Foot posture as a risk factor for lower limb

overuse injury: a systematic review and

meta-analysis
Bradley S Neal​1,2​, Ian B Griffiths​1​, Geoffrey J Dowling​3​, George S Murley​3,4​, Shannon E
Munteanu​3,4​, Melinda M Franettovich Smith​5​, Natalie J Collins​6 ​and Christian J
Barton​1,2,4,7*

Please see related article:

http://www.jfootankleres.com/content/7/1/53

Abstract

Background: Static measures of foot posture are regularly used as part of a clinical examination to determine
the need for foot level interventions. This is based on the premise that pronated and supinated foot postures
may be risk factors for or associated with lower limb injury. This systematic review and meta-analysis
investigates foot posture (measured statically) as a potential risk factor for lower limb overuse injuries.
Methods: A systematic search was performed using Medline, CINAHL, Embase, SportDiscus in April 2014, to
identify prospective cohort studies that investigated foot posture and function as a risk factor for lower limb
overuse injury. Eligible studies were classified based on the method of foot assessment: (i) static foot posture
assessment; and/or (ii) dynamic foot function assessment. This review presents studies evaluating static foot
posture. The methodological quality of included studies was evaluated by two independent reviewers, using
an adapted version of the Epidemiological Appraisal Instrument (EAI). Where possible, effects were
expressed as standardised mean differences (SMD) for continuous scaled data, and risk ratios (RR) for
nominal scaled data. Meta-analysis was performed where injuries and outcomes were considered
homogenous. Results: Twenty-one studies were included (total n = 6,228; EAI 0.8 to 1.7 out of 2.0). There
was strong evidence that a pronated foot posture was a risk factor for medial tibial stress syndrome (MTSS)
development and very limited evidence that a pronated foot posture was a risk factor for patellofemoral pain
development, although associated effect sizes were small (0.28 to 0.33). No relationship was identified
between a pronated foot posture and any other evaluated pathology (i.e. foot/ankle injury, bone stress
reactions and non-specific lower limb overuse injury). Conclusion: This systematic review identified strong
and very limited evidence of small effect that a pronated foot posture is a risk factor for MTSS and
patellofemoral pain respectively. Evaluation of static foot posture should be included in a multifactorial
assessment for both MTSS and patellofemoral pain, although only as a part of the potential injury risk profile.
Whilst the included measures are clinically applicable, further studies are required to determine their
relationship with dynamic foot function.
 Keywords: Lower extremity, Foot, Pronation, Supination, Prospective studies, Risk factors, Musculoskeletal
diseases, Review

* Correspondence: ​[email protected] ​1​Pure Sports

Medicine, London, UK 2​​ Centre for Sports and Exercise Medicine, Queen

Mary University of London, London,

​ UK Full list of author information is
available at the end of the article

2014 Neal et al.; licensee BioMed Central. This is an Open Access article distributed under the terms of the Creative Commons Attribution License
(http://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and reproduction in
any medium, provided the original work is properly credited. The Creative Commons Public Domain Dedication
waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article,
unless otherwise stated.

Two recent reviews have evaluated the relationship

Background ​Identifying lower extremity musculoskeletal between foot posture and lower extremity injury [11,12]. Tong
injury risk factors is important for sports medicine clinical and Kong [11] concluded that both pronated and supinated foot
practice and research, potentially allowing for the development types are significantly associated with lower extremity injury,
of more effective and efficient prevention and manage- ment although the strength of this re- lationship was low, and the
strategies. Several risk factors have been suggested to increase authors did not provide a breakdown of individual pathologies
lower extremity injury risk, including in- creased body mass or outcome mea- sures. Additionally, this review included
index [1], female sex [2] and altered hip mechanics [3]. studies that were not prospective in nature, which limits the
Foot pronation as a potential lower extremity overuse ability to differentiate between cause and effect. Chuter and
injury risk factor has received great attention in research and Janse de Jonges [12] narrative review suggested that ex-
clinical practice. Historically, foot mechanics are considered to cessive foot pronation increased the risk of exercise re- lated
contribute to lower extremity malalign- ment and pathology lower leg pain and MTSS, but not patellofemoral pain.
proximal to the foot via joint coupling with tibial internal However, this review was not systematic in nature, making
rotation [4]. Research has suggested that rearfoot motion conclusions potentially open to bias. Addition- ally, Chuter and
(eversion) closely corre- sponds with tibial motion (internal Janse de Jonge [12] focused on dynamic function, and did not
rotation) [5,6] and is potentially associated with transverse include studies related to static foot posture.
plane rotations at the hip [7]. Based on this model of lower To the authors knowledge, there has not been a sys-
extremity joint coupling, there has long been a theoretical link tematic review investigating the relationship between static
between foot pronation and lower extremity pathologies includ-foot posture or dynamic foot function and lower extremity
ing exercise related lower extremity injury, medial tibial stressinjury development using only prospectively designed studies.
syndrome (MTSS) and patellofemoral pain [1,8,9]. At the otherTherefore, the aim of this systematic review was to (i) identify
end of the spectrum increased foot supi- nation has been linkedand appraise the current evi- dence for the prospective link
to lower extremity injury via a mechanism of increased limbbetween foot posture and lower limb overuse injury and (ii)
stiffness and subsequent vertical loading rates [10]. provide guidance for future research in this area. This paper,
Considering the hypothesised link between foot pos- turefocusing on static foot posture measures, represents the first
and lower extremity injury, static foot posture is frequentlycompo- nent of a two-part systematic review on foot function-
assessed in the clinical setting, with a belief that this mayrelated risk factors for lower limb overuse injury.
provide indications for biomechanical in- terventions (e.g. foot
orthoses). Commonly employed as- sessment methods to assessMethods ​The protocol for this systematic review was
foot posture include, but are not limited to, navicular drop,developed using guidelines provided by the Preferred
resting calcaneal eversion, the longitudinal arch angle and theReporting of Systematic Reviews and Meta-Analysis
Foot Posture Index (FPI) [11]. (PRISMA) State- ment [13] (Additional file 1).
 were: (i) prospective cohort study design; (ii) quantitative
Search strategy ​MEDLINE, CINAHL, Embase, measurement of foot posture or function at baseline (static or
SPORTDiscus and Goo- gle Scholar were searched from dynamic); and (iii) prospective collection of specific or
inception until April 2014. Medical Subject Headings (MeSH) non-specific lower limb overuse injury sur- veillance data over
were exploded to encompass relevant subheadings, as well as a specified time period. No exclusion was made relative to any
relevant keywords (Additional file 2). The search strategy given population. Two authors (BSN and IBG) reviewed all
limited findings to adult human participants and English lan- abstracts to determine eligi- bility. Full texts were screened to
guage publications. We hand searched reference lists of confirm eligibility, and where there was uncertainty regarding
identified systematic and narrative reviews and contacted field eligibility from the abstract alone. A third reviewer (CJB) was
experts (e.g. physiotherapists, podiatrists) regarding known available for any discrepancies.
important publications. Additionally, a cited ref- erence search Studies that fulfilled the initial eligibility criteria were
for each included paper was undertaken in Google Scholar. separated into those that investigated static measures of foot
posture and those that investigated dynamic mea- sures of foot
Eligibility criteria ​A single investigator (GJD) exported all posture (i.e. measured during walking or running). This review
studies identified by the search strategy to Endnote version X5 focused on static measures, while dynamic measures are
(Thomson Reuters, Philadelphia). Initial eligibility criteria addressed in the accompanying
​ 014, ​7​:55 Page 2 of 13 http://www.jfootankleres.com/content/7/1/55
Neal ​et al. Journal of Foot and Ankle Research 2
consensus meeting. Average scores across the 43 items were
paper [14]. Any study that included both static and dy- namic calculated, with a maximum possible score of 2.0. Studies were
measures of the foot was included, but only data pertaining to then classified as high quality (​≥​1.4), moderate quality (1.1 to
static measures was used for this part of the review. Studies <1.4), or poor quality (<1.1) [26].
that included static foot posture mea- sures that were not
quantitative in nature were excluded [15-25]. We defined Data management ​Data regarding study characteristics were
specific lower limb overuse injuries as those with a single extracted from each study by two independent investigators
diagnosis and non-specific lower limb overuse injuries as those (BSN and IBG). This included publication details (year, author,
without a specific diagno- sis or where multiple overuse country), participant characteristics (number of parti- cipants
injuries were pooled. injured and uninjured, age, sex, eligibility cri- teria, population
[i.e. military]) and study methods (foot posture measurement,
Quality assessment ​The Epidemiological Appraisal Instrument examiner details, injury outcome, duration of study, covariates
(EAI) [26] was used to evaluate the methodological quality of investigated) (Table 1). For continuous scaled foot posture
the included studies. The EAI was designed specifically for variables means and standard deviations (SD) were extracted
cohort studies and consists of 43 items across five domains (i) for injured and uninjured participants. For nominal scaled
reporting, (ii) subject/record selection, (iii) measure- ment variables raw counts of injured and uninjured participants (e.g.
quality, (iv) data analysis and (v) generalisation of re- sults injury incidence in categories of foot types) were extracted.
[26]. Individual items were scored as Yes (score of 2), Partial Corresponding authors were contacted for additional data if
(score of 1), No (score of 0), Unable to determine (score of 0) adequate data were not provided in the publica- tion. For
or Not Applicable (item ex- cluded). Previous studies have studies that described particular foot posture variables but did
found the EAI to have ad- equate external validity and good to not publish data, this was recorded as
excellent intra-rater (Kappa coefficient range 52 to 60), and not reported (NR) and it was assumed that no signifi-
inter-rater (Kappa coefficient=90% [95% CI; 87-92%]) cant differences were observed between those who were
reliability [26]. The wording of the 43 items was modified injured and uninjured.
slightly for this re- view to improve clarity and rater
interpretation. To main- tain validity, no items were removed Statistical methods ​Inter-rater reliability of EAI scores
(Additional file 3). between the two raters was evaluated descriptively using
Two raters (BSN and IBG) who were blind to the authorpercentage agreement. Differences between scores for Yes ,
and publication details independently evaluated each study.Partial , No , and Unable to determine were calculated, with
Discrepancies between the raters were resolved during aperfect agree- ment indicated by zero difference. Ratings for
the not applicable response were excluded from analysis, as no 1.20) [29]. A RR>1.0 indicated that the lower limb overuse
rater interpretation was required. injury was more likely to be found in partici- pants with the
Extracted means and SDs for continuous scaled risk factor present. A small effect was in- dicated by a RR ​≥
variables were used to calculate standardised mean dif- 2.0, and a large effect ​≥ ​4.0 [29]. Effects were considered to be
ferences (SMD) with 95% confidence intervals (CIs). Extracted statistically significant if the asso- ciated 95% CI for SMD did
nominal scaled data was used to caclulate risk ratios (RR) with not contain zero, and the 95% CI for RR did not contain one.
95% CIs. Data for men and women were analysed separately
where this information was provided. Data for right feet only Evidence-based recommendations ​Based on previous work
were entered when studies provided a breakdown for both feet, by van Tulder ​et al. ​[30], levels of evidence were assigned for
to maintain independence of data [27]. All analyses were each foot posture measure evaluated, incorporating statistical
completed in Review Manager 5.0 (The Cochrane outcomes and metho- dological quality of included studies.
Collaboration, Copenhagen, Denmark). Meta-analysis (data
pooling) was performed where homogeneity between studies Strong evidence ​Pooled results derived from three or more
was deemed to be adequate (i.e. outcome measures were studies, in- cluding a minimum of two high quality studies that
performed and reported in a similar fashion for the same are statistically homogenous; may be associated with a sta-
pathology). The level of statistical heterogeneity for pooled tistically significant or non-significant pooled result.
data was established using ​I​2 ​statistics and asso- ciated p values
(heterogeneity defined as ​I2​ ​> 50%) [28]. Moderate evidence ​Statistically significant pooled results
Calculated individual or pooled SMDs were cate- derived from mul- tiple studies that are statistically
gorised as small (​≤ ​0.59), medium (0.60 to 1.19) or large (​≥ heterogeneous, including
​ 014, ​7​:55 Page 3 of 13 http://www.jfootankleres.com/content/7/1/55
Neal ​et al. Journal of Foot and Ankle Research 2
Table 1 Summary of study characteristics Willems ​et al​., 2006 [​34​] Physical education students 12
Population Observation period months Medial tibial stress
(activity, duration) syndrome
Injury outcome Injured group Uninjured group Foot posture 46 (29) NR 354 (130) NR Resting calcaneal position
measure (degrees)

Reinking, 2006 [​35​] Female collegiate athletes One athletic

N total (n females) season Medial tibial stress
Age (mean SD)
syndrome
N total (n females)
20 (20) NR 56 (56) NR Navicular drop
N total (n females)
Age (mean SD)
Age (mean SD) Reinking ​et al.​ , 2007 [​36​] Collegiate athletes One athletic
Age (mean SD) season Medial tibial stress
syndrome
Bennett ​et al.​ , 2001 [​31​] Cross country runners 8 weeks 60 (31) NR 28 (13) NR Navicular drop
Medial tibial stress
syndrome
Plisky ​et al.​ , 2007 [​37​] High school runners 13 weeks Medial
15 (13) 15.3 (1.0) 21 (8) 15.7 (1.5) Resting calcaneal
tibial stress
position
syndrome
(degrees)
16 (11) NR 88 (29) NR Navicular drop
Yates and White, 2004 [​32​] Naval recruits 10 weeks basic
training Medial tibial stress
Hubbard ​et al​., 2009 [​38​] Collegiate athletes One athletic
syndrome season Medial tibial stress
40 (18) NR 84 (22) NR FPI-8
syndrome
29 (9) 19 (0.98) 117 (72) 19.9 (1.8) Navicular drop
Burne ​et al.​ , 2004 [​33​] Military cadets 12 months Medial
tibial stress
Bennett ​et al.​ , 2012 [​39​] Cross country runners Cross
syndrome country season Medial tibial stress
23 (11) NR 135 (25) NR Resting calcaneal position
syndrome
(degrees) 26 (13) NR NR 33 (15) Navicular drop
 Thijs ​et al.​ , 2008 [​42​] Recreational runners 10 weeks
Yagi ​et al.​ , 2013 [​40​] High school runners 3 years Medial Patellofemoral
tibial stress pain
syndrome 17 (16) 39.4 (10.3) 85 (73) 37.6 (9.4) FPI-6
102 (44) NR 142 (54) NR Navicular drop

Boling ​et al​., 2009 [​43​] Naval recruits 1-2.5 years

Hetresoni ​et al​., 2006 [​41​] Infantry recruits 14 weeks basic Patellofemoral
training Patellofemoral pain
pain 40 (16) NR 1279 (489) NR Navicular drop
61 (NR) NR 344 (NR) NR Resting calcaneal position
(degrees)

Beynnon ​et al​., 2001 [​44​] Collegiate athletes One college season Foot/ankle injury 20 (13) NR 98 (55) NR Longitudinal arch angle

Cain ​et al.​ , 2007 [​45​] Male Futsal players One Futsal season Foot/ankle injury 33 (0) NR 43 (0) NR FPI-6
12 (12) NR 89 (89) NR Subtalar joint ROM
Winfield ​et al.​ , 1997 [​46​] Female marines 10 weeks basic
(Goniometry)
training Bone stress
reaction

Kaufman ​et al​., 1999 [​47​] Male Navy Seal candidates 2 Years LL overuse injury 149 (0) NR 300 (0) NR Longitudinal arch angle
Table 1 Summary of study characteristics ​(Continued)

Burns ​et al.​ , 2005 [​48​] Triathletes 10 weeks LL overuse injury 37 (NR) NR 91 (NR) NR FPI-8

Rauh ​et al​., 2010 [​49​] Female marines 13 weeks LL overuse injury 104 (110) NR 644 (634) NR Longitudinal arch angle

Buist ​et al.​ , 2010 [​50​] Novice runners 13 weeks LL overuse injury 100 NR 476 NR Navicular drop

Nielsen ​et al.​ , 2014 [​51​] Novice runners 12 Months LL overuse injury 252 (NR) NR 478 (NR) NR FPI-6
LL = lower limb; NR = not reported; FPI = foot posture index.
Results ​Search results ​The electronic database search
at least one high quality study; or from multiple moderateyielded a total of 33,518 citations across the two parts of this
quality or low quality studies which are statisticallysystematic review (static foot posture and dynamic foot
homogenous. function). Follow- ing the sequential review of titles, abstracts
and full texts, as well as removing studies that were not pro-
spective cohort studies, 32 studies that evaluated static
Limited evidence ​Results from one high quality study or measures of foot posture were identified [15-25,31-51] (Figure
multiple moderate or low quality studies that are statistically 1). Full text versions of these were assessed for eligibility
heterogeneous. based on static foot posture assessment, and 21 studies met the
eligibility criteria [31-51], which were grouped according to
injury type.
Very limited evidence ​Results from one moderate quality study
or one low quality study.
Quality assessment of included studies ​Based on EAI
evaluation, quality scores ranged from 0.8 to 1.7 (out of a
No evidence P ​ ooled results insignificant and derived from possible score of 2.0), with the majority of studies included in
mul- tiple studies regardless of quality that are statistically this review being of moderate
heterogeneous.
Neal ​et al. Journal of Foot and Ankle Research ​2014, ​7​:55 Page 6 of 13 http://www.jfootankleres.com/content/7/1/55
Records identified by search strategy:

8,300 Medline and Embase

11,616 SportDiscus

13,602 CINAHL

35,518 titles and abstracts screened

35,437 excluded no relevance to research question

81 full text obtained

14 excluded as relating to dynamic foot function

67 studies relating to static foot posture

35 studies excluded as not prospective in study design

32 studies relating to static foot posture

 11 excluded due to non-quantitative method of foot posture

21 studies investigated quantitative static foot

posture and overuse injury

Figure 1 ​Search results throughout the review process.

11.5% to 44.1% [31-40,49]; patellofe- moral pain 3.0% to
quality (MQ) (n=13, 62%) [31-34,36,40,42-45,47,49,50] 15.7% [41-43]; foot and ankle injury 16.9% to 32.0% [44,45];
(Additional file 4). Five studies (24%) were classified as high bone stress reaction 11.5% [46] and a pooled group of
quality (HQ) [35,37,39,48,51], and three studies (14%) as low non-specific lower limb overuse injuries 13.9% to 37.5%
quality (LQ [38,41,46]). In terms of inter- rater reliability [47-51].
across 35 items included in the quality assessment, 14 items
had perfect or near perfect agree- ment. That is, these items Outcome measure of choice varied E ​ ight studies investigated
were awarded the same score or there was a maximum of one navicular drop [35-40,43,50], five studies investigated the foot
point difference in sco- ring. For a further 15 items, the raters posture index [32,42, 45,48,51], four studies investigated
had near perfect agreement for >80% of the studies reviewed. resting calcaneal pos- ition [31,33,34,41], three studies
Item 35 ( is prior history of disease and/or symptoms collected investigated the longi- tudinal arch angle [44,47,49] and one
and included in the analysis ) displayed the lowest agree- ment, study investigated subtalar joint goniometry [46].
with perfect or near perfect agreement for only 11 of 21Medial tibial stress syndrome (MTSS) ​Eleven studies
studies. Percentage agreement across the 35 items ranged from[31-40,49] investigated foot posture as a risk factor for the
33% to 100%. development of MTSS, nine of which provided data suitable
Common themes relating to categories of methodo-for meta-analysis [31,32,34-40,49]. Strong evidence from
logical quality were identified using the EAI [26]. High qualitycontinuous scaled measures of foot posture (including
studies scored well for relevant descriptions (e.g. hypothesis,navicular drop, calcaneal ever- sion and FPI) indicated that
risk factors, participants), statistical para- meters and resultindividuals exhibiting a more pronated foot posture were more
reporting, as well as adherence to pro- spective methodology.likely to develop MTSS (3 HQ [35,37,39], 4 MQ [31,34,36,40]
2​
Poor quality studies generally failed to perform a powerand 1 LQ [38]), with a small but significant pooled SMD (I​ =
calculation with regards to sample size [31-36,38,40,42-50];0%, ​p = ​ 0.56, SMD 0.28, 0.14 to 0.42) (Figure 2). When strati-
demonstrated inad- equate or absent reporting of reliability andfying for foot posture measure, a significant risk asso- ciation
validity, both for outcome measurewas seen for all three measures, including the FPI (very limited
[31-34,38,40-42,44-47,49-51] and injury determinantevidence, medium SMD 0.62, 0.23 to 1.02), calcaneal eversion
[31-36,38,39,41-46,48-50]; inadequate or absent description of(limited evidence; I​2 ​= 0%, ​p =
​ 0.51; small SMD 0.33, 0.05 to
intrinsic and extrinsic variables [39,41,43,46,47]; and0.61), and navicular drop (Strong evidence; I​2 ​= 0%, ​p = ​ 0.82;
inadequate adjustment for these variablessmall SMD 0.19, 0.01 to 0.36). Limited evidence from pooled
[31,32,36,39,41-44,46,47]. dichotom- ous measures (2 MQ [36,49]) indicated no
association between foot type (defined by navicular drop
Study characteristics ​Foot posture variables as risk factors magnitude >10 millimeters) and increased risk of MTSS
for lower limb overuse injuries ​The 21 included studies develop- ment (RR 1.09, 0.78 to 1.52) (Figure 3).
incorporated a total of 6,228 participants. The participant
population varied, with ten studies investigating recreational Patellofemoral pain ​Four studies [41-43,49] investigated foot
level runners [34-38,40,42,44,50,51], seven studies posture as a risk factor for the development of patellofemoral
investigating mili- tary personnel [32,33,41,43,46,47,49], two pain, three of which provided data suitable for effect size
studies in- vestigating cross-country runners [31,39], and singlecalculation [42,43,49]. Very limited evidence from con-
studies investigating futsal players [45] and triathletes [48]. Thetinuous measures indicated that individuals exhibiting
types and incidence of lower limb overuse in- jury were: MTSSincreased pronated foot posture measured using navicular drop
are more likely to develop patellofemoral pain (1 MQ [43]), of the ankle affecting performance or limiting sporting par-
with a small SMD (0.33, 0.02 to 0.65) (Figure 4). Limited ticipation [45]). One study provided data suitable for risk ratio
evidence from pooled dichotomous measures (2 MQ [42,49]) calculation [44] (Figure 6). Very limited evidence from
indicated no association between a pro- nated foot posture dichotomous measures (1 MQ [44]) indicated no association
(defined by FPI and navicular drop) and increased risk of between foot posture (defined by longitu- dinal arch angle) and
patellofemoral pain development (RR 1.22, 0.73 to 2.02) increased risk of foot/ankle injury development (RR 0.92, 0.38
(Figure 5). to 2.24).

Foot/ankle injury ​Two studies [44,45] investigated foot posture Bone stress reaction ​One study investigated foot posture as a
as a risk factor for the development of foot/ankle injury risk factor for the development of bone stress reaction, using
(defined as any tissue damage, pain and/or physical complaint subtalar joint goniometry [46]. This measured the gross total
​ 014, ​7​:55 Page 7 of 13 http://www.jfootankleres.com/content/7/1/55
Neal ​et al. Journal of Foot and Ankle Research 2

Non-specific lower limb overuse injury ​Four studies

investigated foot posture as a risk factor for non-specific lower
limb overuse injury [47,48,50,51], two of which provided data
suitable for effect size calculation [48,51]. Limited evidence
from continuous measures indi- cates no association between
individuals exhibiting a more pronated foot posture (defined by
FPI) and injury develop- ment (1 HQ [48], SMD 0.50, 2.28 to
1.28). Limited evi- dence from dichotomous measures (1 HQ
[51]) found no association between a foot posture (defined by
FPI) and
increased risk of non-specific lower limb overuse injury (RR
1.18, 0.68 to 2.04) (Figure 7).

Discussion ​This is the first systematic review and

meta-analysis of prospective research regarding static foot
posture and its relationship to lower limb overuse injury
development. Findings showed that a pronated foot posture was
a risk factor for the development of both MTSS and patellofe-
moral pain. However, associated effect sizes were small,
indicating this relationship is weak and only a part of the
multifactorial etiology.
Across the 21 studies included in this review, four dif-
range of rearfoot motion from supination to pronation in aferent measures of static foot posture were employed, in-
non-weight bearing position, with a higher available rangecluding navicular drop (n = 9), the FPI (n = 5), calcaneal
indicating increased static pronation. We were unable toeversion (n = 4) and the longitudinal arch angle (n = 3).
calculate effect sizes due to inadequate data reporting.
​ 014, ​7​:55 Page 8 of 13 http://www.jfootankleres.com/content/7/1/55
Neal ​et al. Journal of Foot and Ankle Research 2
Figure 2 ​Forest plot detailing SMD for medial tibial stress syndrome.
 Figure 3 ​Forest plot detailing RR for medial tibial stress syndrome.
drop and the FPI [52,53]. They also indicate that calca- neal
eversion may be less sensitive in identifying risk of lower limb
injury development when compared to the FPI and navicular
drop. Further prospective research concurrently collecting
multiple foot posture measures is needed to confirm this.
Direct comparison of findings with the review of Tong
and Kong [11] is difficult, due to analogous data being
unavailable within their results. They concluded that both high
arched (supinated) and flat foot (pronated) types are risk factors
for lower extremity injury, but did not provide a breakdown for
individual pathologies or outcome measures. Our findings are
Interestingly, the findings of this review, which link foot in agreement with the MTSS systematic review of Newman ​et
posture measured using navicular drop and FPI to injury risk al​. [54], which reported greater navicular drop magnitude to be
(MTSS and patellofemoral pain respectively), are in conflict a risk factor (SMD = 0.26 for continuous scaled data; risk
with reported findings from two studies that were not included ratio=1.99 for nominal scaled data). Two additional studies
within the meta-analysis due to an ab- sence of adequate data completed by Reinking and colleagues [35,36]
[33,41]. Specifically, calcaneal eversion was reported to beare included in our review, which may explain the small
unrelated to both MTSS [33] and patellofemoral pain [41] risk.variance in our statistical findings (SMD=0.19) com- pared to
These conflicting findings may be explained by the varyingNewman ​et al​. (SMD 0.26) [54]. Importantly, these additional
method of foot posture measurement, and the inferior reliabilityfindings provide further confirmation of a relationship of small
of the calcaneal eversion measurement compared to navicular
effect between greater navicular drop and risk of MTSS.
Limited and very limited evidence indicated that static Clinical implications ​Although a relationship between a
foot posture may not be a risk factor for the develop- ment of a pronated foot posture and greater risk of MTSS and
pooled group of non-specific lower limb over- use injuries or patellofemoral pain was identified, the associated pooled
foot and ankle injuries, respectively. In these cases the broad SMDs indicate a small effect (0.28 and 0.33, respectively).
and ambiguous definitions of pathology may have made Therefore, whilst a pronated foot posture may provide an
determining precise relation- ships with foot posture difficult. indication of in- jury risk, other factors should also be
Rather than combining all lower limb overuse injuries in considered. Both MTSS and patellofemoral pain are considered
analyses, future studies should prioritise evaluation of discrete, to have a multi-factorial etiology [54,55]. It is important that
well-defined con- ditions, which will enable more accurate clini- cians consider additional established risk factor variables
identification of foot posture risk factors for specific injuries. such as altered hip kinematics [3,56], increased body
​ 014, ​7​:55 Page 9 of 13 http://www.jfootankleres.com/content/7/1/55
Neal ​et al. Journal of Foot and Ankle Research 2

Figure 4 ​Forest plot detailing SMD for patellofemoral pain.

Figure 5 ​Forest plot detailing RR for patellofemoral pain.
 studies eligible for inclusion in the current sys- tematic review
provided data suitable for meta-analysis, and obtaining this
through corresponding author contact was unsuccessful in all
instances. Therefore, the meta- analysis did not encompass all
potentially available data, reducing confidence in its results.
Complete reporting of
all available data (i.e. group means, standard deviations, as well
as participant numbers) in future prospective studies evaluating
the potential risk of foot posture to lower limb injury is
encouraged to facilitate future meta- analyses.
The average methodological quality of studies in this
review was moderate, suggesting a dearth of high quality
research in this area. Less than 50% of the studies included in
mass index [1] and limited running experience [54] in this review [33,36,37,39,43,48] reported the reproducibility of
their outcome measures; a metho- dological limitation that
evaluating possible risk factors.
should be addressed in future research. Additionally, many
Another possible reason for the limited relationship bet-
studies failed to estimate their sample size based on a power
ween foot posture and injury risk may be the limitation of static
calculation, cite validity and reliability data for injury
measures to predict dynamic function. This has been the
determinant, or ad- equately adjust for covariates.
subject of much research, with differing conclu- sions drawn
Unfortunately, this further reduces the confidence in the results
regarding any association; seemingly depen- ding upon the
of our meta- analysis, but these methodological issues were
static measure implemented [57-60]. Static measures of
taken into account during the allocation of levels of evidence
navicular height are not strongly correlated with dynamic
for each finding. Future studies should seek to improve upon
navicular motion [61] and although the FPI has been shown to
the above limitations, as it will increase the strength of
correlate with dynamic measures of foot function, the strength
evidence than can be recommended.
of this correlation has varied from weak to strong [62,63].
Length of follow up varied greatly (eight weeks to three
Additionally, Barton ​et al​. [64] found that dynamic measures
years), which may have an impact on injury rates and thus may
were predictive of foot orthoses outcomes in patellofemoral
affect the validity of data pooling. Future studies should seek to
pain whilst static measures of foot posture were not. Dynamic
employ a longer duration of follow up with consideration of
measures of foot function may well have a stronger relationship
multiple time points to facilitate com- parison between trials.
and as such may be of greater priority during clinical exami-
An additional consideration related to data pooling is the
nation. This is explored in the accompanying dynamic re- view
variation in populations studied (e.g. military and running
[14], which indicates plantar loading variables are risk factors
athletes), which affect loa- ding volumes and subsequent injury
for both patellofemoral pain and Achilles tendinopathy.
risk. Nonetheless,
Limitations and recommendations for future research ​Not all
​ 014, ​7​:55 Page 10 of 13 http://www.jfootankleres.com/content/7/1/55
Neal ​et al. Journal of Foot and Ankle Research 2
Figure 6 ​Forest plot detailing SMD for foot and ankle injury.
 Figure 7 ​Forest plot detailing SMD for non-specific lower limb overuse injury.
pronated foot type to be associated with increased risk.
considering the paucity of research currently available, it wasInterestingly, one of the included studies containing insufficient
felt that the pooling conducted was valuable to strengthen data for meta-analysis reported on supinated foot postures and
findings of the review. Only studies from sporting and militaryinjury risk [45], reporting that a supi- nated foot type based on
populations were found to be eli- gible and future studies the FPI is a risk factor for foot and ankle injury. Considering
these findings, it is recom- mended future studies consider
investigating the impact of foot posture on injury risk in other
occupational settings is warranted to determine the categorising individuals with supinated foot postures to
generalizability of these findings to other populations. evaluate the potential link between this foot posture and
The majority of findings in this review indicating a link increased injury risk. This would allow similar reviews and
between foot posture and lower limb injury risk, found a more appropriate meta- analysis to evaluate the potential link
between a supi- nated foot posture and injury risk.
 To improve the clinical applicability of results achieved,competing interests.
future studies should seek to describe participants in rela- tion
Authors contributions GSM, MMFS, BSN, IBG, CJB, SEM and NJC
to both intrinsic (e.g. body mass index) and extrinsic (e.g.conceived the idea for this review. GSM, MMFS, BSN, IBG designed and
footwear) covariates and report risk factor statistics based onpiloted the search strategy. GJD undertook the search. Title and abstracts
combining static foot posture data with such co- variates. Inwere reviewed by BSN and IBG. Quality appraisal was undertaken by BSN
comparison to nominal scaled outcome mea- sures, continuousand IBG. Study information extracted by BSN and IBG. Data was extracted
and meta-analysis was completed by BSN and CB. The manuscript was
scaled outcome measures appear to be stronger predictors ofdrafted by BSN, IBG, CJB, GJD, GSM, MMFS, SEM and NJC. All authors
injury development, particularly in relation to a pronated foothave read and approved the final manuscript.
posture. However, because it is simpler to relate injury risk to a
defined value, nominal scaled measures may be moreAuthor details ​1​Pure Sports Medicine, London, UK. ​2​Centre for Sports and
applicable when screening for injury risk in clinical practice.
Exercise Medicine, Queen
​ Mary University of London, London, UK.
Future studies that use both continuous scaled and nominal3​Department of Podiatry, Faculty of Health Sciences, La Trobe University,
scaled data from an outcome measure (where possible) may beMelbourne, Australia. 4​​ Lower Extremity and Gait studies program, Faculty
5​
useful in this regard, to allow for both statistical and clinicalof Health Sciences, La Trobe University, Melbourne, Australia. ​ School of
Physiotherapy, Australian Catholic University, Brisbane, Australia.
conclusions to be reached. 6​
Department of Mechanical Engineering, Melbourne School of Engineering,
The University of Melbourne, Melbourne, Australia. 7​​ Complete Sports Care,
Conclusions ​Strong and very limited evidence indicates thatMelbourne, Australia.
a pro- nated foot posture increases the risk of MTSS and patel-
Received: 27 August 2014 Accepted: 30 November 2014
lofemoral pain, respectively. However, this relationship is of
small effect, indicating that a pronated foot posture may only
be a minor component of the injury risk pro- file for these References 1. Moen MH, Bongers T, Bakker EW, Zimmerman WO, Weir A,
conditions. Foot posture was not found to be associated with Tol JL, Backx FJG:
Risk factors and prognostic indicators for medial tibial stress syndrome.
the risk of foot and ankle injury, bone stress reactions or a
Scand J Med Sci Sports 2 ​ 012, 22:34 39. 2. Neely FG: Intrinsic risk factors
pooled group of non-specific lower limb overuse injuries, for exercise related lower limb injuries.
although caution with inter- pretation is needed here since only Sports Med 1​ 998, 24:253 263. 3. Noehren B, Davis I,
very limited to limited evidence exists. Of the measures used inHamill J: Prospective study of the biomechanical factors associated
the cur- rently available prospective research, it appears that with iliotibial band syndrome. ​Clin Biomech ​2007, 22:951 956. 4.
Tiberio D: The effect of excessive subtalar joint pronation on
navicular drop and FPI can predict lower limb overuse injury,
patellofemoral mechanics: a theoretical model. ​J Orthop Sports Phys
however dynamic measures of foot function may Ther ​1987, 9:161 165. 5. Cornwall MW, McPoil TG: Footwear and foot
display stronger relationships with injury risk. Static measures orthotic effectiveness
of foot posture should be used as part of a multifactorial injury research: a new approach. ​J Orthop Sports Phys Ther ​1995,
21:337 344. 6. Nigg BM, Cole GK, Nachbauer W: Effects of arch height
risk assessment and not considered in isolation.
of the foot on
angular motion of the lower extremities when running. ​J Biomech ​1993,
26:909 916. 7. Souza TR, Pinto RZ, Trede RG, Kirkwood RN, Fonseca
Additional files
ST: Temporal
couplings between rearfoot shank complex and hip joint during walking.
Additional file 1: ​PRISMA statement checklist. ​Additional file 2: ​ 010, 25:745 748. 8. Sharma J, Golby J, Greeves J,
Clin Biomech 2
Search strategy. ​Additional file 3: ​Epidemiological Appraisal Spears IR: Biomechanical and lifestyle risk
Instrument used to rate the quality of the 21 included studies.
factors for medial tibial stress syndrome in army recruits: a prospective
Additional file 4: ​Results from quality assessment using the
study. ​Gait Posture ​2011, 33:361 365. 9. Powers CM, Bolgla L,
Epidemiological Appraisal Instrument (21 included studies).
Callaghan M, Collins N, Sheehan F: Patellofemoral pain: proximal, distal
and local factors. ​J Orthop Sports Phys Ther ​2012, 42:A1 A20.

Competing interests The authors declare that they have no

​ 014, ​7​:55 Page 11 of 13 http://www.jfootankleres.com/content/7/1/55
Neal ​et al. Journal of Foot and Ankle Research 2
injuries: systematic literature review with meta-analysis. ​J Orthop Sports
Phys Ther ​2013, 43:700 714. 12. Chuter VH, de Jonge XAK J: Proximal
10. Williams DS, Davis IM, Scholz JP, Hamill J, Buchanan TS:
and distal contributions to lower
High-arched
​ 012, 36:7
extremity injury: a review of the literature. ​Gait Posture 2
runners exhibit increased leg stiffness compared to low-arched runners.
15. 13. Liberati A, Altman DG, Tetzlaff J, Mulrow C, Gotzsche PC,
Gait Posture ​2004, 19:263 269. 11. Tong JWK, Kong PW: Association
Ioannidis JP, Clarke M, Devereaux PJ, Kleijnen J, Moher D: The PRISMA
between foot type and lower extremity
statement for reporting systematic reviews and meta-analyses of studies factors associated with exertional medial tibial pain: a 12
that evaluate health care interventions: explanation and elaboration. ​Ann month prospective clinical study. ​Br J Sports Med 2​ 004,
Int Med ​2009, 151:W65 W94. 14. Dowling, GJ, Murley, GS, Munteanu, SE, 38:441 445.
Franettovich Smith, MM, Neal, BS, Griffiths, IB, Barton, CJ, Collins, NJ: 34. Willems TM, De Clerq D, Delbaere K, Vanderstraeten G, De
Dynamic foot function as a risk factor for lower limb overuse injury: a Cock A,
systematic review. ​J Foot Ankle Res i​ n press. 15. Reinking MF, Austin TM, Witvrouw E: A prospective study of gait related risk factors for exercise
Hayes AM: Risk factors for self reported exercise ​ 006, 23:91 98. 35. Reinking M:
related lower leg pain. ​Gait Posture 2
​ 010,
related leg pain in high school cross country athletes. ​J Athl Train 2 Exercise related leg pain in female collegiate athletes. ​Am J
45:51 57. 16. Witvrouw E, Lysens R, Bellemans J, Cambier D, Sports Med 2 ​ 006, 34:1500 1507. 36. Reinking MF, Austin TM,
Vanderstraeten G: Intrinsic Hayes AN: Exercise related leg pain in collegiate cross country athletes:
risk factors for the development of anterior knee pain in an athletic extrinsic and intrinsic risk factors. ​J Orthop Sports Phys Ther ​2007,
population. ​Am J Sports Med 2​ 000, 28:480 489. 17. Baumhauer JF, 37:670 678. 37. Plisky MS, Rauh MJ, Heiderscheit B, Underwood FB,
Alosa DM, Renstrom PAFH, Trevino S, Beynnon B: A Tank RT: Medial tibial
​ 995,
prospective study of ankle injury risk factors. ​Am J Sports Med 1 stress syndrome in high school cross country runners: incidence and risk
23:564 570. 18. Mei-Dan O, Kahn G, Zeev A, Rubin A, Constantini N, ​ 007, 37:40 47. 38. Hubbard TJ,
factors. ​J Orthop Sports Phys Ther 2
Even A, Nyska M, Mann Carpenter EA, Cordova ML: Contributing factors to medial
G: The medial longitudinal arch as a possible risk factor for ankle sprains: atibial stress syndrome: a prospective investigation. ​Med Sci Sport Ex
​ 005,
prospective study in 83 female infantry recruits. ​Foot Ankle Int 2 2009, 41:490 496. 39. Bennett JE, Reinking MF, Rauh MJ: The
26:180 183. 19. Dahle LK, Mueller M, Delitto A, Diamond JE: Visual relationship between isotonic
assessment of foot type plantar flexor endurance, navicular drop and exercise related lower leg
and relationship of foot type to lower extremity injury. ​J Orthop Sports pain in a cohort of collegiate cross country runners. ​Int J Sports Phys Ther
Phys Ther ​1991, 14:70 74. 20. Cowan DN, Jones BH, Robinson JR: Foot 2012, 7:267 278. 40. Yagi S, Muneta T, Sekiya I: Incidence and risk
morphologic characteristics and factors for medial tibial
risk of exercise related injury. ​Arch Fam Med ​1993, stress syndrome and tibial stress fracture in high school runners. ​Knee
2:773 777. 21. Twellar M, Verstappen FTJ, Huson A, van Surg Traumatol Arthrosc 2 ​ 013, 21:556 563. 41. Hetresoni I, Finestone A,
Mechelen W: Physical Milgrom C, Ben Sira D, Nyska M, Radeva-Petrova D,
characteristics as risk factors for sports injuries: a four year prospective Ayalon M: A prospective biomechanical study of the association between
study. ​Int J Sports Med ​1997, 18:66 71. 22. Michelson JD, Durrant DM, foot pronation and the incidence of anterior knee pain among military
McFarland E: The injury risk associated with recruits. ​J Bone Joint Surg (Br) ​2006, 88-B:905 908. 42. Thjis Y, De Clercq
​ 002, 23:629 633. 23. Lun V,
pes planus in athletes. ​Foot Ankle Int 2 D, Roosen P, Witvrouw E: Gait related intrinsic risk
Meeuwisse WH, Stergiou P, Stefanyshyn D: Relation between factors for patellofemoral pain in novice recreational runners. ​Br J Sports
running injury and static lower limb alignment in recreational runners. ​Br J Med 2​ 008, 42:466 471. 43. Boling MC, Padua DA, Marshall SW,
Sports Med 2 ​ 004, 38:576 580. 24. Esterman A: Foot shape and its effect Guskiewicz K, Pyne S, Beutler A: A
on functioning in Royal Australian prospective investigation of biomechanical risk factors for patellofemoral
air force recruits. Part 1: Prospective cohort study. ​Mil Med ​2005, pain syndrome. ​Am J Sports Med ​2009, 37:2108 2116. 44. Beynonn BD,
170:623 628. 25. Di Caprio F, Buda R, Mosca M, Calabro A, Giannini S: Renstrom PA, Alosa DM, Baumhauer JF, Vacek PM: Ankle ligament injury
Foot and lower limb diseases in runners: assessment of risk factors. ​J Sci risk factors: a prospective study of college athletes. ​J Orthopt Res 2 ​ 001,
Med Sport 2 ​ 010, 9:587 596. 26. Genaidy AM, LeMasters GK, Lockey J, 19:213 220. 45. Cain LE, Nicholson LJ, Adams RD, Burns J: Foot
Succop P, Deddens J, Sobeih T, morphology and foot/ankle
Dunning K: An epidemiological appraisal instrument - a tool for ​ 007, 10:311 319. 46. Winfield
injury in indoor football. ​J Sci Med Sport 2
​ 007, 50:920 960. 27.
evaluation of epidemiological studies. ​Ergonomics 2 C, Bracker M, Moore J, Johnson CW: Risk factors associated with
Menz HB: Two feet, or one person? Problems associated with statistical stress reactions in female marines. ​Mil Med ​1997, 10:698 702.
analysis of paired data in foot and ankle medicine. ​Foot ​2004, 14:2 5. 28. 47. Kaufman KR, Brodine SK, Shaffer RA, Johnson CW, Cullison TR: The
Higgins JP, Thompson SG, Deeks JJ, Altman DG: Measuring effect of foot structure and range of motion on musculoskeletal overuse
inconsistency njuries. ​Am J Sports Med 1 ​ 999, 27:585 593. 48. Burns J, Keenan A,
in meta-anslyses. ​BMJ ​2003, 327:557 60. 29. Hume P, Hopkins Redmond A: Foot type and overuse injuries in
W, Rome K, Maulder P, Coyle G, Nigg B: Effectiveness of ​ 005, 95:235 241. 49. Rauh
triathletes. ​J Am Podiatr Med Assoc 2
foot orthoses for treatment and prevention of lower limb injuries. ​Sports MJ, Macera CA, Trone DW, Reis JP, Shaffer RA: Selected static
Med 2 ​ 008, 38:759 779. 30. van Tulder M, Furlan A, Bombardier C, Lex B: anatomic measures predict overuse injuries in female recruits. ​Mil Med
Updated method guidelines 2010, 175:329 335. 50. Buist I, Bredeweg SW, Lemmick KAPM, van
for systematic reviews in the cochrane collaboration back review group. Mechelen W, Diercks RL:
Spine ​2003, 28:1290 1299. 31. Bennett JE, Reinking MF, Pluemer B, Predictors of running related injuries in novice runners enrolled in a
Pentel A, Seaton M, Killian C: Factors systematic training program. ​Am J Sports Med ​2010, 38:273 280. 51.
contributing to the development of medial tibial stress syndrome in high Nielsen RO, Buist I, Parner T, Nohr EA, Sorenson H, Lind M,
school runners. ​J Orthop Sports Phys Ther ​2001, 31:504 510. 32. Yates Rasmussen S:
B, White S: The incidence and risk factors in the development of Foot pronation is not associated with increased injury risk in novice
medial tibial stress syndrome among naval recruits. ​Am J Sports Med runners wearing a neutral shoe: a 1 year prospective cohort study. ​Br J
2004, 32:772 780. 33. Burne SG, Khan KM, Mallet RJ, Newman PM, Sports Med 2 ​ 014, 48:440 447. 52. Razeghi M, Batt ME: Foot type
Steinman LJ, Thornton E: Risk classification: a critical review of current
 methods. ​Gait Posture ​2002, 15:282 291. 53. Jarvis HL, consensus statement from the 3rd international patellofemoral pain
Nester CJ, Jones RK, Williams A, Bowden PD: Inter-assessor research retreat held in Vancouver, September 2013. ​Br J Sports Med
reliability of practiced based biomechanical assessment of the foot and 2014, 48:411 414. 56. Noehren B, Hamill J, Davis I: Prospective evidence
ankle. ​J Foot Ankle Res ​2012, 5:1 10. 54. Newman P, Witchalls J, for a hip etiology in
Waddington G, Adams R: Risk factors associated patellofemoral pain. ​Med Sci Sport Ex ​2012, 45:1120 4. 57.
with medial tibial stress syndrome in runners: a systematic review. ​Open McPoil TG, Cornwall MW: Relationship between three static angles of the
Access J Sports Med 2 ​ 013, 4:229 241. 55. Witvrouw E, Callaghan MJ, rearfoot and the pattern of rearfoot motion during walking. ​J
Stefanik JJ, Noehren B, Bazett-Jones DM, Willsom J, Earl-Beohm JE, Orthopt Sports Phys Ther ​1996, 23:370 375.
Davis IS, Powers CM, McConnell J, Crossley KM: Patellofemoral pain:
​ 014, ​7​:55 Page 12 of 13 http://www.jfootankleres.com/content/7/1/55
Neal ​et al. Journal of Foot and Ankle Research 2
Neal ​et al. Journal of Foot and Ankle Research 2​ 014, ​7​:55 Page 13 of 13 http://www.jfootankleres.com/content/7/1/55

58. Cashmere T, Smith R, Hunt A: Medial longitudinal arch of the foot:

static
versus walking measures. ​Foot Ankle Int ​1999, 20:112 118. 59.
Franettovich MM, McPoil TG, Russell T, Skardoon G, Vicenzino B: The
ability to predict dynamic foot posture from static measurements. ​J Am
Podiatr Med Assoc 2 ​ 007, 97:115 120. 60. Bandholm T, Boysen L,
Haugaard S, Zebis MK, Bencke J: Foot medial
longitudinal arch deformation during quiet standing and gait in subjects with
​ 008, 47:89 95. 61. Deng
medial tibial stress syndrome. ​J Foot Ankle Surg 2
J, Joseph R, Wong CK: Reliability and validity of the sit to stand
navicular drop test: do static measures of navicular height relate to
dynamic navicular motion during gait. ​J Student Phys Ther Res ​2010,
2:21 28. 62. Chuter V: Relationships between foot type and dynamic
rearfoot frontal
​ 010, 3:1 6. 63. Nielsen RG,
plane motion. ​J Foot Ankle Res 2
Rathleff MS, Moelgaard CM, Simonsen O, Kaalund S, Olesen CG,
Christensen FB, Kersting UG: Video based analysis of dynamic midfoot
function and its relationship with foot posture index scores. ​Gait Posture
2010, 31:126 130. 64. Barton CJ, Menz HB, Levinger P, Webster KE,
Crossley KM: Greater peak
rearfoot eversion predicts foot orthoses efficacy in individuals
with patellofemoral pain syndrome. ​Br J Sports Med 2 ​ 011,
45:697 701.

doi:10.1186/s13047-014-0055-4 Cite this article as: Neal ​et al.​: Foot

posture as a risk factor for lower limb overuse injury: a systematic
review and meta-analysis. ​Journal of Foot and Ankle Research 2 ​ 014
7:55.
 Submit your next manuscript to BioMed Central
and take full advantage of:

? ​Convenient online submission

? ​Thorough peer review

? ​No space constraints or color ?gure charges

? ​Immediate publication on acceptance

? ​Inclusion in PubMed, CAS, Scopus and Google Scholar

? ​Research which is freely available for redistribution

Submit your manuscript at

www.biomedcentral.com/submit

View View publication

publication stats stats