For many years, runners and coaches have been told that injuries build gradually from accumulated training load, and watch algorithms have been designed around that premise. A large study published in the British Journal of Sports Medicine in July 2025 challenges that view. The Garmin-RUNSAFE Running Health Study followed more than 5,000 recreational runners across 87 countries for 18 months, and its findings remain among the most discussed pieces of recent injury research.
The researchers, from Aarhus University in Denmark and the Luxembourg Institute of Health with other partners, found that weekly workload and the acute-to-chronic workload ratio, long-standing measures used to flag injury risk, showed no clear link with the development of injuries in this cohort. By contrast, injury risk rose sharply as the distance of a single session grew relative to the longest run the athlete had completed in the previous month.
The implication is that many injuries may occur suddenly after one session that goes too far beyond what the body has recently handled, rather than emerging slowly from weeks of overload. That is a notable shift in emphasis, because it moves attention from cumulative totals to the shape of individual sessions. The authors say the results appear to support the familiar 10% guideline against large jumps in mileage, though applied to single runs rather than weekly volume.
There are caveats. The study was sponsored by Garmin, which is relevant given that the findings bear on how wearable companies build their training-load metrics, and it concerns recreational runners rather than elite athletes or those on structured plans. Observational data of this kind can show association but cannot prove that a long run caused a specific injury. Still, the team has released a free algorithm that they hope technology companies will adopt to flag unusually long sessions.
For autumn marathon runners, the practical reading is simple. Peak long runs should be approached through progression, and a one-off, much longer outing, such as a charity run or a day out with faster friends, deserves more caution than a slightly higher weekly total. Coaches may wish to check long-run increments against the previous month’s longest effort rather than relying solely on weekly figures.