Acceptance and commitment therapy for autism spectrum disorder : evaluation of feasibility, effectiveness, and validity of a novel contextual behavioral treatment

Abstract

Background: Autism spectrum disorder (ASD) is neurodevelopmental condition, characterized by challenges in reciprocal social behavior, restricted and repetitive behaviors and interests, and sensory hyper- and hyposensitivity. ASD is associated with executive dysfunction, perceived stress, and psychiatric symptoms, reducing quality of life and adaptive functioning. Acceptance and commitment therapy (ACT) has been proven effective for complex and chronic conditions, although not evaluated in ASD. ACT-consistent instruments, such as Action and Acceptance Questionnaire (AAQ) and Cognitive Fusion Questionnaire (CFQ), measuring psychological inflexibility and cognitive fusion, have not been assessed in autistic individuals. Hence, evaluating ACT and the psychometric properties of AAQ and CFQ in autistic individuals is paramount.

Aims: The overarching aim was to evaluate the feasibility, preliminary effectiveness, and validity of ACT adapted to ASD. Specifically, the aims were to evaluate (1) the feasibility and preliminary effectiveness of group-delivered ACT for autistic adolescents and young adults in a special school setting (Study I), (2) the feasibility and preliminary effectiveness of group-delivered ACT for autistic adults in a psychiatric outpatient setting (Study II and III), and (3) the psychometric properties of AAQ and CFQ in autistic adults (Study IV).

Methods: An adapted ACT protocol (NeuroACT) was evaluated in a quasi-experimental randomized trial (Study I), an open pilot trial (Study II), and a randomized controlled trial (Study III). Study I included 28 ASD adolescents and young adults (13-21 years) utilizing wait-list controls having school classes as usual. Assessments were done at pre, post, and two months follow-up, evaluating self- and teacher-rated stress, self-perceived depression, anxiety, anger, hyperactivity/inattention, prosociality, and conduct and peer problems, analyzed using rmANOV A. Study II included 10 ASD adults (25-65 years), assessing treatment credibility, self- perceived stress and quality of life (primary outcomes), symptoms of depression and anxiety, disability (social; vocational; family), psychological inflexibility, and cognitive fusion at pre, post, and three months follow-up. Data were analyzed using paired t-tests. Study III included 39 ASD adults (21-72 years) randomized to NeuroACT or treatment as usual (TAU), evaluating treatment credibility, self-perceived stress and quality of life (primary outcomes), symptoms of depression and anxiety, sleep problems, disability (social; vocational; family), cognitive and behavioral avoidance, psychological inflexibility, cognitive fusion, autistic core challenges, and executive dysfunction, at pre, post, and six months follow-up, compared to TAU. Data were analyzed using rmANOVA and clinically significant change. Study IV evaluated the construct (convergent and divergent) validity and reliability of AAQ and CFQ in 54 autistic adults (21-72 years) in a psychiatric outpatient setting, using explorative factor analysis and Pearson's correlation coefficient.

Results: In Study I, all participants completed NeuroACT, and treatment satisfaction was high. Overall (pre-post-2-month) statistically significant improvements were found in self- and teacher-rated stress, overall psychiatric symptoms, anger, prosocial behavior, and hyperactivity/inattention (d = 0.70-0.81, 95% CI), and a statistical trend for depression (d = 0.67, 95% CI), in NeuroACT compared to wait-list. No statistically significant interaction effect or statistical trend was found in conduct problems, peer relation problems, or anxiety symptoms. In Study II, 90 % completed NeuroACT, and treatment credibility was high (M=7.7/10, SD = .8). Statistically significant improvements or statistical trends were found (pre-post or pre-3-month) in perceived stress and quality of life (primary outcomes), depressive symptoms, social disability, psychological inflexibility, and cognitive fusion (d = 0.27-0.92, 95% CI). Anxiety or work and family-related disability were not statistically significantly improved. In Study III, 85 % completed NeuroACT and treatment credibility was high (M=7.3/10, SD = 2.5). Overall (pre-post-6-month follow-up) statistically significant improvements or statistical trends were observed in perceived stress and quality of life (primary outcomes), depressive symptoms, sleep quality, cognitive and behavioral avoidance, psychological inflexibility, cognitive fusion, and autistic core challenges related to autistic mannerism (i.e., cognitive and behavioral inflexibility) and social motivation (d = 0.57-1.24, 95% CI) in NeuroACT compared to TAU. Between group clinically significant changes were in favor of NeuroACT. No statistically between group significant change or statistical trends were found in breathing problems, fatigue during daytime, awakening difficulties, social, work, or family-related disability, social awareness, social cognition, communication, or executive dysfunction. Dropout rates were higher in NeuroACT compared to TAU. In study IV, parallel analysis indicated a one-factor solution for AAQ and CFQ. Both instruments showed one-factor solution, explaining 64% of AAQ variance (α = .92) and 67% of CFQ variance (α = .93). Statistically significant positive correlations were found between AAQ and CFQ, and measures of psychiatric symptoms and autistic traits, except social awareness, supporting convergent validity. Statistically significant negative correlations were observed between the AAQ and the CFQ, and quality of life, supporting divergent validity.

Conclusion: ACT adapted to autism is feasible in autistic adolescents and adults and appears to improve stress and mental health. Also, it may help overcome aspects of autistic core challenges. Common instruments to assess ACT are preliminarily valid and reliable for autistic adults. However, more extensive research is needed to further evaluate ACT in ASD. This thesis adds to the growing awareness and empirical support of contextual behavioral models for autistic individuals.