The curvature and variability of wrist and arm movements.

The control of wrist rotations is critical for normal upper limb function, yet has received little attention. In this study, we characterized path shape of wrist rotations in order to better understand the biomechanical and neural factors governing their control. Subjects performed step-tracking wrist rotations in eight directions "at a comfortable speed" and "as fast as possible." For comparison, we also analyzed subjects' arm movement paths in a similar task. We found significant differences between wrist and arm movements. Wrist paths were more curved and more variable than arm paths (p < 0.001). The increased curvature and variability can be explained in part by neuromuscular noise (in actuation and sensing) which is known to increase from proximal to distal in the upper limb. The curvature and variability of wrist paths increased with movement speed (p < 0.001), further implicating (signal-dependent) noise. However, noise cannot explain all of our observations. For example, we found that wrist rotations exhibit a systematic pattern: outbound and inbound paths between the same two targets tend to veer to opposite sides of a straight line. We provide evidence indicating that this type of systematic pattern is not likely caused by noise or other neural causes, but may be explained by the unique biomechanics of the wrist.