Risk perception by patients with anxiety disorders.

Studies by social psychologists have recently shown that normal people have systematically distorted perceptions of risk. We extend these studies to patients with anxiety disorders to see how they differ from normal subjects, and, more specifically, to test the hypothesis that anxiety disorders might result from excessively objective risk assessments. Fifty patients with anxiety disorders and 50 matched controls estimated the likelihood that 22 events would happen to the average person or to the self. On both measures, and on accuracy, the groups were essentially identical. In confirmation of previously reported patterns of risk distortion, both groups overestimated rare risks, underestimated common risks, and consistently underestimated the relative risks to self. These results suggest that cognitive estimates of specific risks are normal in patients with anxiety disorders.