Br J Sports Med: first published as 10.1136/bjsm.2010.077644 on 16 November 2010. Downloaded from http://bjsm.bmj.

com/ on October 25, 2024 at Laval AFMC Consortia. Protected by

Original article

Greater peak rearfoot eversion predicts foot orthoses

efficacy in individuals with patellofemoral pain
syndrome
C J Barton,1,2 H B Menz,1 P Levinger,1 K E Webster,1 K M Crossley3

1Musculoskeletal Research ABSTRACT

Centre, Faculty of Health Objective There is growing evidence for the provision What is already known on this topic
Sciences, La Trobe University,
Bundoora, Australia of foot orthoses when treating individuals with patel-
2School of Physiotherapy, lofemoral pain syndrome (PFPS), and prescription is
Faculty of Health Sciences, frequently based on the assessment of foot posture/ There is growing evidence for the prescription of
La Trobe University, Bundoora, function. However, evaluation of the link between abnor- foot orthoses for individuals with patellofemoral
Australia
mal foot posture/function and foot orthoses outcomes pain syndrome (PFPS). However, not all individu-
3Department of Mechanical
has previously been limited to static alignment measures als with PFPS benefit equally from foot orthoses
Engineering, University of prescription.
Melbourne, Melbourne, and has produced inconsistent findings. In this study, the
Australia ability of baseline foot kinematics associated with prona-
tion to predict marked improvement 12 weeks following
Correspondence to
Mr Christian Barton, foot orthoses prescription in individuals with PFPS was What this study adds
Musculoskeletal Research evaluated.
Centre, La Trobe University, Methods 26 individuals with PFPS were issued with
Bundoora, Victoria, prefabricated foot orthoses, and patient-reported level This study has identified that individuals with
Australia 3068;
of improvement was documented at 12 weeks. Potential PFPS demonstrating signs of excessive rearfoot
[email protected]
predictors of marked improvement at 12 weeks were eversion magnitude during walking are most
Accepted 4 October 2010 measured during walking at baseline and included fore- likely to benefit from prefabricated foot orthoses
Published Online First foot dorsiflexion and abduction, and rearfoot eversion. prescription.

copyright.
16 November 2010 Results Of the 25 participants who completed the
study, seven (28%) reported marked improvement with
the foot orthoses after 12 weeks. Discriminant function hip adduction and dynamic Q angle.10 These tib-
analysis revealed a greater peak rearfoot eversion to ial and femoral kinematic variations are thought
be the only significant independent predictor of marked to be detrimental to the PFJ owing to the associ-
improvement. ated reduced contact area and increased lateral PFJ
Conclusion These findings provide preliminary evi- compression.11
dence that greater peak rearfoot eversion is predictive of Although the majority of previous studies eval-
marked improvement 12 weeks following prefabricated uating foot orthoses for individuals with PFPS
foot orthoses prescription in individuals with PFPS. have used observations of excessive foot prona-
Therefore, foot orthoses may be most effective in the tion as part of their inclusion criteria, 2 – 5 7 evi-
subgroup of people with PFPS and increased dynamic dence linking foot structure or function to clinical
foot pronation. success in PFPS populations is limited. Previous
clinical prediction studies have been limited to
static assessment of posture and reported con-
There is growing evidence for the efficacy of foot fl icting fi ndings relating to the association of foot
orthoses prescription when treating individu- posture with foot orthoses efficacy in individuals
als with patellofemoral pain syndrome (PFPS).1–7 with PFPS. Vicenzino et al12 reported that greater
Traditionally, foot orthoses have been advocated foot mobility (change in arch height from non-
for PFPS based on the premise that they are weight-bearing to weight-bearing) was associated
needed to control excessive foot pronation.8 9 with marked improvement at 12 weeks. Contrary
Tiberio10 provided a theoretical mechanism under- to this, Sutlive et al6 reported a less pronated foot
pinning therapeutic foot orthoses effects in indi- type as measured by navicular drop and calcaneal
vidual with PFPS, by describing how altered foot angle was associated with greater than 50% pain
function during gait may detrimentally affect the reduction at 3 weeks. Additionally, we recently
patellofemoral joint (PFJ). It was proposed that found that static foot posture as measured by the
excessive or prolonged foot pronation (rearfoot foot posture index and normalised navicular drop
eversion) during the stance phase of gait would were unable to predict marked improvement in a
result in greater tibial internal rotation. This group of 60 individuals with PFPS following 12
would in turn delay or reduce the tibial external weeks of wearing foot orthoses.13 Inconsistent
rotation relative to the femur required to allow fi ndings relating to the association between foot
knee extension through midstance. To compen- posture and foot orthoses efficacy in individuals
sate, the hip (femur) would need to rotate inter- with PFPS need to be considered in the context of
nally to a greater degree, thereby also increasing one significant methodological limitation—static

Br J Sports Med 2011;45:697–701. doi:10.1136/bjsm.2010.077644 697

 Br J Sports Med: first published as 10.1136/bjsm.2010.077644 on 16 November 2010. Downloaded from http://bjsm.bmj.com/ on October 25, 2024 at Laval AFMC Consortia. Protected by
Original article

posture assessments may not accurately represent dynamic

foot function. To date, no previous study has evaluated the
association between dynamic foot function and foot orthoses
efficacy in individuals with PFPS.
The aim of this study was to identify kinematic predictors
of foot orthoses efficacy in individuals with PFPS during walk-
ing. Specifically, the predictive value of kinematics associated
with foot pronation (rearfoot eversion, forefoot dorsiflexion
and forefoot abduction) measured at baseline was evaluated
for marked improvement 12 weeks following foot orthoses
prescription. It was hypothesised that a subpopulation of Figure 1 Prefabricated foot orthoses (Vasyli Pro, Vasyli International)
PFPS individuals with signs of excessive foot pronation during issued to each participant.
walking would be the most likely to benefit from foot orthoses
prescription. orthoses whenever possible. Additionally, a diary was com-
pleted, describing when the orthoses were and were not worn
METHODS during physical activity.
Participants
Twenty-six individuals with PFPS (five males and 21 females) Primary outcome measure
were recruited via advertisements placed at La Trobe At 12 weeks following foot orthoses prescription, each par-
University, the University of Melbourne, and in the greater ticipant rated the perceived improvement in symptoms using a
Melbourne area. This group of PFPS individuals form a sub- five-point Likert scale, consistent with previous PFPS RCTs.1 14
population (without additional inclusion/exclusion criteria) The five responses included markedly better, moderately bet-
of a larger clinical prediction rule study. The mean (SD) age, ter, same, moderately worse and markedly worse. The asso-
height, weight and usual weekly sport or recreational physi- ciation of each kinematic variable with those who reported
cal activity time were 25 (5) years, 1.69 (0.09) m, 67 (14) kg marked improvement was evaluated.
and 320 (210) min respectively. The study was approved by La
Trobe University’s Faculty of Health Sciences Human Ethics Kinematic analysis
Committee, and each participant gave written informed con- Data collection included lower-limb kinematic evaluation of
sent prior to participation. Diagnosis of PFPS was based on def- each participant’s symptomatic limb (those with unilateral

copyright.
initions used in previous high-quality randomised controlled symptoms) or most symptomatic limb (in those with bilateral
trials (RCTs).1 14 Inclusion criteria were: aged 18–35 years old; symptoms) during walking. Motion analysis was collected
insidious onset of peripatellar or retropatellar knee pain of at using a three-dimensional motion-analysis system (Vicon MX
least 6 weeks’ duration; worst pain in the previous week of system; Oxford Metrics, Oxford) combined with 10 cameras
at least 30 mm on a 100 mm visual analogue scale; pain pro- (eight MX3 and two MX40) operating at a sampling frequency
voked by at least two activities from running, walking, hop- of 100 Hz. Thirty-six retroreflective markers were placed on
ing, squatting, stair negotiation, kneeling or prolonged sitting; specific anatomical landmarks (outlined below) to form fore-
pain elicited by patellar palpation, PFJ compression or resisted foot, rearfoot, tibial, femoral and pelvic segments. Ground
isometric quadriceps contraction. Exclusion criteria were: use reaction forces were collected using two force plates (type
of foot orthoses in the previous 5 years, physiotherapy treat- 9865B; Kistler, Winterthur, Switzerland; and AMTI, OR6,
ment in the previous 6 months, current use of anti-infl amma- Watertown, MA, USA) at a sampling frequency of 1000 Hz.
tory medications, concomitant injury or pain arising from the The Oxford Foot Model (OFM)15 was used to perform
lumbar spine or hip, knee internal derangement, knee ligament kinematic evaluation of each participant during walking.
insufficiency, previous knee surgery, PFJ instability or patellar For the purpose of static calibration, plug-in gait (PIG) was
tendinopathy. added to the model,15 and each participant’s height, weight,
inter anterior–superior–iliac–spine (ASIS) distance, ASIS
Intervention to greater trochanter distance, knee width and ankle width
Participants attended a single treatment (15 min) and data- were recorded. Retroreflective markers were placed over the
collection (60 min) session. During this session, all baseline following anatomical landmarks by the same investigator
data were collected prior to issuing each participant with a pair for each participant: midpoint of the sacrum between the
of prefabricated foot orthoses. The orthoses were commer- posterior– superior–iliac–spines, and bilaterally on the ASIS,
cially available three-quarter-length devices with lateral cut- lateral aspect of the femur (5 cm wand), head of the fibula,
outs (Vasyli Pro, Vasyli International), made of ethelene-vinyl tibial tuberosity, anterior border of tibia, lateral aspect of
acetate of medium (Shore A 55) density, containing built-in tibia (5 cm wand), medial and lateral malleoli, three markers
arch supports and 4° rearfoot varus wedging (see figure 1). No bisecting the heel (distal, wand and proximal), lateral calca-
customisation of the orthoses was performed, as the aim was neus, sustentaculum tali, base of fi rst metatarsal, head of fi rst
to provide the same level of support for all participants in the metatarsal, proximal fi rst phalanx, head of fi fth metatarsal
absence of a scientifically validated customisation approach and base of fi fth metatarsal (see figure 2). A relaxed stand-
for PFPS. Additionally, these foot orthoses were chosen, as ing calibration trial was then captured with knee-alignment
they could be fitted into most participants’ footwear without devices (KADs) in situ. Prior to the walking trials, the KADs
discomfort. were removed and replaced with lateral femoral condyle
Participants were then instructed to wear the orthoses for markers, and the anatomical markers used to defi ne segment
a period of 12 weeks. To assist with compliance, participants axes were removed (medial malleoli, proximal heels and fi rst
were asked to wear suitable footwear to accommodate the metatarsal heads).

698 Br J Sports Med 2011;45:697–701. doi:10.1136/bjsm.2010.077644

 Br J Sports Med: first published as 10.1136/bjsm.2010.077644 on 16 November 2010. Downloaded from http://bjsm.bmj.com/ on October 25, 2024 at Laval AFMC Consortia. Protected by
Original article

for their association with marked improvements following 12

weeks of wearing the foot orthoses. First, each variable was
tested for its univariate relationship using independent-samples
t tests comparing those reporting marked improvement with
those reporting moderate improvement, same, moderate wors-
ening or marked worsening. All variables with a significance
level of p<0.20 were retained for further discriminant analysis.
Discriminant analysis with significance set at p<0.05 was then
completed entering each retained variable to determine which
were predictive of marked improvement at 12 weeks.

RESULTS
Twenty-five of the 26 participants enrolled completed the study
at 12 weeks. The one drop-out was unable to be contacted at 12
weeks. Participants who completed the study wore their foot
orthoses for 81%±17 of their physical activity performed during
Figure 2 Anterior view of Oxford foot model and plug-in-gait marker the study period. A total of seven (28%) participants reported
placements including the knee alignment devices (KADs) (A) and that they were markedly better after 12 weeks of wearing the
posterior view of Oxford foot model marker placements for the static foot orthoses. There was no difference (p=0.998) in gait veloc-
trial (B). ity between those with marked improvement (1.36±0.13 m/s)
and those with moderate improvement, no change, moderate
Participants performed practice walking trials to allow worsening and marked worsening (1.36±0.14 m/s).
familiarisation with the instrumentation and environment. Univariate analysis comparing those with marked improve-
Once participants felt they were comfortable, and the inves- ment with those with moderate improvement, no change,
tigator deemed they were walking with consistent velocity, moderate worsening and marked worsening for peak angles,
kinematic data collection commenced. Each participant was timing of peak angles and ranges of motion for each kinematic
asked to walk at their natural comfortable speed across a variable evaluated can be found in table 1. Greater peak rear-
12 m walkway. Five successful trials were collected for each foot eversion relative to the laboratory (p=0.045) was the only
participant. A trial was deemed successful when the partici- variable retained for discriminant analyses (p<0.20). No uni-

copyright.
pant’s instrumented foot landed within the borders of the variate associations were found for either timing of peak angles
fi rst force plate they traversed, which was used to identify the or ranges of motion for any of the kinematic variables evalu-
commencement of the gait cycle. To ensure a natural walking ated. Discriminant analysis revealed that greater rearfoot ever-
pattern, participants were not made aware of the force plates sion relative to the laboratory was independently predictive of
which were hidden within the floor. Only the investigator marked improvement at 12 weeks (p=0.045), producing a stan-
knew of their existence and position. Starting positions prior dardised canonical discriminant function coefficient of 0.975.
to each trial were adjusted to optimise the likelihood of a suc-
cessful trial.
DISCUSSION
Despite growing evidence that foot orthoses are an effective
Data analysis intervention for PFPS,1–7 their longstanding theoretical biome-
Each trial was reconstructed, and the retroreflective markers chanical rationale10 lacks validation from empirical evidence.
identified and labelled within the Vicon Nexus software. Gait Foot orthoses are often recommended for individuals with PFPS
events (heel strike and toe off) were identified to allow gait- possessing excessive foot pronation.10 However, previous clini-
velocity measurement using force plate data, and the OFM cal prediction rule studies6 12 have reported inconsistent links
model was applied to the captured markers. Data were then between static foot posture measures and clinical outcomes
exported to a purposely developed Excel template for analysis. following foot orthoses prescription. The current study is the
Variables of interest included magnitude and timing of peak fi rst to evaluate the association between baseline dynamic foot
angles and ranges of motion during stance for: function during walking and clinical outcomes following foot
(i) forefoot relative to rearfoot: dorsiflexion (sagittal plane) orthoses prescription in individuals with PFPS. Importantly,
and abduction (transverse plane) defi ned by the OFM;15 fi ndings show that a subpopulation with greater peak rearfoot
(ii) rearfoot relative to the laboratory: eversion (frontal eversion is likely to benefit from the provision of prefabricated
plane) defi ned by the OFM;15 foot orthoses, while those with normal foot function may not.
(iii) rearfoot relative to tibia: eversion (frontal plane) defi ned We previously found that static foot posture was unable
by the OFM.15 to predict clinical outcomes following 12 weeks of wearing
Additionally, gait velocity was compared between those with prefabricated foot orthoses in 60 individuals with PFPS.13
marked improvement and those with moderate improvement, However, in the current study, which used a subpopulation of
no change, moderate worsening or marked worsening. the same 60 individuals, we found that greater baseline peak
rearfoot eversion was able to predict marked improvement fol-
Statistical analysis lowing 12 weeks. In doing so, this study’s fi ndings support
Statistical analysis was performed using SPSS Version 17.0. the traditional theoretical rationale behind foot orthoses pre-
Prior to statistical analysis, all variables were assessed for nor- scription for individuals with PFPS.10 Additionally, this fi nd-
mality and found to be normally distributed based on graphi- ing implies that evaluating foot function dynamically may be
cal observation and a skewness statistic of between –1.0 and more valid than using static foot posture measures when con-
+1.0. Gait velocity and each kinematic variable were evaluated sidering foot orthoses prescription for individuals with PFPS.

Br J Sports Med 2011;45:697–701. doi:10.1136/bjsm.2010.077644 699

 Br J Sports Med: first published as 10.1136/bjsm.2010.077644 on 16 November 2010. Downloaded from http://bjsm.bmj.com/ on October 25, 2024 at Laval AFMC Consortia. Protected by
Original article

Table 1 Univariate comparisons for magnitude (°) and timing (percentage gait of cycle) of peak angles
and ranges of motion (°) during stance, defining clinical improvement as markedly better.
Moderately better, same,
Markedly better moderately worse or Mean difference
Variable (SD) (n=7) markedly worse (SD) n=18 (95% CI) p Value

Forefoot relative to rearfoot motion dorsiflexion

Magnitude of peak 14.1 (5.2) 11.3 (5.1) 2.8 (−1.9 to 7.5) 0.230
Timing of peak 51.8 (2.6) 51.0 (2.5) 0.8 (−1.5 to 3.1) 0.489
Range of motion 7.9 (0.9) 7.9 (2.5) 0.0 (−1.4 to 1.4) 0.987
Forefoot relative to rearfoot motion abduction
Magnitude of peak 3.0 (7.5) −0.8 (7.4) 3.8 (−3.0 to 10.7) 0.262
Timing of peak 39.2 (8.8) 41.2 (11.6) −2.0 (−12.1 to 8.1) 0.681
Range of motion 7.3 (2.5) 7.5 (2.6) −0.2 (−2.6 to 2.2) 0.856
Rearfoot relative laboratory eversion
Magnitude of peak −5.4 (2.1) −3.1 (2.5) −2.3 (−4.5 to −0.1)† 0.045*
Timing of peak 29.6 (6.6) 32.0 (6.0) −2.4 (−8.0 to 3.4) 0.406
Range of motion 5.2 (1.2) 6.0 (1.8) −0.8 (−2.3 to 0.8) 0.313
Rearfoot relative to tibia eversion
Magnitude of peak −8.7 (3.1) −7.6 (3.0) −1.1 (−4.0 to 1.7) 0.412
Timing of peak 34.4 (4.2) 32.3 (6.5) 2.1 (−3.5 to 7.6) 0.444
Range of motion 9.0 (1.7) 9.5 (3.1) −0.5 (−3.1 to 2.1) 0.676
†Retained for discriminant analysis.

*p<0.05.

Previous studies evaluating rearfoot eversion in individu- factors behind efficacy, additional evaluation of the associa-
als with PFPS have consistently measured rearfoot motion tion between kinematic changes produced by foot orthoses
relative to the tibia.16 –18 The current study evaluated rear- and clinical outcomes in the future may be of value. Second,
foot motion relative to both the tibia and laboratory (floor). owing to the relatively small sample size (n=26), only seven
Interestingly, only rearfoot eversion relative to the labora- participants reported a marked improvement at 12 weeks.

copyright.
tory was able to predict a marked improvement at 12 weeks. Therefore, it is possible that a larger number of predictors may
A likely explanation for the discrepancy between the two be identified using a larger sample size. Third, this study did
methods of rearfoot motion evaluation may be the presence not contain a control group. Therefore, further validation of
of joint coupling between the rearfoot and tibia, where the baseline peak rearfoot eversion as a predictor of foot orthoses
tibia is forced into adduction by the everting rearfoot. In the outcomes in individuals with PFPS requires confi rmation in a
presence of such a coupling relationship, differences in rear- larger RCT. Finally, the subpopulation with marked improve-
foot eversion relative to the tibia may not be detected despite ment in this study was identified using a three-dimensional
the presence of greater rearfoot eversion relative to the labo- marker based analysis. Therefore, directly applying fi ndings to
ratory. Therefore, measuring rearfoot eversion motion rela- a clinical setting may not be possible. Clinical application may
tive to the laboratory (floor) may be more appropriate than be improved by the development of valid and reliable clini-
measuring relative to the tibia in future research and clini- cal foot function tests able to identify individuals with exces-
cal practice if the goal is to predict outcomes following foot sive rearfoot eversion. Until this is completed, sports medicine
orthoses prescription. practitioners may choose to rely on clinical observation to
Baseline forefoot kinematics were not found to be associ- determine the presence of excessive rearfoot eversion before
ated with marked improvement following prefabricated foot prescribing prefabricated foot orthoses.
orthoses prescription in the current study. This could imply
that evaluating baseline forefoot motion may not assist in CONCLUSION
determining the likelihood of improvement when prescribing This study is the fi rst to evaluate the association between foot
prefabricated foot orthoses prescription to individuals with or lower-limb kinematic assessment and clinical outcomes
PFPS. However, it must be considered that the foot orthoses with foot orthoses in individuals with PFPS. Results provide
used in this study were not individualised, and while they preliminary evidence that measuring dynamic foot function
contained rearfoot varus wedging (4°), they did not contain (peak rearfoot eversion) predicted prefabricated foot orthoses
any intrinsic forefoot wedging. Therefore, they may not have efficacy in individuals with PFPS. Therefore, foot orthoses
been appropriate to correct the presence of any abnormal fore- may be most effective in the subgroup of people with PFPS
foot motion which might contribute to PFPS development. and increased dynamic foot pronation.
Additional research is needed to determine if baseline lower-
limb kinematics including forefoot motion are predictive of Acknowledgements The authors would like to acknowledge Vasyli International
clinical outcomes when using different foot orthoses prescrip- for providing the prefabricated foot orthoses used in this study. HBM is currently an
NHMRC research fellow (Clinical Career Development Award, ID: 433049).
tion approaches such as forefoot posting.
Competing interests None.
Limitations Ethics approval Ethics approval was provided by La Trobe University’s Faculty of
Health Sciences Human Ethics Committee.
The fi ndings of this study need to be considered in the con-
text of several limitations. First, kinematics were only mea- Provenance and peer review Not commissioned; externally peer reviewed.
sured at baseline. Therefore, to better understand mechanistic Patient consent Obtained.

700 Br J Sports Med 2011;45:697–701. doi:10.1136/bjsm.2010.077644

 Br J Sports Med: first published as 10.1136/bjsm.2010.077644 on 16 November 2010. Downloaded from http://bjsm.bmj.com/ on October 25, 2024 at Laval AFMC Consortia. Protected by
Original article

REFERENCES 10. Tiberio D. The effect of excessive subtalar joint pronation on patellofemoral
1. Collins N, Crossley K, Beller E, et al. Foot orthoses and physiotherapy in the mechanics: a theoretical model. J Orthop Sports Phys Ther 1987;9:160 – 5.
treatment of patellofemoral pain syndrome: randomised clinical trial. BMJ 11. Wilson T. The measurement of patellar alignment in patellofemoral pain
2008;337:a1735. syndrome: are we confusing assumptions with evidence? J Orthop Sports Phys
2. Eng JJ, Pierrynowski MR. Evaluation of soft foot orthotics in the treatment of Ther 2007;37:330 – 41.
patellofemoral pain syndrome. Phys Ther 1993;73:62– 8; discussion 68–70. 12. Vicenzino B, Collins N, Cleland J, et al. A clinical prediction rule for identifying
3. Amell TK, Stothart JP, Kumar S. The effectiveness of functional foot orthoses patients with patellofemoral pain who are likely to benefi t from foot orthoses:
as a treatment for patellofemoral stress syndrome: the clients’ perspective. a preliminary determination. Br J Sports Med 2010;44:862– 6.
Physiother Can 2000;52:153. 13. Barton C, Menz H, Crossley K. Clinical Predictors of Foot Orthoses Efficacy
4. Johnston LB, Gross MT. Effects of foot orthoses on quality of life for individuals in Patellofemoral Pain Syndrome. Baltimore, MD: American College of Sports
with patellofemoral pain syndrome. J Orthop Sports Phys Ther 2004;34:440 – 8. Medicine Annual Meeting 2010.
5. Pitman D, Jack D. A clinical investigation to determine the effectiveness of 14. Crossley K, Bennell K, Green S, et al. Physical therapy for patellofemoral
biomechanical foot orthoses as initial treatment for patellofemoral pain syndrome. pain: a randomized, double-blinded, placebo-controlled trial. Am J Sports Med
J Prosthet Orthot 2000;12(4):110 –16. 2002;30:857– 65.
6. Sutlive TG, Mitchell SD, Maxfield SN, et al. Identification of individuals with 15. Stebbins J, Harrington M, Thompson N, et al. Repeatability of a model
patellofemoral pain whose symptoms improved after a combined program of for measuring multi-segment foot kinematics in children. Gait Posture
foot orthosis use and modified activity: a preliminary investigation. Phys Ther 2006;23:401–10.
2004;84:49 – 61. 16. Callaghan MJ, Baltzopoulos V. Gait analysis in patients with anterior knee pain.
7. Eng JJ, Pierrynowski MR. The effect of soft foot orthotics on three-dimensional Clin Biomech 1994;9:79 – 84.
lower-limb kinematics during walking and running. Phys Ther 1994;74:836 – 44. 17. Levinger P, Gilleard W. Tibia and rearfoot motion and ground reaction forces
8. Gross MT, Foxworth JL. The role of foot orthoses as an intervention for in subjects with patellofemoral pain syndrome during walking. Gait Posture
patellofemoral pain. J Orthop Sports Phys Ther 2003;33:661–70. 2007;25:2– 8.
9. Powers CM. The influence of altered lower-extremity kinematics on 18. Duffey MJ, Martin DF, Cannon DW, et al. Etiologic factors associated
patellofemoral joint dysfunction: a theoretical perspective. J Orthop Sports Phys with anterior knee pain in distance runners. Med Sci Sports Exerc
Ther 2003;33:639 – 46. 2000;32 :1825 – 32.

copyright.

Br J Sports Med 2011;45:697–701. doi:10.1136/bjsm.2010.077644 701