A follow-up study of cognitive bias in generalized anxiety disorder.

Patients with generalized anxiety disorder (GAD) without concurrent depression (n = 11) and normal controls (n = 17) were tested twice, about 2 months apart, on a modified Stroop colour-naming task, which presented anxiety-related, depression-related and neutral words in masked and unmasked exposure conditions. GAD patients received cognitive behaviour therapy in the test-retest interval, and were also retested at follow-up, about 20 months after initial testing. GAD patients showed interference in colour-naming negative words across both masked and unmasked conditions before treatment, but not post-treatment, compared with controls. Reduced interference effects of masked threat words over time correlated with reduced ratings of anxious thoughts at post-treatment, and at follow-up, in GAD patients. Thus, the preconscious bias for threat information in GAD appears to vary over time in association with changes in anxious thoughts and worries.