Giving children roots and wings: the struggle to understand autism

On the wall of Dr. Evdokia Anagnostou’s office hangs a framed photo of a lone toddler. The boy, a patient of hers, is captured pedalling away from the camera on a plastic orange tricycle.

A quote attributed to U.S. journalist Hodding Carter Jr. floats above his head, and Dr. Anagnostou reads it out loud: “There are only two lasting bequests we can hope to give our children. One of these is roots; the other wings.”

It’s a gift she received years ago from the boy’s father, she explains. It’s also a reminder of the challenge she faces. Roots and wings are a tall order for a doctor who’s striving to understand and find new treatments for autism spectrum disorder. Dr. Anagnostou, a senior clinician scientist at Toronto’s Holland Bloorview Kids Rehabilitation Hospital, has spent more than a decade trying to unearth the causes of autism, with the aim of improving the lives of children affected by it.

But it’s become clear that she and others in her field have hit a wall, the soft-spoken child neurologist tells me.

Although her young patients share the same diagnostic criteria for autism spectrum disorder, including difficulties with social interactions and communication and repetitive behaviours, individual cases often look very different. One child might thrive at school, while another might be nonverbal and prone to hitting himself. And almost all the children she sees also have additional disorders, like attention-deficit hyperactivity disorder (ADHD), obsessive-compulsive disorder (OCD), and intellectual disability.

Further confounding her: In study after study, potential therapies that have shown promise in some children fail to yield the same results when applied to larger groups of individuals with the same diagnosis. And not a single drug has been approved that could treat the core symptoms of autism, which by some projections affect as many as one in 63 people.

“I am frustrated,” she says, then adds with a puckish smile, “but I am also feeling that this is the right time to do something different.”

About five years ago, Dr. Anagnostou began to consider some unsettling questions: What if she and her colleagues had it all wrong? What if by grouping all their patients under an autism or ADHD or OCD label, they were missing critical clues about the roots of their symptoms? What if they needed to rethink their diagnostic categories?

She and her colleagues are now taking a new tack. In a trailblazing project called the Province of Ontario Neurodevelopmental Disorders (POND) Network, sponsored by the provincially funded Ontario Brain Institute, they’re studying multiple disorders together, with the hope that such an approach will lead to a better understanding of their biology.

“We basically say that these concepts may be overlapping. So autism spectrum disorder may not be a distinct biological entity,” Dr. Anagnostou, principal investigator of POND and Canada Research Chair in translational therapeutics in autism, explains.

The POND Network is part of a paradigm shift in how researchers around the world are beginning to examine mental and neurological disorders. Many are finding the current diagnostic categories don’t quite reflect what genetics and neuroscience suggest about the underlying biology of dysfunction. So rather than study each disorder in isolation, they’re compiling wide pools of data and seeing where the patterns emerge.

If they are on the right path, this approach may lead to more effective, targeted treatments and, ultimately, to entirely new ways of classifying the disorders. But a diagnosis is more than just a label for a set of symptoms. For some patients, it’s a core part of their identity. What would new diagnostic categories mean for them?

‘The big poke in the eye’

When I first heard Dr. Anagnostou discuss her work at an Ontario Brain Institute presentation in December, my mind reeled. Does this mean, I thought, dumbstruck, autism spectrum disorder may not exist?

On the contrary, the symptoms of autism obviously exist. Dr. Anagnostou and her colleagues are clear on that. The issue is they may not exist in isolation. “So the construct [of autism] may not be very useful as a diagnostic category,” Dr. Anagnostou explains.

I’m reminded of childhood sorting games, where a category such as “things that hop” might include kangaroos, frogs and grasshoppers, even though to get a better grasp of the nature of each creature, you’re better off grouping them with other mammals, insects and amphibians. This oversimplifies the matter, of course. But throughout history, people have categorized mental illness and neurological conditions in various ways that don’t necessarily best reflect their nature. The ancient Greeks, for example, explained them as an imbalance of bodily humours or fluids.

“It’s human nature to try to categorize things because lumping helps us understand, especially in the face of great diversity,” says Dr. Russell Schachar, a co-investigator of POND and senior scientist at Toronto’s SickKids hospital.

Lumping things together may seem like an intellectual exercise, but the ramifications of doing so are very real. Current diagnostic categories, developed over the past century, have provided criteria for distinguishing one type of disorder from another: autism from intellectual deficiency, schizophrenia from bipolar disorder, and ADHD from conduct disorder. And it has allowed doctors to communicate and policymakers to collect statistics and determine public-health strategies. “That helped us in every which way,” Dr. Schachar says.

But the discovery in recent years of genetic mutations that cut across disorders have presented “the big poke in the eye, if you will – the thing that made it impossible to go any further,” he says.

Genome sequencing, for instance, has pointed to not just one autism mutation, nor even a manageable handful, but hundreds of possible mutations that could be interacting with an infinite number of environmental factors to determine whether an individual develops the disorder. What’s more, these mutations aren’t exclusive to autism. Many of them also appear to be involved in a variety of other neurodevelopmental disorders.

To get anywhere in understanding what’s going on in these neurodevelopmental disorders, Dr. Schachar says, scientists must figure out what is shared among them and what is distinct.

Instead of conducting separate studies on each disorder, POND’s approach breaks down the boundaries between them. For example, a POND study published in the Journal of the American Academy of Child and Adolescent Psychiatry last year examined how well participants diagnosed with OCD, ADHD, and autism could read other people’s emotions by looking at their eyes. Although the inability to read emotions is assumed to be an autistic trait, Dr. Anagnostou explains the researchers found it is common among children with ADHD too.

“It looks like it has nothing to do with disorder,” she says. “If you’re an ADHD kid with impaired social function, your ability to read emotions is as bad as a kid with ASD [autism spectrum disorder]. So it cuts across disorders.”

POND researchers are also beginning to find similar structural connectivity in the brains of children who have ADHD and autism.

Other studies within POND are looking at participants’ cognitive processes, neurochemistry, and the role of epigenetics or how environmental exposures and other factors may interact with genes.

Now in its third year, the POND project is collecting data on the genetics, behaviours and brain imaging of children with various disorders, including autism spectrum disorder, ADHD, OCD, intellectual disability, Rett syndrome, Tourette syndrome, Fragile X syndrome or Down syndrome. It’s a whole new approach to conceptualizing the problem and conducting research, Schacher says.

Thus far, investigators from several participating institutions, including Holland Bloorview and SickKids in Toronto, and McMaster University and University of Western Ontario, have recruited some 1,300 children. Their goal is to bring that to a total of 2,500 within the next two years.

Reliable but not valid?

This new approach to research is like tackling a puzzle from the opposite end. Traditionally, investigators have asked “what can we find out about a specific disorder?” whereas the paradigm shift involves sorting through all the various puzzle pieces – the genetics, neuroscience and behaviours – and asking, “what categories do these form?”

The POND investigators aren’t the only ones finding that current diagnostic constructs don’t seem to align with the latest findings in genetics and neuroscience.

In the United States, the National Institute of Mental Health is taking a similar approach. Its Research Domain Criteria, an initiative launched in 2009, encourages investigators seeking research funding to think beyond the existing diagnostic categories of mental disorders of all kinds.

“Nature, the brain, behaviour, symptoms don’t abide by those categories, and so we shouldn’t pretend that they do any more because it’s getting in the way of progress,” says Dr. Sarah Morris, acting director of the initiative.

Dr. Morris says she and her colleagues acknowledge that the Diagnostic and Statistical Manual of Mental Disorders (DSM) and the International Classification of Diseases (ICD) are the standard diagnostic tools for clinical care. But they hope that by opening up the research to other conceptualizations of mental disorders, the results may inform better diagnostic practices. So, for example, instead of conducting a study on depression, a research team might focus on the various genes, molecules, brain circuits or behaviours that may interfere with people’s responses to reward.

The thing about the existing categories, whether it’s autism, ADHD, depression or schizophrenia, is they’re reliable in a scientific sense, Dr. Morris notes, meaning the same results can be generated repeatedly. Individuals with a specific set of symptoms, for instance, can reliably be diagnosed with the same disorder by multiple doctors. But, she says, the real question is whether the current diagnostic categories are valid in terms of understanding the nature of mental illness. That is, do they accurately reflect the phenomenon being studied?

“There’s evidence to think that they’re not,” Morris says.

Neurodiversity and Identity

In recent years, the concept of “neurodiversity” has grown as a movement in the world of autism advocacy. Neurodiversity – the idea that autism and other neurological differences are normal variations of human life – has gained widespread public attention, thanks to the likes of renowned autism activist Temple Grandin, whose 2010 TED Talk on the importance of autistic minds such as her own has received more than 3.7 million online views, and journalist Steve Silberman, author of the recent best-seller NeuroTribes: The Legacy of Autism and the Future of Neurodiversity. Many proponents of neurodiversity view autism and ADHD not as deficiencies that require fixes or cures, but as natural differences that require acceptance and inclusion in society.

Susan Cosgrove, a Toronto autism advocate who has three children with neurodevelopmental disorders, says she fully believes in neurodiversity. But at the same time, she says, her children need treatment to be able to cope with the world around them. The city won’t get any quieter just because they’re overwhelmed by sounds, she says, nor will buses clear out just because they find it too crowded.

“I absolutely, 100 per cent believe in accepting them for who they are, but the society is not built for them right now,” she says. “Therapy and treatment is super important and is just as important now as it was before neurodiversity was accepted.”

Cosgrove says the questions the POND researchers are posing make sense to her. Her family offers proof that the lines between neurodevelopmental disorders are hazy.

Between her two sons, Liam, 12 and Phoenix, 2, both diagnosed with autism, “there are definitely more differences than similarities,” she says. At the same time, both also share symptoms with their sister, Kaya, 10, who has been diagnosed with ADHD. Those include difficulty with sensory processing and working memory. At Phoenix’s age, for example, Liam barely spoke. And when he did, he mostly uttered words he made up himself. By contrast, Phoenix has an astonishingly precocious ability to read and an uncanny affinity for numbers.

During a recent visit to the family’s apartment, Liam, a shy, sweet preteen, quietly watches his favourite YouTube video on the features of a model excavator. For him, having autism means “I’m scared of strangers,” he explains to me, and “it can be hard to communicate.”

Meanwhile, Phoenix, an energetic, cherubic toddler reads from a stack of pictureless flashcards, shown to him in random order. Without even having to sound the letters out, he bounced happily as he declared in rapid-fire: “Truck!” “Orange!” Hockey!” “Fifty” “Dinosaur!” “Clown!” “Diamond!” “Egg roll!”

So, if it turns out their diagnostic labels aren’t valid, what does this mean for her children’s care?

A new classification

Dr. Anagnostou emphasizes POND has a long way to go before it can offer any answers. But eventually, she says, if it turns out the researchers can conclude they “have no confidence that our constructs exist,” it could lead to new diagnostic labels.

“Maybe instead of having autism, ADHD, intellectual disability, OCD, we may have [separate clusters of patients with] these types of mutations, these types of brain findings, and these types of behaviours. And they’re more likely to respond to medications A, B, C,” Dr. Anagnostou says.

This will, however, have significant consequences for patients and those who provide services for them, she says.

At present, for instance, even though there is great variability between children with autism, a school teacher may at least have a general idea of what to expect of a student with an autism diagnosis, Dr. Anagnostou says. “But if I told them this child has mutation X, Y, Z in this particular cluster, what will the teacher do with that?”

In addition, she says, people often incorporate their diagnoses as part of their identity.

“So what happens if somebody identifies as an autistic person and we come out and say, ‘We don’t know that autism exists as an entity’? That would be something we need to figure out as we proceed.”

The aim, she says, is not to take away patients’ diagnoses, but to take away their suffering.

Cosgrove, whose two sons are enrolled in the POND study, says she isn’t holding her breath, but she would have no hesitation accepting new diagnostic labels, as long as they meant her children would receive more appropriate treatment that would improve the quality of their lives.

“Whatever gives them the best chance of a successful future,” she says.

In Britain, Dr. Declan Murphy, managing director of the European Autism Interventions – A Multicentre Study for Developing New Medications (EU-AIMS), doesn’t believe doctors will stop diagnosing people with autism in the future. Rather, he says, sub-types of autism may eventually emerge, which will allow for better targeted treatments.

“It wouldn’t be that someone who is autistic today would be… no longer called autistic,” Dr. Murphy says.

But, he says, what has become clear is that researchers need to collaborate to collect and share data, since it’s only by studying large numbers of individuals that meaningful subgroups can be identified. “The days of small studies are just over, really. They’ve gone far enough.”

From research to patient

For families and advocacy groups, the promise of better treatments and explanations seems to outweigh any complications that may arise out of reconsidering diagnostic constructs.

Dr. Suzanne Lewis, chair of the professional advisory committee for the advocacy group Autism Canada, which is not involved in POND, says she applauds POND’s approach.

“It’s not so much what it means to those who are defining the labels. It really is: What does it mean to those individuals and families living with the condition?” Dr. Lewis says, adding she has no doubt that research into the biology of the condition will benefit families. “It’ll remove burdens of them wondering ‘What did I do or what didn’t I do that may have contributed to my child having autism or schizophrenia or whatever?’ ”

Back at Holland Bloorview, Dr. Anagnostou acknowledges the POND project is a gamble. While she is confident they’re onto something, she admits it’s possible she and her colleagues may end up being wrong about their assumption that the differences and similiarites among neurodevelopmental disorders will lead to any breakthrough in treatments.

“We may fail,” she says. But, she notes, in the end, whatever data they end up with will help inform future research one way or another.

“If we haven’t made a big difference with an alternative approach, we’ll have learned something so we aren’t where we started. But the kids are where they were,” she says.