Running shoes, and the three things the research actually settles

The shoe conversation is unusual in running medicine because the confident version of it is mostly wrong and the evidence-based version is mostly modest. Here is what has actually been tested.

Pronation does not predict injury

This is the idea the whole in-store fitting ritual is built on: your foot rolls in, that is a fault, and a shoe should correct it. It has been tested directly, at scale, in Denmark.

The study's own conclusion is careful: it contradicts the belief for moderate pronation, and the highly pronated group was too small to settle. But the version of the belief being sold to you, that a normally pronating foot needs correcting, did not survive contact with 326,803 kilometres.

Cushioning, and a result nobody expected

The other standard piece of shop advice is that heavier runners need more cushioning. That one has been tested in a randomised trial, which is rare for footwear.

Two things follow. Softer came out better than harder in this trial, which is worth knowing. And the group it helped was the light runners, not the heavy ones, which is the reverse of what you will be told at the counter.

One trial does not settle a field, and both shoes here were prototypes rather than models you can buy. But it is the strongest single piece of evidence on the question, and it does not support the advice built on body weight.

Rotating shoes

The most practical finding in the footwear literature, and one people rarely act on.

This is a cohort study, not a trial, so it shows an association and cannot prove the shoes caused it. The authors' own explanation is the sensible one: different shoes distribute load slightly differently, and so does cross-training, which was protective in the same analysis. Variation in how the load arrives, rather than any one shoe being correct.

So how should you choose

If pronation does not predict injury and shoe category barely changes what your leg does, something has to fill the gap.

That is a proposal rather than a proven mechanism, and the authors present it as such. But it sits alongside a measured finding: when motion control, neutral and cushioned shoes were compared directly, the differences in what the leg did were statistically detectable but small, and they shrank the further up the leg you looked⁵.

The practical version: run in them, not stand in them. Take several pairs seriously, take the ones that disappear on your feet, and treat a shoe that needs breaking in as a shoe that does not fit.

Carbon plates

A separate question, because this one is about performance rather than injury, and here the effect is real.

Two to three percent is a large effect in this field. It is also a performance effect with no injury evidence attached in either direction, and these shoes feel different enough underfoot that dropping straight into a race in a pair you have not run in is a poor idea.

Changing shoes is a change in load

The part that causes most of the trouble, whatever you buy.

A new shoe with a different drop, stack or stiffness changes where the work goes. A lower drop shifts more to the calf and the Achilles. A stiffer forefoot changes what the toes do. None of that is dangerous, and all of it is a change your tissues have to be given time for.

When it is worth asking