Functional Data Analysis of Running Kinematics in Chronic Achilles Tendon Injury

Abstract

Chronic Achilles tendon (AT) injuries are common, but kinematic studies confirming the exact mechanisms of injury and how orthoses are effective are lacking. Existing analysis often relies on discrete measures and provides an incomplete analysis because many of the data are discarded. Functional data analysis (FDA) views the entire dataset as a function, thus retaining the main features of the curve. This study uses FDA to examine the mechanisms of chronic AT injury and the functional effects of orthoses. Twelve subjects with a history of chronic AT injury and 12 controls ran on a treadmill with and without customized orthoses. Three-dimensional kinematic data were obtained using Qualisys motion capture systems operating at 200 Hz. Ankle dorsiflexion (ADF), knee flexion (KF), eversion (EV), calcaneal, and leg abduction angles were calculated across stance. These angle data were represented as functions, and functional principal components were extracted to describe the factors accounting for variation in the data. These components were compared in AT versus control groups and orthoses versus no-orthoses conditions. Kinematic differences were observed, with the AT group showing greater EV, ADF, and KF during stance, whereas orthoses reduced ADF but increased EV. Different patterns of frontal plane variation distinguished between groups and conditions. Results provided additional information about movement patterns compared to traditional approaches and identified the first half of stance as the most relevant period in injury occurrence. The study showed evidence that variability is related to the presence of injury in this clinical population. Further FDA focusing on within-subject variation is recommended to gain greater insight into the role of variability in injury occurrence.