Effects of deep brain stimulation on implicit learning in patients with Parkinson's disease

Abstract

[ACCESS RESTRICTED TO THE UNIVERSITY OF MISSOURI AT REQUEST OF AUTHOR.] This thesis extends previous work examining the effects of deep brain stimulation (DBS) of the subthalamic nucleus (STN) on cognitive functioning by explicitly testing a prominent model (Frank, 2006). In study one, a procedural learning task involving probabilistic classifications was utilized to explore the contributions of the STN to reward learning and motor planning. Event-related potentials (ERP) were recorded from elderly controls (N=16) and patients (N=9) "on" and "off" stimulation. Stimulation was found to speed responses but at the cost of greatly reduced accuracy. Psychophysiological measures indicate an impaired ability to update working memory and use reward related feedback to drive learning during stimulation. In contrast, motor and pre-motor activation improved during DBS. In study two, parallel rule based and information integration tasks were used to examine if DBS impairs accurate response selection thru insufficient time for integration. Mean level analysis of accuracy and reaction time (RT) data collected from elderly controls (N=18) and patients (N=12), indicated that STN stimulation speeded responses at a cost of decreased accuracy. Trial based analysis, however, indicated that RT changes were consistent across the distribution, and that attentional and cognitive flexibility deficits contributed to performance impairments. Therefore, early responding was not responsible for the poor accuracy seen during DBS. In contrast to the prominent Frank (2006) theory, these findings suggest that the motor and cognitive effects of DBS are dissociable and occur thru separate cortico-basal gangliar loops