妊娠中の歩行関節モーメントの推定における体節慣性パラメータ誤差の感度

Intersegmental mass gains between pregnant individuals poses challenges for estimating the correct segment mass parameters as inputs into biomechanical models for inverse dynamics. This study evaluated whether joint moment calculations are sensitive to physiologically reasonable variability in segment mass distributions during pregnancy. Eleven pregnant participants were assessed longitudinally at gestational weeks 10 and 38. We calculated hip and knee joint moments using inverse dynamics. Five segment mass distribution scenarios were tested by systematically altering the torso, pelvis, thigh, and shank masses, but keeping total body mass constant, as these areas have the most mass gains and variability during pregnancy. Sensitivity of joint moments to anthropometric inputs was quantified for the whole gait cycle and within the gait cycle. Hip joint moments were more sensitive to segmental mass changes compared to the knee. Greater variations were observed during the swing phase across all joints at early and late pregnancy. In early pregnancy, adjustments to torso and shank masses led to the highest deviations in hip and knee moments (maximum RMSE: 1.5 % BW; minimum curve similarities; 0.956; R² ≥ 0.999, p < 0.001). In late pregnancy, sensitivity increased slightly, with hip sagittal moments showing the largest deviations (maximum RMSE: 1.5 % BW; minimum curve similarities: 0.954; R² ≥ 0.999, p < 0.001). But most importantly, all sensitivity metrics were relatively low. These findings indicate errors in anthropometric models within realistic body mass distributions associated with pregnancy should not create significant gait joint moment differences, supporting the use of a general anthropometric model at each gestational timepoint to calculate joint moments.