(PS4-D77) A Pilot Feasibility Study for Integrated Attention Bias Modification Training for Social Anxiety Disorder: A Randomized Controlled Experiment

Backgrounds: Attention bias (AB) is a crucial cognitive mechanism underlying social anxiety disorder (SAD). Attentional bias modification (ABM) training has been developed to help individuals redirect their attention away from threats, but its effectiveness has been inconsistent in previous studies, primarily due to a limited scope of training, distractions during training, and boredom. To address these limitations, we have designed a more engaging ABM for SAD that incorporates relevant top-down cognitive control processes beyond a mere repetition of threat disengagement, in line with the recent cognitive-motivational framework (Mogg & Bradley, 2018) that emphasizes both goal-directed cognitive control and automatic stimulus-driven processes. This pilot study sought to test a mobile-based integrated ABM (I-ABM) for SAD and examine its feasibility.

Methods: Participants diagnosed with SAD (N=34) were randomly assigned to the I-ABM (N=18) or placebo training (PLT; N=16) group. The I-ABM offers systematic training on integrated processes of threat disengagement, attention control, and response inhibition processes, consisting of four levels of training per session. The PLT offers the identical structure of training, but it did not include any active training elements such as threat disengagement or inhibitory control. Both the I-ABM and PLT trainings were designed using a dot-probe paradigm that includes facial stimuli featuring diverse races and genders. Training sessions were completed on participants' smartphones using Inquisit 6. Participants underwent a baseline assessment, 9 training sessions over 3 weeks, a post-training assessment, and a two-week follow-up assessment. We evaluated the changes in SAD symptoms using the Liebowitz Social Anxiety Scale (LSAS) and Mini-Social Phobia Inventory (Mini-SPIN).

Results: Of the 34 participants enrolled in the study to date, data analysis included those who completed at least six training sessions and the post-training assessment (I-ABM: 8, PLT: 12; mean age=25; female 95%; White 70%; 8.19 training sessions completed on average). A repeated measures ANOVA showed a significant main effect of time on the LSAS total score (F(1,18)=9.33, p=.007, η² =0.342), indicating an overall significant reduction (=13.6%) in social anxiety symptoms in our sample, with no group by time interaction. There was a significant time by group interaction on the Mini-SPIN scores (F(1,18)=5.57, p=0.03, η² =0.236), with the I-ABM group showing significantly lower scores compared to the PLT group at post-training. Specifically, the I-ABM group showed a 22% reduction in symptoms on the Mini-SPIN, whereas the PLT group showed a 7% increase in symptoms.
Conclusion: This study is an innovative attempt to potentiate traditional ABM training with enhanced top-down cognitive process components. The I-ABM outperformed the PLT in reducing social anxiety symptoms on the Mini-SPIN, indicating the potential benefits in ABM training that incorporate high-level cognitive control processes. The training outcomes, as well as the high training completion rate (= 91.7%), show that I-ABM training via personal mobile device is a feasible, accessible, and cost-efficient platform for developing ABM interventions.