The loss of a loved one is one of the most traumatic events in a person’s life. In spite of this, most people cope with the loss with minimal morbidity. Approximately 2.5 million people die in the United States every year, and each leaves behind about 5 bereaved people.1,2 By age 65, more than half of American women and 10% of American men have been widowed at least once.3 To better understand the usual course of coping with such losses, it helps to define some basic terms and become familiar with the experiences that individuals move through following a loss.
Bereavement is the reaction to a loss by death. Grief is the psychological and emotional reaction to a significant loss, not limited to death. Mourning is the social expression of bereavement or grief, often formalized by custom or religion. Complicated grief, previously referred to as traumatic grief, is typically not diagnosed until at least 6 months have passed after the loss. It is a distressing condition in which common symptoms include disbelief about the death, anger and bitterness, pangs of painful emotions with intense yearning for the deceased, preoccupation with thoughts of the deceased, and avoidance of reminders of the loss.2
There is some literature on the typical stages of grief. For example, Clayton4 defined 3 stages:
• The first, numbness, typically lasts a few hours to a few days; essential things get done, but most are poorly remembered; anxiety and depressive symptoms may begin.
• The second, depression, lasts a few weeks to a year; irritability and restlessness are prominent, but all depressive symptoms commonly occur.
• The third, recovery, often begins within 4 months after the loss of a loved one; the bereaved accepts the death and returns to some earlier level of functioning.
In reality, little empirical evidence exists to support the stage theory of grief, and it is a mistake to think that all individuals move through such stages in an orderly way.
GRIEF COURSE PATTERNS
In a review of grief course patterns, Bonanno and Kaltman5 found that in the first year of bereavement, most people demonstrate 4 types of disrupted functioning: cognitive disorganization, dysphoria, health deficits, and disruptions in both social and occupational functioning. In the group of cognitive changes, they include difficulty in accepting the loss, a sense of loss of part of oneself, uncertainty about the future, and a search for meaning. In most bereaved persons, these 4 domains of disrupted functioning are common in the first few months of bereavement and generally decline during the first year.
Many longitudinal follow-up studies of the recently bereaved have focused on the widowed. In a study by Clayton,6 many persons reported crying, low mood, sleep problems, loss of appetite, restlessness, fatigue, poor memory, and loss of interest at 1 month. Psychomotor retardation is rare in bereaved persons, but restlessness and anxiety are common.6 By the end of a year, many vegetative symptoms had improved, but insomnia, restlessness, and periodic low mood tended to persist. Overall, somatic symptoms (such as weight loss and fatigue) tended to improve gradually while the psychological symptoms (such as feelings of hopelessness and worthlessness) tended to persist.
At 1 month after the loss, about 40% of the bereaved meet criteria for a major depressive episode.7 However, DSM-IV-TR lists bereavement as a V code and recommends not diagnosing depression related to bereavement until at least 2 months have passed after the loss. This convention is not supported by all researchers, and some argue that bereavement should not be singled out as virtually the only life stressor that excludes a diagnosis of depression.8 The current DSM lists several symptoms that are not typical in a normal grief reaction and may help differentiate a major depressive episode from bereavement. These are extreme guilt, suicidal ideation, severe feelings of worthlessness, marked psychomotor retardation, severe and continued functional impairment, and persistent hallucinations.9
At 2 months, about 24% of the bereaved meet criteria for major depression, at 1 year after the death, about 15% of the bereaved are depressed, and at 2 years, the rate of depression is about 7%.6,7,10,11 In addition to those who met the full criteria for major depression, Zisook and colleagues11 found rates of subsyndromal depression (either dysphoria or loss of interest but no more than 3 additional depression criteria) of 27% at 2 months, 19% at 13 months, and 12% at 25 months.
Several reports have focused on the elderly. In the Cache County study, the point prevalence and lifetime prevalence of major depression were estimated in men to be 2.7% and 9.6%, respectively, and in women, 4.4% and 20.4%, respectively.12 The point prevalence of bereavement-related major depression was 0.3% in men and 0.5% in women. In a study of the relationship between depression and the loss of a spouse in the elderly (70 years and older), there was a strong association between bereavement and depression in both men and women.13 The newly bereaved had an almost 9-fold greater risk of syndromal depression compared with married persons. Rates of depression in the long-term widowed group were much lower and similar to rates in the married group.13
The best predictor of the development of a chronic depressive syndrome is poor physical or mental health before the loss. Clayton6 found that factors such as relationship to the deceased, degree of closeness of that relationship, sex, age, race, social support, family history of a mood disorder, financial status, and religious involvement do not predict long-term outcome. Poorer immediate response, however, is associated with younger age, lack of social support, and low income. Rates of bereavement depression are similar in men and women, unlike major depression, which has a 2-fold greater prevalence in women. Zisook and Shuchter14 report that a personal history of major depression, a family history of major depression, and alcohol(Drug information on alcohol) consumption in the first 2 months after the loss are associated with depression secondary to bereavement.
Mortality associated with bereavement has been a somewhat controversial area. A review by Stroebe and colleagues15 updates their previous work and focuses on longitudinal studies published since 1993. They note that most studies report that widowers are at greater risk for death than widows when both are compared with married same-sex counterparts. Also, mortality risk seems to be greater for younger than for older individuals who have lost a spouse, and those bereaved for a short duration are at greater mortality risk than those bereaved for a longer time.
Aside from depression, there have been occasional reports of manic symptoms related to bereavement, hence the term “merry widow.” Such a switch to mania or hypomania could occur in a person with bipolar-spectrum illness who may not have had a previous episode of depression. The work on sleep deprivation by Wehr16,17 has repeatedly demonstrated the risk of switch into mania. Columbo and colleagues18 found that the switch rate for antidepressants was comparable to that of sleep deprivation in bipolar depression; the rate of switch to mania was 4.85% and the rate of switch to hypomania was 5.83%.
Mr D is a 55-year-old widower who lost his wife to an aggressive form of breast cancer 6 months ago. He visits his primary care physician for his annual examination and reports that he is doing fairly well in spite of his loss. He praises his family for being supportive and underscores the love he has for his 2 adolescent children. Following his wife’s death, he had episodes of crying, difficulty in falling asleep, feeling tired during the day in spite of adequate hours of sleep, and loss of appetite. He says he lost about 10 lb in the first 2 months. He reports that his crying has gradually diminished and he can think about his wife without becoming overwhelmed with emotion. He still has trouble sleeping, but he has regained the weight he lost and has begun an exercise and weight training program at a local gym. He has also started attending a bereavement support group through a local hospital. He still feels somewhat overwhelmed raising 2 children alone and seems to feel that he will not be a good parent.
Because this is an example of normal bereavement, no psychiatric intervention is necessary. Mr D should continue to see his primary care physician as needed, and he should continue going to his support group.
TREATMENT OF BEREAVEMENT-RELATED DEPRESSION
Our group has primarily focused on the psychopharmacological treatment of bereavement-related major depression, but it is helpful to say a few words about psychotherapy for bereavement and grief. Husaini and colleagues19 have shown the effectiveness of group therapy in reducing depressive symptoms in elderly women. Although additional studies are needed for the use of group therapy in bereavement-related depression (BRD), it offers promise.
Three comprehensive reviews provide important insight into the effectiveness of therapy and counseling for bereavement and grief. The first is a meta-analysis based on 35 studies of grief therapy.20 Findings from the analysis indicate a moderate effect size for intervention, and it was noted that studies with self-selected participants had the highest effect sizes.20
The second examined 4 qualitative and quantitative reviews and concluded that formal psychotherapeutic interventions for bereavement demonstrate low efficacy.21 The investigators recommend that clinicians focus interventions on high-risk mourners, including men who are widowed (especially if older and isolated), mothers who have lost a child, people who survive sudden and/or violent traumatic losses, those with previous psychiatric problems, and those with “high-distress grief,” including individuals with complicated grief, early in bereavement.
The final review examined various forms of intervention for the bereaved.22 Aside from good evidence supporting pharmacological treatment of BRD, other forms of intervention (such as support groups/ counseling and psychotherapy-based treatments) designed to diminish grief showed no consistent pattern of benefit.
Most reports on pharmacological treatment in this field target BRD. Several open-label studies, one double-blind parallel group study, and one double-blind, placebo-controlled trial using antidepressant therapy and psychotherapy have been reported. In these trials, patients with BRD have generally shown improvement in both depressive and grief symptoms, although the improvement in symptoms of grief is often less than that seen with depression. An additional controlled trial specifically examined use of a benzodiazepine. Several reports have focused on treatment of complicated grief.
Totaling treatment of just 55 participants, the demographics and results of the 3 open-label studies and one parallel group study are summarized in the Table. Most of the treated patients were older than 60 years. All were grieving the loss of a spouse. The open-label study undertaken by Jacobs and colleagues23 used a tricyclic antidepressant (desipramine for 9 patients, nortriptyline(Drug information on nortriptyline) for 1). Their results showed that depressive symptoms diminished significantly in 7 of 10 patients and that grief symptoms improved in 7 of 8 patients, but to a large degree in only 3 of the 8.
The second open-label trial used nortriptyline and treated patients until they achieved a response, defined as a Hamilton Depression Rating Scale (HAM-D) score of 10 or less for 3 or more weeks and a therapeutic blood level of medication.24 The results showed that mean depression ratings improved by 67% to 68%, and mean grief ratings improved by 9%.
In a third open-label trial, Zisook and colleagues25 treated participants early in bereavement with bupropion SR. In their sample of completers, 86% were responders on the HAM-D, mean depressive ratings fell by 73%, and mean grief ratings on 1 of 2 scales decreased by 22%. In a double-blind trial of both nortriptyline and sertraline(Drug information on sertraline), results were pooled for the 2 medications.26 The investigators found that depression ratings improved by 57% for the sample; they did not measure grief independently.
Warner and colleagues27 evaluated the use of a low-dose benzodiazepine (diazepam, 2 mg/d) within 2 weeks of spousal bereavement in a randomized, double-blind, placebo-controlled, parallel-group trial. Depression was not targeted in this study, and the investigators found no evidence that diazepam(Drug information on diazepam) affected the course of bereavement. In fact, the individuals randomized to placebo showed more ease in getting to sleep quickly and had fewer bad dreams than the benzodiazepine group.
The largest published study to date involving medication treatment of BRD used both interpersonal psychotherapy (IPT) and pharmacotherapy and included individuals grieving the loss of a spouse, child, parent, or sibling.28 IPT is a brief manual-based psychotherapy that focuses on interpersonal issues in the treatment of depression. It addresses 4 problem areas of interpersonal interaction: grief, role transition, interpersonal conflict, and interpersonal deficit. In the study, 22 men and 58 women, most of whom were in their 60s, were randomized to 16 weeks of treatment with 1 of 4 treatments. Those individuals who did not achieve at least a 50% reduction in HAM-D score by week 8 of double-blind treatment were considered nonresponders and were switched to open treatment. Numbers of patients and rates of remission (defined as a HAM-D rating of 7 or less for 3 consecutive weeks) and relapse are shown in the Table. Statistical analysis revealed a significant drug effect when examining for remission but no main effect of IPT. Rates of relapse in 16 weeks of continuation therapy were low across all conditions, and rates of decline in bereavement intensity were not different across the 4 treatments. Overall, grief ratings declined by 23% over 16 weeks of treatment.
Mrs R is a 63-year-old who lost her mother. Immediately after the loss, she started to have problems with restlessness, difficulty falling and staying asleep, depressed mood, loss of interest in her family, fatigue, guilty feelings about her past relationship with her mother, and low appetite. These symptoms have persisted over the past 4 months. She is now also having difficulty concentrating at work and reports little interest in her career. She is looking forward to the summer because she works in a school and will have several months off. However, she worries that she will spend her days in bed trying to nap because she is still sleeping poorly.
Mrs R presented for treatment of BRD in an outpatient clinical setting. We initially prescribed fluoxetine(Drug information on fluoxetine), but she had difficulty with it because of GI adverse effects. We switched her medication to paroxetine(Drug information on paroxetine), 20 mg/d, which she tolerated well. She gradually started sleeping and eating better. Her energy improved as well. Her husband noticed improvement in her mood next, but it took 2 more weeks for her to acknowledge mood improvement. By the end of 7 weeks, she felt that her depression was essentially gone, but she still talked about some guilt feelings related to an argument she had had with her mother over her mother’s financial support of Mrs R’s younger, irresponsible brother. She continued medication for 1 year, and then asked to be tapered off. When her symptoms of depressed mood and poor sleep started to return, she resumed her usual dosage of medication, and she continues to do well 2 years later.
TREATMENT OF COMPLICATED GRIEF
About 10% to 20% of bereaved persons have complicated grief; common co-occurring psychiatric disorders are major depression and posttraumatic stress disorder (PTSD).2 Published
reports of treatment for complicated grief have included both medication and psychotherapy. Zygmont and colleagues29 treated 4 men and 11 wom-en who had complicated grief (then called traumatic grief) with paroxetine plus a new form of psychotherapy that targets complicated grief symptoms. They compared this group with a group of 22 persons who received nortriptyline in the study by Reynolds and colleagues28 of BRD. In the paroxetine plus psychotherapy group, scores on the Inventory of Complicated Grief (ICG) scale fell by 53% and depression scores on the HAM-D fell by 54%. Comparable benefits were seen with nortriptyline, but detailed results were not reported.
In a case series, 4 women with a primary diagnosis of complicated grief were treated with escitalopram(Drug information on escitalopram) for 10 weeks.30 Three of the 4 women also had PTSD, all had a history of depression, and 2 were currently depressed. Mean ratings of complicated grief on the ICG fell by 76%, and scores on the HAM-D decreased by 75%.
A psychotherapy targeting complicated grief, called complicated grief treatment (CGT), combines IPT techniques with cognitive-behavioral therapy—based techniques to address trauma. Much of the work involves retelling the story of the death and confronting avoided situations. In a randomized controlled trial by Shear and colleagues,2 patients received either CGT or IPT for about 16 weeks. Almost half of the patients were concurrently taking antidepressant medication. Results of the trial showed that CGT outperformed IPT in treating complicated grief, with response rates of 51% and 28%, respectively (P = .02). Completion rates for treatment were 73% for CGT and 74% for IPT. Complicated grief therapy seems to be an effective and specific treatment option.
Most people who experience the loss of a loved one will recover gradually without intervention. About 15% of the bereaved are depressed at 1 year following the loss. For those who become depressed and/or experience complicated grief, intervention is critical. The strongest data for treatment of BRD are from studies of antidepressant therapy. Symptoms of depression seem to improve more than symptoms of grief in these pharmacological studies. CGT is the best supported treatment for complicated grief. Additional preliminary data support the use of antidepressant therapy in complicated grief as well. This field of study requires more randomized placebo-controlled trials to further define optimal timing and length of intervention.