Grief and the body: why we fall ill more often after a loss

“Since my father died I’ve been ill three times. Before that I went years without a single sick day. My doctor says everything is fine. Nothing about me is fine.”

I’ve heard versions of this sentence so often that I stopped treating it as coincidence. After a loss, people don’t only describe sadness. They describe a burning pain in the chest that frightens them into going to A&E. Colds that drag on. Digestion that falls apart. Back pain. Exhaustion that sleep doesn’t touch. They tell the doctor they “started falling apart after mum died”, and they get results that are fine.

The results often genuinely are fine. That does not mean the body is inventing things. It means things are happening that a routine test doesn’t measure — and some of them are reasonably well researched.

To the body, loss is a stressor, not a mood

The first thing that needs shifting in how we understand this: the body doesn’t separate “emotional” from “physical” danger the way our language does. When a person loses someone they depended on — emotionally, practically, economically — the same system fires that responds to physical threat. The stress axis between the hypothalamus, pituitary and adrenal glands activates, cortisol output rises, and sympathetic activity increases: faster heart rate, higher blood pressure, muscle tension.

In the short term this is useful. The problem is duration. A system built for a few hours of danger sometimes stays switched on for weeks or months in grief. And persistently raised cortisol is not innocent: it affects sleep, metabolism, blood pressure and — most important for this piece — how the immune system works.

Sleep goes first

In almost everyone who comes to me after a loss, sleep is the first thing to break. Not necessarily falling asleep — more often waking at three or four in the morning, when the head switches on and starts replaying the last week of the illness, or a conversation that should have been had.

Sleep is the most underrated variable in this whole story. Even a few consecutive nights of shortened sleep measurably changes immune markers and worsens blood sugar regulation. Add that a grieving person often eats less, goes outside less and drinks more, and you get a picture that isn’t mysterious at all: a body without the conditions for repair.

This is why I treat sleep as a priority in grief, not as a side symptom. I’ve written about what happens during those middle-of-the-night wakings, and how the nervous system weaves itself into insomnia, in the piece on insomnia and the nervous system.

What gets measured in the immune system

The story began in 1977, when Australian researchers measured a reduced lymphocyte response in widows and widowers some weeks after their partner’s death. A few years later an American group observed a similar change in men who had lost their wives, appearing after the death rather than during the illness that preceded it.

Those early studies were small and nobody today would publish them without caveats. So a systematic review from recent years, gathering the research on the link between bereavement and immune markers, matters more. The picture from it is not dramatic, but it is fairly consistent: signs of mild systemic inflammation and poorer cell-mediated responses appear more often in bereaved people, more markedly in those who also have more severe depressive symptoms or badly disrupted sleep.

It matters to understand what this is not. It is not evidence that grief causes illness. It is evidence that during grief, markers shift which in other conditions are linked with greater susceptibility. The difference between those two sentences is large, and I’ll sharpen it further in the second-to-last section.

A heart that responds literally

The most surprising part of this story is the heart. In 2005 a group of researchers described a condition that until then had seemed more literary than medical: in people after a severe emotional shock — often precisely the death of someone close — a picture appeared that looked like a heart attack, but with no blocked artery. The left ventricle temporarily ballooned into a shape that gave the condition its name, takotsubo. The cause is thought to be a surge of stress hormones that briefly stuns the heart muscle.

In parallel came research on the risk of an actual heart attack. One study interviewed patients after a heart attack about when in the preceding days they had lost someone important. Risk was markedly raised on the first day after the death and then fell rapidly over the following days and weeks.

A caveat is essential here, and I’ll repeat it: these are relative risks. If the baseline probability that someone will have a heart attack today is very small, then multiplied by twenty it is still small. This figure is no reason to panic. It is a reason for someone with known heart disease not to skip their medication in the first days after a death, and for someone to check on them during that time.

What research on widowhood and mortality shows

Folk wisdom knows the story of a spouse who dies soon after the other. Research suggests there is something in it: reviews of large numbers of studies consistently find moderately raised mortality in the period after a partner’s death, more marked in the first months and somewhat more marked in men.

The reasons are probably intertwined. Part is biological, and part entirely down-to-earth: spouses monitored each other’s health, reminded each other about check-ups, cooked, remembered the tablets. When one goes, that invisible health infrastructure goes too. Many a man who has sat with me after his wife’s death had no idea when he had last seen a doctor, because she had always handled the appointments.

Behaviour is often the main carrier

If someone asks me which of all this matters most in practice, I don’t answer with cortisol. I answer with behaviour.

After a loss people often stop eating regularly, or eat mainly what requires no preparation. They stop going out. They cancel check-ups, because these seem unimportant. They drink more, because for the first hour alcohol really does settle things. They forget to take blood pressure or diabetes medication, because the whole shape of the day has collapsed. Smokers smoke more.

These are not signs of weakness but consequences of being overloaded. They are also the part that is easiest to support — which is exactly why I talk about it first. Nobody can lower another person’s cortisol with advice. But someone can bring them a cooked meal and remind them about a doctor’s appointment.

This is not “all in your head”

When results show nothing, people are often told the whole thing is “psychological”. That sentence is badly put and it does harm. The symptoms are real, measurable and arise from bodily processes — the processes are simply different from the ones a basic blood test looks for.

I’ve written about what it means when pain is real but the results are fine in the piece on psychosomatic symptoms. In grief there is often a further layer: the body holds the memory of the last weeks of the illness, of hospital corridors, of the phone call. I’ve said something about that in the piece on trauma in the body.

Where I stay cautious and where this story doesn’t hold

Now against myself, because this topic tips easily into scaremongering.

First: most bereaved people do not become ill. A statistically raised risk in a group says nothing about what will happen to you. If after this piece you read every skipped heartbeat as a consequence of your loss, I haven’t helped you — I’ve added another worry.

Second: nearly all of this research is observational. It shows that things occur together, not that one causes the other. People who lose a partner are as a rule also older, often with existing illness, and often having spent months or years as carers, which exhausts a body by itself. Separating the effect of grief from the effect of carer exhaustion is very hard, and many studies have not managed it.

Third: the old immunological studies were small and methodologically weak, and the markers are often only indirect. “A poorer immune response in a test tube” is not the same as “you will fall ill”. The field treated that distinction too loosely for a long time, and some popular articles still promise more than the data allow. I’ve explained why the effects of psychological processes are so hard to demonstrate at all in the piece on why therapy is harder to research.

And fourth, most practically: none of these explanations may become a reason to attribute a new physical symptom after a loss to grief and leave it unchecked. Chest pain after your father’s death may be grief. It may also be a heart attack. The order is always the same: doctor first, explanation afterwards.

What actually helps

None of this is surprising, and that is the point. Sleep first, even if all you manage is a regular time for getting up. Short walks every day, because movement acts on sleep and mood at once. Food at regular times, even if it is plain and even if you can’t taste it. Medication continued, without exception. A visit to your GP in the first months, even if nothing is wrong.

And the thing that doesn’t look like a health measure but is: people. In bereavement research, the presence of other people is consistently linked with a better course, physically as well. The conversations don’t have to be deep. It’s enough that someone comes.

What to remember

Grief is not only an emotional state. It is a period in which the body runs for a long time in threat mode, sleep falls apart, and the everyday habits that kept you well quietly drop away. That is why we fall ill more often after a loss — and why results that come back fine don’t mean you invented the symptoms.

The reverse holds at the same time: raised risk in a group is not a forecast for you. Most people walk through this period, and the body settles when the life around it settles.

The most useful thing you can do isn’t managing cortisol but protecting three simple things: sleep, medication and contact with people. If physical symptoms drag on or frighten you, go to a doctor first and reach for explanations afterwards. I’ve described how grief itself reshapes over the months in the piece on sadness as a process.