Depression or Bereavement? Defining the Distinction

Michael B. First, MD; Ronald W. Pies, MD; Sidney Zisook, MD


April 08, 2011

In This Article

Removing the Exclusion: Is There Justification?

Dr. First: As Drs. Pies and Zisook note in their thoughtful responses to my summary of the cost-benefit analysis of the proposed removal of the bereavement exclusion from the criteria for an MDE,[1] we agree on many points but disagree what action to take. For example, we agree that when evidence casts a substantial doubt about the validity of a category, it is reasonable to consider eliminating it. Pies and Zisook claim that "several reviews and population-based empirical studies have found that bereavement-related depressive episodes and nonbereavement-related depressive episodes are more similar than different" but these findings are not relevant to the central question of the validity of the bereavement exclusion in the first place. The exclusion was added to the DSM-III criteria for MDE in recognition of the work of Paula Clayton and others did examining the occurrence of depressive symptoms in widows and widowers shortly after the death of their spouses. They found that it was common for them to develop symptoms that resembled those of a depressive episode[6,22,23] and that measures should thus be put into place to discourage clinicians from labeling individuals who were likely to be experiencing normal grief as having a mental disorder. In order to justify removal of the bereavement exclusion, one would have to demonstrate either that the original concerns about there being a false-positive problem were (and continue to be) unjustified (ie, that individuals experiencing normal grief do not in fact have symptoms that could qualify for a diagnosis of an MDE) or that the exclusion creates a false-negative problem that can only be solved by its elimination.

Besides data from Clayton's pioneering work on widows and widowers, Dr. Zisook's own 1991 study in fact provides unequivocal evidence that some normally grieving individuals meet full criteria for an MDE and, were it not for the exclusion, would have been inappropriately diagnosed as having major depression. In that study, Zisook and Shuchter interviewed men and women whose spouses had recently died 2 months, 7 months, and 13 months after the death.[24] They reported that 30 (8.6%) of the 350 widows and widowers met full symptomatic criteria for an MDE immediately after the death of their spouse but were no longer depressed by the end of 2 months. Without the bereavement exclusion in place, these clear-cut cases of uncomplicated bereavement could have been mislabeled as being clinically depressed. Extrapolating to the general population of those who experience bereavement (ie, all of us), this roughly 10% rate of misdiagnosis represents an enormous false-positive problem.

Pies and Zisook's likely response to this estimate of potential false positives is that although the bereaved individuals in the Zisook and Schucter study qualified for a syndromal diagnosis of MDE according to the DSM-III-R criteria, they would likely not have qualified for a diagnosis under the more stringent DSM-IV/DSM-V criteria because those criteria sets impose an additional requirement of meeting a "clinical significance criterion" (ie, criterion C). My anticipation of Pies and Zisook's response is derived from the comment in their reply casting doubts about the validity of Clayton's classic studies of the 1970s because the Feighner criteria for major depression were used which "are similar to present-day DSM criteria but do not include the important component of significant social/vocational impairment." The implication here is that had the DSM-IV/DSM-V requirement that the depressive syndrome cause significant social/vocational impairment been applied, then none of those individuals experiencing normal grief would have met criteria for an MDE, obviating the need for the bereavement exclusion.

There are several problems with this claim, however. First of all, the actual DSM-IV requirement is not that the symptoms "significantly interfere with social/vocational function" as claimed by Pies and Zisook, but rather that the symptoms "cause clinically significant distress or impairment in 1 or more important areas of functioning," a much lower threshold and thus considerably less likely to screen out false positives since most acutely grieving individuals are in significant distress. Moreover, epidemiological data[25] suggest that only a tiny proportion (2.8%) of individuals who meet the symptomatic criteria for an MDE would not also meet the "clinically significant distress or impairment" criterion, indicating that the vast majority of normally grieving individuals who met the symptomatic criteria for an MDE would have also met the clinical significance criterion. Thus its inclusion in DSM-IV/DSM-V does little to correct the false-positive problem stemming from the symptomatic resemblance between normal grief and mild depression.

Given the body of evidence suggesting that the rationale for including the bereavement exclusion in the MDE criteria is basically sound (ie, that were it not for the exclusion, a significant proportion of normally grieving individuals would meet the symptomatic MDE criteria during the first 2 months after the loss), the only justification for removing it would be evidence demonstrating that doing so would have such a large impact in terms of promoting the appropriate diagnosis of depression that the benefits of reducing false negatives outweigh the risk for false positives. As noted in their reply, however, Pies and Zisook agree that there is no evidence that removing the exclusion would in any way contribute towards correcting the general problem of underdiagnosis and undertreatment of depression in primary care and other settings, which instead stems from "inadequate assessment, rushed diagnosis and the failure to apply DSM criteria."

By arguing against the removal of the bereavement exclusion in DSM-V, I am not saying that the validity of the bereavement exclusion in terms of differentiating between uncomplicated bereavement and bereavement complicated by major depression could not be improved. The current guidelines (ie, persistence for longer than 2 months or being characterized by marked functional impairment, morbid preoccupation with worthlessness, suicidal ideation, psychotic symptoms, or psychomotor retardation) are based on a rather limited empirical foundation and could perhaps be strengthened (eg, by requiring 7 out of 9 MDE symptoms, rather than 5[26] or prior history of depressive episodes[27]). Nonetheless, as long as the definitional criteria for an MDE remain relatively mild (ie, 5 out of 9 symptoms lasting only 2 weeks) so that a significant proportion of grieving individuals meet the criteria, retaining the bereavement exclusion in DSM-V in some form remains essential.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.