Blurry photograph of a crowd of people on a street.
Composite condition: A variety of social and biological factors influence whether a person is ultimately diagnosed with autism.
d3sign / Getty Images

Some who lack autism diagnosis carry variants tied to the condition

The variants are associated with slight differences in measures of intelligence, income and employment, but the relationship may not be causal.

About 1 percent of people not diagnosed with autism have rare variants in genes associated with the condition, a new study finds.

The results support the idea that a variety of social and biological factors help determine whether a person is ultimately diagnosed as autistic.

“If you carry a genetic variation, there is very often a diversity of outcomes that will depend on the genetic and environmental context of each carrier,” says lead investigator Thomas Bourgeron, director of the Human Genetics and Cognitive Function Unit at the Institut Pasteur in Paris, France.

More than 100 genes are associated with autism. But there has been little research on the prevalence of rare variants in these genes in people not diagnosed with autism, or about the effects they may have in this group.

Bourgeron and his colleagues analyzed genetic data on 13,091 people who have an autism diagnosis and 213,558 who do not, including 19,488 first-degree relatives of the autistic participants. The data came from the Simons Simplex Collection, SPARK, the Lundbeck Foundation Initiative for Integrative Psychiatric Research (iPSYCH) and the UK Biobank projects. (The Simons Simplex Collection and SPARK are funded by the Simons Foundation, Spectrum’s parent organization.)

Among 185 genes linked to autism with high confidence, 134 show loss-of-function variants in at least one person not diagnosed with autism, the researchers found.

Carriers of these variants had, on average, slightly lower scores on tests of fluid intelligence, or abstract reasoning, than did people who lack the variants, the study shows. Similar group discrepancies were seen for income, the highest educational or professional qualification reached, and metrics related to wealth, such as employment and home ownership.

“Such metrics are important if we want to understand the cumulative effect of the genetic variants on outcome,” says study investigator Thomas Rolland, a researcher at the French National Center for Scientific Research in Paris. He and his colleagues detailed their findings 26 June in Nature Medicine.

“The relationship between income, cognition, material deprivation and variants is very interesting and previously unexplored,” says Stephan Sanders, professor of pediatric neurogenetics at the University of Oxford in the United Kingdom, who did not contribute to the work.


he variants’ links to the outcomes may not be causal, the researchers caution. A variety of social and biological resilience factors could explain why some people with these variants flourish or not.

“Rather than claiming that genes hold an exceptional explanatory power when it comes to developmental diversity, the authors embrace the complexity of development,” says Kristien Hens, a bioethicist and professor of philosophy at the University of Antwerp in Belgium, who was not involved in study but co-authored a commentary about it. “Straightforwardly linking specific genetic variants with human flourishing in general and with neurodivergent flourishing in particular is next to impossible.”

For instance, social environment may influence whether people with autism traits receive a diagnosis. Educational settings may not always be tailored to undiagnosed people with autism traits, which could affect their outcome in adulthood, the researchers note. Previous research also suggests a “female protective effect” may exist that makes girls inherently more resistant to factors that predispose someone to autism.

“It is very tempting to interpret findings in population genetics in terms of a simplistic ‘genes for IQ’ or ‘genes for autism’ narrative,” Hens says. But “genetic information is a small part of what it means to function or flourish, and while genetic research is part of what helps us understand organisms, it is specifically the entanglement with other factors, such as environment and sociocultural expectations, that is in need for further study.”

Although the new study analyzed more than 200,000 people, Rolland notes it still lacks the statistical power to identify definitive links between mutations and specific outcomes.

“In the future, we need more participative research on data collected in a more unbiased way — on diagnosed and undiagnosed individuals,” Rolland says. “Such studies could identify genetic, social and environmental resilience factors that influence how individuals with certain characteristics can flourish better.”