How an Age of Anxiety Became an Age of Depression

Abstract

Context: During the 1950s and 1960s, anxiety was the emblematic mental health problem in the United States, and depression was considered to be a rare condition. One of the most puzzling phenomena regarding mental health treatment, research, and policy is why depression has become the central component of the stress tradition since then.

Methods: This article reviews statistical trends in diagnosis, treatment, drug prescriptions, and textual readings of diagnostic criteria and secondary literature.

Findings: The association of anxiety with diffuse and amorphous conceptions of “stress” and “neuroses” became incompatible with professional norms demanding diagnostic specificity. At the same time, the contrasting nosologies of anxiety and depression in the Diagnostic and Statistical Manual of Mental Disorders III (DSM-III) extended major depressive disorder to encompass far more patients than any particular anxiety disorder. In addition, antidepressant drugs were not associated with the stigma and alleged side effects of the anxiolytic drugs.

Conclusion: Various factors combined between the 1970s and the 1990s to transform conditions that had been viewed as “anxiety” into “depression.” New interests in the twenty-first century, however, might lead to the reemergence of anxiety as the signature mental health problem of American society.

Anxiety was at the forefront of medical and psychiatric attention in the United States during the 1950s and 1960s. Yet since that time, depression—considered a rare disease in the post–World War II period—has become the focus of mental health concern. One of the most puzzling phenomena in the recent history of psychiatric diagnosis is why depression replaced anxiety as the most commonly treated and researched mental health condition associated with the stress tradition.

The stress tradition encompasses a diffuse and multifaceted array of psychic, somatic, and interpersonal problems that often arise as responses to the strains of everyday life (Selye 1968). The common psychological features of these problems include a mélange of symptoms involving nervousness, sadness, and malaise. The typical physical symptoms consist of headaches, fatigue, back pain, gastrointestinal complaints, and sleep and appetite difficulties, often accompanying struggles with interpersonal, financial, occupational, and health concerns. These complaints account for a large proportion of cases found in outpatient psychiatric and, especially, in general medical treatment.

Before the twentieth century, professionals and laypeople alike were likely to regard this varied combination of symptoms as a problem of “nerves,” emphasizing the somatic side of complaints (Shorter 1992). For much of the twentieth century, the equally amorphous terms stress and nervous breakdown captured the same heterogeneous range of psychic and somatic conditions (Swindle et al. 2000). During this era, anxiety and its sibling condition, “neuroses,” became the central themes of what came to be called the stress tradition, famously captured by the poet W.H. Auden's term age of anxiety for the fear and malaise afflicting the population after World War II (Auden 1947/1994).

In contrast, before the 1970s, depression was usually considered a relatively rare condition involving feelings of intense meaninglessness and worthlessness often accompanied by vegetative and psychotic symptoms and preoccupations with death and dying (Shorter 2009). Moreover, depression was more likely to be associated with hospitalized patients than with clients of general physicians or outpatient psychiatrists. But beginning in the 1970s until the present, depression rather than anxiety has become the common term used to indicate the breadbasket of common psychic and somatic complaints associated with the stress tradition. Depression now dominates clinical practice, treatment, and research in psychiatry as well as images of mental health problems in the broader culture (Horwitz and Wakefield 2007).

Why did depression replace anxiety as the featured condition in outpatient diagnosis and treatment as well as in the public consciousness during the last part of the twentieth century? It is difficult to even imagine any “real” cause—whether biological, psychological, or social—that could explain why the actual prevalence of one condition has risen at the same time as the other has fallen. Instead, several factors, including changing norms of psychiatric classification, professional and political advantage, and economic organization and marketing, came together toward the end of the twentieth century to transform an “age of anxiety” into an “age of depression.”

The Transformation of Anxiety into Depression

Before the 1970s, anxiety was the common term used to capture the nonspecific nature of the most common mental health problems seen in outpatient psychiatry and general medical practices (Herzberg 2009; Tone 2009). During this period, the cultural conception of anxiety was not so much as a particular type of psychiatric illness as a general psychic consequence of the demands and pace of modern conditions of life. Dominant theories emphasized how a variety of psychosocial stressors, especially family- and work-related problems, caused “stress,” “nerves,” and “tension,” all of which were manifestations of anxiety. Indeed, the ubiquitous nature of anxiety made it a symbolic condition of American society, as well as of psychiatry, in the post–World War II era.

Anxious patients were especially likely to be found in the offices of general physicians. As a leading expert on the treatment of psychiatric problems in primary medical care noted in 1968, “an abundance of tensions, fears, worries and anxieties confront mankind today, and, in fact, anxiety is seen in the majority of patients visiting the physician's office” (Rickels 1968, p. 10). One overview of the kinds of problems found in general medical practice asked the question, “What illnesses are being treated?” and answered, “Most of what primary care physicians see, they label ‘anxiety’” (Blackwell 1975, p. 29).

Anxiety conditions dominated the presentation of problems in both outpatient psychiatric practices and family medicine. Psychiatric diagnoses in the Diagnostic and Statistical Manual of Mental Disorders I (DSM-I, 1952) and DSM-II (1968) reflected the centrality of the “psychoneuroses,” which were grounded in anxiety. In 1962, for example, anxiety was the most prevalent psychoneurotic condition: According to the National Disease and Therapeutic Index, about 12 million patients received diagnoses of anxiety reactions, compared with just 4 million with diagnoses of neurotic depression (Herzberg 2009, p. 260). One large study at the time indicated that three-quarters of neurotic patients received an anxiety diagnosis, whereas most of the rest were simply considered “neurotic.” In contrast, depression was “absent from the diagnostic summaries” (Murphy and Leighton 2008, p. 1057).

The global conception of stress-related problems in the 1950s and 1960s affected mental health research as well as treatment, so the most prevalent categories of research in the major psychiatric journals explored both general topics (e.g., behavioral science) and policy issues (e.g., mental health services) (Pincus et al. 1993). Particularly during the last half of the 1960s, these journals featured publications using a psychosocial framework. Research on mental problems in the community also relied on measures that reflected a nonspecific view of psychic disturbance, although they emphasized symptoms of anxiety (Horwitz 2002). But because depression was associated with psychotic symptoms, questions about this condition were rarely found in epidemiological surveys.

In addition, the prevailing drug treatments during the 1950s and 1960s were directed at conditions considered to reflect problems of “anxiety.” A revolution in the treatment of mental health problems had begun in the 1950s when the development of meprobamate (Miltown) created the first mass market for treating problems of generalized stress (Healy 1997; Herzberg 2009). Miltown was called a tranquilizer and was marketed for the relief of the anxiety, tension, and stress associated with anxiousness and its accompanying somatic symptoms.

Miltown became the most popular prescription drug in U.S. history. By 1965, physicians and psychiatrists had written 500 million prescriptions for it (Smith 1985, p. 316), and as early as 1960, about three-quarters of all American physicians were prescribing Miltown (Tone 2009, p. 90). By the late 1960s, however, the spectacular success of the benzodiazepine Librium, which was introduced in 1960, displaced Miltown. In turn, Valium succeeded Librium as a blockbuster antianxiety drug, becoming the single most prescribed drug of any sort: 20 percent of all women and 8 percent of all men reported using a minor tranquilizer each year (Parry et al. 1973).

During the 1950s and much of the 1960s, the concept of “depression” barely existed for submelancholic conditions, and “antidepressant” medications were reserved mainly for serious depressive conditions found in hospitalized patients (Shorter 2009). In the DSM-I and DSM-II, nonpsychotic forms of depression were regarded as defense mechanisms used to allay underlying feelings of anxiety. In contrast, these manuals prominently featured depressive psychoses. “Thus, in these years,” according to epidemiologist Jane Murphy and psychiatrist Alexander Leighton, “depression was usually thought of as a psychotic rather than a neurotic disorder” (2008, p. 1056). General physicians rarely prescribed antidepressants, and they were far overshadowed by the tranquilizers in the public consciousness. Because physicians and psychiatrists treated less severe forms of depression with the minor tranquilizers, no market existed for antidepressant drugs aimed at depressive conditions that were not serious. Depression was also of relatively minor importance in American popular magazines. “In the 1960s,” reports historian Laura Hirshbein, “there were three times as many articles about anxiety as there were about depression” (2009, p. 59).

Then, beginning in the 1960s, clinicians and researchers started to pay more attention to depression, especially emphasizing its prevalence among patients in primary medical care (Ayd 1961). This led advertisers to begin to place ads for the antidepressant tricyclics and monoamine oxidase inhibitors (MAOIs) in medical and psychiatric journals. By the end of the decade, the disparity between anxiety and depressive diagnoses thus had narrowed, although anxiety was still far more common than depression. In fact, depressive diagnoses in outpatient treatment grew to 8 million, whereas those of anxiety remained at around 12 million (Herzberg 2009, p. 260).

Treatment statistics during the 1970s reflected the growing interest in depression. During the first half of that decade, the management of depression became as common as that of anxiety (IMS America 1976, pp. 125–26), and by 1975, the 18 million diagnoses of depression surpassed the 13 million diagnoses of anxiety. From 1980 to the present, the upward trajectory of depressive diagnoses has been especially apparent. Between 1987 and 1997, the proportion of the U.S. population receiving outpatient treatment for conditions called “depression” increased by more than 300 percent (Olfson, Marcus, Druss, Elinson, et al. 2002). In 1987, 0.73 persons per hundred adults in the United States were treated for depression, but by 1997, these rates had leaped to 2.33 per hundred. While 20 percent of patients in outpatient treatment in 1987 had a diagnosis of some kind of mood disorder, most of which were major depressive disorder (MDD), depressive diagnoses nearly doubled by 1997 to account for 39 percent of all outpatients.

In contrast, the rates of any anxiety diagnosis for treated patients rose much more slowly, from 10.5 percent in 1987 to 12.5 percent in 1997 (Olfson, Marcus, Druss, and Pincus 2002). By 1996/1997, however, diagnoses of mood disorders were more than three times as common as anxiety diagnoses in office-based psychiatry (Mojtabai and Olfson 2008). A large study of psychiatric practice that the American Psychiatric Association (APA) conducted in 1997 is illustrative, finding that more than half of patients had mood disorders and about a third had a principal diagnosis of MDD, whereas just 10 percent had received a diagnosis of an anxiety disorder (Pincus et al. 1999).

Recent figures present a mirror image of the overwhelming dominance of anxiety in general medicine and psychiatry during the 1950s and 1960s. In 2002, 51.7 million outpatient visits were for mental health care. Depression accounted for 21 million of these, compared with only 6.2 million for anxiety (see http://www.cdc.gov/nchs/fastats/mental.htm). Likewise, by the early part of the twenty-first century, general physicians were more than twice as likely to make diagnoses of depression than anxiety (Schappert and Rechtsteiner 2008). In sharp contrast to the much faster growth of depressive than anxiety diagnoses, epidemiological studies indicate that rates of the actual amounts of both depression and anxiety remained relatively constant from the early 1990s through the early 2000s (Kessler et al. 2005). For whatever actual problems people sought mental health care, the treatment system and, in all likelihood, the patients themselves were calling them “depression.” For example, depression is the single most common topic of online searches for pharmaceutical and medical products, attracting nearly 3 million unique visitors over a three-month period in 2006 (Barber 2008, p. 14).

The takeover of the stress marketplace by the “antidepressant” class of selective serotonin reuptake inhibitors (SSRI) medications strengthened the association between common mental health problems and depression. When the SSRIs came on the market in the late 1980s, antianxiety drugs were about twice as likely to be prescribed in outpatient visits as were antidepressants (Olfson and Klerman 1993). But at that point, the trends changed abruptly. Between 1985 and 1993/1994, prescriptions for antianxiety drugs plunged from 52 to 33 percent of all psychopharmacological visits, and the number of users of antianxiety drugs grew very slowly after that, rising from 5.5 million to 6.4 million in 2001 (Zuvekas 2005).

Conversely, from 1996 to 2001, the number of users of SSRIs increased rapidly, from 7.9 million to 15.4 million. By 2000, antidepressants were the best-selling category of drugs of any sort in the United States; fully 10 percent of the U.S. population was using an antidepressant (Mojtabai 2008). In fact, these drugs were used so widely in general medical practice that in 2003/2004, 310 of every 1,000 female patients received a prescription for an antidepressant (Raofi and Schappert 2006). Prescriptions for SSRIs continued to grow, and by 2006, Americans had received more than 227 million antidepressant prescriptions, an increase of more than 30 million since 2002 (IMS Health 2006). Antidepressants were prescribed for mood and anxiety disorders alike, gaining unchallenged control of the market once held by the anxiolytic drugs (Mojtabai and Olfson 2008).

The conditions and treatments of the stress tradition thus underwent a widespread transformation between 1955 and the present. The heyday of anxiety during the 1950s and 1960s was followed by its steep decline beginning in the 1970s, accelerating during the 1980s and 1990s, and stabilizing in the early 2000s. Over the past half century, those mental health conditions in physicians’ offices, psychiatric clinics, research, and popular culture that were seen as problems of “anxiety” came to be called “depression.” Likewise, antidepressants replaced anxiolytics for their treatment. What factors account for this major relabeling of mental health problems?

How Depression Captured the Stress Marketplace

Diagnostic Specificity

Diagnostic specificity has been the master trend in the recent history of psychiatric classification. For most of history, only a few imprecise categories, such as mania, melancholia, and hysteria, were used to describe severe psychiatric conditions. No distinct diagnoses or treatments were given for common mental problems that did not feature serious symptoms. But during the twentieth century, scientific norms increasingly demanded that medicine, including psychiatry, treat specific diseases. “This modern history of diagnosis,” according to historian Charles Rosenberg,

is inextricably related to disease specificity, to the notion that diseases can and should be thought of as entities existing outside the unique manifestations of illness in particular men and women: during the past century especially, diagnosis, prognosis, and treatment have been linked ever more tightly to specific, agreed-upon disease categories. (Rosenberg 2008, p. 13)

The stress condition's variable and fluctuating mixture of psychic distress, somatic problems, and life difficulties (i.e., “problems of living”) lacked the diagnostic specificity needed to give disease entities medical legitimacy. Although pure forms of anxiety and depression do exist, they are the exceptions rather than the rule. Indeed, the simultaneous presence of anxious and depressed symptoms is far more common than isolated forms of each condition, as more than two-thirds of people with major depression also report an anxiety disorder (Kessler et al. 2003, p. 3101). Nevertheless, beginning in the 1970s, the psychiatric profession was pressured to embrace the norms of diagnostic specificity accepted in the rest of medicine as the standard for definitions of their subject matter (Horwitz 2002).

Before the 1970s, the ill-defined, amorphous, and protean conditions that patients brought to general physicians and mental health specialists did not pose a major problem for the psychiatric profession. The supremacy of psychodynamic perspectives meant that diagnostic norms did not dictate sharply bounded, discrete categories of disorder. Instead, explanations emphasized unconscious mechanisms that were manifest in a variety of overt symptom formations, and treatments used psychotherapeutic methods that were not specific to particular symptoms (Rycroft 1968). Moreover, at that time, most clients paid directly for their therapy, so there were no third-party payers to require specific diagnoses. Neither theoretical nor financial concerns forced psychiatry to differentiate among various types of disorders.

During the 1970s, however, this situation began to change rapidly as demands for specificity placed tremendous pressure on psychiatry to alter its diagnostic system. Generalized conceptions—whether “psychoneuroses,” “stress,” or “nerves”—became a millstone around the neck of the profession (although they remained common in popular discourse). The unreliability of the DSM-II's cursory diagnoses subjected psychiatry to much criticism, ridicule, and even questions about its legitimacy. Prominent critics such as Thomas Szasz (1974) mocked psychiatry because it could not even define its central domain of “mental illness.” Others, like D.L. Rosenhan (1973), conducted highly publicized studies purporting to show that psychiatric labeling worked to hospitalize people who were not sick at all. Psychiatry was under attack from many fronts, including the libertarian right, the Marxist left, and feminists, all of whom focused on its perceived suppression of individual freedom.

Discontent was growing as well among mainstream members of the psychiatric establishment as stinging critiques from within its own ranks questioned its knowledge base. Many academic studies, most prominently the U.S.–U.K. project that systematically compared the diagnostic practices of American and British psychiatrists, indicated that even the most basic psychiatric categories had appallingly low reliability (Cooper et al. 1972). Moreover, the reigning psychosocial model did not provide a solid grounding for why psychiatrists—as opposed to many other professionals, including clinical psychologists, counselors, social workers, and nurses—should have professional dominance over the treatment of mental illnesses. Psychiatry, which in the twentieth century always had a shaky position in the prestige hierarchy in medicine, was in danger of losing both its legitimacy as a scientific discipline and its authority in the broader culture. It became clear that the maintenance of psychiatric authority depended on replacing conceptions of “psychoneuroses” and “stress,” which were at the heart of the DSM-II's diagnostic system.

The National Institute of Mental Health (NIMH) also faced a serious crisis in the 1970s. During the 1950s and 1960s, the agency had emphasized the study of general personality, developmental, and social issues, which were more closely related to the stress tradition than to specific types of mental illness. It awarded 60 percent of its grant funding to psychologists and social scientists and less than 40 percent to psychiatrists and other medical and biological scientists (Grob 1991, pp. 66–67). After Richard Nixon became president in 1968, his administration and Congress began to attack the NIMH for sponsoring research on social problems such as poverty, racism, and violence. Although this type of research accounted for about a fifth of the institute's portfolio, it was a lightning rod for attacks on its overall mission. Psychosocial research thus had become a political liability in the institute's efforts to secure funding from Congress and the executive branch (Baldessarini 2000).

By the late 1970s, biologically oriented researchers had joined the fight against psychosocial research by the NIMH. They were deeply concerned that research on social problems and generalized psychosocial conditions would damage the institute's reputation and subject it to a backlash against all its research programs (Kolb, Frazier, and Sirovatka 2000). These researchers argued that a narrower focus on the study of specific mental disorders would both enhance the quality of scientific research and justify the institute's mission in the face of political opposition.

Around the same time, family advocacy groups became a major lobbying force in the NIMH (Schooler 2007, p. 60). These groups, like the National Alliance for the Mentally Ill, were primarily composed of family members with children suffering from severe mental illnesses. They lobbied the NIMH to shift its focus from broad social research to the study of the biological underpinnings of and treatments for specific mental disorders. These efforts culminated in a 1982 directive from Congress ordering the NIMH to stop its support of social research (Kolb, Frazier, and Sirovatka 2000, p. 223). The transformation from research on general psychosocial problems to specific, biologically based diseases was a great success, and beginning in the early 1980s, funding for the institute sharply increased.

Another spur toward specificity of diagnosis was the mandate from the Food and Drug Administration (FDA) to the pharmaceutical industry to target psychoactive drugs to specific biomedical conditions (Healy 1997). During the 1950s and 1960s, the popularity of the benzodiazepines stemmed from their effectiveness as remedies for general life stresses and protean conditions of anxiety, with little consideration of whether or not they treated explicit disease states. Studies during the 1950s and 1960s found that only about a third of the minor tranquilizers were prescribed for specific mental disorders, while the rest were given as a response to more diffuse complaints and psychosocial problems (Cooperstock and Lennard 1979; Raynes 1979; Shapiro and Baron 1961). For example, a review of psychoactive medication at the time concluded that “only about 30 percent of use is in identified mental disorders and the remainder covers the rest of medicine” (Blackwell 1973, p. 1638). The vocabulary of the era dictated that these drugs would be called antianxiety or tranquilizing drugs, and the problems they treated were considered problems of generalized anxiety, although they often involved co-occurring depression.

Pharmaceutical companies presented the tranquilizers to physicians and psychiatrists as drugs that treated a variety of nonspecific complaints, including anxiety, tension, depression, and mental stress. Advertisements (which at the time were directed at physicians, not consumers) emphasized that these drugs provided relief for such common problems as dealing with unruly children, traffic jams, demanding bosses, and housekeeping (Herzberg 2009). In the 1970s, however, government regulators began to enforce more stringently the legislative requirement dating from 1962 that drug companies target the marketing of their products to particular biomedical conditions (Smith 1985, chap. 9). Moreover, the FDA also began to require that drugs be efficacious as well as safe (Hamilton 1997). In addition, growing coverage by private and public insurance meant that few patients paid the bulk of their mental health treatment costs, and third-party payers reimbursed practitioners only for treating a specific disease. These factors placed pressure on mental health providers to call the conditions they treated diseases or disorders rather than more amorphous “problems of living.”

The emphasis on generalized conditions had been suitable for an era when psychodynamic explanations that emphasized unconscious mechanisms were dominant; the most seriously ill patients with more specific conditions such as schizophrenia and manic depression were concentrated in inpatient institutions and so were rarely found in outpatient settings; clients paid for outpatient treatment out of their own pockets; and therapies were nonspecific. By 1980, though, it was apparent that classifications focusing on specific disease entities were needed to increase psychiatry's professional legitimacy and meet regulatory and insurance standards. For psychiatry, however, the recategorization of the nebulous conditions in the stress tradition as specific diagnostic entities had to support the specialty's ascendant position in the huge market of stress-related conditions. By the late 1970s, the need for specificity in psychiatric diagnosis was clear, so the next question is why depression rather than anxiety took center stage in psychiatry's reinvention of its diagnostic system.

The Rise of Biological Psychiatry

Professional competition within psychiatry is one reason for the rise of depression and the decline of anxiety as the discipline's central point of reference. Psychodynamically oriented psychiatrists emphasized anxiety-related conditions but paid relatively little attention to depression. During the 1970s, however, a group of biological psychiatrists became intensely concerned about the unscientific nature of psychoanalysis and the damage it was doing to the development of psychiatry as a branch of medicine. Research-oriented psychiatrists who generally favored biological perspectives led the opposition against the DSM-II and its etiologically based and unreliable diagnostic categories. They were far more interested in studying specific diseases than amorphous stress conditions, and at that time, because of the close connection of anxiety to the psychodynamic tradition, depression was a more effective vehicle than anxiety to realize the scientific aspirations of the biological psychiatrists. “Depression,” asserts historian Laura Hirshbein, “became a phenomenon around which professionals in the latter part of the twentieth century made claims about psychiatry's status as a scientific specialty” (2009, p. 28).

Depression fit the professionally desirable conception of a severe and specific disease that could be associated with biological causes. Indeed, depression was considered to be a very serious disease connected to suicide and psychosis and so, for the most part, lay outside the stress tradition. In addition, far more than anxiety, depression was theoretically grounded in brain chemistry and conceptions of chemical imbalances. The two most significant biological articles in psychiatry during the 1960s explored the relationship between low levels of biogenic amines in the brain and depressive illness (Bunney and Davis 1965; Schildkraut 1965). These early breakthroughs cemented the coupling of biological approaches and depressive conditions. The biological grounding of depression heightened its appeal to the research-oriented psychiatrists who were in charge of revising the DSM.

The DSM-III

Although biological psychiatry and its central vehicle of depression were gaining ground during the 1970s, the implementation of the third edition of the Diagnostic and Statistical Manual (DSM-III), which the APA issued in 1980, was the central turning point leading to the transition from anxiety to depression. This manual radically changed the nature of psychiatric diagnoses, based on the foundational principle that diagnostic criteria should not assume any particular etiology of symptoms. This strategy allowed its advocates to claim theoretical neutrality and so mitigate the opposition of clinicians who did not adhere to the core group's biological orientation (Horwitz 2002). The goal of purging etiological assumptions from the new manual was especially consequential for the anxiety disorders.

The unifying concept of DSM-I and DSM-II was that the symptoms of all psychoneuroses were defenses against underlying anxiety. A successful attack on this etiological concept required the wholesale destruction of its global concept of anxiety. As an alternative, the DSM-III developed definitions of various specific conditions underscoring that each was a discrete and qualitatively distinct disease (APA 1980). Unlike the DSM-I and DSM-II, which had placed both depression and anxiety within the same psychoneurotic category, the DSM-III formulated anxious and depressive conditions as completely different. It also carved away conditions such as hysteria and hypochondriasis, which had previously been core aspects of anxiety-related states, putting them into distinct groups. The psychoneuroses were split into four separate general categories: anxiety, affective, dissociative, and somatoform disorders.

Four particular aspects of the differential definitions of the anxiety and affective disorders in the DSM-III facilitated the desirability of using major depressive disorder (MDD) rather than any single anxiety condition as a diagnosis for what had been considered to be general stress conditions. The first was the very different way in which the new manual differentiated among the various conditions of anxiety and depression. No single category of anxiety was preeminent. The DSM-III also separated conditions that were formerly viewed as reflecting underlying anxiety, such as depression, hysteria, and somatization, into their own categories. In addition, the anxiety classification was divided into phobic states, anxiety, and posttraumatic stress disorder (PTSD), and numerous subtypes of each (APA 1980, pp. 225–39). For example, there were several types of phobias, including simple phobia, social phobia, and agoraphobia, which itself was divided into conditions that did or did not display panic disorder. Generalized anxiety disorder, which on its face might be viewed as the core anxiety condition, was instead made a residual category, to be diagnosed only when symptoms of phobic, panic, or obsessive compulsive disorders were not present.

Subsequent research illustrates the impact of the differentiation of anxiety into a number of disorders without a focus on any particular condition. After 1980, the vast majority (83 percent) of studies published in psychiatry, psychology, and related fields centered on a single anxiety disorder (Norton et al. 1995). These journals featured panic disorder/ agoraphobia (36 percent), PTSD (28 percent), and obsessive compulsive disorder (27 percent) to an almost equal degree. Less than 10 percent of the articles were on generalized anxiety disorder, which had been the central anxiety condition in psychodynamic theory. Current studies of anxiety remain balkanized and without a central focus (Boschen 2008).

The DSM-III's treatment of depression sharply contrasted with the division of the anxiety disorders into many distinct conditions. Major depressive disorder (MDD) was the only significant category of nonpsychotic depression among the affective disorders. Psychotic forms of mood disorders were identified with bipolar disorder, which was the sole psychotic state of any note within the larger affective disorders category. Unipolar states of psychotic depression were virtually indistinguishable from MDD. Melancholic depression—the central depressive condition before the DSM-III—became a subcategory of MDD (APA 1980, p. 215). People could qualify for a diagnosis of melancholy, which required symptoms of greater severity in the morning, early-morning awakening, marked psychomotor retardation, weight loss, and excessive guilt, only if they already had met the criteria for MDD. The submersion of melancholia into the broader MDD category ensured its fall into obscurity (Zimmerman and Spitzer 1989).

The condition of dysthymia supposedly was created to be a form of minor depression that would contrast with MDD (Shorter 2009). While the three necessary symptoms of dysthymia (raised to four symptoms in subsequent editions of the manual) might have led it to become a suitable label for many people with conditions linked to the stress tradition, the diagnosis was given only to those adults whose symptoms had lasted for at least two years (APA 1980, p. 222). This precluded anyone except those with the longest-standing conditions from receiving a diagnosis of dysthymia.

MDD was clearly the singular nonpsychotic diagnosis in the affective disorders category, encompassing a range of conditions spanning from melancholia through the depressive neuroses to short reactive depressions. The MDD symptoms captured both the amorphous and short-lived psychosocial problems that marked the stress tradition and the serious and chronic conditions that in the past had been associated with melancholic depression.

Research on depression subsequently reflected the overwhelming dominance of the MDD category. Beginning in 1981, MDD began a steep upward trajectory and by 2000 had a citation rate about five times higher than that of all other depressive labels combined (Blazer 2005, p. 28). Citations for conditions of “melancholic,” “endogenous,” or “psychotic” depression fell dramatically beginning in the early 1980s and had almost disappeared by 2000. Likewise, dysthymia never gained traction as a central psychiatric diagnosis (McPherson and Armstrong 2006). Unlike anxiety, with its multiple fields of research and publication, depression was almost completely identified with MDD. Major depressive disorder was unquestionably the core nonpsychotic affective disorder, which helped it replace anxiety as the heir to the stress tradition.

A second reason why depressive diagnoses captured the stress tradition from anxiety had to do with the DSM-III's allocation of the most general symptoms of distress to the different major diagnostic categories. The definition of MDD included such global symptoms as sadness, sleep and appetite difficulties, fatigue, and lack of concentration, which afflicted many people with mental health problems that fell into the stress tradition (APA 1980, pp. 213–14). The capacious MDD criteria thus could cover a heterogeneous group of people ranging from irritable adolescents who constantly sleep, eat little, are uninterested in school, and do not concentrate on their schoolwork to morose elderly people who cannot sleep, overeat, are fatigued, and feel worthless (Murphy 2006, p. 329). In contrast, the diagnostic criteria for the various anxiety disorders were far more specific and centered on narrower manifestations such as intense fears of specific objects or situations, obsessions and compulsions, and posttraumatic stress.

The anxiety diagnosis that could have encompassed the generalized aspect of symptoms of stress—generalized anxiety disorder (GAD)—became a mere phantom in the DSM-III. The hierarchical system of diagnosis in the DSM-III privileged diagnoses of depression over those of anxiety: anxiety diagnoses would not be made in the presence of coexisting depressive disorders. Because of the extensive co-occurrence of depressive and anxious symptoms, this increased the likelihood of making depressive rather than anxiety diagnoses. Moreover, GAD could not be diagnosed in the presence of other anxiety conditions (APA 1980, p. 232). Because GAD almost always was found together with these conditions, it was rarely diagnosed at all. Finally, the criteria for diagnosing GAD bewildered clinicians. “In fact,” psychologist David Barlow summarized, “the category of GAD in DSM-III produced so much confusion that few clinicians or investigators could agree on individuals who would meet this definition” (1988, p. 567). The new diagnostic criteria therefore made MDD a more appropriate label than anxiety for the ubiquitous symptoms of stress that so many patients displayed.

Third, the duration criteria for the anxiety conditions were considerably longer than those for MDD. Most anxiety diagnoses required “persistent” symptoms, usually of at least six months’ duration (APA 1980, p. 227), which ruled out diagnoses of short-lived responses to stressful conditions. In contrast, symptoms that endured for a mere two weeks met the MDD qualifications (APA 1980, p. 213). Transient responses to stress, therefore, could meet diagnostic criteria for depression but not anxiety.

Finally, and perhaps most important, the disparate treatment of the contextual basis of anxiety and depression favored diagnoses of depression over those of anxiety. A very high proportion of patients enter mental health and, especially, primary medical care settings with psychosocial problems of stress that are often the proximate reasons for their symptoms. Yet the diagnostic criteria for the anxiety diagnoses were hedged with many qualifiers that distinguished them from contextually appropriate symptoms. For example, only “irrational” or “unreasonable” fears qualified for diagnoses of phobias, thus ruling out proportionate and reasonable fears (APA 1980, pp. 227–30). Or panic disorders had to occur “unpredictably” and could not be responses to life-threatening situations (APA 1980, p. 230). The treatment of anxiety according to the DSM-III, therefore, ruled out proportionate responses to dangerous situations as possible diagnoses.

In contrast, many patients reacting to stressful psychosocial contexts could meet the MDD criteria. Bereavement was the sole relevant exclusionary criterion for depression: someone grieving the death of an intimate who otherwise met the MDD criteria would not be so diagnosed so long as his or her symptoms were not especially severe or long lasting (APA 1980, p. 213). But no comparable exclusions were made for people who met the criteria after they were laid off from jobs, rejected by romantic partners, or informed of a serious medical diagnosis for themselves or an intimate. Unlike the diagnostic criteria for the anxiety disorders, the MDD criteria did not preclude diagnoses even when the symptoms were proportionate responses to the losses that provoked them. The range of conditions in the stress tradition that featured mixed depressive and anxious symptoms thus became more amenable to depressive than anxious diagnoses.

Whether the problems that people bring to therapy have changed much over the past half century is questionable (Swindle et al. 2000), although their labels have dramatically altered. The DSM-III unintentionally created the conditions for depression, rather than anxiety, to incorporate the disparate manifestations of stress and thus become the central diagnosis in the mental health system.

From Anxiolytics to Antidepressants

A major consequence of the DSM-III's new categorizations was to make depression a more promising target for the new class of antidepressants—the selective serotonin reuptake inhibitors (SSRIs)—that came on the market in the late 1980s. The SSRIs now dominate the treatment of nonpsychotic mental disorders, including MDD and the various anxiety disorders as well as many other conditions. In practice, there is little evidence that the SSRIs’ efficacy has any relationship to the diagnostic categories in the DSM. They act very generally to increase levels of serotonin in the brain that both raise low mood states and lower levels of anxiety, so when they first appeared in the late 1980s, the antidepressant SSRIs could just as easily have been marketed as antianxiety medications (Healy 2004).

By the late 1980s, however, it made more sense for drug companies to market products aimed at the wide array of stress conditions as “antidepressants” rather than “anxiolytics.” A sharp backlash against the anxiolytic drugs had developed in the early 1970s when the media turned sharply against their use, showing in many stories their addictive potential, use in suicide attempts, and other negative side effects. In response, patients, with backing by organized advocacy groups, filed numerous lawsuits against the manufacturers of these drugs (Gabe 1990). In addition, the rise of the feminist movement, which harshly assailed these drugs because of their assumed role in upholding patriarchal norms and keeping women confined in oppressive social roles, was another nail in the tranquilizers’ coffin (Herzberg 2009).

The result was, according to historian Edward Shorter, a “general hysteria about addiction from pharmaceuticals that swept American society in the 1970s” (2009, p. 116). Stimulated by hostile congressional hearings, government agencies, including the FDA and Drug Enforcement Administration (DEA), confronted the pharmaceutical industry and attempted to restrict the use of the benzodiazepines. This backlash resulted in their classification in 1975 by the DEA as Schedule 4 drugs, which required physicians to report all prescriptions written for them and limited the number of refills a patient could obtain.

These developments changed the use of anxiolytics. After twenty years of steadily rising sales since their introduction in the mid-1950s, consumption of this class of drugs plunged. From a peak of 104.5 million prescriptions in 1973, the number dropped to 71.4 million by 1980 and continued to plummet throughout the 1980s (Smith 1985, p. 33). Pharmaceutical companies had difficulty marketing antianxiety drugs. “By the mid-1980's,” writes David Healy, “it had become impossible to write good news stories about the benzodiazepines” (2004, p. 225). Moreover, because their patents had expired, pharmaceutical companies had no interest in either promoting the anxiolytic drugs or conducting new trials that could show their safety and efficacy.

Despite the growing interest in depression in the 1960s and 1970s, antidepressant drugs did not gain any traction in the general marketplace of stress conditions and were rarely prescribed in general medical practice but usually were reserved for the most seriously ill patients. Despite their relative invisibility compared with that of the tranquilizers, antidepressants had several marketing advantages over the anxiolytic drugs. Unlike the tranquilizers, which became popular because they could be used to treat a wide array of common psychosocial problems of people in the community, the early antidepressant drugs—the tricyclics and MAOIs—were prescribed for the problems of severely depressed populations. This connected them with the newly desirable notions of specificity, in contrast to the tranquilizing drugs’ ubiquitous range of effects. In addition, the antidepressants were not linked to the problems of addiction and dependency associated with tranquilizers.

Because the FDA required manufacturers to prove a drug's efficacy for some biomedical condition, the SSRIs could not be marketed for generalized distress, but only for specific diseases. In contrast to the many particular anxiety disorders of the DSM-III, the unification of depression around the MDD criteria put depression in the best position to encompass the amorphous symptoms of the stress tradition. Given the hostile cultural and regulatory climate surrounding antianxiety drugs when the SSRIs came onto the market in the late 1980s, it made much more marketing sense for manufacturers to promote them as antidepressants than as antianxiety agents.

As has so often happened in psychiatric history, the development of a treatment shaped the nature of the illness that it was supposedly meant to treat. Network television shows, national newsmagazines, and best-selling books widely featured the SSRIs as antidepressant medications. In particular, the publication in 1993 of Peter Kramer's wildly popular Listening to Prozac cemented the coupling of the SSRIs with the treatment of major depression. Advertisements for Prozac soon began using the imagery depicted in Kramer's book, using slogans such as “better than well” and showing women cheerfully fulfilling both work and family roles. Although this imagery differentiated the SSRIs from the clientele of the older antidepressants, it positioned this class of drugs as the direct heir of the tranquilizers. Because a drug was called an antidepressant, depression seemed to be the condition that was being treated. Much as “anxiety” had during the 1950s and 1960s, “depression” came to refer to the disparate experiences of suffering connected to the stress tradition during the 1990s and early 2000s.

The FDA's loosening of restrictions on direct-to-consumer drug advertisements in the late 1990s both enhanced the popularity of the SSRIs and reinforced their link to depressive illness. Many of these ads were aimed at selling the disease of depression itself, rather than a particular type of antidepressant (Healy 1997; Hirshbein 2009). They relentlessly pushed the view that “depression is a disease” linked to deficiencies of serotonin in the brain. Advertisements typically connected the most general symptoms of depression from the DSM's diagnosis—sadness, fatigue, sleeplessness, and the like—with common situations involving interpersonal problems, workplace difficulties, or overwhelming demands, themes similar to the messages of ads in the 1950s and 1960s. What is different is that the psychic consequences of these problems were now being called depression instead of anxiety, tension, nerves, or stress.

The Return of Anxiety?

The transition of the age of anxiety into the age of depression demonstrates that diagnoses are contingent on the impact of changing social circumstances. The emphasis placed on any particular type of mental illness also may be influenced by the relative amount of attention that other types of mental health problems receive. If so, the rise (or decline) of one type of diagnosis may lead another type to fall (or increase). There are some signs, in fact, that anxiety could displace depression and recapture its hold on the stress tradition.

Many of the patents for SSRIs to treat depressive conditions have expired, so that far cheaper generic drugs are threatening to take market share and drastically lower the profits derived from the trademarked brands (Druss et al. 2004). In addition, the rapid rise of bipolar conditions that are treated with second-generation antipsychotic drugs is fracturing the market for depression treatments. In the past, what would have been diagnosed as MDD are now called bipolar conditions that can be treated with lucrative patented medications. Indeed, by 2008 antipsychotics were the most profitable class of any kind of drug (IMS Health 2008). Economic considerations seem likely to drive the pharmaceutical market away from depression as the condition to be treated.

Anxiety should become a particularly attractive target for trademarked SSRIs. More than a quarter of the population experiences enough symptoms of anxiety disorders to meet the DSM's criteria, making the group of anxiety disorders the most prevalent of any general category of mental health conditions (Kessler et al. 2005). As one marketing report points out,

Anxiety disorders are considered the most prevalent of psychiatric disorders. However, poor diagnostic rates and treatment outcomes mean there is still considerable scope for manufacturers to move into the anxiety market… . Despite a fifth of the total population across the seven major markets suffering from an anxiety disorder only a quarter of these individuals are diagnosed and therefore treated. As a result, drug manufacturers are failing to maximize revenues from the anxiety disorders market. Investment in awareness campaigns is essential. (Rose 2006, p. 471)

The boundaries between depression and anxiety are permeable enough that the same drugs can easily be marketed as responses to anxiety rather than to depressive disorders.

While the differentiation of the many forms of anxiety in the DSM-III initially enhanced the appeal of the unitary condition of MDD, each form of anxiety now can become a segmented market. For example, in 1999 the FDA approved Paxil for the treatment of social anxiety disorder (SAD) and Zoloft for PTSD; two years later Paxil and Effexor gained approval for the treatment of generalized anxiety disorder (GAD). Old drugs can seem innovative and up-to-date when they are prescribed for new indications. Different brands can target a variety of specific types of anxiety conditions and capture distinct niches. The extraordinary success of GlaxoSmithKline's efforts to promote Paxil as a treatment for SAD indicates the huge potential of anxiety conditions as pharmaceutical targets. A vast advertising campaign blitzed the media shortly after Paxil was approved in 1999 for treating SAD, which previously had been viewed as a rare disorder. Paxil became the largest-selling antidepressant at the time, with sales of $3 billion a year, since consumers now widely recognized anxiety as a reason to seek drug treatment.

Transformations also may move from the class of antidepressants to that of anxiolytic drugs. The same sort of reaction to the anxiolytic drugs that occurred in the 1970s shows signs of reemerging against the SSRIs, with questions being raised about the efficacy, side effects, potential addictiveness, and safety of this class of drugs (e.g., Bass 2008; Shorter 2009). If this backlash against the SSRIs grows, drug companies could develop a marketing strategy that will once again emphasize the anxiolytic drugs as first-line treatments for problems of stress. Indeed, while still dwarfed by the antidepressant market, benzodiazepine prescriptions grew from 69.4 million in 2002 to 80.1 million in 2006 (IMS Health 2008).

The diagnosis of depression is no longer as useful to psychiatry as it was over the past quarter century. The profession's scientific credibility is now far greater than it was in the 1970s; its diagnostic system is generally regarded as reliable; and its biological models are widely accepted. Most important, the drugs used to treat depression have lost their patents. The reasons for psychiatry's turn away from anxiety—its association with the indefinable category of psychoneurosis, with psychodynamic treatments, and with the presumed addictive nature of the tranquilizers—have long been forgotten, leaving no traces in cultural memory. But anxiety has never gone away, and it will be surprising if such a ubiquitous and universal condition does not once again come to the forefront of the stress tradition.

Conclusion

Before the 1970s, a broad conception of mental health problems, with stress and anxiety at its core, dominated mental health treatment, research, and policy. Drug treatments for these problems were widespread but coexisted with a variety of psychotherapeutic and more general social approaches. Beginning in the 1970s and accelerating since that time, protean stress conditions were transformed into the particular diagnostic categories that are now foundational in psychiatric classification. The political and economic circumstances the profession confronted in this era led depression to be a more attractive vehicle than anxiety for realizing psychiatry's ambitions to become a scientifically respectable branch of medicine. At the same time, the needs of pharmaceutical companies led depression to become the focus of its marketing efforts.

The movement from generalized conditions grounded in anxiety to specific disease categories dominated by depression has had major consequences for mental health policy. Because the targets of mental health treatment came to be viewed as specific diseases, the use of drug treatments, particularly antidepressants, soared. Responses to common mental health problems became equated with the prescription of medication at the expense of alternative psychotherapeutic approaches. At the same time, the use of psychological and social options for common mental health problems has substantially declined, despite evidence that they are at least as effective treatments as pharmaceuticals (and that the combination of various therapies can be the most successful of all).

Psychiatric classifications inevitably reflect the social forces prevailing in any particular historical era (Brown 1995). Which conditions are diagnosed and how they are treated depend not only on the symptoms that patients display but also on factors that include professional fashions in diagnoses, the financial rewards from various treatments, the activities of various interest and regulatory groups, cultural images of disorder, and the concerns of funding agencies. The amorphous psychic, somatic, and interpersonal problems that bedeviled humans long before the emergence of standardized diagnostic categories will continue to underlie whatever specific labels are used to classify them. Anyone interested in mental health policy issues should consider the extent to which the kinds of diagnostic labels that are most marketable in a particular historical context drive treatment, research, and policy, sometimes to the detriment of optimal mental health care.