Can your mind make you sick?

• About
• Latest Posts

Adam Foster

Director at The Fibro Guy

Adam Foster is the founder of The Fibro Guy and one of the leading voices in hypermobility, EDS, and chronic pain rehabilitation. A former Reconnaissance Soldier, he developed chronic pain after an IED blast in Afghanistan and built his approach from the ground up when the medical system had no answers. Since 2009 he has worked with thousands of people across eight countries, drawing on pain neuroscience, motor learning, and sensorimotor rehabilitation. He co-authored "No Pain, No Pain" with trauma psychologist Dr Mairi Harper, directed the documentary "Invisible Monster," and has been featured in Forbes, BBC Radio 4, ITV News, LiveScience, and The Telegraph.

Latest posts by Adam Foster (see all)

• Hypermobile Neck Exercises - 8 November 2026
• Swollen legs and hypermobility: pooling, lymphoedema and lipoedema - 4 November 2026
• Walking with fibromyalgia - 2 November 2026

Yes, your brain can produce real physical symptoms, and long running stress really does change how your body deals with an infection. Both of those have been measured in people, deliberately, under controlled conditions. What that doesn’t tell you is whether a particular symptom of yours was invented, because every symptom you have ever had, a broken ankle included, was assembled by your nervous system before you felt a thing. Brain involvement is true of the lot, which is exactly why it settles nothing about any one of them.

So the question worth asking is which processes are producing this one, and what can be done about each. Real or in my head is a question with no answer in it.

How your brain reaches the rest of you

Your brain isn’t sitting in a box on top of a body it occasionally sends instructions to. It’s wired into the rest of you, and most of the traffic is coming up from the body rather than going down to it.

Some of that wiring is literal. The autonomic nervous system, the part that runs your heart rate, your gut, your blood vessels and your sweat glands without ever consulting you, reports back from all of those places as well as giving the orders. Hormones carry the rest of the outbound traffic, cortisol among them, going out into the blood and reaching any tissue with a receptor waiting for one. And your immune system talks upwards.

Cytokines, the chemical messengers immune cells use to coordinate a response, reach the brain through the bloodstream and along the vagus nerve, which runs from the brainstem down into the chest and gut. The brain answers back by changing what the immune system does next. That loop is why an infection feels the way it does. Aching, foggy, flat, off your food, unwilling to get off the sofa: that’s sickness behaviour, a coordinated programme the brain runs when it picks up inflammation, and it’s there because lying still and eating nothing is a sensible thing for an animal to do while its immune system gets on with the job. So your brain already has a switch that produces physical symptoms, and it uses it on everybody, several times a winter.

Stress reaches your health along those same biological routes and along the duller ones too, what you eat, whether you sleep, whether you smoke ^[1].

Inflammation

The cleanest evidence for any of this comes from making healthy volunteers mildly ill on purpose. Give somebody a small dose of bacterial endotoxin, which is a fragment of bacterial cell wall the immune system reacts to without any actual infection being present, and mood drops and anxiety rises in proportion to the dose, alongside the cytokine, temperature and hormone changes you’d expect ^[2].

Mood is only part of it though. After an endotoxin challenge people become less willing to pick the high effort option in a task, without their appetite for the reward itself dropping off ^[3]. The reward still looks worth having, and what’s gone up is the price of getting off the sofa to fetch it, which is close to what people describe when they say they want to do the thing and can’t make themselves start it.

A typhoid vaccine does a gentler version of the same thing: fatigue, muddled thinking and trouble concentrating, arriving with no fever at all, so there’s nothing to point at and say that’s why I feel wrecked. The fatigue tracked activity in the insula and cingulate cortex, the parts of the brain that deal with the body’s own internal state ^[4].

The symptoms are usually short lived though ^[5], so what these experiments show you is that the machinery exists and is strong enough to flatten a well person on purpose, rather than that a rough week is the same thing as a rough decade. The dose matters as well. An influenza vaccine raised cytokines with no change in the group’s mood or social behaviour ^[6], so mild inflammation doesn’t reliably move any of it.

Stress and infection

A virus still has to get into you from somewhere, so what stress changes is what happens once it has. When people were exposed to a cold virus under controlled conditions, those under long running stress were more likely to actually come down with it, and the mechanism offered is that immune cells become less responsive to cortisol, which is the signal that would normally keep inflammation in check ^[7].

Cortisol gets sold as the stress hormone, which makes that sound backwards. One of its main jobs is putting the brakes on inflammation, so if the cells it’s shouting at gradually stop listening, the brakes come off while the shouting carries on. So it’s the long grinding sort of stress that matters here rather than a bad afternoon, and taken across a very large body of work, chronic stress goes with suppression of several immune measures, while short bursts of stress sometimes shift things in a useful direction instead ^[8].

If you’ve lain awake wondering whether you stressed yourself into this, two things are worth having. Stress is in the picture for heart disease and it’s a long way from the biggest thing in it: its effects in adults are generally smaller than smoking, blood pressure and cholesterol, and it looks most important as a trigger in people who are already vulnerable ^[9]. The size of your own stress response is a poor guide to your health years later too, with 68% of the associations coming back null, and both unusually large and unusually small responses going on to predict different things ^[10].

Anxiety is a physical event

Anxiety arrives with a body attached, and the body is most of what you actually notice. A panic attack is chest discomfort, palpitations, breathlessness, dizziness, abdominal distress and tingling, which is a list anybody with POTS will read with a certain weariness ^[11]. Part of the dizziness has a plain mechanical explanation: breathing hard lowers carbon dioxide in the blood, and low carbon dioxide narrows the vessels feeding the brain ^[11]. Less blood arriving at the head, and the room goes swimmy. There is nothing imagined anywhere in that sequence.

Can expecting a symptom bring it on?

This is the nocebo effect, the placebo effect’s unpopular twin, and it’s the half of the story nobody sells anything on the back of. Tell somebody a sensation will hurt and it hurts more than the identical sensation does without the warning, and in healthy volunteers that effect on pain and itch sits somewhere between small and moderate ^[12].

It’s measurable below the neck, which is the part that surprises people. Put somebody in a scanner and deliver identical heat, and after a nocebo instruction the dorsal horn of the spinal cord, which is the first relay the signal reaches on its way up, is working harder ^[13]. The amplification has already happened before any of it gets near the part of the brain that would decide what to tell you.

That makes sense once you know how perception works. Your brain runs a constant best guess about what’s arriving and what it’s likely to mean, and it weighs the incoming traffic against that guess rather than reporting it raw. A warning changes the guess, and the dial it turns sits a long way below the bit of you that would be choosing words for it.

Most of these experiments amplify something that was already there rather than generating a symptom out of nothing ^[14]. There’s a real stimulus, a real bit of heat or a real itch, and expectation turns the volume up on it. Attempts to bring on symptoms purely out of a warning have a patchier record, and several of them simply haven’t worked ^[15].

Which symptom you’re asking about matters as well. The clearest effects are on pain and itch, with fatigue and nausea considerably less settled than either ^[16], and nausea is what people mean when they ask whether worry can make them feel sick.

So: worry turns the volume up on something that’s already there, and that shows up in the spinal cord rather than only on a rating scale. Conjuring a symptom out of nothing is a bigger claim, and a thinner one.

Real, or in my head

The trouble starts when brain involvement gets read as a verdict on whether a symptom is genuine. The neatest demonstration of how wrong that is came from running both processes in the same people at the same time. In people made ill with endotoxin, a positive treatment label reduced the sickness symptoms they reported and did nothing at all to their inflammatory markers, which carried on doing what inflammation does ^[17]. Expectation moved how ill they felt, and the immune system carried on regardless.

The same holds for the conditions that get labelled functional, which is the word medicine uses when a system is working wrongly without anything turning up broken on a scan. Functional neurological symptoms, which includes weakness, tremor and seizure like episodes, are genuine and involuntary ^[18], and involuntary means what it says: a leg that won’t take weight is outside the reach of anybody’s will, yours included. Psychological difficulty is neither necessary nor sufficient for the diagnosis, which means somebody can have one of these with the psychological part absent ^[19].

IBS is where this gets argued hardest, and the gut is demonstrably doing things too: altered motility, meaning the muscle wall moving things along too fast or too slowly, altered permeability of the gut lining, immune activation and changes in the bacteria living there, none of which a purely psychological account covers ^[20]. And once a mental health label is already on the file, physical symptoms start getting read through it. That has a name, diagnostic overshadowing, and a measurable cost to the care people end up getting ^[21].

Fibromyalgia, hEDS and POTS

What turns up on the tests

Fibromyalgia has been imaged more than most conditions in this family. What comes up repeatedly is reduced grey matter volume or cortical thickness in the anterior cingulate, the insula and the prefrontal cortex ^[22], regions heavily involved in monitoring the body’s own state and in deciding how much attention a signal gets. On tasks, the networks handling nociception, the incoming traffic from tissue that may or may not end up being felt as pain, respond more than they should, while the descending control system, the one meant to turn that traffic down before it arrives, works less well than it should ^[22].

Resting state findings are all over the place and give no stable signature ^[22]. None of that tells you which came first, and it does tell you there’s something there to find, on a scan, in a condition people get sent home from clinics being told has nothing in it.

There’s a peripheral side as well, and it took a long time to get taken seriously. Around 49% of people with fibromyalgia show small fibre pathology on a skin biopsy or corneal imaging ^[23]. Small fibres are the thinnest nerves you’ve got, carrying temperature, nociception and the autonomic traffic out to blood vessels and sweat glands, so losing them turns up as burning, odd temperature sensations and some of the trouble with standing upright. Call it about half, which makes it a genuine finding in a condition supposed to have none, and leaves plenty of people sitting on the other side of it.

In hEDS the figure comes from a much narrower group. Among those referred on to a specialist for neuropathic pain, which is pain generated by the nerve itself rather than by the tissue it reports on, or for dysautonomia, meaning an autonomic nervous system running badly, 58% had definite small fibre neuropathy on both nerve fibre density and sensory testing, with another 33% possible ^[24]. Everybody in that sample had already been sent on because something looked wrong, so the number belongs to them rather than to everyone with hEDS.

In POTS, imaging shows structural changes in brain tissue and reduced blood flow to the brain at the sites handling autonomic control ^[25]. Brains work badly on a reduced blood supply, which is the dull and obvious explanation for the greying out and the fog that arrives with being upright.

One caution about a tool you may have been scored with. The Central Sensitization Inventory, the questionnaire used to screen for amplified pain processing, tracks depression, anxiety and catastrophising closely, and tracks the laboratory measures of pain sensitivity weakly or not at all ^[26]. I’d treat a high score on it as information about something, and probably not about the thing the name on it promises. Central sensitisation itself is a separate question and it has its own article.

Being told it’s psychological

In a review of hEDS case records, 94.4% had been told their symptoms were in their head, that they were making it up, or had been given a conversion disorder diagnosis ^[27]. Conversion disorder is the older name for functional neurological disorder, and that label is the one rejected hardest of the lot, at 95% in one hEDS group ^[28]. In a survey of young people with POTS at one centre, 65% had been told they had a psychological or psychiatric disorder before the POTS diagnosis arrived ^[29]. Most estimates of this kind come from surveys people opt into, which will over represent the ones who had a bad time of it ^[30], and that caveat belongs with every figure here.

Delay is the other half of it. Mean time to a fibromyalgia diagnosis in real world data is 6.42 years ^[31]. A global survey put the mean for hEDS at 22.1 years ^[32]. Women with POTS wait longer than men, 7.0 years against 3.8, and that gap was put down to factors on the clinician’s side rather than to women being slower to go and ask ^[33]. Twenty two years is long enough to be told it’s growing pains, then that you’re anxious, then that you’re stressed, and to quietly stop bringing it up at all.

It leaves something behind, as well. Over 80% of people reporting a psychosomatic misdiagnosis also reported damage to their self worth, and they were significantly more likely afterwards to under report symptoms and avoid healthcare altogether ^[34]. If you’ve ever gone quiet in an appointment because the last one went badly, that’s the mechanism, written down.

There’s a loop in all this that keeps the argument circular. Anxiety is a genuine comorbidity in these populations and not only ever a misdiagnosis ^[35]. The overlap between what anxiety does to a body and what POTS does to a body is real, and the two are close to indistinguishable without objective testing ^[36]. So the clinician in front of you is looking at a genuinely ambiguous picture, which explains how the error happens and does nothing to excuse where it lands. Having anxiety alongside POTS doesn’t make the POTS less real, and having POTS doesn’t make the anxiety imaginary.

The question worth asking instead

If somebody tells you stress plays a part in your symptoms, what they’ve named is one process out of several, and the follow up worth having is which ones, and what can be done about each. Nervous system involvement covers everything from a sprained ankle to a flare, so on its own it sorts nothing.

Where we’d put the effort is on the processes that actually move. Sleep, because every bit of machinery described above runs worse on a bad night. How much of the day your body is spending in a state it’s treating as an emergency. And how clearly your joints are reporting where they are, which is the one we’d start with, and the only part of the list that shifts in a studio rather than a clinic. Your joints are full of receptors telling your brain where each limb is and how much load is going through it, and the brain leans on that to hold you upright without bothering you about it. When that report arrives late, or arrives vague, the brain compensates: it grips, it guesses, and it pays far more attention to the area than the area deserves.

And if you’re being sent away with an anxiety label that doesn’t fit what’s happening to you, the question to take back in is what would be done differently if it weren’t anxiety, and whether that test has been done yet.

So, your brain can make you sick. It does that to everybody, and it has no bearing at all on whether the person it’s happening to is telling the truth. Ask which processes are running, and you’ve got a question an appointment can do something with.

Adam

References

[1] O’Connor, D.B., Thayer, J.F. and Vedhara, K. (2021) ‘Stress and Health: A Review of Psychobiological Processes’, Annual Review of Psychology. https://doi.org/10.1146/annurev-psych-062520-122331

[2] Grigoleit, J.S., Kullmann, J.S., Wolf, O.T., Hammes, F., Wegner, A., Jablonowski, S. et al. (2011) ‘Dose-Dependent Effects of Endotoxin on Neurobehavioral Functions in Humans’, PLoS ONE. https://doi.org/10.1371/journal.pone.0028330

[3] Draper, A., Koch, R.M., van der Meer, J.W., AJ Apps, M., Pickkers, P., Husain, M. et al. (2017) ‘Effort but not Reward Sensitivity is Altered by Acute Sickness Induced by Experimental Endotoxemia in Humans’, Neuropsychopharmacology. https://doi.org/10.1038/npp.2017.231