DSM-5 grief scorecard: Assessment and outcomes of proposals to pathologize grief


  • Jerome C. Wakefield

    1. Department of Psychiatry, School of Medicine, New York University, New York, NY
    2. Silver School of Social Work, New York, NY
    3. InSPIRES (Institute for Social and Psychiatric Initiatives–Research, Education and Services), Bellevue Hospital/New York University, New York, NY
    4. Department of Psychiatry, Division of Clinical Phenomenology, Columbia University College of Physicians and Surgeons, New York, NY, USA
    Search for more papers by this author

Where does grief stand diagnostically, now that the dust has settled and the DSM-5 has been approved? The DSM-5 Task Force considered an unprecedented series of proposals to identify grief-related mental disorders where now there are assumed to be normal variations. The proposals taken together had the potential to transform psychiatry's conceptualization of grief and the clinician's response to bereaved patients. Targets for pathologization included both depressive symptoms during grief and grief itself – the yearning, disbelief, and other experiences distinctive of grief.

Four grief-related proposals made it to the final leg of the DSM-5 revision process, a major event in itself. Here I review the proposals, assess their validity, and present the Task Force's final decisions, providing an overview of the status of grief post-DSM-5.


This was perhaps the most controversial diagnostic proposal since depathologization of homosexuality. Grief sometimes triggers a major depressive disorder (MDD). However, some depressive symptoms, such as depressed mood, insomnia, decreased interest, decreased appetite, and lack of concentration, are general-distress symptoms that frequently occur in normal grief [1]. Thus, normal grief can satisfy the DSM's 5-symptoms-for-2-weeks criterion for MDD, yielding a mistaken “false positive” MDD diagnosis. The bereavement exclusion (BE) rectified this situation by distinguishing as normal those “uncomplicated” grief-related depressive episodes that included only general distress symptoms and remitted quickly. “Complicated” episodes were classified as MDD, despite the recent loss, if they included pathosuggestive symptoms such as psychomotor retardation, suicidal ideation, sense of worthlessness, or lengthy duration. The BE's elimination means that two weeks of general-distress depressive symptoms after death of a loved one falls under MDD.

The main argument for the BE's elimination was that excluded cases are just like other MDD on pathology validators [2]. However, when reviewed, claims that research evidence supports such similarity were shown to be unfounded [3]. Several new studies falsified the similarity claim, showing, for example, that depression recurrence and development of anxiety disorders, which occur at high rates after MDD, occur no more frequently in BE-excluded episodes than in populations who have never had MDD, demonstrating the BE's strong predictive validity [4-6]. Warnings that excluded depressive episodes would contain elevated rates of suicidal cases turned out to be groundless [7]. Two studies demonstrated that uncomplicated grief-related depression is similar to uncomplicated reactions to other stressors, raising the question of whether the BE should be eliminated or expanded to other stressors [8, 9]. Recent studies answer that a broadened exclusion applied to uncomplicated reactions to all major stressors has both concurrent and predictive validity, with recurrence and other predictive validators not different from background population levels, unlike other MDD [10, 11].

Assessment: This is an invalid and empirically unsupported proposal. The BE's rules have been demonstrated to be both concurrently and predictively valid with ample, replicated, high-quality evidence. Speculative claims supporting elimination have been empirically falsified.

Outcome: The proposal to eliminate the BE was accepted by the DSM-5 Task Force. The BE has been eliminated in DSM-5. It has been replaced by a vague note stating that normal grief and reactions to other stressors can have depressive symptoms, and the clinician must judge the diagnosis, but with no guiding criteria, making research virtually impossible and the note likely to be ignored.


Until DSM-5, non-depressive grief feelings were not targeted by any category of disorder. However, two grief research groups have been working to validate intense, lengthy grief as pathology, called “prolonged” or “complicated” grief disorder [12, 13]. Validation rested either on risk of future harms such as disorders, thus potentially confusing risk of disorder with disorder, or on the claim that grief in the identified group is “derailed” or “frozen” in ”interminable” grief, a claim unsubstantiated by longitudinal evidence.

The two research groups proposed different diagnostic criteria for the proposed category, both claiming empirical support. The DSM-5 resolved this conflict by creating diagnostic criteria combining elements from both proposals along with some new elements, and recommending placement in section 3 for further study. Moreover, the grief researchers' criteria do not require that symptoms have been continuous since the acute grief stage, whereas the DSM-5 requires the symptoms to be present more days than not since the death. Finally, DSM-5 increased the post-loss duration threshold from the grief researchers' 6 months to 12 months, a much more defensible cut-point, though likely still too short given evidence that many individuals are still on a healing trajectory and are not “derailed” or “frozen” in their grief at that point [14].

Symptom criteria for the DSM-5 grief disorder require at least one out of four “separation distress” symptoms (yearning/longing, intense sorrow, preoccupation with the deceased, preoccupation with the death's circumstances) and at least six out of 12 additional symptoms including difficulty accepting, shocked/stunned/numb, difficulty positively reminiscing, bitterness/anger, self-blame, avoidance of reminders, difficulty trusting, wanting to join the deceased, loneliness/detachment, meaninglessness/emptiness, role confusion or feeling part of oneself died, and difficulty pursuing interests or plans. Note that all of these phenomena can occur normally during acute grief, so it is the prolonged intensity rather than a trajectory of resolution that suggests pathology.

Assessment: In principle, adding a suitably formulated category for enduring intense grief without a normal trajectory of adaptation makes sense. The DSM-5 criteria improved grief researchers' proposals in terms of face validity and greater consistency with durational evidence. However, the original proposals each had an empirical track record, whereas the DSM-5 compromise proposal has no research history. Moreover, many case examples suggest that grief at durations suggested in these criteria sets may represent a plateau along a normal but slower healing trajectory, especially when the loss or its interaction with personality or contextual variables is particularly difficult. Grief researchers' proposals seem to err on the side of caseness, whereas caution is warranted because this category has high potential for abuse, especially if grief becomes targeted for medication development.

Outcome: Persistent complex bereavement-related disorder was accepted for inclusion in DSM-5's section 3, for further study. This allows immediate diagnosis under “other specified” categories.


The DSM-IV also contained a bereavement exclusion for adjustment disorder (AD): “the symptoms do not represent bereavement”. Because of the anticipated demise of the major depression BE, it was proposed that the AD exclusion also be eliminated.

However, AD and MDD are not analogous in this regard. AD diagnosis includes the specifier “with depressed mood” (“when the predominant manifestations are symptoms such as depressed mood, tearfulness, or feelings of hopelessness”), but unlike major depression, there are no duration or symptom thresholds. Consequently, eliminating the AD bereavement exclusion would mean that any transient subsyndromal depressive symptoms such as sadness and insomnia within the first weeks or months post-loss would qualify for AD diagnosis. Such symptoms are almost universal in early normal bereavement [1]. No research has examined the AD bereavement exclusion [15].

Assessment: This is an invalid and empirically unsupported proposal.

Outcome: This proposal was rejected by the DSM-5 Task Force. The DSM-5 AD criteria include the bereavement exclusion.


Anticipating elimination of the AD bereavement exclusion, thus the perceived need to include grief symptoms among AD symptoms, a new category of “AD related to bereavement” was proposed to diagnose persistent non-depressive grief symptoms. This proposal offered a back-door way to introduce complicated/prolonged grief into the manual.

The proposed symptom criteria required that “for at least 12 months following the death of a close relative or friend, the individual experiences on more days than not intense yearning/longing for the deceased, intense sorrow and emotional pain, or preoccupation with the deceased or the circumstances of the death. The person may also display difficulty accepting the death, intense anger over the loss, a diminished sense of self, a feeling that life is empty, or difficulty planning for the future or engaging in activities or relationships”.

This definition requires only one symptom, either yearning, sorrow, “or” preoccupation; the others “may also” be present. Whether one or a few symptoms are required, there is no research on such a category, and existing evidence strongly suggests invalidity, with many or most grievers qualifying for diagnosis in multiple studies [16-20]. For example, Prigerson et al [16] found that the average yearning frequency for all grievers at 1 year post-loss is about every other day, not distant from the DSM's proposed pathological AD threshold of yearning “more days than not”.

Assessment: This is an invalid and empirically unsupported proposal.

Outcome: This proposal was rejected by the DSM-5 Task Force. No bereavement-related AD category appears in DSM-5.


Of the four proposals, the two that would have pathologized virtually all grief as adjustment disorder were rightly rejected. The bereavement exclusion to major depression was eliminated despite excellent evidence supporting its validity, a triumph of DSM politics over science. Finally, the new category of persistent complex bereavement-related disorder, in principle a needed category if properly formulated, was incorporated into section 3 of the manual for further study, with adjusted criteria that are more rigorous than the original proposals but still lack adequately demonstrated specificity.

In all, the Task Force made three reasonably wise decisions, and one major error with the BE that should be rectified as soon as possible. Post-DSM-5, normal grief remains safe from diagnosis when it includes few depressive symptoms. However, normal grief reactions that include several general-distress depressive symptoms have been mistakenly pathologized. Given how common such depressive feelings are as part of normal grief, this puts a sizable percentage of grievers at risk for false positive diagnosis.

  • Jerome C. Wakefield1-4

  • 1Department of Psychiatry, School of Medicine, New York University, 550 First Avenue, New York, NY 10016;

  • 2Silver School of Social Work, New York, NY 10003;

  • 3InSPIRES (Institute for Social and Psychiatric Initiatives – Research, Education and Services), Bellevue Hospital/New York University, New York, NY 10016;

  • 4Department of Psychiatry, Division of Clinical Phenomenology, Columbia University College of Physicians and Surgeons, New York, NY 10032, USA