Copyright
Corrective
Flexible
BY
DEAN
DENNIS
R.
MORGAN,
1989
of Bone
WENGER,
M.D.t,
DALLAS,
From
the Division
SAN
DIEGO,
TEXAS,
AND
for three
flexible
flatfoot
years
Joint
Surgery,
the Texas
of Orthopedics,
DONALD
Scottish
University
does not influence
the course
Rite
as Treatment
for
and Children*
L.
LIEBER,
Hospital,
that the condition
should
be vigorously
treated
during
early
childhood57.
Flat feet are often detected
by a child’s
parents,
who
may be concerned
because
they know
of adults
for whom
the condition
is painful.
The concern
is a common
reason
for orthopaedic
ment because
consultation.
Many
the concept
of corrective
established
in our culture.
rective
shoes
in childhood,
*
No benefits
in any
form
parents
shoes
request
has been
treatfirmly
Many
of the parents
wore corand they may assume
that the
have
been
received
or will
be received
from
a commercial
party related
directly
or indirectly
to the subject
ofthis
article.
Funds
were received
in total or partial
support
of the research
or clinical
study presented
in this article.
The funding
sources
were grants
from the
Prescription
Footwear
Association
and the Research
Fund
of the Texas
Scottish
Rite Hospital.
t Department
ofOrthopedic
Surgery,
Children’s
Hospital,
8001 Frost
Street,
San Diego,
California
92123.
Please
address
requests
for reprints
to Dr. Wenger.
1:Texas Scottish Rite Hospital,
Dallas,
Texas
75219.
800
SPECK,
M.D4,
SAN
DIEGO,
CALIFORNIA
Diego.
shoes
were
responsible
have
noted.
Despite
concerted
San
Diego
for any
improvement
attention
that
to the subject,
they
may
no scientific
study
shoes
has established
whether
corrective
shoes or inserts
in
affect the course
of flexible
flat feet. Previous
studies
either
have
been
retrospective
spective,
have
not
treatment
and
control
or,
included
when
described
randomization
groups
or the
of
use
as propatients
to
of matched
con-
trols23.
In 1977
whether
,
we conducted
treatment
with
a prospective
corrective
study
shoes
to determine
and inserts
changes
the course of flexible
flatfoot
in children.
We enrolled
forty
patients,
but we were forced
to abandon
the trial because
the parents
of the patients,
who had been selected
from a
clinical
population,
and the compliance
were also criticized
tion
because
had a poor understanding
of the study,
with treatment
was unsatisfactory.
We
by the Prescription
Footwear
Associa-
the shoes
were
not
fitted
by certified
pedorth-
ists.
In
1978,
we
redesigned
the
cruited
To minimize
only
referred
by
This
termine
of flexible
problems
who
children
private
with
had typical
randomized
Materials
the
the
for fitting
compliance,
flat feet
trial
whether
corrective
shoes
flatfoot
in children.
for
incorporating
Association
we re-
and who
were
pediatricians.
prospective
Eligibility
study,
Footwear
of the Prescription
the shoes.
a disorder
in children
an anatomical
variant,
may
not need
treatthat flatfoot
is a prefoot in adulthood
and
GAIL
at San
expertise
Flexible
flatfoot
is so common
that it has been thought
by some to be
related
to ligamentous
laxity,
that
ment”7’#{176}”’3. Others
have believed
cursor
to a painful
problem
with the
M.D4,
PH.D.t,
Dallas.
of California
of
in children.
Incorporated
MAULDIN,
RICHARD
ABSTRACT:
We performed
a prospective
study to
determine
whether
flexible
flatfoot
in children
can be
influenced
by treatment.
One hundred
and twenty-nine
children
who had been referred
by pediatricians,
and
for whom the radiographic
findings
met the criteria for
flatfoot,
were randomly
assigned
to one of four groups:
Group I, controls;
Group II, treatment
with corrective
orthopaedic
shoes; Group III, treatment
with a Helfet
heel-cup;
or Group IV, treatment
with a custom-molded
plastic insert. All of the patients
in Groups
II, III, and
Iv had a minimum
of three years of treatment,
and
ninety-eight
patients whose compliance
with the protocol
was documented
completed
the study.
Analysis
of radiographs
before treatment
and at the most recent follow-up
demonstrated
a significant
improvement
in all
groups
(p < 0.01),
including
the controls,
and no significant difference
between
the controls
and the treated
patients
(p > 0.4).
We concluded
that wearing
corrective
shoes or in-
serts
and
Shoes and Inserts
Flatfoot
in Infants
C.PED4,
and
by The Journal
was
designed
or inserts
affect
to de-
the course
and Methods
Study
One to six-year-old
children
who had flexible
flat feet,
as diagnosed
by their pediatricians,
were
referred
to the
Texas Scottish
Rite Hospital
Flatfoot
Clinic,
Dallas,
which
was organized
for the sole purpose
of conducting
this study.
One
hundred
and
twenty-nine
children
met
the
criteria
for
entry
and were enrolled
in the study.
The children
had to
be less than six years old at the time of enrollment
(Fig.
1), but if they passed
their sixth birthday
during
the period
of follow-up
they still
Enrollment
began
remained
in the study.
late in 1978 and was completed
in 1980,
children
when
enough
a large enough
sample
three-year
follow-up
until
early
1984.
had
been
for each group.
for all patients,
A move
THE
by the
JOURNAL
late
to ensure
To allow a minimum
the study
continued
senior
OF
enrolled
BONE
one
AND
of us (D.
JOINT
SURGERY
R.
CORRECTIVE
SHOES
AND
INSERTS
AS TREATMENT
FOR
FLEXIBLE
801
FLATFOOT
‘I)
C
0
20
..0
C,)
.0
0
I,.-
0
10
.
E
0
12-32
months
Ftc.
Bar graph
depicting
the age
distribution
in each
treatment
group
the
presentation
at the time
32-52
52-72
months
months
I
of entry
into
the study.
UCBL
= University
of California
Biomechanics
Laboratory.
to another
W.)
institution
delayed
of this
most
The
private
about
formed
asked
these
to refer
patients
was
of concern
ceptive
assigned,
was
in our
prospective
children
who
were referred
to the
disregarded.
thirty-one
parents,
the
(I)
C
0
families
were
trial
had flexible
flat
for the specific
during
had to be dropped
in-
and
were
were
very
from
and
a few
examined
because
their
to confirm
that
The diagnosis
was based
on a
and poor formation
of the arch,
as observed
by the referring
physician
and by us. The patient
was first observed
from the rear while
weight-bearing,
to
re-
that we
schedule
the six-year
of non-compliance
they had flexible
flat feet.
valgus
position
of the heel
feet. Because
problem
that
the specific
treatment
study,
the follow-up
However,
patients
community
randomized
in terms of following
and, early
in the
rarely
period
pediatricians
our
because
families
moved
from the state.
Initially,
131 children
were
report.
confirm
that the heel was in a valgus
child was asked to stand on tiptoe,
and
study
heel
the study,
shifted
into
the subtalar
joint.
a varus
position,
position.
in every
confirming
Then
patient
the
the
flexibility
of
30
4-
0
0
C,)
.0
0
20
0
I.-
0
.0
E
1
Low Laxity
High
FIG.
Bar graph
high
laxity
California
VOL.
71-A,
depicting
the number
of patients
(2 to 5, according
to the system
Biomechanics
NO.
6. JULY
Laboratory.
1989
who had low
of Wynne-Davies)
laxity
Laxity
2
(0 or 1 , according
to the system
of Wynne-Davies)
in each treatment
group
at the time of entry
into
the
compared
with
study.
UCBL
those
who
University
had
of
802
D.
R.
WENGER
FIG.
Figs. 3-A and 3-B: A typical
patient.
Fig. 3-A: Initial
lateral
radlograph
of the foot. made
foot (46 degrees)
and the talus and the first metatarsal
with the patient
(24 degrees).
ET
3-A
standing.
Tightness
of the heel cord was documented
by measuring the degree
of passive
dorsiflexion
of the ankle with
the talonavicular
joint
locked
in inversion.
Laxity
was
scored using a scale of 1 to 5 , based on the criteria
of WynneDavies
(Fig.
documented,
2). In addition,
including
the
calf
Also,
study,
or foot.
we measured
when
several
parental
subjective
reports
the patients
were
the foot-progression
factors
were
of pain in the
entered
angle
into
the
(to document
FIG.
Initial
anteroposterior
foot) and 36 degrees
radiograph
(left foot).
of the
feet.
made
with
the patient
AL.
standing.
demonstrating
the
lateral
the degree
of toeing-in
varum or genu valgum.
The
a person
in relation
angles
between
or toeing-out)
the
talus
and
the
and the degree
sole
of the
of genu
foot-progression
angle is defined
as follows.
For
walking
in a straight
line, the angle of each foot
to that line is measured
in degrees.
If the person
is toeing-in,
the value in degrees
is stated
example,
25 degrees).
If the person
value is stated as a positive
(for example,
-
as a negative
(for
is toeing-out,
the
+ 25 degrees).
3-B
demonstrating
anteroposterior
THE
talocalcaneal
JOURNAL
angles
OF BONE
of 40 degrees
AND
JOINT
(right
SURGERY
CORRECTIVE
SHOES
AND
INSERTS
AS
TREATMENT
TABLE
CHARACTERISTICS
AT
OF NINETY-EIGHT
PATIENTS
THE
TIME
FOR
WHO
ENTRY
INTO
of patients
Mean
length
follow-upt
(mos.)
98
of
36.8
37.0
±
THE
TRIAL
STUDY
II
Group
(Corrective
Shoe)
21 (21%)
5.04
±
THE
COMPLETED
Group
No.
803
FLATFOOT
I
OF
Group
I
(Control)
Total
FLEXIBLE
28 (29%)
36.5
2.4
III
(Helfet
Heel-Cup)
±
Group
IV
(UCBL*)
27 (28%)
37.0
1.4
±
22 (22%)
3.1
36.8
±
1.7
Sex
Male
Female
60 (61%)
38 (39%)
Mean
age at
entryt
(mos.)
UCBL
*
= University
1- Mean
and
Any
condition
29.3
of California
standard
that was known
be associated
joints
(Down
or Ehlers-Danlos
prior treatment
with corrective
entry
into
Special
Biomechanics
±
Laboratory
32.2
11.6
the
to have,
disease)
a neurological
or a syndrome
with excessive
laxity of the
syndrome),
or who had had
footwear
or inserts,
was ex-
were
study.
initially
at the
Time
of Entry
and lateral
standing.
radiographs
Every child
ofthe feet were
met one of two
as the talohorizontal
angl&3.)
is 26.5
± 5.3 degrees2.
2. An angle
between
of more than
The
normal
the talus
10 degrees.
value
and the first
In a normal
or naviculocuneiform
metatarsal
foot, this
talocalcaneal
angle
(Fig.
has
3-B)
due
to poor
in the study,
the strict
for three
standards
was
not always
double
steel
shank.
The subjects
is wearing
were
custom-designed
surface
plantar
photographed
mirror-table,
sessment
of the front,
during
weight-bearing.
of the
foot
Assignment
to Groups
allow
leather
shoes
were
had no corrective
in contour
(not
that
normal
-
shoes
The patients
shoes,
a
was
and
that
that
are
has
standing
a
on a
simultaneous
data
after
with
According
as-
long
VOL.
71-A,
to a protocol
NO.
6. JULY
review
1989
that
for subsequent
the
approved
patients
by the
were
of patients.
feet)
(thirty-nine
is, so-called
features
straight-last)
built
wore
orthopaedic
in. The shoes
and had a steel
in the other three groups
the following
heel-cup8
medial
as-
exceptions.
The
in shoes
that
The
twenty-two
counter.
wore identical
twenty-eight
pa-
had
shoes that
Laboratory
a Thomas
patients
heel
and
a
(thirty-eight
had a University
custom-molded
of Calplastic
Interim
All patients
in the Flatfoot
research
nurse,
treatment.
had been
board,
fol-
patients,
Follow-up
wear
institutional
were
I). In sixteen
tients
(fifty-four
feet) in Group
II wore shoes
that had a
Thomas
heel, a long medial
counter,
and a navicular
pad5.
The twenty-seven
patients
(forty-nine
feet) in Group
III wore
dorthist.
hospital’s
and
of a
pa-
criteria
for entry into the
were included
in the anal-
that
-
relo-
shank.
rear views of the foot
a clear view of the
and provided
before
shoe
while
to
top, side, and
This allowed
of the foot
comparison
radiographs
a corrective
(Table
controls
standard-last
Laboratory
from the
patients),
the number
I, twenty-one
assigned
or development
Thus,
ninety-eight
for compliance
or more
feet) in Group
IV wore
ifornia
Biomechanics
insert9.
anteroposterior
a child
(twenty-five
met
years
were
with corrective
and thirty-one,
Biomechanics
later dropped
lowed
preting
measuring
were
tients
a Helfet
and
thirty-one
were treated
heel-cups8;
compliance
also recorded
initially,
but it was not used as a criterion
at
the time of interim
follow-up
because
of difficulty
in interwhile
10.7
±
III, and IV) by a nurse
Of the 129 patients
who
of the state (five patients),
disorder
(one patient).
In Group
in children
joint
enrolled
II,
cation
out
neurological
ysis
angle is 0 degrees
or slightly
negative5”3.
A cavus foot has
a markedly
negative
value.
A flatfoot
that is associated
with
a sag at the talonavicular
positive
angle.
The conventional
groups
(I,
randomly.
only one foot met the radiographic
study,
so the number
of feet that
angle between
the talus and the sole of the foot
3-A) of more than 35 degrees.
(This angle has also
described
as the plantar
flexion
angle of the talus’4 and
made
28.2
13.5
with
University
of California
inserts9.
Thirty-one
patients
study
Anteroposterior
with the child
3-A)
±
to the control
group;
thirty-two
shoes;
thirty-five,
with Helfet
1 . An
(Fig.
28.7
signed
to treatment
who picked
numbers
criteria:
(Fig.
been
17.0
±
insert.
the study.
One hundred
and twenty-nine
of the
who were referred
to us met the criteria
for
Studies
made
27.2
9 (41%)
13 (59%)
deviation.
child who had, or proved
(cerebral
palsy or muscular
cluded
from
13 1 patients
13.6
±
22 (81%)
5 (19%)
16 (57%)
12 (43%)
13 (62%)
8 (38%)
Through
Association,
were
re-examined
arrangement
the pedorthist
to ensure that all ofthe corrective
to the standards
at three-month
Clinic
by the study team,
an orthopaedic
surgeon,
and
specifications
intervals
which
included
a
and a certified
pe-
with the Prescription
was present
for
Footall visits
shoes were fitted according
of that
association.
804
D.
R.
WENGER
FIG.
Lateral
radiograph
of the
foot,
made
with
the child
standing
and
wearing
Great
care was taken
to maintain
a proper
fit of the
shoes,
since one of the criticisms
of the earlier,
failed trial
had been that the fit of the shoes might not have been correct
or might
not have been maintained
satisfactorily,
The fit of the shoes and of the corrective
inserts
firmed
at each visit.
If it was not correct,
new
inserts,
mented
or both.
was conshoes
or
or both,
were
provided.
Compliance
was docuby observing
the wear of the shoes.
Non-compliant
patients
were
To allow
dropped
efficient,
from the study.
correct
shoe-fitting,
of
of shoes was kept on portable
carts in the central
storage
area of the hospital
and was brought
for each followup visit. All of the shoes were provided
by the Prescription
Footwear
Association.
Data on growth
of the feet and the
frequency
that the shoe size changed
because
of the growth
recorde&5.
Every
six
and
also
repeated
4
a shoe
repeat
anteroposterior
at six-month
To ensure
unbiased
and
radiographs
were
tween
while
wearstudy
all records,
photographs,
in a confidential
file
by
the
Recent
recent
follow-up.
This
was
done
in a blinded
manner
that is, without
knowledge
of the patient’s
name or treatment group.
The paired
sets of photographs
(initial and most
recent)
were graded
on a scale of 1 to 4, with 1 meaning
greatly
improved;
2, improved;
3, no change;
and 4, worse.
This was done to determine
if there was a correlation
be-
impressions
Biomechanics
Laboratory
and the radiographic
insert.
parameters.
Analysis
were
tested
explicitly
of variance
between
groups
The log of the mean
was
by testing
using
plotted
BMDP
against
for equality
program
the log
P7D4.
of the
standard
deviation
for the four experimental
groups.
On the
basis of this analysis,
it was determined
that none of the
data sets required
transformation
for variance
stabilization.
In addition,
the distribution
of all data sets was tested
for
and symmetry
by calculating
g2 for each experimental
group.
variables
were
log-transformed,
skew
g and kurtosis
As a result,
all radiographic
which
resulted
in a nor-
mally
distributed,
symmetrical
assumptions
of the parametric
types
To assess the clinical improvement
in the appearance
of the foot, the orthopaedic
surgeons
and certified
pedorthist
compared
the clinical
photographs
that had been made at
the beginning
of the study with those that were made at the
most
clinical
of California
The data from the radiographic
measurements
were
transcribed
from standardized
forms onto a PDP- 1 1/73 computer
(Digital
Equipment,
Maynard,
Massachusetts)
for
analysis
using the BMDP
statistical
package4.
Assumptions
of the parametric
statistical
methods
that were used (as will
data
set
and
satisfied
the
tests.
In order
to test the null hypothesis
of treatment
had no effect
on the
that the various
change
in the ra-
diographic
angles,
a one-way
analysis
of covariance
was
used,
specifying
the initial
radiographic
angle
as the covariant
(BMDP
program
P2V).
This parameter
was implemented
as a covariant
the initial radiographic
groups
-
a University
Statistical
intervals.
scoring,
stored
lateral
research
nurse.
At the time of the three-year
follow-up,
clinical,
radiographic,
and photographic
analyses
were completed
with the examiners
in ignorance
of the child’s
treatment group.
Most
with
normality
months,
radiographs
of the feet were made
for each patient
the child stood barefoot
and also while the child stood
ing the corrective
shoes
(Fig. 4). The photographic
was
AL.
be described)
an inventory
500 pairs
were
ET
(p < 0.01).
on the basis of the observation
that
angle differed
significantly
between
Assumptions
iance (equality
of slopes
variances
between
groups)
of the
between
groups
were explicitly
analysis
of covar-
and
tested
equality
of
and shown
to be satisfied.
Two separate
models
for analysis
of covariance
were
used because,
in a few patients
(sixteen
of ninety-eight),
only one
both feet
dren).
In
the foot
foot was measured.
The first model was used when
were evaluated
(eighty-two
of ninety-eight
chilthis model,
the data were grouped
by treatment,
(left or right)
was used as the within
factor,
and
the initial radiographic
the second
model,
the
angle
data
was used as the covariant.
In
were
grouped
by treatment
THE JOURNAL
OF BONE
AND
JOINT
SURGERY
CORRECTIVE
SHOES
AND
INSERTS
AS TREATMENT
TABLE
RADIOGRAPHIC
Group
II
MEASUREMENTS
I
Group
II
Group
III
Group
of children
21
28
27
22
39
54
49
38
the
talus and
(degrees)
foot*
the
39.8
34.7
-5.48
between
the talus and
first metatarsal*
(degrees)
Initial
Final
Deltat
angle*
Deltat
and
standard
t Corrected
±
0.71
40.5
±
0.70
39.5
±
0.6
41.8
±
0.78
±
0.73
34.0
±
0.66
34.7
±
0.61
34.2
±
0.84
±
0.71
-6.47
±
0.59
-5.47
±
0.63
±
0.72
16.7 ±
11.8 ±
-5.78
±
0.87
19.1
11.7
-6.80
±
16.8 ±
11.5 ±
-6.65
±
19.7
±
0.83
±
0.75
0.84
0.70
0.76
0.91
0.83
0.67
11.3 ±
36.3
31.5
-4.50
±
0.81
±
0.99
36.2
±
1.2
±
1.2
29.4
±
0.74
±
0.91
±
0.78
-7.36
±
0.84
36.8
±
0.97
±
0.77
30.1
±
0.82
±
0.81
±
0.92
value
obtained
from
analysis
ment group at the time of entry
recent
follow-up
is documented
levels
are discussed
that
multivariate
obtained
the clinical
was related
discriminant
using
the
groups,
using
was
performed
F-to-enter
particular
=
radiographic
at p
<
To protect
angle.
Results
were
the sample
population
considered
sig-
0.05.
Type-Il
error
(accepting
a false
null
analysis
was perthat was needed
for
equations,
size of
was calculated
using four treatment
groups,
a
standard
deviation
of 5 degrees,
a desired
dif-
ference
to detect of 5 degrees,
a significance
level
and a power ofO.9.
Under these conditions,
twenty
were needed
for each group.
Expecting
that some
would
not complete
the study
because
of failure
erate,
we planned
to enroll
at least thirty children
group,
hoping
would
complete
null hypothesis
of 0.05,
samples
children
to coopin each
that approximately
twenty
in each
group
the study.
In this way, if we accepted
the
when p > 0.05,
we would
be 90 per cent
we were
not
committing
a Type-Il
error.
Data
were
collected
on all
year period of study.
Thirty-one
for all of the follow-up
visits,
129 patients
patients
did not
ularly,
or moved
from the state were
The remaining
ninety-eight
patients
VOL.
71-A,
NO.
6.
JUL?
1989
over
the
six-
who did not return
wear the shoes reg-
dropped
from the study.
had a minimum
follow-
and
records
The
complete
(Table
radiographic
An
analysis
clinical,
angle
between
severity
of the
and
and
flatfoot
in each
treat-
into the study and the most
in Figures
5, 6, and 7.
of covariance
the talus
radiographic,
II).
between
using the initial radiographic
angle
no significant
difference
between
the
four
groups,
as the covariant,
revealed
groups
for change
in the
the sole
of the foot
(p
correlation
between
radiographic
angle (p <
the initial angle and
0.001 ), demonstrating
itive
who
had
independent
ysis
between
the
angle
sole
equation.
variance.
initial
experimental
derived
from
discriminating
caneal
a large
of the method
angle
groups,
the photographs,
function.
Only
and change
of the
foot)
had
of treatment.
(p >
8, 9,
a pos-
the change
in
that the pa-
the
most
change,
Discriminant
based
anal-
on the impression
resulted
in a relatively
poor
two variables
(initial
talocal-
in the angle
were
0.4),
>
in the angle between
the talus and the first metatarsal
0.5),
or in the talocalcaneal
angle (p > 0.5). (Figs.
and 10). For all radiographic
parameters,
there was
entered
between
into
Most radiographic
angles
had
The ability of the discriminating
the
the talus
and
discriminating
a great deal of coequation
to clas-
sify various
classes
into the correct
group
retrospectively
ranged
from a high of 60 per cent (for the group
that was
clinically
described
as worse)
to a low of 12 per cent (for
the group that was classified
as having
no change).
Stepwise
linear regression
again revealed
that the greatest contributor
to change
in the radiographic
angle was the
initial
Results
years
photographic
tients
against
hypothesis),
a priori
statistical
power
formed
to select the size of the sample
the experimental
design.
With standard
that
=
2.995.
Finally,
in order to unthe course
of the flatfoot
better,
stepwise
linear
was used to determine
which
of the numerous
variables
contributed
most to the change
in a
derstand
regression
measured
sure
first
impression,
as deto the radiographic
analysis
experimental
3.000 and F-to-remove
nificant
were
in the text.
determine
whether
from photographs,
four
-6.44
of covariance.
the
the
0.92
30.0
-6.65
up of three
between
0.98
±
37.1
(Groups
I to IV), there was no within
factor,
and the initial
radiographic
angle
was again
used as the covariant.
The
two methods
yielded
identical
results.
Thus,
for simplicity,
variable,
-6.44
error.
mean
significance
-6.62
(degrees)
Final
To
termined
IV
of feet
Talocalcaneal
Initial
model
805
FLATFOOT
Number
Angle
the
Mean
FLEXIBLE
Number
Angle
between
sole of the
Initial
Final
Deltat
*
FOR
radiographic
angle.
Interestingly,
contributed
to change
in the radiographic
patients
who had the greatest
initial laxity
improvement.
Serial
correlation
coefficients
initial
angle,
having
for
laxity
also
with the
the most
the three
variables
ranged
from 0.33 (change
in the angle
between
the talus and the first metatarsal)
to 0.47
(change
in the
talocalcaneal
angle),
indicating
that these correlations
with
laxity were relatively
weak.
806
D.
of Foot Anq,
Tolus-Sol.
--
Control
R.
WENGER
ET
AL.
(fl
GrOUD
Tolus-Sol.
of Foot
Angi.
Sho. Gp. (H)
Corr.ct.
--
k.y
=
30
C
0
U
:1
K
40
E
30
-‘S
20
8
j
I
10
n
-‘
20
%%
10
:r
fl
:
--
34-47
d.gr..s
21-34
d.gr..s
ll
47-60
21-34
d.gr.ss
34-47
d.gr..s
47-60
d.gr..s
Talus-Sol.
of
Foot
Angi.
d.gr..s
UCBL Ins#{149}rt
Group
--
(IV)
k.y
--.
n
30
C
0
U
20
2
8
0
0
.2
.2
E
E
10
3
3
z
z
21-34
34-47
d.gr.es
FIG.
Bar
time
graphs
documenting
of both
entry
into
the
the
severity
study
and
Tolo-Ist
of flatfoot
the
most
recent
--
Anal.
Ustatorsol
in each
group.
follow-up.
Control
GrOUD
=
by the
University
lateral
angle
between
of California
Talo-lsf
(l’l
C
0
the
talus
Biomechanics
and
the
sole
--
Corr.ct.
Sho.
zz
d
of the
foot,
at the
Laboratory.
#{149}totarsoI AngI.
Ky
=
d.gr..s
5
as measured
UCBL
47-60
d.gr..s
Gp.(II)
K.,1
-
=
30
C
0
20
0
-.-.-..
20
0
0
--‘
0
..-%
.2
10
.2
E
3
z
E
z
3
r-i
10
:Ii’’
--
0-14
14-25
d.gr..s
d.gr..s
.t M.totarsol
Tab-i
28-42
Mgls
--
30
H..I
Group
Cup
Tab-i
(III)
=
st .totarsa
Angi.
..qrss.
-- UCOL Inwf
Op.
(IV)
kip
=
-
fld
-%
-
C
20
%-.
I
.
20
d.gr..s
1k.,1
U
0
14-28
d.gr..s
d.gr..s
C
0
0-14
0
.2
E
z
I
10
nI-0-14
d.gr..s
14-28
d.gr..s
On--n
0-14
d.grss
28-42
d.gr..s
FIG.
Bar graphs
of both entry
documenting
into the study
the severity
of flatfoot
and the most recent
in each
follow-up.
10
14-28
d.gr..s
28-42
d.gr..s
6
group.
as measured
by the lateral angle between
UCBL
= University
of California
Biomechanics
the talus and
Laboratory.
THE JOURNAL
the
first
OF BONE
metatarsal,
AND
JOINT
at the time
SURGERY
CORRECTIVE
ToIo-Colcon.aI
Anal.
--
SHOES
AND
Control
Groua
INSERTS
AS TREATMENT
FOR
(II
FLEXIBLE
TaIo-Colcan.oI
807
FLATFOOT
AngI.
--
Corr.ctlv.
Shoi Group (H)
Ik.p
-
K.p
=
30
r;
30
C
0
20
.
0
20
I
I
10
10
r1i_
10-30
30-50
d.gr..s
-.
50-70
10-30
d.gr..s
ToIo-Calcanal
MgI.
d.gr..a
H.&
--
Cup Group
-
40
TaIo-Calcan.ol
(III)
30-50
50-70
d.gr..s
d.gr..s
Arigl.
UCBL Ins#{149}rf
Group
--
-
k.p
I
k.p
=
30
n
j
C
0
(IV)
0
30
U
U
20
-
“S
‘S
.0
-‘
0
0
20
0
.2
.2
E
E
z
r-i
II
I
]
II
U
n__
f1
10-30
50-70
30-50
d.gr..s
d.gr..s
d.gr...
FIG.
Bar graphs documenting
the severity
the study and the most recent
follow-up.
The initial
foot-progression
foot-progression
angle
recent
the
heel
cord
(p
0.6),
>
who
Subjective
were
of flatfoot
UCBL
(p
and
or genu valgum
(p > 0.5)
in predicting
improvement
patients
>
initial
angle
0.3),
in each group,
as measured
= University
of California
(p >
initial
degree
0. 15), most
tightness
of
of genu
varum
were not found to be significant
in either the control
group or the
treated.
Some parents
in the control
group
reported
that the children
even
had less pain in the feet, fewer
in the
better
calves,
-
and
gait
The reports
were
related
to parental
seemed
after
they
difficult
to
personality.
began
validate
Cheerful,
those
aches
wearing
the
and often
coopera-
tive parents
were more likely to report relief of pain. Also,
the recipient
of the data seemed
to affect
the report.
The
research
nurse often received
a somewhat
different
report
than did the orthopaedic
surgeon
or the pedorthist.
Because
reported
factors
we
were
unable
improvement
were
not
in gait
analyzed
to
accurately
or decrease
quantify
in aching,
the
these
further.
study
demonstrated
that
normal
preschool
children
commonly
have poor formation
of the longitudinal
arch and that the arch gradually
improves
with growth.
Whether
children
who have marked
flatfoot
also have spontaneous
improvement
with growth,
and whether
the degree
of improvement
can be affected
by treatment,
have not been
VOL.
71-A,
NO.
6.
JULY
1989
30-50
50-70
d.gr..s
d.gr..s
7
by the anteroposterior
Biomechanics
Laboratory.
previously
Bleck
talocalcaneal
determined,
and Berzins
angle,
to our
reported
at the time
knowledge.
the results
University
of California
Biomechanics
and Helfet
heel-cups
in 122 children,
domly
assign
the patients
to the two
However,
only
twenty-two
patients
no controls
were included.
The results of the present
of both
entry
oftreatment
into
with
Laboratory
inserts
but they did not ranmethods
of treatment
completed
study
the study
and
that
flex-
demonstrated
ible flatfoot
in children,
quantitated
by measurement
of the
angles
between
the talus and the sole of the foot, the talus
and the first metatarsal,
and the talus and the calcaneus,
naturally
improves
over a three-year
period.
In our series,
the degree
of improvement
was not affected
by wearing
a
corrective
shoe or a shoe with
Because
flatfoot
is thought
ity,
we
provement
children
Discussion
A recent
10-30
d.gr..s
or include
matched
controls.
Bordelon
treated
fifty children
with custom-molded
inserts
and reported
improvement
in
the talometatarsal
angle, as measured
on lateral radiographs.
Impressions
shoes.
‘
10
3
10
analyzed
the
an insert.
to reflect
relationship
ligamentous
between
of flatfoot.
Interestingly,
who had greater
laxity,
laxity
laxand
im-
our data suggested
as measured
using
that
the
Wynne-Davies
scale when they entered
the study,
had somewhat greater
improvement
than did children
who had less
laxity.
This
somewhat
surprising
correlation
was
only
weakly
supported
by statistical
analysis.
Also,
the radiographic
angles
who had more
improved
more
with
severe
flatfoot
initially
who
values.
had
lower
Thus,
a child
time
than
who
for the patients
for the children
had
greater
laxity
808
D.
R. WENGER
FIG.
Lateral
radiograph
angle between
the
of the foot seen in Fig. 3. made with
talus and the sole of the foot decreased
and more severe
ment.
However,
strated
that the initial
radiographic
severity,
were far more significant
radiographic
improvement.
In
selected
peak at
lowered
phasized
improvedemon-
becomes
irreversible’
‘ ‘
older
the feet.
percentage
practice.
angles,
which
reflected
than laxity in predicting
the corrective
talus and the
children
who
shoe.
after three years of treatment.
first metatarsal,
to 15 degrees.
had
extremely
severe
The
pronation
of
Also,
these younger
children
account
for a large
of the flat feet that are encountered
in clinical
We chose an upper
age-limit
of six years old be-
after that age peer pressure
to wear sneakers
or other
fashionable
footwear
makes
treatment
with shoes or inserts
impractical.
Although
we considered
three years to be adequate
for
initial assessment,
we thought
that we might subsequently
extend
the period oftreatment
to five or six years.
However,
cause
‘
.
We did not want
to study
number
of patients
who
returned
for
the
most
recent
Final
J50
40
the
Key
Initial
-
a,
and not wearing
that between
the
only
this study,
the age range
(one to six years)
was
because
the clinical
problem
and parental
concern
this time.
The age for entry
into the study
was
to one year at the urging
of pedorthists,
who emthat treatment
must be started
early,
before
the
deformity
AL.
8
the patient
standing
to 34 degrees
and
flatfoot
was likely to have more
the serial correlation
coefficients
ET
=
=
=
0)
C
:
0
0
L&
I20
0
a)
0
i
10
Control
Corrective
Shoe
Heel
UCBL
Cup
Insert
FIG. 9
Comparison
the four
groups
of the
(p >
initial
0.4).
and
Each
the
bar
most
recent
represents
lateral
the
mean
angles
and
between
the
standard
the talus
error.
and
UCBL
the
sole
=
of the
University
foot. There
was no significant
of California
Biomechanics
TIlE
JOURNAL
OF
BONE
difference
between
Laboratory.
AND
JOINT SURGERY
CORRECTIVE
SHOES
AND
INSERTS
AS TREATMENT
FOR
FLEXIBLE
809
FLATFOOT
U)
a)
20
a)
a,
#{149}0
a,
0)
C
1
0
0
4-
0
a,
1
4-
4-
U)
0
0
I-
a)
:
Control
Corrective
Shoe
FIG.
Comparison
four groups
of the initial and the most recent
lateral
angles
between
(p > 0.5).
UCBL
= University
of California
Biomechanics
follow-up
corrective
confirmed
that children
shoes
for only a few
the families
were
and well adjusted,
the children
dren
stable
some
to wear
were
dropped
can be persuaded
years.
Although
and the
parents
parents
were
were unable
the corrective
from
the
footwear,
study.
The
ninety-eight
children
who completed
mented;
however,
almost
none wanted
the shoes
after
the completion
responsible
to convince
and these
compliance
chilof the
the study was docuto continue
wearing
of the three-year
trial.
For these reasons,
we doubt that a carefully
study requiring
diligent
wear of the corrective
be easily
prolonged
for more than three years.
study,
the rate of non-compliance
would
probably
become
so high as to negate
the validity
of the results.
Similarly,
even if a stoical
group
of children
could
be coerced
into
participating
in a five or ten-year
treatment
plan and the
outcome
suggested
formation
American
shoes
that
the
inserts
beyond
the
age
of entry
We do not recommend
shoes
tennis
who
have
typical
the
child
wear
that
the
in-
school.
corrective
flexible
either
shoes or inserts for
flatfoot.
Instead,
shoes
or soft,
no
a smooth,
flexible
shoes
or inserts
we adflexible
(rubber-soled
most children.
on carpeted
or other abrasive
surfaces,
better wearing
lightweight,
soft leather
a toddler
shoes that
sole.
We still occasionally
tive
into
to protect
the soles of the feet. Sneakers
shoes)
seem
to be satisfactory
for
However,
may walk
have
effective,
would be of little practical
value.
Average
North
children
cannot
be persuaded
to wear corrective
children
vise
were
has
severely
symptomatic
flatfoot,
whether
the symptoms
are in the foot or in the calf.
This is done only after known
causes
of painful
flatfoot
(tarsal
coalition
or subtalar
rheumatoid
arthritis)
have been
ruled out. Since we could not document
any long-term
structural
benefit
to the foot (that
is, improvement
in radiographic
Cup
Insert
first
metatarsal.
There
is no significant
difference
between
the
angles),
we suggest
that children
wear corrective
shoes only
if the parents
note a functional
benefit.
For the rare older
child (five to fifteen years old) who has flatfoot
and aching
in the foot or calf, widely
available
high-top,
lightweight
athletic
shoes
symptoms
paedic
generally
and
are
provide
better
footwear.
Parental
perceptions
that
application
themselves
wear,
must
adequate
accepted
of
be treated
are convinced
that
to relieve
traditional
flexible
flatfoot
remain
a major
of our recommendations.
had flat feet as children,
and
support
than
ortho-
as a serious
roadblock
to
Frequently,
the parents
wore corrective
foot-
the footwear
for any improvement
that occurred.
to dissuade
the parents
from buying
was
responsible
A physician’s
attempt
corrective
footwear
is
often met with serious
skepticism.
Unfortunately,
often the
shortest
and easiest
escape
for the physician
is to write a
prescription
for corrective
shoes or inserts.
When possible,
we avoid this by carefully
examining
the child and the feet,
out abnormalities,
and emphasizing
to the parents
that the flatfoot
is a normal
variation
that will improve
over
time.
We state that it is unlikely
that the child will have
symptoms
in adult
life. When
one or both parents
have
ruling
severe,
symptomatic
flat feet, we clarify
that the child may
indeed
have a similar
condition
when
he or she is fully
grown,
but that it can be treated
with a supportive
insert at
that time. We also give the parents
an illustrated
booklet
explaining
foot and
advise that a child wear suppor-
if he or she
the
disorder
controlled
shoes could
In a longer
UCBL
10
the talus and
Laboratory.
to wear
generally
Heel
current
opinion
its treatment.
In conclusion,
flexible
growth
flatfoot
and that
or inserts
history.
NOTE:
the patients
authors
in the study
the nature
of flexible
flat-
the results of this study confirmed
in young
intensive
for a three-year
The
about
thank the clinical
and Judy I.each.
children
treatment
period
research
R.P.T.
.
slowly
improves
with corrective
does
not
alter
the
that
with
shoes
natural
nurses for their help in enrolling
and maintaining
for her help in preparing
the pap
810
D.
R.
WENGER
ET
AL.
References
1 . ASHER,
2.
BLECK,
C.: Postural
E. E. , and
Variations
BERZINS,
in
Childhood,
U. J. : Conservative
pp.
76-101
. London, Butterworth,
Management
of Pes Valgus
1975.
with Plantar
Flexed
Talus,
Flexible.
Clin.
Orthop.
,
122:
85-94,
1977.
3.
4.
5.
6.
BORDELON,
DixoN,
W.
R. L.: Correction
of Hypermobile
Flat Foot in Children
by Molded
Insert.
Foot and Ankle.
1: 143-149,
J.: BMDP
Statistical
Software.
Los Angeles,
University
of California
Press,
1983.
N. J.: Foot Disorders.
Medical
and Surgical
Management.
Ed. 2, pp. 149-152.
Philadelphia,
Lea and
GIANNESTRAS,
GILADI,
MICHAEL;
AHARONSON,
ZEEv:
84,
MILGROM,
The Low
CHARLES;
Arch,
STEIN,
a Protective
MICHAEL;
Factor
KASHTAN,
in Stress
HANOCH;
Fractures.
MARGULIES,
A Prospective
JOSEPH;
Study
of 295
CHISIN,
Military
1980.
Febiger,
ROLAND;
Recruits.
1973.
STEINBERG,
Orthop.
and
, 14: 81-
RAYA;
Rev.
1985.
R. I., and BEATH,
THOMAS:
Hypermobile
Flat-Foot
with
Short Tendo
Achillis.
J. Bone and Joint Surg. , 30-A:
116-140,
Jan. 1948.
8. HELFET,
A. J. : A New Way of Treating
Flat Feet in Children.
Lancet,
1: 262-264,
1956.
9. HENDERSON,
W. H. , and CAMPBELL,
J. W.: UCBL
Shoe Insert Casting
and Fabrication.
Technical
Report 53. Biomechanics
Laboratory,
University
of California,
San Francisco
and Berkeley,
1967.
10. STAHELI,
L. T., and GRIFFIN,
LAURA:
Corrective
Shoes for Children:
A Survey
of Current
Practice.
Pediatrics,
65: 13-17,
1980.
11. STAHELI,
L. T.; CHEW,
D. E.; and CORBETT,
MARILYN:
The Longitudinal
Arch. J. Bone and Joint Surg. , 69-A: 426-428,
March
1987.
12. STEWART,
S. F.: Footgear
Its History,
Uses,
and Abuses.
Clin. Orthop.,
88: 119-130,
1972.
13. TACHDJIAN,
M. 0. : The Child’s
Foot, pp. 556-597.
Philadelphia,
W. B. Saunders,
1985.
14. VANDERWILDE,
RUSSELL;
STAHELI,
L. T. ; CHEW,
D. E. ; and MALAGON,
VALENTIN:
Measurements
on Radiographs
of the Foot in Normal
Infants
and Children.
J. Bone and Joint Surg. , 70-A: 407-415,
March
1988.
15. WENGER,
D. R.; MAULDIN,
D. ; MORGAN,
D.; SOBEL.
M. G. ; PENNEBAKER,
M. ; and THALER,
R. : Foot Growth
Rate
in Children
Age One to Six
Years.
Foot and Ankle,
4: 207-210,
1983.
16. WYNNE-DAvIES,
RUTH:
Acetabular
Dysplasia
and Familial
Joint Laxity:
Two Etiological
Factors
in Congenital
Dislocation
of the Hip. J. Bone
and Joint Surg.,
52-B(4):
704-716,
1970.
7.
Nov.
HARmS,
THE JOURNAL
OF BONE
AND
JOINT
SURGERY