Gait & Posture 61 (2018) 423–430

Contents lists available at ScienceDirect

Gait & Posture

journal homepage: www.elsevier.com/locate/gaitpost

Review

Effects of high heeled shoes on gait. A review T

⁎
M.M. Wiedemeijer , E. Otten
Center for Human Movement Sciences, University Medical Center Groningen, Groningen, The Netherlands

A R T I C L E I N F O A B S T R A C T

Keywords: Background: Walking in high heels (HH) may alter gait in various ways, which may be of importance to de-
Shoes signers and physicians.
Footwear Research question: How does walking in high heels alter gait and how can this be explained from a biomechanical
Walking and control point of view.
High heels
Methods: Relevant literature has been collected in which high heeled walking was studied, after which the
Gait
results were bundled and interpreted in a framework of biomechanics and control.
Results: Major changes were found in the rollover function of the feet, the ankle and knee joints and the lower
back, while step length and balance were compromised. An increase in heel height forces the foot in an increased
plantar flexion, which in its turn increases knee flexion and lordosis of the lower back. All changes can be related
with each other in a plausible pattern of movement and control.

1. Introduction
Presently it is mostly women who wear high heels and claim their
purpose is for aesthetic reasons only. On top of that Guéguen [7] stu-
died the effect of men's behavior on women wearing HH shoes. He
found that men's helping behavior increases as heel height increases.
Although there seem to be some advantages, there are many more
disadvantages in walking in HH shoes. For instance, apart from the
discomfort during walking in HH shoes, the long term effects have been
linked to many medical problems. Many studies in HH gait are done
from a medical point of view, and with specific parts of the body (e.g.
knee, ankle, foot) as their main focus. However, in order to fully un-
derstand how different parts of the body are affected by HH gait, we
need to bundle the different findings and find the underlying me-
chanism. That is the purpose of this review: to increase our under-
standing of high heeled gait. In this way, we can more thoroughly
understand the separate changes in gait since they are part of the whole
pattern of gait itself. This can be relevant information for the users of
HH shoes, shoe designers and physicians. For instance, they could ad-
dress complaints of pain that may not seem logically linked to HH gait
at first sight. Next to that, designers could adjust a shoe to make it more
comfortable to wear.
2. Methods
This critical review is about the biomechanical effects and control of
high heeled shoes on gait. Both authors agreed on using the following
selection criteria: the focus in the search of articles was to collect a
group of studies of which an as large as possible part was handling the
field of high heeled gait. Shoes with a heel height of 5 cm or higher
were considered as being shoes with high heels. This choice is made
based upon the overall used heel height within the studies that com-
pared high heels with low heels or barefoot walking. There was no
selection made on the sort of heels (wide based or stiletto heels) that
were used within the studies. This is due to the facts that some studies
are very concise in explaining the type of heels used and that the di-
versity in shoe types used was large between the studies. The same
accounts for the diversity of the types of shoebox and the fact that there
were differences in whether he experimenter provided the shoes or the
own shoes were brought by the subject.
Gait was considered to occur at a natural walking pace on an even
and horizontal surface. Studies on the effect of HH shoes in gait within
the framework of metabolism, energy consumption or from a medical
viewpoint were excluded. After using the Smartcat (https://rug.on.
worldcat.org/discovery) literature search engine on the following
keywords: [(shoes OR footwear) AND walking AND high heels AND
gait], 218 articles were found. Of these articles abstract and title were
screened after which 184 references were found to be irrelevant, based
on the above arguments. Twenty articles were excluded because they
did not match with the selection criteria or only contained an abstract.
After reading the studies, 5 new relevant papers were included from
their references. Finally, 19 studies formed the basis of this review
(Fig. 1). Since this review is about the principles of alterations in HH
gait, no attempt is made to select on certain demographics, ethnicity or

⁎
Corresponding author.
E-mail address: mickey.wiedemeijer@gmail.com (M.M. Wiedemeijer).

https://doi.org/10.1016/j.gaitpost.2018.01.036
Received 21 April 2017; Received in revised form 28 January 2018; Accepted 29 January 2018
0966-6362/ © 2018 Elsevier B.V. All rights reserved.
M.M. Wiedemeijer, E. Otten
Fig. 1. Flowchart literature search.
age groups of the female subjects in the papers. All moments of force in
the present paper are internal moments of force: the moments of force
produced by the muscles. When papers use external moments of force,
they have been transformed to internal moments of force, to make
comparisons consistent.
3. Results
An overview of the extracted data and test conditions can be found
in the appendix. The data from the literature have been reorganized and
bundled in several subjects, biomechanically from feet upward (foot
pressure and rollover curve, the ankle, knee, and hip joints and above),
and after that the CoM and gait parameters were covered.
3.1. Foot pressure and the rollover curve in high heeled gait
Speksnijder et al. [18], Snow et al. [17] and Eisenhardt et al. [5]
studied foot pressure differences in walking with high heels compared
to low heeled shoes and a barefoot condition. Speksnijder et al. [18]
provided shoes and Snow et al. [17] asked the women to bring their
own shoes. Eisenhardt et al. [5] studied barefoot and shoe conditions
paced by a metronome that was set at a cadence of 100 steps per
minute. Speksnijder et al. [18] divided the foot in 7 regions: the heel,
the midfoot, the forefoot (medial, central, and lateral), and the hallux
and toes. Snow et al. [17] studied 6 regions by excluding the midfoot.
Speksnijder et al. [18] found in walking with HH a significant de-
crease in the contact area of the heel, midfoot, lateral forefoot and
central forefoot [18]. The contact time increased in the midfoot and
lateral forefoot [5,18].
The time to maximum pressure for the hallux increased compared to
the barefoot condition [17]. In the gait cycle the time to maximum
pressure under the forefoot was significantly reduced with increasing
heel height [17]. The maximum force decreased in the heel and mid-
foot, but increased in the medial forefoot and the hallux. The force time
integral decreased in the midfoot and increased in the central forefoot
and hallux with an increasing heel height [17,18].
Peak pressure increased in the central forefoot, medial forefoot and
hallux [18] between the shoe conditions and compared to the barefoot
condition peak pressure increased in all the areas [17]. Maximum
pressure under the hallux tended to be the highest, followed by medial
forefoot, central forefoot and lateral forefoot with the least pressure and
increased under the forefoot with an increase in heel height [17]. The
pressure time integral decreased in the midfoot and increased in the
heel, central forefoot, medial forefoot and hallux [18]. A linear corre-
lation was found between the increase of heel height and the increase of
peak pressure and pressure time integral in the medial forefoot and
peak pressure in the central forefoot [18].
Hansen et al. [8] examined the effects of heel height on the rollover
424
Gait & Posture 61 (2018) 423–430
characteristics [8] in gait with the use of markers. Women were asked
to walk in shoes of their own, one pair with medium high heels and one
pair with high heels. Hansen et al. provided a pair of shoes with heels of
0 cm high. They found that the rollover radii [8] did not change with a
significant effect with increased heel height. A roll over curve is the
effective curve of contact of the shoe foot combination in the coordinate
system of the lower leg. In the anterior–posterior direction no sig-
nificant difference was found between the LH and medium high con-
ditions. In the HH condition a shift of the rollover curve [8] was found,
resulting in a more posterior position with a significant effect. In the
vertical direction the rollover curve changed significantly for all the
three heel heights. With an increasing heel height, the rollover curve
moved down with respect to the ankle marker on the lateral malleolus.
3.2. The ankle joint in high heeled gait
Barkema et al. [2], Cronin et al. [4], Esenyel et al. [6], Kerrigan
et al. [10], Mika et al. [13] Simonsen et al. [16] and Stefanyshyn et al.
[19] investigated the net joint moments and or muscle activity in HH
walking around the ankle joint. Barkema et al. [2] found that the ankle
moments were strongly affected by changing heel height. In walking in
the LH shoes, the first 80% of the stance phase showed an inversion
moment. In the HH conditions a large eversion moment was seen right
after heel strike. In late stance the higher the heel, the more eversion
moment was found [2]. However Kerrigan et al. [10] found that in the
frontal plane the ankle eversion moment was reduced in the HH con-
dition compared to the barefoot condition and they state that the
eversion angle of the foot decreased as heel height increased, however
without significant differences [19]. In the sagittal plane, plantar
flexion angles increased at heel strike with a significant effect when
heel height increased [13]. Just after heel strike the ankle joint showed
a significantly higher dorsiflexor moment in the HH condition com-
pared to the barefoot condition [16]. Later in the stance phase, at push
off, a significantly lower plantar flexion moment was seen
[4,6,10,16,19]. According to Esenyel et al. the plantar flexor muscle
work was reduced by 29% compared to the LH condition [6]. Regarding
the ankle angle in the HH condition the shoe forced the foot into a more
plantar flexed position and therefore the ankle angle was significantly
different in the first part of the stance phase and after toe off [16]. The
range of motion (RoM) of the ankle was smaller in the HH condition
and the ankle showed more plantar flexion during ground contact
[4,10,13,19]. Muscle activity of muscles spanning the ankle joint ap-
peared significantly higher in the HH condition compared to the
barefoot condition [13]. Cronin et al. [4] found that the tibialis anterior
and the soleus muscles showed an increase in activation in the HH
condition, which indicates co-activation. No stance phase was in-
dicated. However in the study of Mika et al. [13] the tibialis anterior
muscle in the HH condition only showed an increased activity at heel
strike. In the same study the rectus femoris and the medial gastro-
cnemius muscles exhibited a significant effect in activation in the HH
condition compared to the barefoot condition, both at heel strike and
toe-off [13]. Cronin et al. [4] found that the medial gastrocnemius
muscle did not change in activation when averaged, but its activation
was prolonged during stance phase in the HH condition [4]. Stefany-
shyn [19] found that the activity of the soleus muscle increased as heel
height increased and the muscle activity of the gastrocnemius and ti-
bialis anterior muscles revealed no significant differences between the
conditions.
3.3. The knee in high heeled gait
In the sagittal plane several authors [6,9,13,16,19,20] found that
the angle of knee flexion at heel strike and midstance increased as well
as the extensor moment with an increasing heel height. This corre-
sponded with the found increased muscle activation of the rectus fe-
moris muscle to control this increased knee flexion, by increasing knee
M.M. Wiedemeijer, E. Otten
extensor moments [19]. After limb loading the concentric knee extensor
muscle work increased by 200% [6]. In barefoot walking a flexor mo-
ment was present just before reaching the push off phase. In the HH
condition this part of the stance phase showed an extensor moment
[16]. Simonsen et al. [16] found that knee extensor moments in the
stance phase were twice as high in the HH condition than without shoes
[16], and with a longer duration [6]. Ho et al. [9] also found that with
an increase of heel height the knee peak extensor moment increased
significantly during weight bearing. At toe-off the maximum knee ex-
tension moment decreased in HH condition compared to the LH con-
dition, while body loading showed no effect [16]. The kinetic internal
knee extensor pattern shifted to a little later in the gait cycle from the
peak in early stance in the HH condition and showed a prolonged ex-
tensor knee moment. During midstance phase the knee showed a sig-
nificantly decreased flexor moment compared to barefoot walking [10].
The maximum knee flexion moment appeared to increase significantly
in the HH condition after correction for walking speed in terminal
stance [20]. Wearing higher heels decreases the maximum abduction
moment around heel strike, and increases the maximum abduction
moment during midstance and at toe-off [20,16]. In the study of Bar-
kema et al. [2] the peaks in the knee abduction moments at early and
late stance increased as heel height increased in both walking condi-
tions. In the frontal plane, the peak knee adduction moment increased
in the HH condition during early load bearing, but without a significant
effect [6]. In the study of Kerrigan et al. [10] the knee showed an in-
creased adduction moment throughout the stance phase and in the
beginning and end of stance phase, these differences were significant
compared to the barefoot condition. In the HH condition the knee RoM
was greatly reduced compared to the barefoot condition [1,13]. In the
swing phase there was less knee flexion in the HH condition [10].
3.4. The hip in high heeled gait
Simonsen et al. [16] found that in the frontal plane the hip ab-
duction moment increased moderately but with a significant effect in
the HH condition just after heel strike and just before the push-off phase
and in the middle of stance phase compared to the barefoot condition.
Barkema et al. [2] found a hip abduction moment during most of the
stance phase, except at heel strike. In both the conditions of preferred
and fixed speed, the hip abduction moment patterns resembled each
other. Only the hip abduction peak just after heel strike, in fixed
walking speed, in the HH condition appeared significantly larger
compared to that in the other shoe conditions [2]. In the study of
Esenyel et al. [6] the hip flexion moment in the sagittal plane prolonged
in the HH condition during late stance phase. This resulted in a 23%
increase in concentric hip flexor muscle work. In the frontal plane an
increase of 11% in peak hip abduction was found in the HH condition,
however this effect was not found to be significant. Kerrigan et al. [10]
found a larger hip flexor moment during early stance in walking with
HH. Hip angles were larger during stance phase in the HH condition,
except for the time just after heel strike. For a brief moment the hip
angles were equal in size [16]. Also the hip flexion showed a greater
peak in stance in the HH condition compared to the barefoot condition
[10]. Mika et al. [13] found that middle-aged women showed sig-
nificantly lower hip RoM values compared to the younger group in the
sagittal plane.
3.5. The pelvis in high heeled gait
The study of Mika et al. [12] described the pelvic tilt angle, pelvic
range of motion (RoM), and pelvic alignment angle at initial contact
and toe-off. During HH gait the group of young women exhibited an
increase in pelvic RoM in the sagittal plane compared to LH and
barefoot condition. The middle-aged women did not show this increase
of RoM.
425
Gait & Posture 61 (2018) 423–430
3.6. The lumbar spine in high heeled gait
In the study of Lee at al. [11], in which the lumbar flexion angle
(expressing lordosis) was examined, they found that this angle (ap-
proximately 150°) at stance increased with an additional 16° while
walking. During the walking conditions they found that the lumbar
flexion angle decreased approximately 1° for each 1 cm increase in heel
height. In two studies (Mika et al. [12] and Lee at al. [11]) the effect of
HH and LH gait on the erector spinae muscle activity together with the
kinematics of the pelvis was examined. A significant increase in activity
of the erector spinae muscles in the LH and HH condition at toe-off and
in the initial contact phase was found, compared to the barefoot con-
dition. The middle-aged women showed at toe-off and in the initial
contact phase a significant increase of activation in these muscles only
in the HH condition. Lee at al. [11] found a significant correlation
between the EMG data of the erector spinae muscles of the L4/L5 seg-
ment and the vertical displacement of the CoM.
3.7. The shoulders and neck in high heeled gait
Annoni et al. [1] reported that at heel strike the shoulders were
more inclined at heel strike in the frontal plane. Mika et al. [14] were
interested in the effect of walking in HH on the cervical paraspinal
muscle activity. They measured two different age groups of women
while walking in three different conditions. The women were instructed
to walk barefoot, in an LH and in HH condition while surface electro-
myography (sEMG) was used to record muscle activity. In both the
groups of women, Mika et al. [14] found that in the HH condition
muscle activation of the cervical paraspinal muscles was significantly
larger at toe-off and at heel strike compared to the barefoot condition.
3.8. The center of mass, ground reaction forces and gait parameters in high
heeled gait
Annoni et al. [1] and Lee at al. [11] assessed the modifications of
the CoM movement in walking in HH compared to LH. They found an
increase in vertical displacements of the CoM with increasing HH.
Vertical velocity modifications (time curve changes), were found while
the CoM remained in a more anterior position. During the whole gait
cycle the average position of the CoM was 3% lower in HH-gait nor-
malized to body length, including the shoes. The velocity of the CoM, at
heel strike, in the vertical direction did not differ between the two
conditions. During the rest of the cycle it reached larger absolute values
in the HH condition. Stefanyshys et al. [19] found larger braking forces
as the heel height increased, causing a greater deceleration of the CoM.
This deceleration of the CoM was counteracted by a moderate increase
in peak propulsion force at push off. With an increasing heel height,
they found the CoM in a more anterior position. Chien et al. [3] in-
vestigated the influence of the surface of the base and height of the
heels on the motion of the CoM. This was done by studying the in-
clination angles (IA) of the CoP (center of pressure)-CoM line with re-
spect to the vertical. They also studied the rate of change of the in-
clination angle (RCIA). They compared the results of barefoot walking
and walking with LH and HH shoes. Chien et al. [3] and Rao et al. [15]
found that gait speed and stride length decreased in the shoe conditions
compared to bare foot walking. Stance time and double leg support
(DLS) time were increased in all the shoe conditions. Only in the HH
condition the single leg support (SLS) time was shorter compared to the
barefoot condition. During DLS, the combination of heel height and the
reduced base of support (BoS) contributed to reduced RCIA in the
frontal and sagittal plane. In the sagittal plane they found a similar
strategy of RCIA control. However, after the first half of DLS they found
that the sagittal RCIA was increased. At increased walking speed in HH
gait Rao et al. [15] found an increased knee power generation in the
first half of stance phase in HH gait compared to LH gait. A decrease in
hip flexion was associated with an increase of walking speed in HH gait.
M.M. Wiedemeijer, E. Otten Gait & Posture 61 (2018) 423–430

The determinants of increased stride length in the sagittal plane were

mainly an increased ankle power generation, an increased hip range of
motion (RoM), and a decreased hip extension moment. Barkema et al.
[2] found that stride length (SL) decreased and cadence increased sig-
nificantly as heel height increased. Cronin et al. [4] compared the
spatio-temporal gait parameters between the control and within the
two groups of which one group consisted of habitual wearers of HH
shoes. The HH group showed a significantly shorter step length in the
HH condition compared to barefoot walking. Esenyel et al. [6] also
found that walking in HH decreased the velocity of walking by 6% from
1.3 to 1.2 m/s because of a stride length reduction from 1.37 to 1.31 m.
During limb loading the peak vertical ground reaction force (GRF) was
5% larger in the HH condition, although this effect did not appear to be
significant. Stefanshyn et al. [19] found larger ground reaction forces
(GRF's) in the anterior and posterior directions, which corresponded
with an increase in the peak deceleration and acceleration forces in the Fig. 2. (A) The displacement of the heel requires a higher internal dorsiflexor moment at
vertical direction in HH. The HH condition had the smallest vertical heel strike, a2 > a1. (B) The forefoot is closer to the ankle joint in the HH condition
causing a decreased plantar flexion moment, a1 > a2.
ground force at heel strike and during loading while the LH condition
showed a significant larger vertical force. The time of propulsion in the
final stance phase was shorter compared to barefoot and decreased even
more with an increasing heel height [5]. The support time in the
barefoot condition was significantly shorter than in all the shoe con-
ditions [5,17]. Midstance showed to be more prolonged in the HH
condition [5]. Simonsen et al. [16] found no significant differences
between the two groups and between the two heel conditions regarding
the duration of the gait cycle, the duration of the stance phase, the
length of the stride, velocity and cadence.

4. Discussion

Although the quality of the papers showed some variation, the final
selection of 19 papers showed a coherent picture, so that variation was
of little importance. From this coherent picture the following in-
tegrative account can be given.
A single change in posture of the foot due to wearing high heels,
may affect the whole gait pattern. Gait being a system property does not Fig. 3. (A) The HH condition elongates the leg and effect the moments around the knee
have a particular order of events. For convenience, the following order with k2 > k1 and the hip with h2 > h1. (B) The gastrocnemius muscle activation
generates a knee flexion moment, counteracted by the rectus femoris muscle which also
of discussion will be used: biomechanically from feet upward and after
creates a lordosis.
that the CoM and gait parameters will be covered.

4.1. CoP and balance
The vertical impact force decreases at heel strike in the HH condi-
tions [19]. Due to more flexed knees, the leg is more compliant.
Compliance of a leg is the inverse of its stiffness. When legs are more
bent, moment arms of the ground reaction force increase, making the
leg more compliant. Flexion in the knee will reduce the vertical impact
force. Moreover, possibly a late weight transfer occurs in the early load
bearing phase of the gait cycle. The peak of the vertical impact force
during limb loading increases with 5% (although this was not with a
significant effect) [6]. It could be that this is the result of a faster weight
transfer and corresponds with the findings of a reduction of SLS time
[3,5]. A reduction of SLS time is the result of a balance strategy. Some
studies mention a longer midstance and shorter toe-off (late stance)
phase [3,5,18]. The time for the foot to roll from the heel to the forefoot
is shorter. Because of this, less balance deviations in the medio-lateral
direction could be the result. Time to maximum pressure under the
hallux increased, and the contact area under the medial forefoot and
hallux increased [17,18]. This corresponds also with a need for balance
in the medio-lateral direction. Maximum pressure under the forefoot
increases most under the hallux, followed by medial forefoot, central
forefoot and lateral forefoot [17,18]. Pressure time integral and force
time integral decreased in midfoot [18]. This is related with a weight
transfer from heel to forefoot in a shorter time, which improves bal-
ance. Balance is partly controlled by the control of the CoP. Since that
CoP is now closer to the new foot position, balance is improved. Also
this explains the found patterns of co-activation of the tibialis anterior
muscle and soleus muscle [4]: they serve anterior–posterior balance. In
early stance, a larger knee extension moment was seen [6] and also
during the rest of the stance phase [9,16,20]. Before toe-off [16] and at
toe-off [20] extension moments decrease. Both sets of observations can
be explained, like the above, by an early shift of the CoP to the forefoot.
4.2. The ankle joint
An increase in plantar flexion was found throughout the cycle [10]
and the medial gastrocnemius muscle activation is prolonged during
stance phase [4]. This is an inevitable result from the fact that the heel
height forces the foot in a more plantar flexed orientation. The found
increased plantar flexion in heel strike [13] is needed to overcome the
heel height during the placement of the foot on the ground. This is
followed by a greater internal dorsiflexor moment at early stance [16].
This greater dorsiflexor moment is needed due to the increased lever
arm from the GRF vector to the ankle joint in the HH condition

426
M.M. Wiedemeijer, E. Otten
(Fig. 2A). A reduced plantar flexion moment was found at late stance
[4,6,10,16,19]. With a higher heel, the forefoot is closer to the ankle
joint, which causes the plantar flexion moment to decrease. This ex-
plains also the lower plantar flexion moment found at toe-off [16]
(Fig. 2B) and the more posterior found HH rollover curve, since most of
the rollover curve is formed by the forefoot rolling motion [8]. The RoM
of the ankle joint decreases in the frontal and the sagittal plane [4,13].
In the sagittal plane this is due to the already plantar flexed foot and co-
activation of the tibialis anterior muscle and the triceps surea muscles.
In the frontal plane in early stance this can be attributed to the reduced
heel base which does not allow rotation of the foot without instability.
This decrease of ankle RoM is compensated with a knee power gen-
eration increase [15]. There is an increase in knee RoM in mid stance
phase and more power production [4].
4.3. The knee joint
The rectus femoris muscle and medial gastrocnemius muscle acti-
vations increased at heel strike and toe-off [13,19]. These muscles are
antagonists. Because the medial gastrocnemius muscle originates from
the medial condyle of the femor bone, knee flexion is one of its func-
tion. Knee flexion is found to increase [9,12,14,13,19,20]. The rectus
femoris muscle is activated to prevent too much knee flexion at heel
strike. At toe-off the rectus femoris muscle is activated to form an an-
tagonist against the medial gastrocnemius muscle, with is active for
plantar flexion of the foot but is also a knee flexing muscle. The reduced
RoM at the ankle joint (due to plantarflexion), is also compensated at
the knee joint by introducing knee flexion. The compliance of the ankle
is reduced due to the heel height. The foot is in a more plantar flexed
position. This reduces the moment lever arm of the ground reaction
force, making the joint less compliant, which is compensated by an
increased compliance at the knee joint. During stance a decreased and
prolonged knee extensor moment was found [10]. Again this can be
attributed to more knee flexion in midstance. In the swing phase less
knee flexion occurs [10]. This is made possible by an elevation of the
hip of the swing leg.
4.4. The hip joint
An increased internal knee abduction moment is present just after
heel strike and before toe-off [2,10,16,20]. The hip abduction moment
increases during most of the stance phase, except at heel strike [2,6,16].
These two findings are due to the fact that the leg is elongated with the
height of the heel (Fig. 3A). During stance more hip flexion occurs and
is prolonged during late stance phase [6,10]. This is associated with an
increased lumbar flexion (resulting in lordosis, see below). The found
increase of RoM of the hip in the frontal plane [16] is due to an in-
creased anterior pelvic roll.
4.5. CoM and gait parameters
The CoM makes larger vertical displacements [1,11] during the gait
cycle. Stefanyhys et al. [19] found larger anterior–posterior braking
forces as the heel height increased, causing a greater deceleration of the
CoM. This can be explained by an early shift of the CoP to the forefoot.
Thereby the CoP is more in front of the CoM. Gait becomes less fluent
Appendix A
See Table A1.
427
Gait & Posture 61 (2018) 423–430
with increasing heel height because of the breaking and propulsion
forces. The CoM is found in a more anterior position [1,19] and the
CoM was found to be 3% lower [1] compared to the LH condition. The
lowering of the CoM might be explained by the fact that during HH gait
a combination occurs of lordosis and knee flexion. Due to prolonged
period of CoP under the forefoot the CoM moves to a more anterior
position. The cervical muscles [14] and the erector spinae muscle ac-
tivation was found to increase at the height of L4/L5 [11] and L3/L4
[12]. As explained, larger decelerations and acceleration are found in
HH gait in anterior–posterior direction [19]. This implies that the mass
of the trunk and head have to be stabilized in anterior–posterior di-
rection. It is reported that velocity and stride length decreases, while
cadence remains the same [1–4,6,15,20]. As explained before, the early
shift of the CoP to the forefoot may reflect a need for balance. This early
shift is only possible with a limited step length. At an increase of knee
flexion, the gastrocnemius muscle pulls on the femur bone. The rectus
femoris muscle is activated as a counter force, to prevent the knee from
flexing too much. The rectus femoris muscle originates from the ante-
rior inferior iliac spine, and when active pulls at the pelvis, increasing
lumbar lordosis. Indeed Lee at al. [11] found that lumbar flexion de-
creased with increasing heel height (Fig. 3B).
With a single change in footwear the whole gait pattern changes.
Due to an increase in heel height an increase in plantar flexion of the
foot is enforced. An increase in plantar flexion causes an increase in
knee flexion. This knee flexion has three consequences:
1. The loss in ankle joint power production is compensated by an in-
crease in knee power production.
2. An increased knee flexion enhances shock absorption which is par-
tially lost in the ankle joint.
3. More knee flexion allows for an early forefoot loading which im-
proves balance.
This knee flexion also invites an anterior pelvic tilt, which in its turn
increases lordosis and lumbar muscular activity.
Walking in high heels compromises gait, both in terms of perfor-
mance and balance. Our conclusions could be further substantiated and
our insight into the effect of HH on the dynamics of gait increased by
future studies with experimental work on balance control, for instance
by using an unstable computer controlled floor and the effect of redu-
cing gravity.
5. Conclusion
Gait patterns are system properties. Therefore, a single change, like
increasing heel height, changes the whole pattern. It is shown that both
performance and balance are compromised when walking in HH. The
most significant changes can be found around the ankle and knee joints.
The consulted literature shows good coherence in the changes mea-
sured, which can be explained from a biomechanical and control point
of view.
Conflict of interest
The authors declare no conflicts of interest.
 Table A1
Measuring methods and focus of research of the studies. Test conditions and outcome measures of the studies.

Reference number Annoni 2014 Lee 2001 Chien 2013 Chien 2014 Rao 2013 Barkema Cronin 2012 Esenyl 2003 Mika 2011 Mika 2012a Gefen 2001
2011

Number of subjects 13 5 15 30 31 15 19 15 31 46 8
M.M. Wiedemeijer, E. Otten

Measure method
EMG x x x x x
Ultrasound x
EDG
Force plates x x x x x x
Optical motion capture x x x x x x x x x

Test conditions
cm HH 10 8 7.3 7.3 7.6 9 11 6 10 10 no
cm HH 1 4.5 6.3 1.9 sp.s 4 4
cm HH 3.9
cm HH
cm HH B B B B B B
Own shoes x x x’ x
Heel base in cm2 1 nn 3.2 3.2 1 30.3 1 1
Heel base in cm2 3.2 3.2
Heel base in cm2 3.2 3.2
Walking velocities ss 1.1 ss ss ss ss/1.3 ss ss ss ss
Wide vs narrow heel
Increased weight

Outcome measures
CoP x x x x

428
CoM–CoP x x x x
CoM x x x x x x x
CoP/BoS x x x x
Forefoot pressure x x
Roll over characteristics
Muscle activity x x x
Medio-lateral stability x
Step length x x x x x x x
Step width x
Walking velocity/cadence x x x x x x x x
Stance phase characteristics
Muscle activity x x x x x x x x
Ankle angles/moments x x
Knee angles/moments x x x x x x x x
Hip/pelvic angle/moments x x x x x x x
Trunk angle/moments x x x x x x x
Stability/balance x x x x x x x

Reference number Mika 2012b Stefanyshyn Speksnijder Snow 1992 Eisenhardt Hansen Simonsen Kerrigan Titchenal Ho 2012 Kerrigan
2000 2005 1995 2004 2012 1998 2014 2001

Number of subjects 46 13 10 45 30 10 14 20 14 11 20

Measure method
EMG x x x
Ultrasound
EDG x
Force plates x x x x x x x
Optical motion capture x x x x x x

Test conditions
Gait & Posture 61 (2018) 423–430
 Table A1 (continued)

Reference number Mika 2012b Stefanyshyn Speksnijder Snow 1992 Eisenhardt Hansen Simonsen Kerrigan Titchenal Ho 2012 Kerrigan
2000 2005 1995 2004 2012 1998 2014 2001

cm HH 10 8.5 5.9 8.3 8.7 7.1 9 6 8.3 9.5 7

cm HH 4 5.4 2 5.1 5.72 3.7 3.8 1.3 7
cm HH 3.7 1.9 3.12 0 sp.s 6.3
M.M. Wiedemeijer, E. Otten

cm HH 1.4 1.75
cm HH B B B B B B
Own shoes x’ x x
Heel base in cm2 32.3 0.7 1 1 1 1 1.2 wide
Heel base in cm2 50.7 1.8 4.5 wide
Heel base in cm2 1 16.6
Walking velocities ss 1.4 ss 1.4 nn sl-ss-fst 1.1 ss ss ss ss
Wide vs narrow heel x
Increased weight x

Outcome measures
CoP x x
CoM–CoP
CoM x x x
CoP/BoS x x x
Forefoot pressure x x x x x
Roll over characteristics x
Muscle activity x
Medio-lateral stability
Step length x x
Step width x
Walking velocity/cadence x
Stance phase characteristics x v

429
Muscle activity x x x x
Ankle angles/moments x x x
Knee angles/moments x x x x x x
Hip/pelvic angle/moments x x x
Trunk angle/moments x
Stability/balance x

nn = not known, x’ = time to practice with the shoes, B = barefoot, sp.s = sport shoes, ss = self selected, sl = slow and fst = fast.
Gait & Posture 61 (2018) 423–430
M.M. Wiedemeijer, E. Otten
References
[1] I. Annoni, A. Mapelli, F.V. Sidequersky, M. Zago, C. Sforza, The effect of high-
heeled shoes on overground gait kinematics in young healthy women, Sport Sci.
Health 10 (2014) 149–157.
[2] D.D. Barkema, T.R. Derrick, P.E. Martin, Heel height affects lower extremity frontal
plane joint moments during walking, Gait Posture 35 (2012) 483–488.
[3] H.L. Chien, T.W. Lu, M.W. Liu, Control of the motion of the body's center of mass in
relation to the center of pressure during high-heeled gait, Gait Posture 38 (2013)
391–396.
[4] N.J. Cronin, R.S. Barrett, C.P. Carty, Long-term use of high-heeled shoes alters the
neuromechanics of human walking, J. Appl. Physiol. 112 (2012) 1054–1058.
[5] J.R. Eisenhardt, D. Cook, I. Pregler, H.C. Foehl, Changes in temporal gait char-
acteristics and pressure distribution for bare feet versus various heel heights, Gait
Posture 4 (1996) 280–286.
[6] M. Esenyel, K. Walsh, J.G. Walden, A. Gitter, Kinetics of high-heeled gait, J. Am.
Podiatr. Med. Assoc. 93 (2003) 27–32.
[7] N. Guéguen, High heels increase women's attractiveness, Arch. Sex. Behav. 44
(2015) 2227–2235.
[8] A.H. Hansen, D.S. Childress, Effects of shoe heel height on biologic rollover char-
acteristics during walking, J. Rehabil. Res. Dev. 41 (2004) 547.
[9] K.Y. Ho, M.G. Blanchette, C.M. Powers, The influence of heel height on patellofe-
moral joint kinetics during walking, Gait Posture 36 (2012) 271–275.
[10] D.C. Kerrigan, M.K. Todd, P.O. Riley, Knee osteoarthritis and high-heeled shoes,
Lancet 351 (1998) 1399–1401.
[11] C.M. Lee, E.H. Jeong, A. Freivalds, Biomechanical effects of wearing high-heeled
430
Gait & Posture 61 (2018) 423–430
shoes, Int. J. Ind. Ergon. 28 (2001) 321–326.
[12] A. Mika, L. Oleksy, P. Mika, A. Marchewka, B.C. Clark, The effect of walking in
high-and low-heeled shoes on erector spinae activity and pelvis kinematics during
gait, Am. J. Phys. Med. Rehabil. 91 (2012) 425–434.
[13] A. Mika, Ł. Oleksy, P. Mika, A. Marchewka, B.C. Clark, The influence of heel height
on lower extremity kinematics and leg muscle activity during gait in young and
middle-aged women, Gait Posture 35 (2012) 677–680.
[14] A. Mika, Ł. Oleksy, E. Mikołajczyk, A. Marchewka, P. Mika, Changes of bioelectrical
activity in cervical paraspinal muscle during gait in low and high heel shoes, Acta
Bioeng. Biomech. 13 (2011) 27–33.
[15] S. Rao, R. Ripa, K. Lightbourne, Predictors of walking speed and stride length in
high-and low-heeled footwear, Footwear Sci. 5 (2013) 179–184.
[16] E.B. Simonsen, M.B. Svendsen, A. Norreslet, H.K. Baldvinsson, T. Heilskov-Hansen,
P.K. Larsen, T. Alkjær, M. Henriksen, Walking on high heels changes muscle activity
and the dynamics of human walking significantly, J. Appl. Biomech. 28 (2012)
20–28.
[17] R.E. Snow, K.R. Williams, G.B. Holmes, The effects of wearing high heeled shoes on
pedal pressure in women, Foot Ankle Int. 13 (1992) 85–92.
[18] C.M. Speksnijder, R.J. vd Munckhof, S.A. Moonen, G.H. Walenkamp, The higher the
heel the higher the forefoot-pressure in ten healthy women, Foot 15 (2005) 17–21.
[19] D.J. Stefanyshyn, B.M. Nigg, V. Fisher, B. O Flynn, W. Liu, The influence of high
heeled shoes on kinematics, kinetics, and muscle EMG of normal female gait, J.
Appl. Biomech. 16 (2000) 309–319.
[20] M.R. Titchenal, J.L. Asay, J. Favre, T.P. Andriacchi, C.R. Chu, Effects of high heel
wear and increased weight on the knee during walking, J. Orthop. Res. 33 (2015)
405–411.