. From their results, we estimate that
Figure 4. The number of head impacts for individual players of each team during A and B, all practices, and C and D, all games. The data
are plotted as a cumulative histogram with the number of A, practice and C, game impacts plotted on the x-axis and the percentage of
players on each team with the number of impacts, or greater, plotted on the y-axis. B and D, Ordinary histograms of the same data.
554Volume 45 N Number 6 N December 2010
the average number of impacts per player per session was
approximately 4, which would be in the lower 20% of the
188 players from our study. However, this prediction of
impacts per player per session likely is an underestimate
considering that 18 players were not instrumented each day
for the entire study. Using similar technology, Schnebel et
al22 reported 54154 impacts for 40 players over 105
sessions at 1 NCAA Division I school during 1 season.
Their overall average number of player head impacts per
session was approximately 13, which was greater than our
median value of 9.4 impacts per player per session. Mihalik
et al24reported that the total number of impacts sustained
Figure 5. The number of head impacts for individual players A and B, per practice, and C and D, per game for players on each team. The
data are plotted as a cumulative histogram with the number of impacts A, per practice, and C, per game plotted on the x-axis and the
percentage of players on each team with the number of impacts or greater plotted on the y-axis. B and D, Ordinary histograms of the
Journal of Athletic Training 555
in full-contact practice (28610) was about twice the
number of those sustained in games (12873). This ratio is
roughly consistent with our findings.
We found that player position affected both head impact
frequency and location. Other researchers have suggested
similar trends. Schnebel et al22reported that their nonline-
men (‘‘skill positions’’) received only 25% of the total
impacts, in contrast to linemen, who received 75% of the
total impacts. In another study of 1 collegiate football team
over 2 seasons, the largest percentages of impacts were
recorded in OFs (36%) and DLs (22%),24 which is
consistent with our findings. In that study, LBs received
only one-third of the impacts that the linemen received,
whereas in our study, DLs, OLs, and LBs received
approximately the same number of impacts per practice
and per game.
We found that most impacts occurred to the front of the
helmet for all player positions except QBs. The OLs had
the highest percentage of impacts to the front of the helmet,
which is consistent with the observation that OLs are more
likely to initiate and control the site of impact than other
position groups. The highest percentage of impacts to the
back of the helmet occurred in QBs, suggesting that the
QBs most often were hit from behind or were tackled,
falling backward and hitting the backs of their heads on the
ground. These explanations are based upon general
observation of football and have not been confirmed by
video analysis. Mihalik et al24did not examine impact
location by player position, but their overall results on
impact location are in general agreement with our findings
for all players.
We focused our analysis on head impact frequency and
impact location for individual players. We chose this focus
because this analysis for individual players has not been
reported and the resulting data are crucial in establishing
baseline exposures for the mechanism and the risk of
concussion injury, as well as any risk of cumulative
subconcussive injury. Accordingly, a substantial number
of data from our project were not reported in this study.
The severity (magnitude) of the linear and rotational
acceleration and the duration of head acceleration during
impacts were not reported because these are the subject of
an ongoing analysis of specific biomechanical input
variables and their relationship to symptoms and cognitive
function. In addition, cumulative measures of head impacts
have not been formulated and, hence, were not included in
this analysis. We did not report concussion injuries or any
measure of long-term cognitive deficits. Our study also was
limited to 3 teams during a single football season. Our
multiyear study is ongoing, and we will analyze any
differences among seasons as the study continues. We
selected a lower range cutoff of 10g of peak linear
acceleration of the head for inclusion as an impact to be
consistent with data-collection thresholds across the 3 test
sites. Given the size of the data set and number of levels of
within-subjects (5 [location] 3 2 [activity]) and between-
subjects (3 [team]) factors, not all sources of heterogeneity
of variance could be tested in our statistical analyses.
Although we believe that the numbers of samples would
minimize this effect, heterogeneity of variance across some
factors could affect the mixed-model analyses.
Head impact in sports continues to be an important and
growing concern at all levels of football and other sports
because of the known adverse outcomes in some cases and
the potential for long-term detrimental cognitive effects.
The exact mechanisms for and variability of concussion
Figure 6. The mean (61 SD) numbers of impacts A, per practice, and B, per game across player positions did not differ with team and
were grouped together. Impacts per game were typically 2.4 times greater than the impacts per practice across these various positions.
556Volume 45 N Number 6 N December 2010
signs, symptoms, and long-term sequelae from head
impacts, particularly in helmeted sports, are not well
understood. Few data address possible differences in
mechanisms and susceptibilities among athletes of different
ages, including children, and between sexes. Estimates of
concussion injury thresholds based on laboratory recon-
structions using animal, cadaver, and manikin surro-
gates15,28,29have been inadequately predictive of injury
when compared with measurements of actual head
acceleration.26 To appropriately evaluate the risk of
concussion injury and the potential for interventions likely
to reduce the incidence of concussions in sports, as well as
the potential role of accumulated subconcussive events, a
detailed understanding of the exposure and the mechanism
of injury is needed. Using animal models, researchers have
suggested that multiple factors likely influence the risk of
Figure 7. A–H. The mean (61 SD) percentage of season head impacts at each helmet location (front, left, right, back, top). Most players
had the highest percentage of impacts to the front of the helmet. Offensive linemen had the greatest percentage of impacts to the front of
the helmet, whereas quarterbacks had the greatest percentage of impacts to the back of the helmet.
Journal of Athletic Training557
neurologic and somatic symptoms after concussive head
impacts and that these might include previous head impact
events, location or direction of head impact, and other
mechanical and physiologic factors.7–12The data that we
presented begin this process of quantifying head impact
exposure in collegiate football players by focusing on head
impact frequency and location.
We found that an individual player can receive as many
as 1400 head impacts during a single season. The average
number of head impacts per game was nearly 3 times
greater than the average number of head impacts per
practice. We noted differences in impact frequency and
impact location among different player positions. We also
demonstrated differences in head impact frequency among
teams, but it is unclear if this is related to differences
among the players themselves, coaching approaches, or
other factors that remain to be identified. We found no
difference among teams in the average number of head
impacts per game. These data documented head impact
exposure in terms of frequency and location sustained by
individual players in college football, which varies accord-
ing to practice versus game, player position, and team.
These data could aid football-helmet manufacturers in
establishing design specifications and governing bodies in
setting testing criteria and, with further studies, could
provide clinicians and scientists with a more complete
understanding of the relationship among head impact
exposure, concussion injury, and long-term cognitive
This study was supported in part by research grant R01
HD048638 from the National Center for Medical Rehabilitation
Research at the National Institute of Child Health and Human
Development at the National Institutes of Health (Dr Greenwald)
and by research grant R01 NS055020 from the National Institute
for Neurological Disorders and Stroke at the National Institutes
of Health (Dr McAllister). The HIT System technology was
developed in part through research grant R44 HD40473 from the
National Institutes of Health (Drs Greenwald and Crisco) and
with research and development support from Riddell Inc (Elyria,
OH). We thank the researchers and institutions from which the
data were collected: Mike Goforth, ATC, Virginia Tech Sports
Medicine; Steve Rowson, MS, Virginia Polytechnic Institute and
State University; Dave Dieter, Edward Via Virginia College of
Osteopathic Medicine; Jeff Frechette, ATC, and Scott Roy, ATC,
Dartmouth College Sports Medicine; Mary Hynes, RN, MPH,
Dartmouth Medical School; and David J. Murray, ATC, and
Kevin R. Francis, Brown University. We thank Lindley Brainard
of Simbex for coordination of all data collection. We also thank
Tor Tosteson, PhD, and Jason T. Machan, PhD, for their
assistance with statistical analysis.
Joseph J. Crisco, PhD; Jonathan G. Beckwith, MS; Jeffrey J.
Chu, MS; Richard M. Greenwald, PhD, and Simbex reported
having a financial interest in the instruments (HIT System,
Sideline Response System [Riddell Inc]) that were used to collect
the data in this study.
1. Gerberding JL. Report to Congress on Mild Traumatic Brain Injury in
the United States: Steps to Preventing a Serious Public Health
Problem. Atlanta, GA: Centers for Disease Control and Prevention;
2. Thurman DJ, Branche CM, Sniezek JE. The epidemiology of sports-
related traumatic brain injuries in the United States: recent
developments. J Head Trauma Rehabil. 1998;13(2):1–8.
3. Powell JW, Barber-Foss KD. Traumatic brain injury in high school
athletes. JAMA. 1999;282(10):958–963.
4. Powell JW, Barber-Foss KD. Injury patterns in selected high school
sports: a review of the 1995–1997 seasons. J Athl Train. 1999;34(3):
5. Collins MW, Lovell MR, McKeag DB. Current issues in managing
sports-related concussion. JAMA. 1999;282(24):2283–2285.
6. Guskiewicz KM, Weaver NL, Padua DA, Garrett WE Jr. Epidemi-
ology of concussion in collegiate and high school football players.
Am J Sports Med. 2000;28(5):643–650.
7. Raghupathi R, Mehr MF, Helfaer MA, Margulies SS. Traumatic
axonal injury is exacerbated following repetitive closed head injury in
the neonatal pig. J Neurotrauma. 2004;21(3):307–316.
8. Laurer HL, Bareyre FM, Lee VM, et al. Mild head injury increasing
the brain’s vulnerability to a second concussive impact. J Neurosurg.
9. Gennarelli TA, Thibault LE, Adams JH, Graham DI, Thompson CJ,
Marcincin RP. Diffuse axonal injury and traumatic coma in the
primate. Ann Neurol. 1982;12(6):564–574.
10. Gennarelli TA, Thibault LE. Biomechanics of acute subdural
hematoma. J Trauma. 1982;22(8):680–686.
11. Gennarelli TA. Animate models of human head injury. J Neuro-
12. Prange MT, Margulies SS. Regional, directional, and age-dependent
properties of the brain undergoing large deformation. J Biomech Eng.
13. Dick R, Agel J, Marshall SW. National Collegiate Athletic
Association Injury Surveillance System commentaries: introduction
and methods. J Athl Train. 2007;42(2):173–182.
14. Pellman EJ, Viano DC, Tucker AM, Casson IR; Committee on Mild
Traumatic Brain Injury, National Football League. Concussion in
professional football: location and direction of helmet impacts. Part
2. Neurosurgery. 2003;53(6):1328–1341.
15. Pellman EJ, Viano DC, Tucker AM, Casson IR, Waeckerle JF.
Concussion in professional football: reconstruction of game impacts
and injuries. Neurosurgery. 2003;53(4):799–814.
16. Naunheim RS, Standeven J, Richter C, Lewis LM. Comparison of
impact data in hockey, football, and soccer. J Trauma. 2000;48(5):
17. Moon DW, Beedle CW, Kovacic CR. Peak head acceleration of
athletes during competition: football. Med Sci Sports. 1971;3(1):
18. Reid SE, Tarkington JA, Epstein HM, O’Dea TJ. Brain tolerance to
impact in football. Surg Gynecol Obstet. 1971;133(6):929–936.
19. Crisco JJ, Chu JJ, Greenwald RM. An algorithm for estimating
acceleration magnitude and impact location using multiple non-
orthogonal single-axis accelerometers. J Biomech Eng. 2004;126(6):
20. Manoogian S, McNeely D, Duma S, Brolinson G, Greenwald R.
Head acceleration is less than 10 percent of helmet acceleration in
football impacts. Biomed Sci Instrum. 2006;42:383–388.
21. Beckwith JG, Chu JJ, Greenwald RM. Validation of a noninvasive
system for measuring head acceleration for use during boxing
competition. J Appl Biomech. 2007;23(3):238–244.
22. Schnebel B, Gwin JT, Anderson S, Gatlin R. In vivo study of head
impacts in football: a comparison of National Collegiate Athletic
Association Division I versus high school impacts. Neurosurgery.
23. Brolinson PG, Manoogian S, McNeely D, Goforth M, Greenwald R,
Duma S. Analysis of linear head accelerations from collegiate football
impacts. Curr Sports Med Rep. 2006;5(1):23–28.
558Volume 45 N Number 6 N December 2010
24. Mihalik JP, Bell DR, Marshall SW, Guskiewicz KM. Measurement
of head impacts in collegiate football players: an investigation of
positional and event-type differences. Neurosurgery. 2007;61(6):
25. Duma SM, Manoogian SJ, Bussone WR, et al. Analysis of real-time
head accelerations in collegiate football players. Clin J Sport Med.
26. Greenwald RM, Gwin JT, Chu JJ, Crisco JJ. Head impact severity
measures for evaluating mild traumatic brain injury risk exposure.
27. Shankar PR, Fields SK, Collins CL, Dick RW, Comstock RD.
Epidemiology of high school and collegiate football injuries in the
United States, 2005–2006. Am J Sports Med. 2007;35(8):1295–1303.
28. Gurdjian ES, Lissner HR, Hodgson VR, Patrick LM. Mechanism of
head injury. Clin Neurosurg. 1964;12:112–128.
29. Ommaya AK, Yarnell P, Hirsch AE, Harris EH. Scaling of
experimental data on cerebral concussion in sub-human primates to
concussion threshold for man. In: Proceedings of the 11th Stapp Car
Crash Conference. New York, NY: Society of Automotive Engineers;
Address correspondence to Joseph J. Crisco, PhD, Bioengineering Laboratory, Department of Orthopaedics, The Warren Alpert Medical
School of Brown University and Rhode Island Hospital, CORO West, Suite 404, 1 Hoppin Street, Providence, RI 02903. Address e-mail to
Journal of Athletic Training 559