Different mechanisms underlying decision making impairment on Iowa gambling task in patients with eating disorders and healthy young female participants

Abstract

Patients with eating disorders are often reported to show impaired decision making ability on a commonly used decision making task called “Iowa Gambling Task” (IGT). However, the underlying mechanisms of IGT impairments have not been fully understood. The aim of the present study is to investigate the mechanisms underlying impaired IGT performance in patients with eating disorders through the perspective of altered reward and/or punishment sensitivity and explicit and/or implicit learning ability since the existing literature provided limited information in this perspective while reward and punishment sensitivity and learning ability are vital in performing IGT. Reward and punishment sensitivity were examined by using a modified version of IGT (mIGT) and a risky decision making task “Risky Gain Task” (RGT). Explicit and implicit learning ability were tested by the Category Learning Task (CLT). In order to conduct a more valid study, the hypothesized relationships were examined in a healthy young female sample (HC) first in Study 1, then in patients with eating disorders and HC as comparison in Study 2. In Study 1, forty healthy young females participated in the research. A total of 47.5% of them performed poorly on the IGT. Poor IGT performance was negatively correlated with the frequency of choosing risky number 40 after previous punishment outcome (f40_punishment) in the RGT and positively correlated with the sensitivity to punishment score in the Sensitivity to Punishment and Sensitivity to Reward Questionnaire. In Study 2, seventeen patients with eating disorders and twenty HC attended the research. Both groups performed poorly on the IGT. Generally, poor performance on the IGT and mIGT suggested that myopia to the future might be associated with the poor IGT performance in patients. Moreover, the frequency of choosing risky number 80 after previous punishment outcome (f80_punishment) in the RGT and efficient explicit learning ability in CLT were negatively correlated with poor IGT performance. In HC, both hypersensitivity to reward and hyposensitivity to punishment were found to be associated with poor IGT performance. In addition, a sub-group of HC also showed “myopia to the future”. Contrary to patients with eating disorders, poor IGT performance in HC was found to be associated with impaired explicit learning ability. In conclusion, patients with eating disorders do show decision making impairment on the IGT. Myopia to the future, hyposensitivity to punishment, and efficient explicit learning ability are associated with the poor IGT performance. However, since this is the first study to examine the performance of IGT and modified IGT together, and also the first study to examine RGT and CLT performance in patients with eating disorders, future research is needed to validate the findings found here. Moreover, future research may also consider targeting specific type of eating disorder to further explore the relationships observed here. Finally, based on the high rate of poor performers in healthy population, the interpretation of “impaired” decision making ability in healthy population and clinical population should be more cautious and evidence of real life decision making impairment should be included into consideration.