We know death in the abstract, we understand it intellectually as we watch parents get older, hear about sudden accidents, and see relatives ageing. Yet when death arrives at our door, it rarely feels like something we were expecting. Instead, it can feel as though all the laws governing the world had changed, and with extraordinary violence. It might come after long periods of anticipatory mourning, when a loved one has been slowly disappearing through illness, or cancer. Even then, the moment of death has an unexpected finality: years preparing can still bring disbelief.

It may revisit our opening a wardrobe that still contains a particular scent, hearing the door because your body still clocks them in at certain times. Increasingly, it is social media reminders, or old texts. Yet each encounter carries the same tiny shock followed by a despairing reality check.

Many people think grief is something between sadness and depression but clinicians have come to understand that it is much larger than losing someone we love. 

Relearning the world
More than a century ago, Sigmund Freud writing in Mourning and Melancholia (1917), suggested that the mind slowly and painfully reorganises itself around an absence — that mourning demanded an enormous amount of psychological work. While this insight has proved durable, the language of psychology has become much more nuanced. 

Neurologist and writer Oliver Sacks observed that much of our nervous system develops around the people closest to us, and that after such an injury we are forced to remap our world. We gradually realise that this internal map is far more detailed than we had imagined, containing countless tiny landmarks that organised our days. A partner sleeps on one side of the bed, answers the telephone at a particular time, or walks through the front door at six o’clock. These expectations become so deeply embedded that we rarely notice them until they are no longer true.

This may help explain why clinicians sometimes compare bereavement to a phantom limb. After an amputation, the missing limb can still feel present because the brain’s map of the body has not yet caught up with reality. 

In The Grieving Brain (2022), Mary-Frances O’Connor argues that our brains are constantly making predictions about the people we love. Shaped by years of shared life, this predictive system continues to expect their return. Bereaved people may smell a familiar perfume, hear footsteps on the stairs, or wake in the night expecting someone to be beside them before remembering they are gone. These moments can be deeply unsettling, yet they speak to the extraordinary intimacy of human attachment. They are not signs of mental illness, but part of the bewildering task of relearning the world after someone has gone.

When someone close to us dies, we lose far more than companionship. We lose someone whose presence helped regulate our nervous system. Grief is therefore not simply emotional; it is profoundly physiological. 

No man an island
Attachment theory begins with a simple observation: human beings are not designed to navigate the world alone. Developed by the British psychiatrist and psychoanalyst John Bowlby in the 1950s and ‘60s, attachment theory proposes that our earliest relationships with caregivers shape how we seek comfort, safety and connection throughout life. As infants, we rely on parents or caregivers not only for food and shelter but also to regulate our emotions and physiology. Over time, these attachment figures become a ‘secure base’ or safe haven from which we explore the world and to which we return in times of stress.

Contemporary neuroscience has deepened this understanding by showing that close relationships become biologically embedded in our nervous system. Long-term partners, family members and close friends help regulate our heart rate, cortisol levels, stress responses and even sleep. Studies have shown, for example, that simply holding the hand of a trusted partner can reduce activity in brain regions associated with threat and pain, while seeing the face of a loved one can dampen the body’s physiological stress response (Coan, Schaefer and Davidson, 2006). In other words, the people closest to us become part of the body’s regulatory system. When they die, we lose far more than companionship. We lose someone whose presence helped our brain and body feel safe. This is one reason why grief is not simply an emotional experience but a profoundly physical one, requiring the nervous system to adapt to a world in which one of its most important sources of regulation is suddenly absent.

Decades of research have expanded on that insight, showing that trusted relationships help regulate heart rate, stress hormones, sleep and emotional responses. Couples’ physiological states often become synchronised over years of living together, and laboratory studies suggest that simply holding the hand of someone we trust can reduce neural responses to threat. 

This helps explain the long-recognised ‘widowhood effect.’ Colin Murray Parkes found mortality among recently bereaved widowers was about 40 percent higher during the first six months after a wife’s death than among married men of the same age. Insurance companies in the 1950s and ‘60s knew bereavement’s correlation with increased illness and premature death, particularly heart disease, during the first year after loss. The phrase ‘dying of a broken heart’ turns out to be more than metaphor. 

Everyone grieves differently
One of the most important developments in bereavement research over the past two decades has been recognising that grief itself is not a mental illness, but that, for a small minority of people, it can become so persistent and disabling that it requires clinical treatment. This distinction lies at the heart of Prolonged Grief Disorder (PGD), which is now recognised in the Diagnostic and Statistical Manual of Mental Disorders (DSM-V).

The difference between normal grief and PGD is not simply a matter of how long someone grieves, but whether they are gradually able to adapt to a world without the person they have lost. In normal bereavement, grief often comes in waves. One moment, a bereaved parent may be overwhelmed by tears after hearing their child’s favourite song; the next, they may find themselves laughing with a friend over coffee. This oscillation between confronting the loss and engaging with everyday life is considered a healthy part of mourning and forms the basis of Margaret Stroebe and Henk Schut’s influential 1999 Dual Process Model. Although the pain may never disappear entirely, most people slowly begin to rebuild routines, relationships and a sense of meaning.

In PGD that adaptation becomes stalled. Intense yearning for the deceased, difficulty accepting the death, emotional numbness, a sense that life has lost its purpose, or an inability to imagine a meaningful future persist for many months and significantly impair daily functioning. 

Crucially, the diagnosis is never made on time alone. Many people continue to miss a spouse, parent or child decades after their death without meeting the criteria for a mental disorder. What distinguishes PGD is not enduring love but enduring impairment. Recognising this difference allows clinicians to validate the profound pain of ordinary grief while identifying those who may benefit from specialised, grief-focussed psychological treatment.