The impacts of loss and risk in decision making of depressed patients and their underlying neural mechanisms

Abstract

Depression is characterized by negative bias in mood, and often results in impairment in decision making. Alteration in the prefrontal cortex and serotonin system has been implicated in depressed patients, particularly suicide attempters. Iowa gambling task (IGT) has been widely used in experimental studies with depressed patients and suicide attempters, but the deficit in decision making was still not well characterized due to complexity in IGT. In addition, neural substrates underlying valuation process in depressed patients were also poorly understood yet. Here, I aimed to characterize impairment in decision making of depressed patients using probability discounting, loss aversion and reversal learning tasks. I also examined the influence of medication, suicidal ideation and suicide attempt history as well as potential correlation between behavioral performance and clinical questionnaire scores. In the fMRI experiment, I aimed to investigate the brain regions related to valuation and choice in decision making under risk with comparison between depressed patients and healthy controls using probability discounting and loss aversion tasks. I hypothesized that depressed patients might have disturbed neural activation characterized by decreased cortical activity and increased subcortical activity. Depressed patients, particularly ones with suicide attempt history, exhibited significantly higher risk aversion and loss aversion than healthy controls. Depressed patients were impaired in both acquisition and reversal stage in reversal learning. Depressed patients tended to less stay with a preceding choice and poorly responded to feedbacks, thus received significantly lower amount of total earning. In summary, three decision making tasks characterized different aspects of decision making deficit in depressed patients which were not completely predicted by conventional questionnaires. In fMRI experiments, Depressed patients tended to recruit brain regions related to ...