Implicit category learning performance predicts rate of cognitive decline in nondemented patients with Parkinson's disease.

Nondemented patients with Parkinson's disease (PD) are impaired in learning to categorize simple perceptual stimuli when category membership is defined by a nonlinear relationship between stimulus dimensions but not when the relationship is linear (J. V. Filoteo, W. T. Maddox, D. P. Salmon, & D. D. Song, 2005). In the present study, the authors examined whether performance in either of these 2 category learning conditions was predictive of global cognitive decline following a mean of 1.6 years since the time patients were 1st seen. Results indicated that final block accuracy in the nonlinear condition, but not the linear condition, predicted global cognitive decline. Performance on the Wisconsin Card Sorting Test (WCST) did not significantly predict global cognitive decline, although there was a trend for this to be the case. In addition, the association between nonlinear category learning and global cognitive decline was not impacted by patients' performance on the WCST. Results suggest that nonlinear category learning predicts cognitive decline in nondemented patients with PD and that nonlinear category learning and WCST performances may provide independent measures of integrity of the posterior and anterior caudate, respectively.