Subtalar joint moments of children with cerebral palsy

Children with cerebral palsy (CP) often develop foot deformities [1], such as equinovarus, planovalgus non-midfoot break (PNMFB) and midfoot break (MFB) [2]. The aetiology of the development and progression of these deformities remains unclear. Since they are often associated with abnormal subtalar joint (STJ) anatomy and orientation, altered STJ loading may be important. Personalized musculoskeletal (MSK) modelling employs anatomically accurate bone geometries to define the subtalar axis, including any deformity, and can be used to investigate STJ kinematics and kinetics during gait. How do STJ moments in CP children with equinovarus, PNMFB and MFB deformities compare with typically developing children? Three children demonstrating the three deformities described, were compared with five typically developing controls (1 foot per participant). Weight bearing cone beam CT images were acquired (Multitom Rax, Siemens) and foot bones were semi-automatically segmented (Mimics 24.0 Materialize). Using shape fitting techniques, the bone surfaces were employed to automatically generate personalised skeletal models of the lower limb (STAPLE toolbox [3][4]). Kinematics and ground reaction forces of 6 gait trials per participant were recorded using a 12-camera Vicon system (sample rate: 100 Hz). Externally applied STJ moments were computed in OpenSim with Inverse Dynamics analysis (OpenSim 4.1) using the experimental ground reaction forces and joint angles from the Inverse Kinematics tools. Moments were normalized by weight and temporally normalized over the stance phase and 1D SPM paired t-tests [5] were used to compare CP patients with controls. Mean peak subtalar joint moment of typically developing controls was 0.34±0.10 Nmkg-1 in eversion at 73% of stance phase. Average peak PNMFB STJ moment was 0.29±0.02 Nmkg-1 in eversion occurring earlier in the stance phase at 22% (p<0.001). Average peak STJ moment in the equinovarus foot was 0.17±0.03 Nmkg-1 in inversion, occurring at 84% of stance phase (p<0.001).Download : Download high-res image (145KB)Download : Download full-size image External STJ moments of typically-developing controls acted to evert the foot throughout the majority of stance, agreeing with the literature [3,6]. The greatest difference in pathological STJ moments was observed in equinovarus feet, with moments acting to invert across 78% of the stance phase which can be attributed to the amount of STJ inversion (15° vs 2° in typically developing). PNMFB STJ moments were greater than healthy controls over 0-36% of the stance phase, due to high forefoot loading observed in the experimental ground reaction forces. Unexpectedly, the MFB foot demonstrated no statistically significant differences from typically developing feet, perhaps due to a greater range of motion at the midfoot that our model was not representing. In conclusion, there is a measurable difference between STJ moments in equinovarus, planovalgus and typically developing feet in our case studies, however further analyses of a greater range of CP pathological feet are needed to confirm these differences between groups.