Abnormal Grief: Should We Consider a More Patient-Centered Approach?

The loss of a significant other is a highly stressful interpersonal event (Holmes & Rahe, 1967; Osterweis, 1984) accompanied by psychological pain. This pain, felt during any normal grieving process, can be marked by the desire to reunite with the deceased, a feeling of emptiness, and depressive symptoms, such as sadness, loss of interest and pleasure, fatigue, loss of appetite, sleep disturbance, and loss of concentration (Clayton, Desmarais, & Winokur, 1968). The evolution of grief is usually benign, with a progressive decrease in psychological pain and a recommitment to life within six months (Clayton et al., 1968; Clayton, Halikas, & Maurice, 1972; Holman, Perisho, Edwards, & Mlakar, 2010; Horwitz, 2007; Zisook & Shuchter, 1991a). Not to pathologize bereavement, which is a natural life event, the DSM-IV allowed the diagnosis of major depressive disorder (MDD) in bereavement only when the criteria for MDD persisted more than two months or presented characteristics that were more intense or more debilitating than those usually observed in bereavement. In this case, DSM-IV offered the term grief-related major depression (GRMD). The DSM-IV did not recognize complicated grief (CG) as a nosological entity; the DSM-5 proposed such a diagnosis, but in the end relegated it to the category of “Condition for Future Study” under the name “persistent complex bereavement disorder” (APA, 1994, 2014).

The DSM-5 considers bereavement akin to any other stressor likely to trigger MDD, and permits making the diagnosis once the criteria of MDD are met within two weeks after the loss. To help physicians not conflate grief and MDD, the DSM-5 explains (however insufficiently) the differences between GRMD and grief. It only takes into consideration the symptom profile differences, neglecting the personal factors that make a bereaved individual vulnerable to grief.

Various therapies have been offered to help, support, and treat patients experiencing abnormal grief (Jacobs, Nelson, & Zisook, 1987; Prigerson & Jacobs, 2001; K. Shear, Frank, Houck, & Reynolds, 2005; Wetherell, 2012; Zygmont et al., 1998). Antidepressant pharmacotherapy seems to help reduce CG symptoms (Zygmont et al., 1998). Grief-related major depression also responds to antidepressants, decreasing depressive symptoms without improving the grieving process (Jacobs et al., 1987; Wetherell, 2012). Whereas pharmacological treatments seem just to reduce symptom severity, various psychotherapeutic approaches have shown efficacy in treating CG and GRMD. Interpersonal psychotherapy (IPT) has been recognized as an efficient treatment for GRMD (Klerman, 1984; Weissman, 2000). Interpersonal psychotherapy helps the patient experience a normal grief process: feeling and expressing sadness and other emotions linked to bereavement, letting go of the deceased, and re-establishing social contacts (Klerman, 1984; Weissman, 2000). The more recently developed complicated grief treatment (CGT) uses cognitive behavioral as well as interpersonal methods. In one head-to-head study it yielded a better response to treatment than IPT in treating CG (Shear et al., 2005).

Research demonstrates that the personal history of a grieving person, the nature of the person’s relationship with the deceased, and the nature of the loss affect the evolution of bereavement (Anderson, Arnold, Angus, & Bryce, 2008; Chiu et al., 2010; Dodd et al., 2008; Johannesson et al., 2009; Johnson, Zhang, Greer, & Prigerson, 2007; Keesee, Currier, & Neimeyer, 2008; Meert, Thurston, & Thomas, 2001; Nelson et al., 2010; Neria et al., 2007; Pfefferbaum et al., 2001; Robinson & Marwit, 2006; Shear & Shair, 2005; Shear, Jackson, Essock, Donahue, & Felton, 2006; Siegel, Hayes, Vanderwerker, Loseth, & Prigerson, 2008; Simon et al., 2005; Simon et al., 2007; Tomarken et al., 2008; van Doorn, Kasl, Beery, Jacobs, & Prigerson, 1998; Vanderwerker, Jacobs, Parkes, & Prigerson, 2006; Wiese et al., 2010). Some authors, mainly those following the IPT school, have highlighted the influence of interpersonal factors on the grieving process. Interpersonal psychotherapy emphasizes the need to understand how the grieving person perceives the loss, experiences the relevant emotions, commits to the immediate interpersonal environment in adapting to the new reality of life, and reestablishes interpersonal relations to recover his or her mood.

We present a clinical case to illustrate the importance of focusing care and support on the individual needs of each patient, taking into consideration the patient’s history and the manner in which he or she experiences the loss of a significant other. At the time of this clinical case, the DSM-IV was the diagnostic gold standard. Based on this case we will ask a number of questions:

Losing a “significant other” can severely disrupt a person’s life (Holmes & Rahe, 1967; Osterweis, 1984) and ranks as the most severe stressor rated on the Holmes and Rahe stress scale (Holmes & Rahe, 1967). Grief, the psychological reaction to such loss, is commonly considered normative, with stages that have been described in the literature (Bowlby, 1961, 1980; Kübler-Ross, 1969; Maciejewski, Zhang, Block, & Prigerson, 2007; Parkes, 1972; Parkes 1983). Normal grief, the most common reaction to the loss of a significant other (Maciejewski et al., 2007; Prigerson, 2005), progresses from acute to integrated grief. Acute grief encompasses intense psychological suffering, a desire, usually, to re-encounter the deceased, repeated thoughts of the deceased, sadness, and a tendency to neglect daily life. Integrated grief embodies the acceptance of loss, decreased intensity of psychic suffering when remembering the deceased, and a recommitment to life (Kendler, Myers, & Zisook, 2008; Zisook & Kendler, 2007; Zisook, Paulus, Shuchter, & Judd, 1997; Zisook, Schneider, & Shuchter, 1990; Zisook & Shear, 2009; Zisook, Shear, & Kendler, 2007; Zisook & Shuchter, 1991b).

Traditionally, five successive stages have characterized the evolution of the normal grieving process: denial, anger, bargaining, depression, and acceptance (Johannesson et al., 2009; Kübler-Ross, 1969; K. M. Shear et al., 2006; van Doorn et al., 1998; Wiese et al., 2010) complementing an earlier, four-stage model defined by numbness, yearning, disorganization, and reorganization (Bornstein, Clayton, Halikas, Maurice, & Robins, 1973; Bowlby, 1961, 1980; P. H. G. Parkes C.M., 1972; W. R. S. Parkes C.M., 1983). A prospective research studied the five stages of grief. the research showed that during the six months after a loss, in disbelief over the loss of a loved one gradually decreases and an acceptance of the loss increases. Yearning, anger, and depression peak four, five, and six months respectively after the loss, after which there begins a decline in these stages. After six months, the intensity of all these stages continues to diminish, and the individuals settles into acceptance. Consequently, the study authors maintained a six-month cut-off point, after which persistence of the above indicators show a need for evaluation for abnormal grief (Maciejewski et al., 2007). Other prospective studies have shown that after the loss of a significant other depressive symptoms frequently appear, and these could temporarily meet criteria for MDD if the DSM-IV restrictions are ignored (Maciejewski et al., 2007). Such symptoms decrease spontaneously for most people with the passage of time and without requiring professional assistance (Prigerson, 2005).

Risk factors for CG have been extensively studied. The literature proposes different risk factors associated with complicated grief, outlined in the Table 1.

The literature considers additional factors concerning the abnormal evolution of grief that partly relate to the bereaved person’s emotional, cognitive, and behavioral reactions to the loss. These factors concern the here and now of the grieving process and the manner in which the bereaved coped with the loss. This has been mostly discussed by IPT authors. The left-hand column of Table 2 details these factors, complementing the risk factors in Table 1. Some factors concern the individual’s grieving experience in the present moment, which the patient can address with the therapist’s aid, rather than risk factors moderating but not inherent in the grieving process. Important interpersonal factors directly linked to the grieving process include: the inability to feel or having difficulty in dealing with emotions linked to the loss (item 2); avoidance of rituals, ceremonies, and other activities that facilitate experiencing bereavement in its social and symbolic dimensions (item 3); the tendency to preserve the environment as it was before the death (item 6); and absence of family or social support during the grieving process (item 7). All four factors can hold back the bereaved from returning to the environment and accepting the new reality of life without the deceased. Clinicians can quickly identify these factors during the grieving process and can address them with the patient through IPT.

Although the DSM-IV did not recognize CG as an entity, the literature continues to debates its signs and symptoms, treatment, prognosis, and comorbidities. Initially, the DSM-5 proposed a CG diagnosis as a full nosological entity, but ultimately relegated it to the category of “Condition for Future Study” under the name “Persistent Complex Bereavement Disorder” (APA, 1994, 2014). According to the litterature, approximately 10 percent of grieving individuals appear to develop CG (Middleton, Raphael, Burnett, & Martinek, 1998; Prigerson et al., 2009). Complicated grief is associated with sleep disturbance (Germain, Caroff, Buysse, & Shear, 2005; Hardison, Neimeyer, & Lichstein, 2005; McDermott et al., 1997), social-professional dysfunction (Boelen & Prigerson, 2007; Simon et al., 2005; Simon et al., 2007), suicidal ideas and behaviors (Latham & Prigerson, 2004; Szanto et al., 2006), addictive behavior (Middleton et al., 1998; Zisook S, 1990), and higher general mortality (Helsing & Szklo, 1981; Stroebe, Schut, & Stroebe, 2007).

The DSM-IV precluded making a MDD diagnosis when depressive symptoms were related to the grieving process. It allowed an exception to this rule, however, if the depressive symptoms persisted more than two months after the loss or if the symptoms included guilt not related to a person’s regret over what he or she could/should have done to save the deceased, morbid ideas of uselessness, significant psychomotor retardation, substantive social dysfunction; or psychotic symptoms other than hearing the deceased person’s voice or briefly seeing an image of the deceased (APA, 1994).

The DSM-5 removed the grieving exclusion from MDD diagnosis but added a footnote explaining differences between GRMD and grief, trusting the physician to distinguish between them based on clinical common sense. Supporters of this modification emphasized that bereavement is a major stressor, and, like other stressors, could produce a depressive state similar to MDD in non-grief contexts (Kendler et al., 2008; Lamb, Pies, & Zisook, 2010; Zisook et al., 2012; Zisook & Kendler, 2007; Zisook et al., 1997; Zisook et al., 1990; Zisook & Shear, 2009; Zisook et al., 2007; Zisook & Shuchter, 1991b). Those who supported the modification stressed the potentially important consequences of abnormal grief in morbidity/mortality (Chen et al., 1999; Kaprio, Koskenvuo, & Rita, 1987; Osterweis M., 1984), increased use of alcohol and other substances (Zisook S, 1990), as well as a suicidal state and social-professional dysfunction (Hensley P.L., 2008; Kendler et al., 2008; Zisook & Kendler, 2007; Zisook et al., 2007). Moreover, the literature suggested that 20% to 40% of bereaved persons met criteria for MDD during the first two months after the loss, and 16% met the criteria a year thereafter (P. Clayton et al., 1968; Zisook, Shuchter, Sledge, Paulus, & Judd, 1994). Other authors observed a 9% prevalence of MDD four months after the loss (Barry, Kasl, & Prigerson, 2002).

Based on the literature—and as the DSM-5 footnote explains—though grief and MDD symptoms overlap, the experience of grief seems to differ from depression. In a normal grief process, bereaved individuals combine depressive humor and gloomy ideas with positive memories and positive feelings about the deceased, whereas in MDD humor is sparse and ideas remain consistently gloomy. Furthermore, self-esteem seems preserved during grief, whereas feelings of worthlessness and self-loathing are common in MDD (Zisook et al., 2010). Table 3 outlines GRMD and Grief symptom profile differences according to the DSM-5.

Although some bereaved individuals express significant suffering that fulfills the MDD criteria, and others express at risk behavior compromising their health, the DSM-IV separated GRMD from MDD resulting from any other stressor, and limited the possibilities for treating it before two months after the loss. On the other hand, the DSM-5 considers GRMD in the same way as MDD related to any other stressor, but emphasizes the necessity of distinguishing it from a normal grief process by explaining differences between the two entities. This offers the possibility of treating the bereaved, in the case of GRMD, during the first weeks of the grieving process, but does not clearly explain when and how normal grief becomes GRMD. It fails to consider the individual’s personal risk factors that moderate vulnerability to developing abnormal grief. DSM-5 also ignores the personal capacity of an individual to cope with loss based on emotional, interpersonal and social ressources. It does not encourage the physician to identify the individual risk factors for abnormal grief based on the patient’s personal history, nature of the loss, and type of relationship with the deceased. Thus it asks the physician to distinguish grief from GRMD based on general symptom profile differences. Consequently, it remains a general, symptom-based approach to grief tht is not patient centered. Although risk factors for abnormal grief have been studied, neither of the approaches in the DSM-IV or DSM-5 recommend that medical practitioners consider them when deciding whether or not to treat the bereaved. Yet, evidence indicates the need to distinguish clinically between individuals with and without risk factors (Table 1). Without considering such individual clinical specificities, a practitioner cannot offer a patient-centered approach to treatment. We shall discuss the above challenges through a clinical case presentation.

Mr. A, a retired, 65-year-old man of the Christian faith, was referred for suspicion of depressive symptoms five weeks after the death of his ex-wife. His chief complaint was “sadness and loss of interest in life.” Mr. A was retired, did not have a psychiatric history, and suffered from insulin-dependent diabetes. Mr. A faced an unbearable feeling of emptiness. He had settled into his late ex-wife’s apartment, felt sad, progressively isolated himself, lost his appetite, and began taking his diabetic treatment irregularly. He retrieved his ex-wife’s ashes, created a “chapel” in the living room, and passed several hours each day meditating beside the ashes. Despite increasing isolation, he had managed to perform his responsibilities and social activities. Two weeks later, he was hospitalized on an internal medicine ward for general impairment and complications of diabetes. Following his hospitalization, which was five weeks after his ex-wife’s death, he spoke with the psychiatric consultant at the hospital. He expressed guilt at not having done enough for his ex-wife while she was alive. All such ideas were directed toward the deceased. However, he denied gloomy or suicidal ideas.

Mr. A had been divorced from Ms. B for 19 years and was the father of an 18-year old son with whom he was on very good terms. Mr. A did not know his own father. His mother died when he was five years old. His grandmother brought him up until he was nine years old; following her death, a neighbor raised him. He has two older brothers. When 25 years old, he married a woman of the same age (Ms. B) and together they managed a restaurant that they purchased. After a few years, his wife became pregnant but he asked her to have an abortion because he needed her at the restaurant. She agreed. Their shared professional life was difficult to manage. The conflicts that arose motivated Mr. A to have extramarital affairs, including with the woman with whom he had a child, his only son. This affair precipitated his divorce. His difficult relationship with his child’s mother soon led to their separation and her departure. He gained full custody of his son, whom his ex-wife, Ms. B, agreed to love and help raise. Mr. A never remarried. During the previous two years, he had passed most of his time in the hospital caring for his ex-wife, who had become paralyzed following a stroke. He visited the hospital five days a week and shared at least one meal per visit with her. The hospital caregivers were impressed by his devotion to her and extolled his faithfulness. Motivated by guilt for having pushed his ex-wife to the abortion and for having caused their divorce through his affair, he derived the principal meaningful role in his life by visiting the hospital every day and caring for Ms. B.

When Ms. B died, Mr. A reported not feeling much emotion, despite his having lost his most important role and the most significant person in his life. He organized her funeral and described himself as being more active than usual.

At first contact with the psychiatrist his symptoms included sadness, loss of interest and pleasure, fatigue, poor concentration, sleep disorder, loss of appetite, the loss of four kilograms (almost nine pounds) within five weeks, mild psychomotor retardation, and guilty feelings toward his ex-wife. He denied gloomy or suicidal ideas. Had the DSM-5 been established at that time, Mr. A could have received the diagnosis of MDD with bereavement as stressor. From the total of 9 he fulfilled eight DSM-IV MDD criteria, including both A criteria, five weeks after the death of his ex-wife.

Despite meeting most DSM-IV criteria for MMD, Mr. A could not technically receive this diagnosis. This was because his symptoms had not lasted longer than two months since the loss, he did not have major psychomotor retardation, and his social functioning was not significantly disturbed. His guilt was solely directed toward the deceased, and he denied suicidal ideation and psychotic symptoms. His Hamilton Depression Rating Scale (Ham-D, 17 items) scale was 14 at the initial session, consistent with mild to moderate depression (Hamilton, 1960).

Despite the diagnostic technicality, the psychiatrist decided to treat Mr. A with IPT because he, recognized Mr. A had multiple risk factors for abnormal grief (Table 1), significant distress, and a grieving process showing evidence of abnormal grief (Table 2). Although the patient’s symptoms had lasted less than two months, the psychiatrist proposed psychotherapeutic treatment to the patient, who agreed. Mr. A had suffered early losses of significant others (his mother and grandmother), and his ex-wife had died from a difficult physical ailment. He lacked a supportive social environment, and was living in his ex-wife’s apartment, preserving the environment as it was before she died.

Interpersonal psychotherapy (Klerman G, 1984; Weissman M, 2000) is a time-limited therapy (12 to 16 weekly sessions) recognized in numerous treatment guidelines as efficacious for treating MDD (Cuijpers, Dekker, Hollon, & Andersson, 2009; Cuijpers et al., 2011; also see the Practice guideline for the treatment of patients with major depressive disorder (revision). American Psychiatric Association, 2000). In IPT there is recognition of the close link between the nature of the interpersonal relationship with the deceased and depressive symptoms (Brown G.W, 1978; Brown, Harris, & Peto, 1973; Coyne, 1976; Henderson, Byrne, Duncan-Jones, Scott, & Adcock, 1980), WithIPT death is seen as a troubling life event that may push a vulnerable individual into a major depressive episode. The death may have varying interpersonal consequences, including loss of a crucial social support or guilt caused by patient’s hatred for someone who has died. Interpersonal psychotherapy techniques facilitate the expression and understanding of grief, and these help the patient to reestablish interests and relationships in substitution for the lost relationship (Weissman, 2000). As Table 2 shows, IPT identifies seven signs of abnormal grief.

According to IPT, elements in Mr. A’s grief process indicated an abnormal evolution of grief. During the days after his ex-wife’s death, and particularly during the funeral, Mr. A functioned in a hyperactive state without registering or expressing the emotions normally associated with the loss of a loved one. Thereafter, Mr. A increasinglyisolated himself socially, settled into his ex-wife’s apartment, and refused to alter the “chapel” dedicated to his ex-wife, where he passed more and more time “in the presence.” Among the ther risk factors present for possible abnormal grief were based on history, including (1) Ms. B’s death from a serious disease that caused major disabilities requiring months of hospitalization, and (2) Mr. A’s childhood marked by repeated early bereavements including the death of his mother when he was five years old and the death of his grandmother (who had raised him after the death of his mother) when he was nine.

During the 14 IPT sessions with Mr. A, all of the above factors were actively identified and addressed. Mr. A expressed emotions that he had not experienced during the first weeks of his mourning. Much sadness arose, and he cried for his ex-wife. He described feeling lonely after her death, and his need to be with her. Reviewing the history of his relationship with Ms. B, he was able to stand back, look at it in a more nuanced manner, and contextualize the guilt he felt toward her. Encouraged by his therapist, Mr. A found another place for her ashes, and renovated her apartment by repainting it, changing the carpet, and replacing the furniture. To increase his social support, Mr. A started to attend his local church, joined social groups, and made new friends. He gradually recovered his good mood and ended the psychotherapeutic treatment. His Ham-D score was 3 at week 12, consistent with remission. Although still grieving for Ms. B (as would be expected), he felt more comfortable in doing so. His diabetes improved because of better adherence to treatment.

The case of Mr. A demonstrates that when aspects of the bereavement process in an individual’s history raise the possibility of the abnormal bereavement, therapeutic intervention can address it. This raises a question: Had Mr. A’s grieving been allowed to fester longer, would it have worsened and proved harder to ameliorate? Based on the DSM-5, one can discuss the need or lack thereof for offering him treatment. Mr. A had had depressive symptoms for five weeks. His depressed mood was associated with thoughts and reminders of the deceased. Mr. A could remember positive moments he had shared with Ms. B and feel positive emotions. However, negative thoughts and emotions predominated. His thought content was marked by preoccupation with memories of the deceased. He tended to blame himself because of what he had not had the opportunity to do for the deceased, and because of what he had done that had harmed her.

Based on the DSM-5 footnote, even if Mr. A suffered from depressive symptoms for a few weeks, these symptoms most likely fullfilled criteria for grief than GRMD. On the other hand, one could interpret these symptoms as evidence of GRMD and offer treatment.

This single case obviously does not provide sufficient evidence to modify the diagnostic criteria for MDD to remove the bereavement exclusion. By confusing grief and MDD there is a risk of denying the “natural” character of bereavement suffering and pathologizing a normal life event. The DSM-5 offers the possibility of earlier and potentially more systematic intervention by allowing diagnosis of GRMD two weeks post bereavement. However, while DSM-5 emphasizes the need to distinguish grief from MDD, it does not clearly discuss when and how grief becomes depression. As no one may know the answers to these questions, would it not have been better for DSM-5 to discuss the personal capacity of the individual to cope with the bereavement, instead of focussing on symptoms to explain differences between grief and GRMD? Two weeks post loss, the physician must decide to treat or wait based on clinical common sense. Even though the DSM-5 stresses the need to consider the personal history of the patient, it does not clearly discuss risk factors exposing an individual to abnormal grief. Although most bereaved individuals will not develop abnormal grief, there is a need to detect patients more vulnerable to it. Research on this issue could occur on two levels: studying risk factors for abnormal evolution of grief (Table 1), and level of interpersonal dysfunction (Table 2), both of which negatively affect the evolution of the grief.

The debate over eliminating the bereavement exclusion from the MDD diagnostic criteria makes the factors in Table 2 more interesting, as these factors affect the bereaved individual in the here and now and, if not corrected, could result in abnormal bereavement. Inadequate grieving during the bereavement period, avoidance of thoughts of death, enshrining the environment as it was when the significant other died, and absence of family or other social supports could all be detected early in an individual’s grieving period. These are factors IPT actively works to correct in MDD. Interpersonal psychotehrapy postulates that correcting such dysfunctions helps to treat GRMD and to restore mood. Interpersonal psychtoehrapy tries to facilitate the patient’s grieving process by giving the patient the possibility to express emotions relevant to the loss and understand why it feels like such a loss. It emphasizes the importance of experiencing the reality of the loss and suffering it rather than avoiding it. In the next stage it encourages the patient to accept the reality of life without the loved one.

Having observed the cathartic effect of expressing emotions linked to the loss, IPT encourages the patient to step back from the loss and revisit his relationship with the deceased to have a more nuanced view of the lost relationship and individual, positive and negative. The patient’s view of his or her relationship with the deceased becomes less polarized, closer to reality. Interperonsal psychotherapy encourages the patient to change his or her environment and to adjust to the new life. Moreover, it actively works with the patient to recreate social ties and to decrease social isolation. When an individual presents such signs and symptoms during grief, should therapists only consider the international classification recommendations to decide whether to intervene? Is there not a better chance to correct the evolution of the grief through early intervention if the patient manifesting risk factors, presents in addition a history revealing signs, as described by IPT, of a abnormal evolution of grief? We believe more research may be needed to test this on targeted, high risk individuals who express general and interpersonal risk factors of abnormal grief.

A patient-centered vision could potentially make our diagnostic and therapeutic approach more nuanced. It is important to consider a bereaved patient’s individual history and risk factors likely to expose the patient to CG or GRMD. The nature of the bereavement event (for example, traumatic or non-traumatic) and the nature of the patient’s relationship with the deceased (for example, type of interpersonal attachment) are also factors to weigh. As IPT shows, how a person experiences bereavement in the interpersonal context is a major factor affecting, positively or negatively, the evolution of grief.

Today we attempt to detect psychopathology in a prodromal state to prevent its aggravation, improve prognosis, and reduce health care and other costs. This is true, for example, in programs attempting to identify and manage prodromal psychosis, or programs for patients at high risk for postpartum depression or PTSD. We know that CG and GRMD can increase morbidity, mortality, and consumption of health care services. Knowing whether a person is likely to develop abnormal grief, and intervening with a brief therapy, would allow a more patient-centered approach, responsive to the patient’s specific history and characteristics. Impact, efficacy, and cost-effectiveness of such intervention require validation by further research.