Autism spectrum debate: Is the diagnosis too broad?

The question seems simple enough at first: what does it mean to be autistic? But ask it in the wrong room, and you'll start a war.

Elon Musk, Greta Thunberg, Robbie Williams—these are the faces people conjure when they think of autism now. But then there is James Fitzpatrick, 34 years old, who does not speak, who has a learning disability, who requires constant care. He too carries an autism diagnosis. The same diagnosis. The same umbrella term that supposedly describes both the world's richest man and a man who cannot communicate his own needs. This gap—this chasm, really—is what has set off one of the most bitter arguments in modern medicine.

Three years ago, England had roughly 700,000 people with autism on their medical records. Today that number has climbed to 1.1 million. The National Autistic Society calls this evidence of long-overdue recognition, particularly among women and older people who went undiagnosed for decades. But others see something different: a diagnostic net cast so wide it has begun to catch people who may not actually be autistic at all. This is where Dame Uta Frith enters the story. At 85, she is one of the most influential voices in autism research since the 1960s. And she has begun to ask a question that has made her both celebrated and reviled: what if some of these newly diagnosed people have been misdiagnosed?

Frith is reluctant to name a number. "In my worst moments, I think it's a great number, but in my best moments, I think it is a small number," she said. What troubles her most is that people with severe autism—those who need the most support—are being overshadowed by a surge in late diagnoses, particularly among women. She worries that the definition has become so broad it has lost meaning. She suggests that some people diagnosed as autistic in adulthood may actually have anxiety, depression, or obsessive-compulsive disorder. Others, she hints, may simply be seeking an identity. Her solution: a "pruned" spectrum that would exclude many of the newly diagnosed.

The response has been swift and brutal. The National Autistic Society called her views "misinformation" and "false narratives." Autism researchers and advocates have accused her of being dangerous, of pitting autistic people against each other, of calling into question the very identity of millions. She receives hate mail. One message tells her she will have blood on her hands. Yet she continues, driven, she says, by "almost a duty" to represent those with the most severe needs—a group that makes up about a third of all autistic people but comprises only 6 percent of research participants.

The heart of the disagreement is not really about numbers. It is about what autism is, and whether a single diagnosis can meaningfully describe people with such radically different lives. Michael Fitzpatrick, James's father and a retired GP, cannot fathom how the same label applies to his son—who has injured himself, bitten himself, attacked people without warning—and to someone living independently. "Extremely different types of people," he calls them. Yet others argue that this logic is flawed. Dr. Rachel Moseley, an autistic academic who was diagnosed in her late twenties, points out that while people with severe support needs face real hardship, autistic people without learning disabilities have the highest suicide rates among the autistic population. "Everyone's challenges are different," she says. "But you can't make the assumption that the person with high support needs is always having a harder time."

One proposal gaining traction is to divide autism into subtypes—autism with learning disability, autism with language needs, autism with co-occurring ADHD or depression. This could allow for more precise targeting of support. But the National Autistic Society worries it could lead to discrimination, to people losing diagnoses and the services attached to them. Kayleigh, diagnosed in her thirties, captures the exhaustion many feel with this entire debate. She has been called high-functioning, given the label Asperger's, told she has AuDHD. "The goalposts keep moving," she says. "As an autistic person, I don't even know what I should be called." What she wants is not a perfect definition. She wants help. She wants to be able to leave her house without her family. She wants society to have recognized her struggle years ago. "You could call it 'silly brain disease'," she says. "I still know what I'm going through."

The definition of autism has shifted before. In 1911 it was a symptom of childhood schizophrenia. By 1943 it described children with peculiar interests who seemed indifferent to the world. In the 1960s, most autistic children were thought to have intellectual disabilities. By the 1980s, researchers embraced the idea of a spectrum—that autistic people could have average or very high intelligence. In 2013, all these variations were folded into a single diagnosis: Autism Spectrum Disorder. Now, after a decade of explosive growth in diagnoses, the question is whether that umbrella has become too large to be useful. The government has commissioned an independent review. But whatever it finds, it is unlikely to settle a question that cuts to the heart of identity, support, and what we owe to people whose needs are profound and whose voices are often unheard.