Adapted Cognitive Behavior Therapy to Treat the Anxiety of Children with Autism and Low Verbal Ability

Friday, May 15, 2015: 5:30 PM-7:00 PM
Imperial Ballroom (Grand America Hotel)
J. Danial, UCLA, Woodland Hills, CA
Background:   Research investigating Cognitive Behavioral Therapy (CBT) programs for youth with autism has been limited to participants with average IQ scores. The present study describes results of a second multiple-baseline investigation of an adapted CBT program to treat the anxiety of three youth with autism (ages 10-12) and verbal IQ scores < 70. The first study revealed statistically significant decreases in anxiety across all symptoms for three participants. The program was adapted from the Building Confidence manual, which addresses anxiety in highly verbal youth with autism. Adaptations included using play-based stories with preferred characters, using simplified rating techniques, and equipping participants with mantras as coping strategies. The manual was further refined based on feedback of participating families in the first study and prepared for the current study. 

Objectives:   The current study aims to further support the effectiveness of the adapted manual in treating the anxiety of youth with autism. 

Methods:   A multiple-baseline design was employed across three participants with baseline sessions of 5, 7, and 9 respectively. Three anxiety symptoms for each participant were treated and rated throughout the 16-session program. Following treatment, participants’ parents completed anxiety ratings during a 5-week follow up period. Also, an anxiety interview was conducted before and after the treatment program. Effect sizes will be calculated using multiple recommended approaches for single-case data: standard mean difference, visual inspection, and simulation modeling analysis. Simulation modeling analysis is a computer-based statistical procedure that allows researchers to evaluate the statistical significance of between phase changes (i.e. between baseline and intervention) by generating thousands of data streams to determine the probability of the observed effect size being due to chance (Borckardt et al, 2008). 

Results:   treatment at the end of January. For the first participant, standard mean difference, visual inspection and simulation modeling analysis all revealed decreases in severity for all three of the participant’s anxiety symptoms. Standard mean difference effect sizes ranged from 1.72 to 3.24. Simulation modeling analysis calculations revealed effect sizes of r=.407, p=.0001), (r=.465, p =.08), and (r=.551, p=.0001). Lastly, visual inspection also suggested decreases for all three anxiety symptoms which were maintained during the 5-week follow-up period. The previous study of the manual revealed similar results with standard mean difference effect sizes ranging from 5.2 to 18.0 and simulation modeling analysis coefficents ranging from .522 to .961 at statistically significant levels. The two participants currently enrolled in the program are displaying similar progressions to previous participants who demonstrated decreases in anxiety by the end of their respective programs.


Results of the participant currently finished with treatment and results from the previous study both indicate potential for CBT programs to be effectively adapted to treat anxiety in children with less verbal capabilities. This can be a valuable avenue of treatment to a population who currently has few options for anxiety-focused behavioral interventions.