The Flip-Side of “Good Grief” May be Missed Depression

The Flip-Side of “Good Grief” May be Missed Depression

My colleague Allen Frances is rightly concerned with the risk of over-calling normal grief as major depression - - that is, the risk of  "false positives" - - if the DSM-IV "bereavement exclusion" is dropped in the DSM-5 while the 2-week minimum duration criterion is retained. Indeed, Sidney Zisook and I1 have argued that the 2-week minimum is usually too brief to make a firm diagnosis of major depressive disorder (MDD -- whether or not in the context of recent bereavement). For most non-melancholic, non-suicidal cases of bereavement-associated depressive symptoms, we would like to see a 4-week minimum duration used.

Unfortunately -- as I think Dr. Frances would agree -- it seems unlikely that the DSM-5 will increase the minimum duration criterion for MDD. So we will probably be left with the question, is it still better to get rid of the "bereavement exclusion" (BE)? On balance, both Dr. Sidney Zisook and I believe it is.

Indeed, Dr. Frances's blog does not deal at all with the flip-side of the coin; that is, the potentially catastrophic risks of missing an episode of true MDD by mis-labeling it as simple bereavement, and sending the patient on his way. Given the completed suicide rate of roughly 4 in 100 among patients with MDD, [see] I believe the risk of creating "false negatives" must also be carefully considered.
And, while Dr. Frances notes that the DSM-IV provides a group of "severity" specifiers that may permit an MDD diagnosis during bereavement -- pathological guilt, clear suicidality, marked sense of worthlessness, marked psychomotor retardation, marked functional impairment, and psychosis -- it has not been demonstrated that these particular features have real predictive value or differ in frequency between bereaved and non-bereaved depressed patient.2 Furthermore, many depressed, suicidal patients do not acknowledge - - and may deliberately deny -- feeling "suicidal,"in order to avert an involuntary hospitalization.

There are other problems and paradoxes that arise from applying the DSM-IV bereavement exclusion. Consider the following case:

Mr. A visits your office complaining of 4 weeks of a persistently low mood, inability to enjoy his usually enjoyable activities, difficulty falling and staying asleep, daytime fatigue, inability to concentrate or make decisions, and poor appetite. He is otherwise in excellent health. Five weeks ago, his wife passed away. Mr. A shows none of the “severity” features listed above (no pathological guilt, clear suicidality, marked sense of worthlessness, etc).

I don't know if Dr. Frances would consider this a case of "well-established" depression or not -— since there is no uniformly accepted or validated definition of that term. But using the present DSM-IV exclusionary rules -- which Dr. Frances wants to retain-- Mr. A. would receive a diagnosis of "bereavement."

In contrast, a doctor using the ICD-10 criteria would give Mr. A. a diagnosis of major depression [F32.0], probably of mild to moderate severity. That's because the ICD-10 does not recognize a "bereavement exclusion"; ie, if you meet criteria for major depression, you've got major depression -- whether or not you have had a recent loss. Notably, Dr. Frances has called for harmonizing the DSM and ICD systems, in these same pages.

Given the high risk of suicide associated with major depression, I would prefer to err on the side of caution and go with the ICD-10 approach. This will get Mr. A. involved in the mental health system, as opposed to being sent home and  told, "You are perfectly normal. You just need time to grieve."

Contrary to widespread belief, diagnosing Mr. A. as having a major depression does not obligate the clinician to start an antidepressant; indeed, I recommend reserving medication for more severe, melancholic presentations, and instead beginning frequent meetings and psychotherapy with Mr. A. If the patient shows marked improvement over subsequent weeks, I may "downgrade" my initial diagnosis of MDD -- and no great harm has been done, so long as I protect the patient's confidentiality. If the patient's depression persists or worsens, I might then consider a trial of an antidepressant.

Sure, I know -- in a 15 minute meeting, the average PCP is likely to reach for the Rx pad, not provide psychotherapy. And, alas, most prescriptions for antidepressants are not written by psychiatrists, but by primary care doctors. But that is a matter for intensive continuing medical education and public health policy to address -- not a rationale for gerrymandering our diagnostic criteria. By the way, there is no credible evidence that antidepressant medication "interferes with "the process of working through grief; on the contrary, some open data suggest that measures of grief decline in parallel with those of depression, when an antidepressant is used in bereavement-related major depression.3 In contrast, much clinical experience suggests that severe depression itself can severely hamper "working through" of grief.

Here is yet another paradox that arises from the DSM-IV "bereavement exclusion", as noted in the Oxford Textbook of Psychopathology.4 A woman who experiences major depressive symptoms a few weeks after the death of her husband will not be diagnosed with depression, using DSM-IV exclusion rules; yet a woman who experiences the exact same depressive symptoms because her husband has abandoned her will be diagnosed with depression, under the DSM-IV's conventions. Is this any way to run a diagnostic system?

Finally, Dr. Frances "...challenge[s] anyone to distinguish clinically between 2 weeks of normal grief and 2 weeks of mild MDD..." in the course of bereavement. True: if one simply uses the symptom checklists of the DSM-IV, this can be tricky, and 2 weeks is too short,  as I've acknowledged. However, by exploring more of the patient's "inner world"-- what philosophers call "phenomenology" -- I  believe one can make the distinction in many cases.

In sorrow, grief, and ordinary bereavement, we are still capable of feeling closely connected with others; in major depression, the patient usually feels "alone," isolated, or outcast. In ordinary bereavement, the grieving person often experiences intermittent "positive" thoughts, such as pleasant memories of the deceased; in major depression, thoughts are almost uniformly gloomy. In ordinary grief and bereavement, we usually feel that someday, "life will be good again"; in major depression, hopelessness and nihilism predominate. And, importantly, as Dr. Kay Jamison notes in her book, Nothing Was the Same, the grieving person is "consolable" -- she can brighten up for a few hours with distraction, music, poetry, or social contact. This is unusual in major depression, in which mood is  relatively "autonomous."5 Going beyond the DSM checklists can thus help the clinician arrive at the correct diagnosis.

In summary: Dr. Frances and I agree that the 2-week minimum duration is usually too brief to make any confident mood disorder diagnosis. We agree that "watchful waiting" over the next few weeks may tell the tale. We agree that there is some risk of overcalling grief as major depression during the first few weeks of bereavement. Nevertheless, I believe that the risks of missing an incipient episode of MDD far outweigh the risks of diagnosing it "prematurely." I also believe that the risks of retaining the bereavement exclusion for those with 3 or more weeks of depression outweigh the risks of eliminating it, and create paradoxical conflicts with the ICD-10 system and within the DSM system itself.

Finally, I believe the APA and DSM-5 work group needs to be much more  proactive in communicating its stance to the general public, who views the DSM-5 process with mistrust and cynicism--as indeed, it views much of organized psychiatry [eg,]
Thus, I would like to get this message out to the general public:

“Psychiatrists don’t regard ordinary grief as a disorder that needs treatment, nor do we believe that all bereaved individuals ought to be “over” their grief within a 2 week period. We do not automatically give them a “mental disorder” diagnosis simply because they are intensely sad, frequently tearful, sleeping poorly, and feel distraught, 2 weeks, 2 months, or even longer, after the death of a loved one. Those are common elements of grief after bereavement, and grief is a useful and adaptive response to loss. Psychiatrists don’t want to “take away” anybody’s grief! But, unfortunately, the loss of a loved one does not “immunize” the bereaved person from the very serious consequences of major depression, which can literally be life-threatening. It’s only when certain worrisome criteria are met -- 2 or more weeks after the death of a loved one--that the bereaved person would be diagnosed with major depression. And that “call” can always be revised, if the person greatly improves over the next few weeks. Furthermore, knowledgeable psychiatrists don’t reflexively start a patient on antidepressants just because he or she has been diagnosed with major depression. Medication is best reserved for the more severe and enduring cases of major depression. However, we do follow all patients with bereavement-related depression very carefully and provide appropriate psychological support.”


1. Lamb et al. The Bereavement Exclusion. Psychiatry MMC, July, 2010.
2. Karam EG, Tabet CC, Alam D, et al. Bereavement related and non-bereavement related depressions: a comparative field study. J Affect Disord. 2009;112:102–110.
3. Zisook S, Reynolds CF 3rd, Pies R, et al. Bereavement, complicated grief, and DSM, part 1: depression. J Clin Psychiatry. 2001;62:227-230.
4. Oxford Textbook of Psychopathology. Blaney & Millon, 2008.
5. Pies R. Depression or “Proper Sorrows”— Have Physicians Medicalized Sadness? Prim Care Companion J Clin Psychiatry. 2009;11(1):38–39. Online at]. Accessed August 24, 2010.
Loading comments...
Please Wait 20 seconds or click here to close