P 256. Augmented set-shifting in Parkinson’s disease with noisy Galvanic Vestibular Stimulation

Abstract

Introduction Cognitive impairment and executive dysfunction, in addition to overt motor deficits, are common in Parkinson’s disease (PD). Perseveration of behavioural paradigms in PD reflects cognitive inflexibility ( Cools, 2001 ) and is related to the inability to stop and switch motor paradigms ( Stoffers, 2001 ). Noisy Galvanic Vestibular Stimulation (GVS) is a technique that has improved motor metrics in PD ( Samoudi et al., xxxx , Yamamoto, xxxx ). Widespread vestibular areas-both cortical and subcortical-are activated by GVS ( Lopez and Blanke, 2011 ), perhaps explaining the positive effects of GVS on PD motor symptoms and also on cognitive effects, such as enhancing visual memory recall ( Wilkinson, 2008 ). Objectives We investigated whether noisy GVS has effects on PD switching abilities in a set-shifting motor task. Unlike other switching paradigms (e.g., Wisconsin Card Sorting Task), our task does not additionally depend on other cognitive components such as working memory and rule learning, thereby predicting a more sensitive measure of switching abilities per se. Materials and methods We tested 4 patients with mild to moderate PD (Hoehn and Yahr stages: 1–2), and 4 healthy age-matched control subjects. All PD patients were asked to stop taking dopaminergic medication at least 12 h prior to the study. Subjects performed 8 trials of a sinusoidal visual tracking task which alternated between 2 modes depending upon whether the displayed cursor position underestimated the actual error by 30% (’Better’) or overestimated by 200% (’Worse’). The switching was covert, unbeknownst to the subject. Either a subthreshold, noisy GVS (0.1–10 Hz, 1/f-type power spectrum) or sham current was applied during each trial in a pseudorandom order. Using Multiple Model Adaptive Estimation (MMAE) ( Oishi, 2010 ), we were able to predict which type of performance mode (‘Better’ or ‘Worse’) the subject was applying to the task. The MMAE algorithm essentially determines the likelihood a subject is behaving in a given ‘Better’ or ‘Worse’ performance mode at specific time stamps throughout a given trial. Results Noisy GVS had no significant effect on the switching detection frequency in control subjects ( Fig. 1 ). PD patients off medication demonstrated notably less switching than control subjects. Patients also responded to GVS delivery with significantly greater frequencies of switching (p = 0.02, Fig. 1 ). Conclusion Noisy GVS increased overall switching behaviour in PD patients without any effect on healthy control subjects, suggesting that Vestibular Stimulation may be able to influence cognitive flexibility deficits in PD. However, further work is needed to see if GVS can, in fact, also result in excessive switching between motor programs.