The Impact of Participant Characteristics on the Effectiveness of Facial Emotion Training in Children with Autism Spectrum Disorders

Background:  Struggling to identify emotions from facial expressions can have a detrimental impact on a child’s social development and future success. Individuals with autism spectrum disorders (ASD) often fail to notice the emotions of those around them. Targeted interventions have been shown to improve the ability of children with ASD to recognize emotions from facial expressions.

Objectives:  The primary aim of this study is to evaluate how participant characteristics impact the effectiveness of facial emotion training in children with ASD. Specifically, we were interested in evaluating how pre-training facial emotion recognition skill, intellectual ability, age or autism severity impact the success of a coach-assisted, computer-based facial emotion training intervention. The intervention included didactic instructional videos of adults demonstrating seven basic emotions (happy, sad, angry, disgust, fear, surprise, and contempt), repeated practice, imitation exercises and a post-session assessment. Screen modifications directing children to important changes in facial features (eyes, eyebrows, and mouth) were also utilized during training.

Methods:  Children with ASD were pre-screened for eligibility based on age, average intellectual ability (WASI-II; IQ ≥ 80), autism status (based on history of diagnosis, SCQ, ADI-R, and ADOS scores) and facial emotion recognition deficit (based on several direct assessments including the MiX, DANVA, and CATS). Once determined eligible, a block randomization design was used to assign participants to either a wait-list control (WLC) or active intervention (AI) group.  Twenty-four children with ASD (12=WLC; 12=AI; ages 8-15 years) participated. Children in the AI group participated in one-hour treatment sessions twice weekly until s/he reached a set criterion. Upon completing training, children in the AI group completed an assessment battery of direct and indirect facial emotion recognition assessments and general social functioning. Participants assigned to the WLC group followed the same assessment schedule, but received the intervention following the completion of outcome measures. ANCOVAs were used to analyze the effects of pre-training facial emotion recognition skill, intellectual ability, age and autism severity on outcomes.

Results:  Children in the AI group improved in their facial emotion recognition. Data showed that the impact of the intervention on post-training facial emotion recognition skill was different at different levels of children’s pre-training skill. Specifically, the intervention had statistically significant effects on post-training MiX scores in participants with lower pre-training MiX (F(1,18)=9.809, p=.006) and DANVA scores (F(1,17)=6.300, p=.022). There were no significant effects of intellectual ability, age or autism severity on the impact of the intervention.

Conclusions:  These data support that the more severe the facial emotion recognition deficit, the greater the impact of facial emotion recognition training. That there were no effects of intelligence, age or autism severity on the results may indicate that children of a broader range of ability may be equally successful after training. Further analyses will be conducted to explore other potential contributors to the success of facial emotion recognition training.