Site-specific Concurrent Validity of the ActiGraph GT9X Link in the Estimation of Activity-related Skeletal Loading

The aims of this project were twofold: 1) to assess the concurrent validity of raw accelerometer outputs with ground reaction forces (GRF) and loading rates (LR) calculated from force plate across a range of simulated habitual PA and 2) to identify the optimal wear site among the ankle, hip, and wrist with the strongest relationships between accelerometer and force plate and/or skeletal outcomes.

Thirty healthy young adults (23.0 ± 4.5 yr, 50% female) wore a triaxial accelerometer at the right ankle, hip, and wrist while performing eight trials of walking, jogging, running, low box drops, and high box drops over an in-ground force plate. Repeated-measures correlations and linear mixed models were used to assess concurrent validity of accelerometer and force plate outcomes across wear sites.

Strong repeated-measures associations were observed between peak hip resultant acceleration and resultant LR (rrm 1169 = 0.74, P < 0.001, 95% confidence interval = 0.718, 0.769) and peak hip resultant accelerations and resultant GRF (rrm 1169 = 0.69, P < 0.001, 95% confidence interval = 0.660, 0.720) when data were combined across activities. By contrast, small to moderate associations were seen between ankle-based outcomes and corresponding GRF and LR during walking and jogging (rrm 209 = 0.17-0.34, all P < 0.001). No significant associations were seen with wrist-based outcomes during any activity. In addition, linear mixed models suggested that 24%-50% of the variability in peak GRF and LR could be attributed to measured accelerations at the hip.

Peak accelerations measured at the hip were identified as the strongest proxies for skeletal loading assessed via force plate.