Autism and face processing

Neuhaus E, Kresse A, Faja S, Bernier RA, Webb SJ. Face processing among twins
with and without autism: social correlates and twin concordance. Soc Cogn Affect
Neurosci. 2016 Jan;11(1):44-54.

Abstract

Autism spectrum disorder (ASD) has a strong heritable basis, as evidenced by twin concordance rates. Within ASD, symptom domains may arise via independent genetic contributions, with varying heritabilities and genetic mechanisms. In this article, we explore social functioning in the form of (i) electrophysiological and behavioral measures of face processing (P1 and N170) and (ii) social behavior among child and adolescent twins with (N = 52) and without ASD (N = 66). Twins without ASD had better holistic face processing and face memory, faster P1 responses and greater sensitivity to the effects of facial inversion on P1. In contrast, N170 responses to faces were similar across diagnosis, with more negative amplitudes for faces vs non-face images. Across the sample, stronger social skills and fewer social difficulties were associated with faster P1 and N170 responses to upright faces, and better face memory. Twins were highly correlated within pairs across most measures, but correlations were significantly stronger for monozygotic vs dizygotic pairs on N170 latency and social problems. We suggest common developmental influences across twins for face processing and social behavior, but highlight (i) neural speed of face processing and (ii) social difficulties as important avenues in the search for genetic underpinnings in ASD.



Families of twins were recruited from community schools, groups for parents of multiple births and a University autism center registry. Twin pairs between the ages of 5 and 30 years with and without ASD were invited to participate based on parent report that both twins were typically developing, both twins had ASD, or one twin had ASD...

Among the 56 twin pairs, both twins received ASD diagnoses in 19 pairs (ASD-concordant pairs); only one twin received an ASD diagnosis in 12 pairs (ASD-discordant pairs) and neither twin received an ASD diagnosis in 25 pairs (non-ASD pairs). In one of the triplet families, none of the individuals had ASD; in the second, two of the three siblings had ASD....

Our findings yield several insights into face processing and social behavior in ASD. Consistent with expectations, we observed differences between participants with and without ASD on P1 latency to human faces, suggesting a slower P1 response among individuals with ASD...

In contrast to previous studies and to our hypotheses, however, we did not find effects of ASD diagnosis on the N170 response to faces. Although earlier work suggested reduced N170 speed and amplitude in ASD, we found that the sample as a whole evidenced more negative N170 to faces and upright images, with no effect of ASD diagnosis...

Electrophysiological and behavioral measures of face processing were significantly correlated with social skills and social difficulties in our sample when participants with and without ASD were considered together. Faster processing (i.e., faster P1) and more accurate processing (i.e., higher accuracy scores on holistic and memory tasks) were associated with better social skills and fewer social difficulties, in accordance with hypotheses. Interestingly, these associations emerged in the absence of correlations between P1 latency and behavioral indices of face processing (i.e. holistic face task and face memory). This may be due to the fact that these tasks draw upon a range of skills, including early attention (indexed in part by P1), memory and executive function. Although all of these would likely contribute to social functioning, the P1 response captures only a small component of these demands and thus might not appear to be correlated. Also surprising, we did not observe correlations between response amplitudes and social skills. To the extent that amplitude reflects neural effort, this pattern of association may suggest that successful social functioning relies more upon speed of face processing rather than neural effort. Indeed, social cues (e.g. facial expressions, changes in eyegaze) are typically fleeting and dynamic, and slowed processing of these cues could result in missed or inaccurate social inferences... 

With regard to twin correspondence, we found strong correlations between twins on nearly all measures of interest, including P1 and N170 latency and amplitude, face processing tasks and social measures. Although strong correlations on parent-report measures could be due in part to the same parent reporting on both of their twins, the breadth and strength of our observed correlations also suggest shared developmental influences on twins' social behavior and face processing. For many variables, twin correspondence was similar for MZ and DZ pairs, suggesting that many of these influences on social development may be environmental in nature (e.g. family factors, peer groups, intrauterine environment), rather than strictly genetic. Twin similarity may have been especially strong in our sample because our analyses of twin correspondence were restricted to sex- and diagnosis-concordant pairs....

Our findings provide preliminary evidence of (i) differential P1 but similar N170 responding to faces among twins with and without ASD, (ii) tentative links between neural and behavioral face processing, and social functioning, and (iii) potential heritability of both neural speed in face processing and social difficulties. Taken together, these findings reinforce the importance of face processing, both in predicting meaningful social outcomes and in the search for potential genetic underpinnings in ASD.

http://www.medscape.com/viewarticle/857360_7?nlid=98563_3001