Sports Medicine

https://doi.org/10.1007/s40279-019-01126-5

SYSTEMATIC REVIEW

Factors Influencing the Relationship Between the Functional

Movement Screen and Injury Risk in Sporting Populations:
A Systematic Review and Meta‑analysis
Emma Moore1   · Samuel Chalmers2,3   · Steve Milanese1   · Joel T. Fuller4 

© Springer Nature Switzerland AG 2019

Abstract
Background  Studies investigating the association between the Functional Movement Screen (FMS) and sports injury risk
have reported mixed results across a range of athlete populations.
Objectives  The purpose of this systematic review was to identify whether athlete age, sex, sport type, injury definition and
mechanism contribute to the variable findings.
Study design  Systematic review and meta-analysis.
Methods  A systematic search was conducted in October 2018 using PubMed, EBSCOhost, Scopus, EmBase and Web of
Science databases. Studies were included if they were peer reviewed and published in English language, included athletes
from any competition level, performed the FMS at baseline to determine risk groups based on FMS composite score, asym-
metry or pain, and prospectively observed injury incidence during training and competition. Study eligibility assessment
and data extraction was performed by two reviewers. Random effects meta-analyses were used to determine odds ratio (OR),
sensitivity and specificity with 95% confidence intervals. Sub-group analyses were based on athlete age, sex, sport type,
injury definition, and injury mechanism.
Results  Twenty-nine studies were included in the FMS composite score meta-analysis. There was a smaller effect for junior
(OR = 1.03 [0.67–1.59]; p = 0.881) compared to senior athletes (OR = 1.80 [1.17–2.78]; p = 0.008) and for male (OR = 1.79
[1.08–2.96]; p = 0.024) compared to female (OR = 1.92 [0.43–8.56]; p = 0.392) athletes. FMS composite scores were most
likely to be associated with increased injury risk in rugby (OR = 5.92 [1.67–20.92]; p = 0.006), and to a lesser extent American
football (OR = 4.41 [0.94–20.61]; p = 0.059) and ice hockey (OR = 3.70 [0.89–15.42]; p = 0.072), compared to other sports.
Specificity values were higher than sensitivity values for FMS composite score. Eleven studies were included in the FMS
asymmetry meta-analysis with insufficient study numbers to generate sport type subgroups. There was a larger effect for
senior (OR = 1.78 [1.16–2.73]; p = 0.008) compared to junior athletes (OR = 1.21 [0.75–1.96]; p = 0.432). Sensitivity values
were higher than specificity values for FMS asymmetry. For all FMS outcomes, there were minimal differences across injury
definitions and mechanisms. Only four studies provided information about FMS pain and injury risk. There was a smaller
effect for senior athletes (OR = 1.28 [0.33–4.96]; p = 0.723) compared to junior athletes (OR = 1.71 [1.16–2.50]; p = 0.006).
Specificity values were higher than sensitivity values for FMS pain.
Conclusion  Athlete age, sex and sport type explained some of the variable findings of FMS prospective injury-risk studies.
FMS composite scores and asymmetry were more useful for estimating injury risk in senior compared to junior athletes.
Effect sizes tended to be small except for FMS composite scores in rugby, ice hockey and American football athletes.
Protocol registration CRD42018092916.

1 Introduction
Electronic supplementary material  The online version of this
article (https​://doi.org/10.1007/s4027​9-019-01126​-5) contains Although injury mechanisms are complex and multifactorial,
supplementary material, which is available to authorized users.
there are a variety of intrinsic risk factors that can poten-
* Emma Moore tially lead to increased injury risk; for example, history of
emma.moore@mymail.unisa.edu.au previous injury, joint instabilities, muscular strength imbal-
Extended author information available on the last page of the article ances and anatomical impingements [1]. Musculoskeletal

Vol.:(0123456789)
 E. Moore et al.

concluded that only three prospective studies in military

Key Points  and police populations were sufficiently homogenous to
attempt a meta-analysis. Pooled results from these studies
FMS composite scores were more strongly associated suggested that an FMS composite score threshold of ≤ 14
with injury risk in senior compared to junior athletes, was associated with a small increase in all-cause injury
and in rugby, ice hockey and American football com- risk. The remaining 20 prospective studies identified in the
pared to other sports. search were considered too heterogenous in study population
The association between FMS asymmetry and injury risk (i.e. age, sex, sport) and injury definition to be included in
was stronger in senior compared to junior athletes. a meta-analysis.
All previous systematic reviews investigating the relation-
FMS composite scores and FMS pain were more specific ship between FMS composite score and subsequent injury
than sensitive for injury risk, whereas FMS asymmetry have identified substantial heterogeneity in study population
was more sensitive than specific. type, follow-up time and injury definition [12–14]. Review
authors have either ignored this heterogeneity and attempted
to draw conclusions from a potentially inappropriate meta-
screening tests are designed to identify these risk factors analysis [12, 13] or not attempted to statistically investigate
so that medical professionals can implement appropriate the effect of this heterogeneity on study outcomes [14].
training strategies (prehabilitation) in an effort to reduce Therefore, the primary aim of this review was to identify
the incidence of injury. factors that contribute to the contradictory findings of stud-
The Functional Movement Screen (FMS) is a battery ies investigating the relationship between FMS composite
of seven tests that assess fundamental movement patterns score and subsequent injury risk in sporting populations.
to identify dysfunctional, asymmetrical and painful move- Furthermore, all previous systematic reviews of FMS
ments that could contribute to future injuries [2, 3]. The research have focussed on the relationship between poor
seven tests are the deep squat, in-line lunge, hurdle step, movement quality (i.e. low FMS composite score) and injury
shoulder mobility, trunk stability push up, active straight risk without considering the relevance of FMS asymmetry or
leg raise and rotary trunk stability. Each test is scored on pain. Recent studies have reported mixed results regarding
a ranking scale of 1–3 to produce a composite score out the relationship between FMS asymmetry and injury risk
of 21, with higher scores indicating better movement. FMS in sporting populations [15, 16], and, to the authors knowl-
pain and asymmetry are dichotomous outcomes based on the edge, there has been no attempt to systematically summarise
presence or absence of pain during FMS testing and at least studies that have investigated the relationship between FMS
one FMS test difference between left and right sides of the asymmetry and injury. Additionally, pain findings are rarely
body, respectively. Early research conducted by Kiesel et al. reported in relation to injury in FMS research even though
[4] on a group of 46 professional American footballers over painful tests receive a score of zero and have a large impact
one season found that an FMS composite score of ≤ 14 was on the FMS composite score. Therefore, this review sum-
associated with more than 11-fold increased odds of future marises the effects of asymmetry and pain on subsequent
injury (sensitivity: 0.54; specificity: 0.91). Subsequent stud- injury risk in sporting populations.
ies have continued to use the ≤ 14/21 cut-off despite dem-
onstrating poor sensitivity [5, 6]. However, several studies
have also attempted to identify their own cut-off threshold 2 Methods
due to the poor sensitivity of the ≤ 14 FMS composite score
threshold [7–11]. 2.1 Design
Three recent systematic reviews have reported conflicting
results regarding the ability to infer injury risk from FMS This review followed the Preferred Reporting Items for Sys-
findings. Dorrel et al. [12] pooled results from six prospec- tematic Reviews and Meta-Analysis (PRISMA) statement
tive cohort studies and demonstrated that the ≤ 14 FMS com- for the reporting of systematic reviews and meta-analyses
posite score threshold had limited ability to predict all-cause [17]. This review was prospectively registered with PROS-
injuries in a combined military, firefighting and sporting PERO (CRD42018092916).
population (area under the ROC curve: 0.58). In contrast,
a review by Bonazza et al. [13] including nine prospective 2.2 Search Strategy and Selection Criteria
cohort studies reported a pooled odds ratio (OR) that sug-
gested scoring ≤ 14 was associated with a small threefold Medline, Scopus, EBSCOhost (including SportDiscus, Aca-
increase in all-cause injury odds in an athlete, firefighting demic Search Premiere, Health Source: Consumer Edition,
and military population. Most recently, Moran et al. [14] Health Source: Nursing/Academic Edition), Embase and
Functional Movement Screen and Sports Injury Risk

Web of Science databases were searched from inception appraisal item were agreed upon by the review team. Two
until 20 October 2018 using the following search strategy, reviewers appraised each study based on these appraisal
which was adapted for each database: definitions. Any discrepancies were resolved by a discus-
sion amongst reviewers. A grade of ‘Yes’, ‘No, ‘Unclear’ or
functional movement screen or fms or functional
‘Not Applicable’ was issued for each appraisal items. ‘Yes’
movement screen* and injur* or injury prediction
and ‘Not Applicable’ answers were indicative of a lower
or injury prevention or injury risk or injury preven-
risk of bias; therefore, the total frequency of ‘Yes’ and ‘Not
tion screening and sport* or football* or baseball* or
Applicable’ scores were tallied to indicate overall methodo-
cricket* or soccer or rugby or basketball* or hockey or
logical quality. The level of agreement between reviewers
ice hockey or volleyball* or athlet* or swim* or run*
(inter-rater reliability) was calculated via Cohen’s Kappa
or walk* or jog* or skat* or ski* or cycling or track or
(κ) analysis, using SPSS© (IBM SPSS Statistics v22.0,
field or hurdling or row* or netball or surf* or tennis or
Armonk, NY, USA). As SPSS© does not provide confi-
cross country or handball or diving or golf or softball.
dence intervals (CIs) for κ analysis, the CI was calculated
Database search results were exported to Endnote© (ver- separately using Microsoft Excel© (Excel 2016, Micro-
sion 8.2, Thomson-Reuters, Toronto, CA, USA) and then soft, Washington, USA), and interpreted using the values
uploaded to Covidence© Systematic Review Software defined by McHugh [20]. Agreement was considered to be
(Veritas Health Innovation, Melbourne, VIC, Australia). none (κ < 0.20), minimal (κ 0.21–0.39), weak (κ 0.40–0.59),
All duplicates were removed before two reviewers indepen- moderate (κ 0.60–0.79), strong (κ 0.80–0.89) and almost per-
dently screened titles and abstracts for eligibility. Full-texts fect (κ > 0.90).
were obtained for the remaining articles and independently
assessed for eligibility by two reviewers. Results from each 2.4 Data Extraction
reviewer were compared after each stage and any discrepan-
cies were resolved by a discussion amongst all four review- Data were extracted by one reviewer and entered in a stand-
ers. Reference lists of all eligible studies and any previous ardised Microsoft Excel© spreadsheet. These data were
systematic reviews were checked to identify additional eli- independently cross-checked by another reviewer and any
gible studies that were not identified by the primary search. discrepancies were resolved through discussion. Further
Inclusion criteria were as follows: (1) peer reviewed information was sought from study authors if all informa-
and published in the English language, (2) participants tion could not be obtained from the full-text article. The
were competing at any level of sporting competition, (3) extracted information included publication details (author
prospective cohort study design that assessed FMS perfor- information, publication date, country of origin), study
mance at baseline using the complete FMS test battery and methodology (sample size, FMS tester experience, follow-
subsequently observed participants during sports training up period, injury definition, injury mechanism), participant
and competition, (4) identified risk groups based on FMS characteristics (age, sex, height, body mass, sport, competi-
composite score, asymmetry or pain, and (5) outcome meas- tion level), FMS risk classification (FMS composite score
ures were injury incidence that could be categorised within threshold, asymmetry, pain) and statistical outcomes (2 × 2
the six injury level classifications provided by Orchard and contingency table data, injured athlete FMS composite
Hoskins [18]. For this review, the injury level classifications score, uninjured athlete FMS composite score, effect size,
were condensed into three levels, whereby level 1 was tissue confidence intervals, p values).
damage or presentation to medical staff, level 2 was limited
or loss of training or match, and level 3 was limited or loss 2.5 Statistical Considerations
of match only. Studies were excluded if data were presented
in formats such as theses or conference abstracts. Summary effect sizes were presented for each study using
OR for comparing injury risk between risk groups, and
2.3 Risk of Bias standardised mean differences (SMD) for comparing FMS
composite scores between injured and uninjured athletes.
An assessment of methodological quality for the selected The precision of the effect sizes was described using 95%
studies was undertaken using the Joanna Briggs Institute CIs whenever sufficient information was reported by study
(JBI) Critical Appraisal Checklist for Cohort Studies [19]. authors. For the purpose of this review, effect sizes were pre-
The JBI Critical Appraisal Checklist was used because it sented for each study as defined by Hopkins [21, 22]. Effect
is specific to the prospective cohort design of the stud- sizes were considered trivial (OR 1.00–1.49; SMD < 0.20),
ies included in this review. It was developed to determine small (OR 1.50–3.49; SMD 0.20–0.49), moderate (OR
the potential for bias in study design, content and analy- 3.50–8.99; SMD 0.50–0.79), and large (OR ≥ 9.00;
sis. Before commencing assessment, definitions of each SMD ≥ 0.80).
 E. Moore et al.

Random-effects meta-analyses were performed using the were unclear whether the FMS test results were valid and
Metafor statistical package in R© software (version 3.4.3, R reliable. Nine of 36 studies did not explicitly state that par-
Foundation for Statistical Computing). Dependent variables ticipants were injury-free at the time of testing. Five of 36
were OR, sensitivity and specificity based on risk groups studies were deemed to not describe the injury surveillance
determined by FMS composite score ≤ 14, asymmetry on method in sufficient detail. Four of 36 studies used a fol-
≥ 1 subtest or pain on ≥ 1 subtest. Sub-group analyses were low-up period that was less than one complete competitive
based on athlete age (senior vs. junior), athlete sex (male season. Twenty-three studies were unclear whether follow-
vs. female), sport type, injury definition (level 1 vs. 2 vs. 3) up was completed for all participants, and 18 were unclear
and injury mechanism (all-cause injury vs. non-contact-only whether strategies to account for incomplete follow-up were
injury). Sensitivity is defined as the probability of the test implemented. Six studies did not utilise statistical analysis
(FMS) being positive given the participant has the condition that resulted in the reporting of injury-risk statistics.
(i.e. reports an injury in the subsequent follow-up period),
while specificity is defined as the probability of the test 3.3 Qualitative Analysis
(FMS) being negative given the participant does not have
the condition (i.e. does not report an injury in the subse- Seven studies were unable to be included in the meta-
quent follow-up period) [23]. Individual study effects were analysis and therefore have been included in a qualitative
weighted using the inverse variance method. Statistical het- synthesis (Table 2). Four studies only compared the mean
erogeneity within each meta-analysis was investigated using FMS composite scores of injured and uninjured athletes. Of
I2 statistics, which indicate the consistency of study effects these four studies, one study ([46], n = 20, soccer) found a
between the included studies [24]. Statistical heterogeneity trivial harmful difference and three studies ([47], n = 119,
(i.e. inconsistency) was considered as low (I2 < 25%), moder- rugby; [48], n = 238, soccer; [49], n = 137, soccer) found
ate (I2 = 25–49%) and high (I2 > 50%) [24]. large harmful differences in senior athletes. Two studies
([9], n = 100, mixed sports; [11], n = 144, American foot-
ball) found a trivial to small harmful association between
3 Results scoring ≤ 17 and injury risk in senior athletes. One study
([50], n = 84, soccer) found a small beneficial association
3.1 Search Results between scoring ≤ 15 and injury risk in junior athletes. All
results were non-significant at the p = 0.05 level.
Systematic database searches identified 1028 potential stud- An additional three studies used multiple FMS compos-
ies and pearling of previous reviews identified three more ite score thresholds and were included in both qualitative
potential studies for inclusion. Following the removal of and quantitative analyses. One study in juniors ([7], n = 185,
duplicates and ineligible articles, 36 studies were included mixed sports) found a non-significant trivial harmful asso-
in this review. Twenty-nine studies provided sufficient ciation between scoring ≤ 13 and future injury and one study
information to be included in the meta-analysis, while the in seniors ([10], n = 62, rugby) found a significant moderate
remaining seven could only be included in a qualitative syn- harmful association between scoring ≤ 13 and future injury.
thesis. A complete list of articles identified for inclusion in The third study ([8], n = 257, mixed sports) found a non-
the meta-analysis can be found in Table 1, while the articles significant small harmful association between scoring ≤ 15
included in the qualitative synthesis can be found in Table 2. and future injury in seniors.
A complete overview of the screening process can be found
in Fig. 1. 3.4 Meta‑analysis

3.2 Risk of Bias Assessment 3.4.1 FMS Composite Score Threshold (≤ 14), Sport

and Injury Risk
Reviewers initially agreed on 364 out of 396 (91.9%) JBI
Cohort Analysis items (κ = 0.865 [0.819–0.910] p < 0.001) Twenty-nine studies provided information about an FMS
and achieved consensus on the remaining items after discus- composite score threshold of ≤ 14 and subsequent injury
sion and consideration of the operational definitions. The risk (Fig. 2). An FMS composite score of ≤ 14 was associ-
inter-rater reliability of the operational definitions can be ated with a significant (p = 0.006) moderate harmful effect
defined as strong based on the kappa analysis. in rugby athletes (k = 2) that was not affected by any sta-
Results of the risk of bias separated by category can be tistical heterogeneity and a moderate harmful effect that
found in Online Supplement 1. The mean score of the 36 approached significance (p = 0.059) in American football
studies was 7.9 ± 2.0 items. There were some noteworthy athletes (k = 2) that was affected by high statistical hetero-
results from the risk of bias analysis. Seven of 36 studies geneity. Ice hockey (k = 2) was also associated with a trend
Functional Movement Screen and Sports Injury Risk

Table 1  Summary of studies used in the meta-analysis

Author(s), year N Age (years) ± SD Sport FMS certified Follow-up period Injury defini-
tion/mecha-
nism

Avery et al. [25] 36 9 ± 1 Junior Ice Hockey Yes 1 season 1/A
Azzam et al. [26] 34 NR Senior Basketball Yes 4 seasons 3/A
Bardenett et al. [7] 167 15 Junior Mixed Sports Yes 1 season 2/A
Bond et al. [27] 119 21 College Basketball Unknown 1 season 1/A
Bring et al. [28] 183 18 College Runners Yes 1 season 2/A
Chalmers et al. [15] 237 17 ± 1 Junior Australian Footballers Yes 1 season 3/A
Chalmers et al. [16] 277 17 ± 1 Junior Australian Footballers Yes 1 season 3/A
Chorba et al. [5] 38 19 ± 1 College Mixed Sports Unknown 1 season 1/A
Clay et al. [29] 37 NR College Rowing Yes 1 season 2/A
Dorrel et al. [8] 257 NR College Mixed Sports Yes 1 season 2/A
Dossa et al. [30] 20 18 ± 1 Junior Ice Hockey Yes 1 season 3/A
Duke et al. [31] 68 22 ± 3 Senior Rugby Yes 1 season 2/A
Garrison et al. [32] 160 NR College Mixed Sports Unknown 1 season 2/A
Hotta et al. [33] 84 20 ± 1 College Running Unknown 1 season 3/NC
Kiesel et al. [4] 46 NR Senior American Football Unknown 1 season 3/A
Kiesel et al. [6] 238 NR Senior American Football Unknown 1 preseason 2/A
Kolodziej and Jaitner [34] 83 24 ± 4 Senior Soccer Yes 1 season 1/NC
Lee et al. [35] 55 15 ± 2 Junior Baseball Unknown 1 season 1/A
Martin et al. [36] 27 NR Junior Cricket No 1 season 1/A
Mokha et al. [37] 84 20 ± 1 College Mixed Sports Yes 1 season 1/A
Moran et al. [38] 117 35 ± 10 Senior Cross Fit Unknown 12 weeks 2/A
Philp et al. [39] 24 19 Senior Soccer Yes 1 season 1/A
Rusling et al. [40] 120 13 Junior Soccer Unknown 1 season 1/NC
Schroeder et al. [41] 96 24 ± 4 Senior Soccer Yes 1 season 2/NC
Slodownik et al. [42] 30 23 ± 4 Senior Handball No 1 season 1/A
Smith and Hanlon [43] 89 23 ± 4 Senior Soccer Yes 1 season 2/NC
Tee et al. [10] 62 25 ± 3 Senior Rugby Yes 1 season 2/NC
Walbright et al. [44] 36 NR College Mixed Sports Yes 1 season 2/A
Warren et al. [45] 167 NR College Mixed Sports Yes 1 season 1/NC

Injury definition: Category 1 (1—perceived injury and 2—medical presentation); Category 2 [3—limited participation (training or match) and
4—full time loss (training or match)]; Category 3 [5—limited participation (match only) and 6—full time loss (match only)]
Injury mechanism: A all cause injury, NC non contact

towards a moderate harmful effect (p = 0.072) that was not 3.4.2 FMS Composite Score Threshold (≤ 14), Age, Sex
affected by statistical heterogeneity. Pooled effects for soccer and Injury Risk
(k = 5), Australian football (k = 2), basketball (k = 2), run-
ning (k = 2) and mixed sport sub-groups (k = 7) were not An FMS composite score of ≤ 14 was associated with a sig-
statistically significant (p > 0.298) and of trivial beneficial nificant (p = 0.020) small harmful effect in senior athletes
or harmful magnitude. There was no statistical heterogene- (k = 20) that was affected by high statistical heterogeneity
ity for Australian football or basketball sub-groups but high and a non-significant (p = 0.549) trivial harmful effect in
statistical heterogeneity for running, soccer and mixed sport junior athletes (k = 8) that was affected by moderate statisti-
sub-groups. Only single studies providing non-significant cal heterogeneity (Fig. 2). An FMS composite score of ≤ 14
results were available for cricket (p = 0.406; small benefi- was associated with a small harmful effect that was signifi-
cial effect), CrossFit and handball (p > 0.181; small harmful cant in male (p = 0.024) but not female (p = 0.392) athletes
effect), and rowing (p = 0.159; moderate harmful effect). A (Fig. 2). Statistical heterogeneity was high for both male and
single baseball study provided significant results (p = 0.006) female subgroups.
with a moderate harmful effect magnitude.
 E. Moore et al.

Table 2  Qualitative analysis results

Author(s), N Age Sport Tester’s Follow-up Injury defini- Composite Effect statistic Effect size
year (years) ± SD FMS certi- period tion/mecha- score thresh- (95% CI) descriptor
fied nism old

Armstrong 119 20 College U 1 season 2/A NR SMD = 0.41 Small

and Greig Rugby (NR)
[47]
Bardenett 185 15 Junior Mixed Y 1 season 2/A 13 SMD = 0.05 Trivial
et al. [7] Sports (NR)
Dorrel et al. 257 NR College Y 1 season 2/A 15 OR = 1.50 Small
[8] Mixed (0.92–2.48)
Sports
Hammes et al. 238 44 ± 7 Senior Soccer U 1 season 2/A NR SMD = 0.96 Large
[48] (NR)
Letafatkar 100 23 ± 3 College U 1 season 2/A 17 OR = 1.56 Small
et al. [9] Mixed (0.70–3.47)
Sports
Newton et al. 84 13 ± 1 Junior Soccer Y 1 season 2/NC 15 RR = 0.66 Small
[50] (0.40–1.10)
Svensson et al. 137 NR Senior Soccer U 1 season 2/A NR SMD = 1.06 Large
[49] (NR)
Tee et al. [10] 62 25 ± 3 Elite Rugby Y 1 season 3/A 13 OR = 5.20 Moderate
(2–14)
Wiese et al. 144 19 ± 1 College U 1 season 2/A 17 OR = 1.42 Trivial
[11] American (0.60–3.20)
Football
Zalai et al. 20 23 ± 3 Senior Soccer U 6 months 2/A NR SMD = 0.13 Trivial
[46] (NR)

N no, NR not reported, OR odds ratio, RR relative Risk, SMD standardised mean difference, U unknown, Y yes
Injury Definition: Category 1 (medical presentation); Category 2 (time loss—training or match); Category 3 (time loss—match only)
Injury Mechanism: A all cause injury, NC non contact

Fig. 1  PRISMA flowchart for screening of articles

Functional Movement Screen and Sports Injury Risk

Fig. 2  Forest plot of athletes for strength of association (odds ratio) female combined, A all-cause, NC non-contact, CS composite score,
between FMS composite score threshold of ≤ 14 and subsequent TP true positive, TN true negative, FP false positive, FN false nega-
injury risk. Sn senior, Jn junior, M male, F female, Mix male and tive
 E. Moore et al.

3.4.3 FMS Composite Score Threshold (≤ 14), Injury sensitive, but non-contact injury had similar sensitivity and
Mechanism and Injury Risk specificity. Injury definitions 1 and 2 were more specific than
sensitive while injury definition 3 had similar sensitivity and
An FMS composite score of ≤ 14 tended (p = 0.058) to have specificity.
a small harmful effect with high statistical heterogeneity
for all-cause (k = 22) injury mechanisms while non-contact 3.4.5 FMS Test Asymmetry and Injury Risk
injury (k = 7) was associated with non-significant (p = 0.112)
small harmful effects with moderate statistical heterogeneity. Eleven studies provided information about ≥ 1 asymmet-
Pooled effects for injury definition 1 (k = 10), and 3 (k = 6) rical sub-tests and subsequent injury risk (Fig. 3a). Hav-
were not statistically significant (1 p = 0.524; 3 p = 0.261) ing ≥ 1 asymmetrical test was associated with significant
and of trivial to small harmful effects; injury definition 2 (p = 0.008) small harmful effects in senior (k = 7) athletes
(k = 13) was associated with significant (p = 0.025) small and non-significant (p = 0.432) trivial harmful effects in jun-
harmful effects. All injury definitions were affected by high ior (k = 4) athletes. The senior and junior athlete subgroups
statistical heterogeneity. demonstrated moderate statistical heterogeneity. Having ≥ 1
asymmetrical test was associated with significant (p = 0.043)
3.4.4 Sensitivity and Specificity of FMS Composite Score small harmful effects for all-cause (k = 8) injury and non-
Threshold (≤ 14) and Injury Risk significant (p = 0.214) trivial harmful effects for non-contact
(k = 3) injury mechanisms. All-cause injury was affected
A meta-analysis identifying the sensitivity and specificity by high statistical heterogeneity, while non-contact injury
of the ≤ 14 FMS composite score threshold was performed demonstrated no statistical heterogeneity. Having ≥ 1 asym-
on 25 studies (Table 3). Meta-analysis of sport subgroups metrical test was associated with non-significant (p > 0.177)
indicated varied sensitivity and specificity results. Australian trivial to small harmful effects for injury definitions 1 and
football was the only sport where sensitivity was greater than 3, while injury definition 2 was associated with significant
specificity. Specificity was higher than sensitivity regardless (p = 0.002) small harmful effects. Definitions 1 (k = 5) and
of athlete age or sex. All-cause injury was more specific than 3 (k = 3) were affected by moderate statistical heterogeneity,

Table 3  Pooled ≤ 14 composite Subgroup k n Sensitivity [95% CI] Specificity [95% CI]

score threshold sensitivity and
specificity meta-analysis results American Football 2 284 0.37 [0.16–0.65] 0.87 [0.82–0.91]
Australian Football 2 514 0.52 [0.41–0.63] 0.43 [0.33–0.54]
Baseball 1 55 0.58 [0.39–0.75] 0.79 [0.61–0.90]
Basketball 2 153 0.38 [0.04–0.89] 0.58 [0.07–0.96]
Cricket 1 27 0.20 [0.05–0.54] 0.65 [0.40–0.83]
Cross Fit 1 71 0.46 [0.22–0.72] 0.64 [0.51–0.75]
Handball 1 30 0.54 [0.28–0.78] 0.71 [0.46–0.87]
Ice Hockey 2 56 0.38 [0.21–0.59] 0.85 [0.44–0.98]
Mixed Sports 7 908 0.49 [0.38–0.60] 0.57 [0.48–0.68]
Rowing 1 37 0.29 [0.15–0.50] 0.92 [0.61–0.99]
Rugby 2 158 0.60 [0.15–0.93] 0.77 [0.14–0.99]
Running 2 267 0.53 [0.18–0.85] 0.58 [0.52–0.65]
Soccer 5 292 0.35 [0.16–0.60] 0.55 [0.15–0.89]
Senior Athletes 21 2033 0.43 [0.35–0.52] 0.67 [0.56–0.77]
Junior Athletes 7 819 0.50 [0.43–0.57] 0.59 [0.44–0.72]
Male Athletes 17 1534 0.46 [0.37–0.56] 0.66 [0.51–0.78]
Female Athletes 3 110 0.41 [0.24–0.60] 0.71 [0.38–0.90]
All-Cause Injury Mechanism 22 2243 0.43 [0.36–0.50] 0.66 [0.56–0.76]
Non-Contact Injury Mechanism 6 609 0.53 [0.30–0.74] 0.60 [0.42–0.75]
Injury Definition 1 9 580 0.38 [0.26–0.51] 0.67 [0.50–0.81]
Injury Definition 2 13 1574 0.42 [0.32–0.53] 0.68 [0.55–0.78]
Injury Definition 3 6 698 0.56 [0.47–0.65] 0.55 [0.34–0.74]

CI confidence interval, k number of studies, n number of participants

Functional Movement Screen and Sports Injury Risk

Fig. 3  Forest plot of athletes for strength of association (odds ratio) between injury risk and a ≥ 1 asymmetrical test and b ≥ 1 painful FMS test.
Sn Senior, Jn Junior, A all-cause, NC non-contact, AS asymmetrical, TP true positive, TN true negative, FP false positive, FN false negative

while definition 2 (k = 3) demonstrated no statistical 3.4.7 Sensitivity and Specificity of Asymmetry or Pain

heterogeneity. and Injury Risk

3.4.6 FMS Pain and Injury Risk A meta-analysis investigating sensitivity and specificity for
the relationship between the presence of ≥ 1 asymmetrical
Four studies provided information about ≥ 1 painful FMS sub-tests or ≥ 1 painful sub-tests was performed (Table 4).
testing and subsequent injury risk (Fig. 3b). Meta-analysis of The presence of ≥ 1 asymmetrical test was more sensitive
these studies presented a significant (p = 0.007) small harm- than specific when identifying athletes at increased risk of
ful effect with no statistical heterogeneity. The effect was not injury regardless of athlete age, injury mechanism and injury
statistically significant (p = 0.723) for senior athletes (k = 2) definition. The presence of ≥ 1 painful FMS tests was more
but was statistically significant (p = 0.006) for junior athletes specific than sensitive when identifying athletes at increased
(k = 2). Both subgroup effects were of trivial to small harm- risk of injury regardless of the age group being tested.
ful effect size with no statistical heterogeneity.
 E. Moore et al.

4 Discussion present review identified the greatest number of studies of

any FMS review, indicating a growing interest in the use of
The overall aim of this review was to identify factors that the FMS and its relationship with injury risk. This is also
contribute to the contradictory findings of studies investi- the first review to examine the relationship between injury
gating the relationship between FMS composite score and risk and FMS asymmetry or pain.
injury risk in sporting populations. Secondary aims were
to analyse the results of studies investigating the relation- 4.1 FMS Composite Score (≤ 14) and Injury Risk
ship between prospective injury risk and FMS asymmetry
and pain. Thirty-six studies were identified by this review 4.1.1 FMS Composite Score (≤ 14), Sport and Injury Risk
with 29 included in the meta-analysis. Overall, mixed results
emerged for the relationship between FMS composite score An FMS composite score of ≤ 14 was the most commonly
threshold and increased injury risk, although some key used injury-risk identification threshold and this meta-
findings were evident. First, the ≤ 14 FMS composite score analysis found this threshold was associated with minimal
threshold was associated with a small increase in injury increase in injury risk amongst several sporting cohorts.
risk for senior but not junior athletes and was more likely to Additionally, the generally poor sensitivity across sports
be associated with increased injury risk in rugby, and to a suggests that FMS composite scores are not effective at
lesser extent American football and ice hockey, compared to identifying athletes who are at high injury risk. As a result,
other sports. Second, the FMS composite score of ≤ 14 was caution should be observed if the FMS is used to separate
associated with a small increase in injury risk for male but athletes into high-risk and low-risk injury risk sub-groups
not female athletes, although statistical heterogeneity was as the increases in injury risk observed show an associa-
high and sex-specific data limited. Third, FMS asymmetry tion between factors without limited predictive ability [51].
was associated with a small increase amongst senior but not Therefore, the FMS should not be considered a complete
junior athletes. Fourth, the relationship between pain dur- injury-risk estimation tool.
ing FMS testing and injury has been sparsely investigated. Pooled results from Australian football, basketball and
Fifth, FMS composite score and pain findings were more soccer were all found to have consistent, non-significant
specific than sensitive and FMS asymmetry findings were trivial associations with injury risk and thus were likely to
more sensitive than specific. Last, the majority of effect sizes indicate true null effects. It is highly likely that the FMS
that described relationships between FMS groups and injury alone is not be appropriate for injury-risk assessment in
risk were only small in magnitude and unlikely to be clini- these sports and a more specialised protocol may be more
cally meaningful. effective for determining accurate injury-risk estimation.
To the authors’ knowledge, this is the first review to One likely explanation is that the FMS composite score is
isolate sporting populations, which are the most popular based on combined scores from multiple subtests that each
population for FMS research and could benefit greatly from assess different movement qualities (i.e. range of motion,
an effective injury-risk screening tool [2]. Previous FMS strength, balance, etc.) and involve different body regions.
systematic reviews [12–14] have included studies involving Hence, one athlete might score below 14 due to poor range
military, firefighter and police cohorts, which limits the abil- of motion at the hip and another might score below 14 due
ity to generalise the findings to sporting populations. Despite to poor shoulder strength balance. The injury risk profiles
the narrower inclusion criteria for the study population, the of these athletes are likely to be different, and considering

Table 4  Pooled asymmetry and Subgroup Variable k n Sensitivity [95% CI] Specificity [95% CI]
pain sensitivity and specificity
meta-analysis results Senior Athletes Asymmetry 7 763 0.69 [0.61–0.76] 0.40 [0.29–0.53]
Junior Athletes Asymmetry 4 577 0.60 [0.44–0.73] 0.43 [0.38–0.48]
All-Cause Injury Asymmetry 8 1000 0.64 [0.53–0.74] 0.45 [0.34–0.56]
Non-Contact Injury Asymmetry 3 340 0.71 [0.62–0.78] 0.36 [0.29–0.43]
Injury Definition 1 Asymmetry 5 344 0.66 [0.50–0.79] 0.44 [0.34–0.55]
Injury Definition 2 Asymmetry 3 398 0.67 [0.53–0.79] 0.40 [0.17–0.69]
Injury Definition 3 Asymmetry 3 598 0.63 [0.49–0.76] 0.40 [0.35–0.46]
Senior Athletes Pain 2 254 0.08 [0.02–0.32] 0.94 [0.69–0.99]
Junior Athletes Pain 2 514 0.42 [0.33–0.52] 0.70 [0.57–0.80]
All Studies Pain 4 768 0.24 [0.08–0.53] 0.85 [0.60–0.96]

CI confidence interval, k number of studies, n number of participants

Functional Movement Screen and Sports Injury Risk

only their FMS composite score overlooks this. To date, a and Greig [47], where the mean FMS composite scores in
small amount of studies have compared individual subtests injured rugby athletes were significantly lower than those of
to overall injury risk [10, 33, 40, 41], finding that individual uninjured athletes. Given that elite rugby populations have
subtests may have a stronger association with prospective some of the highest time loss injury rates in sport [54, 55],
injury risk. However, it may be possible that certain subtests effective injury screening protocols would be a beneficial
have stronger associations with specific injury types, and this process for teams. FMS results could inform rugby athlete
could be considered in future research. preparation programs that aim to improve movement defi-
Running and mixed athlete cohorts were both found to cits to potentially reduce the number of injuries [31]. It is
have non-significant trivial harmful effects with high sta- unclear why results were stronger and more consistent for
tistical heterogeneity across 2–7 studies. Upon further elite rugby athletes and this should be a focus of future FMS
investigation, the mixed athlete cohorts all varied in the sex research; it has been suggested by Tee et al. [10] that poor
and age of the athletes being tested, the sports that were FMS composite scores might lead to poor technical tackling
included, and the injury definitions being used. In addition, ability and an associated increase in injury risk, irrespective
most studies used cohorts that included more than four dif- of the number of tackles performed [56]. Furthermore, we
ferent sports. Notably, the running studies used different speculate that poorer movement quality may reduce the abil-
methodological approaches that may have contributed to ity of rugby athletes to maneuverer their bodies in order to
the inconsistent results. For example, Bring et al. [28] used avoid or absorb collisions.
both male and female athletes that competed in both track Some sports (baseball, cricket, cross fit, handball and
and cross-country running and were tested by FMS certified rowing) were only represented by one study; effect sizes var-
testers, whereas Hotta et al. [33] used only male track run- ied from small beneficial (cricket), small harmful (cross fit
ners who were tested by non-certified FMS testers and used and handball) and moderate harmful (baseball and rowing)
a non-contact injury mechanism as opposed to an all-cause but only the effect for baseball was significant. These sports
mechanism. Both studies used different injury definitions are under-represented and a conclusion cannot be defini-
that could have impacted the results, although admittedly the tively reached within the scope of this review. The moderate
present meta-analysis found that injury mechanism or defini- effect size for the single rowing and baseball studies suggests
tion only had a small effect on the relationship between FMS some promise and could be cause for further research in
composite score and injury risk. Type of sport and athlete upper-body dominant sports.
sex can influence injury risk as well as the types of injuries
incurred [52, 53]. This is further supported by the varying 4.1.2 FMS Composite Score (≤ 14), Age and Injury Risk
results of different sporting cohorts across the whole meta-
analysis. Therefore, the variability in the results of the mixed An FMS composite score threshold of ≤ 14 was found to
athletes and running cohorts indicates that multiple sports have non-significant trivial harmful effects for junior athletes
should not be combined in injury-risk prediction studies with moderate statistical heterogeneity across eight studies,
because it makes it challenging to draw strong conclusions. whereas, it was found to have a significant small harmful
American football, ice hockey and rugby subgroups dem- effect size on senior athletes with high statistical heterogene-
onstrated moderate harmful associations between FMS com- ity across 21 studies. The high heterogeneity across studies
posite score threshold (≤ 14) and injury based on two studies in senior athletes is likely a result of the variety of sports
each, although results were non-significant for ice hockey contributing to this analysis, which were shown to demon-
and non-significant with high heterogeneity for American strate different relationships with injury. However, the more
football. Within the two American football studies, varying consistent null finding in junior athletes is notable because it
injury definitions were used, and Kiesel et al. [6] featured suggests the ≤ 14 FMS composite score threshold is not use-
a smaller time frame of one pre-season instead of one full ful or appropriate for estimating injury risk. Typically, junior
competitive season as used by Kiesel et al. [4]. The two ice athletes have been found to have lower mean FMS composite
hockey studies used varying injury definitions and varying scores where a larger amount of athletes score ≤ 14 [57, 58];
age groups; while both studies used junior cohorts, Dossa this increases the number of false positives identified and
et al. [30] used adolescent athletes and Avery et al. [25] reduces the usefulness of the screen for the estimation of
used pre-pubescent athletes. Rugby had significant results injury risk. The exploration of lower FMS composite score
with no statistical heterogeneity (both studies used elite male thresholds for junior athletes would be a beneficial area of
athletes tested by FMS certified testers using the same injury future FMS research. This has previously only been com-
definition), which indicates that the FMS is a suitable test for pleted by Bardenett et al. [7] using a threshold of ≤ 13; how-
injury risk identification in elite rugby players in this con- ever, this study used a mixed population of athletes, which
text, with a sensitivity of 0.60 and a specificity of 0.77. This makes interpretation of results difficult based on the known
was further supported by the study completed by Armstrong differences in results across sports.
 E. Moore et al.

4.1.3 FMS Composite Score (≤ 14), Injury Mechanism meta-analysis found small harmful effects on injury risk
and Injury Risk with ≥ 1 asymmetrical test for senior but not junior ath-
letes. It is often suggested that asymmetrical movement,
The principal focus of the FMS is identifying movement especially in simple body weight tasks such as the sub-
pattern limitations [43] that could contribute to non-contact tests of the FMS, is indicative of motor control deficits [61]
injury risk to a greater extent than contact injury risk [59]. that may subsequently increase injury risk. If correct, this
Therefore, it is plausible that non-contact injury mechanisms would likely be more apparent in senior compared to junior
may be more strongly associated with injury risk rather than athletes, due to the fact that senior athletes are exposed to
all-cause definitions. However, the present meta-analysis higher game speeds [62] and increased force impacts [63].
found that all-cause (k = 22) and non-contact (k = 7) injury This may explain the stronger relationship between FMS
mechanisms both had non-significant trivial to small harm- asymmetry and injury risk for senior compared to junior
ful effects. Due to the limited number of studies that utilised athletes in this review. Nevertheless, the effect sizes in senior
a non-contact mechanism, further research could be war- athletes were still small and unlikely to be clinically useful.
ranted to compare the injury risk between all-cause and non-
contact mechanisms to determine if one mechanism is more 4.3 FMS Pain and Injury Risk
closely related to the FMS than the other. Additionally, the
FMS was most strongly associated with injury in collision Painful movement often leads to movement compensations
sports (i.e. rugby and American football) and the reasons for as part of pain-avoidance strategies, which could increase
this should be considered in future research. the risk of injury [64]. Only two studies in junior athletes
Variations in injury definition affect the amount of inju- and two studies in senior athletes provided FMS pain infor-
ries that are captured in a study. A broad definition (i.e. med- mation. Overall, painful tests had a significant small harmful
ical presentation only) captures more injuries than a narrow effect on injury risk with no heterogeneity in junior athletes
definition that requires a game to be missed [18]. However, only. This significant finding for junior athletes may reflect
a narrow definition could be more reliable as it captures sig- the less developed motor coordination of junior athletes [65].
nificant injuries and may be less prone to subjectivity [18]. Pain is associated with alterations in neuromuscular control
Notably, the present meta-analysis suggested that injury [66] and these alterations might be less effective in junior
definition has minimal effect on injury risk when using a athletes as they do not possess the same level of control
composite threshold of ≤ 14, with all definitions reporting as senior athletes [67, 68]. Athletes who undertake neuro-
trivial to small effects with high heterogeneity. The high muscular training have been found to experience decreased
heterogeneity in the narrow definitions could be influenced subsequent injury risk [69]. Such strategies might be par-
by the time-loss specification of each study. For example, ticularly important for junior athletes that are experiencing
Azzam et al. [26] used a 7-day time loss, Chalmers et al. pain with movement.
[15] and Chalmers et al. [16] used a similar time loss of
one match (approximately 7 days between matches), Kiesel 4.4 Limitations and Future Research
et al. [4] used a 3-week time loss, and Hotta et al. [33] used
a 4-week time loss. Using a strict time loss definition greatly This review has limitations that should be considered when
reduces the amount of reported injuries; it has been found interpreting the findings. The average critical appraisal score
that between 70 and 92% of injuries might not be captured was 8/11, which would appear to indicate low overall risk of
when using the narrowest definition in comparison to the bias for studies included in this review. However, there were
broad definition of medical presentation [60]. Due to the several aspects of the risk of bias assessment that a large
relatively smaller number of studies that incorporate the number of studies failed to meet; particularly, accounting for
narrowest definition, further research into the variation of drop out, ensuring participants were not injured at the time
within-study results caused by time loss definitions (1-week of FMS testing, and using strategies to reduce the impact of
vs. 3–4 weeks) could be warranted to better identify how confounding variables. In addition to addressing these issues
the various time loss definitions impact the association of within study design, future studies should ensure thorough
FMS composite thresholds with injury risk within individual reporting of methodology as in many cases it was not possi-
studies. ble to assess methodology quality due to insufficient report-
ing. The present meta-analyses were a series of univariate
4.2 FMS Asymmetry and Injury Risk analyses; if larger study numbers were available, it could
have been possible to assess the interaction between vari-
It has been suggested that athletes with movement asym- ables using multivariate analysis. Additionally, only three
metries establish compensatory movement patterns that studies included in the meta-analysis used a female only
could increase their risk of injury [2, 3]. The present cohort compared to 22 studies using a male only cohort.
Functional Movement Screen and Sports Injury Risk

As a result, findings from this review can be more strongly Conflict of interest  Emma Moore, Joel T. Fuller, Steve Milanese and
related to male compared to female athlete populations. Samuel Chalmers declare that they have no conflict of interest.
There are areas of research found in the interpretation of Data availability statement  The datasets generated during and/or ana-
this meta-analysis that warrant further investigation. First, lysed during the current systematic review are available in the Online
determining the factors involved in rugby that have contrib- Supplementary Material.
uted to the FMS being more consistently related to injury
risk with a threshold of ≤ 14. Second, further research into
determining whether one injury mechanism (all cause vs. References
non-contact vs. contact) is more closely related to injury
risk. Third, further research involving female only study 1. Bahr R, Krosshaug T. Understanding injury mechanisms: a key
cohorts are required to address the current sex bias in the component of preventing injuries in sport. Br J Sports Med.
FMS literature. Last, whether poor scores in specific subtests 2005;39(6):324–9.
2. Cook G, Burton L, Hoogenboom B, Voight M. Pre-participation
are associated with an increased risk of injury for the body screening: the use of fundamental movements as an assessment
regions that the subtest assesses. of function-part 1. N Am J Sports Phys Ther. 2006;1(2):62–72.
3. Cook G, Burton L, Hoogenboom B, Voight M. Functional move-
ment screening: the use of fundamental movements as an assess-
ment of function-part 2. Int J Sports Phys Ther. 2014;9(4):549–63.
5 Conclusion 4. Kiesel K, Plisky P, Voight M. Can serious injury in professional
football be predicted by a preseason Functional Movement
The systematic review identified 36 prospective studies Screen? N Am J Sports Phys Ther. 2007;2(3):147–58.
investigating the relationship between the FMS and injury 5. Chorba R, Chorba D, Bouillon L, Overmyer C, Landis J. Use
of a functional movement screening tool to determine injury
risk in athletes, and 29 of these studies were suitable to be risk in female collegiate athletes. N Am J Sports Phys Ther.
included in a meta-analysis. Overall, the majority of meta- 2010;5(2):47–54.
analysis effect sizes were only small in magnitude and 6. Kiesel KB, Butler RJ, Plisky PJ. Prediction of injury by limited
unlikely to be clinically meaningful. However, some impor- and asymmetrical fundamental movement patterns in American
football players. J Sport Rehabil. 2014;23(2):88–94.
tant factors that influence the association between the FMS 7. Bardenett SM, Micca JJ, DeNoyelles JT, Miller SD, Jenk DT,
and injury risk were evident. There were consistent null find- Brooks GS. Functional Movement Screen normative values and
ings for studies involving Australian football, basketball and validity in high school athletes: can the FMS™ be used as a pre-
soccer athletes, whereas studies focussed on rugby athletes dictor of injury? Int J Sports Phys Ther. 2015;10(3):303–8.
8. Dorrel B, Long T, Shaffer S, Myer GD. The functional movement
demonstrated a consistent and moderately strong harmful screen as a predictor of injury in National Collegiate Athletic
association between an FMS composite score of ≤ 14 and Association Division II athletes. J Athl Train. 2018;53(1):29–34.
injury risk. American football and ice hockey demonstrated 9. Letafatkar A, Hadadnezhad M, Shojaedin S, Mohamadi E. Rela-
similar moderately strong associations, but they were not tionship between functional movement screening score and his-
tory of injury. Int J Sports Phys Ther. 2014;9(1):21–7.
statistically significant. Irrespective of sport type, FMS com- 10. Tee JC, Klingbiel JF, Collins R, Lambert MI, Coopoo Y. Presea-
posite score and presence of asymmetry were stronger pre- son Functional Movement Screen component tests predict severe
dictors of injury risk in senior compared to junior athletes. contact injuries in professional rugby union players. J Strength
Injury mechanism (all-cause vs. non-contact) and definition Cond Res. 2016;30(11):3194–203.
11. Wiese BW, Boone JK, Mattacola CG, McKeon PO, Uhl TL.
did not have a substantial impact on the relationship between Determination of the Functional Movement Screen to predict mus-
injury risk and an FMS composite score threshold of ≤ 14 or culoskeletal injury in intercollegiate athletics. Athl Train Sports
with asymmetrical tests. Screening specificity was greater Health Care. 2014;6(4):161–9.
than sensitivity for FMS composite score and pain findings, 12. Dorrel BS, Long T, Shaffer S, Myer GD. Evaluation of the Func-
tional Movement Screen as an injury prediction tool among active
whereas screening sensitivity was greater than specificity for adult populations: a systematic review and meta-analysis. Sports
FMS asymmetry findings. Health. 2015;7(6):532–7.
13. Bonazza NA, Smuin D, Onks CA, Silvis ML, Dhawan A. Reliabil-
Author contributions  All authors contributed to the conception and ity, validity, and injury predictive value of the Functional Move-
design of the review and completion of the search strategy. Joel T. ment Screen: a systematic review and meta-analysis. Am J Sports
Fuller was responsible for the meta-analysis. Emma Moore drafted Med. 2016;45(3):725–32.
the manuscript. All authors edited and revised the manuscript and 14. Moran RW, Schneiders AG, Mason J, Sullivan SJ. Do Functional
approved the final version of the manuscript. Movement Screen (FMS) composite scores predict subsequent
injury? A systematic review with meta-analysis. Br J Sports Med.
2017;51(23):1661–9.
Compliance with ethical standards  15. Chalmers S, Fuller JT, Debenedictis TA, Townsley S, Lynagh M,
Gleeson C, et al. Asymmetry during preseason Functional Move-
Funding  This research received no specific grant from any funding ment Screen testing is associated with injury during a junior Aus-
agency. tralian football season. J Sci Med Sport. 2017;20(7):653–7.
 E. Moore et al.

16. Chalmers S, Debenedictis TA, Zacharia A, Townsley S, Gleeson preparative period and competitive period in high school baseball
C, Lynagh M, et al. Asymmetry during Functional Movement players. J Exerc Sci Fit. 2018;16(2):68–72.
Screening and injury risk in junior football players: a replication 36. Martin C, Olivier B, Benjamin N. The Functional Movement
study. Scand J Med Sci Sports. 2018;28(3):1281–7. Screen in the prediction of injury in adolescent cricket pace bowl-
17. Moher D, Liberati A, Tetzlaff J, Altman D. Preferred reporting ers: an observational study. J Sport Rehabil. 2017;26(5):386–95.
items for systematic reviews and meta-analyses: the PRISMA 37. Mokha M, Sprague PA, Gatens DR. Predicting musculoskel-
statement. Int J Surg. 2010;8(5):336–41. etal injury in National Collegiate Athletic Association Divi-
18. Orchard J, Hoskins W. For debate: consensus injury definitions sion II athletes from asymmetries and individual-test versus
in team sports should focus on missed playing time. Clin J Sport composite Functional Movement Screen scores. J Athl Train.
Med. 2007;17(3):192–6. 2016;51(4):276–82.
19. Moola S, Munn Z, Tufanaru C, Aromataris E, Sears K, Sfetcu R, 38. Moran S, Booker H, Staines J, Williams S. Rates and risk factors
et al. Chapter 7: Systematic reviews of etiology and risk. In: Aro- of injury in CrossFit™: a prospective cohort study. J Sports Med
mataris E, Munn Z, editors. Joanna Briggs Institute Reviewer’s Phys Fit. 2017;57(9):1147–53.
Manual: the Joanna Briggs Institute. Adelaide: Joanna Briggs 39. Philp F, Blana D, Chadwick EK, Stewart C, Stapleton C, Major
Institute; 2017. K, et al. Study of the measurement and predictive validity of
20. McHugh ML. Interrater reliability: the kappa statistic. Biochem the Functional Movement Screen. BMJ Open Sport Exerc Med.
Med. 2012;22(3):276–82. 2018;4(1):e000357.
21. Hopkins WG. A scale of magnitudes for effect statistics. Sports 40. Rusling C, Edwards KL, Bhattacharya A, Reed A, Irwin S, Boles
Sci. 2002. http://www.sport​sci.org/resou​rce/stats​/effec​tmag.html. A, et al. The functional movement screening tool does not predict
Accessed 3 May 2018. injury in football. Prog Orthop Sci. 2015;1(2):41–6.
22. Hopkins WG. Statistics used in observational studies. Sports 41. Schroeder J, Wellmann K, Stein D, Braumann KM. The Func-
Injury Research. 1st ed. Oxford: Oxford University Press; 2010. tional Movement Screen for injury prediction in male amateur
23. Fritz JM, Wainner RS. Examining diagnostic tests: an evidence- football. Dtsch Z Sportmed. 2016;67(2):39–43.
based perspective. Phys Ther. 2001;81(9):1546–64. 42. Slodownik R, Ogonowska-Slodownik A, Morgulec-Adamowicz
24. Higgins JPT, Thompson SG, Deeks JJ, Altman DG.
N. Functional Movement Screen™ and history of injury in assess-
Measuring inconsistency in meta-analyses. Br Med J. ment of potential risk of injury among team handball players. J
2003;327(7414):557–60. Sports Med Phys Fit. 2018;58(9):1281–6.
25. Avery M, Wattie N, Holmes M, Dogra S. Seasonal changes in 43. Smith PD, Hanlon MP. Assessing the effectiveness of the Func-
functional fitness and neurocognitive assessments in youth ice tional Movement Screen in predicting non-contact injury rates in
hockey players. J Strength Cond Res. 2018;32(11):3143–52. soccer players. J Strength Cond Res. 2017;31(12):3327–32.
26. Azzam MG, Throckmorton TW, Smith RA, Graham D, Schol- 44. Walbright PD, Walbright N, Ojha H, Davenport T. Validity of
ler J, Azar FM. The Functional Movement Screen as a predictor functional screening tests to predict lost-time lower quarter injury
of injury in professional basketball players. Curr Orthop Pract. in a cohort of female collegiate athletes. Int J Sports Phys Ther.
2015;26(6):619–23. 2017;12(6):948–59.
27. Bond CW, Dorman JC, Odney TO, Roggenbuck SJ, Young SW, 45. Warren M, Smith CA, Chimera NJ. Association of the Functional
Munce TA. Evaluation of the Functional Movement Screen and Movement Screen with injuries in Division I athletes. J Sport
a novel basketball mobility test as an injury prediction tool for Rehabil. 2015;24(2):163–70.
collegiate basketball players. J Strength Cond Res. 2017;. https​:// 46. Zalai D, Panics G, Bobak P, Csaki I, Hamar P. Quality of
doi.org/10.1519/JSC.00000​00000​00194​4. functional movement patterns and injury examination in elite-
28. Bring BV, Chan M, Devine RC, Collins CL, Diehl J, Burkam B. level male professional football players. Acta Physiol Hung.
Functional Movement Screening and injury rates in high school 2015;102(1):34–42.
and collegiate runners: a retrospective analysis of 3 prospective 47. Armstrong R, Greig M. Injury identification: the efficacy of the
observational studies. Clin J Sport Med. 2017;21:21. Functional Movement Screen in female and male rugby union
29. Clay H, Mansell J, Tierney R. Association between rowing inju- players. Int J Sports Phys Ther. 2018;13(4):605–17.
ries and the Functional Movement Screen™ in female collegiate 48. Hammes D, Aus der Funten K, Bizzini M, Meyer T. Injury predic-
Division I rowers. Int J Sports Phys Ther. 2016;11(3):345–9. tion in veteran football players using the Functional Movement
30. Dossa K, Cashman G, Howitt S, West B, Murray N. Can injury in Screen™. J Sports Sci. 2016;34(14):1371–9.
major junior hockey players be predicted by a pre-season Func- 49. Svensson K, Alricsson M, Olausson M, Werner S. Physi-
tional Movement Screen—a prospective cohort study. J Can Chi- cal performance tests–a relationship of risk factors for muscle
ropractic Assoc. 2014;58(4):421–7. injuries in elite level male football players. J Exerc Rehabil.
31. Duke SR, Martin SE, Gaul CA. Preseason Functional Movement 2018;14(2):282–8.
Screen predicts risk of time-loss injury in experienced male rugby 50. Newton F, McCall A, Ryan D, Blackburne C, aus der Fünten K,
union athletes. J Strength Cond Res. 2017;31(10):2740–7. Meyer T et al. Functional Movement Screen (FMS™) score does
32. Garrison M, Westrick R, Johnson MR, Benenson J. Association not predict injury in English Premier League youth academy foot-
between the Functional Movement Screen and injury development ball players. Sci Med Football. 2017;1(2):102–6.
in college athletes. Int J Sports Phys Ther. 2015;10(1):21–8. 51. McCall A, Fanchini M, Coutts AJ. Prediction: the modern-day
33. Hotta T, Nishiguchi S, Fukutani N, Tashiro Y, Adachi D, Morino sport-science and sports-medicine “quest for the holy grail”. Int J
S, et al. Functional Movement Screen for predicting running inju- Sports Physiol Perf. 2017;12(5):704–6.
ries in 18-to 24-year-old competitive male runners. J Strength 52. Ristolainen L, Heinonen A, Waller B, Kujala UM, Kettunen JA.
Cond Res. 2015;29(10):2808–15. Gender differences in sport injury risk and types of injuries: a
34. Kolodziej M, Jaitner T. Single Functional Movement Screen items retrospective twelve-month study on cross-country skiers, swim-
as main predictors of injury risk in amateur male soccer players. mers, long-distance runners and soccer players. J Sports Sci Med.
Ger J Exerc Sport Res. 2018;48(3):349–57. 2009;8(3):443–51.
35. Lee C-L, Hsu M-C, Chang W-D, Wang S-C, Chen C-Y, Chou 53. Ristolainen L, Heinonen A, Turunen H, Mannström H, Waller
P-H, et al. Functional Movement Screen comparison between the B, Kettunen JA et al. Type of sport is related to injury profile: A
study on cross country skiers, swimmers, long‐distance runners
Functional Movement Screen and Sports Injury Risk

and soccer players. A retrospective 12‐month study. Scand J Med 62. Burgess D, Naughton G, Norton K. Quantifying the gap between
Sci Sports. 2010;20(3):384–93. under 18 and senior AFL football: 2003 and 2009. Int J Sports
54. Brooks JH, Fuller C, Kemp S, Reddin DB. Epidemiology of inju- Physiol Perf. 2012;7(1):53–8.
ries in English professional rugby union: part 1 match injuries. Br 63. Romiti M, Finch CF, Gabbe B. A prospective cohort study of
J Sports Med. 2005;39(10):757–66. the incidence of injuries among junior Australian football play-
55. Chalmers DJ, Samaranayaka A, Gulliver P, McNoe B. Risk factors ers: evidence for an effect of playing-age level. Br J Sports Med.
for injury in rugby union football in New Zealand: a cohort study. 2008;42(6):441–6.
Br J Sports Med. 2012;46(2):95–102. 64. O’Sullivan P. Diagnosis and classification of chronic low back
56. Gabbett TJ, Jenkins DG, Abernethy B. Physical collisions and pain disorders: maladaptive movement and motor control impair-
injury in professional rugby league match-play. J Sci Med Sport. ments as underlying mechanism. Man Ther. 2005;10(4):242–55.
2011;14(3):210–5. 65. Balyi I, Hamilton A. Long-term athlete development: trainability
57. Fuller JT, Chalmers S, Debenedictis TA, Townsley S, Lynagh M, in childhood and adolescence. Victoria: National Coaching Insti-
Gleeson C, et al. High prevalence of dysfunctional, asymmetrical, tute British Columbia & Advanced Training and Performance Ltd;
and painful movement in elite junior Australian Football players 2004.
assessed using the Functional Movement Screen. J Sci Med Sport. 66. Bonacci J, Chapman A, Blanch P, Vicenzino B. Neuromuscular
2017;20(2):134–8. adaptations to training, injury and passive interventions. Sports
58. Portas MD, Parkin G, Roberts J, Batterham AM. Maturational Med. 2009;39(11):903–21.
effect on Functional Movement Screen™ score in adolescent soc- 67. Ozmun JC, Mikesky AE, Surburg PR. Neuromuscular adaptations
cer players. J Sci Med Sport. 2016;19(10):854–8. following prepubescent strength training. Med Sci Sports Exerc.
59. Shimokochi Y, Shultz SJ. Mechanisms of noncontact anterior 1994;26(4):510–4.
cruciate ligament injury. J Athl Train. 2008;43(4):396–408. 68. Ramsay JA, Blimkie C, Smith K, Garner S, Macdougall JD, Sale
60. Hodgson L, Gissane C, Gabbett TJ, King DA. For debate: consen- DG. Strength training effects in prepubescent boys. Med Sci
sus injury definitions in team sports should focus on encompass- Sports Exerc. 1990;22(5):605–14.
ing all injuries. Clin J Sport Med. 2007;17(3):188–91. 69. Myer GD, Ford KR, Brent JL, Hewett TE. Differential neuro-
61. Cholewicki J, Silfies SP, Shah RA, Greene HS, Reeves NP, Alvi muscular training effects on ACL injury risk factors in” high-
K, et al. Delayed trunk muscle reflex responses increase the risk risk” versus” low-risk” athletes. BMC Musculoskelet Disord.
of low back injuries. Spine. 2005;30(23):2614–20. 2007;8(1):39–45.

Affiliations

Emma Moore1   · Samuel Chalmers2,3   · Steve Milanese1   · Joel T. Fuller4 

1 3
International Centre for Allied Health Evidence (iCAHE), Sport and Exercise Science, School of Science and Health,
University of South Australia, GPO Box 2471, Adelaide, Western Sydney University, Sydney, NSW, Australia
SA 5000, Australia 4
Faculty of Medicine and Health Sciences, Macquarie
2
Exercise and Sport Science, Faculty of Health Sciences, The University, Sydney, NSW, Australia
University of Sydney, Sydney, NSW, Australia