Publication Date



Deficits in muscles of the lumbo‐pelvic region, such as a relatively small multifidus muscle, have been used to predict lower limb injuries in professional football players. Results have been less consistent for the size of the quadratus lumborum (QL) muscle. Changes in size of the multifidus and QL muscles could be functionally related to each other, and modeling this relationship could improve prediction of lower limb injuries. Ultrasound imaging examinations were performed on male elite football players at the start of the Australian Football League (AFL) pre‐season and playing season. Injury data were obtained from records collected by each club. Results indicated that the cross‐sectional area of the multifidus muscle was related to the occurrence of an injury in the pre‐season (odds ratio [OR] = 2.08/cm2decrease below the mean; OR for dichotomized measure = 12.2) and in the season (OR = 2.43/cm2). The size of the QL muscle was significantly related to an injury in the pre‐season (OR = 2.12/cm2 increase above the mean; OR for dichotomized measure = 7.26) but not in the season. A significant link was found between the ratio of the multifidus and QL muscles, and the incidence of pre‐season (OR = 14.71) and season injuries (OR = 5.29). The sensitivity and specificity of the model in the pre‐season were 75% and 85.7%, respectively; values for the playing season were 88.4% and 62.5%. A model was developed for prediction of lower limb injuries in football players. Combining size measurements of the multifidus and QL muscles improved predictive power. This information may have clinical implications for injury screening and prevention.


Mary MacKillop Institute for Health Research

Document Type

Journal Article

Access Rights

ERA Access

Access may be restricted.