For decades the standard injury-prevention rule passed around running clubs has been the so-called 10% rule: increase weekly mileage by no more than 10% to stay healthy. A new analysis of the 5,205-runner Garmin-RUNSAFE cohort, now published in the British Journal of Sports Medicine, suggests the unit of measurement runners should be tracking is not the week but the individual session. The paper finds that exceeding 10% of your longest single run from the previous 30 days, in any one workout, is associated with a sharply elevated rate of overuse injury, regardless of how the rest of your training week looks.
The cohort is one of the largest of its kind in running science. Across 588,071 monitored sessions, 35% of participants sustained a running-related injury during the study window. When the authors broke results down by the size of the single-session jump above each runner's 30-day longest, they found a hazard rate ratio of 1.64 for small spikes, 1.52 for moderate spikes and 2.28 for large spikes. In plain terms, doubling the length of your longest recent run on a given day more than doubles your overuse-injury rate over the surrounding period, even if your weekly volume looks otherwise tame.
The finding builds on earlier results from the same Garmin-RUNSAFE programme published last year, which argued that running injuries arrive suddenly rather than accumulating gradually. The new paper sharpens that argument: it identifies the specific training behaviour most associated with that sudden onset and offers a personal, data-driven threshold any runner can compute from their own watch. Importantly, the threshold is relative; a 10% jump for someone whose longest recent run is 20 minutes looks very different from a 10% jump for an ultrarunner whose longest is four hours.
For coaches, the practical implications are straightforward but counterintuitive. Build the long run gradually, rather than skipping forward to a target distance based on race-plan templates, and treat occasional weekend "social" long runs that overshoot recent training as a discrete risk event rather than a free hit. The authors note the rule applies symmetrically to tempo and intervals if those sessions push beyond the runner's recent volume envelope. The result is a simple heuristic that travels well across watches and apps: before any session, check that the planned distance does not exceed 110% of the longest run logged in the prior month.
The research does not invalidate the classical 10% weekly rule so much as relocate it. Risk in runners, the new paper argues, is not a smooth accumulation but a series of discrete spikes, and the spikes that matter are session-by-session. With wearables now logging that data automatically, the takeaway for the average club runner is that the watch on the wrist already has the information needed to flag a high-risk run before it starts. As the authors put it in the conclusion, the goal of injury-aware coaching should shift from managing the week to managing the session.