Running Shoe Recommendations are NOT Based on Science

Below are excerpts from one of my favorite articles of all time, by Steve Magness of Science of Running. He details how running shoe recommendations such as neutral, stability, etc. are not based on science. He also discusses how our understanding of things like cushioning and pronation work are also distorted by propaganda from shoe companies. When I first read this while managing my running store, this rocked my world, and I decided to use the information to help my customers, rather than be afraid of it because it changed our business model and approach. 

Our bodies are smarter and more adaptive than we give them credit for. This is a must read for every runner in my opinion!

Why Running Shoes Do Not Work—Looking at Pronation, Cushioning, Motion Control, and Barefoot Running

By Steve Magness,

The running shoe model needs to be fixed. Pronation, Motion Control, Cushioning, and Stability shoes? Get rid of them all.

It’s not just barefoot running and minimalism versus running shoes, the either/or situation many portray it to be. It’s much deeper than that. It’s not even that running shoe companies are evil and out to make a profit. Shoe companies may be accomplishing the goals they set out for, but maybe the goals their aiming for are not what need to be done. The paradigm that running shoes are built upon is the problem.
Running shoes are built upon two central premises, impact forces and pronation. Their goals are simple, limit impact forces and prevent overprontation. This has led to a classification system based on cushioning, stability, and motion control. The problem is that this system may not have any ground to stand on. Have we been focused on the wrong things for 40+years?
I’ll start with the customary statistic of 33-56% of runners get injured every year (Bruggerman, 2007). That is kind of mind blowing when you think about it. Since there are a ton of injuries going on, let’s look at what shoes are supposed to do.

As said earlier, shoes are built upon the premise that impact forces and pronation are what cause injuries. Pronation, in particular has been constructed as the bane of all runners. We have become inundated with limiting pronation via motion control shoes. The central idea behind pronation is that overpronating causes rotation of the lower leg (i.e. ankle,tibia, knee), putting stress on the joints and therefore leading to injuries. Running shoes are therefore designed to limit this pronation. Essentially, running shoes are developed and designed to put the body in “proper” alignment. But do we really need proper alignment?
This paradigm on pronation relies on two main things: (1) Over-pronation causes injuries and (2) Running shoes can alter pronation.
Looking at the first premise, we can see several studies that do not show a link between pronation and injuries. In an epidemiological study by Wen et al. (1997), he found that lower extremitly alignment was not a major risk factor for marathon runners. In another study by Wen et al. (1998), this time a prospective study, he concluded that “Minor variations in lower extremity alignment do not appear conclusively to be major risk factors for overuse injuries in runners.” Other studies have reached similar conclusions. One by Nigg et al. (2000) showed that foot and ankle movement did not predict injuries in a large group of runners.
If foot movement/pronation does not predict injuries or is not a risk factor for injuries, then one has to question whether the concept is sound or working…
Looking at the second premise, do shoes even modify pronation? Motion control shoes are designed to decrease pronation through a variety of mechanisms. Most choose to insert a medial post or a similar device. In a study by Stacoff (2001), they tested several motion control shoe devices and found that they did not alter pronation and did not change the kinematics of the Tibia or Calcaneus bones either. Similarly, another study by Butler (2007) found that motion control shoes showed no difference in peak pronation when compared to cushioning shoes. Lastly, Dixon (2007) found similar results showing that motion control shoes did not reduce peak eversion (pronation) and didn’t change the concentration of pressure.
This is sort of a double whammy on motion control shoes. If excessive pronation does not cause injuries to the degree that everyone thinks, and if motion control shoes don’t even alter pronation, what’s the point of a motion control shoe?

Impact forces are the other major scoundrel of running injuries. The thinking goes like this, the greater the impact force on the lower the leg, the greater stress the foot/leg takes, which could potentially lead to injuries. To combat this fear, running shoes, particular cushioning ones, are to the rescue. Let’s take a look.
The first question is, do cushioning shoes do their job?
Wegener (2008) tested out the Asics Gel-Nimbus and the Brooks Glycerin to see if they reduced plantar pressure. They found that the shoes did their job!….But where it reduced pressure varied highly—Meaning that pressure reduction varied between forefoot/rearfoot/etc. This led to the interesting conclusion that there should be a shift in prescribing shoes to one based on where plantar pressure is highest for that individual person. It should be noted that this reduction in pressure was based on a comparison to another shoe, a tennis shoe. I’m not sure that this is a good control. Basically, this study tells us that cushioned running shoes decrease peak pressure when compared to a Tennis shoe.
In a review on the subject, Nigg (2000) found that both external and internal impact force peaks were not or barely influenced by the running shoes midsole. This means that the cushioning type does not change impact forces much, if at all. But how can this be? I mean it’s common sense if you jumped on concrete vs. jumped on a shoe foam like surface, the shoe surface is softer right? We’ll come back to this question in a minute.

Underestimating Our Body: Impact forces as feedback:
Back to the question I asked earlier: How can impact forces not change based on shoe sole softness and why isn’t running on hard surfaces lead to more injuries?
The problem is, once again, we underestimate the human body! It’s an amazing thing, and we never give it the credit it deserves. The body adapts to the surface that it’s going to strike, if you give it a chance. The body adapts to both shoe and surface adjusting impact forces via changes joint stiffness, the way the foot strikes, and a concept called muscle tuning.
An example of this can be seen with barefoot running, the diminished proprioception (sensory feedback) of wearing a shoe negates the cushioning of the shoe.  Studies using minimal shoes/barefoot have shown that the body seems to adapt the impact forces/landing based on feedback and feedforward data. When running or landing from a jump, the body takes in all the sensory info, plus prior experiences, and adjusts to protect itself/land optimally.  As mentioned above, it does this through a variety of mechanisms. Thus, you stick some cushioned running shoe on the bottom of your foot and the body goes “Oh, we’re okay, we don’t need to worry about impact as much, we’ve got this soft piece…on our foot.
One concept that needs to be further discussed is muscle tuning. It’s a concept recently proposed by Nigg et al. in 2000. He sees impact force as a signal or a source of feedback, as I stated earlier. The body then uses this information and adjusts accordingly to minimize soft tissue vibration and/or bone vibration. His contention is that impact force is not the problem, but rather the signal. Muscle tuning is essentially controlling these vibrations via a variety of methods. One potential mechanism is pre-activation. Pre-activation is activation of the muscles prior to impact. In this case it serves as a way of muscle tuning to prepare for impact and in addition can alter muscle stiffness, which is another way to prepare for impact. Pre-activation has been established with multiple EMG studies.
Shoes not only impact this, but surface type does too. As mentioned previously, the change in running surface did not impact injury rates. Why? Probably because the body adapts to running surface. In an interesting study measuring muscle activity, O’Flynn(1996) found that pre-activation changed based on surface. To prepare for impact, and presumably to minimize muscle/bone vibration, when running on concrete pre-activation was very high, when running on a soft track, not so much.

What all of this means is that the body adapts via sensory input. It has several different adaptation methods. A shoe influences how it adapts. The shoe is not doing anything to alter cushioning, it is simply altering how the body responds to impact. It’s a significant mindset jump if you think about it. Here’s the summary:
The type of shoe and material of the shoe changes impact NOT because of alignment of the lower leg or because of changes in cushioning. Instead it changes impact characteristics because it alters the sensory feedback
In conclusion on the cushioning concept. Well, what are we trying to cushion? Heel impact forces have not been shown to relate to injuries, in fact in one study low impact runners had a 30% injury rate compared to a 20% injury rate in high impact runners. Shoe midsoles do not change, or marginally change impact forces anyway. So, not only may cushioning not be the answer, the shoes might not even be doing their job. But what about those shoe cushioning studies showing improved cushioning with their new midsole?! Well, the majority of that testing is done by using a machine to simulate the impact forces that you experience during running. That means, yes it may cushion an impact more, but it doesn’t take into account the role of the body adjusting impact based on feedback.
The reason cushioning doesn’t work? Because the body adapts based on feedback and feedforward information. These results prompted one notable researcher(Nigg,2000) to call for the reconsideration of the cushioning paradigm for running shoes.

Barefoot running?
Quickly, this topic could not be complete without a brief mention of barefoot running. An interesting thing to note is that the initial peak impact force is absent in barefoot running when compared to running with shoes. What this means is that, the impact forces look like (A) for (raised heel) shoes and (B) for barefoot. That initial little blip in A is the initial impact force. There is a hypothesis that this initial impact force is related to injuries.

Impact Transient Heel vs Forefoot

A=Heel Strike/Heeled Shoes B=Natural Landing


A recent study by Squadrone et al.(2009) compared running shoes, barefoot running, and running in Vibram Five Fingers. They demonstrated reduced impact forces, shorter ground contact and stride length, but increased stride frequency while running barefoot (and in Vibrams) as compared to running with shoes. This is not unexpected, but shows that running shoes do in fact alter our normal strides. An interesting point is the reduction in stride length but increase in stride frequency. Shoes tend to promote this longer stride at a consequence of ground contact times and frequency. This happens because of changes in feedback signaling, increased likelihood to land on heel stretched out, increased weight, all of which lead to longer times on the ground. It’s interesting to note that elite runners all have short ground contacts and high frequencies (as demonstrated by the often quoted Daniels study of 180 strides per minute).
Tying this to the discussion above on the body controlling things based on sensory information, when running barefoot, there is a higher degree of stiffness in the lower leg. Increased stiffness can result in an increased SSC (stretch shortening cycle) response, resulting in greater force on the subsequent push off (2001). Dalleau et al. demonstrated that pre-activation causing increased stiffness improved Running Economy. In his study, the energy cost of running was related to the stiffness of the lower leg (1998)

Another recent study found that knee flexion torque, knee varus torque, and hip internal rotation torque all were significantly greater in shoes compared to barefoot. What does all of this mean? Potentially, this means more stress on the joints in this area. Jay Dicharry put it best when he said:
“The soft materials in modern running shoes allow a contact style that you would not use barefoot. The foot no longer gets the proprioceptive cues that it gets unshod. The foot naturally accommodates to surfaces rapidly, but a midsole can impair the foot’s ability to react to the ground. This can mute or alter feedback the body gets while running. These factors allow a runner to adopt a gait that causes the elevated forces observed above.”

The one thing that non-barefoot/heel strike proponents use to dismiss midfoot striking/barefoot running is the Achilles tendon. They say, correctly, that the load on the Achilles is higher in midfoot striking runners. The Achilles is meant to take a large load. The problem is we’ve weakened the Achilles through years of wearing shoes with their elevated heels. Essentially, we’ve created the Achilles problem with the shoes meant to prevent it. The Achilles is designed to operate in a rubber band like fashion. . During impact such as the braking or contact phase of running, the achilles tendon stores energy and then subsequent releases that energy via recoil during the take off phase of running. The Achilles, can store and return approximately 35% of its kinetic energy (Ker, 1987). Without this elastic storage and return, the oxygen uptake required would be 30-40% higher! So, in terms of performance why are we trying to minimize the tendonous contribution? It’s like giving away free energy.

Running shoes do not utilize the elastic storage and return as well as barefoot or minimal shoes. More energy is lost with shoes than with barefoot running (Alexander and Bennett, 1989). In addition, in some models of shoes, the arch is not allowed to function like a spring. The arch of the foot can store around 17% of kinetic energy (Ker, 1987). Given these results, its not surprising that running barefoot when compared to running with shoes is more efficient. Several studies have shown a decreased VO2 at the same pace with barefoot running, even when weight is taken into account. This should be no surprise as I mentioned above, without elastic recoil VO2 requirement would be 30-40% higher. Running in a minimal shoe allows for better utilization of this system.

So, the take away message is that shoes change natural mechanics to one that creates mechanical changes that are not optimal for running fast (decreased stride frequency, increased ground contact, decreased stiffness of the system, decreased elastic contribution, and on and on).

Tying it together with elites:
Looking at elite athletes, when racing and training, they generally have higher turnover, minimal ground contact time, and a foot strike that is under their center of gravity. Since the majority of elites exhibit these same characteristics while racing, it makes sense that this is the optimal way to run fast. So, why are we wearing footwear that is designed to increase ground contact, decrease turnover, and promote footstrike out in front of the center of gravity? I have no idea.  (Golden Note: Sounds like an Altra ad to me!)

In conclusion, I’m not some fanatic saying everyone ditch shoes now. Chances are you’ve been running in shoes for 20+ years. Your bodies done some adapting during that time. You’ve got to gradually change if you want to undue some of the changes.

The purpose of this article wasn’t to talk about the benefits of barefoot running. Instead it was to point out the problems with Running Shoe classification. It’s based on a cushioning/pronation paradigm that simply is not as true as they want us to believe. That paradigm needs to be reevaluated. It’s not founded on good science but rather initial ideas that made sense with no science behind them, but upon further review may not stand up to testing. A recent study found that using the good old shoe classification system that everyone uses, had little influence on injury prevention in a large group of Army Basic Training participants (Knapik, 2009). They concluded that selecting shoes based on arch height (like all major running magazines suggest) is not necessary if injury prevention is the goal. I guess that means the systems broken…

Where do we go and how do we fix it? I have no idea. Sorry, no genius answers here. My inclination is that we aim for letting the foot function how it is meant to function, or at least come up with some shoe that may alter foot mechanics but while still allowing feedback/functionality of the body. The first step is looking at the foundation on which running shoes are built upon, the motion control, stability, and cushioning paradigm. My take is that it needs to be reevaluated. I’m going to end with something I’ve already said, but it’s an important concept to get across:

The body is more complicated and smarter than we give it credit.
The type of shoe and material of the shoe changes impact or stride characteristics NOT because of alignment of the lower leg or because of changes in cushioning. Instead it changes impact and stride characteristics because it alters the sensory feedback. The brain is a wonderful thing.’

If you found this article to be informative, I’d appreciate it If you passed it along.  The goal is to get research based data out there so people can be well informed.
-Steve Magness


One comment

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s