How We Get Labeled As Crazy

Pardon the personal question, but have you ever had a sexual fantasy involving the use of a nonliving object — Anna Kournikova's tennis outfit, say, or Tom Cruise's Risky Business skivvies? Actually, don't answer that — we really don't want to know — and you should probably think twice before telling your therapist. She might diagnose you with fetishism, which is listed along with schizophrenia and bipolar disorder in a curious but extremely influential book called the Diagnostic and Statistical Manual of Mental Disorders, or DSM for short.

The DSM lists the criteria used by mental-health professionals to make their various diagnoses, from "mild mental retardation" (the first listing) to "personality disorder not otherwise specified" (the last); there are more than 350 in all. Hence this 943-page doorstop is one of the most important books you've never heard of. And the inscrutable process of writing it is starting up again. The American Psychiatric Association (A.P.A.), the manual's publisher, recently began planning a giant review of the book. The new edition, the fifth — called DSM V — will appear around 2010. Evidently, it takes a long time to figure out all the ways America is nuts.

The first official attempt to measure the prevalence of mental illness in the U.S. came in 1840, when the Census included a question on "idiocy/insanity." From that single category flowered many more disorders, but each asylum classified them differently. The DSM was first published in 1952 so that "stress reaction" would mean the same in an Arkansas hospital as it does in a Vermont one.

The DSM works like this: imagine you are Tony Soprano in the first season of The Sopranos. You have, in DSM-ese, "recurrent, unexpected panic attacks." You also have "persistent concern about having additional attacks," and you fear you're "losing control, having a heart attack, 'going crazy.'" You aren't on drugs (other than all those bottles of Vesuvio's wine), so — presto — Dr. Melfi gives you a diagnosis of panic disorder, DSM No. 300.01. By the way, if you truly think you are Tony Soprano, see No. 295, schizophrenia.

Of course, in the real world, psychiatric diagnosis doesn't — or at least shouldn't — work like a checklist at a sushi counter. Many of the items that appear as diagnostic criteria in the DSM are sometimes symptoms of a disorder and sometimes signs of perfectly normal behavior. An adolescent who "often argues with adults" may have an unusual condition called "oppositional defiant disorder" or a more common condition called "being 14 years old." The DSM includes a cautionary statement saying it takes clinical training to tell the difference. But many nonspecialists use the book too: insurers open the DSM when disputes arise over the proper course of treatment for particular conditions. (If your treatment doesn't jibe with the DSM, you may not get reimbursed.) DSM diagnoses can be used by courts to lock you in a mental hospital or by schools to place your child in special-education classes. A DSM label can become a stigma.

All of which raises a pressing question: What actually goes into defining a disorder? A.P.A. officials take this question seriously, and they understand the high stakes of a DSM diagnosis. That's one reason they so often revise the book to keep it current with the latest research. (Three editions have been published since 1986.) According to Dr. Darrel Regier, chief of A.P.A. research, roughly 1,000 mental-health professionals will help produce DSM V. The A.P.A. will host at least a dozen conferences, review unending piles of literature and conduct new studies to see whether proposed changes would work in clinical settings.

But like the conditions it helps diagnose, the DSM is more than the sum of its symptoms. As the American storehouse of insanity — the dictionary of everything we consider mentally unbalanced — it's a window into the national psyche. And so it bears close reading, and close questioning, by those outside the psychiatric establishment. Why is caffeine intoxication included as a disorder when sex addiction isn't? Why is pathological gambling apparently crazy when compulsive shopping isn't?

More important, can even a thousand Ph.D.s gathered at a dozen conferences ever really know the significance of such vague symptoms as "fatigue," "low self-esteem" and "feelings of hopelessness"? (You need only two of those, along with a couple of friends telling the doctor you seem depressed, to be a good candidate for something called dysthymic disorder.) Though it's fashionable these days to think of psychiatry as just another arm of medicine, there is no biological test for any of these disorders. While imaging techniques have shown abnormalities in the brain of some people with schizophrenia, no scan can diagnose even that severe condition, let alone something opaque like "histrionic personality disorder." (For which the DSM lists the following as a sign: "consistently uses physical appearance to draw attention to self." So I'm sick if I exchange my Aunt Thelma's drab sweaters for flashier ones every Christmas?)

If the DSM is all we've got, why is it inherently flawed? Because many forces besides science shape it, including politics, fashion and tradition. The A.P.A. actually once held a vote among its members to see whether an alleged disorder — homosexuality — existed. (In 1974, being gay was deemed sane by a vote of 5,854 to 3,810.) Women's groups helped excise "self-defeating personality disorder" from the book. The revised third edition, in 1987, said the typical sufferer "chooses people and situations that lead to disappointment, failure, or mistreatment even when better options are clearly available." But feminists successfully argued that battered women could unfairly fit this category.

Other questionable diagnoses stay in the book because no one fights hard enough to remove them. Thus heterosexual men can be diagnosed with a supposed disorder called "transvestic fetishism" if they meet only two criteria: they have sexual fantasies about cross-dressing, and those fantasies cause "impairment in social, occupational, or other important areas." In other words, someone is sick not if he has the fantasies but if he gets caught having them — for instance, if his boss reads a kinky e-mail he sent at work, which then leads to a pink slip ("occupational impairment").

"For some of these, there is an issue of grandfathering," admits Dr. Michael First, editor of DSM IV. "The onus is on the person who wants to change it to prove that we should do so." First also acknowledges that the A.P.A. does not subject every criterion to rigorous scientific testing, "for practical reasons of continuity." Which may be another way of saying some old-timers still bill sessions for "transvestic fetishists," and they don't want to lose the DSM stamp of approval needed for insurance reimbursement.

To be sure, a few disorders are dropped from each edition. First notes that a supposed childhood condition called identity disorder was excluded from DSM IV even though many child psychologists wanted to keep it. Kids could qualify for that disorder if they were "uncertain" about long-term goals, career choice and friendship patterns. "We said, 'Wait a minute. This looks like normal adolescence,'" says First, "and so we eliminated it."

The DSM's critics say this sit-around-the-table-and-jawbone method isn't really science. Jerome Wakefield, a Rutgers professor of social work, says that while the DSM's authors do try to eliminate errors so that normal emotional reactions aren't diagnosed as disorders, "there's no systematic process here. Changes are made on a very ad hoc basis, where people say, 'Oh, my god, we forgot X.'" Others have even harsher criticism. Dr. Paul McHugh, who chairs the department of psychiatry and behavioral sciences at Johns Hopkins University School of Medicine, says the DSM has lost its usefulness partly because it has "permitted groups of 'experts' with a bias to propose the existence of conditions without anything more than a definition and a checklist of symptoms. This is just how witches used to be identified." He cites multiple-personality disorder as an example of an "imagined diagnosis"; while much of the evidence supporting its existence has been debunked, multiple-personality disorder is still listed in the DSM, though today it's called "dissociative identity disorder."

New controversies have already erupted over what to put in DSM V. For instance, the A.P.A. is considering adding "relational disorders"--severe problems between spouses or siblings — to the fifth edition. Relational-disorder sufferers are completely sane except when they are around, say, their spouse. Skeptics contend that marital spats shouldn't be considered mental illnesses. A group of Stanford researchers wants to put "compulsive shopping disorder" into DSM V, but First doesn't seem to like the idea. While a number of studies have shown that pathological gambling exists and can be measured, he says, compulsive shopping "has received virtually no research attention to date." (The same goes for sex addiction, according to other psychiatrists: it's just Clinton-era pop psychology thus far, not a documentable illness.)

How could the DSM be improved? Critics say the A.P.A. should start by holding every diagnosis to tough scientific standards. Antiquated notions about deviant sexuality should be brought up to date or scrapped altogether. McHugh of Johns Hopkins suggests that the DSM become more than a laundry list of symptoms — some of which are always going to be ambiguous — by organizing psychiatric conditions around what he calls their "fundamental natures." Accordingly, he would use four categories of disorders: those arising from brain disease, those arising from problems controlling one's drive, those arising from problematic personal dispositions and those arising from life circumstances. While such groupings are imperfect — is alcoholism caused by a brain disease or a problem in controlling one's drive, or a little of both?--they at least get clinicians focused not only on the symptoms of an illness but on its possible causes as well.

In the end, though, the DSM can't achieve certainty because psychiatry can't. Unless brain researchers discover exactly how neurological mechanisms become abnormal, the DSM will always include more hypotheses than answers. Which means all those guys fantasizing about tennis outfits are probably just weird, not certifiable.