Prediction Errors in Depression: A Quasi-Experimental Analysis.

The overprediction and underprediction of fear, pain, and other constructs have been linked with etiology and maintenance of anxiety disorders. We aimed to extend this research to depression by investigating the overprediction of effort and the underprediction of reward and enjoyment associated with a simple task. Participants with depression ( = 20) and a control group ( = 40) predicted how much effort and reward/enjoyment would be experienced during a short walk around the university. Post-walk, participants reported the actual effort required and the actual experienced reward and enjoyment. Compared to the control group, who demonstrated accurate predictions, depressed individuals significantly overpredicted the required effort associated with the walk even though there were no between-group differences in actual effort required. There was no evidence to suggest that the depressed group underestimated or were inaccurate in their predictions of walk-related reward and enjoyment relative to the control group. Implications for cognitive and behavioral approaches to depression are discussed.