2 - (OP14) Early Response and Activity Engagement in Digital Behavioral Activation with B-ACT

Background: Depression among young adults in Korea is increasing, yet access to evidence-based treatment remains limited. Behavioral activation (BA) is a first-line intervention for depression, but few digital BA programs have delivered core therapeutic components while addressing cultural context. B-ACT is a culturally adapted digital BA intervention designed for Korean young adults. While preliminary findings demonstrated overall clinical benefits, less is known about how and when symptom change occurs during treatment, and how patterns of activity engagement relate to outcomes.

Methods: This single-arm, seven-week pre–post study included 47 young adults with depressive disorders who used the B-ACT app in hospital outpatient clinics and community counseling centers. Primary outcomes included depressive symptoms (PHQ-8, BDI-II, HDRS-17), anxiety (GAD-7), behavioral activation (K-BADS), and quality of life (SF-36). In addition to overall pre–post effects, secondary analyses examined (1) early response (baseline to Week 4 symptom change) as a predictor of post-treatment outcomes and (2) patterns of positive activity engagement. Activity engagement was operationalized as overall activity level (mean) and week-to-week variability, and participants were additionally compared based on relatively consistent versus concentrated (clustered) engagement patterns.

Results: Participants demonstrated significant improvements from baseline to post-treatment in depressive symptoms, anxiety, behavioral activation, and quality of life. Early response significantly predicted post-treatment depressive symptom change across all three depression measures, with the strongest effect for BDI-II (R² = .41), followed by PHQ-8 (R² = .22) and HDRS-17 (R² = .14), indicating that early cognitive–emotional changes were particularly predictive of overall treatment outcome. Comparisons between consistent and concentrated engagement patterns did not yield significant group differences in symptom change across measures. Accordingly, analyses focused on activity level and variability revealed measure-specific patterns: greater mean levels of positive activity were significantly associated with improvement on clinician-rated depression (HDRS-17), whereas activity variability was not independently predictive. In contrast, self-reported depression outcomes (PHQ-8, BDI-II) were more closely related to early symptom response than to cumulative activity levels.

Conclusions: These findings suggest that digital BA may operate through distinct temporal and process-based mechanisms depending on outcome measurement. Early response appears to be a robust indicator of overall treatment success, particularly for self-reported cognitive–emotional symptoms, while cumulative behavioral engagement, rather than engagement pattern per se, is more closely reflected in clinician-rated outcomes. B-ACT shows promise not only in producing symptom reduction but also in elucidating processes of change in digital BA, supporting its relevance for scalable, process-informed CBT interventions.