When grief doesn't pass: what prolonged grief disorder is

“Everyone has moved on. I stayed in that day.”

A woman said this to me three years after her husband died. She went to work, cooked lunches, paid the bills. From the outside it looked as though life had started up again. Inside, nothing had shifted: his wardrobe stayed untouched, his number stayed in her phone, and every day she spent hours going over that last week. She wasn’t sad in the way a person is sad. She was stopped.

In recent years the field has given this state a name. The International Classification of Diseases, ICD-11, has listed it since 2018 as prolonged grief disorder, and the American manual DSM-5-TR added it as a diagnosis in 2022. This is an important change, and a contested one. In this piece I will explain it — with a warning I’m putting right here at the start: this is not a list from which to diagnose yourself. Diagnoses are not made from a blog post.

Why the field needed a new diagnosis at all

For a long time the assumption was that grief is by definition a healthy response and doesn’t need treating. In the large majority of cases that holds. Research consistently shows that most people do not develop a lasting disorder after a loss — the pain reshapes itself over the years, life widens around it, a person can work and love again, even if they never “get over” the loss.

The problem was the minority for whom it didn’t go that way. These people stayed for years in a state that was more severe and more persistent than anything you would describe as sadness, and that did not respond to ordinary support. Because there was no name for it, they often got the wrong one: depression, an anxiety disorder, sometimes nothing at all. They also got the wrong treatment.

A name in that position is not a label but access. A diagnosis means there is a researched way of treating it, that an insurer can cover it, and that nobody says “try to take your mind off it” any more. I’ve written about what a diagnosis says and what it doesn’t in the piece on why a diagnosis is a map and not a name.

What ICD-11 and DSM-5-TR actually describe

Both manuals describe a similar picture. At its core is persistent yearning for the person who died, or a lasting preoccupation with them — a thought that doesn’t move aside and around which the day is organised. Alongside it come several accompanying signs: an inability to accept the death, emotional numbness, a sense that part of you died with them, loss of a sense of who you are, avoidance of anything that recalls the loss, intense anger or guilt, an inability to experience anything good, marked withdrawal from relationships.

Then comes the part that needs reading carefully. The criteria include time — ICD-11 speaks of at least six months, DSM-5-TR of at least a year in adults — and the condition that the response clearly exceeds what is expected in the person’s cultural and religious setting, and that it significantly impairs everyday functioning.

Now the thing that matters more than the numbers: the months are not a deadline. They are not the point at which grief “should” pass, and nobody in the field reads them that way. They are a lower bound below which the diagnosis may not be made at all — a safeguard that stops someone being diagnosed a month after the funeral. If you are still frequently broken eight months after your mother’s death, that means nothing except that your mother mattered.

How often this happens

A meta-analysis of studies among adults who had lost someone to a natural death estimated that roughly one in ten develops this picture. The proportion is substantially higher among people who have lost a child, and after sudden and violent deaths.

Two things follow, and they are worth holding together. First: this is not rare — in a group of bereaved people it is probably one in ten. Second: it is a minority. Nine out of ten people after a loss need neither a diagnosis nor therapy. They need time, people, and permission for it to be hard.

Why this is not depression and not PTSD

In practice the most common confusion is with depression. The difference lies mainly in where the thought is aimed and what the pain is like.

I’ve worked through the difference between sadness and depression in more detail in the piece on why depression is not sadness. Here let this be enough: in prolonged grief the yearning is for a particular person, not a general loss of meaning. The person still wants things — it’s just that all those wants lead to one person who is gone.

What is happening in the brain

An interesting detail from neuroscience: in a study where women’s brain activity was recorded while they looked at images of the person who had died, those with complicated grief showed responsiveness in regions linked to reward and craving — in a nucleus that also responds in longing and in addiction. In those whose grief was following a more ordinary course, this pattern was absent.

This is not evidence that grief is an addiction, and it is certainly no reason for shame. But it is a useful picture for understanding why a person returns to memories even though each return wounds them: the system that drives us towards contact with someone we love stays switched on, because that person never arrives. I’ve described something similar in the piece on rumination, which in grief often weaves itself in as an apparent solving of the unsolvable.

What raises the likelihood

Some circumstances make this picture considerably more likely, though they don’t determine it. The loss of a child. A sudden, violent or unexpected death. A very close or a very complicated relationship with the person who died — including one full of unresolved anger. Dependence on them, not only emotional but practical and economic. An absence of people around you. Depression or an anxiety disorder already present before the loss. Several losses in a short time.

One thing here is often overlooked: not being able to be present in the last days or at the funeral. During the pandemic this happened to a great many people, and I still hear about it in my room today.

Reading a list like this, one thing needs care. Risk factors are not a forecast. Many people with several items from this list grieve without a lasting disorder, and some with not a single one of these circumstances get stuck. The list says where it makes sense to look more closely, not what will happen.

What helps

The good news is that a tested treatment exists and works better than general supportive therapy. Complicated grief therapy, developed in the United States, combines two things: systematically approaching the story of the loss — including telling the death aloud and gradually entering situations that have been avoided — and at the same time working on rebuilding a life, on goals and relationships that could continue. In comparative trials it was clearly more effective than standard interpersonal therapy.

One finding about medication is striking: in a trial where an antidepressant was added alongside therapy, it did not improve the grief itself beyond what therapy achieved on its own, although it helped with accompanying depressive symptoms. That is important information for anyone hoping a tablet will resolve a loss. I’ve written more about the relationship between medication and therapy in the piece on antidepressants and therapy.

In my own room I notice that the hardest part is almost always the second one — rebuilding a life. People are willing to talk about the death. It is far harder to allow yourself the question of what you might still want from your own life, because that feels like a betrayal.

Where I stay cautious and where the criticism is right

Now against myself, because this diagnosis is not innocent.

The first objection is serious: there is a risk that we turn a deeply human experience into an illness. Critics point out that the boundary between intense grief and disorder is in fact continuous rather than sharp, and that a diagnosis can pathologise people whom time and community would have carried forward anyway. This concern is not marginal, and serious grief researchers share it.

The second objection is cultural. The criteria were developed largely on Western samples. What in one setting is a sign of disorder is elsewhere the prescribed way of grieving — anniversaries, visits to the grave, loud lamentation, keeping a room untouched. The classification does explicitly take this into account, but in everyday practice the judgement of “what is expected here” depends quite a lot on who is judging.

The third objection is my own, from the room. People read the criteria online and get frightened. Almost everyone who grieves finds themselves in the list — because the list describes grief. The difference is not in the individual items but in how far life has stopped and for how long. You cannot measure that in yourself by reading.

And fourth: a diagnosis is not an explanation. It tells you what the state is called and what has been shown to help; it doesn’t tell you why this happened to you. The answer to that is in your story, not in a manual.

When it makes sense to talk to a professional

Instead of a list of criteria, I’d rather offer the questions I ask. Are you able to work, look after yourself and be with people — not brilliantly, but at all? Has anything shifted with time, even a little, or is everything as it was in the first month? Are you living your life or only getting through it until you can think about them again? Is avoidance of places, people and things restricting you so much that your world has shrunk?

And independently of all the above: if thoughts appear that you would rather not be here, or if drinking has increased considerably, that is not something to wait out but a reason for an immediate conversation — with your GP, on a crisis line, or with a professional. I’ve described when the first stop is a doctor rather than a therapist in the piece on when to see a psychiatrist first.

What to remember

Prolonged grief disorder is a real condition that affects a minority of bereaved people, and there is treatment for it that works. Its essence is not how long the pain lasts — pain over the loss of someone important is allowed to last a lifetime — but whether life around it has stopped.

The months in the criteria are not a deadline. They are a safeguard against a hasty diagnosis, not a reminder that you are behind. Nobody expects you to have finished something by a particular date.

If you recognise yourself in the description, don’t turn it into a diagnosis. Turn it into one conversation. That is the only step worth taking from reading — and the most this piece can do for you.