Content uploaded by Allison H. Gruber
Author content
All content in this area was uploaded by Allison H. Gruber on Mar 27, 2015
Content may be subject to copyright.
Footwear Science
Vol. 3, No. 1, March 2011, 33–40
Impact characteristics in shod and barefoot running
Joseph Hamill*, Elizabeth M. Russell, Allison H. Gruber and Ross Miller
Biomechanics Laboratory, 23 Totman Building, University of Massachusetts, Amherst, MA 01003, USA
(Received 27 September 2010; final version received 18 November 2010)
Increased impact characteristics are often cited as a cause of running injuries. One method that has been used to
reduce impact characteristics is to increase the thickness of the midsole of running footwear with the intention of
attenuating greater shock from the foot-ground collision. A second method that has been suggested is to run
barefoot. The purpose of this study was to compare the impact characteristics of running footwear of different
midsole thickness to a barefoot condition. Three-dimensional kinematic and kinetic data were collected as
participants ran at their preferred running speed and at a fixed speed. Impact characteristics (impact peak, time to
impact peak and vertical loading rate) were derived from the vertical ground reaction force component.
Ankle and knee joint stiffness during the loading phase of support were derived from the change in moment
divided by the change in angle. The impact parameters were statistically analyzed using a two-way, repeated
measures ANOVA. There were no significant speed by footwear condition interactions. For impact peak, ankle
stiffness and knee stiffness, there was no difference among the shod conditions but there were significant
differences between the shod and barefoot conditions. Based on their strike index, participants in this study
appeared to alter their footfall pattern from a rearfoot to a midfoot pattern when changing from running shod to
barefoot. It may be concluded that the change in the impact characteristics is a result of changing footfall pattern
rather than midsole thickness.
Keywords: impact forces; barefoot running; shod running; loading; running injury
1. Introduction
Injuries to runners have been well documented since
the beginning of the running ‘boom’ in the 1970s
(James et al. 1978). Although a great deal of research
during the intervening decades has focused on running
and running related injuries, the percentage of runners
who get injured has remained essentially the same
(Clement et al. 1981, Taunton et al. 2002). Some
researchers have implicated running footwear as a
possible risk factor for running related injuries.
Robbins and Waked (1997a) suggested that modern
footwear with a thick midsole necessitates a large
impact force in an attempt to transform the interface
into a more stable surface. They proposed that this
mechanism explains the 123% higher injury frequency
found in runners who use more expensive footwear
compared with those who use lower cost alternatives
(Robbins and Waked 1997b). The researchers
concluded that modern running footwear may be
potentially over cushioned.
Other researchers have argued that running shoes
can be part of the solution to running injuries. Divert
et al. (2005a) suggested that the purpose of the shoe is
to ‘protect the foot and leg structure by means of a
damping and low stiffness material’. Milgrom et al.
(1992) reported that basketball shoes were superior to
normal military boots in preventing stress fractures of
the foot and other overuse injuries of the foot. Rome
et al. (2008) concluded that their use in footwear with
shock absorbing insoles may reduce the occurrence of
stress fractures. Hunter et al. (2007) stated that patella
misalignment is potentially modifiable through
footwear.
A solution that is often offered to the quandary of
the effectiveness of footwear in reducing injury risk is
the use of ‘minimal footwear’. Minimal footwear can
be defined as a shoe with a thin, flexible midsole and
outsole and a light, basic upper with little or no heel
counter. In terms of overuse injuries, a more minimal
design may mean a thinner midsole. If the theory
proposed by Robbins and Waked (1997a) suggesting
that the midsole masks the magnitude of impact shock
is correct, a thinner midsole may allow runners to sense
the severity of impacts and adjust kinematics accord-
ingly. In a study by Clarke et al. (1983) using shoes
with different midsole hardnesses, it was shown that
subjects adjusted their running kinematics in such
*Corresponding author. Email: jhamill@kin.umass.edu
ISSN 1942–4280 print/ISSN 1942–4299 online
ß2011 Taylor & Francis
DOI: 10.1080/19424280.2010.542187
http://www.informaworld.com
Downloaded By: [Hamill, Joseph] At: 18:57 15 February 2011
a way that impact forces were not grossly different.
This finding was supported by Snel et al. 1985, Nigg
et al. (1987) and Hennig et al. (1996). In addition,
if a runner lands on the lateral rearfoot portion of the
shoe (as with the rearfoot footfall pattern), a thinner
midsole may reduce the lever arm between the ground
reaction force and the sub-talar joint which, in theory,
may reduce the degree of pronation, and possibly
reduce the pronation velocity (Nigg 1986) which has
been associated with running injuries (Messier and
Pittala 1988).
A currently popular extension of the minimal shoe
concept is the move to an even more minimal state,
barefoot running. Some claim running barefoot allows
the athlete to run ‘naturally’, or as nature intended;
which could potentially take advantage of the body’s
natural shock attenuation and energy return capabil-
ities. However, several researchers have shown that a
runner will alter their kinematics when running bare-
foot versus running shod (De Wit et al. 2000, Divert
et al. 2005a,b). These kinematic alterations may affect
the impact characteristics during the initial foot-
ground contact phase of running. For example,
Oakley and Pratt (1988) reported a 40% reduction in
vertical ground reaction force loading rate when
running in a 63 durometer insole compared to
barefoot.
Recently, footwear manufacturers and running
enthusiasts have increased interest in ‘minimal’ shoe
constructions and also in barefoot running. It is
generally believed that reducing the midsole thickness
of running shoes will result in alterations of the impact
characteristics and also how runners contact the
ground, but these theories have not yet been confirmed
experimentally. Therefore, the purpose of this study
was to determine the impact characteristics as footwear
midsole thickness varied and compare these character-
istics to barefoot running. We hypothesized that no
changes in the impact characteristics would result as
midsole thickness decreased but there would be signif-
icant changes in impact characteristics between shod
and barefoot running.
2. Methods
2.1. Subjects
Ten participants (five females and five males) were
used as subjects in this study. The participants had the
following characteristics: for the males – age ¼
29.6 2.9 years, height ¼1.74 0.37 m, body
mass ¼80.7 5.1 kg; and for the females –
age ¼27.4 3.7 years, height ¼1.64 0.43 m, body
mass ¼60.6 5.8 kg. All participants were regular
runners who ran at least 15 km per week and all
normally used a rearfoot footfall pattern (i.e. initial
contact was made on the heel). None of the partici-
pants had known lower extremity injuries at the time of
data collection. Each of the participants signed an
informed consent form approved by the University
Institutional Review Board.
2.2. Experimental set-up
Kinematic and kinetic data were acquired from the
right lower extremity of all subjects. Three-dimensional
kinematic data were collected at 240 Hz using an
eight-camera Qualisys Oqus capture system (Qualisys,
Inc., Gothenberg, Sweden). Ground reaction force
data were collected synchronously at 1200 Hz using an
AMTI force platform (AMTI, Inc., Watertown, MA,
USA)that was flush with the running surface.
Running speed was monitored by recording the time
between two photoelectric sensors placed 6 m apart.
2.3. Protocol
In order to track the motion of the lower extremities,
clusters of retro-reflective markers on rigid plastic
shells were placed on the thigh, leg and foot (McClay
and Manal 1999). In addition, markers were placed on
the right and left anterior superior iliac spines (ASIS),
the sacrum at the level of the ASIS markers, the heads
of the first and fifth metatarsals and the toe. These
17 markers served as tracking markers for the move-
ment trials. Prior to each individual data collection, a
standing calibration trial was collected with the subject
in quiet stance. To model the lower extremities and the
pelvis, ten calibration markers were placed to identify
joint centers and segment mass centers. For the
standing calibration trial, calibration markers were
positioned to define the individual segment geometries
and segment coordinate systems. The calibration
markers were positioned on the skin overlying:
(1) the right and left iliac crests; (2) the right and left
greater trochanters; (3) medial and lateral femoral
condyles; and (4) medial and lateral maleoli. Following
the standing calibration, the calibration markers were
removed with only the tracking markers remaining.
Three pairs of custom built shoes (size 8M for
women and size 10M for men) with identical uppers
but different midsole thicknesses were used in this
study. All shoes had the same mass, foam midsole,
midsole durometer, last slope and upper. Shoe A had
a heel height of 4 mm and thin outsole only in the
forefoot; shoe B had a heel height of 12 and 8 mm in
the forefoot; and Shoe C had a heel height of 20 and
34 J. Hamill et al.
Downloaded By: [Hamill, Joseph] At: 18:57 15 February 2011
16 mm in the forefoot. In addition, there was a
condition in which the runners ran barefoot (BF). In
all conditions, no instructions were given to the
runners on how to contact the ground during the run.
Participants were dressed in tight bicycle-type
shorts and a tight fitting short-sleeved shirt. They ran
along a 25 m runway in which the force platform was
14 m from the start of the run. Each subject was then
given a familiarization period running in the experi-
mental area for 5–10 min before data collection began.
Participants ran at their preferred speed (mean
3.57 0.41 m s
1
) and at a fixed speed
(4.0 m s
1
5%) in each of the four footwear condi-
tions for a total of eight conditions per subject. The
order of conditions was randomized to prevent an
order effect. Ten satisfactory trials in each condition
were collected. A satisfactory trial was one in which the
subject made a right foot contact on the force platform
within 5% of the prescribed running speed without
modifying their gait.
2.4. Data analysis
Kinematic data for the stance phase of each over-
ground running trial were digitized using QTM soft-
ware (Qualisys, Inc.). Synchronized raw kinematic and
kinetic signals were exported from the QTM software
in .C3D format and processed using Visual 3D
software (C-Motion, Inc., Rockville, MD). Raw kine-
matic data were low-pass filtered using a fourth order,
zero-lag Butterworth digital filter. The cut-off frequen-
cies for the low-pass filtering of kinematic were 12 Hz
which was determined using a residual analysis
(Winter 2005).
Three-dimensional joint angles for the hip, knee
and ankle were calculated using an x(flexion/exten-
sion), y(abduction/adduction), z(axial rotation)
Cardan rotation sequence (Cole et al. 1993). All
angles were referenced to coordinate systems embed-
ded in the distal segment. In addition, the metatarso-
phalangeal angle was calculated as the flexion angle
about an axis from the 1st to the 5th metatarsal.
A Newton–Euler inverse dynamics approach was
employed to calculate the 3D internal moment at the
lower extremity joints. Internal joint moments at
the ankle and knee were calculated and reported in
the coordinate system of the leg segment.
The ground reaction force (GRF) data were filtered
at 75 Hz using a recursive fourth-order low-pass digital
filter and were subsequently scaled to each partici-
pant’s body weight. From the vertical GRF compo-
nent, the following parameters were calculated: first
peak vertical force (IPeak); time to first peak vertical
force (TTP) and average vertical loading rate (VLR).
VLR was determined as the slope of the force–time
profile from 20–80% of the period between initial foot
contact and the impact peak. This section of the
vertical force–time profile was chosen because it is the
most linear portion of the initial loading portion.
Joint torsional loading stiffness was calculated as
the change in joint moment divided by the change in
joint angle (Stefanyshyn and Nigg 1998, Hamill et al.
2010). This stiffness, often referred to as
quasi-stiffness, represents the sum of many individual
stiffness measures (Latash and Zatsiorsky 1993).
Sagittal plane ankle (AStiff) and knee joint stiffness
(KStiff) were determined from initial foot contact to
midstance (i.e. maximum knee flexion). Joint moments
were scaled to body mass prior to the calculation of the
joint torsional loading stiffness.
Strike index (SI) (Cavanagh and Lafortune 1980)
was calculated to confirm the footfall pattern used by
the participants in each of the conditions. Strike index
is the point of intersection of a perpendicular drawn
from the center point of pressure at initial foot contact
to the long axis of the foot. The point of intersection is
then reported as a percent of the total foot length from
the heel. All participants had SI533% indicating that
they were all rearfoot footfall pattern runners during
the shod conditions.
2.5. Statistical analysis
All variables were determined for each of 10 trials per
participant for each condition (footwear, barefoot) and
averaged within the participant and then across con-
ditions. A repeated measures Analysis of Variance
(Speed Conditions Participants) was used to deter-
mine differences between means. A criterion level of
0.05 was employed. A Tukey post hoc test was used
when appropriate. In addition, effect sizes (ES) were
determined for all variables to aid in the interpretation
of any trends (Cohen 1989). An ES ¼0.2 indicated a
small effect, ES ¼0.5 a moderate effect and ES ¼0.8
a large effect.
3. Results
Each participant altered their footfall pattern to a
midfoot pattern when running barefoot from their
natural rearfoot pattern. With an SI533% in the shod
conditions, the indication was that the participants
used a heel-toe or rearfoot footfall pattern in the shod
conditions. However, in the barefoot condition, all of
the participants changed their footfall pattern to a
Footwear Science 35
Downloaded By: [Hamill, Joseph] At: 18:57 15 February 2011
midfoot contact (i.e. SI433% but566%).
Additionally, to verify this change in footfall pattern,
we calculated the average ankle touchdown angle. In
the three shod conditions, a dorsiflexed position was
observed in the shod conditions (11.14 4.46
o
and
10.67 4.43
o
at the preferred and fixed speeds, respec-
tively) and a plantar flexed position in the barefoot
condition (–7.13 3.00
o
and –6.58 2.24
o
at the pre-
ferred and fixed speeds, respectively).
For all of the parameters examined in this study,
there were no statistically significant Speed
Condition interactions (P40.05) indicating that the
parameters showed the same trend across conditions
when the participants ran at either their preferred
speed or the fixed running speed.
An exemplar profile of the vertical ground reaction
force component of a single trial for a male participant
is presented in Figure 1. For IPeak, there was a
significant difference between running speeds
(P50.05) with moderate to large ES (ES ¼0.65–0.77)
for each of the footwear conditions between speeds.
The IP was always greater at the fixed running speed
which was faster than the preferred speed for all
participants (see Figure 2a). However, for both
running speeds, IPeak was significantly different
between footwear conditions (P40.05). The post hoc
test revealed that the locus of the difference was
between the barefoot and the footwear conditions
with a moderate ES between these conditions
(ES40.61). The barefoot condition’s IPeak was
between 0.19 and 0.31 BW less than the three footwear
conditions.
For TTP, the results were similar to the impact
peak (see Figure 2b). There was a statistically signif-
icant difference among conditions with the post hoc
test revealing differences between the barefoot and the
three footwear conditions (P50.05; ES ¼1.35 and
3.40). At each running speed, TTP in the barefoot
condition was approximately 50% less (12.31 ms versus
24.35, 26.67 and 25.61 ms and 11.97 ms versus 21.70,
23.17 and 24.31 ms for the preferred and fixed speeds)
than the footwear conditions.
In terms of VLR (see Figure 2c), there was a
significant difference across conditions (P50.05) again
with the post hoc test indicating that there was a
difference between the footwear conditions and the
barefoot condition at each running speed. When
running at preferred speed, VLR for the barefoot
condition was 30.43 versus 72.23, 65.06 and
59.59 BW s
1
for the 4/0, 12/8 and 20/16 mm shoe
conditions respectfully. At the fixed speed, VLR for
the barefoot condition was 21.26 BW s
1
versus 69.16,
63.03, 55.02 BW s
1
for the 4/0, 12/8 and 20/16 mm
shoe conditions respectfully. While not significant,
there was a trend to decrease LR with decreasing
midsole thickness.
AStiff values are presented in Figure 3a. There was
no significant difference (P40.05) and only small
effect sizes (ES50.18) between running speeds. At
each running speed, there were no differences between
the shod conditions (P40.05; ES50.19) but there were
differences between the shod and barefoot conditions
(P50.05) with large effect sizes (ES41.0).
There was a significant difference between running
speeds (P50.05) for KStiff with the fixed running
speed having greater values than the preferred speed
with small to moderate ES ranging from 0.31 to 0.64
(see Figure 3b). However, there were no differences in
KStiff across footwear conditions (P40.05; small
ES50.29).
0.0
0.5
1.0
1.5
2.0
2.5
3.0
020 40 60 80 100
Percent support
Force (BW)
0.0
0.5
1.0
1.5
2.0
2.5
3.0(a)
(b)
0 20406080
100
Percent support
Force (BW)
4/0 mm 12/8 mm
20/16 mm Barefoot
Figure 1. Exemplar vertical GRF component for all foot-
wear conditions for a single trial of a male participant at:
(a) the preferred running speed; and (b) the fixed
running speed.
36 J. Hamill et al.
Downloaded By: [Hamill, Joseph] At: 18:57 15 February 2011
4. Discussion
The purpose of this study was to examine the impact
characteristics when running in footwear with similar
construction but varied midsole thickness and compare
these characteristics to barefoot running. The hypoth-
esis tested was that there would be no changes in the
impact characteristics (e.g., IPeak, TTP, VLR, AStiff,
KStiff) as a result of midsole thickness but there would
be differences in these impact characteristics between
shod and barefoot running. The hypothesis was, for
the most part, substantiated. For IPeak, TTP, VLR
and AStiff, there were no differences among the shod
conditions but the shod conditions were significantly
different from the barefoot condition. These differ-
ences can be explained by the fact that all participants
used a heel–toe footfall pattern in the shod conditions
but, in the barefoot condition, all of the participants
changed their footfall pattern to a midfoot contact.
When running barefoot, all subjects contacted the
ground with a slightly plantar flexed foot angle, then
contacted the ground with the heel shortly after.
The magnitude of the impact characteristics of the
shod conditions in the current study were similar to
those from other studies. The vertical ground reaction
force component, from which IPeak, TTP and VLR
are derived, is within the magnitude range of compa-
rable studies that used shod running. IPeak, VLR,
AStiff and KStiff all show magnitudes comparable to
other studies (Stefanyshyn and Nigg 1998, Milner et al.
2006, Hamill et al. 2010).
The IPeak and VLR of the vertical ground reaction
force have been associated with running injuries
(Hreljac et al. 2000, Milner et al. 2006, Pohl et al.
2009). Barefoot running has recently been advocated
to reduce running injuries by altering kinematics to
reduce the effect of impact loading (Lieberman et al.
2010). However, several earlier studies have found
increased external loading rates during barefoot run-
ning (Dickinson et al. 1985, Komi et al. 1987, Lees
1988, Oakley and Pratt 1988, De Clercq et al. 1994, De
Wit et al. 2000). These findings are not consistent with
the present study which found reduced impact ground
reaction force parameters during barefoot running at
both preferred and fixed running speeds. It may be
that, while our runners altered their footfall pattern to
a midfoot initial contact during barefoot running, in
the prior studies the runners maintained an initial
rearfoot contact. Oakley and Pratt (1988) reported a
40% reduction in VLR when subjects ran shod with a
rearfoot footfall pattern compared to barefoot running
with a rearfoot footfall pattern. Additionally, these
authors saw no difference in loading parameters when
comparing shod and barefoot running with a forefoot
footfall pattern. This suggests that a change in footfall
pattern is responsible for the differences in loading
parameters in shod versus barefoot running. However,
our findings support those of Squadrone and Gallozzi
(2009) and Divert et al. (2005a) who also observed
greater initial ground reaction force peaks during shod
running.
Figure 2. Mean GRF parameters for all participants for the
preferred (black) and fixed running speeds (white): (a) impact
peak force; (b) time to peak impact force; and (c) loading
rate.
Footwear Science 37
Downloaded By: [Hamill, Joseph] At: 18:57 15 February 2011
Landing on the anterior portion of the foot during
barefoot running has been observed in other investi-
gations (Squadrone and Gallozzi 2009). By altering
their footfall pattern to an initial midfoot contact
during barefoot running, the participants in the current
study reduced the IPeak and the VLR. However, the
IPeak in shod heel–toe running is not synonomous
with the IPeak in barefoot running. The impact in
barefoot running occurs on the midfoot area of the
foot while in the shod conditions it is on the lateral
aspect of the heel. The apparent impact peak in a
midfoot footfall pattern represents the heel touching
down after the initial midfoot contact.
In order to accommodate the alteration of foot
posture to a slightly more plantar flexed position at
foot contact, the midfoot runner employs a necessarily
stiffer sagittal ankle joint at initial impact while a less
stiff ankle joint is seen in shod running. Greater ankle
stiffness is necessary in midfoot patterns to prevent the
heel from impacting the ground. The stiffer ankle joint
was observed during barefoot running compared to the
shod conditions in the current study. While it may
appear beneficial to use a midfoot contact based on the
reduced impact peak of the vertical GRF, modifica-
tions of other impact-related variables (e.g., ankle
stiffness) should be considered as well when assessing
the pros and cons of different footfall patterns.
Impact attenuation has been associated with the
action of knee flexion during the initial portion of the
support phase (Derrick et al. 1998). In the present
study, there was a non-significant trend towards an
increase in knee joint stiffness (KStiff). Divert et al.
(2008) reported greater leg stiffness with barefoot
running. Leg stiffness may be viewed as a combination
Figure 3. Mean ankle (a) and knee (b) stiffness for all participants for the preferred (black) and fixed (white) running speeds.
38 J. Hamill et al.
Downloaded By: [Hamill, Joseph] At: 18:57 15 February 2011
of all stiffness components in the lower extremity
including knee stiffness. Thus, the non-significant
increased trend in KStiff in the barefoot condition
may indicate that shock attenuation may take place by
other means. However, Squadrone and Gallozzi (2009)
reported a reduced stride length while running bare-
foot. Reducing stride length has been shown to
decrease impact characteristics and increase shock
attenuation (Derrick et al. 1996). The reduced stride
length may help to explain why impact characteristics
are less and greater attenuation is found in barefoot
running. However, if the runner maintains the same
stride length (and footfall pattern) during shod and
barefoot running, then the impact characteristics
would be expected to be greater in barefoot running.
Runners will generally adopt a forefoot or midfoot
footfall pattern when running barefoot on a firm
surface possibly to avoid heel contact on the hard
surface. Lieberman et al. (2010) suggested barefoot
running may reduce the incidence of running related
injuries because runners would change their footfall
pattern to either mid- or forefoot landings. It has also
been suggested that forefoot landings may alter the
shock attenuation mechanisms during running (Pratt
1989, Williams and Cavanagh 1987). While results of
the present study appear to support these contentions,
it is not clear what the long-term effect of altering one’s
footfall pattern will be on the risk of running injury.
In a recent paper, Lieberman et al. (2010) suggested
a thick heel midsole will result in a rearfoot footfall
pattern. The converse of this argument may be that
a thinner midsole may lead to a change to mid- or
forefoot footfall pattern. In each of the shod condi-
tions in the current study, the participants, who were
natural rearfoot footfall pattern runners, all main-
tained the rearfoot pattern throughout even in the
footwear condition with no midsole. It would appear
that the participants in the current study altered their
footfall pattern as a function of not being shod rather
than midsole heel height.
5. Conclusions
The current study indicated that impact characteristics
are different between shod and barefoot running.
However, we also showed that the participants in this
study typically altered their footfall pattern from a
rearfoot pattern to a midfoot pattern when progressing
from shod to barefoot running. Impact characteristics
were not affected by a change in midsole height.
However, impact characteristics such as IPeak and
VLR were reduced in barefoot running compared to
the shod conditions. These findings support the
contention that the presence of footwear influences
impact characteristics, but do not necessarily indicate
that running without shoes or with a particular footfall
pattern is beneficial for avoiding injury. Long-term
prospective studies would be useful to this end. It may
be that the change in footfall pattern affects impact
characteristics to a greater degree than midsole
thickness.
References
Cavanagh, P.R. and Lafortune, M., 1980. Ground reaction
forces in distance running. Journal of Biomechanics, 13,
397–406.
Clarke, T.E., Frederick, E.C., and Cooper, L.B., 1983.
Effects of shoe cushioning upon ground reaction forces
in running. International Journal of Sports Medicine,4,
247–251.
Clement, D.B., et al., 1981. A survey of overuse running
injuries. Physician and Sports Medicine, 9, 47–58.
Cohen, J., 1989. Statistical power analysis for the behavioral
sciences. 2nd ed. Hillsdale, NJ: Earlbaum.
Cole, G.K., et al., 1993. Application of the joint coordinate
system to three-dimensional joint attitude and movement
representation: a standardization proposal. Journal of
Biomedical Engineering, 115, 344–349.
De Clercq, D., Aerts, P., and Kunnen, M., 1994.
The mechanical characteristics of the human heel pad
during foot strike in running: an in vivo cineradiographic
study. Journal of Biomechanics, 27, 1213–1222.
De Wit, B., De Clercq, D., and Aerts, P., 2000.
Biomechanical analysis of the stance phase during bare-
foot and shod running. Journal of Biomechanics, 33,
269–278.
Derrick, T.R., Hamill, J., and Caldwell, G.E., 1998. Energy
absorption in conditions of various stride frequencies.
Medicine and Science in Sports and Exercise, 30 (1),
128–135.
Dickinson, J.A., Cook, S.D., and Leinhardt, T.M., 1985.
The measurement of shock waves following heel strike
while running. Journal of Biomechanics, 18, 415–422.
Divert, C., et al., 2005a. Stiffness adaptations in shod
running. Journal of Applied Biomechanics, 21, 311–321.
Divert, C., et al., 2005b. Mechanical comparison of barefoot
and shod running. International Journal of Sports
Medicine, 26, 593–598.
Divert, C., et al., 2008. Barefoot-shod running differences:
Shoe or mass effect? International Journal of Sports
Medicine, 29, 512–518.
Hamill, J., Moses, M., and Seay, J., 2010. Lower extremity
stiffness in runners with low back pain. Research in Sports
Medicine, 17, 260–273.
Hennig, E., Valiant, G., and Liu, Q., 1996. Biomechanical
variables and the perception of cushioning for running
in various types of footwear. Journal of Applied
Biomechanics, 12, 141–150.
Footwear Science 39
Downloaded By: [Hamill, Joseph] At: 18:57 15 February 2011
Hunter, D.J., et al., 2007. Patella malalignment, pain and
patellofemoral progression: the Health ABC Study.
Osteoarthritis Cartilage, 15, 1120–1127.
Hreljac, A., Marshall, R.N., and Hume, P.A., 2000.
Evaluation of lower extremity overuse injury potential in
runners. Medicine and Science in Sport and Exercise, 32,
1635–1641.
James, S., Bates, B., and Ostering, L., 1978. Injuries to
runners. American Journal of Sports Medicine, 6, 40–50.
Komi, P., et al., 1987. Interaction between man and shoe in
running: consideration for a more comprehensive mea-
surement approach. International Journal of Sports
Medicine, 8, 196–202.
Latash, M.L. and Zatsiorsky, V.M., 1993. Joint stiffness:
myth or reality? Human Movement Science, 12, 653–692.
Lees, A., 1988. The role of athlete response tests in the
biomechanical evaluation of running shoes. Ergonomics,
31, 1673–1681.
Liebermann, D.E., et al., 2010. Foot strike patterns and
collision forces in habitulally barefoot versus shod runners.
Nature, 463, 531–536.
McClay, I.S. and Manal, K., 1999. Three-dimensional kinetic
analysis of running: significance of secondary planes of
motion. Medicine and Science in Sport and Exercise,31
(11), 1692–1737.
Messier, S.P. and Pittala, K.A., 1988. Etiologic factors
associated with selected running injuries. Medicine and
Science in Sports and Exercise, 20 (5), 501–505.
Milgrom, C., et al., 1992. Prevention of overuse injuries of
the foot by improved shoe shock attenuation. A rando-
mized prospective study. Clinical Orthopedics and Related
Research, 281, 189–192.
Milner, C.E., et al., 2006. Biomechanical factors associated
with tibial stress fracture in female runners. Medicine and
Science in Sport and Exercise, 38, 323–328.
Nigg, B., ed. 1986. Biomechanics of running shoes.
Champaign, IL: Human Kinetics Publishers.
Nigg, B.M., et al., 1987. The influence of running velocity
and midsole hardness on external impact forces in
heel–toe running. Journal of Biomechanics, 20 (10),
951–959.
Oakley, T. and Pratt, D.J., 1988. Skeletal transients during
heel and toe strike running and the effectiveness of some
materials in their attenuation. Clinical Biomechanics,3,
159–165.
Pohl, M.B., Hamill, J., and Davis, I.S., 2009. Biomechanical
and anatomical factors associated with a history of plantar
fasciitis in female runners. Clinical Journal of Sports
Medicine, 19, 372–376.
Pratt, D.J., 1989. Mechanisms of shock attenuation via the
lower extremity during running. Clinical Biomechanics,4,
51–57.
Robbins, S. and Waked, E., 1997a. Balance and vertical
impact in sports: role of shoe sole materials. Archives of
Physical Medicine and Rehabilitation, 78, 463–467.
Robbins, S. and Waked, E., 1997b. Hazard of deceptive
advertising of athletic footwear. British Journal of Sports
Medicine, 31, 299–303.
Rome, K., Hancock, D., and Poratt, D., 2008. Barefoot
running and walking: the pros and cons based on current
evidence. New Zealand Medical Journal, 121, 109–111.
Snel, J.G., et al., 1985. Shock-absorbing characteristics
of running shoes during actual running. In: D.A. Winter,
R.W. Norman, R.P. Wells, C. Hayes and A.E. Patla, eds.
Biomechanics IX-B. Champain, IL: Human Kinetics,
133–138.
Squadrone, R. and Gallozzi, C., 2009. Biomechanical and
physiological comparison of barefoot and two shod
conditions in experienced barefoot runners. Journal of
Sports Medicine and Physical Fitness, 49, 6–13.
Stefanyshyn, D.J. and Nigg, B.M., 1998. Dynamic angular
stiffness of the ankle joint during running and sprinting.
Journal of Applied Biomechanics, 14, 292–299.
Taunton, J.E., et al., 2002. A retrospective case-control
analysis of 2002 running injuries. British Journal of Sports
Medicine, 36, 95–101.
Williams, K.R. and Cavanagh, P.R., 1987. Relationship
between distance running mechanics, running economy,
and performance. Journal of Applied Physiology, 63,
1236–1245.
Winter, D.A., 2005. Biomechanics and motor control of human
movement. 3rd ed. New York: John Wiley & Sons, Inc.
Zifchock, R.A., Davis, I.S., and Hamill, J., 2006. Kinetic
asymmetry in female runners with and without retrospec-
tive tibial stress fractures. Journal of Biomechanics, 39,
2792–2797.
40 J. Hamill et al.
Downloaded By: [Hamill, Joseph] At: 18:57 15 February 2011