When Lorna Wing wrote her paper on Asperger’s in 1981, figures on the prevalence of autism were scant. She ballparked it15 at about 4 or 5 per 10,000, and guessed that maybe half of those were the kind of patient described by Asperger. When Volkmar’s team conducted their field trials, there were no better statistics for Asperger’s because, of course, the disorder didn’t yet exist to be counted. But even if Volkmar’s group was impressed enough by those fifty-one cases to push for inclusion, and even if they knew that adding Asperger’s would likely add patients to the sick rolls, they certainly didn’t think they’d stumbled on an epidemic. At best, they had identified a small if significant number of cases that could be carved out of another small number.
So imagine their (and everyone else’s) surprise when Eric Fombonne, an epidemiologist working in England, reviewed16 twenty-three surveys of autism prevalence conducted between 1966 and 1999, and concluded that the rates were on the rise. Estimates varied wildly—from 0.7 percent in a 1970 study to 21 percent in 1996, but one factor alone seemed to affect the overall trend. Prior to 1989, the average study turned up 4.3 cases in 10,000 people; after that year, the rate was 7.2 in 10,000. Even accounting for all other factors—age, gender, severity, IQ, and so on—the relationship between year of study and rate of autism was statistically significant. Time alone, in other words, was making autism rates rise—and not, so the epidemiologists said, simply because the clock was ticking or because there was what Fombonne called a “secular increase17 in the incidence of the disorder.” Instead, he thought, the increase was the result of “improved recognition and detection . . . together with a broadening of the diagnostic concept and definitions.”
People like Volkmar might have thought the territory they had established with the Asperger’s diagnosis was sparsely populated, but epidemiologists were reaching a different conclusion. Using the DSM-IV criteria, one group reviewed the medical and educational records of schoolchildren in Atlanta and turned up a rate of 34 per 10,00018, a tenfold increase over the few comparable studies that had been conducted in the United States.
Fombonne, in an editorial accompanying the Atlanta study, speculated that even these results were “an underestimate19.” And subsequent studies proved him right. As of 2002, the CDC reported20, 1 in 150 children had an autistic spectrum disorder, a rate that by 2006 had reached 1 in 110, or just over 90 per 10,000. It was around this time that Lorna Wing started referring to the publication of her original paper as “opening Pandora’s Box21.”
The numbers just kept going up. By 2008, the CDC said, the rate was 1 in 88, and in May 2011, researchers writing in the American Journal of Psychiatry announced they had found a prevalence rate in a city in Korea22 of 2.64 percent. For those of you keeping score, that’s 264 per 10,000—and 374 for every 10,000 boys—a greater than fiftyfold increase since 1981.
And people were keeping score—including the people at Autism Speaks. “The world’s largest autism science and advocacy organization,” according to its website, it sponsored the Korean study and, soon after its publication, blitzed the media23 with the claim that 1 in 38 children was autistic. This, the group said, indicated “the need for improved and wider autism screening24 among the general population”—a move that, to judge from the last twenty years, could only increase prevalence rates, and with them public attention, sympathy, and, of course, money for their cause.
Opinions varied on what was causing this increase. To those who held mercury-based preservatives in common childhood vaccines responsible, the increasing rates were further proof that vaccination policies were ruining kids’ brains. Most explanations, however, focused on the old and elusive question of whether epidemiologists were getting better at detecting a disease that was already there, or if they were charting the rise of an epidemic.
But the researchers in the Korean study offered an interesting alternative explanation. They noted that only 63 percent of the people they had solicited had agreed to participate in the study—a disappointingly low number, far lower than the 80 percent considered optimal, and often achieved, in community surveys. But, they argued, if you consider the tendency among Koreans to think of autism as “a stigmatizing hereditary disorder25” that “impugns the child’s lineage on both sides and threatens the marriage prospects of unaffected relatives,” then this level of participation seems not so bad. In fact, it’s remarkable: something had inspired nearly two-thirds of the parents in a Korean city to risk having their families marked as members of a defective gene pool.
The researchers focused their speculation on Korea’s “highly structured” educational system, with its emphasis on “behavioral regulation” in twelve-hour-long classes that meet six days a week. Some children’s problems would be impossible to conceal in such a setting, but “for quiet, high functioning children with ASDs,” they wrote, “this environment may reduce the likelihood of referrals to special education programs.” Still, even if they remained invisible at school, these children’s problems might have been obvious at home, and the survey “may have provided welcome opportunities” for parents caught between their concern for their kids and their fears for their families’ reputation to get information and maybe even treatment. “It could be argued that in this cultural context,” the researchers wrote, “parents of children with serious but unrecognized developmental problems would be more likely to participate in research than parents of unaffected children.”
This wasn’t the first time a researcher attributed increasing prevalence to motivated parents seizing an opportunity. The Atlanta study’s authors had been disturbed by a finding that the likelihood that a child would qualify for a diagnosis increased with age—but only until they got to the oldest group of children, who were nine and ten. Did this mean that younger children were developing autistic disorders at an increasing rate? Had they found the leading edge of the epidemic?
The scientists didn’t think so. “Lower rates in 9- and 10-year-olds26 may reflect the use of narrower diagnostic criteria for autism before the publication of DSM-IV,” they wrote, “and the increased availability of educational, health and social services for children with autism.” Perhaps the older children hadn’t escaped an epidemic, but rather had been born too late to benefit from the DSM-IV’s relaxed criteria and their intersection with the 1991 mandate issued by the U.S. Department of Education to provide special education services to children with autistic disorders. This, the researchers allowed, was “possibly leading to increases in the number of children classified with autism because of the availability of services.”
“Diagnosis of ASDs27 [autistic spectrum disorders] is an uncertain art,” wrote autism expert Tony Charman in an editorial accompanying the Korean study, “and in the end one has to make a clinical judgment of whether the number, quality, severity, and accompanying impairment of symptoms are sufficient to fulfill DSM criteria.” As in any diagnosis, the need for judgment about ASDs increases when the disorder is milder; a kid who rocks in the corner and can’t talk is much easier to diagnose than someone like Nomi or CC. But milder cases like theirs account for nearly two-thirds of the diagnoses that make up the prevalence rates.
It’s no mystery why parents would seek out a diagnosis, as Nomi’s and CC’s parents did—and, like them, to feel relief when the doctor issues it. After all, it’s not as if parents have to be encouraged to worry about how their children are faring socially, or to wonder if they will ever learn to negotiate all the complex and unspoken rules of their little societies, or to think that disaster awaits when they cannot. Even if they haven’t read Lord of the Flies, parents know just how brutal those politics are, and how ill suited some children are to them. In a marketplace of worried parents searching for resources, a diagnosis like Asperger’s syndrome is bound to be a hot commodity.
And really, how hard is it to sell people on a disorder whose “symptoms” can bestow what Lawrence Osborne called “extraordinary gifts”? Osborne wasn’t
talking about special education dollars, but about Glenn Gould and Vladimir Nabokov and Ludwig Wittgenstein, all diagnosed—posthumously, of course—into a Hall of Fame that, depending on which website you consult, also includes Isaac Newton, Emily Dickinson, and Albert Einstein.
Even for those who will fall short of greatness, the diagnosis has its appeal. And its charms are on the increase, at least according to the Aspie who predicted to Osborne that “society will actually become more and more dependent on people with Asperger’s to usher it through the difficulties ahead.” That was more than a decade ago, and since then, interaction has come to rely less and less on the nonverbal cues that Aspies are so bad at decoding and more and more on the tablets and handhelds, the binaries of emoticons and tweets, that they are more suited to. While the neurotypical among us grow more and more bewildered by the barrage of information fed to us by our devices, the burdens of Asperger’s, at least in the view of some of the diagnosed, increasingly become a gift. Even for those who think this is an ominous sign—like the novelist Jonathan Lethem, who speculated that people with Asperger’s are “canaries [who] sensed before anyone else28 that we’d entered a coal mine”—the fit between disease and society is striking, the temptation to claim the label obvious.
This may be the real surprise of the Asperger’s outbreak: that it is neither a “secular increase” nor an artifact of counting, but rather a supply-side epidemic. Unlike the childhood bipolar outbreak, this one was not set off by a drug company and its shills, exploiting loopholes in the DSM for profit. This one came right from the authors of the book, who wanted nothing more than to help kids help themselves to social resources. They had built a diagnosis, and the patients had come.
• • •
Volkmar thinks that the rush to get Asperger’s into the DSM-IV might have had something to do with the outcome. “Had we had a little more time, we would have been able to have a better definition and criteria for Asperger’s,” he told me, which might have made the diagnostic threshold a little harder to cross. But even so, Volkmar is glad Asperger’s made the cut. “I don’t regret getting it in,” he said. “You need to have a label to get services, and this is a population that needs them.”
Michael First also thinks it was “premature to add Asperger’s29, given the limited amount of data,” he told me. “You could argue that this was a deviation from DSM-IV’s conservative approach.”
That’s an assessment with which Allen Frances agrees. “We probably were premature30 in introducing Asperger’s,” he acknowledged. Childhood disorders, he said, were outside his expertise. “Had a comparable situation arisen on the adult side, I would have been more insistent on the strict application of the DSM-IV rules.” It’s a lapse that he now regrets. “Had I been smart enough to predict that the diagnosis would be made so loosely, that parents would panic and stop vaccinating their kids, and that school services would be based just on the presence or absence of the diagnosis, then I definitely would have insisted on a stricter application of the rules.”
Regret over the Asperger’s diagnosis, or at least over the way it had become ubiquitous, was one thing that Frances and the DSM-5 leadership agreed on. David Kupfer told The New York Times that the swelling ranks of Asperger’s patients were getting to be a problem. “It involves a use of treatment resources31,” he said. “It becomes a cost issue.” You don’t want to reify your diagnoses, but on the other hand, if a diagnosis is nothing more than a label that gets services . . . well, there is such a thing as too much success, especially when resources are scarce, costs are high, and public credibility is slipping.
The DSM-5 proposal sure looked like a way to raise the bar for entry. DSM-IV’s Autistic Disorder had listed twelve criteria under three main headings and required that patients meet only six of them, with at least two coming from the first category and at least one each from the remaining two; the Asperger’s diagnosis required two from Criterion A’s four symptoms and one from the four listed under Criterion B. (One math-minded psychiatrist calculated that there were 1,256 combinations of symptoms that would meet the criteria.) DSM-5, on the other hand, required all three Criterion A symptoms to be met and two of four in Criterion B. But even if it seemed like a no-brainer that moving from a Chinese menu to prix fixe would reduce the numbers of diagnoses, the APA insisted this was not the purpose of the change. “The goal was not to change prevalence32,” work group member Catherine Lord told me. After all, gerrymandering a diagnosis to solve a public policy, or a public relations, problem—that would be cheating.
Besides, Lord told me, “we think these prevalence rates are accurate.” There was a problem with accuracy—not with the numbers of diagnoses, but with what doctors meant when they rendered them. “We want to make the criteria match up better with what clinicians are seeing when they make the diagnosis,” she said. The problem with criteria that were too vague and broad was not that doctors could use them too liberally—say, to make sure a kid who was a little odd got special services in school. Rather, Lord said, when they rendered their diagnoses, it wasn’t clear exactly what, if anything, the criteria had to do with the conclusion. In fact, the most important factor in determining which diagnosis33 a kid got, at least according to a study Lord had conducted, was the hospital that did the evaluation. Doctors had reverted to the old pre-DSM-III way: deciding which illness a patient had through some combination of local custom and intuition, rather than by systematically applying the criteria.
But only astute clinicians could be trusted with this uncertain art. The rest of us—the parents who want their kids to qualify, the geeks who want to inherit the earth from the neurotypicals, the people who would prefer having a mental disorder over being a jerk—needed to have their leash yanked. “People use ‘Asperger’s’ and ‘autism’ colloquially,” Lord told me. Tightening the criteria would give the masses “who don’t know much about autism less of a feeling that everyone who’s socially awkward has autism.”
On the other hand, if those people want to call themselves Aspies, that’s fine by Lord. “We’re not trying to take away that identity,” she said. “It’s very helpful to some people. It’s just not a medical diagnosis.” Still, it was clear that the APA could take away what it had given—not the name, but the rights and privileges it had inadvertently granted by enshrining Asperger’s in the DSM. And, evidently, when the diagnosis got out of hand, that is exactly what it would do.
Lord reassured me that this wasn’t the problem it seemed to be. First of all, she said, “people with Asperger’s vacillate between wanting to describe themselves as having a disorder or not.” So deleting the diagnosis would actually help people like Nomi Kaim and CC Carley by providing the privileges without the burden and by removing the stink of pathology from their identity. As for those privileges, kids in special education have to get reevaluated every three years, so they are always at risk of losing their diagnoses. Of course, generally that would be because their condition has changed, not because it has disappeared from the DSM. But Lord was quite certain that the only problem patients would face was having to move from one diagnostic home to another. The overall prevalence rates, she said, were very unlikely to come down.
That was a question that would have to wait for field trials of the new diagnoses. But as the APA geared up for its annual meeting in May 2011, it was becoming evident that the field trials were in disarray, and even as their significance was increasing, they might not be able to shed much light on questions like these—even to the very few people from whom the results would not be kept a secret.
Chapter 14
In early May 2011, the APA posted its revised draft of the DSM-5. Asperger’s syndrome was still slated for removal, as was the bereavement exclusion. But Psychosis Risk Syndrome (PRS) and Temper Dysregulation Disorder were gone. In their place were two new diagnoses—Attenuated Psychosis Symptoms Syndrome (APSS) and Disruptive Mood Dysregulation Disorder. The ingredients of the new dis
orders were exactly the same as the old ones. Only the labels had changed.
The renaming strategy had its obvious advantages. It got David Shaffer off the hook for suggesting a name that seemed to medicalize ill temper, even as the criteria continued to medicalize ill-tempered kids. It also allowed Will Carpenter to deflect attention from the research showing that kids whose illness consisted of being on the verge of psychosis rarely went on to get psychotic. On further consideration, he told me, his work group had realized that PRS really wasn’t a “prevention concept” at all, but rather “a diagnostic concept,” not an illness about to happen, but one in progress right now.
That explanation raised the question of why, if PRS was a disorder in its own right, the work group hadn’t proposed it as a “diagnostic concept” in the first place. It also didn’t explain why, according to the website, the rationale for introducing the diagnosis was that “young people at risk for later manifestation of a psychotic disorder can be identified,” or why APSS was still described as “a risk syndrome for psychosis.” Perhaps the APA’s marketing department and editorial department weren’t talking to each other.
“People say, ‘What’s in a name1?’” Dilip Jeste, the future president of the APA, told three hundred of his colleagues the week after the changes were posted. “But it seems the name means a lot to lots of people.” Jeste was talking about Mild Neurocognitive Disorder, a disorder once known as Minor Neurocognitive Disorder. He explained that many people who wrote into the website had complained to his neurocognitive disorders work group that minor was a “trivializing” word. “The people pointed out that if you call something minor, insurance companies will not reimburse,” he said.
The Book of Woe: The DSM and the Unmaking of Psychiatry Page 22