Researchers distinguish four types of autism and their genetic drivers

In a study in Nature Genetics, researchers analyzed genetic, phenotypic and clinical data from the database of people with autism spectrum disorder (ASD).

The paper stated that the differing phenotypic and clinical outcomes seen in subjects with autism correspond to different underlying genetic and molecular pathways. And that combined, their findings provide insight into the biology of autism.

“We believe the big picture takeaway from this work is that the complex, varied presentations of autism may arise from distinct underlying biological mechanisms rather than a single common genetic pattern among all individuals with autism,” said Natalie Sauerwald, associate research scientist at the Center for Computational Biology, Flatiron Institute in New York.

Autism has a complex phenotypic structure. It can be difficult untangling core features because they can vary substantially in severity and presentation between individuals. Plus, individuals may also have co-occurring conditions such as ADHD, the paper said.

While more than 1,000 genes have already been identified as playing roles in ASD, clinical testing helps explain autism in only about 20% of patients. Generally, little is known about the biology and actual mechanisms of how genes impact the condition.

To dig into this, the researchers drew on data from the SPARK cohort, a U.S. based nation-wide autism research project and part of the Simons Foundation Research Initiative.  It has more than 380,000 participants including 157,000 people with ASD and their first-degree family members.

For the study, the researchers used data from 5,392 individuals from the SPARK cohort. They were aged four to 18 years, and 80% were male. The researchers used 239 phenotype features derived from standard diagnostic questionnaires and background developmental histories. They then used a computational model to group individuals into different categories and classes based on their combinations of ASD traits.

Once clinically relevant autism subtypes were identified, the researchers then linked these to distinct genetic profiles and established developmental trajectories.

Developmental trajectories included using known brain development patterns. Such information describes which genes are active during different, specific points during brain development and what features they impact. Gene variants can affect how well specific parts of the brain grow and when.

The study findings defined four subtypes or classes of autism, as well as their genetic patterns: Social and Behavioral Challenges (abut 37% of participants); Mixed ASD with Developmental Delay (19% of the cohort); Moderate Challenges (34% of the cohort); and Broadly Affected (10% of the cohort).

There were no significant differences in the proportions of male vs. female within the groups, nobody assigned to multiple subtypes, Sauerwald told Canadian Healthcare Network.

Each of these subtypes has distinct developmental, medical, behavioural and psychiatric traits, along with their own patterns of genetic variation.  And “we are confident now that there are at least four, but we may find more with larger and richer datasets,” she said.

The work helps shed light on biological mechanisms that contribute to the condition. 

It also “helps explain why past genetic studies often fell short—it was like trying to solve a jigsaw puzzle without realizing we were actually looking at multiple different puzzles mixed together. We couldn’t see the full picture, the genetic patterns, until we first separated individuals into subtypes,” she said.

But the study partly reinvents the wheel as previous studies preempt some of their findings, according to Stephen Scherer, a professor of genetics at the University of Toronto, who co-founded both the Autism Genome Project and the Autism Speaks MSSNG 10,000 Project.

For example, there was a North American-wide study he participated in 2020, which was reportedly the largest genetic sequencing study of ASD at that point. Published in Cell, it used 35,584 samples from nearly 12,000 people with ASD. The researchers were able to associate risk genes with clinical features and cellular mechanisms. 

The findings of that study, and others, not only showed which mutations were associated with specific phenotypic differences but described how they affected cellular interactions in the developing human cortex.

As for the new study, Scherer stated in an email exchange with CHN, that the researchers used data from the Simons Foundation, which he described as “an amazing resource.”

“However, some of the statistical methods used may limit its immediate clinical applicability,” and it does not identify and meaningful new biomarkers, he said.