Theories of cognitive therapy have long proposed that those with generalized anxiety disorder (GAD) have inaccurate expectations. By challenging them with objective evidence, symptoms are thought to decrease. To test these premises, this study used ecological momentary assessment (EMA) during the Worry Outcome Journal (WOJ) treatment to determine the percentage of GAD worries that did not come true. We then analyzed the association between participants' untrue worry percentages and GAD symptom change across treatment. Twenty-nine participants with GAD recorded worries when prompted for 10 days, reviewed them online nightly, and tracked their worry outcomes across 30 days. These recordings were then coded by independent raters. Analyses applied bias-correct bootstrapping path analysis on slopes extracted from longitudinal linear mixed models. Primary results revealed that 91.4% of worry predictions did not come true. Higher percentages of untrue worries significantly predicted lower GAD symptoms after treatment, as well as a greater slope of symptom reduction from pre- to post-trial. Participants' average expected likelihoods of worries coming true were much greater than actual observed likelihoods. The most common percentage of untrue worries per person was 100%. Thus, worries in those with GAD were mostly inaccurate. Greater evidence of this inaccuracy predicted greater improvement in treatment. As theorized, disconfirming false expectations may significantly contribute to treatment's effect.
Keywords: cognitive therapy; generalized anxiety disorder; self-monitoring; worry; worry outcome monitoring.
Copyright © 2019. Published by Elsevier Ltd.