Abstract

Single inhalations of a 50% carbon dioxide/50% oxygen gas mixture were administered to 16 subjects with spontaneous panic attacks and to 16 social phobics who did not experience such attacks. Half of each diagnostic category was randomly allocated to either a no explanation condition in which minimal instructions on expected outcome were provided, or an explanation condition in which all possible sensations were described and attributed to the effects of the gas. Subjects with panic attacks who were given no explanation reported a greater proportion of catastrophic cognitions, greater panic, and a greater similarity of the overall experience to a naturally-occurring panic attack than those with panic attacks who received a full explanation. In contrast, both groups of social phobics reported similar effects to each other, regardless of the explanation given. The results provide support for cognitive mediation in the "panic" component of spontaneous panic attacks.