Effect of gravity-like torque on goal-directed arm movements in microgravity.

Gravitational force level is well-known to influence arm motor control. Specifically, hyper- or microgravity environments drastically change pointing accuracy and kinematics, particularly during initial exposure. These modifications are thought to partly reflect impairment in arm position sense. Here we investigated whether applying normogravitational constraints at joint level during microgravity episodes of parabolic flights could restore movement accuracy equivalent to that observed on Earth. Subjects with eyes closed performed arm reaching movements toward predefined sagittal angular positions in four environment conditions: normogravity, hypergravity, microgravity, and microgravity with elastic bands attached to the arm to mimic gravity-like torque at the shoulder joint. We found that subjects overshot and undershot the target orientations in hypergravity and microgravity, respectively, relative to a normogravity baseline. Strikingly, adding gravity-like torque prior to and during movements performed in microgravity allowed subjects to be as accurate as in normogravity. In the former condition, arm movement kinematics, as notably illustrated by the relative time to peak velocity, were also unchanged relative to normogravity, whereas significant modifications were found in hyper- and microgravity. Overall, these results suggest that arm motor planning and control are tuned with respect to gravitational information issued from joint torque, which presumably enhances arm position sense and activates internal models optimally adapted to the gravitoinertial environment.

Online control of anticipated postural adjustments in step initiation: evidence from behavioral and computational approaches.

Anticipatory postural adjustments (APAs) prior to step execution are thought to be immutable once released. Here we challenge this assumption by testing whether APAs can be modified online if a body perturbation occurs during execution. Two directions of perturbation (resisting and assisting) relative to the body weight transfer were used during the execution of APAs. We found that APAs are modified online (increase in both ground pressure and muscle activity) to compensate for resisting perturbations. The outcomes of a biomechanical model confirmed that the early changes in the APAs resulted from an active control of the APAs and were not merely mechanical consequences of the perturbation. However, no modification of the initial feedforward command was observed for assisting perturbations. The motor command changes for the resisting perturbation may originate from the mismatch between passively originated forces and those actively specified by the central command when acting in the opposite direction. The absence of a mismatch in the assisting perturbation might explain why the central nervous system was not prompted to modify the APAs in this condition.

Can prepared anticipatory postural adjustments be updated by proprioception?

Stepping over an obstacle is preceded by a center of pressure (CoP) shift, termed anticipatory postural adjustments (APAs). It provides an acceleration of the center of mass forward and laterally prior to step initiation. The APAs are characterized in the lateral direction by a force exerted by the moving leg onto the ground, followed by an unloading of the stepping leg and completed by an adjustment corresponding to a slow CoP shift toward the supporting foot. While the importance of sensory information in the setting of the APAs is undisputed, it is currently unknown whether sensory information can also be used online to modify the feedforward command of the APAs. The purpose of this study was to investigate how the CNS modulates the APAs when a modification of proprioceptive information (Ia) occurs before or during the initiation of the stepping movement. We used the vibration of ankle muscles acting in the lateral direction to induce modification of the afferent inflow. Subjects learned to step over an obstacle, eyes closed, in synchrony to a tone signal. When vibration was applied during the initiation of the APAs, no change in the early APAs was observed except in the case of a cutaneous stimulation (low frequency vibration); it is thus possible that the CNS relies less on proprioceptive information during this early phase. Only the final adjustment of the CoP seems to take into account the biased proprioceptive information. When vibration was applied well before the APAs onset, a postural reaction toward the side of the vibration was produced. When subjects voluntarily initiated a step after the postural reaction, the thrust amplitude was set according to the direction of the postural reaction. This suggests that the planned motor command of the APAs can be updated online before they are triggered.