Jeroen P H Verharen1, Unna N Danner2, Sabrina Schröder3, Emmeke Aarts4, Annemarie A van Elburg2, Roger A H Adan5. 1. Brain Center Rudolf Magnus, Department of Translational Neuroscience, University Medical Center Utrecht, Utrecht, The Netherlands. 2. Department of Clinical Psychology, Utrecht University, Utrecht, The Netherlands; Altrecht Eating Disorders Rintveld, Zeist, The Netherlands. 3. Altrecht Eating Disorders Rintveld, Zeist, The Netherlands. 4. Department of Methodology and Statistics, Utrecht University, Utrecht, The Netherlands. 5. Brain Center Rudolf Magnus, Department of Translational Neuroscience, University Medical Center Utrecht, Utrecht, The Netherlands; Altrecht Eating Disorders Rintveld, Zeist, The Netherlands; Institute of Physiology and Neuroscience, Sahlgrenska Academy at the University of Gothenburg, Gothenburg, Sweden. Electronic address: r.a.h.adan@umcutrecht.nl.
Abstract
BACKGROUND: Patients with anorexia nervosa (AN) demonstrate aberrations in choice behavior, including impairments in laboratory measures of decision making. Although a wealth of studies suggest that these aberrations arise from alterations in value processing, it remains unclear by which core component of value processing this is mediated. METHODS: We fit trial-by-trial data of patients with AN (n = 60 first cohort, n = 216 second cohort) and healthy control participants (n = 55) performing the Iowa Gambling Task to a computational model based on prospect utility theory. We determined, per participant, the best-fit model parameters and compared these between the groups. RESULTS: Analyses revealed a decreased estimate of model parameter λ in patients with AN, indicative of an attenuation of loss-aversive behavior in the Iowa Gambling Task. In comparison, measures of reward sensitivity, value-based learning, and exploration versus exploitation were unaltered in patients with AN. A measurement in a second independent cohort replicated the finding that loss aversion, typically observed in healthy individuals, is reduced in patients with AN. CONCLUSIONS: We show that patients with AN, in contrast to healthy control participants, demonstrate reduced loss-aversive behavior. This finding provides important fundamental insights into the decision-making capacity of patients with AN, suggesting alterations in the mechanisms involved in value processing related to negative feedback.
BACKGROUND:Patients with anorexia nervosa (AN) demonstrate aberrations in choice behavior, including impairments in laboratory measures of decision making. Although a wealth of studies suggest that these aberrations arise from alterations in value processing, it remains unclear by which core component of value processing this is mediated. METHODS: We fit trial-by-trial data of patients with AN (n = 60 first cohort, n = 216 second cohort) and healthy control participants (n = 55) performing the Iowa Gambling Task to a computational model based on prospect utility theory. We determined, per participant, the best-fit model parameters and compared these between the groups. RESULTS: Analyses revealed a decreased estimate of model parameter λ in patients with AN, indicative of an attenuation of loss-aversive behavior in the Iowa Gambling Task. In comparison, measures of reward sensitivity, value-based learning, and exploration versus exploitation were unaltered in patients with AN. A measurement in a second independent cohort replicated the finding that loss aversion, typically observed in healthy individuals, is reduced in patients with AN. CONCLUSIONS: We show that patients with AN, in contrast to healthy control participants, demonstrate reduced loss-aversive behavior. This finding provides important fundamental insights into the decision-making capacity of patients with AN, suggesting alterations in the mechanisms involved in value processing related to negative feedback.