Grief and Depression Redux: Response to Dr Frances’s “Compromise”

Sep 28, 2010

Clearly, we all share the goals of respecting-not “medicalizing”-ordinary grief; as well as recognizing and treating clinically significant depression. We differ with Dr Frances in how to achieve these goals, while remaining faithful to the best available scientific data.

We thank Dr Allen Frances for his thoughtful exploration of some common ground with our position on the bereavement exclusion (BE). Clearly, we all share the goals of respecting-not “medicalizing”-ordinary grief; as well as recognizing and treating clinically significant depression. We differ with Dr Frances in how to achieve these goals, while remaining faithful to the best available scientific data. In this piece, we try to clarify what is actually being proposed for DSM-5; how we differ or agree with Dr Frances’s most recent  proposals; and how arguments against eliminating the BE from DSM-V are ill-founded.

First, we need to highlight some myths and misunderstandings as regards elimination of the BE from DSM-5. No one who favors this change wants to imply that grief should be “over” within 2 weeks, 2 months, or even longer, after the death of a loved one. Rather, we argue that grief and major depression are separate constructs, although there is some symptomatic overlap-a point we will explicate shortly. Moreover, many bereaved individuals grieve intensely for protracted periods, if not for a life time, whether or not they also meet criteria for a major depression episode (MDE). Furthermore, ordinary grief does not become pathological or mysteriously transformed into major depression after 2 months, 6 months, or any number of months.  Rather, “normal” grief may last, in its various forms, for months to years .1 These points are not adequately explained in the DSM-IV, and need to be clarified in DSM-5.

At the same time, it is critical to note that for some vulnerable individuals, the death of a loved one may precipitate a MDE, and that in no sense is the person “immunized” against clinically meaningful depression by the recent loss-nor should our diagnostic classification imply otherwise. Moreover, when MDE develops in the context of bereavement, grief may be even more severe and protracted than otherwise. (Technically, “MDE” is not a formal DSM diagnosis, and the term major depressive disorder [MDD] should be used; however, to remain consistent with Dr Frances’s terminology, we will retain the term “MDE.”)

Now to some specific statements in Dr Frances’s piece that require comment:

“We [Zisook, Frances, and Pies] agree that the problem is the fuzzy boundary between normal grief and Major Depressive Episode (MDE).” Actually, we don’t quite agree with that formulation. To view the problem merely as one of “fuzzy boundaries” is to misapprehend the qualitative and categorical differences between ordinary grief and MDD, notwithstanding their symptomatic overlap (sadness, tearfulness, disturbed sleep, etc). Ordinary grief is not an “illness” at all, but a universal, adaptive response to the stress of bereavement. If pressed to identify the “problem” with the present diagnostic system, we would point to its predilection for nullifying or undermining the diagnosis of MDD, even when all symptom and duration criteria are met, solely on the basis of bereavement within the past 2 months.

“…severe depressive symptoms during bereavement are already meant to be diagnosed as MDE using the current guidelines in DSM-IV-TR. . .  If the grief symptoms are severe, DSM-IV-TR insists that MDE be diagnosed.” Dr Frances here argues that there is no need to eliminate the BE, because the current DSM framework already guards against missing MDE when it is “severe.” As we interpret the DSM-IV, this characterization is only partly correct. Within the current DSM framework, a clinician is justified in over-riding the BE only if certain “conditional” features of the depression are present, such as severe psychomotor retardation, marked functional impairment, psychosis, and suicidality. However, there are many ways in which a bereaved person’s depression may be “severe” but not conform to DSM's over-ride criteria. Thus, a recently (less than 2 months) bereaved patient meeting symptom and duration criteria for MDE may have profound loss of pleasure, markedly decreased appetite, severely impaired concentration, and severe sleep disturbance, and may score very high on the Hamilton or Beck Depression Inventory scales. All other things being equal, this patient would probably not receive a diagnosis of MDE using current DSM-IV-TR guidelines governing the bereavement exclusion

“…the test case is someone who has lost a spouse or child and has just two weeks of sadness and loss of interest, appetite, sleep, and energy. Such a person would have to be diagnosed with MDE if we were to follow the DSM 5 suggestion to simply remove the Bereavement exclusion.” We respectfully disagree. By DSM-IV’s-and almost certainly, DSM-5’s-rules, one can’t receive a diagnosis of MDE unless there is “… clinically significant distress or impairment in social, occupational, or other important areas of functioning.” So, in the “test case” Dr Frances describes, there would be no requirement-indeed, no justification-for diagnosing MDE, absent “clinically significant distress or impairment” (admittedly, a judgment call). In the absence of such distress or impairment, we believe the clinician would be justified in simply using the current “V code” of bereavement.  In effect, we don’t need the bereavement exclusion to avert a diagnosis of MDD, so long as there is no finding of clinically significant distress or impairment. On the other hand, if there were clinically significant distress and impairment accompanying the symptoms in Dr Frances’s test case, we would not want to preclude diagnosing an incipient bout of MDE-though, as Dr Frances would agree, the 2-week minimum duration (likely to be retained in DSM-5) may be too short a period to diagnose anything with confidence. In ambiguous cases of bereavement with depression, in which the “clinically significant distress or impairment” criterion is questionable, the clinician could diagnose Adjustment Disorder with depressed features.

In short, dropping the BE would not automatically compel clinicians to diagnose MDE in very recently bereaved individuals who meet mild symptomatic criteria for MDE, as in Dr Frances’s “test case.”

“. . . There is no research suggesting that the distinction between normal grief and mild MDE can be reliably and validly made so early after the loss of a loved one [ie, within 2 weeks].” Here we both agree and disagree with Dr Frances. The main problem with the 2-week minimum duration is not that it impairs our ability to distinguish ordinary grief per se from major depression. Rather, the 2-week minimum exacerbates the inherent difficulties of distinguishing mild MDE of short duration from any kind of non-pathological sadness, not just that of bereavement. A 2-week period is also too short, in our view, to gauge adequately the “trajectory” of the patient’s symptoms; eg, whether they are waxing or waning. In our table, we suggest using additional data, such as personal and familial history of MDE, as well as phenomenological data, to make these difficult diagnostic calls.  Nonetheless, our position is generally in agreement with that of Dr Frances, when he writes that “…there is no need to medicate this griever having mild and completely expectable symptoms for so short a period [2 weeks or fewer]. We would both instead recommend a combination of commiseration, empathy, support, and watchful waiting…”.  However, if the depressive symptoms persist or intensify, more active treatment strategies become paramount.

“The diagnosis of MDE may be inappropriate if the individual's culture calls for a more profound expression of grief.” We agree that clinicians should remain aware of and sensitive to specific, culturally-based mourning rituals, such as “sitting shivah” (7-day mourning period) in the Jewish faith, or the 3-day funeral characteristic of Hmong culture.2 However, we are not aware of any clinical studies showing that the standard DSM-IV criteria for MDE-including, it must be repeated, that of clinically significant distress or incapacity-- are invalidated as a consequence of such cultural variations in grief and mourning. Indeed, since we believe that grief is a distinct construct from MDE, we would not expect any substantial confusion to arise in this regard, even when the person’s expression of grief is “profound” and intense.

“…to remove the Bereavement exclusion for MDE would disastrously open the floodgates to the misdiagnosis and overtreatment of normal grief  (especially by hurried primary care physicians who do much of the prescribing)… We can assume that drug advertising to patients and marketing to doctors will result in a flood of treatment that they would also question.” In contrast to the DSM-IV, the ICD-10 does not use or recognize a “bereavement exclusion” for  major depression. If Dr Frances is correct, we should be able to detect a signal of the “floodgates” opening in Europe and other parts of the world where the ICD-10 criteria for major depression are widely used. Yet we are not aware of data showing that the ICD-10 system leads to a vast increase in the prevalence of MDE, or to inappropriate treatment of depression.  On the contrary, European studies generally find similar depression prevalence rates when ICD-10 and DSM-IV criteria are compared.3 Moreover, the actual percentage of patients presenting with bereavement-related depression is likely to be small, in comparison with the total number of depressed patients; eg, Corruble et alfound that, of their total sample of patients with depression, 8.5% would have been “excluded” by virtue of bereavement, using the DSM-IV BE exclusion rules. This is not a trivial number, but it does not suggest to us that the “floodgates” would open in the US if the BE were dropped from DSM-5. Furthermore, we would expect only a small fraction of this percentage to present to a clinician within the first 2 weeks of bereavement, based on our clinical experience; hence, Dr Frances’s “test case” would probably be an exceedingly rare occurrence in the psychiatrist’s office. Rather, depressed and grieving patients presenting only 2 weeks after bereavement would likely be limited to a few hospitalized patients who had developed a psychotic depression or who had made a frank suicide attempt-in which cases the BE would not apply anyway. With respect to “over-treatment of normal grief” by harried PCPs, we regard this hypothetical outcome as a matter best addressed through medical education and consultation with our PCP colleagues-not by retaining a diagnostic scheme that flies in the face of the best available data. More important: most studies of primary care have pointed to high rates of under-diagnosis and under-treatment of major depression-a problem we regard as far more serious than the hypothetical risk of PCPs “over-calling” grief as major depression. Indeed, in primary care settings, up to 50% of MDE cases go unrecognized, and treatment is often inadequate.5 We would invite  Dr Frances to cite contrary studies showing that PCPs frequently over-diagnose “normal grief” as MDE, as we are unaware of such findings.

“It is an unproven and unlikely red herring that the DSM 5 change will save lives (or for that matter have any beneficial effect at all)…” We have never claimed that eliminating the BE from DSM-5 will “save lives.” What we have said is that retaining the BE creates a completely unnecessary and avoidable risk; ie, that potentially serious MDE will be misperceived and misdiagnosed as “just normal grieving.” In principle, this could lead to potentially suicidal individuals being sent home with nothing more than a, “There, there-you just need time to grieve,” from the misguided clinician. What we believe is an “unproven and unlikely red herring” is Dr Frances’s “open floodgates” scenario. It will of course require empirical studies, after DSM-5 is operationalized, to determine the actual effects of eliminating the BE, if that is what the DSM-5 work group decides.

“There is no proven efficacy for medication treatment given after just 2 weeks of mild symptoms in grievers….” Of course, we agree, and we would never advocate antidepressant treatment for uncomplicated, “normal” grief of any duration. Indeed, even for mild-to-moderate cases of MDE, we would usually opt for psychotherapy as the treatment of first choice. That said, we would point out that when grief is part of bereavement-related MDE, there is preliminary evidence that antidepressants may be of benefit and do not interfere with the grieving process.6,7

“Reasonable people can disagree about the precise duration requirement before grief can be considered mild depression… Pies/Zisook suggest 1 month. I am OK with the DSM-IV-TR 2 months…” We believe Dr Frances is conflating several different time frames and conditions. First, we must again note that grief does not somehow metamorphose into depression-mild or otherwise-after a fixed number of days, nor are we advocating a “time limit” on ordinary grief. Grief is grief; depression is depression (Table). Our “1 month” proposal refers to the minimal duration we would like to see applied, in most cases, to the diagnosis of first episode MDE (we allow for shorter duration when melancholic, psychotic, or suicidal features are present). Further research and field trials will be needed to determine the most useful “cut-offs” for these types of depression. The “2 months” Dr Frances alludes to has nothing directly to do with when “grief can be considered mild depression”; rather, the 2 month criterion is simply the outer limit for applying DSM-IV’s BE; ie, when MDD symptoms occur beyond 2 months after bereavement, DSM-IV implies that use of the BE (all other things being equal) is not warranted. We believe all these issues could be set aside if the DSM-5:
•Drops the BE
•Increases the minimal duration to 1 month, for most cases of first episode, nonpsychotic  MDE
•Clearly describes the phenomenology and symptom picture for ordinary grief and bereavement.

To conclude: We believe that the BE is simply unsupportable by the best available data, and that eliminating it from the DSM-5 will remove a confusing and potentially harmful distraction from the clinician’s deliberations. We do not anticipate that any “floodgates” will open by eliminating the BE, nor do we think our recently-bereaved patients who meet MDE criteria will feel disrespected by our honest and empathic judgment that they are also clinically depressed and merit our professional care. 



Major Depressive Episode (moderate to severe) with or without bereavement/ recent loss

Uncomplicated (“Normal”, “Productive”, “Adaptive”) Griefin context of bereavement or other loss

Features in common

Sadness, tearfulness, sleep disturbance, decreased socialization, decreased appetite

Sadness, tearfulness, sleep disturbance, decreased socialization, decreased appetite

Differentiating signs and behaviors*

Usually, prolonged, persistent impairment in ability to carry out most activities of daily living (ADLs); pronounced, persistent social/vocational impairment; social relations often poorly maintained; pronounced psychomotor slowing or agitation; consistent early a.m. awakening, pronounced weight loss are common.   



Variable, but often able to carry out most ADLs after first 1-2 weeks of bereavement; often able to maintain most social/vocational role responsibilities and relations after first 1-2 weeks; no consistent or prolonged psychomotor impairment; initial insomnia more common than consistent early a.m. awakening; weight loss usually mild/ moderate.

Differentiating Symptoms and Phenomenology (contents & structure of  person’s experience, “world-view”) *

Mood depressed all day, nearly every day; depression is autonomous (not mitigated by positive stimuli); commonly: pronounced pathological guilt, self-loathing, markedly lower self-esteem, feels “worthless”; hopelessness, suicidal intentions or plans common; delusional beliefs may be present, often of nihilistic nature.



Person feels depression will never end; sense of bleak future, feeling “trapped”, imprisoned; thoughts constantly negative, self-focused; even with support, often feels “outcast”, alienated from others; inconsolable even with distraction, music, etc.  May feel like “sinner”, “condemned by God,” “rotting away” , “deserve to die”, etc. (psychotic MDE)

Sadness comes in “waves” most days but bereaved also has “good” days; waves of sadness brought on by thoughts/images of the deceased; bereaved not pathologically guilt-ridden (guilt confined to wishing they had “done more” for deceased or been more compassionate); self-esteem preserved, no self-loathing. Hopelessness uncommon; thoughts of death, suicidal ideas confined to reunion fantasies, not frank suicidal urges or plans; not delusional (but visual/auditory hallucinations of deceased are common).

 Bereaved usually believes life will someday get back to normal, he/she will again feel like “old self”;  positive thoughts & memories interspersed with painful recollections of deceased; with appropriate support, bereaved does not feel “outcast”, alienated; often able to make use of and appreciate love, support; often able to be consoled by family & friends, music, literature, etc; does not feel like “sinner”, “condemned”, deserving of death; no nihilistic delusions but may feel “like part of me is missing” or “I lost the best part of me.”

Past mood disorder history/psychiatric comorbidity

Often positive for previous bout of major depression; co-morbid psychiatric disorders (anxiety, OCD, etc) common

May or may not have any history of mood disorder or other psychiatric disorders

Family history of mood disorder

Often positive for depression in first-degree relatives

May or may not have family history of mood disorder

Biological abnormalities

Melancholic cases frequently associated with abnormal HPA axis activity, elevated cortisol; PET studies often show abnormal frontal lobe activity (↑
or ↓)

Insufficient data but no consistent evidence to date of major biological disturbances in the absence of concurrent mood disorder

Response to somatic interventions

Moderate to severe cases show response to antidepressants > placebo; severe/melancholic,
psychotic depression highly responsive to ECT

Insufficient data, but no evidence that somatic treatment is needed; most episodes resolve with time and support. In cases with concurrent MDE, antidepressants do not appear to interfere with grief work

Psychosocial interventions

Responsive to CBT, IPT and other psychotherapies

Support groups and grief counseling may be useful in some cases, but no evidence that psychotherapy is required for resolution of ordinary grief


Good for acute episode, though up to 1/3 cases may be treatment-resistant; major depression sometimes chronic and usually recurrent.  

Excellent unless “complicated” by concurrent MDE or other psychiatric disorder or psychopathology.


 *There is considerable variability in both MDE and uncomplicated grief, and these differentiating features are intended only as general guidelines. While the majority of individuals with these conditions tend to show the differences indicated, exceptions are often seen in clinical practice, and the first few weeks after bereavement may be particularly intense and distressing, leading to diagnostic confusion. In first episodes of MDD, if symptoms are mild and distress equivocal, we generally recommend waiting a month after symptom onset before making a firm diagnosis.  In recurrent cases; in cases when the distress or impairment is obvious;  when the index depressive presentation is characterized by melancholic or psychotic features or suicidality; or when the presentation is  similar to that of a previous, well-documented major depressive episode, we do not feel that waiting even as long as 2 weeks is necessary for making a diagnosis of MDE or instituting treatment.   



1. Zisook S, Shear K. Grief and bereavement: what psychiatrists need to know. World Psychiatry. 2009;8(2):67-74.

2. Lee K. Rituals, Roles, And Responsibilities Included In A Hmong Funeral: A Guidebook For Teachers To Better Understand The Process Their Hmong Students Experience In A Time Of Family Loss. Doctoral dissertation. Available at: Accessed Sept. 24, 2010.

3. Ayuso-Mateos JL, Vazquez-Barquero JL. Depressive disorders in Europe: prevalence figures from the ODIN study. British J Psych. 2001;179:308-316.

4. Corruble E, Chouinard VA, Letierce A, et al. Is DSM-IV bereavement exclusion for major depressive episode relevant to severity and pattern of symptoms? A case-control, cross-sectional study. J Clin Psychiatry. 2009;70(8):1091-1097. Epub 2009 Jun 30.

5. Saver BG, Van-Nguyen V, Keppel G, et al: A qualitative study of depression in primary care: missed opportunities for diagnosis and education. J Am Board Family Med. 2007;20:28-35.

6. Zisook S, Shuchter SR, Pedrelli P, et al. Bupropion sustained release for bereavement: results of an open trial. Clin Psychiatry. 2001;62(4):227-230.

7. Hensley PL, Slonimski CK, Uhlenhuth EH, et al. Escitalopram: an open-label study of bereavement-related depression and grief. J Affect Disord. 2009;113(1-2):142-149. Epub 2008 Jul 2.