Is Grief a Disease? Sometimes.

In a classic 1961 article,¹ Engel posed the question “Is grief a disease?” After a consideration of the typical characteristics of grief—the initial shock and disbelief in response to an important loss followed by painful experiences of loss and sadness, often with a sense of emptiness, hopelessness, and loss of interest in usual activities, followed by the “work of mourning,” an often prolonged phase of restitution and recovery—Engel concluded that “the experience of uncomplicated grief also represents a manifest and gross departure from the dynamic state considered representative of health and well being.”^{1 (p20)} Engel thus viewed grief as an appropriate topic for clinical research.

In a much earlier (1917) classic article² on “Mourning and Melancholia,” Freud reached a different conclusion. He viewed grief as a normal, although very painful, life experience characterized by the necessary but time-limited psychological process of mourning. Such normal grief after a loss was to be distinguished from the pathological state of melancholia, where there is a marked increase in severity and duration of the symptoms of loss and, in particular, a marked fall in self-esteem with bitter self-reproaches, self-reviling, and expectations of punishment. In Freud’s psychoanalytic interpretation, this was due to the unconscious anger at the lost person being “turned against the self.”

Such earlier conceptualizations, based entirely on clinical observations, were the precursors to current views about normal and abnormal grief, still largely based on observational studies, although with more attention to systematic collection of data to reduce potential sources of bias. The current diagnostic manual of the American Psychiatric Association, the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition, Text Revision,³ recognizes that typical depression symptoms (such as sadness, low energy, and sleep and appetite disturbances) are common after the loss of a loved one but that a major depressive disorder (MDD) should not be diagnosed unless the symptoms persist for more than 2 months (as opposed to the 2-week duration requirement when depression symptoms occur in the absence of such a loss) or include marked functional impairment, morbid preoccupation with worthlessness, psychotic symptoms, suicidal ideation, or psychomotor retardation.

“Bereavement” is included in a chapter on “Other Conditions That May Be a Focus of Clinical Attention,”^{3 (p741)} in which the important clinical point is made that transient experiences of thinking that one hears the voice of or transiently sees the image of the deceased person can be part of normal grief. Thus, the current official psychiatric nomenclature distinguishes between normal grief (bereavement) and depressive disorder precipitated by loss of a loved one in a manner that is analogous to the distinction between depression as a normal mood variation and depression as a mood disorder. Namely, “ . . . periods of sadness are inherent aspects of the human experience”^{3 (p355)} that should not be diagnosed as an MDD unless criteria are met for severity, duration, and clinically significant distress or impairment.

Some investigators studying persons who have persisting symptoms and impairment following bereavement have concluded that there is a syndrome of complicated grief (also called traumatic grief) that can sometimes co-occur with MDD or posttraumatic stress disorder (PTSD) but that has distinct phenomenology, clinical course, and outcomes, thus meriting a separate diagnosis.^{4 - 6} Characteristic features include persisting preoccupation with thoughts about the lost loved one along with longing, yearning, and inability to accept the death, distressing intrusive thoughts about the death, and avoidance of reminders of the loss. A helpful summary of similarities and differences among complicated grief, MDD, and PTSD is provided in a table in the article by Shear et al⁷ in this issue of JAMA. Diagnostic criteria for complicated grief have been proposed^{8 - 9} and an inventory to quantify its features has been published.¹⁰

One confusing terminological issue is that the term “pathological grief” was used in earlier literature for various aspects of abnormal grieving but that heterogeneity and the lack of consistent criteria or definitions were major problems.¹¹ Perhaps the most important finding about complicated grief to date from a clinical perspective, and an indication that its clinical importance goes well beyond academic hair-splitting, is that clinical experience and several studies^{7 ,12 - 13} indicate that the complicated grief symptoms do not appear to respond well to either antidepressant medication or psychotherapy, even with improvement in depressive symptoms.

Shear et al⁷ take a logical step in developing a clinically relevant understanding of complicated grief: empirical testing of a form of psychotherapy specifically developed to treat it. This randomized controlled trial enrolled 95 patients who had high scores on the Inventory of Complicated Grief¹⁰ with grief symptoms persisting for at least 6 months following the loss of a loved one. Strikingly, the median number of years since the loss was about 2, with the range extending to more than 36 years, indicating the chronicity that is often a feature of this disorder. Almost half of the patients met criteria for current MDD, about half did for current PTSD, and almost half continued taking antidepressant medication they had started before study entry.

The study compared the effects of a planned 16 sessions of interpersonal psychotherapy (IPT), a form of individual psychotherapy previously shown to be effective for MDD, with the same number of sessions of the newly developed complicated grief treatment (CGT). The IPT sessions focused on grief so that any differences in outcome would not be due to a simple failure to focus on grief issues. As described by the authors, CGT included techniques involving exposure to aspects of the death of the loved one that were persisting problems for the patient despite the passage of time and any prior treatment he/she had received. These exposure techniques were developed because patients with complicated grief typically experience the loss as a trauma, whether or not the circumstances of the death were violent, as was true for only a third of the study patients.

The traumatic aspect of this condition is a “separation trauma” syndrome^{8 - 9} that is not necessarily triggered by physical trauma. Painfully intrusive memories and images related to the death, along with unsuccessful efforts to avoid them, became important and persisting symptoms similar to the intrusive reexperiencing of trauma characteristic of PTSD, for which exposure treatment has been shown to be effective.¹⁴ In CGT, the patient goes through a “revisiting” of the death—retelling the story of the death, listening to a recording of the story, and having an imagined conversation with the deceased as guided by the therapist. These exposure techniques illustrate that the treatment requires considerably more than empathic listening, although having an empathic therapeutic relationship with the patient is obviously essential to assist in completing the painful but eventually liberating process of grieving. The difficulty of this process is illustrated by the fact that 12% of CGT patients in the study discontinued the treatment reporting that it was too difficult and a similar proportion completed the treatment but without participating in the imaginal exposure. Based on blinded independent evaluation ratings, treatment responses of much or very much improved were significantly (P = .02) more common for CGT (51%) than for IPT (28%) with a number needed to treat of 4.3.

The results of the study by Shear et al, while interesting and provocative, obviously leave a number of unanswered questions about complicated grief and its treatment. Although CGT was shown to be superior to IPT, a finding of particular relevance to the distinction of complicated grief from depressive disorder in view of IPT’s demonstrated efficacy for MDD, the 51% response rate could be viewed as disappointing. Does that indicate a need for improvements in the treatment procedures or perhaps a need for longer duration of treatment? Do the psychological demands that the CGT exposure techniques place on patients mean that its acceptability and effectiveness will be limited in nonresearch clinical practice?

In addition, a number of broader issues about complicated grief need resolution. Reliable information is needed about its prevalence, clinical correlates, and risk factors for its development. The use of 2 different but interchangeable terms, complicated grief and traumatic grief, needs to be resolved. In my view, the former would be more appropriate since the use of “traumatic” would very likely perpetuate a misunderstanding that physical trauma must have occurred. Although there is a growing consensus about the clinical characteristics of complicated grief, the same is not true for a required minimal duration of symptoms, with time requirements varying from 2 months,⁹ 6 months,⁷ and 14 months,⁸ perhaps reflecting the wide range of times considered appropriate for mourning in various cultures. The question of whether this condition should be an officially recognized mental disorder separate from MDD and PTSD is an important issue for the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition, currently planned for publication by the American Psychiatric Association¹⁵ in 2012. It is clear that awareness about depression is important for all physicians.¹⁶ Is that also true about complicated grief, since patients and families almost certainly consult primary care physicians about the persisting symptoms and dysfunction associated with it?

A concern that some might raise is that the concept of complicated grief as a disorder warranting treatment is yet another example of the medicalization of various aspects of the human condition.¹⁷ The available evidence that distinguishes complicated grief from normal grief^{9 - 10} and also from MDD^{7 ,9} and PTSD⁷ appears to provide a compelling response to that concern. Thus, the answer to the question “Is grief a disease?” is “sometimes.” The painful process of normal grief following bereavement certainly warrants sympathy and concern, along with the support of family and friends. Complicated grief warrants more research about effective ways to prevent and treat it.