Skip to Main Content
Gravity-induced oscillations of the lower leg in normal and spastic subjects were examined with a view towards evaluating a clinical test of spasticity called the pendulum test. For passive limb motion (in which no reflex excitation occurred), a second-order linear model did not provide an adequate description of the motion for either spastic or normal legs. System equations including nonlinear mechanical properties simulating asymmetries in the swing and amplitude dependent variations in stiffness and damping provided a more accurate description. For spastic limb motion (in which reflex excitation did occur) accurate simulation required components accounting for abnormal reflex activation, coinciding with the time course of EMG activation. These included increased stiffness and damping with their gains related to reflex EMG magnitude, and changes in the rest length of the stiffness. Comparison of numerical with experimental data showed that the nonlinear model simulated the motion accurately, with the variance accounted for usually exceeding 90%.