Clinical Biomechanics 29 (2014) 1056–1062

Contents lists available at ScienceDirect

Clinical Biomechanics
journal homepage: www.elsevier.com/locate/clinbiomech

The immediate effects of foot orthoses on hip and knee kinematics and
muscle activity during a functional step-up task in individuals with
patellofemoral pain
Simon Lack, Christian Barton, Roger Woledge, Markus Laupheimer, Dylan Morrissey ⁎
Centre for Sports and Exercise Medicine, William Harvey Research Institute, Bart's and the London School of Medicine and Dentistry, Queen Mary University of London, England, United Kingdom

a r t i c l e i n f o a b s t r a c t

Article history: Background: Evidence shows that anti-pronating foot orthoses improve patellofemoral pain, but there is a paucity
Received 30 April 2014 of evidence concerning mechanisms. We investigated the immediate effects of prefabricated foot orthoses on
Accepted 14 August 2014 (i) hip and knee kinematics; (ii) electromyography variables of vastus medialis oblique, vastus lateralis and
gluteus medius during a functional step-up task, and (iii) associated clinical measures.
Keywords: Methods: Hip muscle activity and kinematics were measured during a step-up task with and without an anti-
Patellofemoral pain
pronating foot orthoses, in people (n = 20, 9 M, 11 F) with patellofemoral pain. Additionally, we measured
Orthoses
Onset times
knee function, foot posture index, isometric hip abductor and knee extensor strength and weight-bearing
Kinematics ankle dorsiflexion.
Electromyography Findings: Reduced hip adduction (0.82°, P = 0.01), knee internal rotation (0.46°, P = 0.03), and decreased
gluteus medius peak amplitude (0.9 mV, P = 0.043) were observed after ground contact in the ‘with orthoses’
condition. With the addition of orthoses, a more pronated foot posture correlated with earlier vastus medialis
oblique onset (r = −0.51, P = 0.02) whilst higher Kujala scores correlated with earlier gluteus medius onset
(r = 0.52, P = 0.02).
Interpretation: Although small in magnitude, reductions in hip adduction, knee internal rotation and gluteus
medius amplitude observed immediately following orthoses application during a task that commonly aggravates
symptoms, offer a potential mechanism for their effectiveness in patellofemoral pain management. Given the
potential for cumulative effects of weight bearing repetitions completed with a foot orthoses, for example during
repeated stair ascent, the differences are likely to be clinically meaningful.
© 2014 Published by Elsevier Ltd.

1 . Introduction up (Stathopulu and Baildam, 2003), highlighting the need for more ef-
fective management plans to be identified. With the aetiology of PFP
Patellofemoral pain (PFP) is one of the most common presentations widely accepted to be multifactorial in nature (Powers et al., 2012),
in recreationally active and sporting populations (Baquie and Brukner, these poor long-term outcomes may represent a failure to address the
1997; Taunton et al., 2002). Of 2429 injury presentations to a sports specific deficits contributing to the development and persistence of PFP.
medicine clinic over a 12 month period, 668 (27.5%) cases affected the Anti-pronating foot orthoses (APFOS) are commonly prescribed to
knee, with PFP reported to be the most common knee complaint individuals with PFP, and have been reported to effectively reduce
(Baquie and Brukner, 1997). Furthermore, a study of 2002 running pain and improve function (Collins et al., 2008; Eng and Pierrynowski,
injuries over a two-year period in a sports medicine clinic, reported 1993; Mills et al., 2012a). However the mechanism for their effective-
331 patients (16.5%) were diagnosed with PFP (Taunton et al., 2002). ness is poorly understood (Barton et al., 2010). Tiberio (1987) proposed
PFP is commonly aggravated by stair ascent and descent, squatting, sit- that excessive sub-talar joint pronation may lead to greater tibial and
ting for long periods and high impact activity such as running (Kujala hip internal rotation, and consequently increased lateral tracking and
et al., 1993). Despite its high prevalence and positive short term treat- loading of the patellofemoral joint (PFJ). This proposed kinematic cou-
ment outcomes (Collins et al., 2008; Crossley et al., 2002), 80% of indi- pling between lower limb segments has been supported by reports
viduals who complete a rehabilitation programme for PFP still report that greater peak rearfoot eversion is associated with greater tibial
pain, and 74% report a reduction in physical activity at 5 year follow internal rotation during walking in individuals with PFP (Barton et al.,
2012). It is proposed the APFOS may prevent these aberrant movement
⁎ Corresponding author at: Mile End Hospital, Bancroft road, London, England E1 4DG,
patterns and hence reduce pain associated with PFP (Tiberio, 1987).
United Kingdom. Step negotiation was chosen to explore lower limb biomechanics in
E-mail address: [email protected] (D. Morrissey). PFP populations due to higher loading forces reported within the PFJ

http://dx.doi.org/10.1016/j.clinbiomech.2014.08.005
0268-0033/© 2014 Published by Elsevier Ltd.
 S. Lack et al. / Clinical Biomechanics 29 (2014) 1056–1062 1057

during this activity (Andriacchi et al., 1980) and patients commonly inclusion based on; (1) age 18–40 years; (2) insidious onset of anterior
reporting symptoms with stairs (Kujala et al., 1993). Contrary to pre- knee or retropatellar pain of greater than six weeks' duration;
vious findings that reported changes in symptomology (Eng and (3) provoked by at least two of prolonged sitting or kneeling, squat-
Pierrynowski, 1993), foot and knee (Eng and Pierrynowski, 1994) ting, running, hopping, or stair walking; (4) tenderness on palpation
and hip (Lack et al., 2014) kinematics resulting from foot orthoses, of the patella, or pain with step down or double leg squat; and
Mills et al. (2012b) reported no significant changes to hip or knee ki- (5) worst pain over the previous week of at least 30 mm on a 100 mm
nematics in individuals with PFP during running with the addition of visual analogue scale. Exclusion criteria were; (1) concomitant injury
APFOS (Mills et al., 2012b). The lack of consistent findings between or pain from the hip, lumbar spine, or other knee structures; (2) previ-
studies potentially highlights the multifactorial nature of the condi- ous knee surgery; (3) patellofemoral instability; (4) knee joint effusion;
tion, differences in biomechanical response to orthoses during (5) any foot condition that precluded use of foot orthoses; (5) physio-
differing tasks, or possibly a delay in the influence of APFOS on kine- therapy or foot orthoses treatment within the previous year; and
matic variables in individuals with PFP. (6) use of anti-inflammatory drugs (Collins et al., 2008). Ethical approv-
Another proposed mechanism for foot orthoses effectiveness is al- al was obtained from Queen Mary University Ethics of Research Com-
tered neuromotor control. Nigg et al. (1999) proposed that an orthoses mittee and each participant provided written informed consent.
that supports a preferred movement path could minimise muscle activ-
ity and reduce fatigue by providing input through the sole (Nigg et al., 2.2 . Clinical measures
1999). Individuals with PFP have been reported to frequently possess al-
tered neuromotor control with delayed onsets of vastus medialis 2.2.1 . Kujala patellofemoral score
oblique (VMO) (Chester et al., 2008; Lankhorst et al., 2013) and gluteus The Kujala patellofemoral score (KPS) is a 13-item questionnaire
medius (GMed) (Barton et al., 2013) muscle. Despite these identified categorising symptoms and current knee function, such as the ability
deficits and theoretical rationale for foot orthoses to address them, a to negotiate stairs, walk, run, jump and sit for prolonged periods. Each
paucity of research exploring the effects of orthoses on neuromotor var- item is weighted and a total score between 0 and 100 calculated, with
iables exists (Mills et al., 2010). Recent studies exploring proximal higher scores representing greater levels of function (Kujala et al.,
neuromotor effects of APFOS have reported no immediate changes in 1993).
gluteal and quadriceps muscle onsets or amplitudes during running in
individuals with PFP (Mills et al., 2012b) or during a functional step- 2.2.2 . Foot posture index
up task in asymptomatic individuals (Lack et al., 2013). However, with Methods for measuring foot posture index (FPI) have been reported
electromyography (EMG) changes described as being highly variable previously (Lack et al., 2013). Briefly, the lead author with established
within a heterogeneous population (Mundermann et al., 2006) and excellent intra-tester reliability (ICC = 0.94) assessed static foot pos-
PFP widely regarded as having a multifactorial aetiology (Powers ture (SL). Participants were instructed to march on the spot and then
et al., 2012), further work exploring the association of specific EMG stand in a comfortable position as the examiner assessed the dominant
changes with clinically applicable measures is clearly warranted. foot. A score from −12 to +12 was obtained with scores between 0 and
The primary aim of this study was to explore the immediate effects 5 normal, 6 and 9 pronated, 10+ highly pronated, −1 and −4 supinat-
of prefabricated foot orthoses on (i) hip and knee kinematics; and ed and −5 and −12 highly supinated.
(ii) electromyography (EMG) variables of VMO, vastus lateralis (VL)
and GMed. The secondary aim of this study was to identify clinical mea- 2.2.3 . Ankle dorsiflexion range
sures that may be associated with these changes. Weight bearing, knee straight (KSAD) and knee bent (KBAD) ankle
dorsiflexion range was measured using digital inclinometer methods
2 . Methods previously described (Lack et al., 2013). The long axis of the foot was
aligned with a taped line on the floor perpendicular to a wall. Partici-
Symptomatic participants had biomechanical data collected at the pants lunged as far forward as possible whilst keeping the heel on the
knee and hip during a functional step-up task. Clinical measures were ground. The largest KSAD and KBAD angle of the three measures was re-
obtained prior to testing and subsequently analysed to determine any corded from the inclinometer (Baseline® Digital Inclinometer 12-1057;
correlation with changes observed due to orthoses application. Fabrication Enterprises Inc, New York, USA) placed on the anterior tibia
with the knee fully straight and maximally bent respectively.
2.1 . Participants
2.2.4 . Others
Twenty individuals (9 M 11 F; Table 1) were recruited to participate Orebro musculoskeletal pain questionnaires were administered to
in the study through referral from private sports medicine clinics in evaluate the risk of long-term disability (Linton and Boersma, 2003),
greater London. A sports physician or registered physiotherapist with and hip abduction and knee extension maximum voluntary isometric
over 5 years clinical experience assessed all potential participants for strength were measured using a hand held dynamometer method
described previously (Bohannon, 1986).

Table 1 2.3 . Electromyographic recordings

Patient demographics and clinical measures.
Values are mean (SD) unless otherwise stated.
Muscle activity of VMO, VL and GMed of the participant's affected leg
Measure N = 20 was recorded by wireless surface electromyography (sEMG) (Telemyo
Age 28.5 years (4.2) 2400 T G2, Noraxon, USA). Where subjects reported bilateral symptoms,
Height 171.9 cm (7) their most symptomatic leg was used. The subject's skin was prepared
Weight 64.8 kg (9.7) and pairs of Ag/AgCl surface electrodes with an intra-electrode distance
Kujala patellofemoral score (Median (IQR)) 80 (10.75)
of 20 mm (Tyco Healthcare, Germany) were placed over the muscles of
Orebro SCORE (Median (IQR)) 63 (20.75)
Foot posture index 5.4 (3.2) interest according to standard SENIAM guidelines (Freriks et al., 2000).
Knee straight ankle dorsiflexion 39.4° (5.8) The GMed electrode was placed halfway along the line between the iliac
Knee bent ankle dorsiflexion 45° (6.6) crest and greater trochanter, orientated vertically. The VMO electrode
Hip abduction strength 26.2 kg (6.2) was placed at 80% of the distance down the line between the anterior
Knee extension strength 30.8 kg (7.4)
superior iliac spine (ASIS) and the medial knee joint line just anterior
1058 S. Lack et al. / Clinical Biomechanics 29 (2014) 1056–1062

to the medial ligament, orientated ~55° to the vertical. The VL electrode

was placed 2/3 of the distance down the line from the ASIS to the lateral
patellar border, orientated ~ 15° to the vertical. Surface EMG signals
were sampled at 1500 Hz, pre-amplified and band-pass filtered be-
tween 10 and 500 Hz, prior to export to Matlab (version 2009a,
Mathworks, Natwick, MA, USA) for post-processing.

2.4 . Kinematics

During the 5-minute rest following collection of clinical measures,

participants were fitted with the motion capture equipment (Fig. 1). A
modified Helen-Hayes marker protocol was used (Kadaba et al., 1990)
to place active infra-red markers bilaterally over the ASIS and PSIS on
the pelvis, lateral femoral condyle, lateral malleolus, and on the outside
of the shoe in locations to best represent the lateral calcaneus and fifth
metatarsal head. Marker mounting wands were placed over the lateral
femur, and at the level of the tibial tuberosity. Movement data was cap-
tured using four CodaMotion Cx1 sensor units (Charnwood Dynamics,
Rotheley, Leicestershire, UK) sampling at 200Hz.
Fig. 2. Prefabricated foot orthoses (Vasyli Easy Fit, Vasyli International, UK) prescribed to
each participant.
2.5 . Step-up task

Methods for collecting data during the step-up task have been
reported previously (Lack et al., 2013). Briefly, participants stepped a N 10 N threshold) was set up, within which kinematic and EMG data
up onto a Kistler force plate (Type 9281CA, Kistler Corporation, were further analysed.
Switzerland) mounted on a wooden step (combined height 22 cm), in Peak EMG amplitude values were extracted within this time win-
response to a verbal command. The symptomatic leg was always the dow for each subject and averaged across the five trials for each test
lead leg. Participants were randomised into either APFOS–no APFOS or condition.
no APFOS–APFOS test groups by a coin toss. Data was collected during Muscle onsets of GMed, VMO and VL were identified using a novel
five separate repetitions for each test condition, with a 30 second sitting algorithm. No previously identified method of onset determination
break between the two test conditions. Participants wore standardised has been reported as optimal (Uliam Kuriki et al., 2011), and attempts
neutral footwear (Asics Nimbus, Asics, Cheshire, UK) during all test con- to implement on our data were unsuccessful. A rise of EMG activity
ditions. The APFOS devices were unmodified prefabricated 6° varus above a predetermined threshold that was maintained for a period of
posted orthoses (Vasyli easy fit, Vasyli, Essex, UK) (Fig. 2) designed to N30 ms was described as muscle onset. The threshold was calculated
reduce rearfoot pronation, and placed directly under the heel when from the minimum of the means of all trials plus 10% of the range
required. (maxima of means of all trials minus minima of means of all trials).
The algorithm-determined muscle onset was imposed onto the EMG
2.6 . Data analysis record to allow for visual confirmation. Negative values represent mus-
cle onset prior to foot contact, and a positive value was subsequent to
The raw EMG was rectified and smoothed using a 0.02 s running me- foot contact.
dian method. A time window 0.5 s prior to and 0.5 s post initial contact Kinematic data across the three planes and two joints (hip and knee)
(determined from the first positive change in the force record exceeding were averaged across the five trials for each subject, and were extracted

Fig. 1. Demonstrating the CODA motion and electromyography (EMG) set up during completion of the step-up task.
 S. Lack et al. / Clinical Biomechanics 29 (2014) 1056–1062 1059

at four time points (− 100 ms, 0 ms, + 100 ms and + 200 ms) either 3.4 . Relationship of clinical measures to biomechanical changes induced by
side of initial contact for further analysis between conditions. APFOS application

A more pronated foot posture, determined using the FPI, correlated

2.7 . Statistical analysis with greater change to earlier VMO onset with the addition of orthoses
(r = −0.51, P = 0.02) (Fig. 4). Higher Kujala scores, indicating greater
Statistical analysis was performed using SPSS (version 18.0, SPSS Inc, function and less pain, were associated with a greater change to earlier
Chicago, Il). Between-conditions comparisons were made at the four GMed onset with the addition of orthoses (r = 0.52, P = 0.02) (Fig. 5).
extracted time points using paired t-tests for GMed, VMO and VL Ankle dorsiflexion range, hip abduction and knee extension strength,
onset times and peak amplitudes, and for peak hip adduction, hip inter- and Orebro score did not correlate with any of the biomechanical
nal rotation angles, knee adduction and knee internal rotation angles. variables.
Effect size was calculated for significant post hoc findings as a ratio of
mean difference divided by the between-subject standard deviation. It 4 . Discussion
has been suggested that effect size of 0.20 or less represents small
change; 0.60 represents moderate change; and 1.20 represents large This study is the first to evaluate the effects of anti-pronating foot
change (Hopkins et al., 2009). Bonferroni adjustment was not made orthoses (APFOS) on hip and knee kinematics, and electromyography
for pairwise comparisons to ensure potentially clinically meaningful find- during a functional step-up task in individuals with PFP. Results indicate
ings were not missed due to stringent statistical correction (Perneger, a reduction in knee internal rotation (100 ms, 0.46°) and hip adduction
1998). Subsequent Spearman's rank correlation analysis was calculated (200 ms, 0.82°) with an APFOS after initial contact. Additionally, a
between FPI and the change in the biomechanical variables, and within reduction in GMed amplitude in the APFOS condition (0.9 mV, P =
the change in kinematic and EMG variables induced by the APFOS. The 0.043) was observed. Although small in magnitude, these biomechani-
alpha level was set at 0.05. cal changes may offer potential mechanisms to underpin reported
APFOS effectiveness in individuals with PFP (Collins et al., 2008; Eng
and Pierrynowski, 1993; Mills et al., 2012a).
3 . Results Findings from the present study are in contrast with previous re-
search that reported no immediate change in lower limb kinematics
3.1 . Hip and knee kinematics or EMG activity with varying density orthoses in individuals with PFP
(Mills et al., 2012b), the explanation likely being the differing tasks eval-
A significant reduction in hip (femur relative to pelvis) adduction uated. In the current study individuals performed a step-up task rather
(0.82°, P = 0.01, SMD 0.86) was observed 200 ms after initial contact than a running task (Mills et al., 2012b). Biomechanical changes are par-
with the APFOS in situ (Table 2, Fig. 3). A trend towards reduced hip in- ticularly pertinent during this task given reported neuromuscular
ternal rotation at initial contact (−1.4°, P = 0.07) was observed with (Barton et al., 2013) and kinematic (Lankhorst et al., 2013) deficits iden-
the APFOS. A small but significant reduction in knee (tibia relative to tified within PFP populations. Additionally, stair stepping is commonly
femur) internal rotation (0.46°, P = 0.03, SMD 0.53) was observed cited as a clinically relevant pain provoking activity (Kujala et al., 1993).
100 ms after initial contact with the APFOS in situ. No significant chang- The sequential pattern of knee transverse plane followed by hip cor-
es were observed in the sagittal plane at the hip or knee. onal plane modification observed in this study in part validates the the-
oretical mechanism of orthoses effect described previously (Tiberio,
1987). This theoretical paradigm proposes that a reduction in rear foot
3.2 . Electromyography findings pronation duration results in less internal tibial rotation and conse-
quently less compensatory femoral internal rotation (Tiberio, 1987). Al-
No significant changes in VMO (P = 0.26), VL (P = 0.61) or GMed though no significant changes were observed in hip transverse plane
(P = 0.69) muscle onset times measured relative to initial contact kinematics, the significant reduction in knee internal rotation and hip
were observed (Table 3). A significant reduction in GMed peak ampli- adduction evident with the APFOS (Fig. 3) potentially demonstrates
tude was observed with the APFOS (0.9 mV, P = 0.043, SMD 0.49). No changes evoked through distal kinematic change. Collection of rearfoot
significant changes in VMO and VL peak amplitudes were observed. motion capture in future studies could further validate this kinetic chain
mechanism.
The significant reduction in GMed peak amplitude observed is in
3.3 . Clinical measures contrast to a reported increase (Hertel et al., 2005) and no change
(Mills et al., 2012b) in peak amplitude resulting from orthoses prescrip-
The mean and standard deviations of all clinical measures are pre- tion previously described in a different task. Hertel et al. (Hertel et al.,
sented in Table 1. Eight participants were assessed to have a normal 2005) had asymptomatic individuals performing a maximal single leg
(0 to 5) FPI score, 10 had a pronated (6 to 9) FPI score, one had a severe- squat activity in different orthoses conditions, and Mills et al. (Mills
ly pronated (10 to 12) and one had a supinated (−1 to −5) FPI score. et al., 2012b) determined peak amplitude values in a symptomatic

Table 2
Pairwise comparison between orthoses and no orthoses conditions for maximum coronal and transverse plane hip and knee angles at time points before and after initial contact.

Time Hip angle (°) Knee angle (°)

Coronal plane Transverse plane Coronal plane Transverse plane

MD CI P value MD CI P value MD CI P value MD CI P value

−100 ms 0.29 −0.6 to 1.1 0.49 −1.2 −2.7 to 0.3 0.1 −1.57 −4.0 to 0.9 0.2 −0.32 −1.5 to 0.9 0.6
IC −0.51 −1.6 to 0.5 0.32 −1.4 −2.9 to 0.1 0.07 −1.48 −3.5 to 0.5 0.14 0.16 −1.0 to 1.3 0.78
+100 ms 0.28 −0.4 to 1.0 0.41 −1.3 −3.6 to 1.0 0.26 −1.49 −3.7 to 0.8 0.18 0.46 0.06 to 0.9 0.03*
+200 ms 0.82 0.4 to 1.3 0.001* −0.8 −3.0 to 1.4 0.45 −1.46 −3.4 to 0.5 0.14 −0.22 −0.7 to 0.3 0.4

Key; MD = Mean difference, CI = Confidence interval -100 ms = 100 ms before initial contact with the force plate; IC = initial contact with force plate; +100 ms = 100 ms after ini-
tial contact with force plate; +200 ms = 200 ms after initial contact with the force plate.
1060 S. Lack et al. / Clinical Biomechanics 29 (2014) 1056–1062

Fig. 3. Graph showing the kinematic mean and SEM 500 ms either side of initial contact (0) for APFOS (open symbols, SEM grey) and no APFOS (closed symbols, SEM black) and paired
mean difference (grey symbols). Coronal plane; +ve = adduction, −ve = abduction. Transverse plane; +ve = internal rotation, −ve = external rotation.

population during running. Both of these activities are in contrast with 0.02), and higher Kujala scores with degree of earlier GMed onset
the repeated step-up task that our symptomatic participants performed change (r = 0.52, P = 0.02) were observed with APFOS application.
which may explain the differences observed. The reduction in GMed Although it is important to acknowledge that correlation does not
amplitude during the lesser demanding activity of stepping may be
suggestive of this muscle having to be less active to achieve the same
or improved lower limb alignment as a result of APFOS insertion (Nigg
et al., 1999). Further studies investigating the neuromotor changes
about the hip resulting from distal interventions are required to support
this hypothesis.
The secondary aim of this study was to explore the correlation of
easily applied clinical measures with observed biomechanical changes.
Correlation between FPI and earlier VMO onset (r = − 0.51, P =

Table 3
Pairwise comparison between orthoses and no orthoses conditions for muscle onset times
(ms) and muscle amplitudes (mV).

Muscle Muscle onset Muscle amplitude

MD CI P value MD CI P value

VMO −8.95 −25.2 to 7.3 0.26 0.2 −2.0 to 2.0 0.87

VL 2.8 −8.5 to 14.1 0.61 −0.3 −1.0 to 0.7 0.55
GMed 3.05 −12.9 to 19.0 0.69 0.9 0.03 to 2.0 0.04*

Key; MD = Mean difference, CI = Confidence interval, VMO = vastus medialis oblique, Fig. 4. Demonstrating the correlation between vastus medial oblique muscle onset change
VL = vastus lateralis, GMed = gluteus medius. and foot posture index score.
 S. Lack et al. / Clinical Biomechanics 29 (2014) 1056–1062 1061

bigger in the pain population, demonstrating a greater degree of excur-

sion from movement patterns observed within our previous study of an
asymptomatic population (Lack et al., 2013). It is possibly that this
greater range is less amenable to distal intervention alone, offering a po-
tential explanation for the lesser degree of change observed between
the two groups.
Neuromotor changes that are occurring at the hip within symptom-
atic individuals demonstrate a treatment adjunct that may enable clini-
cians to optimise muscle activity that directly and indirectly affect the
PFJ. Importantly biomechanical changes resulting from APFOS may
optimise treatment outcomes with interventions such as exercise in
some individuals with PFP. Within this study, the novel assessment of
clinical measure association with mechanistic changes has direct appli-
cability to the clinical environment. In particular more pronated foot
posture using the FPI and higher Kujala scores associated with VMO
and GMed onsets respectively, demonstrate tools that can help guide
clinicians in delivering a tailored intervention approach for the manage-
ment of PFP. Given the widely accepted multifactorial nature of PFP
Fig. 5. Demonstrating the correlation between Gluteus Medius muscle onset change and (Powers et al., 2012), having measures to create more homogenous
Kujala Questionnaire score. r = 0.52. groups to guide intervention approach is likely to have a direct impact
on outcome success. Further research exploring the longer-term effects
of a tailored approach is clearly warranted.
imply causation (Buldt et al., 2013), the findings of this study suggest
specific clinical measures may serve as secondary indicators of bio- 4.2 . Limitations
mechanical change resulting from APFOS prescription. Currently
there is an absence of validated clinical predictors for the successful Bonferroni adjustments were not made to reduce the risk of type
application of foot orthoses in the management of PFP, however, less 2 error through stringent statistical correction, however, it is impor-
pain has been reported to be a predictive indicator for orthoses inter- tant to acknowledge that this does increase the risk of type 1 error
vention success (Barton et al., 2011a; Vicenzino et al., 2010). This (Perneger, 1998). The small magnitude of kinematic change observed
previously reported predictive relationship is interesting in the con- to be significant may be sufficient to optimise treatment outcomes or
text of our finding as the greater change to earlier GMed muscle result in cumulative larger biomechanical changes given the high vol-
onset times was seen in those with less severe symptomology po- ume of repetitive activity performed using an APFOS. Further studies,
tentially representing an underpinning mechanism for orthoses with longer follow-up are required to explore this potential effect.
treatment success. This suggestion is strengthened in the context A study design that utilised within-subject and session comparisons
of moderate-to-strong level 1 evidence that GMed onset is common- was used to minimise the risk of measurement error. Review of three
ly delayed in individuals with PFP (Barton et al., 2013). dimensional gait analysis reliability studies has reported errors that
A more pronated foot posture, as measured using the FPI, correlated commonly range between 2° and 5° in the majority of gait variables
with change to earlier onset in VMO in the APFOS condition (r = −0.51, (McGinley et al., 2009). The extent with which this degree of variability
P = 0.02). Previously, a more pronated foot type was found to be asso- is evident during stepping tasks, however, has not been evaluated.
ciated with reduced pain during a single leg squat and number of pain Unmodified prefabricated APFOS were used in this study as there are
free rises from sitting following the application of a similar APFOS no clear guidelines for customisation in PFP. However, this is not reflec-
(Barton et al., 2011b), although this association was not evident after tive of clinical practice where APFOS are frequently moulded and/or
a 12 week APFOS intervention (Barton et al., 2011a). The current find- customised with varying wedging. Further studies tailoring the APFOS
ings identify a potential neuromotor mechanism by which previously to the individual might demonstrate greater biomechanical changes
reported immediate pain reduction with APFOS may be achieved than those found in the current study.
(Barton et al., 2011b). The rate of stair stepping, potentially affecting both movement and
Interestingly, ankle dorsiflexion range demonstrated no correlation muscle activation patterns, was not standardised within this study to
with change in hip adduction angle in this pain population, compared maximise the direct correlation with real life daily function, but it may
to previous association identified in an asymptomatic group (Lack be worth considering standardisation in future work. However, with
et al., 2013). An absence of causation may be secondary to the multifac- only within-subject comparisons being made the likely effect of variable
torial presentation of PFP, with limitation in ankle range only one step rate is small.
component. Future research that explores the direction of compensatory change
resulting from orthoses interventions could offer further insight into the
4.1 . Clinical implications individual's response to distal interventions assisting in the delivery of a
target treatment approach. It is important to acknowledge, that given
Although statistically significant, kinematic changes resulting from the dearth of current literature linking EMG and kinematic variables to
APFOS in this study were small, and may be exceeded by the measure- the effects of APFOS interventions, this study offers potential mecha-
ment error of 3D motion capture reported previously (McGinley et al., nisms underlying observed clinical efficacy within PFP populations.
2009). Specifically, a 0.46° reduction, representing a small effect, in
knee internal rotation and a 0.82° reduction, representing a moderate 5 . Conclusion
effect, in hip adduction with an APFOS were observed. Despite the
small magnitude of these biomechanical changes, given the potential Anti-pronating foot orthoses used during a step-up task reduced hip
for cumulative effects of high number/volume weight bearing repeti- adduction and knee internal rotation after initial contact and peak
tions that are completed with an APFOS in situ, they have the potential GMed amplitude in individuals with PFP. The small magnitude of
to be clinically meaningful (Nawoczenski et al., 1995). Additionally, the these biomechanical changes during high repetition activities may
total range of hip adduction and internal rotation were consistently have significant clinical implications and explain potential mechanisms
1062 S. Lack et al. / Clinical Biomechanics 29 (2014) 1056–1062

for treatment success of APFOS in PFP populations. This is exploratory Hertel, J., Sloss, B.R., Earl, J.E., 2005. Effect of foot orthotics on quadriceps and gluteus
medius electromyographic activity during selected exercises. Arch. Phys. Med.
research of several biomechanical variables, only a few of which showed Rehabil. 86, 26–30.
significant change, which therefore needs to be followed up by clinically Hopkins, W.G., Marshall, S.W., Batterham, A.M., Hanin, J., 2009. Progressive statistics for
relevant hypothesis driven research. studies in sports medicine and exercise science. Med. Sci. Sports Exerc. 41, 3–13.
Kadaba, M.P., Ramakrishnan, H.K., Wootten, M.E., 1990. Measurement of lower extremity
kinematics during level walking. J. Orthop. Res. 8, 383–392.
Acknowledgements Kujala, U.M., Jaakkola, L.H., Koskinen, S.K., Taimela, S., Hurme, M., Nelimarkka, O., 1993.
Scoring of patellofemoral disorders. Arthroscopy 9, 159–163.
Lack, S., Barton, C., Malliaras, P., Twycross-Lewis, R., Woledge, R., Morrissey, D., 2014. The
We would like to thank the Private Physiotherapy Education Fund effect of anti-pronation foot orthoses on hip and knee kinematics and muscle activity
(PPEF) (2652861) for their financial support by funding the researchers' during a functional step-up task in healthy individuals: a laboratory study. Clin.
Biomech. 29, 177–182.
time, and Vasyli UK for the provision of orthoses devices.
Lankhorst, N.E., Bierma-Zeinstra, S.M., van Middelkoop, M., 2013. Factors associated with
patellofemoral pain syndrome: a systematic review. Br. J. Sports Med. 47, 193–206.
Linton, S.J., Boersma, K., 2003. Early identification of patients at risk of developing a
References persistent back problem: the predictive validity of the Orebro Musculoskeletal Pain
Questionnaire. Clin. J. Pain 19, 80–86.
Andriacchi, T.P., Andersson, G.B., Fermier, R.W., Stern, D., Galante, J.O., 1980. A study of
McGinley, J.L., Baker, R., Wolfe, R., Morris, M.E., 2009. The reliability of three-dimensional
lower-limb mechanics during stair-climbing. J. Bone Joint Surg. (Am. Vol.) 62,
kinematic gait measurements: a systematic review. Gait Posture 29, 360–369.
749–757.
Mills, K., Blanch, P., Chapman, A.R., McPoil, T.G., Vicenzino, B., 2010. Foot orthoses and
Baquie, P., Brukner, P., 1997. Injuries presenting to an Australian sports medicine centre: a
gait: a systematic review and meta-analysis of literature pertaining to potential
12-month study. Clin. J. Sport Med. 7, 28–31.
mechanisms. Br. J. Sports Med. 44, 1035–1046.
Barton, C.J., Munteanu, S.E., Menz, H.B., Crossley, K.M., 2010. The efficacy of foot orthoses
Mills, K., Blanch, P., Dev, P., Martin, M., Vicenzino, B., 2012a. A randomised control trial of
in the treatment of individuals with patellofemoral pain syndrome: a systematic re-
short term efficacy of in-shoe foot orthoses compared with a wait and see policy for
view. Sports Med. 40, 377–395.
anterior knee pain and the role of foot mobility. Br. J. Sports Med. 46, 247–252.
Barton, C.J., Menz, H.B., Crossley, K.M., 2011a. Clinical predictors of foot orthoses efficacy
Mills, K., Blanch, P., Vicenzino, B., 2012b. Comfort and midfoot mobility rather than ortho-
in individuals with patellofemoral pain. Med. Sci. Sports Exerc. 43, 1603–1610.
sis hardness or contouring influence their immediate effects on lower limb function
Barton, C.J., Menz, H.B., Crossley, K.M., 2011b. The immediate effects of foot orthoses on
in patients with anterior knee pain. Clin. Biomech. 27, 202–208.
functional performance in individuals with patellofemoral pain syndrome. Br. J.
Mundermann, A., Wakeling, J.M., Nigg, B.M., Humble, R.N., Stefanyshyn, D.J., 2006. Foot
Sports Med. 45, 193–197.
orthoses affect frequency components of muscle activity in the lower extremity.
Barton, C.J., Levinger, P., Crossley, K.M., Webster, K.E., Menz, H.B., 2012. The relationship
Gait Posture 23, 295–302.
between rearfoot, tibial and hip kinematics in individuals with patellofemoral pain
Nawoczenski, D.A., Cook, T.M., Saltzman, C.L., 1995. The effect of foot orthotics on three-
syndrome. Clin. Biomech. 27, 702–705.
dimensional kinematics of the leg and rearfoot during running. J. Orthop. Sports
Barton, C.J., Lack, S., Malliaras, P., Morrissey, D., 2013. Gluteal muscle activity and
Phys. Ther. 21, 317–327.
patellofemoral pain syndrome: a systematic review. Br. J. Sports Med. 47, 207–214.
Nigg, B.M., Nurse, M.A., Stefanyshyn, D.J., 1999. Shoe inserts and orthotics for sport and
Bohannon, R.W., 1986. Test-retest reliability of hand-held dynamometry during a single
physical activities. Med. Sci. Sports Exerc. 31, S421–S428.
session of strength assessment. Phys. Ther. 66, 206–209.
Perneger, T.V., 1998. What's wrong with Bonferroni adjustments. BMJ 316, 1236–1238.
Buldt, A.K., Murley, G.S., Butterworth, P., Levinger, P., Menz, H.B., Landorf, K.B., 2013. The
Powers, C.M., Bolgla, L.A., Callaghan, M.J., Collins, N., Sheehan, F.T., 2012. Patellofemoral
relationship between foot posture and lower limb kinematics during walking: a sys-
pain: proximal, distal, and local factors. 2nd International Research Retreat. J. Orthop.
tematic review. Gait Posture 38, 363–372.
Sports Phys. Ther. 42, pp. A1–A54.
Chester, R., Smith, T.O., Sweeting, D., Dixon, J., Wood, S., Song, F., 2008. The relative timing
Stathopulu, E., Baildam, E., 2003. Anterior knee pain: a long-term follow-up. Rheumatol-
of VMO and VL in the aetiology of anterior knee pain: a systematic review and meta-
ogy (Oxford) 42, 380–382.
analysis. BMC Musculoskelet. Disord. 9, 64.
Taunton, J.E., Ryan, M.B., Clement, D.B., McKenzie, D.C., Lloyd-Smith, D.R., Zumbo, B.D.,
Collins, N., Crossley, K., Beller, E., Darnell, R., McPoil, T., Vicenzino, B., 2008. Foot orthoses
2002. A retrospective case–control analysis of 2002 running injuries. Br. J. Sports
and physiotherapy in the treatment of patellofemoral pain syndrome: randomised
Med. 36, 95–101.
clinical trial. BMJ 337, a1735.
Tiberio, D., 1987. The effect of excessive subtalar joint pronation on patellofemoral
Crossley, K., Bennell, K., Green, S., Cowan, S., McConnell, J., 2002. Physical therapy for
mechanics: a theoretical model. J. Orthop. Sports Phys. Ther. 9, 160–165.
patellofemoral pain: a randomized, double-blinded, placebo-controlled trial. Am. J.
Uliam Kuriki, H., Micolis de Azevedo, F., de Faria Negrao Filho, R., Alves, N., 2011.
Sports Med. 30, 857–865.
Comparison of different analysis techniques for the determination of muscle onset
Eng, J.J., Pierrynowski, M.R., 1993. Evaluation of soft foot orthotics in the treatment of
in individuals with patellofemoral pain syndrome. J. Electromyogr. Kinesiol. 21,
patellofemoral pain syndrome. Phys. Ther. 73, 62–68 (discussion 68–70).
982–987.
Eng, J.J., Pierrynowski, M.R., 1994. The effect of soft foot orthotics on three-dimensional
Vicenzino, B., Collins, N., Cleland, J., McPoil, T., 2010. A clinical prediction rule for identify-
lower-limb kinematics during walking and running. Phys. Ther. 74, 836–844.
ing patients with patellofemoral pain who are likely to benefit from foot orthoses: a
Freriks, B., Hermens, H., European Commission, B., Health Research, P., 2000. European
preliminary determination. Br. J. Sports Med. 44, 862–866.
recommendations for surface electromyography results of the SENIAM Project.
Roessingh Research and Development, Enschede.