How to Avoid Running Injuries: What the Evidence Actually Supports

Running-injury advice is full of rules that sound authoritative and have almost nothing behind them. We know, because we published some of them on this page and went back to check. This is the corrected version: what the research supports, what it does not, and where the honest answer is "nobody knows."

First, How Common Are Injuries Really?

You will see "up to 70% of runners get injured every year" everywhere, including in an earlier version of this article. It is not exactly wrong, but it is the top of a very wide range.

A systematic review and meta-analysis found that reported yearly injury incidence ranges from 8.8% to 91.3% across studies, with an overall figure around 26% - roughly 15% in novice runners, 26% in recreational runners, and 63% in competitive runners. Other reviews put recreational runners at 37-56%. The spread is that wide because studies cannot agree on what counts as an "injury": any soreness, or something that stopped you running?

The honest summary: somewhere between a quarter and a half of recreational runners get hurt in a given year, and the number depends more on definitions than on biology.

The 10% Rule Does Not Work

This used to be tip #1 on this page: never increase weekly mileage by more than 10%. It is the single most repeated rule in running, and it has been tested.

Buist and colleagues randomised 532 novice runners into a 13-week graded programme built around a ~10% weekly increase, or a standard 8-week programme that increased volume by ~24% per week. The injury rates came out at 20.8% versus 20.3% - no meaningful difference.

A later systematic review looking at training load and injury put it plainly: no evidence exists for the 10% rule. The same review found that even the broader belief - that sudden training-load spikes cause injuries - rests on very limited evidence.

That does not make ramping up recklessly a good idea. It means nobody has shown that 10% is the magic number, or that load management is the lever we all assumed it was. Anyone selling you a specific percentage is guessing.

Strength Training Is the Strongest Thing You Can Do

Here is what the evidence actually favours, and it used to be buried at #4 on this page.

Lauersen and colleagues meta-analysed 25 randomised controlled trials covering 26,610 participants and 3,464 injuries. Strength training reduced overuse injuries by close to half and acute injuries by around a third - substantially better than stretching or proprioception work, and better than multi-component programmes.

The honest caveat: that meta-analysis covers sport in general, not running specifically, and the effect has not been cleanly replicated in a runners-only trial. It is still the best-supported intervention on this list by a distance.

What to do: squats, single-leg work, calf raises, hip bridges, and core work, two to three times a week. Twenty minutes is enough to matter.

Your Biggest Risk Factor Is Your Own History

The strongest predictor of a running injury is having had one. A prospective cohort study found recreational runners with a previous injury were about twice as likely to get injured again over the following year.

This is useful precisely because it is not something you can train away. If you are coming back from an injury, you are in the high-risk group regardless of how sensible your mileage looks - which is a reason to be conservative and to get the thing properly rehabbed, not a reason to despair.

Run Easy Most of the Time

The 80/20 principle - roughly 80% of your running at an easy, conversational pace - is well established for *performance*, and it is what most coached programmes do. Its injury-prevention case is more of a sensible inference than a proven result: easy running lets you accumulate volume at lower tissue stress.

We are flagging that distinction rather than dressing an inference up as a finding. Heart rate zones are the easiest way to keep easy days genuinely easy, and strength work complements them.

Shoes: The 500-Mile Rule Is Older Than Your Shoes

The familiar "replace at 300-500 miles" traces back to a 1985 study of midsole compression in the EVA foams of the era. Modern midsoles - PEBA, TPU and friends - degrade on different curves, and later work has found midsole hardness rising meaningfully over a few hundred kilometres without landing on any single threshold.

So: it is a ballpark from a different decade, not a rule. Replace shoes when cushioning feels flat or old niggles reappear, and do not treat a mileage counter as a warranty.

Sleep, and the Things We Cannot Cite

Sleep is where tissue repair happens, and 7-9 hours is the standard adult recommendation. We are not going to pretend there is a clean trial showing that a specific sleep duration prevents a specific running injury - there isn't one we would stand behind.

Same for warm-ups. Five minutes of leg swings and high knees before a run is near-universal practice and costs nothing, but the injury-prevention evidence for warm-up routines in runners is thin. Do it because it makes the first mile feel better, not because a study promised you protection.

Where an AI Coach Honestly Helps

An earlier version of this page said sudden load spikes were "the #1 predictor of running injuries" and that Pace Builder could flag them for you. The evidence above does not support that claim, so we removed it.

What a coach - human or AI - can honestly do is keep you consistent, keep your easy days easy, keep strength work actually scheduled instead of aspirational, and adjust when you tell it something hurts. That is real, and it is enough. It is not injury insurance, and nobody who tells you otherwise has the data.

When to See a Doctor, Not an App

If pain persists more than a few days, worsens during a run, changes your gait, or wakes you at night - stop running and see a sports-medicine doctor or physiotherapist. Nothing on this page, and nothing in our app, is a substitute for someone who can examine you.