Extreme response style and symptom return after depression treatment: the role of positive extreme responding.

Objective: Evidence linking extreme response style (ER) to depressive relapse has been mixed. One reason might be high levels of extreme responses that are positive in nature (ER-Ps) compared with those negative in nature (ER-Ns) at posttreatment. ER-Ps likely consist of both maladaptive "style" responses and adaptive "content" responses (i.e., legitimate denials of dysfunction). The composition of ER-Ps might confound measures of total extreme responding as well as conventional scores on cognitive questionnaires. In the current study, we assessed ER in a new sample by (a) disambiguating ER-Ps that reflect style from those that reflect content and (b) assessing the contribution of ER-Ps to the prediction of relapse/recurrence.

Method: Responders (N = 104) to a randomized controlled trial of cognitive therapy versus medications for moderate to severe depression had an average age of 40 years (SD = 12), and they were 58% female, 38% married/cohabitating, and 85% Caucasian. ER variables were calculated using the Dysfunctional Attitudes Scale (DAS; Weissman, 1979), with ER-Ps categorized as either content or style responses. ER indices and DAS scores were used to predict symptom return over 2 years.

Results: No standard extreme responding variables (e.g., an index of total extreme responding) predicted symptom return, but higher relative levels of style ER-P predicted relapse/recurrence. Total DAS scores also predicted relapse/recurrence but only when high levels of style ER-P responses were controlled.

Conclusions: ER-Ps, at least on the DAS, appear to contain indicators of both adaptive and maladaptive positive responses. Future research should attend to the valence of the extreme responses as well as to the content of extreme positive responses, which may reflect either healthy or unhealthy tendencies.