Journal of Strength and Conditioning Research Publish Ahead of Print

DOI: 10.1519/JSC.0000000000001943

The relationship of core strength and activation and performance on three functional
movement screens
Running Head: Relationship of core strength and three functional movement screens
Authors:
Caleb D. Johnson a

D
Paul N. Whitehead a
Erin R. Pletcher a

TE
Mallory S. Faherty b
Mita T. Lovalekar a
Shawn R. Eagle a
Karen A. Keenan a
a. University of Pittsburgh, Neuromuscular Research Laboratory, 3860 South Water St.,
EP
Pittsburgh, PA 15203
b. Michael W. Krzyzewski Human Performance Laboratory, Duke University, DUMC
102916, Durham, NC 27705

Corresponding Author: Caleb Johnson

3860 South Water St.
C

Pittsburgh, PA, 15203

Phone: 412-246-0460
C

Fax: 412-246-0461
cdj20@pitt.edu
A

Funding Source:
We would like to acknowledge the Freddie Fu, MD Graduate Research Award for funding this
project

Copyright ª 2017 National Strength and Conditioning Association

 1

ABSTRACT

Current measures of core stability utilized by clinicians and researchers suffer from several

shortcomings. Three functional movement screens appear, at face-value, to be dependent on the

ability to activate and control core musculature. These three screens may present a viable

alternative to current measures of core stability. Thirty-nine subjects completed a deep squat,

D
trunk stability push-up, and rotary stability screen. Scores on the three screens were summed to

calculate a composite score (COMP). During the screens, muscle activity was collected to

TE
determine the length of time that the bilateral erector spinae, rectus abdominus, external oblique,

and gluteus medius muscles were active. Strength was assessed for core muscles (trunk

flexion/extension, trunk rotation, hip abduction/adduction) and accessory muscles (knee

flexion/extension, and pectoralis major). Two ordinal logistic regression equations were
EP
calculated with COMP as the outcome variable, and: 1) core strength and accessory strength, 2)

only core strength. The first model was significant in predicting COMP (p=.004) (Pearson’s Chi-

Square=149.132, p=.435; Nagelkerke’s R-Squared=.369). The second model was significant in

C

predicting COMP (p=.001) (Pearson’s Chi Square=148.837, p=.488; Nagelkerke’s R-

Squared=.362). The core muscles were found to be active for the majority of screens, with
C

percentages of “time active” for each muscle ranging from 54%-86%. In conclusion,

performance on the three screens is predicted by core strength, even when accounting for
A

“accessory” strength variables. Further, it appears the screens elicit wide-ranging activation of

core muscles. While more investigation is needed, these screens, collectively, appear to be a

good assessment of core strength.

KEYWORDS: FMS, core stability, surface electromyography

Copyright ª 2017 National Strength and Conditioning Association

 2

INTRODUCTION

Core stability can be defined as the ability to achieve and sustain control of the trunk and

hip regions at rest and during precise movement (12, 22). Core stability is accomplished through

a complex interaction of neuromuscular coordination, proprioception, strength and endurance of

the trunk and hip musculature (12, 22). This inherent complexity results in a lack of consistent

D
measurement for core stability across studies. Most objective tests aimed at quantifying core

stability are limited as they either: (a) only measure one aspect of this complex interaction, such

TE
as core strength or (b) assess a more complete picture of this interaction, but in positions that are

functionally irrelevant when the proposed role of core stability during dynamic movement is

considered (3, 12, 19, 20, 29-31). Subjectively scored tests seem to suffer the same shortcomings
EP
as objective tests, along with poor inter- and intra-observer reliability (27-29). To summarize, it

would seem that a more comprehensive and reliable measure of core stability is needed; both for

research and clinical purposes.

C

A test that may better address the multi-dimensional nature of core stability is the

Functional Movement Screen (FMS). This screen involves the subjective scoring of seven
C

different patterns aimed at identifying functional deficits in strength, balance, flexibility, and

neuromuscular control. Several studies have suggested that the FMS Composite score is
A

associated with musculoskeletal injury in collegiate and professional athletes, however the

overall predictive value of the screen has been called into question in recent literature (1, 5, 16,

17). Despite this, the important aspects of the FMS, in relation to the current study, are their

possible application to assessing core stability. Secondly, while the screens are scored

subjectively, they have been shown to have good inter-tester and inter-session reliability (4).

Copyright ª 2017 National Strength and Conditioning Association

 3

Three of the seven movement screens, the Deep Squat (DS), Rotary Stability (RS) and

Trunk Stability Push-up (TSP), are specifically aimed at assessing core stability and involve

dynamic tasks in all three planes of motion (frontal, sagittal, and transverse). Additionally, the

screens assess the core musculature’s ability to keep the lumbopelvic complex in a neutral

position, which has been shown to be optimal for loading of the trunk without sustaining injury

D
(12, 29). These three screens would seem to hold good face-validity in assessing functional core

stability, however their relationship to components of core stability have not been established.

TE
The purpose of the current study was to investigate the relationship between performance

on the DS, RS, and TSP screens (referred to as FMS-3 moving forward) and components of core

stability. First, we examined the relationship between performance on the three screens and
EP
isokinetic strength of the trunk and hip musculature (core muscles) compared to other

musculature (accessory muscles) that may play a role in successful completion of one or more of

the screens. We hypothesized the performance on the screens would be predicted by core muscle

strength and that core muscle strength would explain a greater amount of variance in FMS-3
C

performance than strength of the accessory muscles. Second, we assessed whether the screens

elicit activation of the core musculature, measured through surface electromyography. We

C

hypothesized the three screens would elicit widespread activation of the core musculature. If in

the expected direction, the results of this study would provide evidence for the use of the FMS-3
A

as a viable measure of core muscle function. Further, they would provide justification for further

research into the FMS-3 as a valid measure of core stability.

Copyright ª 2017 National Strength and Conditioning Association

 4

METHODS

Experimental Approach to the Problem

An observational, cross-sectional study design was utilized to test both hypotheses.

Subjects completed all testing on the same day, including: completion of the FMS-3 while core

muscle activity was collected; isokinetic strength for trunk flexion/extension (TFS/TES),

D
bilateral trunk rotation (TRS), and bilateral knee flexion/extension (KFS/KES); and isometric

strength for bilateral hip abduction/adduction (HABS/HADS) and the pectoralis major group

TE
(PMS). Core strength variables were selected based on common definitions of the core region

found in previous literature (12, 29). Accessory strength variables were selected based on our

assessment of the FMS-3 and what muscle groups may also play a role in performance of them.
EP
Pectoralis major strength was included due to the pectoralis major’s role in extension of the arm

and horizontal adduction of the humerus during the TSP. Knee flexion/extension strength was

included due to the hamstrings role in knee flexion, and the quadriceps role in knee extension
C

during the DS screen. Further, co-contraction of the hamstrings and quadriceps is thought to be

important for maintaining the knee joint in a neutral position, one of the criteria the DS screen is
C

scored on (10). Order of testing for the FMS-3 and trunk strength measures was block

randomized.
A

Subjects

Eighteen males (Age= 22.22 ± 2.32 years, Weight= 77.24 ± 6.70 kg, Height= 179.94 ±

8.89 cm) and 21 females (Age= 22.36 ± 3.58 years, Weight= 61.12 ± 8.28 kg, Height= 165.62 ±

7.00 cm) were recruited from an athletic population, with the criteria of having participated in an

organized, land-based sport within the past year. Further, the sport had to involve dynamic

Copyright ª 2017 National Strength and Conditioning Association

 5

movement with changes of direction (i.e. soccer, basketball, volleyball etc..). Subjects were

excluded based on: 1) current musculoskeletal injury or pain, or injury in the past three months,

2) musculoskeletal surgery in the past three years, 3) concussion in the past three months or the

presence of neurological disorders, 4) recent history of screening with the FMS. Musculoskeletal

injury was defined as any impairment that resulted in a limitation to participation in sport,

D
training or activities of daily living for more than a week. Approval by the Institutional Review

Board at the University of Pittsburgh and informed consent by individual subjects was obtained

TE
before any testing procedures were carried out.

Procedures

Core Muscle Activity

EP
Core muscle activity during the FMS-3 was obtained using surface electromyography

(EMG) (Noraxon USA, Scottsdale, AZ). Muscle activity was recorded bilaterally for the erector

spinae (level of L3), external oblique, rectus abdominis, and gluteus medius. Placement sites for
C

each muscle were determined based on SENIAM recommendations (7) and Cram & Kasman

(13). Sites were prepared based on SENIAM recommendations (7) and two bipolar Ag/AC-CI
C

electrodes (AMBU Blue Sensor N electrodes; ABMU, Glen Burnie, MD) were applied to each

site, parallel to the muscle fibers and approximately 2 cm apart. Electrodes were attached to
A

wireless transmitters and placement sites were verified with submaximal contractions prior to

electrodes and transmitters being secured with adhesive tape and cover-all (3M Health Care Ltd.,

Loughborough, UK). A sitting “quiet-trial” was collected to assess the respective resting levels

of muscle activity. A sampling frequency of 1000 Hz was used for all EMG data collection.

Copyright ª 2017 National Strength and Conditioning Association

 6

Functional Movement Screening

The FMS-3 were performed by subjects according to the standardized procedures and

scripts described by Cook (2) and in previous studies (16). The only alteration made to the

screening procedures was to give subjects a countdown before each screen for the purposes of

synchronizing surface EMG data collection. Each screen was scored on a scale of 0-3 points,

D
however a score of zero indicates pain was experienced with the respective movement for a

given screen. Since no subjects experienced pain with screening, the scoring for each screen was

TE
effectively reduced to a scale of 1-3. Screens were administered and scored live by an FMS

certified rater with ample experience. For a sub-sample of subjects (n=29), screens were also

recorded with a digital video camera and then scored a second time, after a minimum of 2
EP
months, and the inter-rater reliability for the scoring is described below. For muscle activity, data

collection was initiated approximately two seconds before the subject initiated the specific

movement for each screen and was terminated when the subject reached the end-point defined

for each screen.

C

Isokinetic Strength Testing

C

All isokinetic strength testing was completed on the Biodex System 3 (Biodex Medical

Systems, Shirley, NY) using concentric/concentric, reciprocal contractions at a speed of 60 °/sec.

A

Subjects were positioned and stabilized according to the manufacturer guidelines, and three

practical/warm-up trials were given at 50% of perceived maximal effort and 100% of perceived

maximal effort. Following a 90 second rest period, subjects completed five trials at 100% effort

for analysis. Isokinetic testing on the Biodex System 3 has shown good reliability and precision

(6).

Copyright ª 2017 National Strength and Conditioning Association

 7

Isometric Strength Testing

A hand-held dynamometer (Lafayette Instrument Co., Lafayette, IN) was used for all

isometric strength testing. Positioning of the subjects and testers was based on manual muscle

testing procedures outlined by Hislop & Montgomery (11), and testing procedures were based

off of established and reliable protocols in past literature, which have also been shown to be

D
reliable (ICC=.76-.989) through previous testing in our lab (15, 26). Hip abduction/adduction

strength and SADS were collected using “make tests”, where subjects were asked to push

TE
maximally against the hand-held dynamometer and testers matched the force being exerted by

the subject. A 50% warm-up was given for each movement and two or three trials, depending on

if the first two were within 15% of each other, were collected at 100% effort for analysis.
EP
Data Reduction

A composite score (COMP) was calculated for the FMS-3, out of a possible 4-12 points,

by summing the four individual scores from each screen. The RS screen is performed separately
C

on either side of the body, therefore the left and right scores were added into the COMP

separately. Isokinetic strength was represented as the average peak torque across the five
C

consecutive contractions for each movement. Isometric strength was represented as the average

of either two or three maximal trials, dependent on the consistency of trials. Averaging across
A

bilateral limbs was appropriate given that none of the movement patterns tested involved

unilateral movements. Further, this limited the number of independent variables in the calculated

regression equations.

A custom Matlab program (Matlab R2014a, MathWorks, Natick, MD) was used for

EMG data reduction. All EMG data were rectified and filtered through a 16-500 Hz band-pass

Copyright ª 2017 National Strength and Conditioning Association

 8

filter. Each trial was cropped to reflect the extra two seconds of data collection prior to the

initiation of a movement by the subject. The level for a muscle being “active” was defined as

three standard deviations above the mean amplitude of the signal during the quiet trial. The

percentage of time that each muscle was active for each screen was calculated by dividing the

time that each muscle met the above criteria for being “active” by the total trial time. Data were

D
averaged for bilateral muscle groups and across the three screens, following similar logic as

expressed for bilateral strength variables.

TE
Statistical Analyses

All statistical analyses were conducted using IBM SPSS Statistics 23 (IBM, Armonk

NY). Descriptive statistics were calculated to examine the central tendency and variability for
EP
percentage of time that core muscle groups were active across the three screens. Frequencies

were calculated for COMP. Reliability coefficients (intra-class correlation coefficients and

Cohen’s Kappa statistics) and percent agreements were calculated to assess the scoring reliability
C

for the FMS scores and COMP.

Two ordinal regression equations were developed to assess the prediction of COMP by
C

core muscle strength. Strength variables were checked for normality by visual inspection of

histograms and Shapiro-Wilk tests of normality. Pearson’s correlation coefficients were

A

calculated to assess the inter-correlation among strength variables, and Spearman’s Rank

Coefficients were calculated to assess the bivariate correlation among strength variables and

COMP. Strength variables were eliminated based on concerns of multi-collinearity as well as the

strength of their bivariate correlation with COMP. The first ordinal regression equation

contained both core (TRS, HABS, HADS) and accessory (KFS, PMS) strength variables. To

compare the relative reduction in explained COMP variance, the second equation only contained

Copyright ª 2017 National Strength and Conditioning Association

 9

core strength (TRS, HABS, HADS) variables. Level for significance was set at α<.05 for all

tests.

RESULTS

COMP scores were found to be effectively reduced to a scale of 6-10, however the range

of scores within the reduced scale was evenly distributed: 6 (20.5%), 7 (25.6%), 8 (15.4%), 9

D
(17.9%), 10 (20.5%).

TE
Core Muscle Activity

Descriptive statistics for core muscle activity during the FMS-3 is presented in Table 1.

The sample size for muscle activity data was reduced (n=38) due to issues with Surface EMG
EP
equipment causing one subject’s data to be discarded. The measured core muscles were found to

be active for the majority (>50%) of the screens. The gluteus medius showed the greatest percent

activation (86.2 ± 10.6%), and the rectus abdominis showed the lowest (54.9 ± 12.8%).

*** TABLE 1 SOMEWHERE NEAR HERE ***

C

Inter-rater Reliability for FMS Scoring

C

Reliability coefficients for the scoring of individual FMS and the COMP are presented in

Table 2. All screens were found to have sufficient reliability (Cohen’s Kappa = .635-.887) except
A

for the Left Rotary Stability screen (Cohen’s Kappa =.018). This low reliability coefficient can

be attributed to a lack of variability in the scorings, with a high number of subjects scoring a two

on the screen. Supporting this assertion is the high percent agreement between scorings

(93.10%). Composite score showed excellent reliability (ICC 2,1 = .917) and percent agreement

(72.41%).

Copyright ª 2017 National Strength and Conditioning Association

 10

*** TABLE 2 SOMEWHERE NEAR HERE ***

Core Strength and FMS-3 Composite Score

Bivariate correlations between strength variables and COMP are presented in Table 3.

The full results of the 1st and 2nd ordinal regression equations are presented in Tables 4 and 5

respectively. The full set of predictors showed significance in predicting COMP for both models,

D
and the relative amount of variance explained by the models, denoted by the Pseudo R- Square

values, was similar (.369 to .362). Finally, TRS was the only significant independent predictor of

TE
COMP for both models.

*** TABLES 3-5 SOMEWHERE NEAR HERE ***

EP
DISCUSSION

This manuscript details the relationship between performance on the FMS-3, and

activation and strength of the core musculature. It was hypothesized that performance on the

screens would be predicted by core muscle strength and that core muscle strength would explain
C

a greater amount of variance in FMS-3 performance than strength of the accessory muscles.

Further, it was hypothesized that the three screens would elicit widespread activation of the core
C

musculature.
A

Core strength measures, as a group, were found to be significantly predictive of COMP

on the FMS-3. While the grouping of core strength and accessory strength variables were also

significantly predictive of FMS-3 performance, the Pseudo R-Squared values were highly similar

between the two equations. While Pseudo R-Squared values cannot be interpreted as the exact

percentage of variance in outcome variables explained by the predictor variables, they can be

interpreted relative to another Pseudo R-Squared value calculated on the same data set (21).

Copyright ª 2017 National Strength and Conditioning Association

 11

Therefore, the minimal difference in values with the elimination of accessory strength variables

can be interpreted as the core strength variables explaining significantly more variance in

COMP.

Few studies have assessed the relationship of performance on the FMS and core muscle

function. Our results are in-line with prior research by Mitchell et al. (24), showing moderate

D
correlations between core muscle endurance measures and composite score on the full FMS

battery. However, correlations were not reported for performance on individual screens, giving

TE
these results little applicability to the current study. Our results are counter to those reported by

Okada et al. (25), showing no correlation between core muscle endurance and individual

performance on any of the FMS. Given the short duration of the screens and quick nature of the
EP
perturbations they provide to the core, it is not overly surprising that correlations were not found

with endurance of the trunk musculature. It would seem natural that trunk muscle strength, which

was measured in the current study, would be a stronger predictor of performance on these

screens.
C

Trunk rotation strength was the only independent predictor of COMP in either of the
C

models. This result should be interpreted with caution however, given the number of variables

that could not be entered into the equation (KES, TFS, TES) due to multicollinearity among the
A

strength variables. Further, TFS and TES (in discussing only the core strength variables) both

showed significant, moderate correlations with COMP (Table 3). However, TRS did show the

highest bivariate correlation (R =.547) with COMP. A possible explanation may be that the

performance of the TRS isokinetic testing elicited recruitment of smaller, deeper muscles,

sometimes termed local stabilizers (8). These muscles are thought to play a larger role in core

stability during low-load conditions; a designation applicable to the screenings included in this

Copyright ª 2017 National Strength and Conditioning Association

 12

study. In contrast, the TFS/TES isokinetic testing fixes the spine in a more rigid position, and

therefore may predominantly elicit recruitment of the larger, more superficial muscles that

mainly function as movers or to provide stability in higher-load conditions (29).

It was found that the core muscles, including the erector spinae, rectus abdominis,

external obliques, and gluteus medius were active for the majority of the FMS-3. The rectus

D
abdominis were found to be active for the lowest percentage of time, at 54.9% of total screening

time. This may have been partially attributed to the adverse effects of increased adipose tissue

TE
around the abdominal area on EMG signals. However, Hamlyn et al. (9) reported no significant

increases in lower abdominal muscle activity between trunk stabilization exercises and normal

resistance exercises. Further, Marshall et al. (23) showed mixed results when assessing
EP
differences in rectus abdominis activity between exercises performed on a stable and unstable

surface. Therefore, it is also possible that the rectus abdominis does not play a vital role in core

stabilization. Further work would be needed to confirm this, however, therefore the current

results can only lead us to conclude that the FMS-3 elicited to lowest activity from the rectus
C

abdominis group.
C

The range of scores was found to effectively reduce the COMP scale from 4-12 to 6-10.

While it could be that subjects whose abilities fall on either end of the spectrum were simply not
A

obtained in the study sample, the subjects recruited ranged from recreational to Division I

athletes. A better explanation would seem to be that alterations are needed in the scoring of

several of the screens to make them more sensitive to a competitive athlete population and more

specific to measuring stability of trunk and hip regions. The vast majority of subjects (82%)

scored a two on the RS screen, in both directions. The scoring criteria for the RS screen is based

on successful completion of a repetition, where the non-moving hand and foot remain in contact

Copyright ª 2017 National Strength and Conditioning Association

 13

with the testing board (2). Adding criteria that places a focus on control of the movement, may

delineate between individuals who can simply perform the motion and those who can perform it

without excessive movement through the trunk and hips. In contrast, the DS screen is scored on a

number of criteria, including depth of the squatting motion, alignment of the lower extremities,

and position of the trunk and dowel rod relative to the hips (2). Altering the scoring of this screen

D
to be primarily dependent on control and position of the trunk and hips throughout the movement

would help to make performance on it more specific to core stability.

TE
There are several limitations to the current study. While the subjective scoring of the

FMS has been shown to have fair to excellent reliability in several studies, the results have been

mixed, with several studies showing poor reliability (4). In general, the reliability of FMS
EP
scoring has been favorable and further, the inter-session reliability for the rater used in the

current study was shown to be adequate for all screens (4). Another limitation inherent to the

FMS, identified in recent literature, is a lack of internal consistency among the seven screens,

indicating that these screens are not measuring a single construct (14, 18). One could argue that,
C

although the FMS is meant to assess overall movement quality, the individual screens are

intended to measure different aspects of overall movement quality, possibly accounting for this
C

lack of internal consistency. This limitation has more implications for the use of the composite

score on the complete FMS, however it establishes a need to assess the internal consistency of
A

the FMS-3 moving forward.

A second limitation lies in the generalizability of the reported results, with the study

sample only including competitive athletes. Our reasoning in recruiting the current sample was

that the role of core stabilization is likely different for individuals with varying levels of athletic

Copyright ª 2017 National Strength and Conditioning Association

 14

skill and overall physical activity. Therefore, the relationship between the variables included in

our analyses would need to be assessed separately for differing populations.

The most important limitation is in the interpretability of the main analyses; delineating

between core strength and core stability. As was stated before, proper core stability is a product

of a complex interaction of core muscle strength and endurance, neuromuscular coordination and

D
control, and proprioception (12). Based on this model, the interpretation of the results is limited

to the relationship between one component of core stability (core strength) and performance on

TE
the FMS-3. Further, while core muscle activation was measured, the level of activation, in terms

of a percentage of each subject’s maximum volitional contraction, was not. Therefore, the

interpretation of the reported muscle activity should be limited to a confirmation that the core
EP
muscles were active for a significant portion of the FMS-3.

In conclusion, the results of this study show a relationship between core muscle strength

and performance on the DS, RS, and TSP screens taken from the FMS. Further, the screens were
C

found to elicit widespread activation of the core musculature. While several alterations to the

screens and their scoring may be required, our results provide evidence that these three screens
C

are a viable and clinic-friendly measure of core muscle strength. Broadening the focus of the

current study, future research is needed to determine the relationship of performance on these
A

three screens and the remaining components of core stability.

PRACTICAL APPLICATIONS

The results show that performance on the DS, RS and TSP screens from the FMS are

capturing some aspect of core strength and core muscle activation. Further, they show that

scoring on these three screens, as well as the COMP score computed by adding the individual

Copyright ª 2017 National Strength and Conditioning Association

 15

scores together, can be done reliably within a single tester. These three screens take several

minutes to administer and require very little, as well as low-cost, equipment. The practical

applications of the current study, then, center on providing evidence for this battery of screens

that can be easily utilized by clinicians in assessing their client’s ability to stabilize the trunk and

hip regions during dynamic movement. While several limitations to the interpretation and

D
generalizability of our results are present, our results still provide compelling evidence that this

is the case.

TE
REFERENCES

1. Chorba RS, Chorba DJ, Bouillon LE, Overmyer CA, and Landis JA. Use of a functional
movement screening tool to determine injury risk in female collegiate athletes. North
American journal of sports physical therapy: NAJSPT 5: 47, 2010.
EP
2. Cook G, Burton L, Kiesel KB, Rose G, and Bryant M. Movement: Functional Movement
Systems. On Target Publications, 2010.
3. Cowley PM and Swensen TC. Development and reliability of two core stability field
tests. Journal of strength and conditioning research / National Strength & Conditioning
Association 22: 619-624, 2008.
4. Cuchna JW, Hoch MC, and Hoch JM. The interrater and intrarater reliability of the
functional movement screen: A systematic review with meta-analysis. Physical therapy
C

in sport : official journal of the Association of Chartered Physiotherapists in Sports

Medicine 19: 57-65, 2016.
5. Dorrel BS, Long T, Shaffer S, and Myer GD. Evaluation of the Functional Movement
Screen as an Injury Prediction Tool Among Active Adult Populations: A Systematic
C

Review and Meta-analysis. Sports health 7: 532-537, 2015.

6. Drouin JM, Valovich-mcLeod TC, Shultz SJ, Gansneder BM, and Perrin DH. Reliability
and validity of the Biodex system 3 pro isokinetic dynamometer velocity, torque and
position measurements. European journal of applied physiology 91: 22-29, 2004.
A

7. Freriks B and Hermens H. European recommendations for surface electromyography:

results of the SENIAM project. Roessingh Research and Development, 2000.
8. Gibbons SG and Comerford MJ. Strength versus stability part 1; concept and terms.
Orthopaedic Division Review 43: 21-27, 2001.
9. Hamlyn N, Behm DG, and Young WB. Trunk muscle activation during dynamic weight-
training exercises and isometric instability activities. Journal of strength and
conditioning research / National Strength & Conditioning Association 21: 1108-1112,
2007.
10. Hewett TE, Myer GD, and Ford KR. Anterior cruciate ligament injuries in female
athletes part 1, mechanisms and risk factors. The American journal of sports medicine 34:
299-311, 2006.

Copyright ª 2017 National Strength and Conditioning Association

 16

11. Hislop H, Avers D, and Brown M. Daniels and Worthingham's muscle testing:
Techniques of manual examination and performance testing. Elsevier Health Sciences,
2013.
12. Huxel Bliven KC and Anderson BE. Core stability training for injury prevention. Sports
health 5: 514-522, 2013.
13. Kasman GS, Cram JR, and Holtz J. Introduction To Surface Electromyography. Aspen
Publeshers, 1998.
14. Kazman JB, Galecki JM, Lisman P, Deuster PA, and O'Connor FG. Factor structure of
the functional movement screen in marine officer candidates. Journal of strength and
conditioning research / National Strength & Conditioning Association 28: 672-678,

D
2014.
15. Kelln BM, McKeon PO, Gontkof LM, and Hertel J. Hand-held dynamometry: reliability
of lower extremity muscle testing in healthy, physically active,young adults. Journal of
sport rehabilitation 17: 160-170, 2008.

TE
16. Kiesel K, Plisky PJ, and Voight ML. Can serious injury in professional football be
predicted by a preseason functional movement screen? North American journal of sports
physical therapy: NAJSPT 2: 147, 2007.
17. Kiesel KB, Butler RJ, and Plisky PJ. Prediction of injury by limited and asymmetrical
fundamental movement patterns in american football players. Journal of sport
rehabilitation 23: 88-94, 2014.
18. Koehle MS, Saffer BY, Sinnen NM, and MacInnis MJ. Factor Structure and Internal
EP
Validity of the Functional Movement Screen in Adults. Journal of strength and
conditioning research / National Strength & Conditioning Association 30: 540-546,
2016.
19. Liemohn WP, Baumgartner TA, Fordham SR, and Srivatsan A. Quantifying core
stability: a technical report. Journal of strength and conditioning research / National
Strength & Conditioning Association 24: 575-579, 2010.
20. Liemohn WP, Baumgartner TA, and Gagnon LH. Measuring core stability. Journal of
C

strength and conditioning research / National Strength & Conditioning Association 19:
583-586, 2005.
21. Long JS. Regression models for categorical and limited dependent variables. Thousand
C

Oaks, CA: Sage Publications, 1997.

22. Majewski-Schrage T, Evans TA, and Ragan B. Development of a core-stability model: a
delphi approach. Journal of sport rehabilitation 23: 95-106, 2014.
23. Marshall PW and Murphy BA. Core stability exercises on and off a Swiss ball. Archives
A

of physical medicine and rehabilitation 86: 242-249, 2005.

24. Mitchell UH, Johnson AW, and Adamson B. Relationship between Functional Movement
Screen Scores, Core Strength, Posture, and BMI in School Children in Moldova. Journal
of strength and conditioning research/National Strength & Conditioning Association,
2015.
25. Okada T, Huxel KC, and Nesser TW. Relationship between core stability, functional
movement, and performance. Journal of strength and conditioning research / National
Strength & Conditioning Association 25: 252-261, 2011.
26. Thorborg K, Petersen J, Magnusson SP, and Holmich P. Clinical assessment of hip
strength using a hand-held dynamometer is reliable. Scandinavian journal of medicine &
science in sports 20: 493-501, 2010.

Copyright ª 2017 National Strength and Conditioning Association

 17

27. Weir A, Darby J, Inklaar H, Koes B, Bakker E, and Tol JL. Core stability: inter- and
intraobserver reliability of 6 clinical tests. Clinical journal of sport medicine : official
journal of the Canadian Academy of Sport Medicine 20: 34-38, 2010.
28. Whatman C, Hing W, and Hume P. Physiotherapist agreement when visually rating
movement quality during lower extremity functional screening tests. Physical therapy in
sport : official journal of the Association of Chartered Physiotherapists in Sports
Medicine 13: 87-96, 2012.
29. Willson JD, Dougherty CP, Ireland ML, and Davis IM. Core stability and its relationship
to lower extremity function and injury. The Journal of the American Academy of
Orthopaedic Surgeons 13: 316-325, 2005.

D
30. Zazulak BT, Hewett TE, Reeves NP, Goldberg B, and Cholewicki J. Deficits in
neuromuscular control of the trunk predict knee injury risk: a prospective biomechanical-
epidemiologic study. The American journal of sports medicine 35: 1123-1130, 2007.
31. Zazulak BT, Hewett TE, Reeves NP, Goldberg B, and Cholewicki J. The effects of core

TE
proprioception on knee injury: a prospective biomechanical-epidemiological study. The
American journal of sports medicine 35: 368-373, 2007.
EP
ACKNOWLEDGEMENTS

We would like to acknowledge the Freddie Fu, MD Graduate Research Award for

funding this project. The authors have no conflicts of interest to report.

C
C
A

Copyright ª 2017 National Strength and Conditioning Association

Table 1. Descriptive Statistics for Core Muscle Activity During Functional Movement Screening

(n=38)

Mean ± SD Minimum Maximum

Erector spinae (% active) 60.89 ±17.94 31.47 91.70

Rectus abdominis (% active) 54.87 ± 12.76 33.14 83.10

D
External obliques (% active) 74.98 ± 11.70 48.69 89.57

Gluteus medius (% active) 86.19 ± 10.64 52.71 99.00

TE
% active= percentage of time the respective muscle was active across the three screens,

SD= standard deviation

EP
C
C
A

Copyright ª 2017 National Strength and Conditioning Association

Table 2. Intra-class Correlation Coefficients for FMS Scoring Reliability (n=29)

Mean- 1st Mean- 2nd Reliability %

Scoring Scoring Coefficient Agreement

Deep Squat 2.10 2.10 .887 93.10

Trunk Stability Push-up 1.90 2.00 .827 89.66

D
Right Rotary Stability 2.03 2.10 .635 93.10

Left Rotary Stability 1.97 2.03 .018 93.10

TE
Composite Score 8.00 8.24 .917* 72.41

*- denotes intra-class correlation coefficient (ICC 2,1); all other reliability coefficients
EP
are Cohen’s Kappa statistics

% Agreement= percentage agreement between 1st and 2nd scorings of respective screen or

composite score
C
C
A

Copyright ª 2017 National Strength and Conditioning Association

Table 3. Spearman’s Rank Coefficients for the Bivariate Correlation Between Strength Variables

and FMS Composite Score (n=39)

FMS Composite Score

Trunk Rotation Strength .547*

Trunk Extension Strength .335*

D
Trunk Flexion Strength .494*

Knee Extension Strength .498*

TE
Knee Flexion Strength .499*

Hip Abduction Strength .103

Hip Adduction Strength .226

EP
Pectoralis Major Strength .377*

*- denotes statistical significance at α<.05

C
C
A

Copyright ª 2017 National Strength and Conditioning Association

Table 4. First Ordinal Regression Equation: Core and Accessory Strength Variables (n=39)

Chi-Square df p-value -2 LL (IO): 124.420

Pearson 149.132 147 .435 -2 LL (F): 107.369
Deviance 107.369 147 .994 Chi-Square: 17.051
df: 5
Test of Parallel Lines p-value: .004

D
Model -2 LL df p-value Pseudo R-Squared: .369
Null 107.369

TE
General 95.255 15 .670

Predictor Variable Estimate SE Wald Statistic p-value

Trunk Rotation Strength .039 .016 5.819 .016
Knee Flexion Strength .011 .018 .352 .553
EP
Hip Abduction Strength -.126 .125 1.018 .313
Hip Adduction Strength .073 .120 .370 .543
Pec Major Strength -.015 .042 .126 .723
C

df= degrees of freedom; -2 LL= -2 Log Likelihood; IO= intercept only; F=final; SE=

standard error
C
A

Copyright ª 2017 National Strength and Conditioning Association

Table 5. Second Ordinal Regression Equation: Only Core Strength Variables (n=39)

Chi-Square df p-value -2 LL (IO): 124.420

Pearson 148.837 149 .488 -2 LL (F): 107.795
Deviance 107.795 149 .995 Chi-Square: 17.051
df: 3
Test of Parallel Lines p-value: .001

D
Model -2 LL df p-value Pseudo R-Squared: .362
Null 107.795

TE
General 97.104 9 .297

Predictor Variable Estimate SE Wald Statistic p-value

Trunk Rotation Strength .041 .012 11.038 .001
Knee Abduction Strength -.137 .123 1.254 .263
EP
Hip Adduction Strength .088 .110 .649 .420

df= degrees of freedom; -2 LL= -2 Log Likelihood; IO= intercept only; F=final; SE=
standard error
C
C
A

Copyright ª 2017 National Strength and Conditioning Association