Motion control running shoes do affect the running injury rate

Earlier today I summed up some of the key papers and discussions at the Footwear Biomechanics symposium in Liverpool (here). In that summary I did promise more on this study as it has the potential to put a real spanner in the works, especially in the context of the continued rhetoric and propaganda that we keep hearing.

This particular study was presented at the conference by Laurent Malisoux from the Sports Medicine Research Laboratory at the Luxembourg Institute of Health (the abstract of the paper is here). It was a prospective, double blinded randomized controlled trial of a running shoe with motion control design features vs one with ‘neutral’ design features.

The presentation and the abstract started off on a bit of a cherry picking bad note with the claim that foot ‘overpronation’ was not a risk factor for injury and cited the Nielsen et al 2014 study as evidence for that (not to mention that I had some issues with that study). There are other prospective studies that show the opposite and the two most recent systematic reviews of all the evidence have concluded the same thing, that ‘overpronation’ is a statistically significant but small risk factor for injury (here and here).

The authors did quite rightly point out that there had not been a randomized controlled trial looking at injury rates comparing motion control shoes to neutral shoes. We do have the previous randomized prospective studies of Knapik et al (that has an alleged issue with it) and Ryan et al that showed that the prescription of running shoes based on the ‘overpronation‘ paradigm is not supported. What the study I am reviewing did was they did not take into account when allocating shoes what the foot posture or alignment was (though they did collect that data).

The study recruited 423 experienced runners who were then randomly allocated to one of two groups. Both groups got an identical ‘looking’ running shoe from the one manufacturer, but one group had a shoe that would be considered ‘neutral’ and the other had motion control design features built into it (TPU and dual density midsole). Not a lot more detail, let alone pictures of the shoes were provided. Hopefully that will come when they publish this in full. They were then followed for 6 months and used the allocated shoe for all their running during that time. A running injury was defined for this study as a running related injury that restricted activity for at least one week, and this was reported by the runners in the study to the researchers via a dedicated online platform.

What did they find? 

  • Fifty-six (15%) of the runners experienced an injury during the 6 month follow-up.
  • While the allocation to the footwear group was not based on foot posture, they did, for the purposes of analysis, divide the groups into pronators and neutralers based on a cut-off point of 7 on the Foot Posture Index. On the basis of that criteria, when combining both groups they found that there was no association between injury and foot posture (however, as I discussed previously a lot of clinicians familiar with the Foot Posture Index would consider 7 being too high a cut off point, even though that is based on normative data).
  • They also found the previously commonly reported finding that the main risk factor for an overuse running injury was a previous injury.
  • When they did a Cox regression analysis to look at difference in injury risk between the two groups, they showed that the injury risk was lower in the group wearing the running shoe with the motion control features.
  • A further sub-analysis of the data showed that when they stratified the feet based on the Foot Posture Index criteria that they used, they showed that the only runners with a pronated foot benefited from the running shoe with the motion control features.

So what do we have?
If we accept the results of the Knapik et al and Ryan et al studies that when you allocate running shoes based on foot posture to motion control, stability or neutral shoes you do not affect the injury risk. However, the above study shows that if you do not allocate them to motion control or stability shoes based on foot posture, then you do affect the injury risk and those that benefit the most are those with a pronated foot.

Who saw that coming? Certainly not me!

As always: I go where the evidence takes me until convinced otherwise, and that is what the evidence shows!

Last updated by .

8 Responses to Motion control running shoes do affect the running injury rate

  1. Marc Schwartz July 10, 2015 at 6:51 pm #


    I do not have access to the full paper.


    1. Is there any indication that the frequency of running and/or the weekly distance run between the two groups was different? The runners were only required to run at least once per week, without any other requirements from what I can see.

    2. For the Cox analysis:

    A. Is there any indication that they tested the proportional hazard assumption?

    B. Is there any indication that any of the runners had more than one injury during the study period? If so, did they only consider the first injury as an event? If any of the runners had more than one injury and they counted each, they violate the independence assumptions of a standard Cox model. They would need to consider a frailty model or adjust the covariance matrix to account for the non-independence.

    C. Was there any indication of the median time to injury between the groups?


    • Craig Payne July 10, 2015 at 7:23 pm #

      Sorry Marc, I really can’t answer any of those as they did not provide much or any more than what is in the abstract (its was only a 15min oral presentation and session running late, so very few questions asked of them). Hopefully we get that info in the full publication and hopefully they do that soon. I have emailed the authors to tell them I wrote the above, hopefully they also see your comment and provide that info in the full publication.

      • Marc Schwartz July 10, 2015 at 7:40 pm #

        Thanks for your reply Craig.

        I will await any further information that may be forthcoming.


  2. Eric Fuller July 24, 2015 at 9:01 pm #

    The results are what I would expect. If a more pronated foot type tended to have a more medially deviated STJ axis then a motion control shoe should do better in preventing injury. A more medially deviated STJ axis foot will tend to have a greater pronation moment from the ground. A dual density midsole shoe will tend to shift the center of pressure under the foot more medially, which will decrease the pronation moment from the ground.

    I agree that the study should correlate foot type with shoe. Although it would be harder, the type of injury should also be correlated with shoe. An anti pronation shoe would tend to increase supination moment and this could lead to an increased tendency for inversion sprains, especially in feet that are already tending in that direction. There are injuries associated with “over supination and there are injuries associated with over pronation.

  3. eric johnson January 9, 2016 at 4:18 pm #

    Hey Craig,

    1) Now that the full publication is out, can you add any thoughts to this initial post?

    I’m coming at it from the perspective of a running store manager.

    2) We fit shoes primarily based upon pain, with comfort being the deciding factor. For example, if they come in with achilles/calf pain, we’ll start with higher drop neutral shoes. For knee pain, we’ll start with lower drop shoes. For medial knee pain, we’ll start with stability shoes. If we aren’t successful in finding a comfortable fit with 5-7 styles of the initial class of shoe, we’ll go into the other categories.

    For no pain, we’ll start with a mix of neutral and low drop shoes. We do look at a few structural things, such as varus/valgus knee, arch flexibility, calf mobility and big toe mobility to help in tricky cases where more info might be helpful.

    It’s not realistic for every customer to try on every shoe in the store so we have to use some sort of factor early on to narrow down the selection. I’d be curious to hear your thoughts on the above process.


  4. eric johnson January 18, 2016 at 9:55 pm #

    I also saw that someone on twitter said there was a 30% dropout in the study.

    Does that affect your interpretation in any way?

  5. Adam January 22, 2016 at 1:49 am #

    But the interesting fact is,tha way that the pronation was corrected in these shoes. This is not the standard way in running shoes, to make a firmer midsole under the first metatarsal. Usually it is somewere in the heel area.

  6. eric johnson April 28, 2016 at 10:31 pm #

    I didn’t know that Adam. Did you see a photo of the motion control shoe?

Leave a Reply