Critique of Foam Rolling Research

Audience: Anyone interested in Foam Rolling I recently read a review of some new foam rolling research (and then read the original paper as well) that concluded that foam rolling resulted in increases in knee flexion range of motion without any decreases in other areas of muscular function that are often associated with stretching.  From quick review of the internets this is going to be quite popular.  I was a bit concerned that the conclusions were running far ahead of what the study actually proved so I thought I would provide a critique of the study.

The abstract for the paper can be seen here and the original review of the paper can be seen on Bret Contreras' website here.

Bret had asked me to write a brief review and he has published that review here.  Bret also has a great interview regarding sprint mechanics here.

But for those on my site you can read my review below.

Foam Roller Critique of MacDonald et al (2012)

Study Title" An Acute Bout of Self Myofascial Release Increases Range of Motion Without a Subsequent Decrease in Muscle Activation or Force

I read the recent study by MacDonald et al (part of David Behm’s great research team in Canada and someone I have asked for research advice in the past a great deal) that Bret reviewed on his site.  I think the study is an excellent start into looking into the prevalent practice of foam rolling and I hope that it was spur a lot of other research.  But, I’m now going to give a mini critique.  I don’t want to be a contrary a-hole (although I know that’s how it looks) but I have always thought it was important to question everything we do.  I am even harder on myself.  I will readily admit that I don’t have a lot of definitive answers and this is why I am indebted to researchers that commit the time and effort into these works.


My Bias


Believe it or not I want foam rolling to be effective.  If proven effective, I think would then be a great tool that empowers patients and athletes to do something actively, that is self directed and that they would get benefits from without paying an arm and a leg.  You don’t have to rely on someone else with a foam roller.  I advocate this self reliance for my patients.  Further, the background behind foam rolling is founded on fascial adhesions, thixotropy and tissue health which as a manual therapist doing ART I would benefit from these theories being proven.  There is nothing like the relief you can feel after having resolving some cognitive dissonance.  But, I continue to question.  I want to be convinced soundly and for this I remain on the fence.





Every study design makes choices and when we make choices those choices lead to some limitations.




The control group was not a true therapeutic control group. 


Meaning there was no “sham” intervention given.  The participants all knew that they were being foam rolled and may have also likely known that it was expected that they would have changes in their function.  The control does not control for the fact that something was done to the subjects limb


The warm up was not consistent with standard practice 


This unfortunately would not replicate what is typically done in usual sport.  If you took these athletes through a standard warm up, with dynamic activities and prepared them for sport would you not expect to have an increase in ROM from that?  Would foam rolling provide any greater benefit from what is considered best practice currently?  Again, this is not a knock on the study design, it helps simplify things for their research question.  But, it does limit its immediate relevance to usual practice



The experimenters were not blind to the interventions. 


This is not very relevant for the strength testing but is quite relevant to the ROM assessment.  The ROM assessment used a goniometer that is often quite subjective.  When you use a goniometer you line up the ends on landmarks that you palpate and being off slightly can change the readings you get.  This is amplified when measuring larger limbs.  Further, the experimenters were the ones actively controlling the amount of knee flexion that the subject had.  The experimenters pulled the knee into flexion to the point of pain experienced by the subject.  This is hugely subjective and can be easily manipulated by the tester (consciously or subconsciously).  If you are the experimenter and you know that they just foam rolled it would be easy to subconsciously coax a few more degrees out the subject.  The subject is also being handled by the experimenter and may feel that they can ‘give” more as well – subject knows they were just foam rolled and will let tester pull them a little more.  There is a huge interaction going on here between subject and tester.  I am suggesting that all of this is not at the conscious level of the experimenters.


The ROM testing was too subjective. 


It is possible (although extremely difficult considering the joint that they chose to study) to create a rig that pulls the subjects leg into flexion with the same force each time.  We could also measure ROM with reflective markers on three points (ankle, knee joint and greater trochanter) and measure this digitally.  Having this objectivity is important because there are too many errors that can occur with testing ROM and with such small changes (10 degrees) we want to make sure that there is really a true change not some study artefact.



Discussion regarding self myofascial release and fascial adhesions


Bret really seemed to like this part of the paper.  I thought it was a nice review of the theory but that it was presented like it was fact.  If you notice the references for this section they are all course notes or peer reviewed papers that actually did not study fascial adhesions, fascial properties or thixotropy etc.  For example:


“When fascia loses its elasticity and becomes dehydrated, fascia can bind around the traumatized areas, causing a fibrous adhesion to form. Fibrous adhesions are known to be painful, prevent normal muscle mechanics (i.e. joint range of motion, muscle length, neuromuscular hypertonicity, and decreased strength, endurance and motor coordination) and decrease soft-tissue extensibility (5, 15, 36).”


Again, I would love for this to be true but it is still a theory and not strongly supported. If anyone has any research that really looks at these beliefs I would love to read it.


An apology


You can certainly say to me “hey jackass, why don’t you do the research yourself?”. Yup, this is a good point and it’s why I am not criticizing the researchers but rather just pointing out the normal gaps that occur when we conduct any experiment.  I know the huge amount of thought and work that went into this good study.  It is very easy to critique something and I don’t doubt that everything I have written above the researchers have also already thought about.  That is why we have follow up studies and multiple research teams.  The researchers made choices and this causes certain limitations.  If they made other choices than different limitations crop up.


Bottom Line – What does this mean for practice?


This paper certainly suggests that there is no decrement in performance variables studied (but who knows about dynamic tests or injury) so it is hard to argue against using foam rollers day to day.  However, you can look at this another way, since there was not change in force parameters and the ROM changes could have been an artefact of the experiment, perhaps foam rolling does nothing at all and there is something else out there that we should be spending our time doing before or after sport.  If we just settle on foam rolling are we missing out on something else?


So if you are already foam rolling you will probably continue and if you aren’t foam rolling your muscles maybe you should? I don’t know. How is that for certainty? I look forward to great research to follow.


Thanks for reading.