Deficit in Emotion Recognition in High Functioning Autism Spectrum Disorder

Background: The ability to recognize emotions is a core deficit of ASD, and difficulty in understanding other people in everyday interaction is widely reported in the individuals with this syndrome. However, the exact nature of the emotional deficits in ASD remains unclear. To date, probably due to the different matching criteria for control group, stimulus type and task demands, no agreement in the investigation of the deficits in recognizing emotional expression in ASD has emerged. Also, the comprehension of social meaning across different communicative channels is still widely unexplored.

Objectives: The aim was to investigate the ability of individuals with high-functioning (HF) ASD in recognizing emotions across a range of different social signals.

Methods: Participants were 20 adults with HF ASD and 20 controls matched for non-verbal IQ, age and gender. They were asked to categorize 24 images of faces, 24 images of whole body, and 30 video clips of Full-light (FLDs) and 30 Point-light display (PLDs) of whole body movements depicting emotional expressions (fearful and happy) or neutral actions. Accuracy and response times (RT) were recorded.

Results: Within group logistic regression analysis showed that, in participants with ASD , the accuracy relative to fearful stimuli was significantly lower in PLDs than in FLDs (p = .01), body images (p = .002) and marginally face (p = .05). Regarding the recognition of happiness, the accuracy in face was higher than that of body images (p = .03), FLDs and PLDs (p = .001). In TD group there was not any difference across emotion neither for face nor for body images recognition. When the two groups were compared, no significant group difference emerged in accuracy in all the classes of stimuli (all p > .2). Repeated measures ANOVAs for the RTs revealed an interaction effect between class of stimuli and emotions (F(2,38) = 4.14, p = .02): fearful faces were recognized by ASD subjects faster than fearful bodies (p = .003) and neutral faces were recognized faster than neutral bodies (p = .001). In TD subjects, no effect of display or emotions were found. Between group comparison showed no group difference in static stimuli (face and body images) (F(1,38) = 0.100, p = .753), while participants with ASD were slower than TD in recognizing dynamic body movements (FLDs and PLDs) (F(1,38) = 5.202, p = .028).

Conclusions: The findings show that individuals with HF ASD are as accurate as TD in recognizing the emotional content of faces and body movements, both static and dynamic. On the contrary, HF ASD were slower than TD in recognizing dynamic body movements, but not static stimuli. Our results are in contrast with prior studies highlighting deficit in face perception in ASD population. Besides, our data reveal a specific difficulty in processing the body movement from dynamic stimuli. Accuracy and RTs did not vary according to the emotional content, in any of display conditions. The ASD impairment in social interaction could be related to the processing of human movement rather than to the recognition of emotion.