AS the New York Times tells the story, Sherri Souza's husband is a National Guardsman posted in Iraq whose long-anticipated return home was canceled after the Pentagon unexpectedly extended his tour of duty. Like most spouses in this situation, Mrs. Souza is acutely disappointed. She misses her husband, worries about his safety, and is anxious for her family's future should he be killed or injured. When his scheduled e-mails are late, she becomes distressed and sometimes crawls into bed to await word of his safety. In the past, she might have described herself as very "sad,""lonely," or "worried." Now, however, she characterizes herself as "depressed." She is taking medication for her symptoms.
The characterization of our emotional reactions to life's challenges as "depression" is more than just a change in colloquial expression. It represents a transformation in psychiatric thinking. Psychiatrists are diagnosing more and more of the population as "depressed," by which they primarily mean the medical condition of major depressive disorder. Psychiatric epidemiological studies indicate that depression now afflicts about 10 percent of adults in the United States each year and about a quarter of the population at some point in their lives. This number has been steadily growing, they say: For the past several decades, each successive generation has reported more depressive disorders than the previous one. These enormous numbers have mobilized psychiatry, general medical practice, and the psychopharmacology industry to mount a coordinated (and profitable) offensive. Today, better recognition of unreported, hidden, or "sub-clinical" depression (that is, depression exhibiting fewer than the number of symptoms usually required for diagnosis) pushes prevalence numbers ever higher.
Not only the number of people said to be depressed but also the number actually treated has increased greatly in recent years. The percentage of the overall population in mental-health treatment for "mood disorders," the category of psychiatric disorder that includes major depression and related conditions, has nearly doubled since the early 1980s. Moreover, in 1997, fully 40 percent of all psychotherapy patients were diagnosed with some mood disorder, compared to 20 percent in 1987. Three times more people were treated for depression in primary medical care in 1997 than ten years earlier. The consumption of antidepressant medications both in absolute numbers and in percentage of diagnosed patients receiving medication has also dramatically expanded; persons treated for depression were four and one-half times more likely to receive psychotropic medication in 1997 than in 1987. At present, three of the seven highest-selling prescription drugs (Prozac, Paxil, and Zoloft) of any sort are antidepressants. W. H. Auden's "Age of Anxiety" appears to have been succeeded by our own "Age of Depressive Disorder."
No plausible theory of depressive disorder, whether genetic, psychological, or social, can explain why rates of depression would have increased so much in such a short period of time. Instead, the explanation appears to lie in changes in the ways that physicians, mental-health professionals, and people themselves characterize and diagnose their mental states. There are, and always have been, true depressive disorders, in which the response to loss goes awry and takes on a debilitating life of its own. But in the past, such disorders were distinguished from normal sadness that arises in response to life's vicissitudes. That traditional, common-sense distinction has broken down in contemporary psychiatry, resulting in the conflation of depressive disorders with normal sadness. The sources and social implications of this breakdown are as yet largely unappreciated.
A history of depression
How did this transformation of sadness into depression occur? To grasp the answer, the current approach to diagnosis must be placed in the historical context of 2,500 years of contrary medical and psychiatric practice. To follow this story to the present, one must also confront the esoterica of modern psychiatric classification as represented by successive editions of the American Psychiatric Association's Diagnostic and Statistical Manual of Mental Disorder (DSM-I through DSM-IV). Often called the "Bible of Psychiatry," the four incarnations of the DSM have offered official diagnostic definitions for all mental disorders.
As long as written records have been kept, Western cultures have recognized that depression can be a mental disorder. Hippocrates, writing in the fifth century B.C., provided the first known definition of the phenomenon as a distinct disorder: "If fear or distress last for a long time it is melancholia." While theories of depressive disorder have changed, the symptoms that indicate the disorder have not. For Hippocrates, its symptoms could include prolonged despondency, blue moods, detachment, nameless fears, irritability, restlessness, sleeplessness, aversion to food, and suicidal impulses, much like today's criteria. But Hippocrates's definition indicates not that such symptoms alone indicate disorder but that such symptoms over an abnormally long duration do.
A century later, Aristotle elaborated the distinction between normal and pathological mood states. He separated melancholic states that arose from the conditions of everyday life from disorders that involved "groundless" despondency of lengthy duration and thus stemmed from internal factors:
We are often in the condition of feeling grief without being able to ascribe any cause to it; such feelings occur to a slight degree in everyone, but those who are thoroughly possessed by them acquire them as a permanent part of their nature.
As in Aristotle's passage, the key distinction in ancient definitions of melancholia was between states of sadness "without cause" and those that had similar symptoms arising from actual losses; only the former were mental disorders. "Without cause" does not mean uncaused, for throughout history depression has been attributed to postulated physical or psychological causes such as excessive black bile, disturbances in the circulation of blood, or depletion of energy. Rather, "without cause" means that the symptoms of depression were not associated with the sorts of environmental events that would appropriately lead to sadness, such as bereavement, rejection in love, economic failure, and the like. Conversely, ancient Greek and Roman physicians would not consider symptoms of depression that occur "with cause" as signs of a mental disorder. Such normal reactions express, in the words of the second- century Roman physician Aretaeus, "mere anger and grief, and sad dejection of mind." The symptoms could be identical in the two conditions; the distinction lay in the relation of the symptoms to the context in which they appeared. Symptoms that arose in contexts that could be expected to produce them, and that abated in a reasonable period of time after the triggering events ended, indicated normal functioning. Comparable symptoms that arose without appropriate triggering events, or had greater duration or intensity than was appropriate to the triggering events, potentially indicated disorder.
The same distinction can be found in Robert Burton's classic work The Anatomy of Melancholy, published in 1621. Burton defined melancholic disorder as "a kind of dotage without a fever, having for his ordinary companions fear and sadness, without any apparent occasion" (emphasis added). Burton considered such states "contrary to nature," and thus disordered. He considered the propensity to self-limiting melancholic feelings in response to loss and disappointment to be a normal part of human nature. He describes
that transitory melancholy which goes and comes upon every small occasion of sorrow, need, sickness, trouble, fear, grief, passion, or perturbation of the mind, any manner of care, discontent, or thought, which causeth anguish, dullness, heaviness, and vexation of spirit.... And from these melancholy dispositions, no man living is free, no Stoic, none so wise, none so happy ... [none] so well composed, but more or less, some time or other, he feels the smart of it. Melancholy, in this sense is the character of mortality.
The traditional distinction between abnormal depression "without cause" and normal depression "with cause" persisted into the twentieth century. In psychoanalytic discussions of depression, the former was called "melancholy" and the latter "mourning." In Freud's central article on depression, "Mourning and Melancholia," he asserted that symptoms associated with mourning, although intense, were a normal and self-healing condition that did not require medical treatment:
Although grief involves grave departures from the normal attitude to life, it never occurs to us to regard it as a morbid condition and hand the mourner over to medical treatment. We rest assured that after a lapse of time it will be overcome, and we look upon any interference with it as inadvisable or even harmful.
Freud added an epicycle to the tradition. He observed that, although the symptoms of mourning and melancholia both include profound dejection, loss of interest in the outside world, inability to feel pleasure, and inhibition of activity, melancholia also frequently includes an extreme and inexplicable ("without cause") decline in self-esteem bey\ond the normal self-recriminations after a loss.
American psychiatry developed successive versions of its own classification system starting early in the twentieth century. Until quite recently, psychiatric definitions of depressive disorder continued to reflect the historical distinction between depression "with" and "without" cause. The first standardized classification system in the United States, the 1918 Statistical Manual for the Use of Hospitals for Mental Diseases, defines "reactive depression" much as Hippocrates did, to contrast it with normal sadness that arises in response to a great variety of losses:
Here are to be classified those cases which show depression in reaction to obvious external causes which might naturally produce sadness, such as bereavement, sickness and financial and other worries. The reaction, of a more marked degree and of longer duration than normal sadness, may be looked upon as pathological.
The DSM-I (1952) and DSM-II (1968) that succeeded the Statistical Manual emphasized psychoanalytic concepts. The DSM-Il defined "depressive neurosis" as follows: "This disorder is manifested by 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." In defining depressive disorders as "excessive" reactions, the DSM-II recognized that they are either disproportionate to actual loss or they involve no loss at all and result from internal causes (assumed to consist of internal conflict). Here, as in the 1918 hospital manual's definition, normal triggers beyond loss of a loved one were recognized, such as loss of a cherished possession.
For two and one-half millennia, psychiatry has held that normal human nature includes a propensity toward potentially intense sadness after certain kinds of losses. Disorder can be judged to exist, it has widely been agreed, only when explanations in terms of triggering events fail to establish a normal cause for the intensity or duration of symptoms.
The revolution in psychiatric diagnosis
The tradition of systematically distinguishing intense normal sadness from depressive disorder was largely abandoned with the publication of the DSM-IH in 1980. This neglect has continued through various revisions to the current DSM-IV (1994). This shift occurred as part of a broader revolutionary transformation of psychiatric classification guided by Robert Spitzer, the DSM-IH's editorin-chief and head of the DSM-III Task Force.
The great weakness of the DSM-I and DSM-II was that their vague definitions were not capable of generating standardized methods of classifying and studying mental disorders. Because psychiatrists had to use considerable personal judgment in fitting a patient to a diagnosis, psychiatric diagnosis was notoriously unreliable. Given the same information about the same patient, different doctors were likely to arrive at different diagnoses.
In addition, the psychoanalytic dominance in psychiatry was waning, and there was a sharp reaction against the psychoanalytic assumptions of the DSM-II. Psychiatric practice and research had split into rival theoretical fiefdoms ranging from behavioral and cognitive theories to the increasingly influential biological approach, each of which had its own idiosyncratic theory-based definitions of mental disorders. This made communication among practitioners difficult. It meant that research within different theoretical schools was based on incommensurable definitions, was not cumulative, and could not be usefully compared. It also put many practitioners and researchers in the uncomfortable position of being unable to accept psychiatry's standard definitions of disorders because of the theoretical assumptions built into them.
On top of all of this, psychiatric diagnosis was under attack from a variety of sources. Behaviorists claimed that all behavior is the result of normal learning processes and that no mental disorders in the medical sense really exist. The "anti-psychiatry" movement, inspired by, among others, the psychiatrist Thomas Szasz and the sociologist Thomas Scheff, portrayed psychiatric diagnosis as a matter of labeling socially undesirable, but not truly medically disordered, behavior with medical terminology to justify using psychiatric intervention to control the behavior. Between psychiatry's theoretical fragmentation, its diagnostic unreliability, and the anti-psychiatry critique, psychiatry's claim to scientific status and even its legitimacy as a medical field seemed in jeopardy.
The DSM-III inaugurated basic changes in psychiatric diagnosis to address these challenges. Explicit diagnostic criteria based on lists of observable symptoms were provided as definitions for each of the disorders, allowing improved reliability. Symptom lists provided criteria on which virtually all clinicians of opposed theoretical schools could agree. References to postulated psychodynamic causes of a disorder (for example, internal conflict, defense against anxiety) were consequently purged. The DSM-IH's diagnostic criteria were theory-neutral in the sense that they did not presuppose any particular theory of the cause of psychopathology, psychoanalytic or otherwise; the criteria were descriptive rather than etiological. Competing theories could henceforth try to prove their claims without begging the question by defining disorders in their own terms, and research became cumulative and comparable across theories. The psychoanalytic dominance of psychiatric diagnosis, unwarranted by the state of the evidence, was effectively ended. Rival approaches were allowed to compete on a flat conceptual playing field.
However, a major drawback of symptom-based criteria was that they eliminated the consideration of the context in which symptoms arose. One reason for the abandonment of context was the quest for reliability. Most depressions occur after some triggering event, so requiring that reactions be disproportionate to context would mean that clinicians would be frequently judging proportionality, substantially reducing reliability. (Never mind that increased reliability at the cost of validity and diagnostic logic is of no real value.) A second reason for leaving out context is that the DSM- III was supposed to be theory-neutral, which was interpreted to mean that it had to be neutral as to how the disorder was caused, although one could argue that distinguishing normal responses to events from disorders is not really a theory-laden distinction. Moreover, there was an impression that psychotropic medication worked on all depressions irrespective of the relation to triggering events, so that the "with versus without cause" distinction was considered irrelevant to treatment decisions. (However, even if medication sometimes works with normal reactions, the normality- versus-disorder distinction could have prognostic implications and thus could be an important consideration in whether or how aggressively to treat a condition.) Finally, there was a fear of misdiagnosing the truly disordered as normal, especially given that depressed patients are subject to suicide risk.
Granting some validity to these rationales, there does not seem to have been an organized attempt to balance potential harms from underdiagnosis and overdiagnosis, as is done with many medical tests. The problem remained that, as the previous 2,500 years of psychiatric diagnostic practice had attested, without an exploration of context and meaning, one simply cannot tell whether someone is likely suffering from intense normal sadness or a depressive disorder.
From sadness to depression
This brings us to how, in the case of depression, the DSM-III criteria went conceptually awry. (For brevity, we focus on "major depressive disorder" and ignore related diagnoses.) The DSM-III replaced DS1M-II's vague criteria with specific symptomatic criteria. These criteria have remained more or less the same through the current DSMIV, on which we focus.
Nearly all recent studies of depressive disorder are based on the DSM-IVs definition. Different specialists use this definition for different purposes: clinicians for diagnosing persons who have sought their help; epidemiologists for determining the number of people in the general population who are depressed; researchers for finding the causes of depression and for evaluating the effectiveness of treatments; economists for estimating the costs associated with depression; pharmaceutical companies for marketing their products; reimbursers for establishing medical necessity and determining quality-of-care standards; and mental-health advocates for quantifying how widespread this condition is. The DSM-IV definition is used not only in the United States but has also come to be the standard definition worldwide. The logic of DSM-IVs definition of depressive disorder is key to understanding depression as a social fact.
The definition requires that five symptoms out of the following nine must be present during a two-week period (the five must include either depressed mood or loss of interest and pleasure): (1) depressed mood; (2) diminished interest or pleasure in activities; (3) weight gain or loss or change in appetite; (4) insomnia or hypersomnia (excessive sleep); (5) psychomotor agitation or retardation (slowing down); (6) fatigue or loss of energy; (7) feelings of worthlessness or excessive or inappropriate guilt; (8) diminished ability to think or concentrate or indecisiveness; and (9) recurrent thoughts of death or suicidal ideation or suicide attempt. In addition, to eliminate rare cases where symptoms are so mild as to be insignificant, symptoms must cause clinically significant distress or impairment in social, occupational, or other important areas of functioning.
Any person satisfying these criteria is today considered to have a depressive disorder, with three exceptions to be considered shortly\. Yet, symptoms such as depressed mood, loss of interest in usual activities, insomnia, lessened appetite, and inability to concentrate might naturally occur for two weeks if a major loss or humiliation is experienced. Such reactions, even when quite intense due to the severity of the trigger, are surely part of normal human experience.
Two exceptions to the DSM-IV depression criteria simply shift the diagnosis to other categories of mood disorder-bipolar (manic- depressive) illness, or depression caused by a medication or a general medical condition. The third exception is the only acknowledgment of the existence of normal sadness-bereavement after the death of a loved one. Depression is diagnosed only when
The symptoms are not better accounted for by Bereavement, i.e., after the loss of a loved one, the symptoms persist for longer than 2 months or are characterized by marked functional impairment, morbid preoccupation with worthlessness, suicidal ideation, psychotic symptoms, or psychomotor retardation.
The bereavement exclusion thus allows that reactions of intense grief are not truly depressive disorders, unless the reaction lasts more than two months or it includes one out of a list of five especially serious symptoms.
The bereavement exclusion is notable for the limited range of normal grief reactions it allows. Surely grief reactions lasting more than two months are not necessarily disordered. And surely the five specified symptoms do not each necessarily indicate disorder: A normally bereaved individual may for two weeks experience "marked functional impairment" (such as not feeling up to usual work or social activities), may think that a lost partner was his or her "better half" and feel worthless or inadequate to life's tasks without him or her, or may entertain the notion that he or she might be better off "joining" the deceased partner, all without necessarily indicating disorder.
The most important thing, however, about the bereavement exclusion is that it offers exclusions only for reactions to the death of a loved one. Yet normal sadness reactions that are symptomatically similar to depressive disorders are not limited to bereavement. They encompass reactions to a wide range of negative events such as betrayal by romantic partners, being passed over for an anticipated promotion, failure to achieve long-anticipated goals, or discovering a life-threatening illness in oneself or a loved one. Of course, to qualify as normal responses that satisfy DSM-IV criteria, such reactions must involve a serious loss, and coping mechanisms must enable the individual to adapt to the new circumstances and get over the symptoms within a reasonable time after the precipitating event ends. If such a trajectory of adaptation does not occur, then one might infer that the original reaction had somehow caused an internal psychological dysfunction that now maintains a reaction that is no longer normal. But there are many intense reactions to loss that, just like normal bereavement responses, might satisfy DSM-IV symptom criteria but are not disorders.
The basic flaw, then, is that the DSM-IV fails to exclude from the disorder category sadness reactions to events other than death of a loved one that are intense enough to meet the DSM-IVs criteria but are still normal reactions. The age of depressive disorder in which we find ourselves today is partly an artifact of a logical error.
Community studies of depression
In a clinical assessment, the clinician can in principle override the DSM-IV diagnostic criteria for depression and judge that an individual satisfying the criteria nonetheless is having a normal reaction. But no such back-up validation procedure exists when questionnaires about symptoms are administered to community members to measure how much disorder exists in the general population. Thus invalid criteria have particular scope for mischief in such studies.
A vast extension of the application of the symptombased concept of mental disorder occurred when DSM criteria became the basis for large epidemiological studies attempting to measure the extent of mental disorder among people in the community who are not undergoing mentalhealth treatment. Because from DSM-IH onward diagnoses were based entirely upon symptoms, epidemiologists could easily construct lists of questions that could be used to determine whether a respondent met DSM criteria for disorder, including depressive disorders. Lay interviewers could administer such questionnaires, allowing researchers to obtain psychiatric diagnoses comparable to those a psychiatrist would obtain, without the prohibitive expense of psychiatric interviewers. The results would presumably provide good estimates of how much untreated mental disorder existed in the community. These estimates were intended to guide policy makers in allocating resources by establishing how much unmet need existed for psychiatric services. The decision to use objective measures of symptoms in community studies largely stems from considerations of practicality and cost, and an uncritical acceptance of the DSM' s symptom-based criteria, not from independent tests showing these methods are accurate in identifying disorder.
Findings from two major national studies, the Epidemiological Catchment Area Study (ECA) conducted in the early 1980s and the National Co-Morbidity Study (NCS) conducted in the early 1990s are the basis for the estimates regarding the prevalence of mental disorder that are now widely cited in the scientific, policy, and popular literatures. The NCS estimates that about 5 percent of subjects had a current (30-day) episode of major depression, about 10 percent had this diagnosis in the past year, about 17 percent had an episode over their lifetime, and about 24 percent reported enough symptoms for a lifetime diagnosis of either major depression or a related disorder, dysthymia.
Are the many cases of putative major depression uncovered in community studies equivalent to treated clinical cases? The odds are against it. These studies follow the DSM in ignoring the context of symptoms, thus confounding ordinary sadness with genuine disorder without recourse to clinical judgment to correct the error. For example, in the ECA study the most common symptoms among those reporting symptoms are "trouble falling asleep, staying asleep, or waking up early" (33.7 percent), being "tired out all the time" (22.8 percent), and "thought a lot about death" (22.6 percent). College students during exam periods (particularly those studying existential philosophy), people who must work overtime, or those worrying about an important upcoming event could all experience these symptoms naturally. Thus conditions that neither respondents nor clinicians would consider reasons for entering treatment can nonetheless indicate disorder in community surveys. Moreover, the symptoms are required to last for only two weeks, allowing many transient and self-correcting symptoms to become the basis for diagnosis. While most studies do not report the context of symptoms, in one study of adolescent depression that did, the single greatest trigger for presumed depression was the breakup of a romantic relationship, suggesting that a potentially large proportion of "disorders" were actually misclassifications of normal responses.
The sorts of experiences that produce normal sadness responses- breakups of romantic relationships and marriages, job losses, disappointed career goals, and the likeare certainly not uncommon in community populations. A survey respondent might recall symptoms such as sad mood, insomnia, overeating, tiredness, and lessened pleasure in usual activities, that lasted for at least two weeks after such an event. Although these symptoms might have dissipated as soon as a new relationship developed, another job was found, or the goal was finally achieved, this individual might satisfy DSM criteria and join the 20 million people who are said to suffer from the presumed disorder of depression each year. In all likelihood, community studies did not so much uncover high rates of depressive disorders as they demonstrated that the natural results of stressful social experiences could meet DSM's symptom criteria for depression. Ironically, reporting such overwhelmingly high rates backfired. It ended up encouraging skepticism and dismissal by policy makers. It also promoted fears of the financial consequences of insurance parity.
The constituencies for depression
Many factors, some legitimate and some less defensible, drove psychiatry and epidemiology to use symptombased, context-free criteria for disorder. Once this conceptually flawed diagnostic system was generally accepted, it became inadvertently advantageous to various constituencies in the pursuit of their agendas. The medicalization of sadness achieved a social impact far beyond the narrow domain of clinical psychiatric diagnosis. That this approach works against the basic psychiatric goal of validly distinguishing disorder from nondisorder has simply been ignored.
The National Institute of Mental Health (NIMH), the major American sponsor of research on mental illness, became a key beneficiary of the new approach. The conflation of mental disorder and ordinary sadness legitimized a broad interpretation of the NIMH's mandated domain and allowed it to argue persuasively for increased funding on the basis that mental disorder is rampant in the population. Allowing some painful but normal psychological reactions to be considered disorders also effectively depoliticized the NIMH's previous concern with the problematic psychological consequences of social problems such as poverty, racism, and discrimination.
This transformation has some benefits. American society is now more compassionate about psychological disorder than previously. Nor is it necessarily unhelpful for nondisordered but disadvantaged people to receive supporti\ve mental-health treatment addressing some of the psychological challenges they face. However, doing so via unreasonably broad definitions of psychological disorder has the potential cost of stigmatizing the disadvantaged as mentally ill, replacing social policy with unwarranted medical treatment, and creating a one-dimensional public discourse that can undermine our capacity for making moral and political distinctions.
Mental-health researchers have also adapted to current criteria and have much to lose if traditional measures of depression should return. Symptom-based criteria are relatively easy to use. They reduce the cost and complexity of research studies, and allow for higher research productivity. Enhanced reliability confers the appearance of a more scientific approach, although in fact considerations of validity should trump reliability. Moreover, the Z)SM's criteria are used in virtually all of the thousands of studies done in recent years on depression, and many researchers' careers are built around these studies. Consequently, any major reconceptualization of diagnostic criteria would throw all that into doubt. Adequately distinguishing normal sadness from depressive disorder could also possibly narrow opportunities for research funding, especially if the NIMH followed suit by focusing its efforts on true disorder. Nevertheless, as researchers certainly appreciate, reaching the goal of understanding the etiology and appropriate treatment of depressive disorder ultimately depends on using a valid definition of disorder as the basis for sample selection.
For mental-health clinicians, symptom-based measures of depression justify reimbursement from third-party insurers for the treatment of a broader range of patients than might otherwise qualify, because insurers generally will pay to treat disorders but not mere problems of living. Individual clinicians are faced every day with patients seeking help who are suffering from conditions that appear to be intense normal sadness, but that satisfy the DSM's criteria for disorder. Many private-practice clinicians will readily admit that a sizable proportion of their "depression" caseload consists of individuals who are psychiatrically normal but experiencing stressful life events. To obtain reimbursement for the treatment of such patients, the clinician must classify the individual within a DSM category of disorder, and depression is one of the more commonly used and easier ones to justify given the ubiquity of its symptoms. The result is a strange case of two "wrongs" seemingly making a "right": The DSM provides flawed criteria that do not adequately distinguish disorder from nondisorder; the clinician, knowingly or unknowingly, incorrectly classifies a normal individual as disordered (Why should the clinician question a diagnosis officially sanctioned by the DSMl); and the patient receives desired treatment for which the therapist is reimbursed.
Such conceptually questionable diagnosis and consequent reimbursement are easily rationalized when the alternative seems insensitive and when rigid reimbursers may refuse desired treatment to people who are suffering. There are complex policy issues lurking here that would benefit from public discussion, however. In other areas of medicine, treatment of the nondisordered is openly debated: Should growth hormones be given to normal but short children? Should post-menopausal women be reimbursed for fertility treatment? When does the use of Viagra represent a medical necessity? In psychiatry, faulty criteria camouflage such treatment and allow the issue to be avoided.
One cost of this avoidance is that some critics use the expansiveness of the DSM criteria to argue against reimbursement parity for mental-health care. When the issue of parity arises, the first objection sure to be heard is that such parity would break the health-care bank because it would allow every instance of normal unhappiness to qualify for treatment. Such objections are buttressed by the ridicule frequently heaped on DSM criteria in the popular media. A more honest discussion of normal versus abnormal conditions and their appropriate rights to reimbursement might help to address some of these objections.
Family advocacy organizations, such as the National Alliance for the Mentally 111 (NAMI), which became an influential political force during the 1980s, have at the top of their political agenda the achievement of insurance reimbursement parity for mental disorder. Their major argument is that certain mental disorders as currently defined, including depression, are biological disorders, just like physical disorders, and deserve to be treated equally with respect to reimbursement. Admitting that current DSM criteria fail to distinguish true depressive disorders from normal sadness reactions would certainly muddy this argument, and appears to go against these groups' agendas. In the long run, however, the case for better treatment might be strengthened if such a distinction was carefully made and the conceptual flaws in current criteria eliminated, so that the criteria did indeed pick out only plausible cases of disorder. The truly disordered would then become the more exclusive focus of treatment and resource allocation.
Pharmaceutical companies are perhaps the most visible beneficiaries of symptom-based diagnoses. Although there is little evidence that these companies influenced the development of the DSM- III, its symptom-based approach created a broader market of disorder for their products to treat. In addition, they are now major sponsors of the activities and research of psychiatrists and advocacy groups. While some concern has arisen about problems with side effects, suicide risk, and dependency, in general, the transformation of sadness into depression has been enormously profitable for these companies. Ubiquitous drug ads now warn the public that common symptoms such as sadness, anxiety, sleep problems, or overeating may be signs of mental disorder. Like clinicians, the drug companies can legitimately explain that they are only using the criteria officially sanctioned by the psychiatric profession, even as they further confuse the public about the boundary between normality and disorder.
Perhaps the most successful effort of pharmaceutical companies has been to increase the diagnosis of depressive disorder and prescribing of medication by general physicians. The inadequate supply of psychiatrists and other trained mental-health professionals to deal with the enormous numbers of presumably depressed people uncovered by epidemiological studies has served as the rationale for an urgent push to have non-mental-health personnel diagnose and treat depression. To aid busy physicians who have no time to perform psychiatric assessments, pharmaceutical companies have supported the development of easyto-use symptom checklists based on DSM criteria. These lists can be routinely administered to patients, whatever the reason for their visit to the doctor. By catching some cases of disorder, such screening can be beneficial. But, obviously, even a patient who is psychiatrically normal may have recently felt intense sadness, for example, in response to a life crisis. The screening instruments physicians use to detect depression follow the DSM in intentionally ignoring all such contextual considerations in reaching a diagnosis.
The result is both unsurprising and shocking: In some studies, levels of mood disorders in general medical practices approach or exceed 25 percent. This includes subclinical diagnoses with fewer than the DSM-required number of symptoms, which are included as potentially treatable depressions in the most widely used physician scale. Because physicians are not trained in psychotherapeutic intervention, medication is the most likely treatment.
Whether medicating normally sad individuals that meet DSM criteria is usually helpful or harmful is unknown and is a topic that has received virtually no study. The reason is simple: The DSM criteria used by researchers do not allow the distinction to be made. Freud's speculation that medical "interference" in normal sadness can be "inadvisable or even harmful" may have some truth; research by the psychologist James Coyne suggests that medical interventions designed to prevent depression may have little effect or can sometimes even worsen depression by disrupting normal coping processes and heightening patient focus on negative experiences. Whatever the answer, surely we ought to be studying this question. That would require making important distinctions between normal and disordered sadness responses. And that would in turn require honestly confronting the conceptual invalidities that now afflict the diagnosis of depressive disorder.
Psychiatry and society
For thousands of years, symptoms of sadness that were "with cause" were separated from those that were "without cause." Only the latter were viewed as mental disorders. The framers of the DSM-III did not explicitly try to expand the domain of disorder and in some ways its definition of depression is more stringent than in previous manuals. The DSM-IH's decontextualized symptom-based criteria for depressive disorder had the unintended consequence, however, of classifying some instances of intense normal sadness as disordered. This greatly expanded the domain of pathological depression, especially when applied to untreated community samples. Reported prevalence rates of depressive disorders have correspondingly increased, triggering attempts to detect and treat depression in new ways, such as in the offices of general medical practitioners.
The medicalization of sadness in psychiatric diagnoses has had broader cultural consequences. When people are constantly exposed to pharmaceutical advertisements, public service messages, and news stories that conflate depression with normal sadness, and are assured th\at no stigma attaches to such diagnoses, they may naturally tend to monitor themselves for such symptoms, reframe their own experiences of sadness as signs of a mental disorder, and seek professional help for their problems. The result has been skyrocketing rates of treatment for depression that give the possibly mistaken impression that actual rates of depression have increased.
While the medicalization of sadness undoubtedly has a number of beneficial aspects both in relieving suffering and occasionally in preventing disorder, it also has many costs that should not be accepted uncritically. Most sadness diminishes with the passage of time, even in the absence of therapeutic interventions. To the extent that sadness results from ordinary human misery, changing or accepting the situations that led to the sadness might be as effective, or more effective, than medication or psychotherapy. Encouraging general physicians and mentalhealth professionals to diagnose and treat mild depression can thus be an inefficient use of their time. The dangers of overintervention with normal conditions are real and need to be balanced against those of undertreatment. One element in deliberatively reaching such a balance is the formulation of a conceptually valid approach to the distinction between disorder and nondisorder.
Depression can be a serious and sometimes life-threatening condition, and depressive disorders must be taken seriously. In addition, one might argue that in many cases there is no good reason why people suffering from normal sadness should be prevented from taking antidepressant medication, if it proves effective and safe to do so. Freud was surely incorrect, at least in relation to our contemporary medical practices, to assert that "it never occurs to us" to treat such painful normal conditions medically. Whether public policy ought to encourage such treatment broadly, and whether such treatment if widely adopted might somehow reduce the depth of human experience of loss and thus indirectly of caring, are further questions in need of exploration. Offering skilled help to those who are suffering and capable of benefiting from such help, whether disordered or nondisordered, is a moral imperative. But it is not clear that relabeling normal forms of misery as depressive disorder is beneficial or yields wise public policy.
Copyright National Affairs, Inc. Winter 2005