In 2023, a remarkable apology was issued. In an interview with the New York Post, the psychiatrist Dr Allen Frances expressed regret for his role in the “massive, careless over-diagnosis” of autism. Frances chaired the taskforce that developed and broadened the criteria for autism in the DSM-IV – the fourth edition of The Diagnostic and Statistical Manual of Mental Disorders published in 1994. Nearly 20 years later, he said he was “very sorry for helping to lower the diagnosis bar”.
The DSM is the encyclopaedia of psychiatric and psychological conditions. It’s a text of huge significance: if a condition is not mentioned in the DSM, private insurers in the US are unlikely to cover the cost of its treatment. But the book’s growing size – as conditions are added with each edition – is indicative of a problem confronting the Western world: overdiagnosis. The manual’s first edition in 1952 listed 106 diagnoses across 132 pages. DSM-V, the latest full update, published in 2013, contains nearly 300 diagnoses; its 947 pages are “thick enough to stop a bullet”, according to one psychiatrist.
The line between mental disorder and normal diversity has become increasingly blurred. “We are not getting sicker – we are attributing more to sickness,” the consultant neurologist Suzanne O’Sullivan writes in her exceptional new book The Age of Diagnosis. With overdiagnosis, the diagnosis might be correct but “does not benefit the patient and may arguably do harm”. The problem detected is at a stage or severity that doesn’t require treatment. It applies to physical as well as mental health problems, and most often occurs in two forms: overdetection – “when new technologies and more sensitive and intensive screening programmes are used to detect earlier and milder forms of disease”; and expanded disease definitions – over time, people once considered healthy are drawn into the group.
Poppy, 24, is one of many patients who feature in the book. She has suffered poor mental health since she was 12: depression, self-harm, an eating disorder. She was bullied at school and tried to kill herself. “Nobody noticed I had autism,” she tells O’Sullivan. Poppy received her autism diagnosis at 20, during a two-hour appointment with a psychiatrist. She doesn’t take medication, “but is a great deal happier”. She has “lots of new friends in the autism community” and the diagnosis defines her. “‘I’m autistic,’ Poppy says. “It’s part of who I am. It’s not something that can be put down like a bag.”
It is accepted that autism has been historically underdiagnosed. But there now is a dispute about whether the correction has gone too far. Between 1998 and 2018, autism diagnoses rose by 787 per cent in the UK; 50 years ago, the disorder was said to affect four in 10,000 people worldwide – it’s now one in 100. Sceptics also point to widening diagnostic criteria. Typically, autism was diagnosed in children who were non-verbal and required help for everyday tasks. But O’Sullivan explains how DSM-V created a new category to account for those who might not meet the changed autism diagnostic criteria. Almost nobody is denied a diagnosis. “There has to be a point at which a person who doesn’t have the right symptoms and behaviours… simply doesn’t have the diagnosis,” she argues.
That point is reached, she says, when the diagnosis is causing more harm than good. Astonishingly, a lot of research doesn’t consider whether treatments might be harmful. A study that looked at 150 (non-drug) early autism-intervention projects found 139 did not measure harms at all – even when people withdrew because of adverse reactions. The researchers seemed certain that their “interventions were always for the best… that few seemed to feel the need to look for downsides”.
Miles, a retired 60-year-old banker, married with three children, received his diagnosis at 54. Like Poppy, autism is part of his identity. He doesn’t see it as a disability and doesn’t accept the label “autism spectrum disorder” either. He sees “the words ‘spectrum’ and ‘disorder’ as pejorative”. He doesn’t accept that “some people with autism are much more disabled than others” or that for them the diagnosis isn’t a positive, but rather, required to receive help.
Diagnoses of attention deficit hyperactivity disorder (ADHD) have also ballooned. Just like autism, diagnostic rates of severe ADHD are relatively stable. Increases have come almost exclusively from mild cases which now “vastly outnumber” those of severe ADHD or autism, O’Sullivan says. While UK rates of ADHD doubled in teenagers between 2000 and 2018, the biggest rise has been in adults. The ADHD Foundation estimates a 400 per cent increase in adults seeking a diagnosis from 2020 to 2023.
The idea that an adult can have ADHD at all is only a relatively recent medical phenomenon. The psychiatrist Alastair Santhouse first heard about it in 2000. Now, he estimates that two or three out of every ten patients he sees “have wondered about, or been advised by friends or family to ask about adult ADHD as an explanation for their troubles”. The NHS, he writes in No More Normal, has “buckled under the tsunami of referrals”. Inevitably, those who can afford it turn to the private sector, where in Santhouse’s experience, “in nearly all cases… they have had an ADHD diagnosis endorsed”.
Underpinning these rises has been a revolution in the patient-doctor relationship. Santhouse has seen “a fundamental shift in our social attitudes towards medical authority and the ownership of knowledge”. Expertise has “become unfashionable”. Diagnoses are no longer made, but perhaps more accurately “agreed with patients”. While “the democratisation of information” is a good thing, he writes, it is only “really going to work if there is a shared understanding of the human body and how it functions”.
Doctors are finding it increasingly hard to say no to those who seek a diagnosis. “Individuals may have started to construct their lives around the narrative of an ADHD diagnosis,” Santhouse explains. He relays a conversation with a specialist, who says that as he becomes “increasingly confident that ADHD is not the diagnosis”, an “anxiety begins to take hold of how he is going to break it to the patient”. There is “a clash between what the patient wants and what the doctor believes”.
To some, this might seem unimportant. Where is the harm in someone believing they have autism or ADHD if it makes them feel better about themselves? For one, writes Santhouse, “this quintessentially modern approach to mental health and mental illness… seems strangely regressive. Why would anyone want a diagnostic label to be the first thing someone sees or knows about them?” And O’Sullivan writes that “being part of a group that is centred around illness” isn’t always positive, as people “develop norms of behaviour that can lead to their symptoms and difficulties being exaggerated”. Often, being in the group is “predicated on being sick”. Recovery becomes almost impossible.
Compassion has driven both doctors to write. “Of the scores of people I spoke to, all of whom were adults, all perceived their lives to be better off thanks to a diagnosis,” O’Sullivan writes. “But almost all had left their job, dropped out of education and lost many old friends. Several were housebound.” She saw “a worrying gap between the perceived benefit of being diagnosed and any actual improvements in quality of life.” Both authors have witnessed the desire to help people go badly wrong.
For Santhouse, “in trying to be kind and inclusive, we have inadvertently done harm”. Not just to those who have received an unnecessary diagnosis. “My patients with severe autism are not represented on TikTok,” O’Sullivan writes. “Those with the greatest need are becoming invisible.” Research into new treatments becomes harder when those with severe conditions “are studied alongside those whose symptoms are so minimal that they were not noticed until late in life”. Grouped together, such patients don’t have enough in common to reach meaningful conclusions, O’Sullivan writes.
The Age of Diagnosis is not solely focused on neurodiversity or mental health. Chapter by brilliant chapter, it raises fundamental questions we should all be asking when thinking about illness, be it cancer or genetic disorders, never shying away from difficult truths, even when such truths are likely to upset people. O’Sullivan, for example, argues that most cases of long Covid are psychosomatic: arising “through a complex mind-body interaction rather than being due to tissue pathology caused by a virus”. The condition has no disease definition and “no specific symptoms”. She is aware that people will misunderstand her, and is at pains to point out that psychosomatic illnesses produce “real physical symptoms” and can be “as debilitating as… cancer”. They do require treatment – just different treatment. The truth can be painful. But it is only by confronting it that people’s lives can be improved – by better research and treatments. Santhouse reminds us that “kindness and clinical utility are not the same thing”.
Both doctors found writing these books difficult; both should be applauded for their bravery. Santhouse “hesitated” to write “for fear of being misunderstood”. But now is the time to act, to ensure diagnoses are saved for those who are, in O’Sullivan’s words, “unequivocally sick”. If we do not, Santhouse may well be justified in fearing “the future backlash of a generation harmed by our passivity, and the accusation, ‘You were meant to be the experts. Why didn’t you say anything?’”
Suzanne O’Sullivan appears at Cambridge Literary Festival on 20 March
The Age of Diagnosis: Sickness, Health and Why Medicine Has Gone Too Far
Suzanne O’Sullivan
Hodder, 320pp, £22
No More Normal: Mental Health in an Age of Over-Diagnosis
Alastair Santhouse
Granta, 272pp, £18.99
Purchasing a book may earn the NS a commission from Bookshop.org, who support independent bookshops
[See also: The Ozempic effect]
Content from our partners
Related
This article appears in the 12 Mar 2025 issue of the New Statesman, Why Britain isn’t working