Do changes in dysfunctional attitudes mediate changes in depression and anxiety in cognitive behavioral therapy?

Abstract

Using structural equation modeling (SEM), four hypotheses about the causal linkages between dysfunctional attitudes (DAs), anxiety, and depression were tested in a group of 521 outpatients treated with cognitive behavioral therapy (CBT) over a 12-week period. The four hypotheses were as follows: (1) changes in DAs lead to changes in depression and anxiety during treatment (the cognitive mediation hypothesis); (2) changes in depression and/or anxiety lead to changes in DAs (the mood activation hypothesis); (3) DAs and negative emotions have reciprocal causal effects on each other (the circular causality hypothesis); and (4) there are no causal links between DAs and emotions—instead, a third variable simultaneously activates DAs, depression and anxiety (the “common cause” hypothesis). Consistent with previous reports, DAs were significantly correlated with levels of depression and anxiety at intake and at 12 weeks; in addition, changes in DAs were significantly correlated with changes in depression and anxiety during treatment. However, the results were inconsistent with the first three hypotheses. There did not appear to be any causal effects linking the DAs with depression or anxiety at intake or at 12 weeks. Instead, the analyses suggested the existence of an unknown variable with simultaneous causal effects on dysfunctional attitudes, depression, and anxiety. This common cause accounted for all the correlations between the attitude and mood variables, and also appeared to mediate the effects of psychotherapy and medication on dysfunctional attitudes, depression, and anxiety.