Because there is no singular cause of autism, scientists say there is therefore no singular driver behind the rise in cases.
But at the heart of the question is an important distinction: are more people exhibiting the traits of autism, or are more people with such traits now being identified? It seems to be both, but researchers really aren’t sure of the maths.
Here is what they know.
An elusive cause
More than 100 genes have been associated with autism, but the disorder appears to result from a complex combination of genetic susceptibilities and environmental triggers.
The CDC has a large-scale study on the risk factors that can contribute to autism, and researchers have examined dozens of potential triggers, including pollution, exposure to toxic chemicals and viral infections during pregnancy.
Some research suggests that babies born to older parents – particularly an older father – may be at increased risk of autism. Other studies hint that premature birth or low birth weight could be associated with autism, which is often linked to high oxidative stress.
The idea that such factors could also be involved in the overall rise in autism is convincing to Juergen Hahn, a professor at Rensselaer Polytechnic Institute who studies the computational systems biology of autism, given that both the number of children surviving premature birth and the average age of new parents are rising.
But testing a hypothetical driver of autism would mean controlling for the endless list of other influences on early development and following the child well into adulthood, when some people now receive their diagnosis.
“These are very complicated studies to conduct, especially if you want statistical certainty, and there are no easy answers,” Dr Hahn said. “Sometimes we just have to say we don’t know. And that always gives people room for speculation.”
The misinformed theory that vaccines are behind the disorder is one such case of speculation. It gained traction when a British researcher named Andrew Wakefield published a study of 12 children in the late 1990s, purporting to reveal a link between the measles, mumps and rubella vaccine and autism.
Many larger studies have since discredited the hypothesis, including one that enrolled the entire Danish child population. Regardless of the types of vaccines, the ingredients or the timing of the schedule, researchers have not been able to establish a link. Dr Wakefield’s article was retracted, and he lost his medical license, but the seeds had been planted in social discourse.
Diagnostic changes
One component in the boom of autism cases is far more certain: the broadening parameters of the spectrum.
Autism first appeared in the Diagnostic and Statistical Manual of Mental Disorders (DSM) in 1980, in the guide’s third edition, but its definition changed in 1987, when it came to include children whose symptoms appeared at older ages – after 30 months. The new edition also expanded the criteria for a diagnosis of autism, to 16 from six, and said children needed to exhibit only half of the 16 criteria listed, rather than all six of the previous criteria.
A fourth edition of the DSM, published in 1997, included Asperger syndrome, a social disorder that can be marked by a preoccupation with a single interest, on the autism spectrum. That was a significant shift, since it meant that all people – including those with average or even above-average intellectual abilities – could receive the diagnosis.
The fifth edition, released in 2013, allowed clinicians to give a combined diagnosis of autism spectrum disorder and attention deficit hyperactivity disorder. It also folded autism, Asperger syndrome and a condition called PDD-NOS – short for pervasive developmental disorder-not otherwise specified – under a single umbrella.
“What we call autism has itself changed to become a broader category,” said Simon Baron-Cohen, the director of the Autism Research Centre at the University of Cambridge. And, he said, with “the growth in private clinicians offering diagnosis – it has become an industry”.
Increased awareness
As those clinical processes were evolving, so was the social infrastructure. Until the 1980s, many people with autism were institutionalised, so parents were far less familiar with the hallmark traits of the disorder and typically did not recognise them or seek a diagnosis when they occurred in their own child.
Then, in 1991, children with an autism diagnosis began to qualify for special services in schools, which gave parents an incentive to seek out and accept diagnoses. Around 2007, the American Academy of Pediatrics recommended that all children be screened for autism at 18 and 24 months, which experts believe led to a major uptick in detection.
Several changes in population-level trends add further evidence that awareness and inclusion could be behind the surge. Historically, boys have been diagnosed with autism at far higher rates than girls, who are thought to “mask” their autism symptoms because of social norms. But in an era of greater acceptance for neurodiversity, the gender gap is narrowing.
Hispanic, Black and Asian children showed major increases in autism diagnoses compared with white children between 2011 and 2022, which could reflect improved access to diagnostic services. And – perhaps most indicative – relative increases in autism diagnosis rates in the last decade were greatest not among children, but among adults ages 26 to 34, many of whom were seeking their own diagnosis for the first time.
Researchers also point to the influence of social media as “one of the biggest unrecognised factors driving the explosion of autism diagnoses”, Dr Baron-Cohen said. Platforms like Reddit, he said, became a space “where people who might wonder if they or their child were autistic could communicate with each other and feel a new sense of belonging and identity”.
Still, for Catherine Lord, a psychologist at the University of California, Los Angeles, who has developed diagnostic tools and longitudinal studies for autism, the maths just isn’t working.
“Some of it is clearly from greater awareness and broader inclusion,” she said. But even if those are the biggest factors driving the surge, and even if biological and environmental factors are contributing as well, she said, “it seems a little far-fetched to think that all of these add up to the increase”.
She is confident, though, from copious data in both the United States and abroad, that heightened autism prevalence does not correspond to childhood immunisations.
“Whatever it is,” she added, “it’s not vaccines.”
This article originally appeared in The New York Times.
Written by: Emily Baumgaertner
Photograph by: Callaghan O’Hare
©2025 THE NEW YORK TIMES