The old advice is to raise weekly mileage by no more than 10 percent. A cohort study of 5,205 adult runners, tracked through their Garmin watches for 18 months, found that the weekly total was not where the risk showed up. The sessions that tracked with overuse injury were single runs longer than the longest run of the previous 30 days: going more than 10 percent past that distance in one outing was associated with a 64 percent higher injury rate. The study was published in the British Journal of Sports Medicine in August 2025.

How the study worked

The runners averaged about 46 years old and 22 percent were women. Their watches logged 588,071 sessions, and 1,820 of them (35 percent) reported an overuse running injury during follow-up. The researchers compared three ways of measuring a jump in training: the distance of each run against the longest run in the past 30 days, the past week against the three weeks before it (the acute-to-chronic workload ratio), and one week against the previous week.

Each run or week was sorted into four groups: no increase or an increase up to 10 percent (the reference), a small spike of more than 10 up to 30 percent, a moderate spike of more than 30 up to 100 percent, and a large spike of more than 100 percent.

What they found

Single-session spikes carried higher injury rates at every size. Compared with runs that stayed within 10 percent of the recent longest run, the adjusted hazard rate ratio was 1.64 for small spikes, 1.52 for moderate spikes and 2.28 for large spikes, meaning more than twice the rate when a run was more than double the recent longest. The weekly measures did not behave the same way: the week-to-week ratio showed no relationship with injury, and the acute-to-chronic ratio showed an inverse one.

The authors declared no competing interests. As an observational study, it shows an association between long-run spikes and injury, not proof that the spike caused the injury, and injuries were self-reported. It also cannot say much about runners who do not wear a watch or who mix in a lot of other training.

What it means for a training plan

The practical reading is to watch the long run rather than only the weekly total. A runner whose longest outing in the last month was 10 miles would stay at or under 11 miles for the next one, by this study's cutoff. That matters most when building toward a race, when the long run tends to grow faster than anything else in the week, and when coming back after time off, when the last month's longest run may be short.

Most watch platforms already show the longest run over a chosen period, so checking it before a weekend long run takes a few seconds. Plans that add a mile or two to the long run each week are worth comparing against the 10 percent figure at the low end: going from 4 miles to 6 is a 50 percent jump, which lands in this study's moderate-spike group.

One cohort is one cohort, and the 10 percent line here came from how the researchers grouped the data rather than from a trial testing it. What it does is shift the focus from the weekly spreadsheet to one number on the watch.

This article reports research and expert guidance. It is not medical advice. Talk to your clinician before changing medication, treatment or a medical diet.

Sources