Autism spectrum disorder (ASD) presents with impairments in communication and social skills, as well as stereotyped, repetitive patterns of behavior.The hallmark deficits of ASD include disturbances of affective reactivity and the innate ability to perceive and appropriately respond to social cues, including emotional facial expressions. This study utilized event-related potential (ERP) recordings in a modification of a “Theory-of-mind” (ToM) test that studies the ability to correctly deduce another individual’s emotional state, the skill known to be underdeveloped in ASD. Our ToM test used recognition of emotional facial expressions to test the emotional responsiveness in children with autism and in typically developing children (CNT).
The goal of the study was to find the differences between the ASD group (N=19, mean age 16.3 yrs) and CNT group (N=21, 14.9 yrs) by examining behavioral (reaction time and accuracy) and ERP correlates of processing emotional information from facial expressions in a ToM test.
All participants were shown a face and were asked to identify either the gender or the emotion of the face in 4 different conditions. A dense-array EEG was recording the brain activity of the patients as they completed the tasks. The ERPs used in the study (e.g., N170, P3a, and P3b) are evoked responses that represent various stages in cognitive processing. The study also used difference waves, which compare ERP differences in two tasks of variant difficulty levels (gender vs. emotion recognition).
The results showed the percent of errors in the gender condition was higher in the ASD as compared to controls (34% in ASD vs. 20 % in CNT, F1,37= 9.12, p=0.005). Comparing the conditions of gender vs. anger/disgust emotion categorization, the CNT group showed less difference of the errors between the two conditions, while the ASD group committed more errors regardless of condition (F1,31= 6.04, p=0.02). ERP measures also yielded group differences. Occipital N170 showed a more negative amplitude in the ASD group when identifying emotional faces (-3.98 μV in ASD vs. -1.10 μV in CNT, F1,35= 5.66, p=0.023). The latency of parieto-occipital N170 was prolonged in the ASD group (e.g., at P4 site: 211 ms in ASD vs. 158 ms in CNT, F1,32= 7.54, p=0.010).The ASD group had a larger frontal P3a amplitude as compared to controls when differentiating emotions (F1,35= 5.15, p=0.030). In the emotion recognition conditions, centro-parietal P3b revealed a larger amplitude in autism ( F1,36= 4.17, p=0.049). At the frontal sites ( 220-350 ms), the ASD group showed a larger positive difference waves when comparing gender and anger/disgust conditions (3.52 μV in ASD vs. 1.9 μV in CNT, F1,32= 5.64, p=0.024).
These results indicate that more effort is required for an individual with autism to recognize emotion rather than gender from viewing a face. The measurement of ERPs along with behavioral responses provides confirmation that differences exist between how those with autism and those without process emotional faces. Abnormal processing of emotional stimuli may provide an explanation for some of the social and communicative deficits observed in autism.