Predictors of pelvic acceleration during treadmill running across various stride frequency conditions

Pelvic running injuries often require extensive rehabilitation and pelvic girdle pain is a barrier to running engagement in population sub-groups, such as perinatal women. However, exploration into how external pelvic loading may be altered during running is limited. This study assessed which biomechanical variables influence changes in external peak pelvic acceleration during treadmill running, across various stride frequency conditions. Twelve participants (7 female, 5 male) ran (9 km∙h^-1) at their preferred stride frequency, and at ± 5% and ± 10% of their preferred stride frequency. Coordinate and acceleration data were collected using a motion capture system and inertial measurement units. Linear mixed models assessed peak tibial acceleration, displacement from hip to knee and ankle, contact time, and stride frequency as predictors of peak pelvic acceleration. Stride frequency and contact time interacted to predict peak vertical (p = .006) and resultant (p = .009) pelvic acceleration. When modelled, short contact times and low stride frequencies produced higher peak vertical (p = .007) and resultant (p = .016) pelvic accelerations than short contact times and average, or high stride frequencies. Increasing contact time, or increasing stride frequency at shorter contact times, may therefore be useful in reducing pelvic acceleration during treadmill running.