Head Games

What are the rules for defining mental illness?

Jacob Sullum from the January 2003 issue

(Page 2 of 3)

Horwitz emphasizes that psychiatry originally focused on the extremes of mood and behavior that today are called psychoses. He cites substantial evidence from twin and adoption studies that a genetic predisposition plays a role in both schizophrenia and bipolar disorder (a.k.a. manic depression). At the same time, most people who receive these diagnoses do not have a family history of the condition. Furthermore, the identical twins of schizophrenics and manic-depressives do not necessarily turn out the same way. The concordance between identical twins is about 50 percent for schizophrenia and 60 percent for bipolar disorder. Clearly, genes are not the whole story.

Horwitz also notes that specific drugs -- phenothiazines and clozapine for schizophrenia, lithium for bipolar disorder -- can help control the symptoms of these particular conditions but are not effective for people with different diagnoses. He concludes that schizophrenia and bipolar disorder seem to represent discrete entities involving characteristic symptoms that can be alleviated with condition-specific treatments. But he does not draw any firm conclusions about what causes these conditions or whether they qualify as brain diseases.

Szasz cautions (and Horwitz acknowledges) that response to a drug is not a sound basis for diagnosis. The fact that you feel better after snorting cocaine does not mean your original state of mind was an illness. Szasz insists that psychoses should not be considered genuine diseases until there is a valid, reliable diagnostic test that indicates a biological defect. He argues that if a schizophrenic's brain could be distinguished from a healthy brain in the same way that the brain of a person with, say, Alzheimer's disease can be, schizophrenia would be treated by neurologists rather than psychiatrists.

"Physical reductionists [predict] that advances in molecular biology will show that many behavioral abnormalities -- now categorized as mental illnesses -- [are] bona fide diseases displaying characteristic lesions on a subcellular level," he writes. "I doubt it. But, supposing that that were to happen, the phenomena so identified would cease to be mental diseases and become instead infectious or neurological diseases -- much as paresis [syphilitic dementia] and epilepsy ceased to be mental diseases once their pathoanatomical and pathophysiological nature became established."

For Szasz, the fact that psychiatrists, politicians, bureaucrats, and activists are constantly insisting that mental illnesses are real diseases, just like cancer or diabetes, indicates that in fact they are fundamentally different. "There are no illnesses outside of the realm of the mental health field whose disease status requires defense by the White House," he slyly notes. "In the end, we come down to the meaning of the term 'mental illness': If we use it to mean brain disease, then psychiatry would be absorbed into neurology and disappear....However, pyromania is plainly not like multiple sclerosis, and treating a patient with schizophrenia without his consent is plainly not like treating a patient with anemia with his consent."

While Horwitz is more inclined than Szasz to believe that schizophrenia and manic depression have a biological basis, he shows that the case for viewing other mental disorders as inherited brain diseases is weak. At most, he suggests, people may be born with a general predisposition to psychological problems that can be manifested in a wide variety of ways, depending upon psychiatric fashion and the individual's sex, culture, and circumstances.

Using detailed examples, he argues that the malleability of mental disorders explains why symptoms of the same condition (depression, say) vary across cultures; why some disorders occur disproportionately among men (substance abuse) or women (bulimia), among people of a certain social class (neurasthenia), or in certain countries (attention deficit disorder); and why diagnoses and the behavior on which they are based suddenly proliferate (as hysteria, anorexia, and multiple personality disorder did) and later become rare or nonexistent.

From Horwitz's perspective, then, the 400 or so disorders listed in the latest edition of the DSM do not represent distinct underlying conditions. Accordingly, it's not surprising that so-called antidepressants such as Prozac and Paxil (known as selective serotonin reuptake inhibitors, or SSRIs) "work equally well for a broad range of disorders including panic, obsessive, and phobic conditions, as well as depressive and anxious states. They are also widely used for substance abuse and eating disorders and for general distress among both adults and children. Moreover, these medications are promoted as ways to enhance the personalities of normal people by improving self-esteem, self-confidence, interpersonal relationships, and achievement." This versatility makes even more sense in light of recent studies indicating that SSRIs may be little more than placebos.

As the range of applications for SSRIs suggests, there is not much (aside from actual diseases) that psychiatrists are not prepared to treat. Horwitz traces this broad purview to Sigmund Freud and his followers. "Neurotic behaviors were conceived as continuous with normal behaviors, the two blurring indistinctly into each other," he writes. "The joining of neurotic with normal behavior served at the same time to make pathological, nonpsychotic behavior ordinary and to pathologize everyday behavior." Hence "dynamic psychiatry transformed the jurisdiction of the mental health professions from people with serious mental illnesses to those with problems in their everyday lives....The potential domain of psychiatric classification was now all of human behavior."

By the 1970s, however, dynamic psychiatry was perceived as unscientific, with vague diagnoses, unfalsifiable etiological theories, and weak standards of effectiveness. Seeking to maintain their status as real physicians, and therefore their advantage over competing mental health professionals and their eligibility for coverage by third-party payers, psychiatrists adopted the precise-seeming approach reflected in the third and subsequent editions of the DSM. To be accepted by psychiatrists, the new DSM had to cover all the sorts of people they were already treating. In addition to distressed individuals seeking guidance and emotional support, these included people coerced into treatment because of their deviant behavior, such as alcoholics, illegal drug users, and juvenile delinquents. "Contrary to the common view that the DSM-III expanded the range of pathology the mental health professions treat," Horwitz writes, "in fact it simply recategorized as discrete diagnostic entities the wide range of problems that dynamic psychiatry had already pathologized."

The result was a bewildering array of diagnoses -- including such handy ones as "antisocial personality disorder," "generalized anxiety disorder," "social phobia," "hypoactive sexual desire disorder," "substance abuse," and "pathological gambling disorder" -- that could be applied to just about anyone a psychiatrist might encounter. "Defining symptom-based entities made these entities seem as if they were real," writes Horwitz, echoing a point Szasz emphasizes. "The focus on reliability," Horwitz continues, "provided the justification for psychiatry to claim it was scientific without having to demonstrate why any of the classified entities ought to be considered instances of mental disorder." Although the DSM is vague regarding the etiology of the conditions it lists, Horwitz notes, "a biomedical model presumes that psychiatric disorders are brain diseases similar to diseases of other bodily organs."

The DSM (now in its revised fourth edition, known as DSM-IV-TR) disparages its own definition of mental disorder, calling it an "anachronism of mind/body dualism," used only because "we have not found an adequate substitute." The implication seems to be that there is no real distinction between physical and mental illnesses because the mind is governed by the brain, which is part of the body. But as Szasz notes, if "everything that happens to or is done by human beings is biological, then saying so is a meaningless truism. Attributing mental illnesses, such as addiction and panic disorder, to biological alterations occurring at a 'subcellular level' is a parody of the denial of free will, choice, and responsibility."

The continued use of the term mental disorder suggests that psychiatrists do not have the courage of their reductionist convictions. Szasz observes that the term "is useful because it enables psychiatrists to waffle: 'It' may be a disease, or may not be; may render the patient incompetent, or may not; may annul intentionality, but not necessarily; may cause dangerousness to self and others, but not always; and be an excuse for crime, and may not be. It all depends on the psychiatrists' interpretation of their so-called clinical observations."

For Horwitz, by contrast, the main problem with the DSM's definition of mental disorder is that psychiatrists do not apply it fully and consistently. Borrowing from the sociologist Jerome Wakefield's condensation of the DSM standard, Horwitz defines mental disorder as "a harmful internal dysfunction" with "socially inappropriate" consequences. Too often, he says, psychiatrists ignore these criteria, improperly treating deviant behavior (such as illegal drug use) and normal, expected responses to stress (such as depression after divorce) as mental disorders.

The problem is even worse when the DSM's symptom lists are applied not to people seeking help from mental health professionals but to samples of the general population diagnosed by survey. Such studies, Horwitz argues, lead to absurdly exaggerated estimates of how many people suffer from particular disorders, or from mental illness in general, because they strip symptoms of their context, which is crucial in determining whether they arise from an "internal dysfunction."