In Scandinavian journal of medicine & science in sports ; h5-index 61.0
Running is a popular form of physical activity with a high incidence of running-related injuries. However, the etiology of running-related injuries remains elusive, possibly due to the heterogeneity of movement patterns. The purpose of this study was to investigate whether different clusters existed within a large group of injured and uninjured runners based on their kinetic gait patterns. A sample of 134 injured and uninjured runners were acquired from an existing database and 12 discrete kinetic and spatiotemporal variables which are commonly associated with running injury were extracted from the ground reaction force waveforms. A principal components analysis followed by an unsupervised hierarchical cluster analysis was performed. The results revealed two distinct clusters of runners which were not associated with injury status (OR = 1.14 [0.57, 2.30], X2 = 0.143, p = 0.706) or sex (OR = 1.72 [0.85, 3.49], X2 = 2.3258, p = 0.127). These results suggest that while there appeared to be evidence for two distinct clusters within a large sample of injured and uninjured runners, there is no association between the kinetic variables and running related injuries.
Senevirathna Angela M, Pohl Andrew J, Jordan Matthew J, Edwards W Brent, Ferber Reed
2022-Oct-25
HCA, Injury, Kinetics, PCA, Running, Unsupervised machine learning
