The Book of Woe: The DSM and the Unmaking of Psychiatry

by Gary Greenberg

  “They had this idea that we had the wrong people doing the DSM, and if we had the neuroscientists and geneticists around, they’d tell us all we need to do to make a paradigm shift,” First said.

  Despite his doubts, First organized the conferences, attended all of them, wrote summaries of the proceedings, and designed a website on which to post them. He might have been sure those scientists would never find what the APA was looking for, but he soldiered on. He no doubt did this out of loyalty—to his profession and to his discipline. But he had another reason. The APA’s aspirations to meet the scientific demands of the day might have been quixotic, its ambitions outsize, but First had an ambition of his own: to lead the effort to produce the DSM-5.

  Chapter 5

  Michael First wasn’t the only old DSM hand skeptical about the prospects for the planning conferences. “When I heard about them1, I was amazed,” Allen Frances told me. “It was absolutely ridiculous from the beginning. There was no way you could force a breakthrough like that.” If scientists had made those seminal discoveries connecting mental disorders with brain function, Frances thought, the paradigm would have already shifted. The attempt to “jump-start science” was part of a “grand ambition that will take many decades to realize.”

  It was that kind of ambition that Frances attempted to dampen with his conservative approach to DSM-IV. And First had set his sights similarly low for DSM-5. “However much we don’t like this paradigm, it’s as good as we have,” he told me. “So let’s find out how to make it more helpful to clinicians.” For all their talk about clinician communication, the defenders of the DSM have precious little idea of exactly how it figures in the way psychiatrists talk to one another, and how it can help them do a better job of it. They’ve spent a lot more time dreaming up new diagnoses than looking into the far less glamorous matter of clinical utility. First thought the APA’s time and money would be best spent tweaking the DSM-IV’s criteria and correcting its obvious errors, while focusing on making the manual maximally useful to clinicians.

  Neither First nor Frances seems conservative by nature. Their politics, to the extent they discuss them, appear to be well left of center. But when it comes to the DSM, they see themselves as stewards of a tradition that, even if imperfect, it is important to uphold, possibly because both men have witnessed the consequences of playing fast and loose with the DSM.

  Perhaps the most disturbing of these consequences was unfolding as the preliminaries to DSM-5 were getting under way. The trouble began at Harvard’s Massachusetts General Hospital, where Joseph Biederman worked as a child psychiatrist. In the 1980s, he developed a stellar reputation as a researcher and clinician working with children diagnosed with Attention Deficit/Hyperactivity Disorder (ADHD). Author of more than six hundred articles, winner of numerous professional awards, a perennial choice as a U.S. News & World Report Top Doctor, and an inductee in the Children and Adults with Attention Deficit/Hyperactivity Disorder Hall of Fame, Biederman was what is known in his industry as a key opinion leader.

  Sometime in the 1980s, Biederman’s opinions turned to a particularly difficult group of his ADHD patients. These were kids who, in addition to being fidgety and distractable, were from a very early age—sometimes, so their parents said, from the time they were born—quick to rage, slow to be comforted, and precocious in all the wrong ways: threatening suicide before their peers even knew what the word meant, acting out sexually before they’d reached puberty. They were defiant and contrary and cranky all the time, hard for parents to parent, teachers to teach, and therapists to treat.

  Biederman thought he detected in these children2 something different from what he saw in other ADHD patients. In particular, he saw what the DSM called a mood disorder in their chronic irritability, their prolonged and frequent tantrums, and the profound sadness that set in after the storm. ADHD diagnostic criteria did not include mood symptoms, but there was a small literature reporting a few cases of “hyperactive children3” who also had “nervous irritability” and who went on to develop one of the best-known and most feared mood disorders of all: what had been called manic-depressive insanity by Kraepelin and, in the 1980s, was renamed Bipolar Disorder (BD). Doctors generally considered it to be neurological or genetic in origin and incurable, although it could be managed with mood stabilizers, usually lithium.

  Biederman thought that he was on the verge of a major discovery: that BD, the onset of which was generally agreed to be in early adulthood, started much earlier. This would indicate that these kids ought to be treated with mood stabilizers rather than the stimulants generally prescribed for ADHD and that they and their parents should be prepared for a lifetime of managing a chronic illness, just as juvenile diabetic patients are.

  The appeal of this hypothesis to doctors and parents was immense. It invoked one of the most time-tested diagnoses in the DSM, it gave clinicians something to call these kids’ condition besides “bad ADHD,” and it pointed in the direction of a new treatment. But there was a problem. According to the DSM-IV, “the essential feature of Bipolar Disorder4 is a clinical course characterized by the occurrence of one or more manic episodes.” And manic episodes have what the DSM calls Criterion A symptoms, features necessary, but not sufficient, for the diagnosis—in the case of BD, “a distinct period of abnormally and persistently elevated5, expansive, or irritable mood, lasting at least 1 week.” But Biederman’s patients didn’t have episodes of mania—or of anything else for that matter. Indeed, the parents’ major complaint was that they never got a break from their children’s biliousness.

  The DSM also lists Criterion B symptoms for most disorders—features that must be in place in addition to Criterion A for the diagnosis to be made. Manic episodes have seven Criterion B symptoms6, four of which are required for the diagnosis. And three of those seven symptoms—excessive talkativeness, distractibility, and fidgetiness—also appeared on the list of possible symptoms of ADHD. So it was hard to distinguish the two disorders. Clinicians following Biederman’s lead might be only slapping a new label on “bad ADHD,” one that moved the patients to a different, but still ill-fitting, category. This kind of ad hoc diagnosing is exactly what the DSM, with its symptom lists, is supposed to preclude.

  So Biederman set out to prove7 that the new label was a winner, that it could reliably gather together particulars once thought to be scattered, and point in the direction of a previously undiscovered natural formation. If he looked only at the chronically irritable patients, he wondered, would their Criterion B symptoms differ from the ones that the rest of the ADHD kids had? If so, then this would be evidence that these kids constituted a different diagnostic group from the others.

  Sure enough, when Biederman and his team looked carefully at the Criterion B profiles of the chronically irritable group, they found that they were more likely to have the four symptoms that BD does not share with ADHD than children who weren’t irritable. The team also looked at levels of impairment among the irritable group—the extent to which the patients’ symptoms interfered with their lives, landed them in hospitals, or led to psychosis—and found a similar grouping: the children who were chronically irritable and scored high on the four unique Criterion B symptoms tended to be more impaired than the ADHD kids. Taken together, Biederman said, patients with this distinct profile account for around 20 percent of children diagnosed with ADHD. One in five of those patients8, Biederman concluded, actually was bipolar and was getting exactly the wrong treatment: stimulants known to aggravate mania.

  Biederman’s announcement provoked9 an outcry from his colleagues. They argued that his proposal flouted hundreds of years of observations about the episodic nature of mania. In addition, they pointed out, a recent epidemiological study, conducted by people other than those trained and supervised by Biederman, had turned up exactly zero children with mania. Biederman’s own research indicated that his patients were not only more irritable than ADHD patients, but also more withdr
awn and prone to sulk—hardly consistent with a diagnosis of mania. Critics also complained that by rejiggering the criteria, he’d lowered the threshold for what was a very serious diagnosis. The Criterion B symptoms that remained after eliminating the ones that overlapped with ADHD—“grandiosity, decreased need for sleep, flight of ideas (i.e., a free-flowing stream of consciousness) and excessive involvement in pleasurable activities that have a high potential for painful consequences”—weren’t these really just a working definition of childhood at its most exuberant? And finally, they cited longitudinal studies, which showed that plenty of “bad ADHD” kids indeed went on to develop various mental disorders, but bipolar was not among them—a finding hard to reconcile with the presumption that BD is a lifelong illness.

  I’ll spare you the ensuing back-and-forth, which is as bitter and rancorous, and as impenetrable, as most academic controversies, and which continues more than fifteen years later. It’s not that it hasn’t been entertaining, at least at points, as when Biederman was moved to liken his critics to people who insist the earth is flat and circled by the sun, and his own discovery to that of Edward Jenner, whose “smallpox vaccine was ridiculed10 when initially proposed”—suggestive comparisons for a man studying a disorder with grandiosity among its symptoms. It also illustrates the bruising politics behind the DSM, the way in which changing it is as much a legislative as a scientific process, and the self-validating nature of diagnosis, by which once you’ve created a diagnostic category, the fact that people fit into it becomes evidence that the disorder exists. But what happened next is of much greater significance, so suffice it to say for now that Biederman proved, to his own satisfaction, that he was correct, that those “bad ADHD” kids really had BD.

  Biederman didn’t try to change the DSM-IV definition of Bipolar Disorder to reflect his findings. The book had just come out when he began his campaign to convince doctors and parents that chronic irritability was the juvenile form of episodic mania, and a revision wasn’t expected for many years. Fortunately for him, however, he didn’t really need it to change. The DSM provides, in addition to the seven variations on BD—Bipolar Disorder, Single Manic Episode; Bipolar Disorder, Most Recent Episode Depressed; and so on—a category called Bipolar Disorder Not Otherwise Specified (BDNOS). (There are NOS categories for every major diagnostic classification in the book.) The BDNOS category is designed for “disorders with bipolar features11 that do not meet criteria for any specific Bipolar Disorder.” In other words, if your patient doesn’t qualify for the diagnosis, but you still think he’s bipolar, you can just go ahead and give the kid the diagnosis anyway.

  • • •

  By the time children get to a doctor of Joseph Biederman’s stature, they’ve generally been through any number of therapists, pediatricians, and psychiatrists. They’ve been the subject of countless meetings at schools, endless testing by psychologists, and home interventions from social service agencies. Their parents have tried everything—drugs and diets, hug therapy and tough love, private schools and residential treatment and family therapy vacations, prayers, and even exorcisms. And in the end they still have a child who seems unhappy nearly every waking moment, who is likely to fly into a rage at the slightest provocation, who holds a knife to his own throat and threatens to plunge it in if he isn’t allowed ten more minutes on PlayStation. They are, in short, devastated and desperate.

  That’s why, when the important doctor tells the beleaguered parents that their child is irrevocably mentally ill, even if they are aware (which they usually aren’t) that he’s stretching the truth, they listen. Besides, the news isn’t all bad. At least now they know. They have been given a name for their (and their child’s) pain. And when the key opinion leader tells his colleagues what he’s doing at Harvard for these kids and offers them the opportunity to provide the same kind of explanation and hope to their own patients, they are quick to follow him through the NOS loophole.

  And by 1999, parents could even go to their doctors and suggest they do exactly that. That at any rate was the advice of Demitri and Janice Papolos, a husband-and-wife team whose bestselling book The Bipolar Child explained that “thoughtful clinical investigators12 are beginning to realize that bipolar disorder in childhood presents a very different pattern” from the adult version. Unfortunately, they wrote, only psychiatrists conversant with the “latest research findings” would recognize the symptoms13—the unsettled infancy, the precociousness and separation anxiety, the nightmares and fear of death and mercurial moods, the sweet tooth and bed-wetting and maybe even the in utero kicking, or any of the forty or so other signs that your child is bipolar that are listed in the book—and issue the proper diagnosis.

  By 2001, parents could comfort themselves and their diagnosed children with books such as Brandon and the Bipolar Bear, in which Dr. Samuel explains to Brandon why he (Brandon, not Dr. Samuel) dismembered his teddy bear. “You have bipolar disorder14,” the doctor says. He explains that Brandon has a harder time controlling his feelings because the chemicals in his brain “can’t do their job right15 so their feelings get all jumbled up inside.” He tells Brandon that he doesn’t have BD because he is bad, nor did he get it from a classmate. Rather, he says, Brandon most likely inherited it16.

  Websites sprang up, self-help groups formed, foundations were funded, and in August 2002, Time devoted its cover to the “Young and Bipolar17” and their families, who had been saved by receiving the diagnosis. Soon, the magazine predicted, a “blood test that will allow bipolar disorder to be spotted as simply as, say, high cholesterol” would shorten the period of incorrect diagnoses and fruitless treatments. By 2003, Biederman’s opinion had led his colleagues to conclude 6.67 percent of office visits18 made by children for mental health problems with a diagnosis of BD—up from less than half a percent in 1994.

  It was possible that the new diagnostic approach had uncovered a previously unknown epidemic. But as Duke University’s chief of child and adolescent psychiatry, John March, told The New York Times when the diagnostic increase was reported, “The label may or may not reflect reality19.” No one questioned that there were many children whose explosiveness and irritability terrified their parents and defied treatment. But did they really have BD?

  But March had it backward. The diagnostic label had been redesigned specifically to reflect reality, and reality had followed suit; those explosive children were now bipolar patients, and the sooner parents thought of them (and taught them to think of themselves) as afflicted with a chronic disease, the sooner they could get help. If Biederman had set out to become the Johnny Appleseed of an entirely new disorder—“scary impossible child disorder,” let’s say—rather than of a new version of an old illness, then the websites and self-help books and cover stories would be offering support for kids diagnosed with that disease, doctors would be arguing about whether or not it reflected reality, and parents would be wondering whether or not their kids warranted the SIC label.

  But that’s not what Biederman did. Neither did he tinker with, say, ADHD or Oppositional-Defiant Disorder (ODD), or simply suggest that clinicians use a perfectly good DSM diagnosis—Disruptive Behavior Disorder, NOS (312.9) seems well suited—and be done with it. He wasn’t determined only to give these children a new diagnostic home. He also had a very specific idea about where they belonged, and if the evidence didn’t support his conviction, then he would change the rules by which the evidence was admitted.

  It’s not clear why Biederman settled on Bipolar Disorder. But it’s easy enough to imagine the possible motives: to ensure that insurance companies would pay for the extensive treatment these kids needed, which a diagnosis of ADHD or ODD did not always justify, but BD does; to make his professional mark by changing one of psychiatry’s most venerable diagnoses; to confirm a hunch. But the label also reflected another reality: that, as Dr. Samuel tells Brandon, there is “some medicine that could help20 you feel better.” Indeed, just as Biederman was star
ting to persuade his colleagues to be on the lookout for childhood mania, new treatments were coming to market to supplement the old standby, lithium. Drug makers were touting anticonvulsants such as Abbott Laboratories’ Depakote, and rebranding atypical antipsychotics21—Zyprexa, Seroquel, Abilify, and Risperdal—as mood stabilizers, the category to which lithium belonged and surely a less terrifying term.

  These treatments were not without their problems. They are sedating—heavily so in the case of the antipsychotics—which means that it is hard to know if they were treating BD or just tranquilizing the children. They are also associated with devastating side effects22: cataracts, obesity, diabetes, tardive dyskinesia (a movement disorder characterized by tics and spasticity), which add up to a twelve-to-twenty-year decrease in life expectancy23 for treated versus untreated patients. And even as the drug companies began to seek (and eventually receive) FDA approval to use the drugs as mood stabilizers and to rechristen them accordingly, they did not study their effectiveness (or their side effects) in children—largely because BD wasn’t thought to affect kids.

  Biederman and others did run some studies indicating that children’s symptoms improved24 in response to the mood stabilizers. But these were short-term trials, often eight weeks or less, and could not possibly explore the consequences for a developing brain of daily use of a powerful drug. Nor could the studies tease out the question of whether the results were due to the overall sedative effects of the drugs or to something specific to these patients. Neither did advocates of mood stabilizers for children talk very much about the obesity and resulting diabetes that were known side effects of the drugs. None of this ignorance put the brakes on the bipolar express, however. Diagnosis and treatment rates continued to soar—by 2003, prevalence of BD among children25 had increased fortyfold over a decade and, hardly coincidentally, by 2005, antipsychotic use in children and adolescents26 had grown by 73 percent in only four years. In 2007 alone, half a million children, twenty thousand of them under six years old, were prescribed drugs that a decade before would have been prescribed only in the most dire circumstances. Biederman’s diagnostic innovation was a runaway hit.