By Eric Vandenbroeck and co-workers

The secrets behind the making of the Diagnostic and Statistical Manual of Mental Disorders(DSM)Part Three

Manufacturing Depression: The Secret History of a Modern Disease and contributor to The New Yorker, Mother Jones, The New York Times, and other publications, Greenberg is a practicing psychotherapist. The Book of Woe: The Making of the DSM-5 and the Unmaking of Psychiatry is his exposé of the business behind creating the new manual.


But there is more than that:


This while in Part One of our overview we focused among others on its protagonist  Robert L. Spitzer and the construction of DSM-III. In Part Two we focused on among others bot the intra-professional external forces involved with DSM-IV and  DSM-5 revision attempts. Yet observable symptoms persist in defining the DSM diagnoses, a situation that general medicine surmounted more than a century ago. The chemical and physical operations of the brain have yet to provide clues that may unravel the mysteries of human consciousness and its distortions. Mental disorders could require understandings that cannot be completely removed from personal life experiences. The future of the DSM is clearly at a crossroads, but the path it should take has no roadmap.


Expanding Pathology?

Each DSM has grappled with how broad a range of pathology the manual should encompass. The DSM’s history is marked by its increasing medicalization of emotions and behaviors in one narrative. This accounts for the growth of DSM diagnoses from 106 entities in the DSM-I to 182 in the DSM-II, 265 in the DSM-III, and 292 in the DSM-IV 5. Much of this expansion involves medicalizing phenomena that had been seen as normal or immoral. 

37Much of this expansion involves medicalizing phenomena that had been seen as normal or immoral.

For some, “medicalization” is a descriptive, not evaluative, term that does not judge whether the trend is desirable or undesirable. This group notes how diagnoses of mental disorders have expanded horizontally to encompass new forms of pathology and vertically to capture milder forms of conditions that previously required greater severity.28 Many critics, however, see growing medicalization as pernicious and strive to “protect normality from medicalization and psychiatry from overexpansion.”29 A third view, illustrated by anthropologist Roy Grinker’s assessment of the growth of autism diagnoses, celebrates the benefits such as better access to treatment and services, lower levels of stigma, and improved research that this trend has brought about: “The prevalence of autism today is a virtue, maybe even a prize.”30

The range of behavior that falls under the legitimate authority of psychiatry has increased, as Spitzer observed. The larger diagnostic categories in the DSM-III than DSM-II reflect the clinical and research need for greater specificity in describing behavioral syndromes.31 It is less clear, however, that the mental disorder itself has grown over the past 20 years.32 The DSM-I and II already encompassed much of the more detailed criteria sets that arose in 1980.33 There is marked variation in the degree of newly medicalized conditions across the major classes of mental disorders. 

Disorders among children and adolescents provide the clearest examples of how the DSM has increasingly medicalized previously nonmedical conditions. The initial DSM paid almost no attention to mental disorders among young people. It mentioned autism as a symptom of childhood schizophrenia and briefly described three categories of adjustment reactions in infancy, childhood, and adolescence. The DSM-II expanded this class, adding seven diagnoses to the original DSM-I conditions. Its category of behavior disorders of childhood and adolescence was 1 of just 11 major classes.34 

The DSM-III greatly accelerated the enlargement of diagnoses among youth with a class renamed “disorders usually first evident in infancy, childhood, or adolescence.” The manual placed this category first among its 15 general diagnostic classes. In contrast to the 3 pages the DSM-II devoted to this group, the DSM-III took 65 pages to describe 46 different criteria sets.35 The following two revisions continued to broaden this class, especially for attention-deficit / hyperactivity disorder, autism, and Asperger's, a condition that first appeared in the DSM-IV. One result was that diagnoses of ADHD surged from less than 1 percent in 1980 to over 10 percent of children in 2011.35 Another was that rates of all forms of autism expanded from just 1 in 2,000 children in the 1970s and 1980s to 1 in 68 when the DSM-5 was published in 2013.36 The DSM-5 did not continue the growing medicalization of mental disorders among youth but moved many of the former age-defined diagnoses to their substantive homes (e.g., separation anxiety went to the anxiety disorders section; conduct disorder to disruptive, impulse, and conduct disorders). Its new class of neurodevelopmental disorders had significantly fewer categories than the child and adolescent class it replaced. 

The bipolar II diagnosis that entered the DSM in 1994 provides another example of expanding pathology. This condition combined the easy-to-meet MDD criteria with a single hypomanic episode of at least four days’ duration. A disorder that historically affected about 1 percent of the population came to encompass about five times that number.37 It became the first DSM diagnosis to create a mass market for antipsychotic drugs. 

The DSM-5’s expansion of the class of substance use disorder to encompass substance-related and addictive disorders provides a third instance of the extension of mental disorder into a new realm. The addition of an addictive category that is not substance-related has huge potential consequences. While the DSM-5 text incorporates only gambling disorders in the new category, the manual includes internet gaming disorder among its conditions for further study. Virtually all (90 percent) Americans are now online, spending an average of between 18 and 24 hours a week on the internet.38

This figure increases to about nine hours a day among teenagers.39 This change potentially extends the notion of “addiction” from its traditional application to substances that induce compulsive consumption to virtually any activity that people frequently engage in, such as sex, shopping, eating, and exercising.40 The DSM’s possible colonization of this huge new realm might realize Karl Menninger’s belief that “most people have some degree of mental illness at some time, and many of them have a degree of mental illness most of the time.”41 

Medicalization displays a less straightforward trajectory for other major DSM classes. The DSM-I is called psychoneurotic disorders, and the DSM-II neuroses had few specific categories but extensive reach. Outpatient psychiatrists and physicians could apply its vague formulations to the psychic consequences of poor marriages, economic worries, failed ambitions, and general nervousness. While the DSM-III and the manuals that followed it contained a far greater number of specific diagnoses, they did not expand the amount of pathology that the broad neurotic categories of the initial DSMs already captured.

The class of anxiety disorders also shows how more diagnoses need not indicate a growing medicalization of previously normal emotions. Although the DSM-I and II had fewer specific anxiety diagnoses than the manuals that followed, they provided a more expansive conception of this condition. The DSM-II conception was exceedingly broad: “Anxiety is the chief characteristic of the neuroses. It may be felt and expressed directly or controlled unconsciously and automatically by conversion, displacement, and various other psychological mechanisms. Generally, these mechanisms produce symptoms experienced as subjective distress from which the patient desires relief.”42 The greater number of anxiety diagnoses in subsequent manuals did not expand pathology but more precisely defined conditions that earlier manuals had incorporated into their broad categories.43

The personality disorders are another case where recent DSMs are not more medicalized than were their predecessors. Although this class's particular types of conditions changed considerably over the DSM era, neither their number nor their range increased. The first two DSMs presented capacious portrayals of 12 personality disorders that spanned from the most introverted to the most antagonistic and many character types in between.44 

Subsequent manuals added some conditions, abolished others, and changed the criteria for still others, but the DSM-5 maintains diagnoses for 12 personality disorders.45 

PTSD illustrates a third type of diagnostic trajectory that fluctuates in scope over the DSM period. Although the cultural presence of trauma-related diagnoses has expanded tremendously since 1980, gross stress reaction, growing out of the professional experiences of military psychiatrists in World War II, was a central diagnosis in the first DSM. Trauma-related diagnoses contracted in the DSM-II only to be resurrected in the DSM-III’s PTSD diagnosis. The broadening definitions of “trauma” in the PTSD criteria of the DSM-III-R and DSM-IV helped bring about an explosion of traumatic diagnoses.46 The DSM-5 changes to the PTSD diagnosis, however, constituted a rare example of an attempt to reduce the number of diagnoses through narrowing the scope of relevant traumas and limiting traumatic exposure to actual events.47 The revised criteria set reversed the consistently growing expansion of PTSD from the DSM-III to the DSM-III-R and the DSM-IV. 

The development of the disruptive mood dysregulation disorder (DMDD) in the DSM-5 is another unusual case of an attempt to reign in diagnostic expansion. The spectacular growth of pediatric bipolar disorders and resulting prescriptions for antipsychotic drugs was a huge embarrassment to the psychiatric profession in the early 2000s. The major proponents of PBD, Harvard psychiatrist Joseph Biederman and his associates, had received more than $4 million from Johnson & Johnson, the maker of the antipsychotic drug Risperdal. The resulting scandal generated widespread negative publicity, including front-page stories in the New York Times and a widely viewed segment on 60 Minutes about a four-year-old who died from an overdose of drugs prescribed for her “bipolar disorder.”48 At the same time, the disruptive, oppositional, and irritable children diagnosed with PBD posed major behavior problems for their parents and others. The DSM-5 workgroups on childhood and adolescent disorders developed an ingenious solution to the combination of the public repudiation of PBD and parental demands for some treatment. After some missteps, they created a DMDD diagnosis that combined symptoms of irritable mood and aggressive behavior.49 They placed it within the depressive disorders, not the bipolar and related disorders class, which severed its connection with the cycling of bipolar conditions that required powerful drugs. This placement signaled a departure from PBD and its attendant need for antipsychotic medication, yet preserved a diagnosis that would help parents deal with their troublesome children.50

A final category encompasses unsuccessful attempts to medicalize behaviors. The failed proposal for a psychosis risk syndrome in the DSM-5 provides perhaps the most prominent example. As the previous chapter discussed, the PRS diagnosis strove to identify persons who did not meet the criteria for psychosis but were thought to be at risk of developing one in the future. This attempt to greatly expand the realm of pathology met intense opposition, and the DSM-5, like its predecessors, lacks any diagnosis related to the risk of becoming mentally ill. 

The depiction of an ever-growing realm of mental disorder that encompasses new forms of pathology and milder forms of previously recognized illnesses is, therefore, an overgeneralization. Childhood and adolescent disorders, the new class of addictive behaviors, and bipolar II illustrate a growing array of pathology to incorporate phenomena previously viewed as mental disorders. Yet, despite alterations in many of their particular criteria sets, the range of other diagnoses, such as anxiety and personality disorders, has not incorporated non-pathological conditions previously. Some conditions, like PTSD, have both expanded and contracted over time. Finally, as PRS shows, some attempts at medicalization have failed. Almost no categories, however, have completely disappeared from the DSM, homosexuality being the rare exception. 

Critics who object to the growing medicalization of previously nonmedical conditions argue that the expansion of mental disorder pathologizes normal experiences, stigmatizes the recipients of diagnoses, and generates unnecessary and often harmful treatments for people who don’t need them. In addition, they note how treating non-disordered conditions takes resources away from those who genuinely could benefit from therapies.56 These critics are unlikely to succeed. Regardless of their validity, these arguments are likely to hold little weight in the face of the substantial benefits that DSM diagnoses reap for many groups. Interests that support their maintenance or expansion include:

The professional, industry and lay interests in preserving or increasing the number of diagnoses will likely continue to enfeeble objections to growing amounts of pathology. 


On the Threshold?

Gerald Grob’s synopsis captures the history of psychiatry: Every generation since the nineteenth century, the specialty has stood on the threshold of fundamental breakthroughs that would revolutionize how mental disorders were understood and treated. In the mid-twentieth century, psychodynamic and psychoanalytic psychiatry became the vehicle by which the mysteries of normal and abnormal behavior would be revealed. At present, the road to salvation is presumably through biological psychiatry, neuroscience, and genetics.57 

As it enters the third decade of the twenty-first century, is psychiatry closer to revealing the mysteries of mental disorder than it was when the first DSM appeared? In one prominent telling, the DSM’s evolution represents a tale of steady, if uneven, progress. The accumulation of new findings has led the manual to ever-better approximations of the reality of mental disorders. Prominent psychiatrists speak of how “scientific evidence” has replaced the charismatic authority of “great professors” in the evolution of the DSM.58 The APA website lauds the DSM-5 Task Force: “Their dedication and hard work have yielded an authoritative volume that defines and classifies mental disorders to improve diagnoses, treatment, and research.”55 Indeed, some developments, the replacement of analytic assumptions with theory neutrality, the recognition that intense social stressors can produce lasting mental disorders, the removal of homosexuality, the acknowledgment of autistic disorders, seem to improve the manual. Few psychiatrists would prefer the always-perfunctory and sometimes analytically infused definitions that prevailed before the DSM-III to the current DSM-5 diagnoses. 

Others paint a more skeptical picture of the history of the DSM. In their portrayals, the medical-like diagnoses that arose in the DSM-III are artifices that disguise psychiatry’s continuing lack of progress in understanding the causes, prognoses, or optimal treatments for any of its major conditions. For example, sociologist Owen Whooley observes how every edition of the manual continues to camouflage psychiatry’s fundamental ignorance about the basic nature of the mental illness.56 Historian Edward Shorter goes further, contending that the current classification is inferior to that of the original DSMs: “The diagnoses that flourished in the middle third of the twentieth century did a better job of cutting Nature at the joints than many of the diagnoses we have today, which are artifacts born of political compromises and sustained by pharmaceutical promotion rather than scientifically accurate descriptions of what is actually wrong with someone.”59 For such critics, the major issues the DSM has had to confront over the past 70 years are no closer to resolution now than when the manual first arose. 

Despite advances in brain-imaging technologies, psychiatry remains as reliant on observable symptoms as it was in the eighteenth century when Thomas Arnold wrote, “When the science of causes shall be complete, we may then make them the basis of our classification, but till then we ought to content ourselves with an arrangement according to symptoms.”60 Like all psychiatric classifications that preceded it, the DSM system must still use reported symptoms as the raw material for constructing its diagnoses. No attempt to develop etiologically informed diagnoses has yet to succeed.

The reliance on external symptoms especially hampers the construction of adequate definitions of the nature of the mental disorder itself. One central, perhaps the central, issue regarding the DSM is what makes any of its diagnoses mental disorders. The DSM-I and II did not try to answer this question. Psychiatry was so well respected in the postwar period that no one questioned the field’s authority to define its subject matter. This situation radically changed in the late 1960s and early 1970s when anti-psychiatry, gay, and feminist activists challenged the field’s authority over what the term “mental disorder” should encompass. At the same time, strong cultural, political, and economic forces pushed psychiatry to conceive of its subject manner as disease entities comparable to those in the rest of medicine. Beginning with the DSM-III and persisting through the present, diagnoses reflect the idea that mental disorders, no less than physical diseases, have their own reality independent of particular life experiences. However, extracting symptoms of mental disorder from the contexts in which they arise is considerably more complicated than isolating symptoms is in the rest of medicine. 

A major obstacle  Robert Spitzer faced was that a core principle of the DSM-III revolution was to use observable symptoms without regard to their underlying causal mechanisms to define each diagnosis. The turn to classifying mental disorders by their outward appearances hindered the ability to separate mental disorders from symptomatically similar but contextually explicable responses and disliked and devalued but not disordered responses. 

Many psychiatrists did not bother with the absence of a definition of “mental disorder.” Why should psychiatry be different from other branches of medicine? This objection persists. For example, when a person has a myocardial infarction (MI), physicians regard it as an instantiation of cardiac disease, regardless of its context.61 For Pies and others, just as a heart attack is a heart attack, a mental illness is a mental illness. Therefore, like the rest of medicine, context is irrelevant to separating psychiatrically disordered symptoms from situationally apt or culturally appropriate expressions. 

Yet Spitzer realized that defining “disorder” is fundamentally different, and far more challenging, in psychiatry than physical medicine. Separating organic symptoms from the settings was a hallmark of diagnostic progress in other medical specialties.62 In psychiatry, divorcing symptoms from context has the opposite impact of hopelessly blurring situationally appropriate psychological phenomena from mental disorders. This is because all mental functions are susceptible to environmental circumstances. Virtually every symptom of various mental disorders can sometimes be biologically and psychologically suitable adaptations to given contexts, culturally explicable expressions, or both. For example, symptoms resembling depression that arise after the death of a loved one indicate that grief mechanisms are working appropriately, not inappropriately. Likewise, a panic attack is an understandable response when facing an impending fall off a cliff but a sign of disorder in the absence of danger.63 Or hearing voices, which can be a hallmark of schizophrenia, is sometimes explicable in particular cultural and religious settings.64 In contrast, a heart attack always signals a failure of natural functioning regardless of the context or culture in which it emerges. Unlike other medical specialties, context is an intrinsic aspect of deciding what a mental disorder is or is not.65 

Spitzer understood the difficulties of developing a general definition of “mental disorder” entailed but realized such a statement was necessary to establish the field’s credibility and protect it from anti-psychiatry attacks. His original formulation has mostly endured through the present DSM-5 version: 

A mental disorder is a syndrome characterized by clinically significant disturbance in an individual’s cognition, emotion regulation, or behavior that reflects a dysfunction in the psychological, biological, or developmental processes underlying mental functioning. Mental disorders are usually associated with significant distress or disability in social, occupational, or other important activities. An expectable or culturally approved response to a common stressor or loss, such as the death of a loved one, is not a mental disorder. Socially deviant behavior (e.g., political, religious, or sexual) and conflicts that are primarily between the individual and society are not mental disorders unless the deviance or conflict results from a dysfunction in the individual, as described above.66 

This characterization both expresses what mental disorders are, dysfunctions of some mental processes that are not working as they should and that lead to distress or disability, and what they are not, culturally defined deviant behaviors or conflicts between individuals and society. The DSM’s succinct definition identifies both essential aspects of mental disorders and separates them from states that are often mislabeled as psychiatric problems.66 

Spitzer knew the definition couldn’t set an exact boundary between dysfunctions and normal-range behaviors because no sharp division exists in nature. The borders between mental disorders and non-disordered states of distress or social deviance are often fuzzy, vague, and ambiguous. Despite this caveat, the statement serviceably separates distressing states from those of mental disorders. When the DSM diagnosis takes this general definition into account, they should strike the best balance between recognizing true disorder cases and weeding out contextually appropriate or socially devalued behaviors.67

The central problem has been that, in practice, many of the particular DSM criteria sets don’t follow the definition.68 Worse, criteria for what makes some core diagnoses mental disorders have deteriorated, rather than progressed, over time. Depression provides probably the most egregious example; succeeding DSMs have increasingly relied on external symptoms without regard to the context in which they arise. The first DSM directed clinicians to consider “the realistic circumstances of the loss,” The DSM-II treated only “excessive” symptoms as signs of a disorder.69 The three DSMs that followed provided far more symptom-based criteria sets for MDD but excluded uncomplicated symptoms that arose from bereavement from diagnosis. The DSM-5 criteria set eliminated even this exception.70

Because depressive symptoms need to last just two weeks, they can easily be short-lived responses to events rather than true disorders. Far from being “on the threshold” of growing understandings, the context-free DSM-5 MDD diagnosis shows even greater heterogeneity and indistinct boundaries with normal sadness than its predecessors. 

In other cases, DSM criteria do not separate mental disorders from social deviance. The history of substance use disorder (SUD) diagnoses recounted in the previous chapter provides the primary example of this type of growing conflation. The first two DSMs placed most SUD diagnoses among conditions that involved brain damage. The DSM-III greatly broadened this category by adding to addictive behaviors a new abuse diagnosis that referred to the problematic social consequences of substance use. Abuse and addiction remained separate diagnoses in subsequent revisions until the DSM-5 combined them into a single category, in defiance of the admonition in the manual’s definition of mental disorder to avoid mislabeling social deviance as dysfunction. Instead, the new criteria for SUD blur dysfunctions that stem from addictions with rule violations and social impairments. 

In other cases, diagnoses do not distinguish dysfunctions from conflicts between individuals and society. The new diagnosis of hoarding disorder in the DSM-5 illustrates this situation. Its essence is “persistent difficulties discarding or parting with possessions regardless of their actual value.”7051 Hoarders typically accumulate so many items that they or others have difficulty navigating their living spaces. This telegenic situation has become the object of a popular television series, Hoarders. Yet, the new diagnosis potentially involves many false-positive problems: hoarders themselves are typically not troubled by their condition unless someone tries to stop their stockpiling. Interventions typically arise after family members, neighbors, or public health departments complain. The justification for calling hoarding behaviors “dysfunctions” or “mental disorders” as opposed to “conflicts between individuals and society” is unclear.71

Although the overall narrative of growing diagnostic progress over the DSM era is, at best, questionable, the DSM-5 did take some steps to distinguish mental disorders from other conditions better. One was replacing the former little-used Axis IV, psychosocial and environmental problems, with a new chapter on other conditions that may focus on clinical attention. The DSM-5 forthrightly states, “The conditions and problems listed in this chapter are not mental disorders.”72 They encompass relational difficulties; abuse and neglect; and educational, occupational, housing, legal, and economic problems. Including a wide range of psychosocial, personal, and environmental difficulties that are explicitly not mental disorders could help clarify the hazy boundaries between mental disorders and contextually appropriate distress that characterized previous manuals. 

The DSM-5 also took steps to incorporate cultural differences in symptomatic expressions into account. It includes a whole chapter on cultural formulation based on the idea that “understanding the cultural context of the illness experience is essential for effective diagnostic assessment and clinical management.” Many of its particular criteria sets note these differences. For example, GAD symptoms can present somatically in some cultures and cognitively in others. In schizophrenia, “ideas that appear to be delusional in one culture (e.g., witchcraft) may be commonly held in another. In some cultures, visual or auditory hallucinations with religious content (e.g., hearing God’s voice) are a normal part of religious experience.”73 Overall, the latest manual is more attentive to errors resulting from mistaking cultural differences with mental disorders. Still, future DSMs will continue to face challenges in distinguishing psychiatric dysfunctions from contextually and culturally explicable responses and social deviance. 


Future DSMs

The DSM’s evolution reveals the huge and possibly unresolvable difficulties in defining mental disorders. At first, the manual’s developers bet that combining organic conditions found in asylums with a psychodynamic model that captured the problems of outpatients was sufficient to meet psychiatry’s diagnostic needs. Next, they embraced a medical model that assumed a basic resemblance between mental and physical disorders but refrained from theoretical speculations about their causes. The DSM-5 unsuccessfully tried to adopt the statistical techniques and dimensional measurements of psychological research to define its conditions. The latest, as yet totally unrealized, efforts assume that neuroscientific research will point the manual in a more brain-based direction. Whether the research domain criteria, the NIMH’s ambitious new diagnostic system grounded in neural circuitry, will succeed is far from assured at present. 

The current DSM contains 22 general classes and nearly 300 specific diagnoses. Yet perhaps the most striking finding from neuroscientific research is that the hundreds of DSM diagnoses reflect variations on a small number of general processes that are loosely related to internalized, externalized, and psychotic disorders.74

However, a DSM organized around just three classes would be professionally unthinkable: such a system would have no medical authority. The lack of alternatives to using external symptoms as the basis for diagnosis ensures that future DSMs are unlikely to result in fundamental breakthroughs in understanding and treating mental disorders. 

The DSM-5 process sunk the credibility of the manual to levels not seen since the 1970s. Critics were not the vociferous anti-psychiatrists who objected to earlier versions but eminent figures within the profession, including Spitzer and Frances, the chairs of earlier DSM revisions, and former NIMH directors Steven Hyman and Thomas Insel. The leaders of the DSM-5 Task Force themselves vigorously critiqued the manual. Only after their efforts at paradigm change failed did they revert to defending the extant diagnostic system. The major guardians of the classification were clinicians who were skeptical about the DSM’s validity but required its diagnoses for practical purposes. Will it be possible to resurrect the DSM enterprise? 

The DSM-III era seems to be over: researchers no longer respect the symptom-based DSM entities and seek neuroscientific alternatives such as the RDoC. On the opposite direction, calls arise to return to the era when psychiatry focused on personal history and interpersonal connections. One group of British psychiatrists calls for junking the DSM system because what they call “good practice in psychiatry primarily involves engagement with the non-technical dimensions of our work such as relationships, meanings, and values.”75 

Despite continuing frustrations over establishing a valid diagnostic classification, the DSM won’t be replaced anytime soon. Patients require their diagnoses to obtain treatment, and many other factors determine eligibility. The current DSM fulfills psychiatry’s need for professional legitimacy; all medical specialties require specific diagnoses. Despite these advantages, the DSM system will persist for years because its diagnostic entities are closely intertwined with too many interests to give up. 

Observable symptoms persist in defining the DSM diagnoses, a situation that general medicine surmounted more than a century ago. The chemical and physical operations of the brain have yet to provide clues that may unravel the mysteries of human consciousness and its distortions. Mental disorders could require understandings that cannot be completely removed from personal life experiences. The future of the DSM is clearly at a crossroads, but the path it should take has no roadmap.


Footnotes upon request by writing to


For updates click homepage here