Internet-delivered interpretation training reduces worry and anxiety in individuals with Generalized Anxiety Disorder: a randomized controlled experiment

Objective: Generalized anxiety disorder (GAD) is a debilitating condition, characterized by negative interpretations about ambiguous situations. This study tested whether entirely internet-delivered interpretation training [cognitive bias modification (CBM)] versus control promotes positive interpretations and reduces worry and anxiety in individuals with GAD, with or without depression. Method: A two-arm (CBM; control) parallel-group randomized controlled experiment. Assessments were preintervention (T0), postintervention (T1), 1-month (T2) postintervention, and 3-month (T3) postintervention. Participants with GAD (with or without comorbid depression) were randomly allocated to either CBM (n = 115) or control (n = 115). Participants, but not researchers, were blind to allocated condition. Participants completed up to 10 online CBM or control sessions across 1 month. Interpretation bias [coprimary outcomes: scrambled sentence test (SST), recognition test (RT)], and number of negative thought intrusions during a breathing focus task were measured at T0 and T1. Self-reported levels of worry [Penn State Worry Questionnaire-trait (PSWQ trait); Penn State Worry Questionnaire-past week (PSWQ weekly)], anxiety [Generalized Anxiety Disorder scale (GAD-7)], depression [Patient Health Questionnaire (PHQ-9)], rumination [Ruminative Response Scale (RRS)], and repetitive negative thinking [RNT; Repetitive Thinking Questionnaire-trait (RTQ-trait)] were assessed at T0–T3. Results: The per-protocol analyses included N = 186 participants (CBM n = 94; control n = 92). As predicted, we found moderate-to-large training effects on the primary outcome of interpretation bias at T1. Secondary outcomes of negative thought intrusions at T1 and self-reported symptoms at T2 were all significantly lower in the CBM versus control condition. All but one effect (trait RNT) were sustained at T3. Conclusions: In this randomized controlled study, we found that fully online interpretation training ameliorated core features of GAD in individuals with or without comorbid depression up to 3 months posttraining.