Genetic risk for autism spectrum disorders and neuropsychiatric variation in the general population

About the study

Autism spectrum disorders (ASDs) are a highly heterogeneous group of neurodevelopmental disorders characterized by impairments in social communication, and by restricted and repetitive interests. Many types of genetic variation influence risk for ASDs, including both common polygenic variation and rare de novo variants. There has been long standing research interest in the relationship between ASDs and more typical variation in social communication ability in the general population. Twin and family studies have shown that social communication can be measured on a continuum in the population, that social communication traits are heritable, and have suggested that there is a familial link between extreme and subthreshold autism-like behaviors. The purpose of this study was to capitalize on new genetic data and directly examine the relationship between genetic risk for ASDs and general population variation in behaviour.

We used two types of genetic tools to examine this relationship. First, a new technique called LD Score Correlation allows one to estimate genetic correlations between traits using the results from genome-wide association studies (GWAS) ( Bulik-Sullivan and Finucane et al., 2015). Genetic correlations estimate the similarity of the common, polygenic influences (those very small and very numerous influences distributed across the genome that operate additively to create risk) on two traits. Traits with a genetic correlation of 1 are estimated to share all of their common, polygenic influences. Those with a genetic correlation of 0 are estimated to share none. In this study, we estimated the genetic relationship between ASDs and population variation in social communication ability using results from three GWAS: 1) a GWAS of the Social Communication Disorders Checklist (SCDC) measured in childhood in the Avon Longitudinal Study of Parents and Children (St. Pourcain et al., 2014) 2) a GWAS of ASD from the Psychiatric Genomics Consortium ASD group and 3) and GWAS of ASD from the iPSYCH-SSI-Broad ASD group. The genetic correlation between the SCDC and each of the ASD samples was approximately 0.3 (p<0.01 for both comparisons). This result suggests a substantial degree of genetic sharing between ASDs and typical variation in social communication ability in the population.

We next looked for evidence of a relationship between the de novo influences on ASD risk and impairments in adaptive functioning in the unaffected siblings of children with ASDs. The Vineland Adaptive Behavior Scales capture differences in social skills, communication ability, and daily living skills. Rate of de novo variants from categories that confer ASD risk was associated with impairments in adaptive functioning in both ASD cases (p=0.008) and their unaffected siblings (p=0.0002). In cases and unaffected siblings with similar levels of impairment in adaptive functioning, de novo rates were not statistically distinguishable.

Both the common and rare variant analyses suggest that genetic risk for ASDs influences a continuum of behavioural and developmental outcomes in the population. These results have significant implications for genetic models of neuropsychiatric disease risk.

The article “Genetic risk for autism spectrum disorders and neuropsychiatric variation in the general population” was published in “Nature Genetics, 2016/48/5”.

Facts about the study

• Social communication ability can be measured in the general population as a quantitative trait. Social communication differences in children in the general population are heritable and linked to genetic risk for ASDs.
• ASD risk reflects several types of genetic influences. This study examined both common, polygenic risk and de novo variant risk. Both were associated with behavioural differences in the general population (controls).
• The correlation between the common, polygenic influences on ASDs and the common, polygenic influences on social communication differences in the pediatric general population was approximately 0.3 (p<0.01). This suggests that the genetic relationship between social communication variation in the population and ASDs is similar in strength to the genetic relationship between obesity and Type 2 diabetes.
• De novo variants associated with ASD risk predicted a continuum of functional impairments across ASD cases and controls. Cases and controls with similar levels of functional impairment did not differ in average de novo variant rate.

Further information

For questions contact Elise Robinson, PhD, the Broad Institute, USA

E-mail: erob@broadinstitute.org