It is Valentine’s Day in Jack’s first-grade class. A few at a time, his classmates get up to hand out their valentines, effortlessly matching names with the faces of their classmates until, empty-handed, they dart back to their seats. Jack glances down at the name on the top of his stack and wanders the maze of desks, staring intently, not at the children’s faces, but at the brown paper bags on each desk.
“He’d look at the kid; then he’d look at the bag to see the name. He didn’t know, ‘Oh that kid who’s been sitting at that desk for six months—I know who that is,’ ” recalled Jack’s mother, Jean Winegardner, who was helping out in his Bethesda, Maryland public school that day.
Once home, Jean took each of the valentines Jack’s classmates had given him and asked him to describe the child who made it. He couldn’t. Jack had been in class with the same children since kindergarten and was quick to correct his mother’s pronunciation of the names, but could not connect any name to a single detail about any of his classmates.
Jack has a lot of trouble with social interaction, clings to strict routines (for a long time he only ate foods that were brown, for example), and had trouble learning to speak when he was younger. Since the age of five he has had a diagnosis on the spectrum of autistic disorders. But in a little more than a year, his specific diagnosis, Pervasive Developmental Disorder Not Otherwise Specified (PDD-NOS), will not exist. Neither will the related and better-known diagnosis of Asperger’s Syndrome.
A group of psychiatrists are working on radical changes to the words that define the diagnoses that make up Autism Spectrum Disorder. The changes at first might seem esoteric, affecting just a few pages in a dense volume known as the Diagnostic and Statistical Manual of Mental Disorders, or DSM. But for autism, words are key. The disorder cannot be diagnosed by a blood test or a brain scan. Change the DSM definition, which is the authoritative guide to the diagnosis of mental disorders, and you change who is considered to have the condition.
The impending changes may make the definition significantly more stringent. To those who believe doctors have created an “autism epidemic” by handing out too many diagnoses to children with marginal impairments, the shift represents an opportunity to bring order to an out-of-control situation. To advocates for autistic children and their families, the proposals threaten to deprive hundreds of thousands of children and adults of badly needed services and a label that has become central to their identities.
The debate, which until recently was largely confined to academic circles, has attracted increasing public attention with the release of new data by two Yale experts on autism. During a presentation at the conference of the Icelandic Medical Association in January, Fred Volkmar and James McPartland announced that their study showed that more than half of children with normal or high IQs who are currently diagnosed with autism spectrum disorders no longer would qualify under the proposed new definitions. The study itself is currently under review by the Journal of the American Academy of Child and Adolescent Psychiatry, and is embargoed until publication in April.
“The sad thing is for many more able kids, because you see how smart they are, you don’t see how socially impaired they are,” said Volkmar, chief of child psychiatry at Yale-New Haven Children’s Hospital and chairman of the Yale Child Study Center.
“One might argue, if you have a system that’s working pretty well, why mess with it?” Volkmar asked. Eighteen years ago, Volkmar headed the committee that last rewrote the definition of autism when the current version of the DSM, the DSM-IV, was written. The main goal then, he said, was to broaden the spectrum so that individuals with higher IQs could receive the benefits of having a diagnosis.
He continues to see the existing definitions as a significant success. In a recent interview, he leaned back in his chair and spread his long arms to indicate the inclusiveness of the spectrum his committee designed eighteen years ago, then folded them over his rotund belly. “It would be nice to be able to get help to people who want it,” he said.
Michael John Carley, for one, believes cutting anyone off the autism spectrum entirely would be a tragedy. Carley is the head of the Global and Regional Asperger’s Syndrome Partnership, or GRASP, which is the largest organization of adults with autistic disorders. He was diagnosed with Asperger’s as an adult just a week after his son was diagnosed with autism. His own experiences and those of the other members of his organization have made Carley keenly aware of the sense of identity a diagnosis brings, and the large, supportive communities it can form. He said in November that he had been reassured about the proposed new standards. “Every expert has told me, don’t worry about it, it’s not going to happen,” he said then.
Several days after Volkmar and McPartland presented their study in Iceland, though, Carley indicated concern. “We’re a little surprised by this turn. I don’t think anybody told me an untruth necessarily, but clearly something inside that room has changed,” he said. “We couldn’t more strongly disagree,” he added.
“The intention of the committee was not to change the rate of autism spectrum disorders, but rather to make the criteria better match up to what clinicians actually do,” said Catherine Lord, a member of the American Psychiatric Association committee that is revising the criteria for the new version of the manual, which will be called the DSM-5. She feels that while “the DSM-IV criteria were a real improvement,” they left too much up to the arbitrary judgment of clinicians. The new criteria are an attempt to more thoroughly describe the patterns of symptoms and behavior that doctors are observing in children and adults with autism.
Lord also questions the ethics of Volkmar’s decision to release information from an embargoed article, calling the amount of publicity it has received “appalling.” When scientific articles are under embargo by a journal, the authors are expected not to discuss the study until it has been reviewed and printed.
According to McPartland, he did not find out that their article had been accepted by the journal until the day Volkmar gave his talk in Iceland, so the embargo may not have gone into effect until after Volkmar’s presentation. In any case, the embargo rule most scientific journals follow allows authors to discuss abstracts of upcoming papers at scientific conferences, as Volkmar did. While authors are not supposed to court media attention, they are not penalized for ensuing news coverage; coverage which, in this case, has put the DSM-5 committee in an awkward position. The members of the committee feel the need to defend the criteria they are developing. Because the DSM-5 committee cannot yet read the Yale researchers’ paper, however, they cannot assess the validity of the results. “It is impossible to talk about it until we’ve seen it, which we’ve not,” Lord said when asked about Volkmar’s and McPartland’s research.
Lined up on a shelf, the previous editions of the DSM, the “Shrink’s Bible,” form a rainbow—each paperback spine adds a new color, a new chapter in psychiatric history and a shift in the perception and diagnosis of autism. Although the disorder was first described in 1943 by a psychiatrist named Leo Kanner, the first two editions of the DSM did not classify it as a distinct illness. Instead, autism was mistakenly said to be a rare subset of schizophrenia that came on in early childhood in reaction to a cold, disconnected “refrigerator” mother. In those days, a child’s diagnosis was a shameful secret to be kept from family and friends. It was considered common knowledge that such children could not be taught, and they were shipped off to institutions, usually for life.
Then, in 1980, the APA released the lime-green DSM-III. Inside was a new section called “Infantile Autism.” For the first time, autism had its own definition and a clear statement of its characteristic symptoms. Researchers began to notice that autism ran in families, across generations, suggesting that genes might play a role in determining who developed it. They also realized that a larger-than-normal percentage of autistic children suffered from seizures, indicating that the disease was rooted in the biology of the brain, not in bad parenting.
Autism as defined by the DSM-III was severely debilitating. Most individuals who received the diagnosis had IQs below 70, in the intellectually disabled range. Few learned to talk or live independently. The individuals who qualified as autistic under this definition made little eye contact and often spent hours alone, rocking back and forth in corners, hands flapping rhythmically by their sides. By the early 1990s, it had become clear that psychologists diagnosing autism in children were leaving out a large group of kids like Jack who have normal and even high IQs, who are often talkative and talented, but who lag behind their peers socially. These children seem to lack an understanding of social rules, fail to form friendships, and often cannot carry on conversations. When they do join conversations, they often talk incessantly about one apparently random topic, such as vacuum cleaners or aqueducts. While these people are better equipped to function in day-to-day life than their more severely affected peers, they often have great difficulty in school, face horrific bullying, and lack the social tact to hold down jobs.
Finding ways to help those people motivated Volkmar and his colleagues as they revised the DSM-III definition. In addition to covering children, the DSM-IV also changed the lives of many adults who had never been diagnosed as they’d grown up with the narrower criteria of the previous DSM.
“Everything in life has been a struggle for me,” said John, who learned only four years ago, at 44, that he has an Autism Spectrum Disorder. John has a steady job as a lineman for Connecticut Light and Power, a wife, and a home, but getting there has been tough. He was held back in the first and tenth grades and eventually dropped out of high school. He finally earned his high school diploma at the age of 23. John’s peers teased him and called him names at school and his alcoholic father did the same at home. Age has done little to reduce the number of taunts thrown at him.
“One guy said to another guy in a conversation in a room, he said ‘Uh, I think John is functionally retarded,’ ” said John, recalling a recent incident with his co-workers. “The other guy said, ‘Well, what do you mean, you think he’s an idiot savant?’ And the other guy says, ‘No, he can’t be an idiot savant, he’s not good at anything, he’s just an idiot.’ ”
Under the DSM-III, John, who asked that his last name not be used for fear of repercussions at work, was too smart, too verbal, and too functional to be considered autistic. As a consequence, John never received help in school. No one ever took the time to try to teach him the social skills he lacked. The DSM-III left him, and many others, out in the cold.
“If you’re not sure what you have, how do you know to handle life?” John asked.
The DSM-IV, with its broadened definition of autism, gave John and others like him a name for their problems. That broader definition came too late to help John in school but has changed his life since then, he said.
“If I run into a situation either at work or at home and it’s a rough course I’m going through,” said John, “I can analyze it and think to myself, ‘Well, wait a minute, that’s the Asperger’s.’ ”
But the breadth that allowed the DSM-IV diagnoses to help people such as Jack and John has led to a large increase in the number of diagnoses. Since the publication of the DSM-IV in 1994, autism and related disorders have gone from being relatively rare, around one in five thousand children at the time of DSM-IV’s release, to alarmingly common. The most recent studies put the prevalence at around 1 in 110 children.
Volkmar believes that the increase in the rate of autism diagnosis means doctors are getting better at identifying the kids who need help. But other psychiatrists disagree. Some believe that the DSM-IV’s expanded definition went too far.
Allen Frances, the former chairman of the DSM-IV Task Force, the umbrella group responsible for the DSM revision process, has repeatedly denounced the DSM-IV for setting off a “false epidemic.” He argues the rate of diagnosis needs to be curtailed.
At stake in the debate is money. Before the DSM-IV, less than 1 percent of students receiving special education services were autistic. Now more than 4 percent are, and the amount districts spend on special education services has skyrocketed. The annual cost of caring for individuals with autism in the United States is now $35 billion.
In 2007, the APA set about revising the criteria again. The association appointed Dr. Susan Swedo, a senior investigator at the National Institute of Mental Health, as the chair of the committee that would deal with the autism criteria. A year later, Volkmar and thirteen other prominent psychiatrists and experts were asked to join the committee under Swedo’s leadership. Soon after, the committee held a small conference for its members at the University of California, Davis.
During the conference, three groups each made recommendations for changing the DSM. One suggested only mild reforms, examining the role of IQ in diagnosis, for example, but made no mention of changing the overall DSM-IV structure. The other two groups, however, came back with much more drastic recommendations. Both advocated eliminating Asperger’s Syndrome altogether and suggested starting throwing away the existing framework entirely. By the spring of 2009, just a year after the conference, Volkmar and another member of the committee had resigned.
Volkmar refuses to discuss his resignation, saying only that he had become “disenchanted with the process.” When asked about the committee, he said, “I’m not saying anything bad about anyone. They’re wonderful people, God love ’em. Just, what a mess.”
Since that conference and Volkmar’s resignation, the DSM-5 committee has followed the recommendations of the groups advocating radical change. In several statements issued to the public, the committee has stated that the distinctions in the DSM-IV among Asperger’s, PDD-NOS, and autism are arbitrary and confusing, like “trying to cleave meatloaf at the joints.” They intend to replace the DSM-IV’s Pervasive Developmental Disorders category with a single disorder, Autism Spectrum Disorder.
Currently, a patient can receive a diagnosis on the autism spectrum if he or she demonstrates only some characteristic autism behaviors. The system is complicated, and there are a total of 2688 combinations of symptoms that would qualify for a diagnosis. The new rules are simpler—there are only six combinations of symptoms that qualify.
While simpler, the new rules may make the standards for determining who receives a diagnosis on the autism spectrum more stringent. A patient will only receive a diagnosis if he demonstrates each and every one of a series of social and communicative impairments. He will also have to have at least two types of restricted or repetitive behaviors—rocking back and forth and obsessively lining toys up in a line, for example.
The committee that has drafted the new definitions maintains that no one who currently has a diagnosis will be cut off. The spectrum of disorders will be just as broad under the new rules as under the current ones, they say. “Really, things looked pretty good,” Lord said, based on the data she has seen. She admits that the field trials examining how the new criteria will affect autism diagnosis rates have been small, and that other, larger studies of the new criteria have not addressed the question of whether the criteria would cut people off the spectrum.
Volkmar calls the committee’s assurances “misleading.” The committee’s public statements have implied “we’re gonna have one big happy family of Autism Spectrum Disorder,” he said. What the committee is not advertising, he said, is that the newly defined Autism Spectrum Disorder will narrow the spectrum back down to a disorder more akin to autism under the DSM-III.
“It’s actually going to be classical autism,” Volkmar says, smacking his palm with the back of his other hand. He is frustrated that the new criteria may narrow the autism spectrum back to intellectually disabled individuals, excluding those with higher IQs but significant social impairments.
Shortly after Volkmar left the DSM-5 committee, he and McPartland decided to go back over diagnostic assessments carried out while testing the DSM-IV criteria and examine how the children who participated in the study would have been scored had they been using the proposed DSM-5 criteria. As they disclosed at the conference in Iceland, 56 percent of patients with normal or high IQs (above 70) who received a DSM-IV diagnosis of autism, Asperger’s, or PDD-NOS would not have qualified for a diagnosis of autism spectrum disorder according to the new standards.
Similarly, this June, a group of researchers from Finland announced that only 46 percent of a group of children who were diagnosed with an autism spectrum disorder under the DSM-IV met the new criteria for autism. In another study, researchers in London examined diagnostic records of children with PDD-NOS. Almost none of the children displayed enough rigid, stereotyped behaviors to meet the proposed criteria for Autism Spectrum Disorder.
Lord argues that these studies all have a “major inherent flaw.” Some of the symptoms the DSM-5 asks clinicians to check for, she says, were not included in the DSM-IV at all. If clinicians never asked about these symptoms when the diagnoses were first made, there is no way to know whether a child evaluated under the DSM-IV had them or not. For this reason, she says it is impossible to determine from previously collected data how the new criteria will affect autism rates. Still, she said, “We really want to know who didn’t meet and why.” She said the committee is open to making changes if its members can be convinced that many patients will be cut off the spectrum. But, “For that, we need information,” she said. “We don’t need polemics.”
Lord said that the DSM-5 is strictly a scientific description of symptoms and behavior and that her colleagues can’t be expected to decide social policy for the country. DSM does not directly determine who gets access to services. “If you need treatment, you should get it, with or without a label,” she said.
But Jennifer Laviano, a special education lawyer in Sherman, Connecticut, noted that bureaucracies such as school districts that regularly care for autistic people depend on authoritative diagnoses. “While there should be no difference in the services a child receives based on the label the child is given, the reality is that there are many important reasons that label matters.” An autism diagnosis can give parents more bargaining power in their meetings with school board officials and can help educators understand the obstacles a child faces.
School districts spend three to six times more on educating a child with autism than they would on educating a child who receives no special services. “School districts generally will use almost any excuse they can not to identify a child, especially if it’s going to be a costly program,” Laviano said. She also said that it is especially difficult to convince school districts to provide services for children who have high IQs but who are socially impaired. “School districts will say, ‘Oh, well, academically, he’s on grade level,’ or ‘If you look at his Connecticut Mastery Test scores, he’s always meeting goal,’ ” she said. Although the laws governing special education cover functional and adaptive skills—such as classroom behavior, cooperation, and learning to take care of oneself—as well as academics, schools do not usually prioritize helping socially impaired children.
Andreana Bellach, a lawyer in Stamford, Connecticut who represents Connecticut school districts, argues that, often, more intelligent children don’t need special education services, even if they are awkward and uncomfortable in social situations. Such children benefit from “high cognitive functioning as well as their ability to be resilient and apply what they’ve learned,” Bellach said.
Often, parents find that the fight for support in school, for classroom aides, therapies, and special social-skills training, goes hand-in-hand with the battle for a diagnosis. The biggest weapon parents have is a federal law, the Individuals with Disabilities Education Act, which guarantees a “free and appropriate public education” to all children with disabilities. In order to be eligible for the school services mandated by the law, however, a child must fall into one of thirteen categories. Autism is one of them.
Decisions about special education services are made by a tribunal of school administrators, teachers, special education providers, the child’s parents, and often lawyers. The group is called a Planning and Placement Team, and it determines what kinds of services are included in each child’s Individualized Education Program. Often, the members of the group disagree on what services a child should receive. Parents want their children to get as much help as possible, and schools need to conserve resources. Parents come to the meetings armed with every shred of information that could help prove that their child has significant challenges. Having an official diagnosis makes a difference.
“The reason that we have diagnoses is to help people who need help to get help,” said McPartland, the Yale researcher who presented alongside Volkmar in Iceland. Lord argued that the committee’s responsibility is to science, not to society, but McPartland wishes that she and the other committee members would think more carefully about the effects their work will have. “We’re at the intersection of clinical decision making and public policy, and I don’t know how it’s going to be decided,” he said.
The qualities that Bellach described are visible in Winegardner’s son Jack, who, despite his social difficulties, has a high IQ and performs above grade-level on his standardized tests.
“We struggled a long time and fought for a long time to get an autism diagnosis,” Winegardner said. Jack, who at six had so much trouble passing out valentines to his classmates, is now eight years old. He was diagnosed with PDD-NOS when he was five, but even under the current DSM-IV criteria, getting the label and more importantly, school services, was a struggle.
“We were told by a lot of people, ‘Oh he doesn’t present as a kid with autism,’” said Winegardner. Still, by the time Jack was two and in a preschool class, she knew something was wrong. “I started to notice that he wasn’t interacting with his peers and he wasn’t talking to the teacher,” she says. In fact, Jack wasn’t talking at all. Jack’s teacher voiced concerns and encouraged Winegardner to seek help. But when Jean took her son to be evaluated by the Maryland Infants and Toddlers Program, the state’s agency for very young children with special needs, he did not qualify.
“They were like, ‘Oh he’s a boy, he’s a second child, he probably doesn’t have to talk as much,’ ” Winegardner recalled. Her older son had been a late talker as well, and the evaluators suggested she was probably overreacting. When Jack turned three and was still struggling in the classroom, Winegardner again sought to qualify her son for special services, this time at Montgomery County’s special education department, Childfind. Again, she was disappointed.
“He works really well one-on-one with adults, and that was what the test was, so he performed really well,” said Jean. “It wasn’t really indicative of how he was functioning in the larger world.” When a psychologist finally visited the school, she agreed with Jean that something was wrong. Jack did not interact with his classmates but spent most of his time playing by himself. He was obsessed with trains. He could speak, but did so rarely. Following the psychologist’s recommendation, the school system moved Jack to a special-education preschool.
Without an autism diagnosis, however, Jack would only be eligible for services until the end of kindergarten. With a diagnosis, he could receive help until his 21st birthday or high school graduation.
Just after Jack turned five, he got a diagnosis. Jean enrolled him in a study at the National Institute of Mental Health in Bethesda, Maryland, and during the screening process for the study, doctors gave Jack the PDD-NOS label. Armed with the diagnosis, Jean was able to get her son an Individualized Education Plan, which provided a classroom aide and accommodation on assignments. Perhaps most importantly, the diagnosis helped Jack’s peers and teachers understand him.
“An autism diagnosis and identifying my son as PDD-NOS has given us a starting place and a map and a path,” said Winegardner. Although coming to terms with Jack’s disability has been difficult, the diagnosis has given her a way to get Jack help. But Winegardner worries a lot about what will happen when the DSM-5 comes out and what will happen if the criteria become narrower.
“No one would look at Jack now and say he’s not autistic,” Winegardner said. “But then again,” she hesitated, “there will always be somebody who falls just outside.” She just hopes it won’t be her son.