Predictors of referential thinking: analyses of clinical subjects and controls

Background. Referential thinking (RT) is a common characteristic of human mentation. In psychopathology, RT has been traditionally associated with psychosis. In this study we analyze RT (self-references, SR) differences between clinical and control samples, and we identify variables to predict RT. METHODS. 120 adults (70 patients and 50 control subjects from the general population), with a mean age of 34.49 (SD, 10.63); 60% female. RESULTS. The number of SR among patients, especially patients with diagnosed psychoses, was significantly greater than that among controls. No significant differences in RT were observed among patients characterized by axis II diagnoses or between patients characterized on different axes. The variables that were most predictive for SR were psychotic thinking (MCMI-II personality inventory, thought disorder), conceptual disorganization (BPRS psychiatric scale), age, and vulnerability indicators of mood disturbances (DAS scale). This set of state and trait variables accounted for 56.4% of the RT variance. CONCLUSIONS. There were more differences between patients and controls in terms of frequency of SR than of content (RT processes occur along a continuum). RT may be both a manifestation of state (with an additive effect on other psychopathologies), and a manifestation of trait (a characteristic of psychotic processes, one possibly associated with vulnerability indicators of mood disorders).