Abstract Summary
We developed models to predict peak vertical ground reaction force (vGRF), vertical impulse, and contact time (tc) from running speed and a waist-mounted accelerometer. 25 collegiate cross country athletes ran at 3.8-5.4 m/s while we measured vGRF and sacral accelerations. We low-pass filtered acceleration data with a 5Hz or 10Hz cut-off and compared each models’ predictive capability. Using a 5Hz cut-off better predicted vertical impulse and tc, while a 10Hz cut-off better predicted peak vGRF when the models were used on 4 new athletes. These models predicted peak vGRF, vertical impulse, and tc with 2-4% mean absolute error.