Perturbation amplitude affects linearly estimated neuromechanical wrist joint properties

System identification techniques have been used to separate intrinsic muscular and reflexive contributions to joint impedance, which is an essential step in the proper choice of patient specific treatment. These techniques are, however, only well developed for linear systems. Assuming linearity prescribes the neuromuscular system to be perturbed only around predefined operating points. In this study, we test the validity of a commonly used linear model by analyzing the effects of flexion-extension displacement amplitude ($2\circ$, $4\circ$, and $8\circ$) on damping, stiffness, and reflex gain of the wrist joint, at different background torque levels (0, 1, and 2 N$\cdot$ m). With displacement amplitude, intrinsic damping increased, while intrinsic stiffness and reflex gains decreased. These changes were dependent on the level of wrist torque. The dependency of the neuromuscular system properties on even small variations in angular displacement is evident and has to be accounted for when comparing different studies and clinical interpretations using linear identification techniques. Knowledge of the behavior of the neuromuscular system around operating points is an essential step toward the development of nonlinear models that allow for discrimination between patients and controls in a larger range of loading conditions.