Prolonged Grief Disorder: Course, Diagnosis, Assessment, and Treatment

Prolonged grief disorder was only recently included as a codable diagnosis in the World Health Organization’s ICD-11 (29); it has recently been approved but not yet published by the American Psychiatric Association for the revision of the DSM-5 (the most recent available diagnostic criteria for ICD-11 and DSM-5 can be found in Table 1) (30). However, several proposed criteria sets for prolonged grief conditions, labeled pathological grief (31, 32), prolonged grief (10), and complicated grief (11), have existed before these recent additions. The first criteria set for a prolonged grief diagnosis in the DSM-5 was persistent complex bereavement disorder (33), included as a provisional diagnosis for further study.

The most recent diagnostic criteria sets for prolonged grief disorder have some similarities and some differences, which are largely reflective of the less proscribed nature of ICD versus DSM criteria in general. Otherwise, the most notable difference is the timeframe criterion, as prolonged grief disorder in the DSM-5 requires the death to have occurred at least 12 months before diagnosis as opposed to 6 months prior. Both criteria sets include gateway symptoms of yearning or preoccupation with thoughts or memories of the deceased. The DSM-5 criteria set requires at least three other symptoms occurring nearly every day for at least the past month. The DSM-5 diagnosis of prolonged grief disorder will likely remain in the Trauma and Other Stressor-Related Disorders section, like the provisional persistent complex bereavement disorder diagnosis.

Common differential diagnoses to consider when evaluating the potential presence of prolonged grief include major depressive disorder, PTSD, and normative or acute grief. Although initial postloss reactions commonly include intense emotions, these emotions do not, for the vast majority of the bereaved, represent the early development of prolonged grief. Acute grief following a loss, particularly the death of a loved one with a strong attachment, is normative and should not be pathologized. Typical signs of acute grief, some of which overlap with prolonged grief disorder, include feelings of shock or disbelief, yearning, waves of sadness or other intense emotions, feeling disconnected from others, and desire to disengage from roles or responsibilities (12, 34). Of note, in the DSM-5, bereavement was no longer listed as exclusionary for assignment of a major depressive disorder diagnosis; however, this diagnosis should be carefully assessed and differentiated from acute grief.

During the acute grief period, bereaved individuals benefit from compassionate support, stabilization of physiological hyperarousal (e.g., sleep interventions), connection to social support systems, and space to oscillate between pain and respite, with encouragement of self-care activities (35). Because prolonged grief disorder cannot be formally diagnosed until 6–12 months after a death, ongoing assessment of an acute grief reaction for those with elevated risk for developing prolonged grief disorder (e.g., preexisting psychiatric history) is recommended; in addition, awareness is needed of the possibility of anniversary reactions, with potential brief heightened grief intensity at important dates, including the 12-month mark. Nonetheless, individuals presenting to clinicians with acute grief may also have other negative outcomes associated with the death; these individuals should be carefully assessed for other common mental health conditions after a major life stressor, such as adjustment disorders, major depressive disorder, PTSD, and elevated risk of suicide (15, 36). As sufficient time passes (at least 6 months in ICD-11 and 12 months in DSM-5) such that prolonged grief disorder may be present, these same conditions, as well as alcohol and substance use disorders, remain the most common to consider in evaluating bereaved individuals.

Major depressive disorder has several overlapping characteristics with prolonged grief disorder, including sadness, crying, social isolation and withdrawal, sleep disturbance, worthlessness and guilt, and suicidal ideation. However, they have been demonstrated to be distinct syndromes despite some overlap in worthlessness and guilt as well as suicidal ideation (37). Typically, sadness, tearfulness, and social isolation are more generalized in major depressive disorder, with evidence of a generalized anhedonia with low reactivity. In prolonged grief disorder, these symptoms are more specifically related to the loss of a close attachment and may include yearning for reunion with the deceased as well as proximity seeking; in addition, avoidance of painful reminders of the permanence of the death can contribute to isolation and functional impairment, such as avoidance of previously shared friends or activities when a spouse dies. This behavior is in contrast to the general withdrawal and low motivation associated with major depressive disorder. Although not diagnostic in prolonged grief disorder, both conditions can also be associated with suicidal ideation and behaviors (15, 37).

PTSD and prolonged grief disorder also have some overlapping features, whether or not a sudden violent or accidental loss occurred, instances that would be considered criterion A traumas within the PTSD diagnostic framework (38). In fact, some models of prolonged grief disorder have described the condition as a combination of traumatic distress and attachment loss (39–41). Overlapping features of PTSD and prolonged grief disorder may include intrusive thoughts or images of the death, avoidance related to the death, and emotional numbness. Although not diagnostic in prolonged grief disorder, both conditions can also be associated with sleep disturbance (42).

As both diagnoses are categorized as trauma and stressor-related disorders in the DSM-5, they have some degree of overlap (37). However, distinguishing features exist. For example, the predominant emotion in PTSD is fear, whereas yearning and sadness are more common in prolonged grief disorder (11, 43). The content of intrusive thoughts in PTSD is related to the traumatic event; in prolonged grief disorder, the content is related to the circumstances of the death, although this content may overlap in the case of a violent or accidental death. Avoidance in PTSD is typically related to safety concerns or reducing potential threat, whereas avoidance in prolonged grief disorder is related to avoiding painful reminders of the loss and its permanence. Furthermore, the presence of hyperarousal and hypervigilance has been more typically associated with PTSD, whereas challenges reengaging in life without the deceased (e.g., meaninglessness, wish to join the deceased, loneliness) are associated with prolonged grief disorder (37).

Psychiatric comorbid conditions are common for individuals with prolonged grief disorder and require careful assessment, especially given the aforementioned challenges with differential diagnosis driven in part by some symptom overlap (not unlike what is seen for other disorders triggered by trauma and stressors). For example, in a sample of 206 individuals with prolonged grief disorder, 75% were diagnosed as having a current comorbid psychiatric condition (44). In this sample, the most common comorbid conditions were major depressive disorder (55%), PTSD (48%), and anxiety disorders (62%), with generalized anxiety disorder being the most common anxiety disorder comorbidity (18%) (44).

It is perhaps not surprising that a significant proportion of comorbid conditions present among patients with prolonged grief disorder onset before the death, as these conditions may interfere with adaptation after a death. Indeed, population-based prospective studies of bereaved caregivers have indicated that preloss grief and depressive symptoms, as well as poorer preloss mental health scores on the 36-item Short Form Survey, predict development of prolonged grief disorder postloss (28, 45). Furthermore, studies have also found higher rates of prolonged grief conditions among bereaved samples presenting with major depressive disorder (25, 46), anxiety disorders (46, 47), and trauma-related disorders (46). Threshold levels of prolonged grief symptoms were also associated with poorer quality of life and greater functional impairment (25, 47), suggesting the need to screen patients presenting with mood, anxiety, or trauma-related disorders for exposure to a death and associated prolonged grief disorder. Furthermore, those with prolonged grief disorder may be at elevated risk for suicidal ideation and behaviors; therefore, suicide risk should be adequately assessed (15, 46).

Other comorbid conditions include alcohol and substance use. As with mood and anxiety disorders, those with preexisting substance misuse may be more vulnerable to developing prolonged grief disorder (47). The presence of prolonged grief disorder may also be linked to increases in smoking (48) or alcohol misuse for those with bipolar disorder (24) and major depressive disorder (25). A recent systematic review of 85 studies (N=12,294 participants) indicated that bereaved individuals are also at elevated risk for experiencing a variety of sleep impairments, including poorer sleep quality, problems falling asleep, shorter sleep duration, and anxious nighttime awakenings (42). Furthermore, it estimated that approximately 80% of individuals with prolonged grief conditions experienced long-term poor sleep (42).

Routine screening for prolonged grief disorder in psychiatric settings could improve familiarity with the disorder and accelerate treatment access for patients (56, 57). The Prolonged Grief Disorder-13 (PG-13) has been used in several studies to assess prolonged grief disorder diagnostic criteria (51, 58, 59). This 13-item, self-report measure examines three facets of prolonged grief symptoms: feelings, thoughts, and actions occurring over the past month (10). Initial questions query the frequency of separation distress, specifically yearning or intense feelings of emotional pain, which must have occurred most days over the past 6 months. Subsequent questions are about the extent of cognitive, emotional, and behavioral problems resulting from a loss; a final question assesses significant impairment due to prolonged grief disorder. The PG-13 questions align closely with the ICD-11 diagnostic criteria for prolonged grief disorder. The reliability of the PG-13 ranges from good to excellent (Cronbach’s α=0.84–0.92) (60–62) with demonstrated construct and criterion validity (60, 62).

If a shorter tool is needed, the Brief Grief Questionnaire (BGQ) is a five-item, self-report instrument designed to screen for prolonged grief disorder (49, 63). The BGQ has good internal consistency (Cronbach’s α=0.82) and discriminant validity, discriminating grief-related symptoms from general psychological distress (56, 57, 63). A BGQ score≥8 indicates probable prolonged grief disorder, and scores from 5 to 7 indicate subthreshold prolonged grief disorder (63).

The Inventory of Complicated Grief (ICG) (50) is a 19-item measure assessing severity of prolonged grief disorder symptoms; it has been widely used in treatment studies before the establishment of prolonged grief disorder as a diagnosis. The ICG has excellent internal consistency (Cronbach’s α=0.94) and good convergent validity with other grief symptom measures (50). The typical score threshold for inclusion in prolonged grief disorder research has been an ICG score of ≥30 (43, 64); however, recommended clinical cutoff can include scores≥25, suggesting a level of grief with significant social, mental, and physical impairments (50, 65).

Clinical evaluation of patients with potential prolonged grief disorder should include consideration of current and lifetime psychiatric disorders, suicidal ideation, mental status, and medical and treatment history (66). This information can be gathered by using a structured assessment, such as the Structured Clinical Interview for DSM-5 (SCID-5) (67) or the Mini-International Neuropsychiatric Interview for DSM-5 (MINI) (68).

However, because prolonged grief disorder was not yet finalized as a diagnosis in the DSM-5, standard assessments such as the SCID-5 and MINI do not yet include prolonged grief disorder (56). Clinicians must therefore assess the severity and intensity of prolonged grief disorder symptoms through separate inquiry. One option is the Structured Clinical Interview for Complicated Grief (SCI-CG) (52), a clinician-administered interview that assesses additional information necessary for a prolonged grief disorder diagnosis (56). The SCI-CG contains 31 symptom ratings about the presence of prolonged grief disorder symptoms over the past month and an optional screening section to assess characteristics related to the death, including relationship to the deceased, cause of death, and time since death (52). Some previous research has used the SCI-CG as a supplement to confirm prolonged grief disorder diagnosis for those above self-report threshold (e.g., 69). The SCI-CG has good to excellent psychometric properties, including interrater reliability (intraclass correlation coefficient [ICC]=0.95), internal consistency (Cronbach’s α=0.77), test-retest reliability (ICC=0.68), and convergent validity with other complicated grief symptom measures (52). However, these properties have been validated within just one study (70); overall, Bui et al. (52) found high acceptability of the SCI-CG, with clinicians reporting that they found the instrument easy to conduct and helpful for assessing patients with prolonged grief disorder.

Additional measures are also being used and developed, such as the Clinician-Administered Prolonged Grief Disorder Scale, and will likely further align with the diagnostic criteria as they become official in the revised DSM-5. ICD-11 criteria are more flexible and can also be applied in clinical practice.

Yearning is considered one of the gateway features of prolonged grief disorder diagnosis and may be integral in starting a cascade of other grief-related symptoms, including emotional pain, preoccupation with the deceased, and feeling life is meaningless without the deceased (37). The Yearning in Situations of Loss Scale–Bereaved (YSL) (53) is a 21-item, self-report measure assessing the extent to which a bereaved person is longing for the deceased. The YSL demonstrates strong psychometric properties, including internal consistency (Cronbach’s α=0.91) and appropriate discriminant validity with constructs such as depression, anxiety, and loneliness (53).

The existence of ruminative, maladaptive cognitions as well as avoidance behavior after a loss have also been shown to contribute to the onset of prolonged grief disorder (71–79). Continued avoidance of emotional pain, thoughts, or places associated with the deceased as well as dysfunctional beliefs may interfere with adaptation to the loss and thus promote the development and severity of prolonged grief disorder (54), impairment, and treatment response (28, 35). As these constructs are not fully captured by self-report or clinician-administered measures, specific self-report assessments, such as the Typical Beliefs Questionnaire (TBQ) (54) and the Grief-Related Avoidance Questionnaire (GRAQ) (55), may be administered in addition to the aforementioned measures.

The TBQ is a 25-item, self-report measure of maladaptive thinking patterns common among people with prolonged grief. Scoring can be done through simple summation of all rating scores or through dichotomous scoring, in which a score of “3” or “4” is considered an endorsed item. The TBQ has good psychometric properties, including internal consistency (Cronbach’s α=0.82) and test-retest reliability (ICC=0.74) (54). Factor analysis revealed five relevant factors: protesting the death, negative thoughts about the world, needing the person, less grief is wrong, and grieving too much (54). Of note, cause of death may affect themes of typical beliefs. For example, those bereaved by suicide may be more likely to endorse self-blame as well as feeling less connected to others and stigma-related beliefs (80).

The GRAQ is a 15-item, self-report measure assessing physical and mental avoidance behaviors of bereaved individuals (55). The GRAQ has shown good internal consistency (Cronbach’s α=0.87–0.89) (55, 81) and convergent validity, differentiating between those with and without anxiety, depression, and PTSD symptoms (81). Factor analysis identified three distinct factors: avoidance of places and things that are reminders of the death (reminders of loss), avoidance of activities that are reminders of the death (reminders of death), and avoidance of situations related to illness or death that evoke loss-adjacent emotions (sympathy situations; e.g., going to funerals) (55, 81). In a GRAQ validation study, loss of a child and death due to short illness resulted in the highest levels of grief-related avoidance compared with other relationships and causes of death (81).

CGT is a 16-session, manualized intervention, which has demonstrated efficacy for reducing prolonged grief disorder symptoms in several randomized controlled trials (RCTs) (18, 43, 69). CGT incorporates components of attachment theory, cognitive-behavioral therapy (CBT), and other approaches to facilitate natural adaptive processes to loss. CGT facilitates adaptation through focus on both loss and restoration, in alignment with the dual process model of grief (83). The loss-focused components include accepting the reality of the death and changing the relationship with the person who died. The restoration-focused elements include working toward aspirational goals in the absence of the individual who died and having a sense of competence and satisfaction in the world without the deceased (35). To meet the goal of adaptation, CGT includes seven core themes: providing information to help patients understand and accept grief, managing emotional pain and monitoring symptoms, thinking about the future, reconnecting with others, telling the story of the death, learning to live with reminders, and connecting with memories. Previous articles have described the essential tools discussed within each of these themes (35), and training in CGT is available (https://complicatedgrief.columbia.edu).

CGT has demonstrated efficacy compared with interpersonal therapy in two trials across the adult age spectrum, with CGT exhibiting better response rates, greater symptom reduction, and less prolonged grief disorder–related impairment (18, 43). In another RCT of 395 patients with prolonged grief disorder, patients were randomly assigned to receive CGT augmented with the antidepressant citalopram, CGT plus placebo, citalopram only, or placebo only. In this study, those receiving CGT (with citalopram or placebo) showed greater reductions in prolonged grief disorder symptoms and suicidal ideation than those on a pill alone (69). Evidence has also been found that CGT is efficacious compared with treatment as usual when delivered in a group setting (84). Overall, these trials have demonstrated efficacy of CGT as a gold-standard intervention for individuals with prolonged grief disorder.

Other therapy protocols with some overlap with CGT that specifically include elements of CBT, such as exposure interventions, may also be useful for patients with prolonged grief disorder. For example, patients randomly assigned to group CBT for grief supplemented by four individual sessions of exposure interventions (exposure to the story of the death) demonstrated greater reductions in grief and depressive symptoms, fewer negative appraisals, and improved functioning compared with group CBT supplemented with four individual supportive counseling sessions (85). In another study comparing two cognitive-behavioral interventions (with six sessions of cognitive restructuring and six sessions of exposure therapy delivered in counterbalanced order) with supportive counseling, the two CBT-based interventions showed greater improvement in grief outcomes (86). Healthy Experiences After Loss, an Internet-based intervention that uses cognitive-behavioral strategies to promote reengagement in positive self-care and with social resources, has demonstrated initial benefits for grief, depression, anxiety, and posttraumatic stress in a pilot study (87). Additional protocols for prolonged grief-specific CBT and suicide bereavement are in development (88, 89).

CBT may also be effective in targeting symptoms that typically co-occur with prolonged grief disorder. For example, a majority of patients with prolonged grief disorder report significant sleep disturbance, which may not always improve with grief-focused treatment (90). For these patients, it may be useful to implement a brief, sleep-focused intervention, such as CBT for insomnia (CBT-I). CBT-I has demonstrated efficacy in improving insomnia symptoms (91) and includes components of sleep hygiene, stimulus control, sleep restriction, and cognitive restructuring regarding negative beliefs about sleep. Little research has specifically examined the effects of CBT-I for bereaved individuals, although some evidence has shown that CBT-I may be useful in improving depressive symptoms (92).

Little systematic research has been conducted on the efficacy of support or bereavement groups on grief-related symptoms, although these groups tend to be accessed by many experiencing loss. For example, in a survey of 47 women who experienced a perinatal loss, 50% accessed support group services (93). For those bereaved by suicide, bereavement groups may help lower intensity of acute grief but may not be sufficient for prolonged grief disorder (94). Aligned with this finding, bereavement groups were qualitatively found to be more helpful and meaningful closer to the loved one’s death, but not all participants reported benefit (95). Overall, bereavement groups may be a useful source of social support after a loss because lack of social support is a risk factor for development of prolonged grief (96, 97); however, more research is required to elucidate the effects of bereavement groups on long-term outcomes for individuals with prolonged grief.

Few studies have investigated the effects of pharmacotherapy alone in the treatment of prolonged grief disorder. A review of the literature identified several early open-label trials of tricyclic antidepressants as well as selective serotonin reuptake inhibitors and other newer generation antidepressants prescribed to bereaved individuals (98). In two small, open-label trials of tricyclics (N=23 total) and one open-label trial of bupropion (N=23), medications significantly improved depression symptoms, but effects on grief intensity were marginal (99–101). In an RCT of 80 older adults with bereavement depression, nortriptyline improved depression symptoms but did not have differential effects on improvement rates of grief intensity compared with placebo (102).

Antidepressants in combination with CBT have also been shown to improve comorbid depression symptoms, but they may have limited effects on grief-specific symptoms. In the large four-arm RCT of 395 patients described earlier, citalopram did not outperform placebo in improving prolonged grief disorder symptoms (69). However, receiving CGT in combination with citalopram showed additive benefit for depression symptoms compared with CGT plus placebo; no differences in grief symptoms were found (69). Of note, suicide-bereaved individuals in the citalopram alone group showed lower completion rates than those bereaved by other causes; thus, pharmacotherapy alone may not be warranted for those bereaved by suicide (64). In a naturalistic study examining effects of antidepressant use for individuals in CGT versus interpersonal therapy, those who remained on antidepressants were less likely to drop out of CGT (91% vs. 58% completion rates) and showed a higher response rate (103). Overall, additional rigorous research is warranted to establish evidence-based targets for pharmacotherapy in prolonged grief disorder and to understand the effects of different classes of medications alone and in conjunction with therapy, especially with respect to improvement in grief-related symptoms. In the meanwhile, given the paucity of evidence for grief-specific efficacy, medications may be most clinically useful in the setting of comorbid conditions or to target specific symptoms if patients do not respond to evidence-based psychotherapy.

As yet, the COVID-19 pandemic has led to more than 400,000 deaths in the United States and more than 2.14 million deaths worldwide. Additionally, because of restrictions such as social distancing, reduced capacity at events, and hospital regulations, typical grief rituals or mourning processes have been significantly disrupted. Therefore, increased prevalence of mental health conditions, including prolonged grief, may be a “second wave” of the pandemic (104). No previous studies have focused solely on individuals bereaved because of historical pandemics (105). However, research with survivors of pandemics, many of whom are bereaved themselves, has identified several themes that emerged subsequent to pandemics, including multiplicity of loss, uncertainty, and disruption of connectiveness (105). Factors influencing bereavement outcomes, which may be disrupted in the current pandemic and contribute to greater chances of developing prolonged grief disorder, include reduced social connections, disrupted grief rituals, and lack of psychosocial support (105).

Research on acute grief during the COVID-19 pandemic has suggested that higher grief symptom levels are present for those bereaved because of COVID-19 than for those bereaved by natural causes in the same time period, but levels were not unlike unnatural bereavement (106). However, acute grief levels may not be predictive of subsequent prolonged grief disorder. In a cross-sectional survey, no significant differences were found between grief levels before or during the pandemic for people bereaved before the pandemic; however, experiencing a loss during the pandemic was shown to elicit more severe acute grief reactions than for acute grief before the pandemic (107). Because COVID-related losses may not have yet surpassed the 6- to 12-month timeframe required for a prolonged grief disorder diagnosis, ongoing research is required to determine the mental health impact of these losses over time; however, it is important to note that research has shown that only a minority of individuals will develop persistent intense grief that causes distress and impairment. Nonetheless, health care workers may be particularly adversely affected because of higher levels of stress, isolation from friends and family, and exposure to death of patients and colleagues; therefore, these individuals should be closely monitored for anxiety, depression, PTSD, and prolonged grief reactions (108). Resources for education about the management of patients with exposure to personal or workplace COVID-related deaths as well as associated traumatic distress are available through several organizations (109–111).

With the new inclusion of prolonged grief disorder in the ICD-11 and DSM-5, future research may utilize advanced computational approaches, such as machine learning, longitudinal trajectories, and network approaches, to better understand differential diagnosis, comorbid conditions, and patterns of prolonged grief disorder development over time. Such approaches have been applied previously with diagnostic criteria sets for complicated grief and persistent complex bereavement disorder (e.g., 37, 112, 113) and from samples assessed with the PG-13 (e.g., 114, 115), which vary slightly from the newly accepted DSM-5 prolonged grief disorder criteria.

Another important direction is to better characterize prolonged grief disorder with a focus on neurobiology and biomarkers. Neurobiological research of prolonged grief disorder using functional magnetic resonance imaging is in its infancy. However, the available evidence suggests neural differences for those with and without prolonged grief disorder. For example, those with prolonged grief disorder demonstrate more activation in the nucleus accumbens when reminded of the deceased, suggesting heightened reward activation associated with the deceased that may impair adjustment to the loss (116). Furthermore, for those with prolonged grief disorder, prefrontal regions may be disrupted during emotion processing (117); in addition, whole brain volume and cognitive function may be different for those with prolonged grief disorder versus normative grief (118).

Few studies have investigated biomarkers of prolonged grief disorder, but initial data suggest future study is warranted. Given the attachment-related nature of prolonged grief disorder, oxytocin, a hormone associated with social connection and bonding, may be implicated. A recent pilot study found higher oxytocin levels among those with prolonged grief disorder compared with those with major depressive disorder; in addition, having threshold levels of prolonged grief disorder on the ICG was associated with higher oxytocin levels (119). Prolonged grief disorder is also categorized as a stress response disorder, which may have implications for immune activation and cortisol responses. Indeed, a review of five studies found that morning cortisol levels were lower for those with prolonged grief disorder, suggesting possible dysregulation of the hypothalamic-pituitary-adrenal axis (120). Overall, more research is needed to understand the neurobiology as well as clinically useful biomarker signatures of prolonged grief disorder. Some potential areas of interest include the biology of stress responses, attachment, and reward.