Reduced neural sensitivity to rapid individual face discrimination in autism spectrum disorder

Abstract

BackgroundIndividuals with autism spectrum disorder (ASD) are characterized by impairments in social communication and interaction. Although difficulties at processing social signals from the face in ASD have been observed and emphasized for many years, there is a lot of inconsistency across both behavioral and neural studies.MethodsWe recorded scalp electroencephalography (EEG) in 23 8-to-12 year old boys with ASD and 23 matched typically developing boys using a fast periodic visual stimulation (FPVS) paradigm, providing objective (i.e., frequency-tagged), fast (i.e., few minutes) and highly sensitive measures of rapid face categorization, without requiring any explicit face processing task. We tested both the sensitivity to rapidly (i.e., at a glance) categorize faces among other objects and to individuate unfamiliar faces.OutcomesWhile general neural synchronization to the visual stimulation and neural responses indexing generic face categorization were undistinguishable between children with ASD and typically developing controls, neural responses indexing individual face discrimination over the occipito-temporal cortex were substantially reduced in the individuals with ASD. This difference vanished when faces were presented upside-down, due to the lack of significant face inversion effect in ASD.InterpretationThese data provide original evidence for a selective high-level impairment in individual face discrimination in ASD in an implicit task. The objective and rapid assessment of this function opens new perspectives for ASD diagnosis in clinical settings.

The right hemispheric dominance for face perception in preschool children depends on the visual discrimination level

“…Given that behavioral measures reflect a wide range of cognitive processes that undergo a long developmental course, and that different tasks and paradigms are used to test different populations, this controversy is difficult to resolve with behavioral studies alone. Here, with a quantitative electrophysiological measure that does not require an explicit face‐related task, we found not only that the individual face discrimination response is qualitatively different (i.e., simpler) than in adults, but that the inversion effect is much smaller: the EEG index of individual face discrimination was only 11% larger in amplitude for upright than inverted faces in the preschool children tested here, while it is about two times larger for upright than inverted faces in 8‐ to 12‐year‐ old children (Vettori et al, ) and almost two and half times larger in adults (Liu‐Shuang et al, ). This relative difference cannot be accounted for by an overall reduction of EEG amplitude, which is typically quite large in young children (e.g., Kuefner et al, ), including in the present study (i.e., the absolute amplitude of the individual discrimination response to upright faces (0.878 µV) was of 99% of the adult response (0.885 µV).…”

“…In addition, we included a condition in the face individuation paradigm in which the same face stimuli were presented upside‐down. In adults, this manipulation greatly reduces the individual face discrimination response (Liu‐Shuang et al, , ; Vettori et al, ), in line with the well‐known behavioral face inversion effect (Yin, ; for review see Rossion, ). Given that this behavioral effect of inversion is either absent in children (e.g., 6‐ and 8‐year‐old children: Carey & Diamond, ; Hills & Lewis, ; Schwarzer, ) or apparent but largely reduced as compared to adults (Carey, ; de Heering, Rossion, & Maurer, ), we expected no or a relatively small inversion effect of the electrophysiological index of individual face discrimination here, providing a platform to study the evolution of this effect during development.…”

“…In the individual face discrimination paradigm, the base images are constituted of an unfamiliar face identity presented repeatedly, varying in size. The deviant faces are different unfamiliar identities (Liu‐Shuang et al, ; see Liu‐Shuang et al, ; Vettori et al, for direct comparison of the two paradigms). In this latter paradigm, external features are removed from the pictures (hair, ears, etc.…”

Saccade Latency Provides Evidence for Reduced Face Inversion Effects With Higher Autism Traits

“…Reports of a lack of face inversion effects in ASD (McPartland et al, 2004;Rose et al, 2007;Vettori et al, 2019) have been used to argue that individuals with ASD use a more feature-based approach to face processing. Typical observers are thought to use a more global or holistic approach to face processing, rather than a feature-by-feature analysis, and it is the disruption to this global analysis from turning the face upside down that produces the face inversion effect (Rossion, 2009).…”

“…However, while most face inversion studies in those on the autism spectrum have measured identity recognition, only a few have utilized electrophysiological measurements to assess more automatic aspects of face processing, finding reduced inversion effects in ASD (McPartland et al, 2004;Vettori et al, 2019). Here, we measured saccadic latency when participants were detecting upright and inverted faces or cars which may also provide a more reflexive measure of attentional bias towards faces.…”