Why Grief Takes So Long: The Neuroscience of Loss and Learning

One of the most common questions people ask after losing someone they love — whether they are grieving themselves or watching someone they care about grieve — is some version of the same thing: why does this take so long?

The question is not impatience. It is bewilderment. The person knows the loss is real. They understand that the person has gone. And yet the grief returns — in waves, in unexpected moments, in dreams, in the automatic reaching for the phone to share something before remembering. The rational mind has accepted the loss. Something else has not.

The neuroscience of grief has an answer to this — and it is one of the most illuminating and compassionate insights to emerge from recent research. Grief takes as long as it does because it is a form of learning. Not a passive experience of sadness, but an active, neurologically demanding process of updating the brain's model of the world around an absence it did not predict and cannot easily incorporate.

TL;DR

• A September 2025 Brain Science Advances narrative review — the most comprehensive recent synthesis of grief neurobiology — confirmed that grief involves distinct neurobiological mechanisms affecting emotional processing, cognitive function, stress regulation, and epigenetic gene expression.
• The "Gone-But-Also-Everlasting" model of grief proposes that the brain holds two simultaneous and conflicting representations: semantic knowledge that the person has died, and deeply embedded episodic memories in which they are still present. Grief is the process of resolving that conflict.
• A January 2026 Nature Neuropsychopharmacology review identified three key brain networks involved in prolonged grief disorder — the Salience Network, Default Mode Network, and Executive Control Network — and proposed a neurobiological model for why some people get stuck.
• Prolonged grief disorder — affecting up to 10% of bereaved people in normal circumstances and up to 30% of those bereaved during the COVID-19 pandemic — is now recognised in DSM-5-TR as a distinct clinical condition requiring specific treatment.
• A 2025 clinical trial of repetitive transcranial magnetic stimulation for prolonged grief disorder began enrolment — the first brain stimulation treatment being tested specifically for grief.

Sleep disturbance and prolonged grief disorder maintain each other in a reciprocal relationship — treating insomnia during grief may be one of the most accessible and underused intervention points available.

Grief Is a Form of Learning

The most useful reframe in contemporary grief science is this: grief is not something that happens to you while you wait for it to be over. It is something your brain is actively doing.

The "Gone-But-Also-Everlasting" model — developed by researchers combining human and animal neuroscience with attachment theory — proposes that grief takes so long because the brain holds two simultaneous and conflicting representations of the person who has died.

The first is semantic knowledge — the rational, explicit understanding that the person has died. This is what the conscious, thinking mind knows and can articulate.

The second is episodic and autobiographical memory — the vast accumulated neural network of memories, associations, expectations, and predictions built around the person across years or decades of relationship. The smell of their coat. The sound of their voice. The automatic expectation of their presence in specific contexts. These representations are not stored in one location in the brain — they are distributed across the hippocampus, the amygdala, the default mode network, and the prefrontal cortex in ways that are deeply integrated into how the brain models the world.

The conflict between these two representations — knowing someone is gone while the brain continues to generate predictions of their presence — is what produces the characteristic experience of grief. The moment of reaching for the phone before remembering. The dream in which they are alive. The involuntary search for their face in a crowd. These are not failures of rational processing. They are the predictive brain encountering reality and updating — slowly, repeatedly, with effort — toward a new model.

Grieving may be a form of learning, requiring time and experiential feedback. The brain's model of the world must change in order for the person to fulfil their goals and achieve a state of safety. This change cannot be accomplished by intellectual understanding alone. It requires repeated exposure to the new reality — each moment of encountering the absence and surviving it is part of the learning process.

This is why time — and specifically the accumulation of real experiences in the world without the person — is the primary mechanism through which grief naturally resolves. Not the passage of calendar time in itself, but what happens during that time.

What the Brain Is Doing During Grief

The September 2025 Brain Science Advances narrative review — one of the most comprehensive syntheses of grief neurobiology published in recent years — identified several brain regions and systems whose altered function during grief explains its characteristic features.

The Amygdala: Heightened Emotional Reactivity

The amygdala — the brain's primary threat and emotional significance detector — shows increased activation and altered connectivity during grief. This heightened amygdala activity explains the emotional intensity of grief, particularly the waves of acute pain that arrive unexpectedly in response to triggers — a song, a smell, a place. The amygdala has assigned high emotional significance to the vast network of stimuli associated with the person who has died. Encountering any of them activates that significance.

Neural alterations are observed in key brain regions associated with memory, emotion regulation, and attachment, including the amygdala, hippocampus, and prefrontal cortex. In people with prolonged grief disorder, amygdala reactivity remains persistently elevated — the emotional significance of loss-related cues does not diminish over time as it does in those who move through grief more naturally.

The Default Mode Network: Self and Other

The default mode network — active during rest, self-reflection, and social cognition — is significantly affected by grief. The DMN is where the brain constructs and maintains the self-concept, autobiographical memories, and the mental representations of important others that form the foundation of relationship.

When someone central to identity is lost, the DMN is disrupted in ways that explain one of grief's most disorienting features: the sense that one's own identity has been altered by the loss. The brain's representation of self was built partly in relation to the person who has died. Their absence is not just external — it disrupts the internal neural architecture of who the grieving person understands themselves to be.

The Hippocampus: Memory and Learning

The hippocampus is central to both episodic memory and learning — the two processes most relevant to grief. It is where new memories are consolidated and where the updating of existing memories occurs.

Stress-induced hippocampal atrophy — damage to the hippocampus from chronic cortisol elevation — can develop after loss and impairs the very learning processes that grief requires. This creates one of the most challenging features of prolonged grief: the stress of grief damages the brain structures needed to resolve it.

A January 2026 review published in Nature Neuropsychopharmacology — examining prolonged grief disorder in later life — confirmed the roles of three key intrinsic brain networks: the Salience Network, which detects emotionally significant stimuli; the Default Mode Network, which maintains self and relationship representations; and the Executive Control Network, which regulates emotional responses and supports goal-directed behaviour. The review proposed a neurobiological model of late-life prolonged grief disorder, grounded in cognitive neuroscience.

The Hormonal Dimension: Oxytocin, Cortisol, and Epigenetics

Beyond the brain's structural and network changes, grief produces significant hormonal disruption with long-term implications.

Hormonal dysregulation, such as elevated oxytocin and cortisol levels, plays a significant role in the physiological response to grief. Epigenetic modifications of stress-related genes further contribute to individual variability in grief responses.

The oxytocin finding is counterintuitive — oxytocin is typically associated with bonding and wellbeing. In grief, elevated oxytocin reflects the brain's active yearning for the lost person — the attachment system remains activated, generating the neurochemical signals associated with seeking proximity to an attachment figure who is no longer present. This is one of the most painful features of acute grief: the very system designed to maintain bonds keeps generating the drive to reconnect with someone who cannot be reached.

Cortisol — the primary stress hormone — is chronically elevated in grief, producing the physiological consequences discussed in the companion Vitae article on grief's physical effects. But the December 2025 University of Pisa systematic review on grief biomarkers adds an important dimension: epigenetic modifications of stress-related genes. Grief does not just produce temporary hormonal changes — it may alter the expression of genes involved in stress regulation in ways that explain why some people's physiological stress response changes durably after significant loss.

This epigenetic dimension is one of the most active areas of current grief research and may eventually explain part of the well-documented finding that bereaved people face higher rates of illness and mortality — not just from the acute physiological stress of grief, but from lasting changes to how their stress-regulation genes are expressed.

Why Some People Get Stuck: Prolonged Grief Disorder

For most people, grief follows a trajectory of gradually decreasing intensity. The acute pain of early loss gives way to integrated grief — a state in which the loss has been incorporated into the person's life narrative and identity, and daily functioning has been restored, even while the loss itself remains present.

For approximately 10% of bereaved people in ordinary circumstances — and up to 30% of those bereaved during the COVID-19 pandemic — this natural trajectory does not occur. In prolonged grief disorder, the brain's reward and attachment circuits remain highly active, as if the brain cannot fully accept the permanence of loss. This "stuck" biological response reinforces yearning and prevents healing.

The Gone-But-Also-Everlasting model identifies several mechanisms through which learning can become stuck:

• Avoidance — avoiding grief-related triggers prevents the repeated encounters with reality that the learning process requires. The short-term relief of avoidance comes at the cost of prolonging the conflict between the brain's two representations of the lost person.
• Rumination — repetitive, passive focus on the loss without the active processing that moves toward resolution. Rumination activates the default mode network without engaging the executive control network that would direct the processing toward adaptive outcomes.
• Stress-induced hippocampal atrophy — chronic cortisol elevation from sustained grief damages the hippocampus, impairing the consolidation and updating of memories that grief resolution requires. This is one of the most important biological explanations for why very intense or very prolonged grief can become self-perpetuating.
• Pre-existing factors — spousal dependency before the loss, pre-existing hippocampal volume, and insecure attachment styles all predict greater difficulty with the grief learning process.

Prolonged grief disorder is now recognised in DSM-5-TR as a distinct clinical condition — not a variant of depression or anxiety, though it can coexist with both — with a characteristic onset six to twelve months after loss and specific symptoms including persistent yearning, difficulty accepting the death, emotional numbness, identity disruption, and significant functional impairment.

The Sleep Connection: An Underused Intervention Point

One of the most practically important recent findings in grief research is the reciprocal relationship between prolonged grief disorder and insomnia.

Research has found that prolonged grief symptoms exacerbate insomnia symptoms, and that insomnia symptoms in turn fuel prolonged grief symptoms — a bidirectional relationship that creates a self-reinforcing cycle. Poor sleep elevates cortisol, which worsens hippocampal function and emotional dysregulation. Emotional dysregulation worsens sleep. And both worsen the capacity for the adaptive processing that grief resolution requires.

This means that addressing sleep during grief — through consistent sleep timing, sleep hygiene, and where necessary clinical support for insomnia — is not merely comfort care. It is directly relevant to whether the grief learning process can proceed. For people whose grief is becoming prolonged, insomnia treatment may be one of the most accessible and clinically underused intervention points available.

What Is Coming Next: Brain Stimulation for Prolonged Grief

The most technologically novel development in grief treatment research is a 2025 clinical trial of repetitive transcranial magnetic stimulation for prolonged grief disorder — the first time brain stimulation has been tested specifically as a grief treatment.

Repetitive transcranial magnetic stimulation is a safe, non-invasive treatment that delivers magnetic pulses to brain areas responsible for mood. rTMS has been approved in Canada to treat mood disorders. The trial, which began enrolment in July 2025 with a completion date of March 2026, targets the brain regions identified in neuroimaging research as dysregulated in prolonged grief disorder — particularly those involved in emotional processing and executive control.

The rationale is direct: if prolonged grief disorder involves identifiable patterns of dysregulated brain network activity, then directly modulating that activity through non-invasive brain stimulation could produce faster and more targeted relief than psychotherapy or medication alone. Results from this trial are anticipated in 2026 and will be an important marker of whether neuroimaging-guided, brain-level interventions for grief are viable.

What Helps: Evidence-Based Approaches

Understanding grief as a learning process suggests a different set of practical principles from the conventional "give it time" advice — while also explaining why time matters.

Allow the encounters rather than avoiding them. Each moment of encountering grief-related triggers — a memory, a place, a piece of music — and surviving it is part of the updating process. Avoidance provides temporary relief but prolongs the conflict. Allowing the encounter, sitting with it briefly, and returning to the present moment is more consistent with how grief resolution actually works.

Prioritise sleep. Given the bidirectional relationship between sleep and prolonged grief, sleep is not a luxury during grief — it is a therapeutic necessity. Consistent sleep timing, limiting alcohol before bed, and seeking clinical support for persistent insomnia are all relevant and underappreciated interventions.

Maintain social connection. Social contact activates the parasympathetic nervous system, reduces cortisol, and supports the hippocampal function that grief resolution requires. Isolation — which grief naturally produces — compounds the neurobiological challenges. Even brief, consistent social connection is meaningful.

Support the body. Gentle physical activity supports cortisol regulation and hippocampal neurogenesis — the growth of new hippocampal neurons that is directly relevant to the memory updating that grief requires. Adequate protein and micronutrient intake supports the biological demands that grief places on multiple systems simultaneously.

Seek professional support when grief is becoming prolonged. Prolonged grief therapy — the most evidence-based specific treatment for prolonged grief disorder — has strong randomised controlled trial support. Recognising prolonged grief disorder as a neurobiological condition rather than a sign of weakness validates patient experience and guides more targeted treatment. If grief is significantly impairing daily functioning more than twelve months after a loss, this is a clinical condition that responds to specific treatment — not a character failing.

Frequently Asked Questions

Why does grief take so long?

Grief takes as long as it does because it is a form of learning — the brain updating its model of the world to incorporate an absence it was not prepared for. The deeply embedded neural networks of memory, expectation, and prediction built around the person who has died do not update through intellectual understanding alone. They require repeated exposure to the new reality — encountering the absence, surviving it, and slowly updating the brain's predictions. This is why time, and specifically the accumulation of real experiences in the world without the person, is the primary mechanism of natural grief resolution.

What does grief do to the brain?

A September 2025 Brain Science Advances review found that grief alters the amygdala, hippocampus, prefrontal cortex, and default mode network — the regions involved in emotional processing, memory, stress regulation, and self-concept. Grief produces hormonal disruption including elevated cortisol and oxytocin, and may produce epigenetic modifications of stress-related genes. Chronic cortisol elevation can cause hippocampal atrophy, impairing the very memory processes that grief resolution requires. A January 2026 Nature review identified three key brain networks dysregulated in prolonged grief disorder.

What is prolonged grief disorder?

Prolonged grief disorder is a recognised clinical condition in DSM-5-TR characterised by intense, persistent grief symptoms — persistent yearning, difficulty accepting the death, emotional numbness, identity disruption — lasting more than twelve months after loss and significantly impairing daily functioning. It affects approximately 10% of bereaved people in ordinary circumstances and up to 30% of those bereaved during the COVID-19 pandemic. It is distinct from depression and anxiety, treatable with prolonged grief therapy, and increasingly understood through a neurobiological lens.

What is the Gone-But-Also-Everlasting model of grief?

The Gone-But-Also-Everlasting model proposes that the grieving brain holds two simultaneous conflicting representations of the lost person — semantic knowledge that they have died, and deeply embedded episodic memories in which they remain present. Grief is the process of resolving this conflict through learning. The model explains why grief takes as long as it does, why avoidance prolongs it, and why pre-existing factors like hippocampal volume and relationship dependency predict complications.

Can sleep affect how long grief lasts?

Yes — significantly. Research shows that prolonged grief disorder and insomnia maintain each other in a reciprocal, self-reinforcing relationship. Poor sleep elevates cortisol, which damages hippocampal function and emotional regulation. Disrupted emotional regulation worsens sleep. Both impair the adaptive processing that grief resolution requires. Treating insomnia during grief is an underused but clinically important intervention that directly affects whether the grief learning process can proceed.

Is there new treatment for prolonged grief disorder?

Prolonged grief therapy remains the most evidence-backed specific treatment. A 2025 clinical trial began testing repetitive transcranial magnetic stimulation — a non-invasive brain stimulation treatment — specifically for prolonged grief disorder, targeting the dysregulated brain networks identified in neuroimaging research. Results are anticipated in 2026. Naltrexone — an opioid receptor antagonist — is also being tested in a separate randomised trial based on the hypothesis that the yearning of grief involves opioid signalling related to social attachment.

The Bottom Line

Grief is not weakness, not failure to move on, and not something a person should be able to think their way through. It is neurologically demanding work — the brain updating deeply embedded representations of a person and a world that no longer exist in the same form.

The neuroscience of 2025 and 2026 makes this clear in ways that are both scientifically precise and humanly compassionate. The amygdala's heightened reactivity, the default mode network's disrupted self-representation, the hippocampus's impaired learning capacity, the conflicting semantic and episodic memories of the Gone-But-Also-Everlasting model — these are not abstractions. They are explanations for the most bewildering features of a universal human experience.

And they suggest something practically important: grief resolution is not passive. It is helped by sleep, social connection, gentle movement, professional support when needed, and the courage to encounter the loss rather than avoid it — because each encounter is a step in the learning process that the brain needs to complete.

For structured support navigating loss through its physical and emotional dimensions, the Grief Reset from the Reset Series™ provides a gentle, practical day-by-day protocol. The Sleep Reset addresses the sleep quality that the research now confirms is directly relevant to whether grief can move toward resolution.

Related reading: The Hidden Physical Effects of Grief: Why Loss Impacts More Than the Mind · Grief and Recovery: What the Science and Experience Both Teach Us · The 7 Minute Connection: Why Giving Someone 7 Minutes Could Be the Most Important Thing You Do Today

Popular Articles