Impaired Face but Not Object Recognition in Young Children with Autism Spectrum Disorders

Background:   Social impairments are inherent in individuals with Autism Spectrum Disorders (ASD), making the processing of social stimuli a primary area of research.  Face recognition specifically has been found to be atypical in older children and adults with ASD. Recently we documented that both scanning and face recognition is atypical in young children with ASD as well. However, it is not clear if deficits in recognition are restricted to faces or represent a generalized deficit in fast and implicit processing of novel stimuli, regardless of their social content.

Objectives:   This study examined whether difficulties in recognition observed in young children with ASD are limited to faces or whether they reflect a global impairment of recognizing novel stimuli, including those void of social information.

Methods:  Fifteen toddlers with ASD (age:M=41.5m,SD=14) and fifteen typically developing (TD) controls (age:M=38.2m,SD=7) were included in this study. All children underwent cognitive and diagnostic testing to confirm appropriate group placement. Each child was tested in a Visual Paired Comparison (VPC) paradigm, an approach used for studying visual stimulus processing that assumes an individual will fixate longer on a novel stimulus than on a paired familiar stimulus.  This novelty preference (NP) is operationalized as successful processing and recognition, and is calculated as the ratio of looking time at the novel stimulus to looking time at both the novel and familiar stimulus. Every child was shown twelve VPC trials split into two conditions:  Objects and Faces. Each trial included a recognition and a familiarization phase.  During familiarization, a photograph of either an object or face was presented on the screen.  A gray screen then appeared for five seconds followed by the recognition phase where the familiar image was displayed next to a comparable, novel image.  Visual responses were recorded using an eye tracker.

Results:  Duration of looking at the stimulus during familiarization did not differ between groups. In the Objects condition, there was no significant difference in the NP between the ASD and TD groups, and both had a NP significantly above chance, suggesting that both groups successfully encoded and recognized the objects (ASD:M=.56(.11),t(14)=2.26,p<.05; TD:M=.58(.09),t(14)=3.56,p<.01). However in the Face condition, a significant difference emerged in the NP between groups (ASD:M=.52(.11); TD:M=.59(.07),p<.05). The TD group had a NP that was significantly above chance (t(14)=4.85,p<.001), while the ASD group showed no preference for the familiar or novel face (t(14)=.619).

Conclusions:   Results of the study indicate that in young children with ASD, visual recognition deficits in faces do not extend to objects. Both groups demonstrated a preference for the novel objects, implying that children with autism are not impaired in encoding and recognizing a complex but nonsocial stimulus. However when viewing faces, children with ASD showed no novelty preference, as compared to their TD peers who were able to recognize the familiar face.  Although children with autism spend as much time looking at a face as TD children they are not encoding and processing in the same manner, which could have implications for development of social interaction skills.