Abstract

Grief is the emotional, cognitive, and physiological response to the death of a person to whom one was attached; MeSH classifies it under bereavement, the objective situation of the loss, although grief names the reaction and not the loss itself. It presents as waves of yearning and sorrow, intrusive thoughts of the deceased, disrupted attention and appetite, and a gradual reorganization of the mental models that had assumed the person's continued presence. Cognitive psychology treats grief less as a fixed sequence of emotional stages than as an extended process of belief updating, in which a deeply overlearned attachment representation is revised against the reality of absence. This article sets out the phenomenology of acute grief, the dual process model of coping, grief as attachment updating, and the boundary between integrated and prolonged grief.

Keywords: grief, bereavement, dual process model, prolonged grief disorder, attachment

Few experiences are as universal, or as cognitively disruptive, as the death of someone close. The bereaved person must function in a world that still presupposes the deceased — a seat at the table, a name in the phone, a half-finished conversation — while the facts contradict every one of those expectations. Early clinical work described the acute state in concrete terms: somatic waves of distress, preoccupation with the image of the deceased, and a sense of unreality, observed systematically for the first time among survivors of the 1942 Cocoanut Grove fire (Lindemann, 1944). The decades since have moved the study of grief from a staged, one-size account toward a model-updating one, in which the central task is not to pass through emotions in order but to revise what the mind takes to be true (O'Connor, 2019). Grief matters to cognitive psychology because it is a natural experiment in how an overlearned representation is unlearned, and in how coping is regulated moment to moment.

Key Takeaways
  • Grief is the response to loss, distinct from bereavement (the loss itself) and from mourning (its cultural and behavioral expression).
  • The staged theory of grief has little empirical support; prospectively, acceptance dominates from the first month and the negative reactions peak and recede rather than arriving in a fixed order.
  • The dual process model describes healthy coping as oscillation between confronting the loss and attending to the demands of ongoing life, with respite from each built in.
  • Neuroscience frames grief as a learning process: the brain slowly updates an attachment representation that still predicts the deceased is present and retrievable.
  • For roughly one in ten bereaved adults, grief remains intense and disabling past the first several months, meeting criteria for prolonged grief disorder, which responds to grief-targeted psychotherapy.

What Grief Is

Grief is distinguished in the research literature from two neighboring terms. Bereavement is the objective situation of loss — the state of having lost someone. Mourning is the public, culturally patterned expression of grief: the rituals, dress, and observances a community prescribes. Grief itself is the internal reaction, spanning emotion, cognition, physiology, and behavior. Keeping the three apart matters because they dissociate: two people in identical bereavement, performing identical mourning, may grieve very differently, and it is the grief reaction — not the fact of loss or the ritual around it — that varies between resilience and disorder.

Acute grief, in its first weeks, is not a single steady feeling but a succession of pangs: episodes of intense yearning and distress, frequently triggered by reminders, that rise sharply and subside, interspersed with calmer intervals (Lindemann, 1944). The bereaved person is typically preoccupied with the deceased, searching for them in crowds or reaching for the phone before the fact of death reasserts itself. These intrusions are not failures of knowledge — the person knows the death has occurred — but the output of an attachment system that has not yet updated its predictions (O'Connor, 2019). Over months, in the ordinary case, the pangs grow less frequent and less consuming, and the person reorganizes goals and identity around the absence without ceasing to value the relationship. Contemporary accounts call this outcome integrated grief rather than resolved grief, because the bond is transformed and carried forward, not erased.

Types of Grief

MeSH indexes grief under the broader descriptor bereavement and lists one narrower descriptor directly beneath grief, shown in Table 1. The MeSH tree is an indexing classification — a scheme for cataloguing the literature — not a theory of how grief carves into natural kinds; the subtypes it names are neither exhaustive nor mutually exclusive, and a single bereaved person's experience can span several descriptive categories at once (anticipatory grief before a death, disenfranchised grief after it).

SubtypeIn brief
Disenfranchised GriefGrief over a loss that is not socially acknowledged, openly mourned, or publicly supported — for example after a stigmatized death, an unrecognized relationship, or a non-death loss — so the mourner lacks the usual scaffolding of ritual and sympathy.

Beyond the single MeSH subtype, the clinical literature routinely distinguishes grief by its timing and course rather than its object. Anticipatory grief begins before a death, during a terminal illness, as the family rehearses the coming loss. Acute grief is the intense early reaction in the weeks and first months after the death. Integrated grief is the reorganized, ongoing form the reaction settles into, in which the loss is accepted and the bond continues in altered form. Prolonged grief disorder is the pathological course in which acute-grief intensity persists, disablingly, well past the expected window — the subject of a later section. These descriptive categories cut across the MeSH tree and are better thought of as points on the trajectory of a single process than as separate entities.

The Phenomenology of Grief Work

The idea that grief requires work — an active confronting and processing of the loss — originates with early psychodynamic theory and was carried into clinical description by Lindemann (1944), who held that the bereaved must perform the labor of detaching from the deceased and that failure to do so produced distorted or delayed reactions (Lindemann, 1944). The grief work hypothesis became the implicit rationale for a great deal of bereavement counselling: that the mourner must express and work through the loss, and that suppression or avoidance is pathogenic.

This hypothesis has not held up cleanly. Reviews note that it was imprecisely defined, assumed rather than demonstrated, and did not generalize across cultures that prescribe very different grieving behavior (Stroebe & Schut, 1999). Prospective data further undercut a rigid version of grief work. In the Yale Bereavement Study, acceptance of the death was the most strongly endorsed response from the first month onward, and the negative reactions — disbelief, yearning, anger, depression — did not arrive in a fixed staged order but each rose to a peak and then declined, with all of them peaking within roughly the first six months (Maciejewski, Zhang, Block, & Prigerson, 2007). The demonstration below reproduces the characteristic decline of acute-grief pangs over time and lets the reader vary how quickly the reaction settles.

The Decline of Acute-Grief Pangs

Acute grief arrives in pangs whose intensity falls off geometrically rather than linearly — steeply at first, then a long tail. Adjust how quickly the reaction settles and watch the curve stretch or compress.

Weeks since lossPang intensity
With a 8-week half-life, pang intensity falls to 50% by week 8, to about 25% by week 16, and to 13% by week 24.
An illustrative exponential-decay model of pang intensity, I(w) = e−kw; representative, not measured. Real courses are noisier and punctuated by reminder-triggered spikes.

The clinical upshot is not that grieving requires no engagement with the loss, but that the engagement is intermittent, and that avoidance — long cast as a failure of grief work — is a normal and even regulatory part of coping, so long as it alternates with confronting the loss. That reframing is the core of the dual process model.

The Dual Process Model

The dual process model of coping with bereavement proposes that the mourner faces two distinct classes of stressor and copes by oscillating between them (Stroebe & Schut, 1999). Loss-oriented stressors are the business of the loss itself: yearning, reminiscing, handling the emotion, contemplating the relationship that has ended. Restoration-oriented stressors are the secondary demands the death creates: mastering new tasks the deceased once handled, reorganizing a daily routine, taking on a changed identity and new roles. Adaptive coping, the model argues, is neither continuous confrontation of the loss nor continuous avoidance of it, but a dynamic alternation — attending to grief, then taking respite from it to deal with living, then returning.

This oscillation does the regulatory work that a one-track grief-work account cannot explain. Dwelling on the loss without pause would be unsustainable; avoiding it entirely would prevent the belief updating the loss demands. By cycling between the two, the mourner titrates exposure, which is why the model treats restoration-oriented avoidance of grief not as denial but as a necessary dosing of a painful process. The demonstration below makes the oscillation explicit, cycling attention between loss-oriented and restoration-oriented coping and showing how time is apportioned between them.

Oscillation in the Dual Process Model

Adaptive coping alternates between loss-oriented attention to the death and restoration-oriented attention to the demands of ongoing life. Set the balance, then step through the sequence to watch the mourner cycle between them.

Loss-orientedRestoration-oriented
Press Step forward to begin the oscillation.
An illustrative scheduler for the dual process model (Stroebe & Schut, 1999); the alternation is schematic, not a measured time-series.

The model also reframes individual and cultural differences. Some mourners weight loss-orientation heavily, others restoration; some cultures prescribe sustained confrontation, others rapid return to function. The dual process model accommodates this as variation in the balance and tempo of oscillation rather than as more or less successful completion of a single prescribed task.

Grief as Attachment Updating

One influential recent reframing treats grief as a problem of learning (O'Connor, 2019). Across a long relationship the brain builds a durable attachment representation that encodes, among other things, the expectation that the attachment figure exists and is retrievable — an implicit model so overlearned that it operates automatically. Death contradicts this model, but a single declarative fact — the knowledge that they have died — does not overwrite an overlearned prediction. The repeated collisions between the model's expectation and the reality of absence — reaching for the phone, setting the second place, scanning the crowd — are prediction errors, and grief is the slow process by which those errors revise the representation. This cognitive-learning view is not the only contemporary account: a constructivist tradition instead treats grief as a search for meaning, in which adaptation turns on narrating the loss into a changed but coherent life story rather than on correcting a prediction (Neimeyer et al., 2014).

This account explains several otherwise puzzling features of grief. It explains why the bereaved can know the death occurred yet still behave as though it has not: declarative knowledge and the implicit attachment model update on different timescales. It explains why reminders provoke pangs: each cue reactivates the old prediction, which reality then violates. And it predicts that the updating is gradual and incremental, like other forms of overlearned relearning, rather than a switch that flips once the fact is accepted. The demonstration below models this updating directly as an error-correcting learning rule, revising a prediction of the attachment figure's presence against the reality of absence.

Grief as Attachment Updating

Each reminder is a prediction error that revises the attachment system’s expectation that the deceased is present, E, toward the reality of absence by a learning rate α: En+1 = En(1 − α). A smaller α stretches integration over far more episodes.

Updating episodesExpectation E
After 0 episodes at α = 0.10, the expectation is 1.000. Reaching the halfway point (E = 0.50) takes about 6.6 episodes. Integration revises the prediction to near zero without erasing the value of the relationship.
An illustrative error-correcting learning rule after O’Connor (2019); the curve is a conceptual model, not measured neural data.

Figure 1

Heterogeneous Trajectories of Grief Over Time

Three grief trajectories plotted as distress against time since loss A line graph with time since loss on the horizontal axis and distress on the vertical axis. A resilient trajectory stays low throughout. A recovery trajectory rises to high distress early and declines gradually toward low. A prolonged trajectory rises early and remains high. Time since loss Distress Resilient Recovery Prolonged
Note. Resilience — a stable low-distress course — is the most common response to bereavement, not a rare or pathological one (Bonanno, 2004). The recovery trajectory shows elevated early distress that declines; the prolonged trajectory remains elevated. Original schematic.

Prolonged Grief Disorder

For most bereaved people the acute reaction softens into integrated grief without clinical intervention; resilience, not chronic distress, is the modal trajectory (Bonanno, 2004). For a minority, grief instead remains intense, preoccupying, and disabling well past the point at which it would ordinarily have begun to settle. A systematic review and meta-analysis of non-traumatic bereavement put the pooled prevalence of this persistent form at roughly one in ten bereaved adults (Lundorff, Holmgren, Zachariae, Farver-Vestergaard, & O'Connor, 2017).

Establishing this as a distinct diagnosis took two decades of psychometric work. A field-trial validation derived a specific symptom set — persistent yearning for the deceased together with symptoms such as difficulty accepting the loss, identity confusion, avoidance of reminders, and a sense that life is meaningless — present to a disabling degree at least six months after the death, and showed it identified a group at elevated risk of enduring distress and dysfunction (Prigerson et al., 2009). That lineage culminated in the formal criteria for prolonged grief disorder (PGD) adopted into the DSM-5-TR and ICD-11, validated alongside a revised screening scale and shown to be statistically distinct from post-traumatic stress disorder, major depression, and generalized anxiety (Prigerson, Boelen, Xu, Smith, & Maciejewski, 2021).

Cognitive-behavioral theory specifies what maintains the disordered course. Three processes are held to be central: the loss is insufficiently integrated into autobiographical memory, so it is not represented as a settled fact; negative global beliefs and catastrophic misinterpretations of one's own grief reactions take hold; and anxious and depressive avoidance strategies prevent the confrontation that would allow updating (Boelen, van den Hout, & van den Bout, 2006). In attachment-updating terms, PGD is a failure of the representation to revise: the prediction of the deceased's presence is never corrected, so the prediction errors — and the pangs they produce — do not diminish.

The disorder responds to treatment that targets grief specifically. A randomized controlled trial found that a dedicated complicated grief treatment produced higher and faster response rates than standard interpersonal psychotherapy (Shear, Frank, Houck, & Reynolds, 2005). A later placebo-controlled trial found that the grief-targeted psychotherapy was the effective ingredient, outperforming antidepressant medication given alone, with the medication adding benefit mainly for co-occurring depressive symptoms (Shear et al., 2016). This specificity mirrors a broader finding: general bereavement counselling offered to all comers shows only small average effects, whereas interventions reserved for those showing marked, prolonged difficulty compare favorably with psychotherapy for other conditions (Currier, Neimeyer, & Berman, 2008).

Worked Example

The attachment-updating account can be made quantitative with the same error-correcting rule that drives the demonstration above. Let E denote the implicit attachment system's expectation that the deceased is present and retrievable, scaled so that E = 1.00 represents the fully intact pre-loss model and E = 0 represents a model that has completely absorbed the absence. Reality now supplies a target of T = 0 at every encounter. On each updating episode — each reminder, each prediction error — the expectation is revised toward reality in proportion to a learning rate α:

En+1 = En + α (T − En) = En (1 − α)

Take a learning rate of α = 0.10, a plausibly slow value for an overlearned representation, and start from E0 = 1.00. Each episode multiplies the expectation by 0.90. After one episode E = 0.900; after two, 0.810; after three, 0.729; after five, 0.590; after ten, 0.349. The expectation decays geometrically rather than linearly, which is exactly the pattern reported for acute-grief pangs — rapid early change, then a long tail — and it reaches the halfway point (E = 0.50) after about 6.6 episodes, since 0.906.6 ≈ 0.50.

The model makes the clinical distinction concrete. A smaller learning rate stretches the same geometric curve out over far more episodes: at α = 0.03, reaching the halfway point takes about 23 episodes rather than 7, and full integration recedes correspondingly. Prolonged grief disorder is, in these terms, the limit in which avoidance and negative appraisal drive the effective learning rate toward zero (Boelen et al., 2006) — the prediction errors keep arriving, but they no longer revise the model, so the expectation, and the pangs it generates, never decline. Note that the rule describes the updating of a prediction, not the erasure of the relationship: an integrated bond with E near zero still carries the full value and memory of the person, now correctly modeled as absent.

Discussion

Treating grief as belief updating rather than as a staged emotional sequence reorganizes several debates in the field. It dissolves the old worry about whether a mourner is doing grief work correctly: what matters is whether the attachment representation is being revised, and the dual process model's oscillation — including its periods of avoidance — is a description of how exposure to the correcting evidence is regulated over time (Stroebe & Schut, 1999). It reframes resilience, long treated as suspicious absence of grief, as the expected outcome for a system that updates efficiently (Bonanno, 2004). And it locates pathology not in the intensity of early grief but in the failure of updating — the stalled revision that cognitive-behavioral theory attributes to poor integration, negative appraisal, and avoidance (Boelen et al., 2006).

The account also clarifies why the staged model was so durable despite weak evidence. The Yale data show that the grief indicators do reach their maxima in roughly the order the stage theory names, even though acceptance dominates throughout and the indicators overlap rather than succeed one another (Maciejewski et al., 2007). A staged narrative is a lossy compression of a set of overlapping rise-and-fall curves — memorable, and not entirely wrong about the ordering of peaks, but misleading as a prescription, because it implies a mourner must pass through discrete phases in sequence and has stalled if they have not.

A limitation worth stating is that the quantitative learning-rule framing is a conceptual model, not a validated measurement instrument; it captures the qualitative shape of grief's decline and the logic of the integrated-versus-prolonged distinction, but the field's actual diagnostic boundary rests on the validated symptom criteria and screening scales, not on an estimated learning rate (Prigerson et al., 2021).

Current Directions

The formal arrival of prolonged grief disorder in the DSM-5-TR and ICD-11 has concentrated recent research on its measurement and boundaries. The validation of the PG-13-R screening scale against the new criteria, across samples at three universities, established that the diagnosis is temporally stable and statistically separable from depression, PTSD, and anxiety — answering a long-standing objection that pathological grief is merely one of those disorders under another name (Prigerson et al., 2021). A parallel strand integrates psychology with neuroscience and psychophysiology, asking which bereaved individuals show the largest neural and physiological signatures of grief and how differences in rumination, inflammation, and cortisol regulation distinguish those who adapt well from those who do not (O'Connor, 2019). The epidemiological work has also sharpened the target for services: with roughly one bereaved adult in ten at risk, and treatment effects concentrated in exactly that group, the current emphasis is on screening to direct grief-targeted therapy to those who need it rather than offering generic counselling to all (Lundorff et al., 2017; Currier et al., 2008).

Common Misconceptions

Grief proceeds through five fixed stages — denial, anger, bargaining, depression, acceptance — in that order.
Prospective data contradict this. Acceptance is the most endorsed response from the first month, and the negative reactions overlap and each rise then fall rather than arriving in sequence (Maciejewski et al., 2007). The five-stage scheme was originally described for dying patients, not the bereaved, and was never validated as a sequence the mourner must pass through.
If someone shows little grief, they are in denial and will pay for it later.
Low, stable distress after loss is the most common trajectory, not suppressed or delayed grief, and it does not predict a later breakdown (Bonanno, 2004). The belief persists because clinical knowledge was built largely from people who sought help, making resilience look rarer than it is.
Grief must be actively worked through, and avoiding it makes it worse.
The grief work hypothesis is poorly supported, and healthy coping involves taking respite from the loss as well as confronting it (Stroebe & Schut, 1999). Avoidance becomes a problem only when it is sustained and total, blocking the belief updating the loss requires (Boelen et al., 2006).
Prolonged grief disorder is just ordinary grief that is taking a long time.
It is a distinct, validated diagnosis with a specific symptom profile — persistent yearning, identity disruption, difficulty accepting the loss — that remains disabling past six months and is separable from depression and PTSD (Prigerson et al., 2021). The distinction matters because the disorder responds to grief-targeted treatment that generic support does not provide (Shear et al., 2005).

Glossary

Acute grief.
The intense reaction in the weeks and first months after a death, marked by pangs of yearning, preoccupation with the deceased, and disrupted functioning.
Anticipatory grief.
Grief that begins before a death, typically during a terminal illness, as the loss is foreseen and rehearsed.
Attachment representation.
The durable, overlearned mental model of an attachment figure, including the implicit expectation that the figure exists and is retrievable.
Bereavement.
The objective situation of having lost a person to whom one was attached, as distinct from the grief reaction to it.
Complicated grief.
An earlier term for the persistent, disabling grief reaction now formalized as prolonged grief disorder; still used in parts of the treatment literature.
Continuing bonds.
The maintenance of a transformed, ongoing relationship with the deceased after integration, rather than a complete severing of the attachment.
Disenfranchised grief.
Grief over a loss that is not socially acknowledged or supported, leaving the mourner without the usual ritual and sympathy.
Dual process model.
The account of coping as oscillation between loss-oriented and restoration-oriented stressors, with respite from each built in.
Grief work.
The hypothesized labor of confronting and processing a loss; a once-central idea now regarded as imprecise and only partly supported.
Integrated grief.
The reorganized, ongoing form grief settles into once the loss is accepted and the bond is carried forward in altered form.
Loss-oriented coping.
Attending to the loss itself — yearning, reminiscing, and processing the emotion of the death.
Mourning.
The public, culturally patterned expression of grief through ritual and observance, distinct from the internal reaction.
Oscillation.
The dynamic alternation between confronting and avoiding the loss that the dual process model treats as adaptive coping.
Prediction error.
The mismatch between the attachment model's expectation that the deceased is present and the reality of absence, which drives the updating of the representation.
Prolonged grief disorder.
The validated diagnosis for grief that remains intense and disabling past six months, with persistent yearning and identity disruption; included in DSM-5-TR and ICD-11.
Resilience.
A stable, low-distress trajectory after loss; the most common rather than a rare or pathological response to bereavement.
Restoration-oriented coping.
Attending to the secondary demands a death creates — new tasks, roles, and routines — and taking respite from the loss itself.
Yearning.
The pang of longing for reunion with the deceased; the dominant negative reaction across the first two years and a core symptom of prolonged grief disorder.

Key Researchers

Paul A. Boelen

Professor of Clinical Psychology at Utrecht University and the ARQ National Psychotrauma Centre, the Netherlands; developed the cognitive-behavioral conceptualization of complicated grief, identifying negative cognitions and avoidance as the mechanisms that maintain a disordered grief reaction. ORCID - Faculty Page

George A. Bonanno

Professor of Clinical Psychology at Teachers College, Columbia University; established the trajectory model of grief, showing that resilience is the most common response to loss and challenging the assumption that absent grief is pathological. ORCID - Faculty Page

Mary-Frances O'Connor

Professor of Psychology at the University of Arizona; pioneered the neurobiology of grief, framing it as a learning process in which the brain updates attachment representations, and documented reward-system activity in prolonged grief. ORCID - Faculty Page

Holly G. Prigerson

Professor of Sociology in Medicine at Weill Cornell Medicine and Director of the Cornell Center for Research on End-of-Life Care; developed the diagnostic criteria for prolonged grief disorder and the PG-13 and PG-13-R screening scales adopted in DSM-5-TR and ICD-11. ORCID - Faculty Page

M. Katherine Shear

Marion E. Kenworthy Professor of Psychiatry at Columbia University and founder of the Center for Prolonged Grief; developed complicated grief treatment and demonstrated in randomized trials that grief-targeted psychotherapy outperforms both standard interpersonal therapy and antidepressant medication alone. ORCID - Faculty Page

Margaret S. Stroebe

Emeritus Professor at Utrecht University and the University of Groningen, the Netherlands; co-originated the dual process model of coping with bereavement, which describes healthy adaptation as oscillation between confronting and avoiding the reality of the loss. ORCID - Faculty Page

Frequently Asked Questions

What is the difference between grief, bereavement, and mourning?

Bereavement is the objective situation of having lost someone, grief is the internal emotional and cognitive response to that loss, and mourning is the public, culturally patterned expression of grief through ritual and observance (O'Connor, 2019).

Does grief really happen in five stages?

No. Prospective research found that acceptance was the most strongly endorsed response from the first month onward and that the negative reactions overlapped and each peaked then declined rather than arriving in a fixed order (Maciejewski et al., 2007).

Is it unhealthy to feel little grief after a loss?

Generally not. A stable, low-distress trajectory is the most common response to bereavement and does not indicate suppressed or delayed grief that will surface later (Bonanno, 2004).

Does grief have to be actively worked through for it to heal?

The grief work hypothesis is only weakly supported; healthy coping alternates between confronting the loss and taking respite from it, and periods of avoidance are a normal part of that oscillation (Stroebe & Schut, 1999).

Why does the mind keep expecting the deceased to be present?

Grief involves updating a deeply overlearned attachment representation, and a declarative fact does not immediately overwrite an implicit model, so the system keeps generating the old expectation until repeated prediction errors revise it (O'Connor, 2019).

How common is prolonged grief disorder?

A systematic review and meta-analysis of non-traumatic bereavement estimated a pooled prevalence of about 9.8%, roughly one bereaved adult in ten (Lundorff et al., 2017).

How is prolonged grief disorder different from depression?

It is a distinct diagnosis with its own symptom profile centered on persistent yearning and identity disruption, and it has been shown to be statistically separable from major depression, PTSD, and generalized anxiety (Prigerson et al., 2021).

Can prolonged grief be treated?

Yes. A grief-targeted psychotherapy produced higher and faster response rates than standard interpersonal therapy and outperformed antidepressant medication given alone (Shear et al., 2016).

References

Boelen, P. A., van den Hout, M. A., & van den Bout, J. (2006). A cognitive-behavioral conceptualization of complicated grief. Clinical Psychology: Science and Practice, 13(2), 109-128. https://doi.org/10.1111/j.1468-2850.2006.00013.x

Bonanno, G. A. (2004). Loss, trauma, and human resilience: Have we underestimated the human capacity to thrive after extremely aversive events? American Psychologist, 59(1), 20-28. https://doi.org/10.1037/0003-066X.59.1.20

Currier, J. M., Neimeyer, R. A., & Berman, J. S. (2008). The effectiveness of psychotherapeutic interventions for bereaved persons: A comprehensive quantitative review. Psychological Bulletin, 134(5), 648-661. https://doi.org/10.1037/0033-2909.134.5.648

Lindemann, E. (1944). Symptomatology and management of acute grief. American Journal of Psychiatry, 101(2), 141-148. https://doi.org/10.1176/ajp.101.2.141

Lundorff, M., Holmgren, H., Zachariae, R., Farver-Vestergaard, I., & O'Connor, M. (2017). Prevalence of prolonged grief disorder in adult bereavement: A systematic review and meta-analysis. Journal of Affective Disorders, 212, 138-149. https://doi.org/10.1016/j.jad.2017.01.030

Maciejewski, P. K., Zhang, B., Block, S. D., & Prigerson, H. G. (2007). An empirical examination of the stage theory of grief. JAMA, 297(7), 716-723. https://doi.org/10.1001/jama.297.7.716

Neimeyer, R. A., Klass, D., & Dennis, M. R. (2014). A social constructionist account of grief: Loss and the narration of meaning. Death Studies, 38(8), 485-498. https://doi.org/10.1080/07481187.2014.913454

O'Connor, M. F. (2019). Grief: A brief history of research on how body, mind, and brain adapt. Psychosomatic Medicine, 81(8), 731-738. https://doi.org/10.1097/PSY.0000000000000717

Prigerson, H. G., Horowitz, M. J., Jacobs, S. C., Parkes, C. M., Aslan, M., Goodkin, K., Raphael, B., Marwit, S. J., Wortman, C., Neimeyer, R. A., Bonanno, G., Block, S. D., Kissane, D., Boelen, P., Maercker, A., Litz, B. T., Johnson, J. G., First, M. B., & Maciejewski, P. K. (2009). Prolonged grief disorder: Psychometric validation of criteria proposed for DSM-V and ICD-11. PLoS Medicine, 6(8), e1000121. https://doi.org/10.1371/journal.pmed.1000121

Prigerson, H. G., Boelen, P. A., Xu, J., Smith, K. V., & Maciejewski, P. K. (2021). Validation of the new DSM-5-TR criteria for prolonged grief disorder and the PG-13-Revised (PG-13-R) scale. World Psychiatry, 20(1), 96-106. https://doi.org/10.1002/wps.20823

Shear, K., Frank, E., Houck, P. R., & Reynolds, C. F. (2005). Treatment of complicated grief: A randomized controlled trial. JAMA, 293(21), 2601-2608. https://doi.org/10.1001/jama.293.21.2601

Shear, M. K., Reynolds, C. F., Simon, N. M., Zisook, S., Wang, Y., Mauro, C., Duan, N., Lebowitz, B., & Skritskaya, N. (2016). Optimizing treatment of complicated grief: A randomized clinical trial. JAMA Psychiatry, 73(7), 685-694. https://doi.org/10.1001/jamapsychiatry.2016.0892

Stroebe, M., & Schut, H. (1999). The dual process model of coping with bereavement: Rationale and description. Death Studies, 23(3), 197-224. https://doi.org/10.1080/074811899201046