Mathematical models that simulate the neuromechanic for maintaining an upright stance are widely used in the research of the dynamics of human movement. A commonly used model is the inverted pendulum with sensory feedback mechanisms that generate corrective torques based on human sway movement. In this paper, the effects on the inverted pendulum model stability are analyzed when additional acceleration feedback and velocity prediction are introduced, based on findings of afferent feedback from muscle spindles from the literature. The robustness of the system was evaluated by comparing the stability regions of the models with and without the addition of the new proprioceptive feedback and prediction. The results showed that the region of stability is considerably larger with acceleration feedback, so that the information on the kinetic state of the leg muscles may have a greater contribution to postural control than the prediction of muscle stretch velocity alone. As the data on acceleration feedback was based on animal data, the demonstration of the power of acceleration feedback obtained in the present work points to the need for future microneurography research on humans to investigate if muscle spindle acceleration feedback indeed occurs with the goal of providing a useful representation of different features.