Modeling equine race surface vertical mechanical behaviors in a musculoskeletal modeling environment

Abstract

Race surfaces have been associated with the incidence of racehorse musculoskeletal injury, the leading cause of racehorse attrition. Optimal race surface mechanical behaviors that minimize injury risk are unknown. Computational models are an economical method to determine optimal mechanical behaviors. Previously developed equine musculoskeletal models utilized ground reaction floor models designed to simulate a stiff, smooth floor appropriate for a human gait laboratory. Our objective was to develop a computational race surface model (two force–displacement functions, one linear and one nonlinear) that reproduced experimental race surface mechanical behaviors for incorporation in equine musculoskeletal models. Soil impact tests were simulated in a musculoskeletal modeling environment and compared to experimental force and displacement data collected during initial and repeat impacts at two racetracks with differing race surfaces – (i) dirt and (ii) synthetic. Best-fit model coefficients (7 total) were compared between surface types and initial and repeat impacts using a mixed model ANCOVA. Model simulation results closely matched empirical force, displacement and velocity data (Mean R2=0.930–0.997). Many model coefficients were statistically different between surface types and impacts. Principal component analysis of model coefficients showed systematic differences based on surface type and impact. In the future, the race surface model may be used in conjunction with previously developed the equine musculoskeletal models to understand the effects of race surface mechanical behaviors on limb dynamics, and determine race surface mechanical behaviors that reduce the incidence of racehorse musculoskeletal injury through modulation of limb dynamics.