A new prospective cohort analysis of 5,205 runners has thrown a sharp question mark over one of running coaching's most reflexive rules. Tracking 588,071 individual training sessions captured through Garmin device data, the study found that 1,820 runners (35 per cent) sustained a running-related overuse injury during the observation period, and that the strongest predictor was not a week-on-week change in volume but a single session that spiked relative to the runner's recent average. The work was published this week in the British Journal of Sports Medicine.
The study defined a "spike" as a session more than 10 per cent longer than the longest run a participant had completed in the previous fortnight. After adjusting for age, body-mass index and prior injury, the relative risk of a new running-related overuse injury within the following 14 days rose by 64 per cent for runners crossing that 10 per cent threshold, and by 127 per cent for those crossing 30 per cent. The week-on-week metrics that have anchored the popular "10 per cent rule" since the 1980s, by contrast, accounted for only a modest fraction of the variance once single-session spikes were included in the model.
The cohort skewed older and more recreational than typical training-load research samples — mean age 45.8 years, with women making up 22 per cent of the population — and the authors are careful to note that the findings should not be over-extrapolated to elite or sub-elite runners whose absolute weekly volumes are an order of magnitude higher. But the data set is unusually clean for a sport that has historically relied on retrospective questionnaires: every spike, every recovery day and every reported injury was time-stamped within the same Garmin Connect platform, removing the recall bias that plagues most observational running research.
For coaches, the practical reading is straightforward. The "long run" remains the place where most overuse injuries are seeded, and the difference between this Saturday's long run and the longest run of the previous two weeks is a more useful number than a generic 10 per cent week-on-week target. Several coaching education bodies, including UK Athletics' endurance pathway, are already understood to be considering whether the long-standing 10 per cent rule should be reformulated around longest-session change rather than weekly mileage change.
The study sits alongside a separate VCU-Harvard analysis, covered on Running Lookout last week, that found mid-pack marathoners who reduced weekly run frequency in the final months of training finished an average of three minutes faster than peers who held volume steady. Together the two papers point to a slow shift in evidence-based recreational running training: the absolute weekly mileage number matters less than the shape of the training week and the size of the jumps within it.