When I started in psychiatry, diagnoses were defined by the opinions of expert clinicians. The good part was that professors insisted on case presentations that included every symptom and historical detail, even heart-wrenching grocery store visits. The embarrassing part was their endless disagreements. They disagreed not only about the diagnosis for specific patients but even about how to define diagnoses. At a staff meeting to discuss a newly admitted hospital patient, one senior psychiatrist proclaimed that the diagnosis was recurrent endogenous depression, another that it was anxiety neurosis, and a third that the problem was obviously pathological guilt following the death of an ambivalently loved father. Brilliant professors wielded extraordinary clinical and rhetorical skills in defense of diagnoses that were little more than opinions.
Such diagnostic inconsistency was an embarrassment for the field. A 1971 study asked psychiatrists in the United States and United Kingdom to watch the same videos of diagnostic interviews.2 For one of the cases, schizophrenia was diagnosed by 69 percent of American psychiatrists but only 2 percent of British psychiatrists. Such wild unreliability made research hopeless. The problem hit the fan in 1973, with publication of an article by the Stanford University psychologist David Rosenhan in the prestigious journal Science. He sent twelve mentally normal “pseudopatients” to emergency rooms, where they reported hearing hallucinated voices saying “empty,” “hollow,” and “thud.” All were admitted to mental wards. They acted normal after admission, but all nonetheless received a diagnosis of schizophrenia.3 Although fake patients could also have duped neurologists or cardiologists, the article made psychiatry a laughingstock. The last straw was in 1974, when the controversial status of homosexuality as a mental disorder was settled—by a vote of APA members. Psychiatry awoke from its long dream to find itself drifting on a psychoanalyst’s couch half afloat in the backwaters of medicine.
A New Diagnostic Manual to the Rescue
Desperate to join the medical mainstream, psychiatry recognized that its diagnostic system was grossly inadequate. For instance, in the 1968 second edition of Diagnostic and Statistical Manual of Mental Disorders (DSM-II), depressive neurosis was defined as “An excessive reaction of depression due to an internal conflict or to an identifiable event such as the loss of a love object or cherished possession.”4 Is moderate depression a week after the loss of a favorite cat “excessive”? One diagnostician would say, “No, not at all, people love their cats”; another, “After a week, it is obviously excessive!” Such disagreements made psychiatry’s scientific aspirations laughable.
The solution was a radical revision, DSM-III, published in 1980.5 Written by a task force of the American Psychiatric Association under the leadership of psychiatry researcher Robert Spitzer, it purged psychoanalytic theory from DSM-II and replaced its 134 pages of clinical impressions describing 182 disorders with 494 pages of symptom checklists that defined 265 disorders. “Depressive neurosis” was eliminated. The definition of a new diagnosis, “major depressive disorder,” said nothing about internal conflict; it only required the presence of at least five of nine possible symptoms for at least two weeks. Every diagnosis was now defined by a checklist of necessary and sufficient symptoms.
DSM-III transformed psychiatry.6 It made possible standardized interviews that epidemiologists could use to measure the prevalence of specific disorders.7 Neurobiologists could now search for brain abnormalities for specific disorders. Clinical researchers at different sites could compare outcomes of alternative treatments, providing the data needed to create treatment guidelines. Regulatory agencies, insurance companies, and funding agencies soon demanded DSM diagnoses. Psychiatrists could finally diagnose specific disorders, just like other physicians. As a solution to the diagnostic unreliability crisis of the 1970s, DSM-III succeeded beyond all expectations.
Controversies Erupt
Despite providing the objectivity so essential for research and scientific respectability, DSM-III aroused vociferous criticism. Instead of fading with time, dissatisfaction has grown. Clinicians say that DSM categories ignore important aspects of many patients’ problems. Teachers of clinicians report that excess reliance on the criteria leads students to neglect careful observation of their patients’ problems.8 Researchers protest that DSM categories do not map well to their hypotheses.9 Physicians in other areas of medicine wonder why psychiatric diagnosis is such a problem. And people outside of medicine who read about these controversies all too often conclude that the whole field of psychiatry is just a bunch of hooey.
DSM-III increased objectivity dramatically, at the cost of discouraging careful clinical evaluations. If Ms. B has five or more symptoms for two weeks or more, she has major depression, too bad about Jack abandoning her the day they were to run away together. Even leading biological researchers are appalled. Nancy Andreasen, the author of The Broken Brain: The Biological Revolution in Psychiatry and the former editor of a leading psychiatry journal, described the “unintended consequences” of DSM-III. “Since the publication of DSM-III in 1980, there has been a steady decline in the teaching of careful clinical evaluation based on a deep general knowledge of psychopathology, and attending to the individual person’s problems and social context. Students are taught to memorize DSM rather than to learn complexities from the great psychopathologists of the past.”10
The problem is not merely theoretical. A psychiatrist in training concluded a grand rounds case presentation by saying, “This patient has sleep problems, low interest, low energy, poor concentration, low appetite, and a seven-pound weight loss, so she qualifies for a diagnosis of major depression. We will begin antidepressant treatment.” When asked, “What set this all off?” the young doctor replied, “Family problems.” “What kind of family problems?” “Her husband left her.” Did she see warning signs about his leaving? Don’t know. Was this her first marriage? Don’t know. Does she have a relationship with another man? Don’t know. Was she abused in childhood? “I didn’t ask about those things because they aren’t relevant. The diagnosis is major depression, and the treatment plan follows established evidence-based guidelines for this brain disorder.” The excessive confidence in and commitment to a narrow ideology were as breathtaking as the willful ignorance about the patient.
DSM-III’s objectivity also exposed other problems. Many patients with one DSM disorder also qualify for several other diagnoses. This problem is so pervasive that the leading psychiatric epidemiologist Ronald Kessler, my former colleague at the University of Michigan, called his biggest project the “National Comorbidity Survey.”11 Not only do many patients meet the criteria for multiple disorders, patients in the same diagnostic category often have very different symptoms. Adding this “heterogeneity” on top of huge comorbidity makes many people wonder if the DSM categories correspond to real natural entities.
The blurry boundaries between different disorders raise still more concerns. For instance, most patients with depression also have anxiety, and vice versa.12,13,14,15 Furthermore, the boundaries that separate disorders from normality are arbitrary. No laboratory tests like those that diagnose cancer or diabetes are available. The authors of DSM-III assumed, in 1980, that better categories would soon be based on new discoveries of brain abnormalities. Almost four decades of intensive research later, no laboratory test is yet available to diagnose any of the major psychiatric disorders.
To their great credit, the leaders of American psychiatry forthrightly acknowledge the problem. Allen Frances, the chair of the task force that wrote DSM-IV, said, “We are at the epicycle stage of psychiatry where astronomy was before Copernicus and biology before Darwin. Our inelegant and complex current descriptive system will undoubtedly be replaced by explanatory knowledge that ties together the loose ends. Disparate observations will crystallise into simpler, more elegant models that will enable us not only to understand psychiatric illness more fully but also to alleviate the suffering of our patients more effectively.”16
> Thomas Insel, a recent director of the NIMH, says, “It is time to rethink mental disorders, recognizing that these are disorders of brain circuits”17 and “Our resources are more likely to be invested in a program to transform diagnosis by 2020, rather than modifying the current paradigm.”18
Frances is less hopeful, saying, “The DSM-V goal to effect a ‘paradigm shift’ in psychiatric diagnosis is absurdly premature. . . . There can be no dramatic improvements in psychiatric diagnosis until we make a fundamental leap in our understanding of what causes mental disorders. The incredible recent advances in neuroscience, molecular biology, and brain imaging that have taught us so much about normal brain functioning are still not relevant to the clinical practicalities of everyday psychiatric diagnosis. The clearest evidence supporting this disappointing fact is that not even 1 biological test is ready for inclusion in the criteria sets for DSM-V.”19
The courage and integrity of these scientists are as remarkable as their vision. All agree that new approaches are essential. The major proposals so far, however, have been to further revise the diagnostic categories and to look yet harder for biomarkers to validate them.
DSM-III was the focus of most complaints, so it was revised to DSM-III-R in 1987, DSM-IV in 1994, and DSM IV-TR in 2000. The task of writing the big revision, DSM-5, was pursued over the course of a decade by a twenty-nine-member APA task force that coordinated the work of six study groups and thirteen work groups.20,21,22,23 After years of sometimes bitter controversies, DSM-5 was finally published in 2013.24 It made a few small structural changes, such as moving personality disorders into the same category as other disorders. It also combined some categories (for example, substance dependence and substance abuse were merged into substance use disorder) and split others (for example, agoraphobia is now separate from panic disorder). These and others sensible changes make DSM-5 more coherent and useful.
Calls for sweeping changes were rejected, however. A proposal to replace categories with scales ranging from mild to severe was rejected as impractical. “The diagnosis is major depression” is simple and definitive compared to “The score on the depression scale is 15.” Categories make for efficient communication and statistical recordkeeping. They also satisfy the human lust for making things seem simpler than they are. We have been trying to map the landscape of mental disorders by drawing lines around clusters of symptoms as if they were islands, but mental disorders are more like ecosystems: areas of arctic tundra, boreal forest, and swamp blend into one another, defying crisp boundaries.
The second strategy has been to push yet harder to find genes, blood tests, or scans that can define diagnoses. No one imagined that we would still have no tests for schizophrenia, autism, and bipolar disorder thirty-seven years after the publication of DSM-III. Continuing the search is essential; it offers our best hope for finding cures. However, after decades of consistently negative results, it is time to step back to ask why specific physical causes for mental disorders are so elusive compared to those for other medical disorders.
The consensus answer is that we have not looked hard enough in the right places. Many neuroscientists suggest that we should shift the focus from molecules and brain locations to “brain circuits.”25 This reflects the growing recognition that diverse brain areas and neurotransmitters are involved even in functions as specific as recognizing faces. Considering circuits highlights adaptive functions, but it perpetuates the misleading analogy of evolved brain systems with human-designed electronic circuits. Circuits designed by engineers have discrete modules with specific functions and defined connections that are all necessary for normal operation. Evolved information-processing systems have components with indistinct boundaries, distributed overlapping functions, intrinsic robustness, and innumerable connections that make them different from anything that an engineer could even imagine. Shifting the focus from molecules and neurons to circuits is a good idea, but neuroscience will succeed faster when it acknowledges that those circuits are organically complex in ways that make them very different from anything an engineer could design.
Revising the diagnostic criteria will not solve the problem. Looking harder for biomarkers will eventually provide a definitive diagnosis for only some disorders. The dilemma has provoked deep thinking about what mental disorders are.
Accepting the Reality of Organic Complexity
The question “What is a mental disorder?” has been addressed by Jerome Wakefield, a social worker, clinician, researcher, and philosopher at New York University.26,27,28 His pithy conclusion is that mental disorders are characterized by “harmful dysfunction.” “Dysfunction” means a malfunction in a useful system shaped by natural selection. “Harmful” means that the dysfunction causes suffering or other harm to the individual. Wakefield’s analysis grounds psychiatric diagnosis in an evolutionary understanding of the normal functions of brain/mind, the same way the rest of medicine understands pathology in the context of normal physiology.29 His cogent analysis has, however, had little influence on how psychiatrists make diagnoses.
The South African psychiatric researcher Dan Stein and I decided to see if a systematic evolutionary analysis could suggest ways to improve the DSM. After wrestling with the problem for several months, we came to a conclusion that surprised us: DSM describes most mental disorders pretty well. We identified some big problems, especially the failure to distinguish symptoms from diseases. But most of the dissatisfaction with DSM diagnoses arises not because they fail to describe clinical realities but because they describe the messy reality of mental disorders all too well. Problems overlap. One disorder can have many causes. One cause can result in many different symptoms. No specific gene or brain abnormality has yet been found to define a mental disorder. Now what?
Toward a Genuinely Medical Model
The so-called medical model in psychiatry usually refers to the view that specific disorders are caused by specific brain abnormalities that are best treated with drugs and other physical therapies. The actual model of disease used in the rest of medicine is subtler. It does not just dive in looking for specific causes for presumably specific diseases. Instead, it tries to understand pathology in the context of normal functioning. Three examples illustrate how a more genuinely medical model could advance psychiatric diagnosis.
First, the rest of medicine recognizes symptoms, such as pain and cough, as protective defenses and carefully distinguishes them from the disorders that arouse them. In psychiatry, by contrast, extremes of emotions, such as anxiety and low mood, are categorized as disorders, irrespective of any situation that might be arousing them. This error is so basic and pervasive that it deserves a name: Viewing Symptoms As Diseases (VSAD). Reforming psychiatric diagnosis will require recognizing negative emotions as responses that can be useful in certain situations—at least for our genes.
Second, the rest of medicine recognizes many syndromes, such as congestive heart failure, which are defined not by specific causes but by failures of functional systems. Physicians know that heart failure can have a dozen different causes. If schizophrenia and autism result from similar system failures, searching for the specific cause is senseless.
Finally, the rest of medicine does not hesitate to diagnose some conditions, such as tinnitus and essential tremor, which have no identifiable specific cause or tissue pathology. Most result from dysregulated control systems. The same may be true for eating disorders and mood disorders.
The core problem for psychiatric diagnosis is the lack of a perspective on normal useful functions that physiology provides for the rest of medicine. Internal medicine doctors know the functions of the kidneys. They don’t confuse protective defenses such as cough and pain with diseases such as pneumonia and cancer. Psychiatrists lack a similar framework for the utility of stress, sleep, anxiety, and mood, so psychiatric diagnostic categories remain confusing and crude.
Carefully distinguishing symptoms from both syndromes and diseases is
crucial for making psychiatric diagnosis like diagnosis in the rest of medicine. Like fever and pain, anxiety and low mood are useful normal responses to some situations. It is time to give up the fantasy that each mental disorder has a specific cause. Instead, many mental disorders are, as in the rest of medicine, extremes of symptoms. Others are system failures that can have many different causes. This does not mean we should give up looking for specific brain abnormalities; they will be found, eventually, for some disorders, and the sooner the better. But the search will be sped by adopting a genuinely medical model.
CHAPTER 3
WHY ARE MINDS SO VULNERABLE?
If the immediate and direct purpose of our life is not suffering, then our existence is the most ill-adapted to its purpose in the world.
—Arthur Schopenhauer, 18511
If the mind were a machine, we would praise its designer to the heavens for creating the most extraordinary device in the universe. It can recognize a thousand faces and instantly come up with names—except for that of the client you want to introduce to your boss at a party. It can learn Chinese, Finnish, or English by age three with no special effort, even tenses, genders, and verb conjugations. The cello virtuoso Yo-Yo Ma plays the thousands of notes in Edward Elgar’s Cello Concerto in E minor in order, fast, from memory. The science rap artist Baba Brinkman makes up hilarious songs on the fly on any topic. A high school student learns calculus. An elderly man recalls the exact rusty pail he and his mother used one sunny morning seventy years ago when they picked blueberries on a sandy hill. A young man rehearses a dozen strategies for getting the beautiful one to go with him to the prom. A young woman, anticipating his invitation and hoping for a better offer, tries to figure out how to postpone giving an answer. What incredible information processing!
Good Reasons for Bad Feelings Page 4