Anyone who reads this blog regularly knows that I’m a big fan of assessments prior to training and repeat assessments along the way to ensure results are coming as desired.  Recently though, it seems that the Functional Movement Screen which was created by Lee Burton and Gray Cook has become extremely popular among my colleagues in the fitness industry.

If you’re not familiar with the Functional Movement Screen, here’s a little excerpt from their website describing the system:

“Put simply, the FMS is a ranking and grading system that documents movement patterns that are key to normal function. By screening these patterns, the FMS readily identifies functional limitations and asymmetries. These are issues that can reduce the effects of functional training and physical conditioning and distort body awareness.

The FMS generates the Functional Movement Screen Score, which is used to target problems and track progress. This scoring system is directly linked to the most beneficial corrective exercises to restore mechanically sound movement patterns.

Exercise professionals monitor the FMS score to track progress and to identify those exercises that will be most effective to restore proper movement and build strength in each individual.”

In short, you do 7 movements, get scored, and based on the outcome your potiential issues are identified and your training program can be designed.

 
Being the skeptic I am, I had to consult the research to determine what has and hasn’t been studied about the FMS.  And while I do believe that not everything worth doing is necessarily validated by science (yet), I’m also cautious when I see a pendulum swinging in one direction and when a great number of people are on board. It seems to me that the FMS has been accepted with little discussion as to whether it is valid or not. The more people that jump on board, the less people are apt to question it (especially when those people are big names).

So I’m not saying the FMS isn’t valid or useful or trying to discredit Lee, Gray, or anyone else who uses the system.  The next few posts are simply meant to be an examination of the existing body of scientific study on the FMS.  More to the point, I’ll be reviewing only the peer reviewed studies that have appeared in journals and not the unpublished stuff that is available on the internet (including at least one doctoral dissertation and a couple conference poster presentations).

Without further ado, let’s get to the first study.

Interrater Reliability of the Functional Movement Screen

Minik KI, et al.  J Strength Cond Res. 2010. 24(2): 479-486

While this isn’t actually the first published study on the movement screen, I thought it was important to present this first as it is indicative of the reliability of the scoring.  In a setting where different individuals would be scoring the FMS, you would need for them to be scoring the same way or the tool would be a lot less useful.

Background

To reduce injury risk, sports medicine professionals have begun to focus on improving movement patterns as opposed to focusing on rehabilitation of a specific joint.  The Functional Movement Screen has been put forth as a potential screening tool for these movement patterns.  The goal of this study is to establish interrater reliability of the FMS by comparing expert raters (who took part in the development of the FMS) with novice raters (who have completed the standardized FMS training program).

Methods

Forty students were filmed performing each of the 7 movements in the Functional Movement Screen.  Each of the subjects’ videos were then viewed by two expert and two novice raters and each of the 7 movements was independently scored as a 0, 1, 2, or 3 using the FMS criteria.  The scores were then compared using the weighted Kappa statistic.

Results

The pair of novice raters demonstrated excellent agreement on 6 of the 17 test components, including the deep squat and shoulder mobility tests, and portions of the trunk stability push-up and ASLR tests. Substantial agreement was evident on 8 of the 17 test components. The right and left components of the lunge and the final component of the rotary stability test each demonstrated moderate agreement.

The pair of expert raters varied more in scoring, with excellent agreement on 4 of the 17 test components, including the shoulder mobility test and the final component of the ASLR. Substantial agreement was seen in 9 of the 17 test components. Two components of the lunge and 2 components of the rotary stability tests demonstrated moderate agreement.

When comparing the average scores of the paired novice and expert raters, 14 of the 17 tests demonstrated excellent agreement. Substantial agreement was evident in 1 component of the rotary stability test and 2 components of the in-line lunge.

Funding

University of Evansville Honor’s Program grant and the University of Evansville’s College of Education and Health Science.

My Thoughts

The kappa statistic is actually a good choice in this case because it is fairly conservative and takes into account possible agreement of the raters due to chance.  They also used a weighted kappa which allowed them to rate larger disagreements between the raters less favorably.   However, it should be noted that the categories listed as excellent, substantial, and moderate are pretty aribitrary and are based pretty much on the personal opinions of some other researchers.  Other arbitrary guidelines exist that rate kappa values differently.   I’m not sure how much this really matters in this case, but it is important to note that these ratings aren’t universally accepted.

The most surprising thing to me was that the novice raters tended to have more agreement with each other than the expert raters.  You’d expect that over time your ratings would become more similar than different.  But why did the novice and expert rater’s results tend to agree with each other when they compared them?  My guess is that some of the differences disappeared when they averaged the two novice and two advanced raters before comparing them to each other which could possibly have masked some of the variability.

The authors did suggest that since the tests were only filmed from two angles there was a third dimension missing and this could have influenced the degree of agreement between raters.  Having evaluated people in person, I can definitely see how this could have an impact.  In person you can move around and get a better view of each movement.  Perhaps one reason the expert raters had more disagreement is because they’re more used to looking at movements in three dimensions that novice raters who are less likely to be able to make as much use of this additional information.

All in all, I think the study was fairly well conducted and the results indicate that the scoring of the FMS is relatively reliable.  It did highlight that perhaps work needs to be done to ensure expert raters are scoring similarly.  Another important point to take home is that both sets of raters tended to struggle with agreement on the lunge and rotary stability tests so if you’re using the FMS you might want to pay extra attention when scoring these movements.

I personally would have liked to have seen the same raters score the same subject numerous times to see if the same rater would come up with the same scores each time.  In personal training settings I believe this scenario would be more likely.  Perhaps this could be a route for future investigation.

One final thing I want to mention is that two of the people conducting the study have a personal stake in the outcome since they are involved with the FMS.  This is not to suggest that they would deliberately alter the outcome, but the possibility is always something to be congizant of.  Testing by independent researchers always carries more weight (at least in my mind).

What are your thoughts?