0198-0211/95/1610-0624$03.

00/0
FOOT& ANKLEINTERNATIONAL
Copyright 0 1995 by the American Orthopaedic Foot and Ankle Society, Inc.

Influence of Malalignment of Feet on the Plantar Pressure Pattern

in Running

Claire Jeanne Louise Sneyers, M.D.,* Roeland Lysens, M.D., Hilde Feys, P.T., and Ruoli Andries, P.T.
Heverlee, Leuven, Belgium

ABSTRACT this evaluation should be carried out. Weightbearing

When overuse injuries of the lower limbs are diagnosed and nonweightbearing examinations of feet, for in-
in athletes, the architecture and function of the foot stance, can have a different outcome.17 Static exam-
should be examined. Foot structure was evaluated in 10 ination of feet, usually based on the height of the foot
male and 14 female athletes. Based on this examination, arch, is easy to obtain clinically, but is not satisfactory
the subjects were classified into three groups: pes pla- for evaluating foot function.6
nus, pes cavus, and pes rectus. While running, the plan-
Evaluation of foot function has been approached in
tar pressure pattern of these athletes was assessed with
pressure-measuring insoles. Using these measure-
totally different ways. With a force plate, the vertical
ments, peak pressures and impulses were calculated for and horizontal components of the ground reaction
different foot anatomical locations. force were studied. Force plate systems, however, are
The plantar heel load was distributed significantly (P< limited in the analysis of foot movement, because the
0.05) more toward the anterior part of the calcaneus in force information is not specific to foot anatomical
the pes planus group compared with the normal group. locations.13Pressure distribution devices provide the
The relative load under the midfoot region was signifi- specific location of pressures as they occur beneath
cantly (Pc 0.05) lower in the pes cavus group compared the moving f 0 0 t . I ~Techniques for the study of pres-
with the other foot types. The relative load of the forefoot sure distribution beneath human and animal feet have
was significantly (P< 0.05) higher in the pes cavus group
been available for almost 100 years2 The first studies
and lower in the pes planus group. Both feet of an athlete
showed a similar plantar pressure pattern. Three suc- of dynamic foot-to-ground contact date back to the
cessive steps were comparable in terms of impulses, but work of Marey and Carlet during the last quarter of the
the peak pressures varied significantly from step to step. 19th century.’ Another early investigator was Elftman,
The local impulse and peak pressure values obtained in who in 1934 managed to obtain a direct observation of
barefoot running differed significantly from the values the momentary pressure distribution underneath the
obtained in running with sport shoes. foot.1x4810 Over the last 50 years, a number of tech-
niques for monitoring dynamic plantar pressure have
been developed, and each of these has a place in the
INTRODUCTION gradual evolution of plantar pressure assessment.’
Floor mounted devices have some disadvantages.
Good alignment of the foot is an important condition
They are limited to measurements of bare feet and
for adequate shock absorption at heel contact and a
they can analyze only one step.16 Often unnatural
stable push-off during locomotion, especially in run-
movements have to be made to place the foot prop-
ning. Several authors agree that, in case of a high or
erly on the platform, the so-called “targeting.” More-
low foot arch, the function of the subtalar and midtar-
over, these devices are not portable, they are relatively
sal joints is i m ~ a i r e dThe
. ~ connection between foot
expensive, and they have to be mounted in special
architecture and sports injuries is often cited in the
literature.6 One can conclude that in all overuse inju- walkways.’*
ries, architecture and foot function should be evalu- More recently, small disk-like transducers have
ated. There is, however, no unanimous stand on how been applied to the foot sole to assess dynamic foot-
to-ground contact. If thin enough, these transducers
can be accommodated within a patient’s shoes. Using
From the Faculty of Physical Education and Physiotherapy,
Katholieke Universiteit Leuven, Tervuursevest 101, 3001 Heverlee, this new technique, we studied the plantar pressure
Leuven, Belgium. Address reprint requests to Dr. Sneyers. pattern of running athletes. Local peak pressures and
* In training for Physical Education and Physiotherapy faculty. impulses were calculated. Peak pressures reflect the
624

Downloaded from fai.sagepub.com at UNIV TORONTO on November 21, 2016

Foot & Ankle International/Vol. 16, No. 1OIOctober 1995 MALALIGNMENT & PLANTAR PRESSURE PATTERN 625

amount of strain put on a certain anatomical region. To

estimate the total load, the impulse (the area under-
neath the force-time curve) has to be calculated."
Local peak pressures and impulses appear to be in-
dependent parameter^.^
Several authors have studied peak pressures and
impulses during ~ a l k i n g . ~ . ~ .However,
~ . ' ~ . ' ~ little infor-
mation is available on these parameters during run-
ning. Since it is known that far more overuse injuries
are present in runners compared with walkers, the
study of pressure distribution underneath the foot in
running could be clinically relevant. Measurement of
the plantar pressures sometimes reveals overload of
certain foot anatomical locations, and this could be
helpful in the treatment of overuse injuries caused by Fig. 1. Evaluation of lower leg-heel alignment.
an unfavorable pressure distribution. Direct measure-
ment of running kinetics, though, is technically diffi- while the subject was standing on a podoscope with
cult. feet in slight abduction (the medial side of the foot was
The purpose of the present study was to determine put along a line, drawn on the glass plate of the
whether a certain type of foot architecture, diagnosed podoscope). A mirror, mounted underneath the glass
by static examination, shows a specific dynamic pres- plate, allowed us to evaluate the plantar contact sur-
sure distribution in running. If so, dynamic overload of face with the glass plate. In cases of low foot arch, the
certain foot anatomical locations can be expected contact surface in the midfoot region is usually wide;
when an abnormal foot architecture is diagnosed. In in high foot arch, the midfoot region barely touches the
the future, the effect of orthotic corrections of the ground, and normal feet usually have intermediary
dynamic pressure pattern in running of abnormal feet midfoot contact surfaces.
could be evaluated using these devices. For evaluation of the lower leg-heel alignment, two
marks were placed on the bisection line of the calca-
MATERIALS AND METHODS neus, one at 0.4 inches from the bottom of the heel
and one at 0.6 inches from the top of the calcaneus,
Twenty-four athletes (14 women and 10 men) were and another mark was placed in the middle of the
used for data acquisition. Their ages ranged from 20 lower leg, 4 inches above the superior calcaneal mark
to 37 years, with an average age of 25.4 (SD 4.5). Their (Fig. 1). The angle between the calcaneal bisection line
weight ranged from 50.5 to 83 kg, with an average and the bisection line of the lower leg was measured
weight of 65.7 (SD 9.2). All subjects participated in with a goniometer while the subject was weightbear-
various sports activities at least 3 hr/week. None of ing. An angle between 0" and 7" was considered nor-
them had a prior history of foot surgery. At the mo- mal. A smaller angle indicated heel varus and a larger
ment of data acquisition, there were no injuries angle indicated heel valgus. For assessment of the
present. The subjects were asked whether they suf- heel-forefoot alignment, the foot was viewed from the
fered from overuse injuries of the lower limbs during back while it was hanging freely over the examination
their sports career. Based on their answers, they were table. The angle between an imaginary line under the
classified into two classes: one that never suffered metatarsal heads and the perpendicular line on the
from injuries or had only one acute injury (less than 4 bisection axis of the calcaneus was measured (Fig. 2).
weeks), and another that had more than one acute An angle between 4" and 8" was considered normal, a
injury or suffered from chronic (more than 4 weeks) or smaller angle was considered a forefoot valgus, and a
recurrent injuries. larger angle indicated a forefoot varus.'
Examination of foot architecture included an evalu- An athlete with a low foot arch, in combination with
ation of the foot arch and measurement of the lower a heel valgus and a forefoot varus, or two of these
leg-heel and heel-forefoot alignment. The method we three symptoms, was included in the pes planus
used was developed between 1985 and 1987 by Dr. group. An athlete with a high foot arch, in combination
M. Kvist of the Sports Medical Research Unit at the with a heel varus and a forefoot valgus, or two of these
Jyvaskyla University in Turku under the leadership of three symptoms, was placed in the pes cavus group.
Prof. P. Komi. The method was commercialized by the If all three criteria were normal or if only one out of
sport shoe manufacturer Karhu under the name of three criteria was aberrant, the athlete was considered
Karhu Ortix System.' The foot arch was evaluated normal.

Downloaded from fai.sagepub.com at UNIV TORONTO on November 21, 2016

626 SNEYERS ET AL. Foot & Ankle International/Vol. 16, No. 10fOctober 1995
FSR
Polymer
Are a

Fig. 2. Evaluation of heel-forefoot alignment. Fig. 4. The basic force-sensing resistor has two sheets of polymer
laminated together. The top sheet is covered with interdigitating
electrodes and the bottom sheet is covered with semiconducting
The dynamic examination was conducted using a
material.
pododynograph (PDG). This device contains a porta-
ble microcomputer, fed by reloadable batteries, and
insoles, each containing 64 pressure sensors (Fig. 3). portable part can be connected with the computer by
A personal computer with a graphics board is needed, a cable for transfer of the measurements. The insoles
preferably with a mouse and two serial lines. The contain tactile sensors that respond on a pressure
stimulus with a variable resistance (force-sensing re-
sistors, or FSR). Each FSR is a small conductive trans-
ducer and is composed of two sheets. The standard
substratum of the sheets is a G.E. Plastics ULTEM R
film, a thermoplastic polyetheremide with a white tem-
perature range (-30"to 170"). One sheet is covered
with interdigitating electrodes and the other is covered
with semiconducting material (Fig. 4). When forces act
on the FSR, the semiconducting material shunts the
interdigitating electrodes to a variable extent. The re-
sulting resistance is proportional to the applied pres-
sure. The resistance ranges from 1 KOhm to 10
MOhm. The acting forces should stay in a range of
1.01 to 100 PA. The mechanical rise times are 1 to 2
msec. Repeatability profiles were carried out by the
producer, and showed that in successive steps, ex-
actly the same area is pressed with the same force.
One can expect little degradation of the signal with
multiple steps because FSRs respond with a variable
resistance, which, in contrast to conductive rubber,
returns easily to the original condition when the ap-
plied pressure is discharged.
In laboratory conditions, a hysteresis of less than
2% was recorded. An array of 64 basic sensors is
lined up on a substratum. Each sensor was calibrated.
The calibration was obtained with a calibration barrel,
which has a flexible membrane on top on which the
sensors are placed. A rigid plate is placed on top of
the sensor and the membrane, and different pressures
are established by filling the barrel with water. Pres-
sure was increased in steps of 50 g/cm2 between 0
Fig. 3. The pododynograph system has a microcomputer con- and 450 g/cm2. The relation between pressure and
nected to the pressuremeasuring insoles. corresponding resistance is nonlinear. Therefore, the

Downloaded from fai.sagepub.com at UNIV TORONTO on November 21, 2016

Foot &Ankle InternationallVol. 16, No. 1010ctober 1995 MALALIGNMENT & PLANTAR PRESSURE PAlTERN 627

curves were linearized mathematically in order to ob-

tain a linear relation between the exerted pressure and
the measured value. After calibration, numerical pres-
sure values can be obtained by a simple multiplication.
The resulting calibration curves were stored in the
computer. The device allows pressure data to be eas-
ily obtained and stored. The insoles are very thin, so
they can be easily placed in different types of shoes.
They are also very durable, although a recalibration
after approximately 300 tests is advised.
The 64 pressure sensors in each insole are distrib-
uted differently in insoles of different sizes, which
means that the center-to-center distance of sensors
differs from insole to insole. The sensor size is 7 mm. I I I
The center-to-center distance is 1.5 cm in length and L7 1 L8 I L9
I
I
I
1.1 cm in width for the smallest insoles, 1.7 cm in 0 0 0 ; O lI . I I
length and 1.2 cm in width for the medium-size in-
soles, and 1.9 cm in length and 1.3 cm in width for the .----------- r----
I I
1
I
I
largest insoles. For clinical use, the spatial resolution
of the device seems acceptable to us. Most devices
with higher resolution are mounted on the floor, with I
I
the disadvantages mentioned earlier. The temporal 0 0 ) . 0
resolution of the device is 150 Hz, and this is border- L5 l L6
line for research of running biomechanics. Other dis- I

advantages of the device are a nonlinear calibration 0 0 ; . 0

I
curve and the fact that only a static calibration was I
I
performed. I
0 0 ; . 0
With the PDG system, plantar pressures were as-
sessed with the patient running barefoot and with ___----
I
-I------
I
running shoes. For the barefoot measurements, the 0 0 ;."4.
insoles were placed in ordinary gymnastic shoes. I
I
All subjects received the same type of gymnastic I
I
shoes. The soles of the gymnastic shoes are made of 0 . I . 0
I
extremely flexible material with a total thickness of 2.5
mm, so that the measurement with these shoes can be
__--_-- I
7------

considered a barefoot measurement. The running

shoe measurements were done with neutral (not cor-
rected for any type of malalignment of the foot) NlKE
running shoes (type Waffle City White/Tahiti/DKRYBL,
NR 902000 130 00). All subjects received a pair of
running shoes and were given the opportunity to break
them in by running around the track once before the
actual measurements were taken. The actual mea-
surements started 2 sec after the subject started run-
ning. For each athlete, at least three successive steps
of the left and right foot were registered.

Data Processing
The procedure described by Clarke4 was used for
The PDG system monitors pressure values at 128 data reduction. Pressure was calculated at plantar
sensors by a rate of 150 Hz for 4 sec. This leaves us regions that have an anatomical significance, as
with 76,800 pressure values for each test. Because it shown in Figure 5. Based upon the ink prints, feet
is difficult to work with these large pressure files, we were divided into heel, midfoot, forefoot, and toe re-
looked for a way to reduce the data while preserving gions, and each of these regions was further divided
the most important aspects from a clinical standpoint. into medial and lateral parts. The heel region was also

Downloaded from fai.sagepub.com at UNIV TORONTO on November 21, 2016

628 SNEYERS ET AL. Foot & Ankle InternationallVol. 16, No. 70/0ctober 1995
TABLE 1 pressures and impulses of barefoot running with those
Ratios of Peak Pressures (PP) and Impulses (I) of Two
of running with sport shoes. A paired t-test was con-
Different Regions or of a Certain Region and the Total Foot
ducted to examine whether the differences in peak
ratio 1 = I of region 2/ I of region 1 pressures and impulses were statistically significant
ratio 2 = PP of region 21 PP of region 1 (laterornedial ratios of
the heel)
(P < 0.05).
ratio 3 = I of region 4/ I of region 3
ratio 4 = PP of region 4/ PP of region 3 (anteroposterior ratios of RESULTS
the heel)
Foot Structure and Injury Rate
ratio 5 = I of region 5 + 6/ total foot I
ratio 6 = PP of region 5 + 6/ total foot PP (ratios of the rnidfoot The results of the foot structure examination are
and the total foot)
summarized in Table 2. Pes planus was diagnosed in
ratio 7 = I of region 8 + 9/ I of region 7 + 8 + 9 + 10
ratio 8 = PP of region 9 + 9/ PP of region 7 + 8 + 9 + 10 four athletes, pes caws was found in four athletes,
(ratios of the middle part of the forefoot and the total and 15 athletes had “normal” feet. One athlete could
forefoot) not be classified because of contradictory measure-
ratio 9 = I of region 7 + 8 + 9 + 10/ total foot I ment results.
ratio 10 = PP of region 7 + 8 + 9 + 10/ total foot PP (ratios of
Data on the injury rate of each foot type are sum-
the forefoot and the total foot)
ratio 11 = I of region 9 + 1O/ I of region 7 + 8 marized in Table 3. A pes planus and a pes cavus both
ratio 12 = PP of region 9 + 10/ PP of region 7 + 8 (rnediolateral resulted in significantly more overuse injuries of the
forefoot ratios) lower limbs (P-value of 0.02 for the right foot and
ratio 13 = I of region 11/ I of region 12 P-value of 0.003 for the left foot).
ratio 14 = PP of region 11/ PP of region 12 (lateromedial toe
ratios)
- - Pressure Data

Figure 6 represents a three-dimensional illustration

divided into anterior and posterior parts, and the fore- of local peak pressures and impulses of a typical pes
foot region was further subdivided into four smaller planus, pes rectus, and pes cavus.
parts. The maximum pressure occurring beneath a In Tables 4 and 5, the mean values of the calculated
certain plantar region was considered the peak pres- pressure ratios for the pes planus, pes cavus, and pes
sure of that region. To estimate the “load,” the impulse rectus group are summarized. We compared the ratios
was calculated from the area beneath each regional of the three different foot types.
force-time curve. Because the running speed was not The foot type did not influence the lateromedial load
imposed and because there was a wide range in distribution underneath the heel significantly. In bare-
weight of the athletes, it was difficult to make an foot running, the anteroposterior impulse distribution
intersubject comparison of peak pressures and im- of the heel was significantly higher (left foot, P =
pulses. Therefore, ratios were made of peak pressures 0.028; right foot, P = 0.001) in the pes planus group
and impulses of two different regions or of a certain compared with the normal group. This tendency was
region and the total foot. These ratios are presented in not apparent for peak pressures.
Table 1. The relative impulses of the midfoot region were
significantly (left foot, P = 0.038; right foot, P = 0.04)
Statistics
lower in the pes caws group compared with the nor-
Fisher’s exact test was performed to examine mal group in barefoot running. In running with sport
whether malalignment of feet resulted in significantly shoes, we noted the same tendency, though only sta-
more overuse injuries. tistically significant for the left foot (left foot, P = 0.02;
A one-way variance analysis with a between-sub- right foot, P = 0.11). The relative peak pressures under
jects design was used to test whether the resulting the midfoot also tended to be lower in the pes cavus
pressure ratios were significantly (P < 0.05) higher or group, in the barefoot condition (right foot, P = 0.04;
lower for a certain foot type compared with the other left foot, P = 0.08) as well as in the sport shoe con-
foot types. dition (left foot, P = 0.005; right foot, P = 0.07).
The device allowed us to measure at least three The relative loads under the second and third meta-
successive steps of each foot. Both a correlation test tarsal head were not affected by the foot type. The
and a variance analysis were used to examine whether relative load under the total forefoot tended to be
peak pressures and impulses varied significantly (P < higher in the pes cavus group compared with the
0.05) from step to step. planus group. For the left foot, this difference is sta-
Finally, we compared peak pressures and impulses tistically significant, for the impulses (P = 0.05) as well
of the left foot with those of the right foot, and peak as for the peak pressures (P = 0.04). The mediolateral

Downloaded from fai.sagepub.com at UNIV TORONTO on November 21, 2016

Foot &Ankle International/Vol. 16, No. 10/0ctober 1995 MALALIGNMENT & PLANTAR PRESSURE PA'TTERN 629
TABLE 2
Results of the Examination of Foot Architecturea
Foot arch Lower leg-heel Heel-forefoot Foot type
Subject
Left Right Left Right Left Right Left Right
1 A A 6 7 4 4 2 2
2 B B 6 7 2 1 2 2
3 A A 13 11 10 11 1 1
4 C C 10 12 7 9
5 C C 0 0 -3 -1 3 3
6 B B 8 7 3 2 2 2
7 A A 10 8 9 10 1 1
8 B B 4 5 5 7 2 2
9 A A 4 6 8 8 2 2
10 A A 7 4 2 0 2 2
11 A A 11 9 14 10 1 1
12 A A 9 6 7 8 2 2
13 B B 8 9 2 1 2 2
14 C C 7 7 1 3 3 3
15 A A 11 6 0 2 2 2
16 C B 6 4 3 3 2 2
17 B B 3 3 -2 -2 3 3
18 B B 7 7 6 7 2 2
19 B B 3 4 2 3 2 2
20 A A 11 10 10 9 1 1
21 B B 4 5 1 0 2 3
22 B B 7 -1 4 0 2 2
23 B B 7 6 3 5 2 2
24 C C 3 -2 0 0 3 3
a Measurements of height of the foot arch (A = low arch, B = normal arch, C = high arch), lower leg-heel angle, and heel-forefoot angle
were used to determine the foot type (type 1 = pes planus, type 2 = pes rectus, type 3 = pes cavus).

load distribution under the forefoot and the load dis- appeared to be clearly different (Table 7), for the left
tribution under the toes were not determined by the foot as well as for the right foot. Some ratios of bare-
foot type. foot running were significantly higher, while others
Statistical tests for repeated measurements showed were significantly lower.
that three successive steps were comparable in terms
of impulses, but only moderately comparable regard- DISCUSSION
ing peak pressures.
When the pressure ratios of the left foot were com- In our study, an unfavorable alignment of feet re-
pared with those of the right foot, we noted that they sulted in significantly more overuse injuries. This con-
were fairly comparable, both in barefoot running and firms our clinical experiences and the findings of many
in running with sport shoes (Table 6). When the pres- other authors.
sure ratios obtained in barefoot running were com- Correlation of pressure distribution data and foot
pared with those obtained in the shoe condition, they type produced some interesting findings. We could

TABLE 3
Frequency Table of the Injury Rate and Foot Type of the Left Foot and Right Foot"
Foot type 1 Foot type 2 Foot type 3 Total

Left Right Left Right Left Right Left Right

Class I injury rate
Observed frequency 1 1 12 11 0 1 13 13
Expected frequency 2.2609 2.2609 8.4783 7.913 2.2609 2.8261
Class II injury rate
Observed frequency 3 3 3 3 4 4 10 10
Expected frequency 1.7391 1.7391 6.521 7 6.087 1.7391 2.1 739
Total 4 4 15 14 4 5 23 23
a Fisher's exact test: left foot, P = 0.003; right foot, P = 0.02.

Downloaded from fai.sagepub.com at UNIV TORONTO on November 21, 2016

630 SNEYERS ET AL. Foot & Ankle International/Vol. 16, No. 10/0ctober 1995

CAVUS

..
Fig. 6. A three-dimensional represen-
tation of peak pressures and impulses of
a typical pes planus, pes cavus, and nor-
mal foot. &I, Impulse (0.1 kg sec/cm'):
peak pressure (1 kg/cm2).

u
RECTUS

not establish mediolateral load shifts under the heel or There was an anterior shift of the impulse under the
forefoot in pes planus subjects in this study. Clarke,4 heel in the pes planus group compared with the nor-
on the other hand, showed that in case of an overpro- mal group. This confirms the results of Cavanagh and
nation in stance, there was a medial shift of impulses R ~ d g e r s ,who
~ detected a shift of peak pressures
and, to a lesser extent, of peak pressures under the toward the anterior part of the heel in running and
heel as well. This discrepancy could be attributed to jumping in cases of pes planus.
the fact that Clarke examined the walking pattern, The relative load under the midfoot in this study
where the heel always strikes before the forefoot, appeared to be lower for pes cavus. These results
whereas we examined running subjects, comprising correspond with those of Cavanagh and R ~ d g e r s , ~
heel strikers as well as forefoot strikers. who found that in cases of pes cavus, the midfoot did

Downloaded from fai.sagepub.com at UNIV TORONTO on November 21, 2016

Foot & Ankle International/Vol. 16, No. 10/0ctober 1995 MALALIGNMENT & PLANTAR PRESSURE PATTERN 631

TABLE 4
Mean Peak Pressure and Impulse Ratios of the Different Foot Types in Bare Foot Running and Statistical Comparison
of These Ratios
- ~~~~~~~~

Mean ratio Comparisona Probability

Calculated ~~ ~~

pressure Foot type 1 Foot type 2 Foot type 3 1-2 1-3 2-3
-
ratios
~ ~~ ~~~~~ ~~

Left Right Left Right Left Right Left Right Left Right Left Right Left Right
~~ ~ ~~~~ .~_____ _ ~ ~~ ~ ____ ~~~~~~~

1 1.35 1.45 1.44 1.42 1.66 1.73 0 65 0.44

2 1.40 1.54 1.47 1.38 1.57 1.62 0.92 0 73
** 0.028 0.001
3 2.75 2.29 1.60 1.34 1.76 1.83
4 1.51 1.36 1.24 1.03 1.61 1.16 0.43 0.24
5 0.11 0.12 0.1 1 0.12 0.03 0.06 0038 0.04
6 0.15 0.17 0.16 0.18 0.07 0.09 * 0.08 o 04
7 0.36 0.33 0.37 0.35 0.39 0.42 0.71 0.049
8 0.34 0.31 0.35 0.33 0.37 0.40 0.62 0.059
9 0.56 0.59 0.62 0.63 0.68 0.69 0 05 0.22
10 0.56 0.61 0.64 0.64 0.73 0.74 0.04 0.18
11 0.69 0.90 0.87 0.97 0.88 0.91 0.38 0.88
12 0.73 0.92 0.88 0.99 0.91 0.90 0.45 0.83
13 0.42 0.46 0.49 0.42 0.38 0.32 0.49 0.44
14 0.45 0.45 0.49 0.43 0.39 0.32 0.68 0.38
~ ____~ ~- ~ ~ ~~ ~~ ~~~ ~~~~~~~

a *P < 0.05; **P < 0.01.

not contact the ground during running and jumping. pressures under the metatarsal heads measured by
They concluded that a pes cavus deforms little during Hennig‘ were twice as high in pes cavus subjects
these activities. Subjects with pes cavus have a higher compared with normal subjects during a downward
rate of injury of the lower limbs, and this could be jump. The mediolateral load distribution under the
attributed to the inadequate shock absorption during forefoot did not seem to be influenced by foot type.
locomotion. Although the alignment of the feet of one subject
The relative load under the forefoot in our study was not always quite the same, the plantar pressure
was, in agreement with other authors, higher in the distribution of the feet appeared to be comparable.
pes cavus group compared with the pes planus group. Another remarkable finding is that most of the im-
Scranton and McMasteri4 studied one subject with pulse and peak pressure ratios obtained in barefoot
p3s cavus and concluded that the metatarsal heads running differ significantly from the same ratios ob-
bore the bulk of the load during running. The peak tained in the shoe condition. In the barefoot condition,

TABLE 5
Mean Peak Pressure and Impulse Ratios for the Different Foot Types in Running with Sport Shoes and Statistical Comparison
of These Ratios
~~~ ~ ~~~~~~~

Mean ratio Comparisona Probability

Calculated ~ ~
~~~~~~~

pressure Foot type 1 Foot type 2 Foot type 3 1-2 1-3 2-3
~-
ratios
~~ ~ ~~ ~~~ ~~ ~~

Left Right Left Right Left Right Left Right Left Right Left Right Left Right
~~~ ~ - ~~~ ~~~~~~~

1 1.23 1.26 1.12 1.13 1.22 1.14 0.53 05

2 1.18 1.18 1.08 1.03 1.21 1 04 0.45 0.27
3 165 177 1.59 1.73 1 51 1.47 0.89 0.54
4 1.35 1.45 1.36 1.38 1.31 1.22 0.95 0.37
5 0.12 0.13 0.1 1 0.12 0.05 0.08 0.02 0.11
** **
6 0.18 0.18 0.16 0.17 0.08 0 12 0.005 0.07
7 0.36 0.35 0.34 0.32 038 033 0.23 0.63
a 0.34 034 0.32 0.31 036 0.32 0.25 0.61
9 0.56 0.52 0.58 0.59 0.60 0.62 075 0.10
10 0.64 0.58 0.61 061 0.60 066 0.77 0.40
11 1.09 0.97 0.96 1.09 1.05 1.18 0.55 0.56
12 1.13 1.0 0.96 1.09 1 07 1.21 0.30 0.45
13 0.54 0 52 0.55 0.47 0.61 0.48 0.82 0.83
14 0.53 0.53 0.55 0.49 0.61 0.50 ~~~~~~
0.75 0.89
~ ~~~

a *P < 0.05: ‘*P < 0 01.

Downloaded from fai.sagepub.com at UNIV TORONTO on November 21, 2016

632 SNEYERS ET AL. Foot & Ankle International/Vol. 16, No. 10/0ctober 1995
TABLE 6 types of structural foot variants (e.g., hallux valgus) on
Left Foot-Right Foot Difference of Peak Pressure and
the plantar pressure pattern.
Impulse Ratios for Barefoot Running and Running
with S ~ o rShoesa
t It has been reported use of biomechanical aids,
such as shoe correction or orthotic control, can com-
Ratio Barefoot pb With shoes P
bat several chronic overuse injuries of the lower
1 -0.032 0.56 -0.002 0.92 limbs.18 A quantitative assessment of the effect of
2 0.013 0.87 0.047 0.06
orthotic devices or different types of running shoes on
3 0.196 0.22 -0.142 0.1 1
4 0.175 0.OY -0.059 0.33 plantar pressure should be the subject of future re-
5 -0.010 0.09 -0.006 0.37 search.
6 -0.012 0.13 -0.009 0.30
7 0.007 0.39 0.019 0.0003** REFERENCES
8 0.003 0.78 0.015 0.002*
9 -0.014 0.20 -0.002 0.86 1. Alexander, I.J., Edrnund, Y.S., Chao, P.D., and Johnson,
10 -0.016 0.21 -0.001 0.93 K.A.: The assessment of dynamic foot to ground contact forces
11 -0.109 0.07 -0.086 0.12 and plantar pressure distribution: a review of the evolution of
12 -0.097 0.09 -0.078 0.13 current techniques and clinical applications. Foot Ankle, 11:
13 0.056 0.19 0.069 0.12 152-1 67, 1990.
14 0.061 0.19 0.050 0.22 2. Cavanagh, P.R., and Michiyoshi, A.: A technique for the dis-
play of pressure distributions beneath the foot. J. Biomech.,
a A paired t-test was used to examine whether peak pressure and
13:69-75, 1980.
impulse ratios of the left foot differed significantly from those of the
3. Cavanagh, P.R., and Rodgers, M.M.: Pressure distribution
right foot.
underneath the human foot. In Biomechanics: Current Interdis-
*P < 0.05: **P< 0.01. ciplinary Research. Dordrecht, Netherlands, Martinus Nijhoff
Publishers, 1985.
comparison of pressure ratios of the three foot types 4. Clarke, T.E.: The pressure distribution under the foot during
yields more statistically significant differences than in barefoot walking (doctoral dissertation). University Park, PA:
Pennsylvania State University Press, 1980, p. 129.
the shoe condition. Therefore, it is probably better to
5. Grieve, D.W., and Rashdi, T.: Pressure under normal feet in
examine the plantar pressure distribution in a barefoot standing and walking as measures by foil pedobarograph. Ann.
condition. We could also draw the conclusion that a Rheumatol. Dis., 43816-818, 1984.
simple running shoe at least partially absorbs the extra 6. Hennig, E.M.: Druckverteilungsmusterin Abhangigkeit von der
loads caused by malalignment of the foot. Fussstruktur bei muskular aktiven und passiven Belastungsfor-
men. Sportverletz. Sportschaden, 4:109-116, 1990.
A subdivision of all feet into three types is probably
7. Hutton, W.C., and Dhanendran, M.: A study of the distribution
too general. Each type is a compilation of different of load under normal foot during walking. Int. Orthop., 4:153-
anatomical variants and their combinations.' Larger 157, 1979.
subject populations, which have yet to be tested, 8. Hyvarinen, T., and Karhu, 0.:The Shoe System of the Future.
would allow assessment of the influence of different A Training Manual. Helsinki, Karhu-Titan Oy, 1988.
9. Jones, L.J., and Todd, W.F.: Abnormal biomechanics of flat-
foot deformities and related theories of biomechanical develop-
TABLE 7 ment. Clin. Podiatr. Med. Surg., 651 1-520, 1989.
Barefoot-Shoe Difference of Peak Pressure and Impulse 10. Lord, M.: Foot pressure measurement: a review of methodol-
Ratios for the Right and the Left FooP ogy. J. Biomed. Eng., 3:91-99, 1981.
Ratio Right P Left P 11. Lord, M.: Foot pressure measurement: A review of clinical
findings. J. Biomed. Eng., 8:91-99, 1986.
1 0.313 0.0002** 0.297 0.0007** 12. Miyazaki, S., and Iwakura, H.: Foot-force measuringdevice for
2 0.359 0.0012** 0.333 0.005**
clinical assessment of pathological gait. Med. Biol. Eng. Corn-
3 -0.071 0.57 0.185 0.16 put., 16429-436, 1978.
4 -0.227 0.009** -0.043 0.71
13. Rodgers, M.M.: Dynamic biomechanics of the normal foot and
5 0.001 0.86 -0.003 0.96
ankle during walking and running. Phys. Ther., 681822-1 830,
6 0.006 0.46 0.006 0.40 1988.
7 0.023 0.002** 0.015 0.06 14. Scranton, P.E.,and McMaster, J.H.: Momentary distribution
8 0.020 0.011* 0.010 0.19 of forces under the foot. J. Biomech., 9:45-48, 1976.
9 0.047 0.0002** 0.035 0.0006** 15. Soames, R.W.: Foot pressure patterns during gait. J. Biomed.
10 0.046 0.003** 0.026 0.04* Eng., 7:120-126, 1985.
11 -0.163 0.004** -0.188 0.0001**
16. Soarnes, R.W., Stott, J.R., Goodbody, A., Blake, C.D., and
12 -0.150 0.005** -0.170 0.0001**
Brewerton, D.A.: Measurement of pressure under the foot
13 -0.058 0.09 -0.078 0.02* during function. Med. Biol. Eng. Comput., 20489-495, 1982.
14 -0.079 0.02* -0.074 0.03*
17. Subotnick, S.I.: The biomechanics of running. Implications for
a A paired t-test was used to examine whether peak pressure and the prevention of foot injury. Sports Med., 2144-153, 1985.
impulse ratios obtained in barefoot running differed significantly 18. Viitasalo, J.T., and Kvist, M.: Some biomechanical aspects of
from those of running with sport shoes. the foot and ankle in athletes with and without shin splints. Am.
*P < 0.05; **P< 0.01. J. Sports Med., 11:125-130, 1983.

Downloaded from fai.sagepub.com at UNIV TORONTO on November 21, 2016