Shoe Compression is Associated with Lower High-Frequency Spectral Power during Drop Landings: An Exploratory Analysis of Six Shoe–Wearer Conditions

Abstract

Time-domain measures of landing impact may fail to capture information in the signal’s frequency spectrum. Four female high-school students performed bilateral drop landings barefoot and while wearing six athletic shoes. For each landing, power spectral density estimates were derived from vertical ground reaction force recordings as well as tibial and sacral accelerometry. The within-participant deviation from barefoot landing was then correlated with the shoe’s peak compression. The relationship was evaluated by frequency and then aggregated into three bands whose boundaries follow the running literature: 0–8 Hz (the movement component), 8–20 Hz (the impact component) and 20–100 Hz (above the impact component). Increased compression was associated with less absolute 20–100 Hz spectral power at the force plate (r = −0.83, leave-one-out [−0.96, −0.77]) and at the sacrum (r = −0.96, [−0.99, −0.88]). The bracketed ranges are the smallest and largest coefficients obtained when each shoe–wearer condition is left out in turn; hence they convey sensitivity and are explicitly not confidence intervals. Movement band power had a positive association with compression, although the force plate coefficient for this band approached zero with the removal of a single wearer-shoe pair. Furthermore, the coefficient for force plate total power reversed sign under leave-one-out, and the two non-saturated sensors yielded opposing directional relationships within the impact band. The most consistent association was a negative association between footwear compression and within-subject changes in 20–100 Hz spectral power. Since absolute 20–100 Hz power changes at the force plate and sacrum covaried closely across footwear conditions (r ≈ 0.8), these two measurement sites appear to reflect a single underlying relationship rather than offering independent corroboration. All findings are exploratory given the small sample size and that each shoe was worn by only one person, so shoe and wearer cannot be separated, and the tibial signal was limited by ±16 g saturation.