43 Effect of dopaminergic medication on risk preference in parkinson's disease. (28th May 2019)

• Record Type: Journal Article
• Authors: Mandali, Alekhya; Chaudhuri, Ray; Rizos, Alexandra; Voon, Valerie
• Is Part Of: Journal of neurology, neurosurgery and psychiatry. Volume 90(2019)Supplement 2
• Page Start: A20
• Page End: A21
• Publication Date: 2019-05-28
• Subjects: Neurology -- Periodicals; Nervous system -- Surgery -- Periodicals; Psychiatry -- Periodicals; 616.8
• DOI: 10.1136/jnnp-2019-BNPA.43
• Languages: English
• ISSNs: 0022-3050

Abstract

Abstract : Introduction: Dopaminergic medication being the standard therapeutic treatment improves motor symptoms in Parkinson's disease (PD) but also implicated in the occurrence of impulse control disorders. Data driven computational models such as drift diffusion model utilize behavioural measures to explain subtle changes that are not sensitive to traditional analysis. Here, we aim to analyse risk preference in PD subjects in OFF and ON medication and the effect of dopamine on risk. Methods: Sixteen patients PD patients during OFF medication and 14 during ON were tested on the 2 step sequential learning task. We calculated the risk associated with each choice (variance of reward probability) and defined the choice with maximum variance as the risky one, for all 134 trials. With behavioural measures (selected choice- risky vs non-risky and response time) as inputs and risk as an independent factor, we extracted threshold ( a ), drift rate ( v ) and response bias ( z ) parameters using a hierarchical drift diffusion model (HDDM) for both groups during ON and OFF drug condition. Statistical analysis on the parameters was analysed using Bayesian factors. Results: Bayesian Independent sample t-test between the 2 groups (ON vs OFF) showed a strong evidence for differences in drift rate (BF10 =34.28) and response bias (BF10 =1.5×10 13 ). We did not observe any evidence for correlation between RL parameters and z for both ON and OFF condition. Behaviourally, with respect to response time, independent sample t-test showed no significance difference between time taken to make risky ( t (28)=−1.28, p=ns) and non-risky choices ( t (28)=−1.06, p=ns). Similarly, no difference was found for change in risky choice selection in presence of the drug ( t (28)=−1.41, p=ns). No differences were found in the traditional reinforcement learning parameters between the groups. Conclusions: Using a novel computational analysis, we showed that dopaminergic medication increased the preference to select a risky choice by modulating drift rate and response bias which was not captured by the behavioural measures. Critically we observe an effect on response bias highlighting the role of apriori information in influencing risky decision making.