A meta-analysis that included data from 24 observational studies (7 cohort studies, 5 cross-sectional studies and 12 case–control studies) and a total of 20,997 individuals has shown that the overall risk of knee OA is significantly increased in those with a history of knee injury (odds ratio 4.20, 95% CI 3.11–5.66). Study design and definition of knee injuries did not affect the significance of the association.