The Bereavement Exclusion and DSM-5: An Update and Commentary

Abstract

The removal of the bereavement exclusion in the diagnosis of major depression was perhaps the most controversial change from DSM-IV to DSM-5. Critics have argued that removal of the bereavement exclusion will “medicalize” ordinary grief and encourage over-prescription of antidepressants. Supporters of the DSM-5’s decision argue that there is no clinical or scientific basis for “excluding” patients from a diagnosis of major depression simply because the condition occurs shortly after the death of a loved one (bereavement). Though bereavement-related grief and major depression share some features, they are distinct and distinguishable conditions. Bereavement does not “immunize” the patient against a major depressive episode, and is in fact a common precipitant of clinical depression. Recognizing major depression in the context of recent bereavement takes careful clinical judgment, and by no means implies that antidepressant treatment is warranted. But given the serious risks of unrecognized major depression—including suicide— eliminating the bereavement exclusion from DSM-5 was, on balance, a reasonable decision.

INTRODUCTION

Controversy continues to surround the removal of the so-called “bereavement exclusion,” (BE) from the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (DSM-5). Without question, this was one of the most contentious decisions the DSM-5 work groups made—and, by some lights, the most controversial decision by the American Psychiatric Association (APA) since homosexuality was removed from the list of psychiatric disorders in 1973. While I was not directly involved with the DSM-5 mood disorders work group, my colleagues and I were participants in a sometimes rancorous debate,^1,2 often fueled by sensational or misleading reports in the lay media; for example, lay media headlines included claims such as, “Psychiatrists want to make normal grief a mental disorder!” and “DSM-5 medicalizes mourning.”

In truth, the DSM-5 criteria for major depressive disorder (MDD) merely say that the subset of persons who meet the full symptom-duration-severity criteria for major depression within the first few weeks after bereavement (i.e., the death of a loved one) will no longer be excluded from the set of all persons with major depression—as many would have been—under DSM-IVs exclusion guidelines. Put another way: DSM-5 recognizes that bereavement does not immunize the patient against major depression, and often precipitates it. Indeed, grief and depression—despite some overlapping symptoms, like sadness, sleep disturbance and decreased appetite—are distinct constructs, and one does not preclude the other.²

The bereavement exclusion was eliminated from the DSM-5 for two main reasons: 1) there have never been any adequately controlled, clinical studies showing that major depressive syndromes following bereavement differ in nature, course, or outcome from depression of equal severity in any other context—or from MDD appearing “out of the blue;”² and 2) major depression is a potentially lethal disorder, with an overall suicide rate of about four percent.³ Disqualifying a patient from a diagnosis of major depression simply because the clinical picture emerges after the death of a loved one risks closing the door on potentially life-saving interventions. The “exclusion” principle also fails to recognize that MDD is often a highly over-determined process, involving multiple, interacting causes (e.g., someone who develops a major depressive syndrome a few weeks after a loved one’s death may also be depressed owing to concomitant hypothyroidism, pancreatic cancer, marital problems, or a recent setback in business).⁴ In such a complex associational context, which factor or factors should be judged “causal”? And how would the old (DSM-IV) bereavement exclusion rules apply?

THE OVERRIDE OPTION

It is true that the DSM-IV criteria provided a way to “override” the bereavement exclusion, (e.g., if the depressed, bereaved patient were psychotic, suicidal, psychomotorically slowed, preoccupied with feelings of worthlessness, or functioning very poorly in daily life). Unfortunately, these override features did not address those bereaved patients whose depressive symptoms were indeed severe, but who did not “qualify” for the specific override criteria (e.g., bereaved persons with profoundly impaired concentration, significant weight loss, or severe insomnia). Under the DSM-IV “rules,” these seriously depressed individuals probably would not have received a diagnosis of MDD and appropriate treatment.

It is sometimes argued that bereaved, suicidal patients would not have been excluded from an MDD diagnosis using the DSM-IV override “rules.” In theory, that was true. But not every depressed patient openly acknowledges suicidal ideation or intentions to a clinician—some fear that doing so will result in involuntary hospitalization. Moreover, the risk of suicide in MDD is not conferred solely by the presence of suicidal ideation; rather, both overall severity of depression and hopelessness also elevate risk of eventual suicide.⁵ Yet neither factor was specifically included in the DSM-IVs list of features that allowed one to override the bereavement exclusion.

The preponderance of data suggest that, compared to MDD emerging in other contexts, such as job loss or recent divorce—or to MDD arising “out of the blue”— bereavement-related major depression (BRMD) differs little in symptom picture, course, outcome, or response to treatment. Thus, there is no strong rationale for “privileging” BRMD for exclusion.

In fairness, there have been some recent epidemiological studies that appear to show lower risk of recurrence for some types of BRMD, compared with “standard” (non-bereavement) MDD.^6,7,8 These data— derived from ECA (Epidemiologic Catchment Area) and NESARC (National Epidemiologic Survey on Alcohol and Related Conditions) surveys of community residents— were obtained by lay interviewers, and, as with all surveys, are subject to recall bias on the part of participants. More important, subjects in the two groups (BRMD vs. “standard” MDD) were not matched for severity, duration, or degree of impairment during the index depressive episodes. This makes it impossible to know whether bereavement per se or some other risk factor for recurrence (such as melancholic features) accounts for the group differences. Furthermore, in the Gilman et al study,⁸ even those with excluded bereavement-related depression had “clinically significant depressive episode[s]” (i.e., episodes that were not clearly “normal grief).” Indeed, the propensity to recur is merely one index of a depressive condition’s clinical significance—and a reduced tendency to recur does not necessarily point to “non-disordered sadness,” as some critics of psychiatric nosology have claimed.¹¹

While definitive controlled studies of BRMD vs. “standard” MDD have not been carried out, one recent study used rigorous methods to tease out the role of bereavement. Hamdan et al⁹ studied a cohort of parentally bereaved youth and non-bereaved controls over approximately five years. Three groups were assessed for symptoms, severity, duration, and risk for recurrence: 1) bereavement-related depression (BRD, n=42), with onset of depressive episode within the first two months after parental death; 2) later bereavement depression (LBD, n=30), with onset at least 12 months after parental loss; and 3) a non-bereaved control group with depression (CD, n=30).

The study found that bereavement-related depression was similar to LBD and CD, with respect to number of symptoms, severity, functional impairment, duration, and risk for recurrence. While the study population was small and limited to youths, the authors concluded that their findings “...support the removal of the bereavement exclusion.”⁹

It is important to understand that the DSM-5 criteria merely allow the diagnosis of MDD when the recently bereaved person meets all required symptom, severity, duration, and impairment criteria for MDD. Nothing in the manual compels a diagnosis of MDD shortly after bereavement. The DSM-5 clearly states that, “...periods of sadness are inherent aspects of the human experience. These periods should not be diagnosed as a major depressive episode unless criteria are met for severity...duration...and clinically significant distress or impairment.” (DSM-5, American Psychiatric Association, 2013 p. 168)

Indeed, if the clinician’s best judgment—and, to be sure, sound judgment is needed!—points to normal, bereavement-related grief, the “V Code” of “Uncomplicated Bereavement” (V62.82) may be used. (The “V” codes, of course, are not “mental disorders”). The DSM-5 provides useful guidance on when to apply MDD in the post-bereavement period (e.g., the footnote on p. 161 of the DSM-5 manual lists several features that help differentiate ordinary grief from a major depressive episode). For example, in bereavement-related grief, self-esteem is usually preserved; in MDD, feelings of worthlessness and self-loathing are common. In ordinary grief, the emotional pain is usually accompanied by positive emotions and fond recollections of the deceased; in MDD, pervasive misery and unhappiness are typical. Another important distinction highlighted by Dr. Kay R. Jamison in her book, Nothing Was the Same is that the normally grieving person is “consolable” by friends, family, and even literature, while the person with MDD usually is not.

ARE TWO WEEKS TOO BRIEF?

To be sure, the two-week minimum duration for diagnosing MDD is often too brief to reach a confident diagnosis of almost anything, particularly in the post-bereavement period. But this two-week duration criterion has applied to all instances of MDD (e.g., after job loss, divorce) since the DSM-III appeared in 1980. (Curiously, few critics of the DSMs objected to the two-week period until the matter of bereavement arose, in advance of the DSM-5). Moreover, in clinical practice, it is quite rare for a patient with “normal” grief to seek professional treatment within two weeks of the death of a loved one. When medical treatment is sought so soon after a death, the patient usually A) has “self-selected” treatment, owing to profound distress or incapacity or B) has been referred by family members, who believe the patient is suicidal, psychotic, or unable to carry out the activities of daily living. Under such dire circumstances, the DSM-IV bereavement exclusion would not have applied anyway. As for the fear that removal of the bereavement exclusion will lead to hordes of bereaved patients being inappropriately diagnosed with MDD, the epidemiological data do not support this view. For example, in the overall NESARC sample, only 0.5 percent of subjects met criteria for “bereavement-excluded depression.”⁸

Finally, nothing in the DSMS will prevent the prudent clinician from undertaking a period of “watchful waiting” during the first few weeks after a bereaved patient presents with depressive symptoms, in order to ascertain the “trajectory” of the patient’s condition. Some patients will show marked improvement in their depressive symptoms, even though their grief—quite understandably—may persist for weeks, months, or years. (Contrary to frequent misrepresentations in the lay press, the DSMS sets no pre-ordained “time limit” on normal grief). Even if post-bereavement MDD is diagnosed, nothing compels the psychiatrist to begin antidepressant treatment. For mild-to-moderate, non-melancholic presentations of MDD, “talk therapy” alone may suffice. As for concerns that primary care physicians will be induced by the new criteria to prescribe antidepressants inappropriately, my colleagues and I believe this hypothetical concern is best addressed through continued medical education and enhanced psychiatric consultation with primary care physicians—not by preemptive gerrymandering of our diagnostic criteria for MDD.^2,10

SUMMARY

In sum, while the studies to date are not conclusive, the best available evidence suggests that the DSMS was justified when it eliminated the bereavement exclusion. No, we must not “medicalize” normal grief, but neither should we “normalize” the serious disorder of major depression simply because it occurs in the context of recent bereavement.

ACKNOWLEDGMENT

Dr. Pies wishes to thank Dr. Carlene MacMillan, Dr. Sidney Zisook, and Dr. M. Katherine Shear for their guidance and support.