AUTOBIOGRAPHICAL MEMORY AND CLINICAL ANXIETY

Previous experiments on the recall of threatening and neutral information in clinically anxious subjects have yielded mixed results. The present study assessed autobiographical memory in patients with generalised anxiety disorder (GAD), and normal controls. Subjects were first presented with neutral cue words and asked to respond with any personal memory. GAD subjects judged the content of their own memories as more consistent with anxious mood, and less consistent with pleasant mood, than that reported by controls. Subjects were then again presented with neutral cue words, but with instructions to provide either "anxious" or "nonanxious" personal memories, on specified trials. Significant interactions between group and instruction condition indicated that anxious subjects recalled more anxiety-evoking memories, and recalled them more rapidly, relative to controls. Ratings by blind judges did not support the hypothesis that this apparent difference in the content of autobiographical recall was an artefact due to anxious subjects labelling memories as being more anxiety evoking than would controls. Possible explanations of the findings, and reasons for the apparent discrepancy between these and earlier results, are discussed.