Journal of Sports Sciences, 2001, 19, 171±179

Impact forces and neck muscle activity in heading by

collegiate female soccer players
JEFFREY A. BAUER,1* TOM S. THOMAS,2 JAMES H. CAURAUGH,2
THOMAS W. KAMINSKI2 and CHRIS J. HASS2
1
Department of Exercise and Sports Studies, SUNY Cortland, Cortland, NY 13045 and 2Center for Exercise Science,
Department of Exercise and Sport Sciences, University of Florida, Gainesville, FL 32611, USA

Accepted 17 October 2000

Three soccer header types (shooting, clearing and passing) and two heading approaches (standing and jumping)
were manipulated to quantify impact forces and neck muscle activity in elite female soccer players. The 15
participants were Division I intercollegiate soccer players. Impact forces were measured by a 15-sensor pressure
array secured on the forehead. The electromyographic (EMG) activity of the left and right sternocleidomastoid
and trapezius muscles was recorded using surface electrodes. Maximum impact forces and impulses as well as
the EMG data were analysed with separate repeated-measures analyses of variance. Impact forces and impulses
did not di˛er among the header types or approaches. Higher values were found for jumping versus standing
headers in the mean normalized EMG for the right sternocleidomastoid. In addition, the integrated EMG was
greater for the right sternocleidomastoid and right and left trapezius (P < 0.05). The sternocleidomastoid
became active earlier than the trapezius and showed greater activity before ball contact. The trapezius became
active just before ball contact and showed greater activity after ball contact. The increased muscle activity
observed in the neck during the jumping approach appears to stabilize the connection between the head and
body, thereby increasing the stability of the head±neck complex.

Keywords: biomechanics, heading, kinetics, soccer.

Introduction (Smodlaka, 1984). These repetitive impacts to the skull

Soccer is the most popular sport in the world and is
one of the fastest growing sports in the United States. Its
rise in popularity is evident by the increase in participa-
tion at youth, high school and collegiate standards.
Women’s college soccer is the most frequently added
sport among intercollegiate institutions. The increased
number of participants at all ages has been accompanied
by a growing concern in the sports medicine community
about the safety of various aspects of the game (Boden
et al., 1998). Heading is one skill that has recently come
under close scrutiny.
On average, soccer players head the ball six times
during a game (Sortland and Tysvaer, 1989). The
potential e˛ects of soccer heading are compounded
when one considers that players perform approximately
5250 headers over the course of a 15-year career
* Address all correspondence to Je˛rey A. Bauer, Room 148, Studio
West, Cortland, NY 13045, USA.
may lead to injury. Heading injuries account for 4±
22% of soccer injuries (Tysvaer, 1992). The potential
side-e˛ects of these repeated impacts to the brain
remain unclear. Several studies have examined single
and multiple heading histories to determine the e˛ect
of heading on chronic symptoms of injury. The clinical
manifestations of these repeated impacts range from
headaches to brain damage (Tysvaer and Storli, 1981,
1989; Burslem and Lees, 1988; Schneider and Zernicke,
1988; Townend, 1988; Fields, 1989; Matser et al.,
1998) as well as acute and chronic cognitive disability
(Putukian et al., 2000).
The head is not the only structure that plays an
important role during heading. The surrounding
musculature in the neck provides a mechanism for
generating power (Burslem and Lees, 1988; Tysvaer and
Storli, 1989) and serves to dissipate energy (Shapiro
and Frankel, 1989) during execution of this skill. More-
over, Tysvaer and Storli (1981) noted that, when soccer
players are `prepared’ to head the ball and they contact

Journal of Sports Sciences ISSN 0264-0414 print/ISSN 1466-447X online Ó 2001 Taylor & Francis Ltd
http://www.tandf.co.uk/journals
172
the ball properly, the e˛ects of the impact force are
e˛ectively reduced by the musculotendinous structures
in the neck. Studies investigating the motion of the head
during soccer heading have concluded that injury
depends more on the acceleration of the head than
on the applied force (Lynch and Bauer, 1996). Soccer
players may reduce the acceleration of the head by
e˛ectively contracting the muscles in the neck, thus
safeguarding against injury. Contracting the neck
muscles increases the e˛ective mass of the body that
contacts the ball and minimizes the acceleratory
e˛ects of the impact forces on the forehead during
heading. Although the existing literature has established
the importance of involving the neck muscles during
heading, no previous studies have examined the acti-
vation patterns of these muscles relative to ball impact.
Schneider and Zernicke (1988) observed an
increased risk of injury during heading for participants
who have a lower ratio of the mass of the head to the
mass of the ball. Children had lower ratios than adults,
which put them at a greater risk of injury. As children
reach adolescence and beyond, there is a growing dif-
ference between the body mass of male and female
soccer players. The typical lower body mass in female
soccer players may increase the likelihood of injury.
Given the relationship of body mass and potentially
dangerous rotational acceleration of the head during
heading, there is surprisingly no literature examining
the e˛ects of heading on female soccer players.
The aims of this study were to assess impact forces
on the forehead and the involvement of the neck
musculature associated with di˛erent types of headers
in female soccer players. Impact force, impulse and the
magnitude and duration of neck muscle activity before,
during and after ball impact during clearing, shooting
and passing headers were investigated.
Impact force and neck muscle activity are considered
to be dependent on the heading technique used and the
speci®c aim of each type of header. A clearing header
requires the ball to be projected high into the air over
a longer distance than the other types of headers.
A shooting header must have su˝cient speed to elude
the goalkeeper, although the ball is not projected over
as large a distance. A passing header advances the ball
over a small distance towards an open area or team-
mate. During the execution of a header, the head is
accelerated forward by the neck musculature to generate
momentum that can be transferred to the ball. Just
before impact, the muscles of the neck must also act to
stabilize the head to dissipate the e˛ects of the contact
with the ball.
Given that the approach run in a jumping header
increases the momentum of the e˛ective striking
mass impacting the ball, we hypothesized (1) that the
jumping header would produce greater impact force
Bauer et al.
on the forehead than a standing header given a similar
incoming ball speed. With a similar incoming ball
speed, we hypothesized (2) that the greater momentum
of the e˛ective striking mass would result in a reduction
in the propulsive requirements of the neck musculature,
but would not change the stabilizing components. Thus,
EMG activity before impact would be less in jumping
headers than standing headers. These hypotheses are
based on the results of Mawdsley (1978), who found
that the contribution of the head and neck in heading
was greater in standing headers than those performed
with an approach run. Given the di˛ering propulsive
requirements of these di˛erent techniques, we hypoth-
esized that impact forces and neck muscle activity (3)
would increase in the following order of header types:
passing, shooting and clearing headers. Based on the
actions of the head and neck during heading, we
hypothesized (4) that the sternocleidomastoids would
be more active before ball contact to generate the
forward velocity of the head. Just before impact, the
trapezius muscles should become active and remain
active following impact to stabilize the head and neck
system.
Methods and materials
Participants
Fifteen healthy female volunteers participated in the
study. All participants were members of the University
of Florida’s varsity soccer team. This sample of athletes
had signi®cant experience in heading and was repre-
sentative of a typical Division I women’s collegiate
soccer team. Descriptive characteristics of the partici-
pants are presented in Table 1.
Instrumentation
A Paromed DataLogger Ò (Paromed Medizintechnik,
Germany) was used to measure impact force and neck
muscle activity during the heading tasks. The mag-
nitude and force distribution of the ball’s impact on
the forehead was measured by a 15-sensor oval array
(see Fig. 1). The array spanned over the left, right,
top and bottom regions of the forehead. Each sensor
contained a silicone-®lled hydrocell in which an
absolute measuring piezoresistive microsensor was
embedded. The sensors themselves had a full range of
measurement of 20±20,000 Hz and were accurate to
within ± 2% of their stated full range measurement
(German TöV), regardless of heat, humidity or loading
frequency. All force data were collected at 200 Hz
providing 3000 discrete data values per second during
data acquisition.
Biomechanics of heading 173

Table 1. Characteristics of the participants

Variable Mean ± s Range

Age (years) 20.3 ± 2.3 18±27

Body mass (kg) 60.1 ± 3.4 56.8±65.9
Height (m) 1.67 ± 0.05 1.60±1.71
College playing experience (years) 1.8 ± 0.9 1±4
High school playing experience (years) 3.7 ± 1.0 0±4
Number of headers in practice* 11.7 ± 9.4 5±40

* Denotes the average number of headers performed in each practice session over
1 week.

Fig. 1. The pressure array used to measure impact forces.

Active bipolar Ag-AgCl surface electrodes were

applied to the skin overlying the muscle bellies of both
the right and left sternocleidomastoid and trapezius
muscles to measure the myoelectric response. A ground
electrode was attached over the clavicle. The ampli®er
gain was set to 800 with the input impedance of 10 GW.
All surface EMG data were recorded at a sampling rate
of 800 Hz.
The impact force and EMG data were stored
on PCMCIA cards, which were subsequently down-
loaded to a computer for storage and analysis. Software
speci®cally developed for the DataLogger Ò (DLS)
was used to transfer and analyse the data. All trials Fig. 2. Placement of electrodes over the neck muscles and
were videotaped using an 8-mm 30-Hz VHS video the water polo cap securing the pressure array to the partici-
camcorder and the tapes were reviewed to determine pant’s head.
successful trials and to ensure the consistency of the
speed at which the balls were tossed towards each
participant. forehead and was held in place by a water polo swim cap
worn over the array and securely fastened to the head
(see Fig. 2). Before placement of the surface electrodes,
Procedures
the skin over the right and left sternocleidomastoid and
All testing was conducted at the soccer practice facility trapezius muscles was slightly abraded and cleaned
at the University of Florida. Each participant was tested with alcohol. The electrodes were then placed over
on a single day. Before testing began, the test protocol the clavicle and the geometric centre of the right and
was explained and individuals read and signed an left sternocleidomastoid and trapezius muscles by the
informed consent approved by the institutional review same investigator for all participants (see Fig. 2). The
board. DataLogger Ò data acquisition system controller was
The sensor array was positioned over the participant’s ®tted securely around each participant’s waist.
174 Bauer et al.

Each treatment condition began with an individual

performing a maximum voluntary isometric contraction
(MVC) of the sternocleidomastoid and trapezius
muscles. The MVC of the sternocleidomastoid was
recorded as resistance was provided to the front of the
head in the posterior direction as the participants
attempted to move their head anteriorly. The MVC
of the trapezius was recorded with the participants
attempting to move their head posteriorly as resistance
was provided in the anterior direction on the back of the
head.
A Wiel CoerverÒ outdoor training goal (Kwik-GoalÒ,
Quakertown, PA; 2.3 ´ 5.5 m in size) was placed 7 m
in front of the participant and served as the target to
Fig. 3. Overhead view of the experimental set-up.
direct all headers. The thrower stood in the middle of
the training goal directly in front of the participant.
A camcorder was placed 10 m from the line of the toss
and provided a sagittal view of the activity (see Fig. 3).
All participants warmed up by performing several
practice headers similar to the types tested. They then
performed the ®rst of the six possible combinations of
headers (three types and two approach conditions). The
types of headers were (1) shooting, (2) clearing and (3)
passing. Each of these headers was performed under
two di˛erent approach conditions: (1) while standing
and (2) while jumping. The order of testing was
randomized using a Latin square design. The ball
was thrown underhand from a distance of 6.5 m by an
assistant women’s soccer coach. Pilot work indicated
that the thrower could repeatedly deliver the ball
to the participant such that it arrived at a speed of
6.8 ± 0.5 m ´ s-1.
When the participants performed a standing
approach, they kept their feet in the same position
throughout the movement (see Fig. 4). When per-
forming the jumping approach, they began two steps
behind the standing header position and jumped
towards the ball before contact. The clearing header
was performed with enough force to head the ball
over the training goal; a shooting header was directed
diagonally towards the ground. The passing header was Fig. 4. Performance of the standing header.
directed anywhere from the waist to the feet of the
thrower. Participants performed six blocks of trials with
three successful headers for each treatment condition were within the range established during the pilot study.
resulting in a total of 18 headers. The trials were The estimated ball speed of the throw was determined
considered successful if the ball was struck such that by dividing the known distance from the thrower to
it was projected back at the thrower within 1 m right the player by the temporal data generated from video
or left of the mark from which the ball was thrown. analysis. Speci®cally, the time from which the ball left
Acceptable ball heights were 0±0.75 m for passes or the coach’s hand to the time it contacted the player’s
shots and > 2.2 m for clears. head was computed. Only trials that met the ball speed
criteria as well as on-®eld accuracy criteria were used for
statistical analysis.
Data processing
Impulse values were calculated using a trapezoidal
Videotapes recorded during testing were evaluated method of integration applied to the force data recorded
before data analysis to ensure that the incoming speeds from all sensors registering applied force during the
Biomechanics of heading 175

impact. Ball contact times for all trials fell within
previously reported values of 15±22 ms (Lynch and
Bauer, 1996). Consequently, these impulse data were
then summed and averaged.
The time during which the muscles were considered
to be active was established to be within 500 ms before
ball impact until 500 ms after impact. The muscles
were considered active when they were at 20% of the
MVC value. The raw MVC EMG signal was centred
and full-wave recti®ed using the DLS 2.1Ò (Paromed,
Germany) data acquisition and analysis software.
The raw signal representing heading trials was ®ltered
at 10 Hz to generate a linear envelope, which was then
®ltered at 20 Hz to remove miscellaneous artifacts.
The ®ltered signals were then normalized using the
mean MVC signal value, which was calculated from
the values recorded during the middle 3 s of a 5-s
data collection trial conducted just before the start of
testing. Integrated EMG data were also determined
using a trapezoidal integration method applied to data
recorded within the 500 ms before and 500 ms after
the impact interval, during which EMG values exceeded
the muscle activation threshold. These values were then
normalized relative to EMG values collected during the
MVC data collection for a similar activation duration.
Data analysis
Measures of central tendency and variability were
determined for impact force data, impulse, and the
EMG activity recorded before, during and after ball
impact. Separate within-individuals analyses of variance
were performed on (a) the peak impact force and
impulse and (b) the EMG data (magnitude and temporal
values) from the right sternocleidomastoid, left sterno-
cleidomastoid, left trapezius and right trapezius, and for
each type of header during the standing and jumping
approach conditions. All statistical tests were conducted
with alpha set to 0.05.
Results
Impact force
The impact force and impulse data were analysed
in separate within-individuals analyses of variance.
The forehead sensors were divided into four regions:
left, right, top and bottom. The header type ´ header
approach ´ forehead area (3 ´ 2 ´ 4) analysis of vari-
ance (ANOVA) failed to identify any signi®cant main
e˛ects or interactions for either impact force or impulse.
The impact force and impulse data are summarized in
Tables 2 and 3, respectively.
Muscle activity
Analysis of muscle activity data with a 3 ´ 2 (header
type ´ header approach) repeated-measures ANOVA
indicated a signi®cant header approach main e˛ect
for muscle activity. The right sternocleidomastoid
showed a higher peak normalized EMG during jumping
headers than standing headers (F1,89 = 4.43, P < 0.05).
No other signi®cant di˛erences were observed in peak
normalized EMG. There was no signi®cant interaction

Table 2. Peak impact force (N) by region (mean ± s)

Jumping Standing

Region Shooting Clearing Passing Shooting Clearing Passing

Left 150 ± 13.2 151 ± 16.6 152 ± 32.3 151 ± 11.4 151 ± 18.0 150 ± 12.4
Top 160 ± 6.33 157 ± 3.89 160 ± 12.4 155 ± 61.9 162 ± 11.4 159 ± 7.63
Bottom 157 ± 18.5 155 ± 23.6 159 ± 30.9 165 ± 29.0 163 ± 20.9 156 ± 22.4
Right 185 ± 25.2 179 ± 29.1 183 ± 30.9 189 ± 30.7 191 ± 37.8 183 ± 30.9

Table 3. Impulse value (N ´ s) by region (mean ± s)

Jumping Standing

Region Shooting Clearing Passing Shooting Clearing Passing

Left 3.10 ± 0.13 3.08 ± 0.07 3.20 ± 0.13 3.13 ± 0.05 3.23 ± 0.27 3.13 ± 0.08
Top 3.32 ± 0.26 3.22 ± 0.13 3.40 ± 0.27 3.31 ± 0.12 3.50 ± 0.45 3.31 ± 0.14
Bottom 3.32 ± 0.33 3.23 ± 0.23 3.43 ± 0.42 3.53 ± 0.31 3.60 ± 0.67 3.32 ± 0.17
Right 3.96 ± 0.46 3.79 ± 0.26 4.02 ± 0.62 4.12 ± 0.43 4.26 ± 0.86 3.95 ± 0.36
176 Bauer et al.

between header approach and header types for muscle
activity.
The within-individuals design revealed signi®cant
increases in integrated normalized EMG for jumping
versus standing headers for the right sternocleido-
mastoid (F1,89 = 7.97, P < 0.01), right trapezius (F1,89 =
7.34, P < 0.01) and left trapezius muscles (F1,89 = 10.21,
P < 0.01). No di˛erences in normalized EMG were
observed for header approach technique for the left
sternocleidomastoid. The peak and integrated nor-
malized EMG values are summarized in Tables 4 and 5,
respectively. There were no signi®cant di˛erences in
peak normalized EMG or integrated normalized EMG
among any of the three header types or a signi®cant
header type ´ approach interaction.
Duration of muscle activation
The sternocleidomastoid muscles became active earlier
than the trapezius muscles before ball contact and had
a higher normalized EMG before contact (F1,89 = 6.72,
P < 0.05). The trapezius muscles remained active
longer than the sternocleidomastoid muscles and had
higher normalized EMG activity after contact (F1,89 =
7.02, P < 0.05). This pattern was observed for both the
right and left sternocleidomastoid and trapezius
muscles. The duration of muscle activation is sum-
marized in Table 6 and the muscle activation patterns
are represented graphically in Fig. 5.
Discussion
Impact force
The impact necessary to cause concussion is similar in
magnitude to that which leads to gross anatomical dam-
age (Ward, 1966). However, injury to the brain can also
result from less impact than that required to concuss

Table 4. Peak normalized EMG values for jumping and standing headers (mean ± s)

Jumping Standing

Muscle Shooting Clearing Passing Shooting Clearing Passing

RSCM 4.57 ± 6.04 3.88 ± 1.72 4.03 ± 1.93 3.60 ± 1.15 3.28 ± 1.24 3.71 ± 3.70
LSCM 5.82 ± 3.62 4.12 ± 5.81 3.88 ± 3.76 3.98 ± 8.68 3.29 ± 2.73 4.10 ± 6.27
RT 5.95 ± 1.04 6.51 ± 3.29 6.86 ± 3.89 4.44 ± 4.26 5.56 ± 1.88 3.70 ± 1.71
LT 11.19 ± 8.31 6.42 ± 4.57 6.37 ± 4.56 5.22 ± 5.32 4.67 ± 3.86 8.40 ± 5.70

Abbreviations: RSCM = right sternocleidomastoid, LSCM = left sternocleidomastoid, RT = right trapezius, LT = left trapezius.

Table 5. Integrated normalized EMG values for jumping and standing headers (mean ± s)

Jumping Standing

Muscle Shooting Clearing Passing Shooting Clearing Passing

RSCM 0.72 ± 0.27 0.67 ± 0.20 0.68 ± 0.20 0.50 ± 0.07 0.47 ± 0.09 0.41 ± 0.05
LSCM 1.86 ± 1.44 0.87 ± 0.51 0.98 ± 0.62 0.85 ± 0.65 0.66 ± 0.18 0.57 ± 0.33
RT 1.29 ± 0.57 1.24 ± 1.07 1.46 ± 1.15 0.81 ± 0.27 0.85 ± 0.43 0.50 ± 0.08
LT 1.28 ± 0.61 1.02 ± 0.27 1.14 ± 0.61 0.77 ± 0.27 0.75 ± 0.46 0.63 ± 0.21

Abbreviations: RSCM = right sternocleidomastoid, LSCM = left sternocleidomastoid, RT = right trapezius, LT = left trapezius.

Table 6. Duration of muscle activation (s) for jumping and standing headers (mean ± s)

Jumping Standing

Muscle Shooting Clearing Passing Shooting Clearing Passing

RSCM 0.98 ± 0.28 0.98 ± 0.20 0.92 ± 0.20 0.78 ± 0.13 0.74 ± 0.12 0.66 ± 0.09
LSCM 1.46 ± 0.36 1.14 ± 0.30 1.22 ± 0.29 1.12 ± 0.41 0.99 ± 0.17 0.89 ± 0.16
RT 1.56 ± 0.42 1.50 ± 0.28 1.34 ± 0.32 1.16 ± 0.47 1.03 ± 0.22 0.79 ± 0.14
LT 1.72 ± 0.37 1.52 ± 0.25 1.51 ± 0.34 1.03 ± 0.22 0.95 ± 0.35 0.89 ± 0.21

Abbreviations: RSCM = right sternocleidomastoid, LSCM = left sternocleidomastoid, RT = right trapezius, LT = left trapezius.
Biomechanics of heading
Fig. 5. The duration of muscle activity. SCM = sternoclei-
domastoid, TR = trapezius.
(Reid et al., 1975). Soccer heading has recently come
under close scrutiny in the sports medicine community
because of the potential side-e˛ects of the repetitive
impacts on the head, although little is known about
the magnitudes of ball impacts associated with heading
and heading types. Based on Mawdsley’s (1978) investi-
gation of the contribution of the approach run to a
jumping header, the recorded impact forces were
expected to be largest during jumping headers. The
approach run was expected to add more force to the
header by increasing the momentum and the e˛ective
mass of the body impacting the ball. However, no dif-
ferences were observed between the impact forces for
jumping and standing headers. A similar two- to three-
step approach before striking the ball was adopted by
most players for both striking conditions; together with
the low speed of the ball toss, this may have contributed
to no statistical di˛erence being observed between
the results for standing and jumping headers. Although
there may have been a slight increase in the speed
with which the head moved forward during jumping
headers, it was not large enough to produce a noticeable
di˛erence in force. The similar magnitudes of forces
observed for standing and jumping headers may also be
attributed to the players developing the largest forces
they were prepared to tolerate during the experiment
for each type of header, thereby resulting in similar force
results.
The order of the increase in magnitude of the impact
force was expected to be pass, shot and clear because of
177
the objectives of each of these three types of headers.
However, the ®ndings did not identify any di˛erences
in impact forces among the three types of headers. This
is surprising, since the objective of each header type
involves a di˛erent combination of power and accuracy.
The lack of a di˛erence may be attributable to the ball
being thrown towards the player slowly and the players
not taking a long enough approach run before ball
contact.
Muscle activity
According to Shapiro and Frankel (1989), the muscles
in the neck are a major source of energy dissipation
and shock absorption. Muscle activity was expected
to be a˛ected by both the header approach technique
and header type. We hypothesized that di˛erences in
muscle activity would be greatest between standing and
jumping headers. During jumping headers, Mawdsley
(1978) found that players are aided by an approach
run that acts to increase the momentum and the mass
of the body, which contributes to propulsion of the ball.
Thus, there are fewer propulsive requirements of the
neck musculature during jumping headers. The present
results showed greater peak and integrated normalized
EMG values for jumping than standing headers for
the left and right sternocleidomastoid and left and right
trapezius muscles. Perhaps the relationship between
header technique and muscle activity is in part depen-
dent on the incoming ball speed and the player’s move-
ment strategy (i.e. a player trying to use her head to
propel the ball rather than to act as a rebound surface).
The relatively slow ball speed in the present study may
have required the player to actively propel the ball; thus
greater neck muscle activity than expected was needed
in the jumping condition to propel the ball the required
distance.
The peak and integrated normalized EMG neck
muscle activities were expected to increase in the
following order: passing, shooting and clearing header.
The passing header was expected to have the least
involvement of the neck muscles because accuracy,
rather than power and distance, is the goal of this
type of header. Clearing headers were expected to have
the greatest involvement of the neck muscles because
the primary goal is to direct the ball forcefully away
from the immediate area. However, we found no dif-
ferences in the normalized EMG neck muscle activity
among header types. One possible reason for this may
be the criterion conditions adopted in the experiment.
Acceptable ball heights ranged from 0 to 0.75 m for
passing and shooting headers and > 2.2 m for clearing
headers. Whether adding a greater distance criterion for
the clearing headers would result in greater neck muscle
activity warrants further investigation.
178
In addition, pressure from an opposing player con-
testing the header may also a˛ect neck muscle activity.
The lower than expected neck muscle activity for
clearing headers may be a result of the lack of simu-
lated match-play provided during this study. Without
opposing players competing for the ball, as is commonly
observed in games, the participants may not have
executed the clearing header with the same muscle
activity used in competition to ensure that the ball is
headed away safely, without danger of being intercepted
by an opponent.
Duration of muscle activation
The sternocleidomastoid muscles were expected to
play a greater role before ball contact and the trapezius
muscles to play a greater role after ball contact. In
preparation for a header, the head is accelerated forward
by the sternocleidomastoid muscles. Just before impact,
there is a brief period of deceleration as the trapezius
muscles become active. This co-contraction at impact
produces rigidity in the head, neck and shoulders in
an attempt to absorb the shock of the impact (Burslem
and Lees, 1988). After impact, there is a period
of deceleration as control of the head is regained
(Mawdsley, 1978). Support for this was observed in
the present study. Our results indicate that the sterno-
cleidomastoid muscles play a large role during the
acceleration of the head towards the ball before con-
tact. After impact, activation of these muscles quickly
decreased to near resting values. The trapezius muscles
became active just before ball contact and continued
to play a large role during deceleration of the head after
ball contact. These ®ndings are consistent with those
of Burslem and Lees (1988) for the role of the neck
muscles in soccer heading.
Overall, this study has contributed a research para-
digm that can be used to test speci®c biomechanical
aspects of heading. The greater muscle activity in the
neck during jumping was contrary to our hypothesis,
indicating that the neck musculature must act to propel
the ball forward as well as to stabilize the connection
between the head and body to dissipate the force of
impact. Impact forces at the forehead were not in¯u-
enced by the type of header (clearance, shot or pass)
or by approach (jumping or standing). The measured
impact forces were much lower than those observed in
other sports involving contact to the head (Atha et al.,
1985). Furthermore, these forces were not su˝cient
to create accelerations that approach the threshold for
acute injury or unconsciousness (Benedek and Villars,
1974).
Finally, soccer is the most popular organized sport
in the world and is rapidly growing in the United States.
With the growing concern for injury from heading
Bauer et al.
among soccer players, more long-term research is
required. Most existing research has concentrated
solely on college-age men and was based on kinematic
and inverse dynamic calculations. However, Schneider
and Zernicke (1988) reported that children and women
are at the highest risk for heading injury because of their
lower body mass. Heading is a fundamental striking
technique used during both o˛ensive and defensive play
in soccer and, as such, it is highly unlikely that this skill
will be removed from the game by any modi®cation
of the rules (Green and Jordan, 1998; Matser et al.,
1998). Moreover, there is no evidence to suggest that
soccer heading (head to ball contact) exposes players to
a high risk for long-term dysfunction. This does not
mean, however, that steps should not be taken to under-
stand the risks heading may pose to players and to make
every e˛ort to decrease or eliminate such risks. The ®rst
step for uncovering any potential dangers of heading
is to identify the biomechanical implications to the
body resulting from repeated heading of the ball. We
have shown that there are biomechanical methods that
can be used to accurately measure the important kinetic
and kinematic parameters. Once su˝cient data on the
types, direction and frequency of headers experienced
by players have been collected and analysed, we may
begin to apply this information to the long-term
prevention of head injury.
References
Atha, J., Yeadon, M.R., Sandover, J. and Parsons, K.C. (1985).
The damaging punch. B ritish Medical Journal (Clinical
Research Education), 291, 1756±1757.
Benedek, G.B. and Villars, F.M.H. (1974). Physics with
Illustrative Examples from Medicine and Biology, Vol. 1,
pp. 4±75. Reading, MA: Addison-Wesley.
Boden, B.P., Kirkendall, D.T. and Garrett, W.E., Jr. (1998).
Concussion incidence in elite college soccer players.
American Journal of Sports Medicine, 26, 238±241.
Burslem, I. and Lees, A. (1988). Quanti®cation of impact
accelerations of the head during the heading of a football.
In Science and Football (edited by T. Reilly, A. Lees,
K. Davids and W.J. Murphy), pp. 243±248. London:
E & FN Spon.
Fields, K.B. (1989) Head injuries in soccer. Physician and
Sportsmedicine, 17, 69±73.
Green, G.A. and Jordan, S.E. (1998). Are brain injuries a
signi®cant problem in soccer? Clinics in Sports Medicine,
17, 795±809.
Lynch, J.M. and Bauer, J.A. (1996). Acute head and neck
injuries. In The U.S. Soccer Sports Medicine Book (edited
by W.E. Garrett, D.T. Kirkendall and S.R. Contiguglia),
pp. 81±85, 185±190. Baltimore, MD: Williams & Wilkins.
Matser, J.T., Kessels, A.G., Jordan, B.D., Lezak, M.D.
and Troost, J. (1998). Chronic traumatic brain injury
in professional soccer players. Neurology, 51, 791±796.
Biomechanics of heading
Mawdsley, H.P. (1978). A biomechanical analysis of heading.
Momentum, 3, 16±21.
Putukian, M., Echemendia, R.J. and Mackin, S. (2000). The
acute neuropsychological e˛ects of heading in soccer: A
pilot study. Clinical Journal of Sports Medicine, 10, 104±109.
Reid, S.E., Epstein, H.M., Louis, M.W. and Reid, S.E., Jr.
(1975). Physiologic response to impact. Journal of Trauma,
15, 150±152.
Schneider, K. and Zernicke, R.L. (1988). Computer
simulation of head impact: Estimation of head-injury
risk during soccer heading. International Jour nal of Sport
Biomechanics, 4, 358±371.
Shapiro, I. and Frankel, V.H. (1989). Biomechanics of the
cervical spine. In Basic B iomechanics of the Musculo-
skeletal System (edited by M. Nordin and V.H. Frankel),
pp. 209±224. Philadelphia, PA: Lea & Febiger.
Smodlaka, V.N. (1984). Medical aspects of heading the ball
in soccer. Physician and Sportsmedicine, 12, 127±131.
179
Sortland, O. and Tysvaer, A.T. (1989). Brain damage in
former association football players: An evolution by cerebral
computed tomography. Neuroradiology, 31, 44±48.
Townend, M.S. (1988). Is heading the ball a dangerous
activity? In Science and Football (edited by T. Reilly, A. Lees,
K. Davids and W.J. Murphy), pp. 237±242. London:
E & FN Spon.
Tysvaer, A.T. (1992). Head and neck injuries in soccer:
Impact of minor trauma. Sports Medicine, 14, 200±213.
Tysvaer, A.T. and Storli, O. (1981). Association football
injuries to the brain: A preliminary report. British Journal
of Sports Medicine, 15, 163±166.
Tysvaer, A.T. and Storli, O. (1989). Soccer injuries to the
brain: A neurologic and electroencephalographic study of
active football players. American Jour nal of Sports Medicine,
17, 573±578.
Ward, A.A. (1966). The physiology of concussion. Clinical
Neurosurgery , 15, 95±111.
Copyright of Journal of Sports Sciences is the property of Taylor & Francis Ltd and its
content may not be copied or emailed to multiple sites or posted to a listserv without the
copyright holder's express written permission. However, users may print, download, or email
articles for individual use.