There are two rationales for changing the way you run. One is to improve performance by reducing the energy cost of running at any given pace. The other is to reduce injury risk. Scientific research going back decades has consistently shown that when runners intentionally alter their natural running form, they do not become more economical. In fact, they often become less efficient. But a separate thread of research has demonstrated that certain changes in running form do reduce the risk of particular injuries.

A new study published in the International Journal of Sports Medicine is the first experiment I know of that has investigated the effects of altered running form on both efficiency and injury risk in the same group of subjects. Sixteen runners, all of them heel strikers and all suffering from patellofemoral pain (a.k.a. runner’s knee), participated. Half of them were trained to switch from heel to forefoot striking while the others served as a control group. All of the runners were subjected to tests of running economy and were asked to subjectively rate the level of pain in their knees during running. Switching from heel to forefoot striking was found to have no effect on running economy either immediately following gait retraining or after one month of practice, but it did reduce knee pain.

These findings affirm the advice I’ve been giving runners for years: Don’t change the way you run for the sake of improving your running economy and performance. It won’t work. Instead, alter your running form only if you have suffered an injury that was caused by a correctible “flaw” (scare quotes used because it’s only a flaw if it causes an injury) in your running mechanics, the heel striking-runner’s knee link being one example.

Now, every time I make this argument, at least one skeptic counters that a month (or six weeks, or however long it is) is not long enough for the energy-saving effects of a form change to manifest. Runners, they say, need more time to adapt to their new running style. There are two problems with this objection. The first is that there is simply no evidence to support it. Studies lasting as long as 12 weeks have shown that the loss of efficiency resulting from modified running form persist despite continued practice. How long is enough? Six months? A year? What is the basis for the faith that switching to a forefoot striking pattern or reducing stride length or reducing vertical oscillation will eventually pay often when it has never been found to do so?

The other problem with the common objection to my advice on running form is that running form tends to improve over time in all runners. Coaches who do teach “proper” form have told me that in their own long-term testing they have observed improvements in running economy in runners whose form they’ve modified. But what these coaches don’t realize is that these runners would have become more efficient anyway, and in fact it they probably would have improved more if their form had been left alone to evolve naturally.

This is precisely why no runner should change his or her running style for the sake of performance. The human running stride is a self-optimizing system that advances automatically toward maximum efficiency through simple repetition. You can’t make it happen any faster through conscious manipulation. Like growing a beard, you have to just let it happen.