The American Journal of Sports

Medicine http://ajs.sagepub.com/

The Functional Movement Screen and Injury Risk: Association and Predictive Value in Active Men
Timothy T. Bushman, Tyson L. Grier, Michelle Canham-Chervak, Morgan K. Anderson, William J. North and Bruce H.
Jones
Am J Sports Med published online December 10, 2015
DOI: 10.1177/0363546515614815

The online version of this article can be found at:

http://ajs.sagepub.com/content/early/2015/12/10/0363546515614815

Published by:

http://www.sagepublications.com

On behalf of:
American Orthopaedic Society for Sports Medicine

Additional services and information for The American Journal of Sports Medicine can be found at:
Published online December 10, 2015 in advance of the print journal.

P<P
Email Alerts: http://ajs.sagepub.com/cgi/alerts

Subscriptions: http://ajs.sagepub.com/subscriptions

Reprints: http://www.sagepub.com/journalsReprints.nav

Permissions: http://www.sagepub.com/journalsPermissions.nav

>> OnlineFirst Version of Record - Dec 10, 2015

What is This?

Downloaded from ajs.sagepub.com at TRENT UNIV on December 17, 2015

 AJSM PreView, published on December 10, 2015 as doi:10.1177/0363546515614815

The Functional Movement Screen

and Injury Risk
Association and Predictive Value in Active Men
Timothy T. Bushman,*y MS, Tyson L. Grier,y MS, Michelle Canham-Chervak,y PhD,
Morgan K. Anderson,y MPH, William J. North,z MS, and Bruce H. Jones,y MD
Investigation performed at the Army Institute of Public Health,
Aberdeen Proving Ground, Maryland, USA

Background: The Functional Movement Screen (FMS) is a series of 7 tests used to assess the injury risk in active populations.
Purpose: To determine the association of the FMS with the injury risk, assess predictive values, and identify optimal cut points
using 3 injury types.
Study Design: Cohort study; Level of evidence, 2.
Methods: Physically active male soldiers aged 18 to 57 years (N = 2476) completed the FMS. Demographic and fitness data were
collected by survey. Medical record data for overuse injuries, traumatic injuries, and any injury 6 months after the FMS assess-
ment were obtained. Sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) were calculated
along with the receiver operating characteristic (ROC) to determine the area under the curve (AUC) and identify optimal cut points
for the risk assessment. Risks, risk ratios (RRs), odds ratios (ORs), and 95% CIs were calculated to assess injury risks.
Results: Soldiers who scored 14 were at a greater risk for injuries compared with those who scored .14 using the composite score
for overuse injuries (RR, 1.84; 95% CI, 1.63-2.09), traumatic injuries (RR, 1.26; 95% CI, 1.03-1.54), and any injury (RR, 1.60; 95% CI,
1.45-1.77). When controlling for other known injury risk factors, multivariate logistic regression analysis identified poor FMS perfor-
mance (OR [score 14/19-21], 2.00; 95% CI, 1.42-2.81) as an independent risk factor for injuries. A cut point of 14 registered
low measures of predictive value for all 3 injury types (sensitivity, 28%-37%; PPV, 19%-52%; AUC, 54%-61%). Shifting the injury
risk cut point of 14 to the optimal cut points indicated by the ROC did not appreciably improve sensitivity or the PPV.
Conclusion: Although poor FMS performance was associated with a higher risk of injuries, it displayed low sensitivity, PPV, and
AUC. On the basis of these findings, the use of the FMS to screen for the injury risk is not recommended in this population
because of the low predictive value and misclassification of the injury risk.
Keywords: military training; epidemiology; injury prevention; Functional Movement Screen; predictive value

Many physically active populations, both civilian and mili-
tary, would benefit from enhanced injury screening methods
to assess the risk of injury. Such populations include (but
are not limited to) the military, athletes, firefighters, and
manual laborers who may use the identification of injury
risk and targeted interventions to prevent injuries, lower
health costs, and reduce time loss due to injuries. For the
military, injuries present a large problem that affects the
health and physical readiness of soldiers during peacetime
and combat operations. In 2012, injuries were the leading
cause of all medical encounters in the United States (US)
Army and resulted in over 1.3 million visits that affected
over 300,000 soldiers (E. Dada and M. Chervak, personal
communication, 2013). Although injuries present a large
The American Journal of Sports Medicine, Vol. XX, No. X
DOI: 10.1177/0363546515614815
Ó 2015 The Author(s)
burden to the US Army and other active populations, they
also create an opportunity for injury screening tools to iden-
tify those at risk for injuries and promote primary preven-
tion. One popular injury screening tool that has been used
to assess the injury risk in a variety of physically active pop-
ulations is the Functional Movement Screen (FMS).§
The FMS is a screening tool intended to identify persons
at risk of activity-related injuries in sports and exercise.6
The FMS consists of 7 movement patterns involving locomo-
tor, manipulative, and stabilizing actions that assess balance,
mobility, and stability.5 Several previous investigations
have touted the effectiveness of using the FMS as an injury
screening tool in nonmilitary populations, associating poor
performance (a composite score of 14 of a possible 21)
with a higher injury risk.3,4,18,19,22,42 An investigation con-
ducted using a military cohort found an association between
§
References 2-4, 18, 19, 22, 24, 28, 35, 40, 42, 43.

1
Downloaded from ajs.sagepub.com at TRENT UNIV on December 17, 2015
2 Bushman et al The American Journal of Sports Medicine

poor FMS performance and an increased risk of injuries.24,28
Some prior studies did not find any association between poor
FMS performance and the risk of injuries.35,40 However,
most of the evidence suggests that there is an association
between low composite scores on the FMS and a greater
risk for musculoskeletal injuries.
In addition to an association with risk, a good screening
tool must also be reliable and valid. The reliability of the
FMS appears to be established.k The validity of the FMS,
however, remains more equivocal.4,11,17,19,21,23,28,41 Krumrei
et al21 conducted a review focused on whether the FMS can
accurately identify highly active patients with an elevated
risk of injuries. Their investigation yielded 3 prospective
cohort investigations that provided moderate-quality evi-
dence that the FMS can accurately identify those with an ele-
vated risk for musculoskeletal injuries among male
professional football players; male Marine officer candidates;
and female collegiate basketball, soccer, and volleyball play-
ers.4,19,21,28 However, previous studies have noted low inter-
nal consistency using the FMS composite score17,23 and low
criterion validity.41 These mixed findings call into question
whether the FMS exhibits adequate validity to be recommen-
ded as a screening tool for the prediction of injuries.
The purpose of this investigation was to (1) evaluate the
association of the FMS with the injury risk in a population
of young physically active men engaged in vigorous physi-
cal training, (2) determine the predictive value of the FMS
as a screening tool for musculoskeletal injuries, and (3)
identify optimal cut points for risk screening using 3 differ-
ent injury types. These analyses are important because
screening tools must not only be associated with the injury
risk but must also accurately identify persons who are
actually at a higher risk for injuries. The current investiga-
tion provides greater insight into these factors from
a robust assessment of injury association and predictive
values using the FMS in a large, active male population.
METHODS
Data Collection
This project was reviewed and approved by an institutional
review board. The FMS assessment was conducted in
September 2011 with soldiers in a light infantry brigade
combat team consisting of 7 battalions: armor, infantry,
cavalry, field artillery, brigade support, brigade special
troops, and headquarters. The brigade physical therapist
k
References 9, 12, 14, 26, 29, 30, 33, 34, 36, 39.
utilized the FMS manual to train approximately 30 sol-
diers over 1 week to facilitate and oversee each FMS
assessment. This group of approximately 30 soldiers was
given a general overview of each FMS test, including
detailed instruction and a video of what constituted a score
of 3, 2, 1, and 0. Soldiers were then given specific instruc-
tions on 1 or 2 stations that they were assigned to for the
test. Soldiers practiced their scoring at these assigned sta-
tions to ensure grading consistency. Discrepancies in grad-
ing were addressed by retraining at that particular testing
station until consistency was reached. These soldiers were
limited to collecting data only at those stations in which
they demonstrated grading competence during the grading
evaluation. All assessments were conducted indoors, and
the full assessment took approximately 15 to 20 minutes
per soldier. Soldiers with limited or permanent medical
restrictions met with brigade physical therapy staff to
determine eligibility for FMS participation. Only those sol-
diers who did not have a medical restriction or were cleared
by brigade physical therapy staff and completed the entire
FMS assessment were included in the analysis. The FMS
involved deep squats, hurdle steps, in-line lunges, shoulder
mobility, active straight-leg raises, trunk stability push-ups,
and rotatory stability. Each movement pattern was scored
on a range of 0 to 3, with a maximum possible composite score
of 21. If pain was present on a movement pattern, the partic-
ipant automatically scored 0 for that particular test. A score
of 1 was given if the participant could not complete the proper
movement but was pain free, 2 points were given if the move-
ment was performed with some difficulty but was pain free,
and 3 points were given if the movement was performed as
instructed with no difficulty or compensation and was pain
free. There were also 3 clearing tests, not scored numerically,
performed to examine for the presence of pain: the shoulder
mobility clearing test, spinal extension clearing test, and spi-
nal flexion clearing test. If pain was experienced on any of
these clearing tests, then the scores from shoulder mobility,
trunk stability push-ups, and rotatory stability were changed
to 0. A more detailed description of all 7 FMS movement pat-
terns can be found in articles by Cook et al.5-8
During the FMS assessment, surveys were adminis-
tered to collect information on personal characteristics
(ie, sex, age, date of birth, rank, race, education level,
and marital status) and fitness data from their most recent
Army Physical Fitness Test (maximum push-ups in 2
minutes, maximum sit-ups in 2 minutes, and a 2-mile
run for time; 98% of the soldiers completed this fitness
test within 1 year of the FMS assessment). Height and
weight were measured before the FMS assessment. Not
all soldiers had complete survey data.

*Address correspondence to Timothy T. Bushman, MS, Directorate of Epidemiology and Disease Surveillance, Army Institute of Public Health, 5158
Blackhawk Road, Aberdeen Proving Ground, MD 21010, USA (email: Timothy.t.bushman.ctr@mail.mil; ttbushman@gmail.com).
y
Directorate of Epidemiology and Disease Surveillance, Army Institute of Public Health, Aberdeen Proving Ground, Maryland, USA.
z
Henry Jackson Foundation, Fort Carson, Colorado, USA.
The use of a trade name does not imply endorsement by the US Army but is intended only to assist in the identification of a specific product. The views
expressed in this article are those of the author(s) and do not necessarily reflect the official policy of the Department of Defense, Department of the Army,
US Army Medical Department, or US Government.
One or more of the authors has declared the following potential conflict of interest or source of funding: This investigation was supported in part by an
appointment to the Postgraduate Research Participation Program at the US Army Institute of Public Health (USAIPH), administered by the Oak Ridge Insti-
tute for Science and Education through an interagency agreement between the US Department of Energy and USAIPH.

Downloaded from ajs.sagepub.com at TRENT UNIV on December 17, 2015

AJSM Vol. XX, No. X, XXXX FMS and Injury Risk: Association and Predictive Value 3

Injury data were obtained from the Defense Medical TABLE 1

Surveillance System to capture injuries occurring 6 Characteristics of the Soldiersa
months after the FMS assessment. Injury data included
all inpatient and outpatient medical encounters for over- Variable n (%)
use injuries, traumatic injuries, or any injury found in sol- Age, y
diers’ electronic medical records with an International 23 853 (34)
Classification of Disease, 9th Revision, Clinical Modifica- 24-26 581 (23)
tion (ICD-9 CM) diagnosis code determined to be an injury. 27-31 574 (23)
Overuse injury–related musculoskeletal conditions, such 32 468 (19)
as stress fractures, Achilles tendinitis, or knee pain syn- Race
dromes, were identified from the ICD-9 CM code series White 1727 (70)
710 to 739. Traumatic injuries, such as acute sprains and African American 297 (12)
Hispanic 306 (12)
strains, fractures, and dislocations, were identified in the
Native American 16 (1)
ICD-9 CM code series 800 to 999.13 Any injury constituted Asian 101 (4)
the combination of both overuse and traumatic injuries. Other 16 (1)
Education level
No high school 21 (1)
Data Analyses High school 1922 (77)
Some college 193 (8)
Descriptive statistics, frequencies, and percentage distribu- Bachelor’s degree 246 (10)
tions were calculated for personal characteristic and injury Master’s degree 39 (2)
information. Age was analyzed in quartiles (23, 24-26, 27- Unknown 55 (2)
Marital status
31, and 32 years). Body mass index (BMI) was calculated
Married 1455 (59)
as kg/m2. FMS performance was analyzed in 2 different Single 930 (38)
ways according to the injury risk cut point of 14 first Other 90 (4)
described by Kiesel et al,19 using a score of 14 compared Body mass index, kg/m2
with .14 and also the scale of 14, 15-16, 17-18, and 19- \18.5 (underweight) 9 (\1)
21. Risks, risk ratios (RRs), and 95% CIs were calculated to 18.5-24.9 (normal) 866 (36)
assess FMS performance with the risk of overuse injuries, 25.0-29.9 (overweight) 1236 (51)
traumatic injuries, and any injury. A 1-way analysis of vari- 30.0 (obese) 306 (13)
ance (ANOVA) was performed to analyze differences in run- Smoking status
ning mileage per week between battalions. Odds ratios (ORs) Nonsmoker 1282 (58)
Smoker 928 (42)
and 95% CIs were calculated with logistic regression using
US Army rank
a forced model to assess the ‘‘any injury’’ risk. Variables E1-E3 412 (19)
included in the model yielded P \ .10 in the univariate anal- E4-E6 1451 (66)
ysis: age, BMI, smoking, muscular and cardiorespiratory fit- E7-E9 98 (4)
ness (maximum push-ups in 2 minutes and a 2-mile run for W1-O6 244 (11)
time), FMS composite score, and battalion. The confounding Battalion
potential of these variables to influence the injury risk has Armor 502 (22)
been established in prior studies.10,13,15,32,37 All of the varia- Infantry 474 (21)
bles, with the exception of BMI, were entered into the model Cavalry 336 (15)
as categorical variables. Field artillery 302 (14)
Brigade support 233 (10)
The sensitivity, specificity, positive predictive value
Brigade special troops 365 (16)
(PPV), and negative predictive value (NPV) for the FMS Headquarters 25 (1)
were calculated.27 Receiver operating characteristic (ROC)
curves were determined including the area under the curve a
Not all soldiers had complete survey data.
(AUC) to test the overall accuracy of the FMS to screen for
the injury risk.1 The ROC curve was also used to determine
an optimal cut point for identifying high- and low-risk per- (77%), married (59%), overweight or obese (64%), non-
sons for each injury type. The formula for measuring the smokers (58%), and lower than E7 rank (85%).
point on the ROC curve that maximized the sensitivity The injury incidence was 28% for overuse injuries
and specificity of the screen is described in the literature.25 (mean [6SD] FMS composite score, 15.0 6 3.1; range, 5-
21), 16% for traumatic injuries (mean FMS composite
score, 15.6 6 2.8; range, 6-21), and 37% for any injury
RESULTS (mean FMS composite score, 15.2 6 3.0; range, 5-21). Sol-
diers who were not injured had a mean FMS composite
A total of 2476 male soldiers completed the FMS assess- score of 16.3 6 2.3 (range, 5-21). Some soldiers experienced
ment. Table 1 displays soldier characteristics at the time both an overuse injury and traumatic injury (which were
of the FMS assessment. A majority of the soldiers were counted for both overuse and traumatic injury types, but
26 years of age (57%), white (70%), high school graduates only once for the any injury type) during the injury period.

Downloaded from ajs.sagepub.com at TRENT UNIV on December 17, 2015

4 Bushman et al The American Journal of Sports Medicine

TABLE 2
FMS Composite Score and 6-Month Injury Riska

Overuse Injury Traumatic Injury Any Injury

FMS Composite n (% of Risk Ratio Risk Ratio Risk Ratio
Score % Total) (95% CI) P Value % n (95% CI) P Value % n (95% CI) P Value

14 43 591 (24) 1.86 (1.50-2.31) \.01 19 591 1.49 (1.07-2.07) .02 52 591 1.73 (1.45-2.07) \.01
15-16 26 718 (29) 1.12 (0.89-1.41) .32 15 718 1.21 (0.87-1.69) .25 35 718 1.17 (0.97-1.42) .10
17-18 21 831 (33) 0.92 (0.73-1.16) .47 15 831 1.22 (0.88-1.69) .22 31 831 1.03 (0.85-1.25) .74
19-21 23 336 (14) 1 13 336 1 30 336 1
14 43 591 (24) 1.84 (1.63-2.09) \.01 19 592 1.26 (1.03-1.54) .03 52 592 1.60 (1.45-1.77) \.01
.14 23 1885 (76) 1 15 1885 1 33 1885 1

a
Bolded values indicate statistical significance (P \ .05). Percentage columns indicate the percentage of soldiers injured within that cat-
egory. FMS, Functional Movement Screen.

Table 2 displays the association of FMS performance with TABLE 3

the injury risk. Soldiers who scored 14 were 1.86, 1.49, and Multivariate Logistic Regression Using a
1.73 times more likely to suffer an injury as those who scored Forced Model: Risk Factors for Any Injurya
19-21 for overuse injuries, traumatic injuries, and any injury,
Variable n Odds Ratio (95% CI) P Value
respectively. Examining participants who scored lower and
higher than the standard cut point (scores of 14 and .14), Age range, y
those who scored 14 were 1.84, 1.26, and 1.60 times more 23 652 1.00
likely to experience an injury compared with those who 24-26 478 0.89 (0.69-1.15) .40
scored .14 for overuse injuries, traumatic injuries, and any 27-31 425 0.91 (0.69-1.18) .49
injury, respectively. The relative risks of injuries for soldiers 32 331 1.01 (0.75-1.36) .91
scoring in the ranges of 15-16 and 17-18 were not much differ- Body mass index 1886 0.98 (0.95-1.02) .51
ent than those scoring in the highest level of 19-21 for overuse Smoking status
Nonsmoker 1110 1.00
injuries, traumatic injuries, or any injury.
Smoker 776 1.26 (1.03-1.53) .02
Table 3 displays the results of multivariate logistic Push-ups, No. of repetitions
regression analysis using a forced model with the following 55 435 0.83 (0.62-1.12) .24
variables: age, BMI, smoking status, muscular and cardiore- 56-66 486 0.95 (0.72-1.26) .76
spiratory endurance (maximum push-ups in 2 minutes and 67-75 482 0.82 (0.62-1.08) .17
a 2-mile run for time), and FMS composite score. Maximum 76 483 1.00
push-ups in 2 minutes (RR, 1.17; P = .04) were used instead 2-mile run time, min.s
of maximum sit-ups in 2 minutes (RR, 1.12; P = .14) for mus- 14.00 491 1.00
cular endurance because they were significant in the uni- 14.01-15.00 520 1.23 (0.93-1.62) .13
variate analysis for any injury and because a bivariate 15.01-15.59 420 1.36 (1.01-1.84) .03
16.00 455 1.66 (1.20-2.29) \.01
correlation revealed significance (P \ .01). Previous injuries
FMS composite score
were not entered into the model because they were not sig- 14 391 2.00 (1.42-2.81) \.01
nificant in the univariate analysis (P = .56). The FMS com- 15-16 551 1.05 (0.76-1.46) .72
posite scoring scale of 14, 15-16, 17-18, and 19-21 was used 17-18 675 0.99 (0.73-1.35) .97
in the model because it yielded a higher relative risk for any 19-21 269 1.00
injury than did the 14/.14 comparison. Soldiers who were
a
current cigarette smokers had a 1.26 times greater risk for Bolded values indicate statistical significance (P \ .05). FMS,
injuries. Soldiers with the slowest run times (16.00 Functional Movement Screen.
minutes) had a 1.66 times greater risk for injuries compared
with those who ran the fastest (14.00 minutes). Soldiers Table 4 displays the sensitivity, specificity, PPV, NPV,
who recorded poor performance on the FMS (ie, composite and AUC for the FMS cut point of 14 and the respective
score of 14) had twice the risk of injuries (2.00) compared optimal cut points derived from the ROC curve. Using the
with those who scored at the highest level of 19-21. cut point of 14, the sensitivity of the FMS was 37%, 28%,
A separate forced model that controlled for battalion and 33%; specificity was 81%, 77%, and 82%; PPV was
was also run and had a minimal effect on the significant 43%, 19%, and 52%; NPV was 77%, 85%, and 68%; and
findings: current smoker (OR, 1.27; P = .01), 2-mile run AUC was 61%, 54%, and 60% for overuse injuries, trau-
time of 16.00 minutes (OR, 1.64; P \ .01), and FMS com- matic injuries, and any injury, respectively.
posite score of 14 (OR, 1.96; P \ .01). No significant The cut points determined to be optimal for injury screen-
between-group differences for battalion were found using ing by ROC analysis were FMS scores of 15, 16, and 15
a 1-way ANOVA to assess the total unit and personal for overuse injuries, traumatic injuries, and any injury,
weekly running mileage (P = .08). respectively. Using the optimal cut point of 15 for overuse

Downloaded from ajs.sagepub.com at TRENT UNIV on December 17, 2015

AJSM Vol. XX, No. X, XXXX FMS and Injury Risk: Association and Predictive Value 5

TABLE 4
Predictive Value of the FMS Using 14 and Optimal Cut Pointsa

FMS Score Overuse Injury No Injury Total Predictive Value

14 256 335 591 Sensitivity: 37%

.14 442 1443 1885 Specificity: 81%
Total 698 1778 2476 PPV: 43%
NPV: 77%
AUC: 61%
15 341 544 885 Sensitivity: 49%
.15 357 1234 1591 Specificity: 69%
Total 698 1778 2476 PPV: 38%
NPV: 78%
AUC: 61%

FMS Score Traumatic Injury No Injury Total Predictive Value

14 110 481 591 Sensitivity: 28%

.14 278 1607 1885 Specificity: 77%
Total 388 2088 2476 PPV: 19%
NPV: 85%
AUC: 54%
16 219 1090 1309 Sensitivity: 56%
.16 169 998 1167 Specificity: 48%
Total 388 2088 2476 PPV: 17%
NPV: 86%
AUC: 54%

FMS Score Any Injury No Injury Total Predictive Value

14 308 283 591 Sensitivity: 33%

.14 612 1273 1885 Specificity: 82%
Total 920 1556 2476 PPV: 52%
NPV: 68%
AUC: 60%
15 421 464 885 Sensitivity: 46%
.15 499 1092 1591 Specificity: 70%
Total 920 1556 2476 PPV: 48%
NPV: 69%
AUC: 60%

a
Values are reported as number of participants unless otherwise indicated. AUC, area under the curve; FMS, Functional Movement
Screen; NPV, negative predictive value; PPV, positive predictive value.

injuries, sensitivity increased from 37% to 49%, but specificity
decreased from 81% to 69%, the PPV decreased from 43% to
38%, and there was a marginal increase in the NPV from 77%
to 78%. The AUC for overuse injuries was 61%. Using the
optimal cut point of 16 for traumatic injuries, sensitivity
increased from 28% to 56%, but specificity decreased from
77% to 48%, the PPV decreased slightly from 19% to 17%,
and the NPV increased slightly from 85% to 86%. The AUC
for traumatic injuries was 54%. Using the optimal cut point
of 15 for any injury, sensitivity increased from 33% to
46%, but specificity decreased from 82% to 70%, the PPV
decreased from 52% to 48%, and the NPV saw a marginal
increase from 68% to 69%. The AUC for any injury was 60%.
DISCUSSION
This investigation identified low sensitivity, PPV, and
AUC using the FMS as an injury screening tool, despite
a significant association between a low FMS composite
score (14) and a higher injury risk (irrespective of injury
type). When looking at injury types, a low FMS composite
score was more strongly associated with overuse injuries
than with traumatic injuries. Using the cut point of 14
did not maximize the sensitivity and specificity of the
screen for any of the injury types. However, the optimal
cut points identified in this investigation had a negligible
effect on improving the predictive value. When controlling
for known risk factors for injuries, multivariate logistic
regression analysis identified poor FMS performance (com-
posite score of 14) as a risk factor for any injury.
The major finding in this investigation was the low sen-
sitivity, PPV, and AUC determined using the FMS as an
injury screening tool (employing a standard cut point of
14 points). The sensitivity for overuse injuries (37%),
traumatic injuries (28%), and any injury (33%) were all
low. The sensitivity from our investigation using overuse
injuries is higher than that reported by O’Connor et al28

Downloaded from ajs.sagepub.com at TRENT UNIV on December 17, 2015

6 Bushman et al
(13%) and Hoover et al16 (8%). The sensitivity from our
investigation using traumatic injuries is between that
reported by Kiesel et al19 (54%) and O’Connor et al28
(12%). The sensitivity from our investigation using any
injury is lower than that reported by Butler et al3 (83%),
Chorba et al4 (58%), Wieczorkowski42 (50%), and O’Connor
et al28 (45%). Only 28% to 37% of soldiers in our investiga-
tion were correctly classified as high risk (composite score
14 points) out of all those who experienced an injury,
thereby missing 63% to 72% of people who were injured
but scored above the injury risk cut point. It should be
noted that these previous studies had smaller sample sizes
(range, 38-874 participants).
The PPV was also low for overuse injuries (43%), trau-
matic injuries (19%), and any injury (52%). Data showed
that only 19% to 52% of soldiers in our investigation who
experienced an injury were correctly labeled as high risk
out of all those who scored 14, thereby misclassifying
48% to 81% of soldiers identified as at risk. The PPV
from our investigation using any injury was lower than
that reported by Butler et al3 (85%). The AUC from our
investigation yielded 61%, 54%, and 60% for overuse inju-
ries, traumatic injuries, and any injury, respectively. The
investigation by O’Connor et al28 reported a slightly lower
AUC of 58%, 52%, and 53% for overuse injuries, traumatic
injuries, and any injury, respectively. Given that 50% indi-
cates that the test offers no advantage over chance, this
result suggests that the FMS had poor accuracy as
a screening tool for all 3 injury indices and was only
slightly better than the ‘‘line of equality’’ or the ‘‘chance’’
line.38 The FMS missed the majority of participants who
actually experienced an injury in our investigation, regard-
less of the injury type. Furthermore, the PPV indicated
that over 40% of those identified as at high risk did not
incur an injury. From a resource perspective, this would
be unacceptable to subject such a high number of soldiers
with false-positive results to remedial training consisting
of corrective exercises in an attempt to prevent injuries
by remedying movement dysfunction.
The use of an optimal cut point in a statistical analysis
depends on the test and disease under study. Analysis
from the ROC curve in this investigation indicated that an
FMS score of 14 was not the optimal cut point, regardless
of the injury type. The ROC curve identified scores of 15,
16, and 15 as the optimal cut points for injury risk
screening for overuse injuries, traumatic injuries, and any
injury, respectively. The optimal cut points had a minimal
effect on the predictive value of the screen for each injury
type. With the upward shift in cut points, there was an
increase in sensitivity (more soldiers with true-positive
results) with a similar decrease in specificity (less soldiers
with true-negative results) and little to no effect on the
PPV and NPV. Previous literature has provided very mixed
results on the value of optimal cut points for the FMS: no
optimal cut point28,40 and FMS scores of 14,3,19,35
14.5,42 16.5,2 and 17.16,22
A low FMS composite score of 14 carried a greater rel-
ative risk of overuse injuries, traumatic injuries, and any
injury compared with a score of .14. Results from our
investigation suggest that the FMS may not be as strongly
Downloaded from ajs.sagepub.com at TRENT UNIV on December 17, 2015
The American Journal of Sports Medicine
associated with traumatic injuries as with overuse injuries
(RR, 1.84 for overuse injuries compared with 1.26 for trau-
matic injuries using 14/.14). The sensitivity (28%) and
PPV (19%) for traumatic injuries were also lower than
the sensitivity (37%) and PPV (43%) found for overuse inju-
ries. The difference in the strength of association may be
caused by the different nature of the 2 injury types; trau-
matic injuries occur in an acute and often more ‘‘acciden-
tal’’ manner and therefore may be less predictable than
overuse injuries in particular. This finding, however, dif-
fers from that of O’Connor et al28 and Lisman et al,24
who found that the overuse injury risk was less strongly
associated with poor FMS performance than were serious
and traumatic injuries. This may be in large part because
of the difference in the definition of a ‘‘serious’’ injury ver-
sus traumatic injury. Also of interest is the finding that rel-
ative risks were similar between those who scored 19-21,
17-18, and 15-16 for each injury type. This may suggest
that the quality of movement is not the primary indicator
of the injury risk and that perhaps other factors such as
pain occurrence are more strongly associated with the
risk than the actual composite score.
Multivariate logistic regression analysis (controlling for
age, BMI, cigarette smoking, and muscular and cardiores-
piratory endurance) showed that a low FMS composite
score of 14 was a significant independent risk factor for
any injury. Our OR of 2.00 for poor FMS performance is
much lower than the OR of 11.67 reported in professional
football players (n = 46),19 8.31 in firefighters (n = 108),3
5.60 in high school basketball athletes (n = 82),42 and
3.85 in National Collegiate Athletic Association (NCAA)
Division II female collegiate athletes (n = 38)4 for scoring
14 on the FMS. This may be because of a number of fac-
tors including sample size, sex, physical training methodol-
ogy, type of activity, injury definition, previous injuries,
and the type of injury under focus. For example, physical
training methodology and type of activity performed may
lead to different exposure considerations. The exposure
period may involve structured to autonomous physical
activity, no contact to high contact activity, or a uniform
volume of activity compared with a nonuniform volume of
activity. Our OR of 2.00 is much more closely aligned
with that of O’Connor et al28 and Lisman et al,24 who found
ORs of 2.00 and 2.04, respectively, for any injury in a cohort
of Marine officer candidates (n = 874). All of these findings
of an association differ from those of Warren et al,40 who
reported no association between the FMS composite score
and injuries (OR, 1.01) using the cut point of 14 in
NCAA Division I collegiate athletes (n = 167). This may
be because of the nonuniformity of their athletic study pop-
ulation, which included athletes from basketball, football,
volleyball, cross-country, track and field, swimming and
diving, soccer, golf, and tennis. In our investigation, ORs
were higher for poor FMS performance (14 points; OR,
2.00) than for poor 2-mile run performance (16.00
minutes; OR, 1.66) and current smoking (OR, 1.26).
Although the purpose of the 2-mile run test is not to assess
the injury risk, results from this investigation suggest that
poor 2-mile run performance is associated with (any) injury
risk similar to poor performance on the FMS.
AJSM Vol. XX, No. X, XXXX
CONCLUSION
Although an FMS score of 14 indicated a higher injury risk
for all injury types, the low sensitivity, PPV, and AUC sug-
gested that the FMS is not a suitable screening tool for iden-
tifying those with a higher risk for injuries in this population
using the composite score. Consideration also needs to be
given regarding the amount of time and resources that are
needed to conduct the screen. Therefore, the use of the
FMS to screen for the risk of injuries is not recommended
in this population because of low predictive values and the
subsequent misclassification of the injury risk found from
this investigation. This may be because of the multifactorial
nature of the injury risk, which is complex and has been
shown to be associated with a number of factors such as aer-
obic fitness and smoking in both military and civilian popu-
lations.20,31,32 Any injury screening assessment must
consider such factors in addition to functional movement def-
icits. The findings from our investigation warrant more scru-
tiny toward the use of the FMS in soldier populations.
Additional analyses of predictive values should be conducted
in the military and other active populations to compare and
contrast findings from this investigation.
Future investigations should focus on quantifying
measures of the predictive value using the FMS as
a screening tool for the injury risk and determine whether
a score of 14 is the optimal cut point for identifying the
injury risk based on the injury type. If other investigations
uncover similar low measures of the predictive value (even
with the shifting of cut points) as well as similar relative
risks for those who score .14, then it may be that a numer-
ical scoring model is insufficient to assess the injury risk
using the FMS.
REFERENCES
1. Boston University School of Public Health. Positive and negative pre-
dictive value. Available at: http://sph.bu.edu/otlt/MPH-Modules/EP/
EP713_Screening/EP713_Screening5.html. Accessed September 4,
2014.
2. Brown MT. The Ability of the FMS in Predicting Injury Rates in Division
I Female Athletes. Thesis. Toledo, Ohio: University of Toledo; 2011.
3. Butler RJ, Contreras M, Burton LC, Plisky PJ, Goode A, Kiesel K.
Modifiable risk factors predict injuries in firefighters during training
academies. Work. 2013;46(1):11-17.
4. Chorba RS, Chorba DJ, Bouillon LE, Overmyer CA, Landis JA. Use of
Functional Movement Screening tool to determine injury risk in
female collegiate athletes. North Am J Sports Phys Ther.
2010;5(2):47-54.
5. Cook G, Burton L, Hoogenboom B. Pre-participation screening: the
use of fundamental movements as an assessment of function, part
1. North Am J Sports Phys Ther. 2006;1(2):62-71.
6. Cook G, Burton L, Hoogenboom B. Pre-participation screening: the
use of fundamental movements as an assessment of function, part
2. North Am J Sports Phys Ther. 2006;1(3):132-139.
7. Cook G, Burton L, Hoogenboom BJ, Voight M. Functional Movement
Screening: the use of fundamental movements as an assessment of
function, part 1. Int J Sports Phys Ther. 2014;9(3):396-409.
8. Cook G, Burton L, Hoogenboom BJ, Voight M. Functional Movement
Screening: the use of fundamental movements as an assessment of
function, part 2. Int J Sports Phys Ther. 2014;9(4):549-563.
9. Elias JE. The inter-rater reliability of the Functional Movement Screen
within an athletic population using untrained raters [published online
Downloaded from ajs.sagepub.com at TRENT UNIV on December 17, 2015
FMS and Injury Risk: Association and Predictive Value 7
July 8, 2013]. J Strength Cond Res. doi:10.1519/JSC.0b013e
3182a1ff1d.
10. Finkelstein EA, Hong C, Malavika P, Trogdon JG, Corso PS. The rela-
tionship between obesity and injuries among U.S. adults. Am J
Health Promotion. 2007;21(5):460-468.
11. Frost DM, Beach TA, Callaghan JP, McGill SM. FMS scores change
with performers’ knowledge of the grading criteria: are general,
whole-body movement screens capturing dysfunction [published
online November 20, 2013]? J Strength Cond Res. doi:10.1519/
JSC.0b013e3182a95343.
12. Gribble PA, Brigle J, Pietrosimone BG, Pfile KR, Webster KA. Intra-
rater reliability of the Functional Movement Screen. J Strength
Cond Res. 2013;27(4):978-981.
13. Grier T, Canham-Chervak M, McNulty V, Jones BH. Extreme condi-
tioning programs and injury risk in a US Army brigade combat team.
US Army Med Dep J. 2013;(Oct-Dec):36-47.
14. Gulgin H, Hoogenboom B. The FMS: an inter-rater reliability study
between raters of varied experience. Int J Sports Phys Ther.
2014;9(1):14-20.
15. Hoffman J. Resistance training and injury prevention. Available at:
http://www.acsm.org/docs/current-comments/rtandip.pdf. Accessed
February 2, 2015.
16. Hoover D, Killian CB, Bourcier B, Lewis S, Thomas J, Willis R. Predic-
tive validity of the Functional Movement Screen in a population of
recreational runners training for a half marathon. Med Sci Sports
Exerc. 2008;40(5):S219.
17. Kazman JB, Galecki JM, Lisman P, Deuster PA, O’Connor FG. Factor
structure of the Functional Movement Screen in marine officer candi-
dates. J Strength Cond Res. 2014;28(3):372-378.
18. Kiesel KB, Butler RJ, Plisky PJ. Prediction of injury by limited and
asymmetrical fundamental movement patterns in American football
players. J Sports Rehabil. 2014;23(2):88-94.
19. Kiesel KB, Plisky PJ, Voight ML. Can serious injury in professional
football be predicted by a preseason Functional Movement Screen?
North Am J Sports Phys Ther. 2007;2(3):147-158.
20. Knapik JJ, Ang P, Reynolds K, Jones B. Physical fitness, age, and
injury incidence in infantry soldiers. J Occup Med. 1993;35(6):598-
603.
21. Krumrei K, Flanagan M, Bruner J, Durall C. The accuracy of the Func-
tional Movement Screen to identify individuals with an elevated risk
of musculoskeletal injury. J Sport Rehabil. 2014;23:360-364.
22. Letafatkar A, Hadadnezhad M, Shojaedin S, Mohamadi E. Relation-
ship between Functional Movement Screening score and history of
injury. Int J Sports Phys Ther. 2014;9(1):21-27.
23. Li Y, Wang X, Chen X, Dai B. Exploratory factor analysis of the Func-
tional Movement Screen in elite athletes [published online December
3, 2014]. J Sports Sci. doi:10.1080/02640414.2014.986505.
24. Lisman P, O’Connor FG, Deuster PA, Knapik JJ. Functional Move-
ment Screen and aerobic fitness predict injuries in military training.
Med Sci Sports Exerc. 2013;45(4):636-643.
25. Medical Biostatistics. ROC curve. Available at: http://www.medical
biostatistics.com. Accessed September 4, 2014.
26. Minick KI, Kiesel KB, Burton L, Taylor A, Plisky P, Butler RJ. Interrater
reliability of the Functional Movement Screen. J Strength Cond Res.
2010;24(2):479-486.
27. NCSSM Statistics Leadership Institute. Sensitivity and specificity.
Durham, North Carolina: North Carolina School of Science and Math-
ematics; 1999. Available at: http://www.ncssm.edu/courses/math/
Stat_Inst/Stats2007/Stat%20and%20Calc/Sensitivity%20and%20
Specificity.pdf. Accessed February 2, 2015.
28. O’Connor FG, Deuster PA, Davis J, Pappas CG, Knapik JJ. Func-
tional Movement Screening: predicting injuries in officer candidates.
Med Sci Sports Exerc. 2011;43(12):2224-2230.
29. Onate JA, Dewey T, Kollock RO, et al. Real-time intersession and
interrater reliability of the Functional Movement Screen. J Strength
Cond Res. 2012;26(2):408-415.
30. Parenteau GE, Gaudreault N, Chambers S, Boisvert C, Gagne G,
Balq F. FMS: a reliable screening test for young elite ice hockey play-
ers. Phys Ther Sport. 2014;15(3):169-175.
8 Bushman et al
31. Reynolds K, Cosio-Lima L, Bovill M, Tharion W, Williams J, Hodges
T. A comparison of injuries, limited duty days, and injury risk factors
in infantry, artillery, construction engineers, and special forces sol-
diers. Mil Med. 2009;174(7):702-708.
32. Sacks JJ, Nelson DE. Smoking and injuries: an overview. Prev Med.
1994;23(4):515-520.
33. Schultz R, Anderson SC, Matheson GO, Marcello B, Besier T. Test-retest
and interrater reliability of the FMS. J Athl Train. 2013;48(3):331-336.
34. Smith CA, Chimera NJ, Wright NJ, Warren M. Interrater and intrarater
reliability of the Functional Movement Screen. J Strength Cond Res.
2013;27(4):982-987.
35. Sorenson E. Functional Movement ScreenTM as a Predictor of Injury
in High School Basketball Athletes. Dissertation. Eugene, Oregon:
Department of Human Physiology, University of Oregon; 2009.
36. Stobierski LM, Fayson SD, Minthorn LM, Valovich McLeod TC, Welch
CE. Clinician scoring of the FMS is reliable to assess movement pat-
terns [published online July 23, 2014]. J Sports Rehabil. doi:10.1123/
jsr.2013-0139.
37. Taimela S, Kujala UM, Osterman K. Intrinsic risk factors and athletic
injuries. Sports Med. 1990;9(4):205-215.
For reprints and permission queries, please visit SAGE’s Web site at http://www.sagepub.com/journalsPermissions.nav.
Downloaded from ajs.sagepub.com at TRENT UNIV on December 17, 2015
The American Journal of Sports Medicine
38. Tape TG. The area under an ROC curve. Available at: http://gim
.unmc.edu/dxtests/roc3.htm. Accessed September 4, 2014.
39. Teyhen DS, Shaffer SW, Lorenson CL, et al. The Functional Move-
ment Screen: a reliability study. J Orthop Sports Phys Ther.
2012;42(6):530-540.
40. Warren M, Smith CA, Chimera NJ. Association of FMS with injuries in
division I athletes [published online September 8, 2014]. J Sports
Rehabil. doi:10.1123/jsr.2013-0141.
41. Whiteside D, Deneweth JM, Pohorence MA, et al. Grading the FMS:
a comparison of manual (real-time) and objective methods [published
online August 26, 2014]. J Strength Cond Res. doi:10.1519/JSC.000
0000000000654.
42. Wieczorkowski M. Functional Movement ScreenTM as a Predictor of
Injury in High School Basketball Athletes. Thesis. Toledo, Ohio:
Department of Kinesiology, University of Toledo; 2010.
43. Zalai D, Panics G, Bobak P, Csaki I, Hamar P. Quality of functional
movement patterns and injury examination in elite-level male
professional football players [published online December 6,
2014]. Acta Physiologica Hungarica. doi:10.1556/APhysiol.101.
2014.010.