Footwear and Flatfeet: Correlation or Causation?

One of the arguments that often get advanced for a barefoot lifestyle is that shoes cause flat feet as they weaken the muscles. Not sure how people make that conclusion as there is no evidence that footwear wearing populations have feet that is any weaker than barefoot wearing populations – you would have thought that if such claims were being made, those making it would have something to back it up. Just stating it and wishing it was true is a logical fallacy. It may or may not be true, but if you want to assert it, then come up with some data to support it (and I addressed the issue about the propaganda on running shoes weakening muscles here).

There is a group of studies that often get cited to “prove” that shoes cause flat feet and this is one of the studies (I was reminded of this study this AM on Facebook, so thought I would write about it – thanks PG):

The influence of footwear on the prevalence of flat foot. A survey of 2300 children.
Rao UB, Joseph B.
J Bone Joint Surg Br. 1992 Jul;74(4):525-7.
We analysed static footprints of 2300 children between the ages of four and 13 years to establish the influence of footwear on the prevalence of flat foot. The incidence among children who used footwear was 8.6% compared with 2.8% in those who did not (p less than 0.001). Significant differences between the predominance in shod and unshod children were noted in all age groups, most marked in those with generalised ligament laxity. Flat foot was most common in children who wore closed-toe shoes, less common in those who wore sandals or slippers, and least in the unshod. Our findings suggest that shoe-wearing in early childhood is detrimental to the development of a normal longitudinal arch.

There are a couple of other similar studies that found something similar. You regularly see them popping up in debates about footwear use and in barefoot lifestyle groups. The real issue with the study is correlation is NOT causation. The above study is wrongly titled based on what the data showed. It should not have been “Influence”; rather it should have been “Relationship” or “Correlation” between “Flatfoot and Shoes”.

Have a look at this graph on a nearly perfect correlation:
There are a whole lot more of these types of spurious correlations at (a lot of them are very funny).

So is the correlation reported in the above study a possible causal relationship or is it just a spurious correlation? As least there is a theoretical mechanism of of a causal relationship (ie the shoe really could weaken the muscles leading to the flat foot), but just because there is a correlation and a potential mechanism does not prove causation. Those with a particular agenda to promote will want to believe it, but again wanting to believe something does not make it true.

What does the data actually support? It could be any one of these explanations:

  1. Yes the shoes do weaken the muscles and cause the flatfoot (but the evidence from a couple of studies that show no correlation between arch height or overpronation and muscle strength would tend not to support this interpretation).
  2. Those who who have a flatfoot wear shoes more often as they feel more comfortable wearing them (which has nothing to do with the shoes causing the problem and hence the correlation).
  3. Another factor, for eg, hard surfaces could cause the flatfoot and perhaps those who spend more time on hard surfaces are more likely to wear shoes (which has nothing to do with the shoes causing the problem and hence the correlation).
  4. Another unknown factor might cause the flatfoot and increase the likelihood that the person wears shoes  (which has nothing to do with the shoes causing the problem and hence the correlation).

Studies like the above one could be interpreted as supporting any one of those possible explanations. It is simply not possible to conclude it is more likely to be (1) above any of the other possible explanations as a correlation is simply not causation. It could be entirely possible that it is (1), but the above study does not show that (despite the authors claiming that it does; and the muscle strength/arch height studies not supporting it). I am not going to make a conclusion or give any more weight to one of those possible interpretations over another until there is some causal data supporting one interpretation over another.

I only raise all this as you often see this study and similar ones being cited as evidence that we should not wear shoes! It does not show or say that at all. I did do a quick check of the websites that were promoting that interpretation and noticed that they were all silent on this more recent study:

Use of footwear and foot condition among rural Ethiopian school children
Emi Watanabe, Colleen M. McBride, Abebayehu Tora, Desta A. Ayode, David Farrell, Gail Davey
Journal of Epidemiology and Global Health; Article in Press
To evaluate whether shoe-wearing affords foot protection among school children living in southern Ethiopia.
Data collectors conducted a standardized foot assessment with children in an elementary school in southern Ethiopia (N=168).
54% reported wearing shoes consistently in the prior three days. Children wearing closed-toed shoes showed less adherent soil and toe nail dystrophy than those wearing open-toed sandals. There were no differences by shoe type with regard to signs of foot trauma or heel fissures.
Shoe wearing provided limited foot protection. Interventions are needed to build behavioral skills, including foot washing and wearing appropriate shoes that maximize foot protection.

This study showed that there was more foot trauma related injuries in the non-shoe wearing population. So even it the interpretation above of the shoes do actually cause the flatfoot, that has to be traded off with the increased risk of trauma related injuries. Six of one, half a dozen of the other.

As always, I go where the evidence takes me until convinced otherwise, and correlation is not causation. Correlation generates useful data to formulate hypotheses that might warrant further investigation.

Watanabe, E., McBride, C., Tora, A., Ayode, D., Farrell, D., & Davey, G. (2014). Use of footwear and foot condition among rural Ethiopian school children Journal of Epidemiology and Global Health DOI: 10.1016/j.jegh.2014.06.001

Rao UB, & Joseph B (1992). The influence of footwear on the prevalence of flat foot. A survey of 2300 children. The Journal of bone and joint surgery. British volume, 74 (4), 525-7 PMID: 1624509

Last updated by .


3 Responses to Footwear and Flatfeet: Correlation or Causation?

  1. Roman R. October 22, 2014 at 1:06 pm #

    Thanks for the study and the nice correlation example 🙂
    Like it alot…
    I guess every footwear affects the body differently with pros and cons.
    Running barefoot you might be more susceptible to traumas, while running in cushioned shoes with shadowed feedback might tempt to overstride.
    I once stubled over a study where they compared the postural control of alpine skiers which compete on different levels (national and regional). Since a good postural control is needed for alpine skiing, one would think that better skiers have a superior postural control. But the opposite was the case (at least without skiboots). They concluded that the extended time which the national skiers spent in their boots influenced the postural ability without them.
    This might be an extreme example but actually the same could be the case on a smaller scale with other shoes.
    A cast can impair the movement of the foot/leg, muscels and sensory systems degenerate. Shoes also influence the foot motion which certainly must have some effect even if they yet have to be proven by a study.

    Here is the Skiers study if someone is interested:
    F. Noé, T. Paillard (2005). Is postural control affected by expertise in alpine skiing? Br J Sports Med 2005;39:835–837. doi: 10.1136/bjsm.2005.018127

    • Craig Payne October 22, 2014 at 6:04 pm #

      Barefoot v shoes –> use different muscles differently; theoretically some might get RELATIVELY weaker and some might get RELATIVELY stronger .
      barefoot tend to forefoot strike –> use tib ant muscle less –> might get RELATIVELY weaker
      shoes tend to heel strike –> use tib ant more –> might get RELATIVELY stronger

      BUT; as runners run more than sedentary –> all muscles are going to be stronger regardless if wear shoes or not as they using ALL muscles more –> no weakness.

  2. Margaret Dunstone November 2, 2014 at 7:28 am #

    Good for a laugh. I really like cheese incl mozzarella. Should I go and get a doctorate? Does eating pizza,give me a better chance of passing,and getting a doctorate? Margaret Dunstone.

Leave a Reply