Toward a validation of a tripartite concept of a putative anxious temperament: psychometric data from a French national general medical practice study

Background: Although generalized anxiety disorder (GAD) is currently described as a time-limited state mental disorder, emerging evidence suggests that it is best considered as an exaggeration of a putative “anxious temperament” (AT). It is presently unknown whether it is a distinct or unitary construct of a mélange of anxious traits related to Cluster-C personality disorders. Methods: As part of a Franco-American collaborative study, we developed the 15-item Operational Criteria for Anxious Personality (OCAP), expanding criteria sets developed earlier by one of us (H.S.A.). The study, which was conducted in the French primary care medical sector, included 1112 young adults (18-40 years), seeking help for isolated anxious complaints, never treated before-and without any diagnosable disorder on the axis I of DSM-IV. As previous papers have reported the preliminary validity of OCAP, especially concurrent validity with the State-Trait Anxiety Inventory (STAI) (Speilberger), in this report, we focus on its full psychometric properties. Results: The present data indicate a normal distribution of AT items, a satisfactory Chronbach's coefficient (0.64), and the presence at intake of three different subtypes of AT: “anxious-avoidant,” “anxious-phobic,” and “anxious-sensitive.” After a prospective 6-month follow-up, the major criteria of AT were stable in 80% of cases, and for specific AT items, the stability rate varied between 65% and 80%; much of the unstable items were accounted by improvement during naturalistic treatment. The latter could explain the different factor structure obtained at follow-up, which tended to be less heterogeneous, and represented by one global factor. Limitation: We used a categorical (yes/no) rather than a Likert-type gradation of frequency and intensity of anxiousness items and relatively low number of items, especially for those involving worrying about one's own health or that of one's loved ones. Conclusions: Anxiousness as a temperamental dimension appears to involve putative subtypes along “worrying,” “phobic,” “sensitive” (and “avoidant”) dimensions.