Matching the shoe to the foot and injury rates in female half marathon runners

In this post I’ll be looking at a paper titled ‘The effect of three different levels of foot wear stability on pain outcomes in women runners: a randomised control trial’, which was written by Michael Ryan, Gordon A Valiant, Kymberly McDonald and Jack E Taunton.

This study divided 81 female runners according to their foot types:  neutral, pronated and highly pronated. It then randomly assigned these runners with a pair of neutral (least amount of support), stability or motion control (greatest amount of support) running shoes. Conventionally a neutral foot would be prescribed with the neutral shoe, pronated with the stability and highly pronated with the motion control. Once supplied with a shoe the women underwent a 13 week half marathon training protocol. The number of injuries and training days missed due to injury were recorded as were pain scores during and after running.

The paper reported that 32% of the women sustained injuries that led to a total of 194 missed training days.  The paper reported there were 51 missed training days by those wearing the stability shoe, 64 by those wearing the neutral shoe and 79 by those wearing the motion control shoes.  It was also found that those wearing the motion control shoe reported greater levels of pain throughout the study and that the neutral shoe had greater levels of pain during running that the stability shoe.

So straight up we can see a high prevalence of running injuries that are sustained by those who run in shoes. Now we need to dive a little deeper and analyze the effectiveness of fitting a shoe type to a foot type. That’s what the fancy footwear shops do when they get you to step on the pad and measure how much your foot rolls in or out.

There were 39 runners with a neutral foot and they were subdivided into three groups: 12 to run in the neutral shoe, 17 in the stability shoe and 10 in the motion control shoe. Of the runners who were assigned to the neutral shoe (i.e. the shoe that in theory would provide the best match) 33.3% sustained injuries, where as 11.8% of runners who were assigned the stability shoe sustained injuries and 40% who were assigned the motion control shoes sustained injuries.

There were 30 runners who were classified as having a pronated foot and they were sub-divided into the three different groups: 8 in the neutral, 13 in the stability (proposed best fit), and 9 in the motion control group. 25% of those who had been assigned the neutral shoes had injuries, 23.1% in the stability group were injured and 44.4% in the motion control group were injured.

The study was only able to amass 12 runners who were classified as highly pronated. They were sub-divided into a group of 5 with neutral shoes, a group of 2 with stability shoe and a group of 5 with the motion control shoe (supposed best fit for this group). 40% of those who ran in the neutral shoe, 0% in the stability shoe and 100% of those with motion control shoe sustained injuries. The fact that there were so few runners in these groups is a major flaw in this study! More on that later.

So what can we draw out from these numbers? Well straight of the bat we can see that there is high incidence of injuries that are associated with women running in shoes. Secondly if we are to consider the conventional method of matching foot type to shoe type only those with the pronated foot and the stability shoe had positive findings. Those findings however are so minimal that they can’t be considered significant. Despite the significant short comings in the design of this study we can clearly see that motion control shoes led to the greatest number of injuries. These findings are consistent with the finding of Craig Richards paper so despite the weakness of the study design we can still infer with reasonable confidence that motion control shoes are potentially hazardous to runners (particularly female!). We can also see that there is no apparent benefit (it may even be harmful) of measuring ones foot posture and matching it with corresponding shoe.

Now it’s important to note that this paper has some very significant floors. Firstly there is no control group!!! Why they didn’t have a group running in their normal running shoes is beyond me. I would love to have seen what injury rates would have been for those who ran in their normal shoes. Secondly there were only 12 runners who were classified as having highly-pronated feet. Having a group so much smaller than the other groups’ means it’s very difficult to compare between group findings and make serious definitive conclusions. This is why I have simply presented the most basic results with which we can make reasonable assumptions, particularly when it’s considered in the context of other studies.

Another interesting side note is that this study was funded by Nike global. Thus we may question whether these results and findings have been watered down so as to not put people off buying their shoes?

So take home message? Distance running in shoes is associated with significant injury rates and the conventional method of fitting the correct shoe to foot shape is not working. We need an alternative approach. Lucky we have one in barefoot running!


By Sam

BM in Sport and Exercise and BA in International studies (German Major) completed at the University of Technology, Sydney.
Master of Physiotherapy at the University of Sydney.

Interests: I am an exercise enthusiast full stop! I play or have played football, tennis, basketball and dabbled in waterpolo and underwater rugby! Recent sporting interests are cycling and barefoot running! I also love watching all codes of football/rugby.

Physiotherapy: Due to my interests in sport and health I am fascinated in human movement and physiology. In addition to musculoskelatal physiotherapy I’m also interested in neurological (stroke, brain injury and spinal cord injury rehab), and cardiopulmonary (heart and lung function and rehabilitation) physiotherapy:

Join the Conversation

Your email address will not be published. Required fields are marked *