Major depressive disorder (MDD) has become psychiatry’s signature diagnosis. Depression is diagnosed in about 40% of patients who see a psychiatrist. This percentage is double that of just 20 years ago, and it is far higher than that of any other diagnosis. The World Health Organization (WHO) estimates that worldwide depression is the leading cause of disability for people in midlife and for women of all ages.

Consumption of antidepressants has soared since 1990. Roughly 10% of women and 4% of men in the United States take antidepressant medication at any time. By 2000, antidepressants were the best-selling prescription drugs of any type. Yet epidemiological studies suggest that there are still vast numbers of untreated depressed individuals. Consequently, primary care practitioners have been recruited as the first line of defense, and many now routinely screen patients for depression. To catch the problem early, a presidential commission has recommended that every adolescent in the country should be screened for depression by the time he or she reaches age 18. Screening is proceeding in some schools.

What accounts for this seeming epidemic of depression? Although depression has been part of the psychiatric canon since the earliest writing of the ancient Greeks, depression was a relatively insignificant diagnosis just 50 years ago. In our recent book, The Loss of Sadness: How Psychiatry Transformed Normal Misery Into Depressive Disorder (Oxford University Press), we argue that the recent pandemic of seeming depressive disorder is the result of changes in the psychiatric diagnostic system presented in DSM-III in 1980 and that persist to the present.

In many respects, DSM-III (and subsequent versions) has been one of psychiatry’s greatest accomplishments. It was the first to use observable symptoms, rather than unobservable (and undemonstrated) etiological processes, to define the various types of mental disorders. Its clear definitions of discrete categories of disorder enhanced diagnostic reliability, thus putting to rest antipsychiatric arguments about the spuriousness of psychiatric diagnosis. These definitions allowed psychiatrists to communicate in a common theory-neutral language, irrespective of theoretical perspectives, that improved the cumulativeness of research.

Yet, these undoubted achievements also entailed some important disadvantages. These drawbacks have become especially apparent in the definition of MDD, and have had substantial social consequences.

A diagnosis of MDD is warranted, according to DSM, when a patient has at least 5 of 9 specified symptoms for at least 2 weeks, and the 5 symptoms include either depressed mood or an inability to derive pleasure from life. The sole exception is that bereaved patients are not considered to have a disorder if they otherwise meet the criteria, as long as their symptoms are not unusually severe and last no longer than 2 months. The reason for the bereavement exclusion seems obvious: people who respond to the loss of an intimate with intense sadness, sleep and appetite difficulties, a loss of concentration on usual roles, and the like, do not have a mental disorder. Rather, they are responding normally to a situation of intense loss. The distinction between sadness that is a normal result of painful losses and depressive disorder is a fundamental one that has been explicitly recognized throughout the 2500-year history of psychiatric medicine.

Yet, the bereavement exclusion raises the question of whether people with enough symptoms to meet the MDD criteria—after, for example, the unexpected loss of a valued job, the collapse of a marriage, the failure to achieve a highly valued goal, or the diagnosis of a life-threatening illness in oneself or a loved one—are similarly reacting normally to situations of intense loss. For thousands of years, until DSM-III, physicians understood that these kinds of situational contexts were an important consideration in determining whether someone was experiencing normal—although intensely distressing—sadness or a depressive disorder in which something has gone wrong with mood processes and the sadness symptoms are no longer linked to the situation or likely to remit over time. Unlike many other diagnoses in DSM, which contain qualifiers that require symptoms to be “excessive” or “unreasonable,” no such qualifiers exist for MDD. Aside from the bereavement exclusion, the diagnostic criteria do not take into account the context in which symptoms arise.