Effects of laterally wedged insoles on knee and subtalar joint moments.

Abstract

To assess the biomechanic effects of wearing a lateral wedge on the knee joint varus moment during gait in elders with and without knee osteoarthritis (OA).
Crossover design whereby subjects walked under 2 different insole conditions: a 0 degrees control wedge and a 6 degrees lateral wedge.
A gait laboratory with 3-dimensional motion analysis and force platform equipment.
Thirteen healthy subjects and 13 knee patients with OA.
Not applicable.
Frontal plane angles and moments at the knee and subtalar joints, ground reaction forces, and center of pressure. Moments were derived by using a 3-dimensional inverse dynamics model of the lower extremity.
The 6 degrees lateral wedge significantly reduced knee joint varus moment and increased subtalar joint valgus moment in both groups when compared with no wedge. All patients had a greater knee joint varus moment with a similar subtalar joint valgus moment compared with the people without OA. There were diverse, sometimes reversed effects with the insole among the patients.
The 6 degrees lateral wedge did not consistently reduce the knee joint varus moment in patients with knee OA. The biomechanic indications and limitations of laterally wedged insoles should be confirmed by a larger study.

Full text

Page 1

ORIGINAL ARTICLE
Effects of Laterally Wedged Insoles on Knee and Subtalar
Joint Moments
Wataru Kakihana, PhD, Masami Akai, MD, Kimitaka Nakazawa, PhD, Takamichi Takashima, PO, PhD,
Kenji Naito, MS, Suguru Torii, MD
ABSTRACT.
Takashima T, Naito K, Torii S. Effects of laterally wedged
insoles on knee and subtalar joint moments. Arch Phys Med
Rehabil 2005;86:1465-71.
KakihanaW, AkaiM,Nakazawa K,
Objective: To assess the biomechanic effects of wearing a
lateral wedge on the knee joint varus moment during gait in
elders with and without knee osteoarthritis (OA).
Design: Crossover design whereby subjects walked under 2
different insole conditions: a 0° control wedge and a 6° lateral
wedge.
Setting: A gait laboratory with 3-dimensional motion anal-
ysis and force platform equipment.
Participants: Thirteen healthy subjects and 13 knee patients
with OA.
Interventions: Not applicable.
Main Outcome Measures: Frontal plane angles and mo-
ments at the knee and subtalar joints, ground reaction forces,
and center of pressure. Moments were derived by using a
3-dimensional inverse dynamics model of the lower extremity.
Results: The 6° lateral wedge significantly reduced knee
joint varus moment and increased subtalar joint valgus moment
in both groups when compared with no wedge. All patients had
a greater knee joint varus moment with a similar subtalar joint
valgus moment compared with the people without OA. There
were diverse, sometimes reversed effects with the insole
among the patients.
Conclusions: The 6° lateral wedge did not consistently
reduce the knee joint varus moment in patients with knee OA.
The biomechanic indications and limitations of laterally
wedged insoles should be confirmed by a larger study.
Key Words: Biomechanics; Gait; Rehabilitation; Shoes.
© 2005 by the American Congress of Rehabilitation Medi-
cine and the American Academy of Physical Medicine and
Rehabilitation
T
deformity of the knee was first reported in 1987.1,2Since then,
kinematic and kinetic analyses of this condition have mainly
focused on a static standing position.1-3Yasuda and Sasaki2
and Wolfe and Brueckmann3reported that wearing a laterally
wedged insole reduced the load in the medial compartment of
HE EFFECT OF WEARING a laterally wedged insole to
reduce symptoms in osteoarthritic patients with a varus
the knee, which was effective for the treatment of knee pain in
patients with osteoarthritis (OA). Since the early 1990s, the
evidence relating to the clinical efficacy of wearing a laterally
wedged insole as a treatment for patients with OA has also
been reported in dynamic conditions,4-7but these studies did
not answer the question of what mechanisms reduced symp-
toms in OA patients. The effect of wearing a laterally wedged
insole has not been sufficiently studied.
Recently, Crenshaw8and Kerrigan9and colleagues reported
that wearing a laterally wedged insole reduced the knee joint
varus moment, suggesting an effective mechanism for knee
pain reduction in patients with OA. However, Maly et al10
found that wearing a laterally wedged insole did not signifi-
cantly reduce the knee joint varus moment during gait. These
studies suggested that the beneficial effects of wearing a later-
ally wedged insole were not consistent for OA patients, and
hence we were interested in the biomechanic details that af-
fected the knee joint varus moment during gait with a laterally
wedged insole.
Keating et al5suggested that the knee joint varus moment
during gait with a laterally wedged insole was associated with
the subtalar joint valgus angle. Our previous study11found that
healthy young adults wearing a laterally wedged insole had
changes in knee joint varus moment and subtalar joint valgus
moment during gait via the more laterally shifted location of
the center of pressure (COP). Our current study focused on
how the knee joint varus moment in patients with OA was
associated with the angle and moment at the subtalar joint
during gait while wearing a laterally wedged insole. In partic-
ular, our objective was to assess frontal plane kinematic and
kinetic effects of wearing a 6° wedged insole on the knee joint
moments. We hypothesized that the same knee joint varus
moment reduction obtained with a 6° laterally wedged insole
during gait in healthy elders would also be present in age-
matched patients with OA.
METHODS
Participants
After informed consent was obtained, 26 elderly women (13
healthy elders, 13 osteoarthritic patients with a varus deformity
of the knee) participated in the experiments. There were no
statistically significant differences in age, height, and weight
(table 1). OA patients were recruited through public advertise-
ments in local community programs about knee OA in the area
around the National Rehabilitation Center. The inclusion cri-
teria for OA patients were: 50 years of age or older, knee pain
for at least 6 months, and bilateral knee OA with a definite
osteophyte indicated by radiograph. In addition to these inclu-
sion criteria, the subjects must have met one of the following
conditions: morning (initial) stiffness for less than 30 minutes,
or crepitus on active motion of the knee (ie, with weight-
bearing, eg, squatting).12Patients with OA were excluded if
they were currently using a wedged insole or other custom
orthotics in their shoes; had had surgery on the lower extrem-
From the Department of Rehabilitation for Movement Functions, Research Institute
of National Rehabilitation Center for Persons with Disabilities, Tokorozawa (Kaki-
hana, Akai, Nakazawa, Takashima); and Graduate School of Human Sciences,
Waseda University, Tokorozawa (Naito, Torii), Japan.
No commercial party having a direct financial interest in the results of the research
supporting this article has or will confer a benefit upon the author(s) or upon any
organization with which the author(s) is/are associated.
Reprint requests to Wataru Kakihana, PhD, Dept of Rehabilitation for Movement
Functions, Research Institute of National Rehabilitation Center for Persons with
Disabilities, 4-1, Namiki, Tokorozawa 359-8555, Japan, e-mail: kaki@rehab.go.jp.
0003-9993/05/8607-9393$30.00/0
doi:10.1016/j.apmr.2004.09.033
1465
Arch Phys Med Rehabil Vol 86, July 2005

Page 2

ity; had concurrent clinically active arthritis of the hip, ankle,
hindfoot, or midfoot; had unilateral knee symptoms; or were
unable to walk without a cane or walker. Healthy, age-matched
elders were also recruited through public advertisements
among the local community-dwelling elders.
Radiologic Assessment
The anteroposterior radiographs of both entire legs were
acquired with subjects standing barefoot and with knees in full
extension. If a subject had a flexion contracture of the knee
joint, the radiographic beam was projected along the joint
space. Based on these weight-bearing radiographs, the femo-
rotibial angles of the less affected legs in healthy elders (all
right legs) and the more affected legs in OA patients (6 right
legs, 7 left legs) were calculated as the static alignment of the
lower extremity. The femorotibial angle was defined as the
lateral angle between the femoral shaft and the tibial shaft. The
radiographic alignment was measured by a trained investigator.
Data Acquisition System
Three-dimensional gait analyses were conducted with a mo-
tion analysis system (Vicon 512 System)aoperating at 60Hz
with 12 infrared cameras and 8 force platforms (9281C)b
operating at 60Hz. Force platforms were situated at the mid-
point of a 7m walkway. Before data collection, each infrared
camera was calibrated, and it was confirmed that the average
spatial resolution was less than 2mm over a measured volume
of 2.4?0.8?1.0m. Infrared cameras were used to measure the
3-dimensional positions of infrared reflecting markers that
were attached to each subject’s skin over bony landmarks (see
Experimental Protocol below). The position of each marker
was calculated by a dynamic linear transformation method in a
fixed laboratory coordinate system.13The vertical and medio-
lateral (ML) components of the ground reaction force and the
location of the COP during the stance phase of gait were also
obtained by using the force platforms.
Laterally Wedged Insoles
Two different laterally wedged insoles were tested: a control
wedge with a 0° lateral angle (N) and a wedge with a 6° lateral
angle (W). The control wedge was placed under the entire foot
and had an even thickness from medial to lateral as well as
from hindfoot to forefoot. The lateral wedges ran along the full
length of the insole from the hindfoot to the forefoot. They had
a coefficient of elasticity of 100 to 300kg/mm2and were made
of ethylene vinyl acetate (EVA 8200)c. Insoles were supplied
in several sizes to accommodate the differences in subjects’
foot size. We needed to attach the infrared reflecting markers to
each subject’s skin over the bony landmarks of the feet. For
this reason, the insoles were attached to subjects’ bare feet
(around the hindfoot, midtarsal, second, or big toe) by using
adhesive tapes after ethanol swab.
Experimental Protocol
Before the trials, reflective markers (diameter, 20mm) were
placed over bony landmarks to minimize skin movement—the
greater trochanter, lateral and medial femoral condyles, antero-
lateral and anteromedial tibial condyles, lateral and medial
malleoli, lateral and medial calcaneal tubercles, head of talus,
and 1st and 5th metatarsal heads—to define the thigh, shank,
and foot segments.11After the markers were attached, subjects
were instructed to stand barefoot for 5 seconds to establish the
relationship among the markers for the subject’s initial ana-
tomic position.14,15In this barefoot standing trial, they were
instructed to stand with knees as fully extended as possible, the
ankles at 0° dorsiflexion, and a comfortable degree of toeing-
out.
For the walking trials, insoles were attached to the subjects’
feet and they were asked to walk at a walking speed of 95
steps/min, indicated by a metronome, to match the cadence of
the patients with OA. This threshold speed corresponded to the
self-selected walking speed of the OA patients (95?4 steps/
min). The insoles with the N and W wedges were randomly
assigned to subjects and no attempt was made to control stride
length, stride width, or foot progression angle during gait.
Data Analysis
A 3-segment rigid link model was used to describe the
motion of the lower extremity in the frontal plane (fig 1).11We
used local reference systems to obtain the angles and moments
at the knee and subtalar joints with respect to the actual rotation
axes of the joints.16,17The rotation of the knee joint was
defined as the rotation of the tibial condyle relative to the
femoral condyle.16The rotation of the subtalar joint was de-
fined as the rotation of the calcaneus relative to the lateral-
medial malleoli.17The axis of rotation of the knee joint was
midway between the markers on the lateral and medial femoral
condyles and the anterolateral and anteromedial tibial con-
dyles. The axis of rotation of the subtalar joint was defined by
using the markers on the lateral calcaneal tubercle and head of
the talus. The moments acting about each joint were calculated
using an inverse dynamics algorithm and expressed as external
Fig 1. The models and coordinate systems of the subtalar joint and
knee joint are shown. Abbreviations: ?, angle; GRF, ground reaction
force; M, moment; W, segment weight. Adapted with permission
from Kakihana et al.11Copyright by Am J Phys Med Rehabil.
Table 1: Anthropometric Data of the Subjects
Data
Healthy Elders
(n?13)
OA Patients
(n?13)P
Age (y)
Height (cm)
Weight (kg)
Femorotibial angle (deg)
64.6?2.3
150.0?3.0
54.1?7.2
172.9?3.7
63.3?5.6
152.7?5.6
57.3?10.3
177.5?3.9
.424
.183
.376
.007
NOTE. Values are mean ? standard deviation. Boldface denotes
significance.
1466
BIOMECHANIC EFFECTS OF WEDGED INSOLES, Kakihana
Arch Phys Med Rehabil Vol 86, July 2005

Page 3

moments. The valgus moment arm of the subtalar joint was
defined as the perpendicular distance from the line of action of
the ground reaction force to the subtalar joint axis in the
subtalar joint coordinate system. Joint moments as well as
vertical and ML ground reaction forces were normalized for
body weight and reported in newton meters per kilograms.
Stance-phase data of the control group were normalized to
100% for their less affected right legs only. The stance-phase
data for the patients with OA were normalized to 100% for
their more affected right or left legs only. The data were
considered to be valid when the markers were detected, that is,
when there was no fragmentation of the marker trajectories.
The mean and standard deviations were obtained from 10 trials
with each insole.
Statistical Analysis
The average moments at the knee and subtalar joints were
equivalent to the division of the area under the moment-time
curve (fig 2) and its time of application, respectively. Evalua-
tions of the average angles at the knee and subtalar joints, ML
Fig 2. Time percentage of
stance phase for the moments
at the knee and subtalar joints,
angles at the knee and subtalar
joints, and the vertical and ML
components of the ground re-
action forces when a subject
walked with a 0° lateral wedge
(N) versus when a subject
walked with a 6° lateral wedge
(W).
1467
BIOMECHANIC EFFECTS OF WEDGED INSOLES, Kakihana
Arch Phys Med Rehabil Vol 86, July 2005

Page 4

and vertical ground reaction force, and valgus moment arm of
the subtalar joint during stance phase were performed by using
the same methods as for those applied for the average moments
at the knee and subtalar joints. A comparison of each average
parameter was performed by using a 2 (groups: healthy elders,
OA patients) by 2 (conditions: insole N, insole W) analysis of
variance with repeated measures on the last factor. Statistical
evaluations were performed with the SPSS software program,
version 11,dwith significance defined as P less than .05. In
addition, simple regression analysis with Pearson coefficients
was used to determine correlation between the femorotibial
angle and the knee joint varus moment during walking.
RESULTS
The effects of wearing lateral wedged insoles on the knee
and subtalar joint moments during stance phase were evident in
each insole condition in both the healthy elders and the OA
patients (fig 2, table 2). The knee joint varus moment was
significantly smaller for insole W compared with insole N (a
10.4% reduction for the healthy elders, a 5.6% reduction for the
OA patients; P?.001), whereas the subtalar joint valgus mo-
ment was significantly greater for insole W compared with
insole N (a 27.7% increase for healthy elders, a 23.5% increase
for OA patients; P?.001).
The angles at the knee and subtalar joints did not show any
changes between the 2 insole conditions. In addition, no obvi-
ous differences in the insole conditions were found for the
vertical ground reaction forces during the stance phase; how-
ever, the ML ground reaction forces during the stance phase for
insole W were significantly greater compared with those for
insole N (a 1.8% increase for healthy elders, a 3.7% increase
for OA patients; P?.035). The most remarkable difference
between the insole conditions was the shift of the COP loca-
tion: the COP location was always parallel to the subtalar joint
axis, but it shifted more laterally when wearing insole W (fig
3A). Consistent with this difference, the valgus moment arm of
the subtalar joint for insole W was significantly greater com-
pared with insole N (a 30.0% increase for healthy elders, a
13.0% increase for OA patients; P?.001) and hence produced
a significantly smaller knee joint varus moment than insole N.
Figure 4 represents the change between the subtalar joint
valgus moment and the knee joint varus moment from insole
condition N to W of each subject. There was a reversed
correlation between the subtalar joint valgus moment and the
knee joint varus moment in all healthy elders (13/13) and in the
majority of the OA patients (11/13). But in 2 OA patients, both
the knee joint varus moment and the subtalar joint valgus
moment were greater for insole W compared with insole N. In
addition, the same 2 subjects showed a medially shifted COP
trajectory when wearing insole W, contrary to all the other OA
patients (fig 3B).
The differences in femorotibial angle between the healthy el-
ders and OA patients (172.9°?3.7° vs 177.5°?3.9°, P?.007)
were accompanied by differences in the varus angle and moment
at the knee joint. Patients with OA had greater knee joint varus
moments than the healthy elders (a 24.1% increase for insole N, a
30.7% increase for insole W; P?.005). There was a stronger
correlation between the femorotibial angle and the knee joint
varus moment with insole N in the healthy elders (r2?.52,
P?.070) than in the OA patients (r2?.30, P?.325), but this
correlation was not significant for either group (fig 5).
DISCUSSION
Our study examined the kinematic and kinetic factors of a 6°
lateral wedge insole (insole W) on the knee joint varus moment
in healthy elders and in OA patients. The knee joint varus
moment was significantly smaller, whereas the subtalar joint
valgus moment was significantly greater with insole W com-
pared with insole N (0° wedge) in both the healthy elders and
the OA patients. With insole W, this finding correlated with a
more lateral shift in the location of the COP during the stance
phase. This could explain the reduced knee joint varus moment
obtained by the laterally wedged insole. In contrast, 2 OA
patients had an increased knee joint varus moment while wear-
ing insole W compared with insole N. This was reflected in a
medial shift in COP location. Despite the reduction in the varus
knee angle and moment when wearing insole W, all values for
the OA patients still differed significantly compared with the
age-matched control group.
This study’s results support in part the hypothesis that the
same reduction in the knee joint varus moment obtained with
insole W in healthy elders would also be present in age-
matched OA patients. Compared with insole N, insole W
significantly reduced the knee joint varus moment regardless of
group assignment (healthy elders or OA patients). There was a
10.4% reduction for the healthy elders and a 5.6% reduction for
Table 2: Stance Phase Average Values Between 0° Lateral Wedge (N) and 6° Lateral Wedge (W)
Variable
Healthy Elders (n?13) OA Patients (n?13)P Value
NWNW123
Moment (Nm/kg)
Knee joint varus
Subtalar joint valgus
Angle (deg)
Knee joint varus
Knee joint valgus
Subtalar joint varus
Subtalar joint valgus
Ground reaction force (N/kg)
ML
Vertical
Distance (mm)
Subtalar joint moment arm
0.29?0.01
?0.18?0.01
0.26?0.02
?0.23?0.01
0.36?0.02
?0.17?0.01
0.34?0.02
?0.21?0.01
?.001
?.001
.005
.548
.325
.331
0.27?0.18
?6.61?1.47
3.38?0.99
?1.02?0.37
0.35?0.18
?6.46?1.49
1.99?0.56
?1.46?0.34
2.23?0.84
?2.68?1.09
1.75?0.38
?1.35?0.40
2.18?0.83
?2.88?1.17
1.96?0.47
?2.14?0.71
.094
.915
.142
.070
.008
.039
.324
.411
.702
.408
.051
.613
0.53?0.03
7.52?0.11
0.54?0.02
7.49?0.10
0.53?0.02
7.58?0.16
0.55?0.01
7.62?0.09
.035
.942
.700
.777
.538
.426
20.72?1.8226.80?1.5222.58?2.2426.08?1.95
?.001
.827.084
NOTE. Values of columns N and W are mean ? standard error. Boldface denotes significance.
Abbreviations: 1, difference between insole conditons; 2, difference between subject groups; 3, interaction between insole conditions and
subject groups.
1468
BIOMECHANIC EFFECTS OF WEDGED INSOLES, Kakihana
Arch Phys Med Rehabil Vol 86, July 2005

Page 5

the OA patients in the knee joint varus moment. These reduc-
tions were associated with a significant increase in the subtalar
joint valgus moment, which to our knowledge have not been
published elsewhere. There was a 27.7% increase for the
healthy elders and a 23.5% increase for the OA patients in the
subtalar joint valgus moment with insole W.
There have been few reports on the effects of wearing a
laterally wedged insole on the angle and moment at the subtalar
joint. Yasuda and Sasaki2found that, in a static standing
position, the laterally wedged insole increased the valgus angle
of the subtalar joint, and therefore reduced the load in the
medial compartment of the knee joint. Keating et al5suggested
that during walking an increased valgus angle of the subtalar
joint obtained by a laterally wedged insole might be effective
for the reduction of knee pain in OA patients. In our study, the
subtalar joint valgus angle did not increase significantly with
the laterally wedged insole. It was assumed that the effect of
wearing the laterally wedged insole on the subtalar joint angle
during gait was not systematic among the subjects, in particular
among OA patients. In contrast, our previous study11and this
one indicate that wearing a laterally wedged insole had the
same effect on the increased subtalar joint valgus moment
during gait in people with and without knee OA. We found that
wearing a laterally wedged insole significantly increased the
valgus moment arm of the subtalar joint, creating a lateral shift
in COP location. It is assumed that this lateral shift in the
location of the COP decreases the length of the knee joint
moment arm resulting in a reduction of the knee joint varus
moment.
This assumption was not completely met. It is generally
known that wearing a laterally wedged insole reduces knee
joint varus moment in the early stage of knee OA. Keating5
Fig 3. The COP trajectory par-
allel to the subtalar joint axis.
(A) All healthy elders and 11 of
the 13 OA patients had a simi-
lar pattern, that is, a more lat-
eral shift in the location of the
COP with insole W. (B) In con-
trast, the 2 OA patients who
had an increased knee joint va-
rus moment when using insole
W; they had a more medial
shift in the location of their
COP. Legend: ?, data for the
insolewiththe0°lateralwedge
(N);
, data for the insole with
the 6° lateral wedge (W).
Fig 4. Relationship between
the subtalar joint valgus mo-
ment and the knee joint varus
moment when
walked with insoles with no
wedge (N, Œ) and a wedge
with a 6° lateral angle (W, ?).
subjects
1469
BIOMECHANIC EFFECTS OF WEDGED INSOLES, Kakihana
Arch Phys Med Rehabil Vol 86, July 2005