Group CBT targeting hostile attribution bias in adolescents and young adults with ASD traits

Background: Adolescence is characterized by heightened self-consciousness and sensitivity to social evaluations. During this period, hostile attribution bias-interpreting ambiguous social cues as hostile-can lower quality of life (QOL) and contribute to future mental health problems. Adolescents with autism spectrum disorder (ASD) show similar difficulties, often more pronounced due to their cognitive style and interpersonal vulnerabilities. Group cognitive behavioral therapy (CBT) aims to correct such biases through structured cognitive and social experiences. This study evaluated the psychological effects of group CBT on hostile attribution bias, social functioning, and QOL in adolescents and young adults with ASD traits. Methods: We conducted an 8-session group CBT program focusing on hostile attribution bias and suspiciousness in 21 adolescents and young adults with ASD traits attending a hospital psychiatric outpatient department. The 15 participants who completed the program were included in analyzes. Psychological indices included hostile attribution bias (Ambiguous Intentions Hostility Questionnaire), social functioning (Social Responsiveness Scale, Second Edition [SRS-2]), and subjective QOL. Pre-post changes were quantified as change rates ((post - pre)/pre × 100). Group-level changes were tested with paired analyzes; exploratory associations among change rates were examined using Spearman correlations. Results: Group CBT significantly improved hostile attribution bias (effect size [ES] = 0.698, p = 0.017), social communication and interaction (SRS-2; ES = 0.780, p = 0.012), and subjective QOL (ES = 0.752, p = 0.011). Exploratory individual-level analyzes showed a discordant pattern: smaller reductions in hostile attribution bias (less negative change rates) were associated with greater increases in subjective QOL (ρ = 0.597, p = 0.019). Conclusions: This pilot study suggests that group CBT may reduce hostile attribution bias and improve QOL and social functioning in adolescents and young adults with ASD traits. Notably, the positive correlation between hostile attribution bias change rates and QOL change rates suggests that greater QOL gains were not necessarily accompanied by larger reductions in hostile attribution bias, indicating that improvements in cognitive bias and perceived well-being may arise through partly distinct or non-linear pathways rather than a simple one-to-one relationship. Clinical trial registry: University Hospital Medical Information Network (UMIN000030140).