What we found

• In a randomised trial of 532 novice runners preparing for a 4-mile (6.7 km) event, a graded 13-week programme built on the 10% rule had 20.8% running-related injuries and a standard 8-week programme had 20.3% (P = .90): the 10% rule programme did not prevent injury (Buist and colleagues, 2008, American Journal of Sports Medicine).
• The difference between the two injury rates is 0.5 percentage points (20.8% minus 20.3%), and it points to slightly more injuries on the 10% programme, not fewer.
• In a one-year cohort of 874 healthy novice runners followed with a GPS watch, 202 sustained a running-related injury (23.1%). Runners were grouped by weekly distance progression: under 10% or a reduction, 10% to 30%, or over 30%. Overall injury rates did not differ significantly between the three groups (Nielsen and colleagues, 2014, Journal of Orthopaedic and Sports Physical Therapy).
• For distance-related injuries only, such as patellofemoral pain and iliotibial band syndrome, runners who progressed by more than 30% had a hazard ratio of 1.59 against runners who progressed under 10% (95% confidence interval 0.96 to 2.66, P = .07): a possible increase, not a statistically significant one. The authors call the study exploratory and ask for randomised trials.
• Both studies are of novice runners, people starting to run. Neither tested experienced runners with high weekly distances.
• A figure often quoted from a one-year cohort of 76 runners is that over 60% of injured men and over 50% of injured women had raised weekly distance by more than 30% at least once in the four weeks before injury. It counts injured runners only; without the share of uninjured runners who did the same, it cannot show that big jumps raise injury risk.