US-inflation in a differential odor-conditioning paradigm is not robust: relevance for medically unexplained symptoms.

Reported somatic symptoms without clear relation to physiological processes are studied. A learning paradigm was used with two odors (CSs) and the inhalation of CO(2)-enriched air (US), while measuring symptom levels and respiratory behavior. After paring one odor with the CO(2)-enriched air and the other odor with air, half of the participants received a US-inflation manipulation (information manipulation and enhanced US). Subsequently, all participants received both odors with air (test). A difference between the odor previously paired with CO(2)-enriched air and the odor previously paired with air was found for the symptom measure, US-expectancy ratings, and, to a lesser extent, for respiratory volume. No differences, however, were found between the group receiving the US-inflation manipulation and a group not receiving this manipulation. The results indicate that, despite a successful learning of reported symptoms, US-inflation does not seem to be robust.