Pointing to remembered targets in 3-D space in Parkinson's disease.

A three-dimensional tracking system was used to examine whether subjects with Parkinson's disease (PD) would show characteristic performance deficits in an unconstrained pointing task. Five targets were presented in a pyramidal array in space to 11 individuals with mild to moderate PD and 8 age-matched controls. After the target was indicated, subjects closed their eyes and pointed to the remembered target locations without vision. Despite the absence of visual feedback during movement, PD subjects were as accurate overall as controls. However, PD subjects showed greater variable errors, more irregular trajectories, and a vertical endpoint bias in which their endpoints were significantly lower than controls. They also showed deficiencies in the compensatory organization of joint rotations to ensure consistency in azimuthal (horizontal) positioning of the arm endpoint. We concluded that, under appropriate task conditions, PD subjects may not show overall deficits in accuracy even when making targeting movements at normal speed without visual feedback. Nevertheless, our findings indicate that there are certain dimensions of performance which are selectively altered in Parkinson's disease even when overall performance is normal.

The interaction of visual and proprioceptive inputs in pointing to actual and remembered targets.

Errors in pointing to actual and remembered targets presented in three-dimensional (3D) space in a dark room were studied under various conditions of visual feedback. During their movements, subjects either had no vision of their arms or of the target, vision of the target but not of their arms, vision of a light-emitting diode (LED) on their moving index fingertip but not of the target, or vision of an LED on their moving index fingertip and of the target. Errors depended critically upon feedback condition. 3D errors were largest for movements to remembered targets without visual feedback, diminished with vision of the moving fingertip, and diminished further with vision of the target and vision of the finger and the target. Moreover, the different conditions differentially influenced the radial distance, azimuth, and elevation errors, indicating that subjects control motion along all three axes relatively independently. The pattern of errors suggest that the neural systems that mediate processing of actual versus remembered targets may have different capacities for integrating visual and proprioceptive information in order to program spatially directed arm movements.