Distinct Facial Phenotypes in Children with Autism Spectrum Disorders and Their Unaffected Siblings

Background:  

The neurodevelopmental hypothesis of autism holds that the cause(s) of autism spectrum disorders (ASD) occur during early prenatal development, affecting the developing brain. The brain and face are intimately related during development, sharing molecular, biochemical, and biomechanical influences. Our previous work has shown that boys with ASD display subtle, but distinct facial phenotypes as compared to typically- developing boys, and that these facial phenotypes are correlated with cognitive and clinical phenotypes. Further, facial phenotypic differences correspond to distinct developmental structures of the face.  Given the shared genetics between siblings, we expect that children with ASD and their unaffected siblings will share facial phenotypic characteristics that distinguish them from typically-developing children.

Objectives:  

Here, we test the hypothesis that unaffected siblings possess a facial phenotype that is intermediate between children with ASD and typically-developing children.

Methods:  

Our sample included a total of 213 participants, ages 4-18 years: 68 boys with ASD, 77 unaffected siblings of probands, and 69 typically-developing children.  3-D photographs were collected using the 3dMD Cranial® system.  3-D coordinates were collected for 17 anthropometric facial landmarks using 3dMD Patient® software. All possible linear distances between these landmarks were analyzed and statistically compared among the groups using Euclidean Distance Matrix Analysis (EDMA) and Principal Coordinates Analysis.

Results:  

Our results indicate that there are some phenotypic differences that distinguish both the affected and unaffected siblings from the typically-developing boys; these differences are localized to the lateral eye and temporal region.  However, unaffected siblings show greater differences from the typically-developing children than the children with ASD. Features that distinguish the unaffected siblings from the typically-developing group that are not observed in the probands include the region around the temples, narrowing of the eyes, and shorter and narrower measures of the philtrum. Further, unaffected siblings have narrower mouths than the children with ASD.

Conclusions:  

Unaffected siblings of children with ASD differ more from their affected siblings than from the typically-developing controls.  Although the cause of this variation cannot be definitively identified from this study, it is clear that the sibling pairs and typically-developing children undergo diverse patterns of head development. Localization of this facial phenotypic variation provides clues to when and where this divergence may arise, and how these divergent patterns correlate with gene expression. Future study will assess the relationships between these distinct facial phenotypes with behavioral phenotypes, and with genetic variation.