POTS and Exercise

• 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)

• Teeth, Gums and Jaw Pain in hEDS, and How to Get Through a Dental Appointment - 28 September 2026
• Why ADHD and Hypermobility Keep Turning Up Together - 16 September 2026
• Hypermobility and Early Development: Milestones, Motor Delay, and Predictions - 12 September 2026

If you’ve had a POTS diagnosis in the last few years, you’ve almost certainly been handed some version of the same advice: you’re deconditioned, you need to exercise, start on a recumbent bike and build up. It isn’t stupid advice, as exercise sits at the top of every POTS list for good reasons, in that it has few side effects, costs almost nothing and is available to everybody. The same people recommending it will also tell you it’s unknown who it actually helps, or by how much ^[1]. And what tends to happen next is that somebody goes away, does exactly what they were told, feels considerably worse, and then gets told they didn’t stick at it.

So here’s what we do instead. The order is the whole point, and the order is the bit nobody gets given.

The Deconditioning Story

The story goes like this. You stopped moving, your heart got smaller, your blood volume dropped, and the racing heart rate is just an unfit body doing its best. Get fit, fix the POTS.

The physiology underneath it isn’t invented. Hearts really are smaller than in people who are equally sedentary without having POTS, and lower blood volume turns up again and again in this group ^[2].

What doesn’t survive is the direction of the arrow. When one American centre asked its own POTS caseload, more than three quarters of them said they’d been exercising regularly before the symptoms that led to the diagnosis ever started ^[3]. Deconditioning of the heart has been argued not to be the main thing driving exercise intolerance here at all, and the brain’s blood supply still drops during orthostatic stress regardless of how deconditioned somebody is ^[1]. Then there’s the timing of the whole thing, as new POTS diagnoses across a very large American health system weren’t trending upwards at all before March 2020, and were rising sharply afterwards ^[4]. Whatever happened there, a sudden national outbreak of laziness isn’t it.

None of which means the deconditioning is fictional. If you’ve spent two years mostly horizontal then you are deconditioned, and that’s real, and it’s part of what you’re dealing with. It just isn’t the cause. It’s a passenger, and treating a passenger like the driver is how people end up doing the right work in the wrong order.

The other thing worth having in front of you is how often hypermobility is sitting in the middle of all this. Among people carrying a POTS diagnosis, around a third also meet the criteria for hypermobile Ehlers-Danlos syndrome, and a decent chunk more are hypermobile without meeting the full criteria ^[5]. That was a self selected group who’d turned up to a conference on dysautonomia, so read it as an overlap worth knowing about rather than a population rate.

What Training Actually Does, And What It Costs

It does something real, it does less than the headlines suggest, and most people can’t finish the programme as written.

The most reliable thing training does is bring your heart rate down, and that turns up everywhere anybody has looked ^[6]. The other things people measured point the right way too, and none of them has enough behind it yet to lean on, and the people who gathered all of it together said plainly they couldn’t turn any of it into a protocol ^[6].

The mechanism side holds up well enough. Three months of the standard programme makes you fitter, makes the heart bigger and puts blood volume back up, and about half the people who did it in a research setting no longer met the objective criteria for POTS at the end ^[2]. Out in the community it was higher still ^[2]. Those are genuinely good numbers and they’re the reason exercise is first line.

Now read the rest of the table, though, because the part everybody quotes is the part at the top. In the one place people were allocated at random, training made them fitter and let them work at a lower heart rate, and the tachycardia on standing didn’t shift at all ^[7]. In the pragmatic trial, counting everybody who was invited rather than only the ones who agreed to take part, about half still met the heart rate criteria six months later ^[8]. And further down that same discussion sits the line nobody quotes, which is that most of them still reported severe fatigue and still rated their quality of life far below healthy levels ^[8]. So you can stop meeting the criteria for POTS and still be knackered and still not have your life back, and both of those things are in the same paper.

The completion figures are the part the advice sheet leaves out. Somewhere around forty five per cent of people get through the programme at all ^[2], and out in the community, rather than in a research setting with somebody checking on you, it’s a good deal worse than that ^[6].

One bit of fairness before anybody uses that as a reason not to bother. Sticking with blood pressure medication, or with cardiac rehab, runs in much the same band ^[2]. That isn’t a unique failure of people with POTS. The difference is that nobody tells a man with high blood pressure that his medication didn’t work because he lacked willpower.

What the dropout does tell you is what it costs. Asked directly, most of them said exercise makes them feel worse, and nearly all of them said it makes them dizzy, and a good many said their joints feel unstable and that they’re afraid of hurting themselves ^[3]. That’s one centre asking its own people, with questions it wrote itself, so hold the exact proportions loosely. The shape of it is hard to argue with.

And if you’re in the hypermobile half of this audience, the position is worse than thin. Somebody went looking specifically for exercise training for POTS in those with joint hypermobility, and found two pieces of work that answered the question ^[9]. Widen it out to exercise for POTS generally and you get ten, not one of which examined exercise training for POTS in people diagnosed with hypermobile EDS ^[9]. Those with hEDS were sitting inside at least one of those trials and nobody analysed them separately, so it’s a missing analysis rather than proof nobody’s ever been looked at. It’s still the entire evidence base for the group most likely to be reading this.

How You’re Likely Breathing

Every protocol above starts the same way, with recumbent cardio and salt and fluids. Not one of them starts with breathing. We do, and this is us telling you what we’ve found rather than what’s been trialled, so take it as our position.

Pretty much everybody we work with who has POTS is overbreathing, and almost none of them know it. It’s rarely dramatic panting, it’s just a bit too much air, a bit too often, mostly through the mouth and mostly up in the chest. And when somebody is doing that, everything we try to progress them onto upright gets harder, slower and considerably more miserable than it needs to be.

It’s common enough that there’s an NHS respiratory physiotherapy service doing nothing but this in POTS, and people who go through it come out breathing slower with lower symptom scores ^[10]. Take that gently, though, as there was no comparison group, everybody in it had already been picked out as having a breathing problem, and the people who ran it call it preliminary themselves.

The reason we think it gets missed is that you genuinely cannot feel it. Ask someone with hEDS to take a breath of a particular size, let them practise with a screen showing them how they’re doing, then take the screen away and ask for the same breath again, and they’re further off than people without hEDS are ^[11]. They also sit and breathe faster at rest without actually moving any more air, and give them something mentally demanding to do at the same time and the whole pattern gets more erratic than it does in everybody else ^[11]. It’s one lab and a small group, so don’t build a cathedral on it. It does line up exactly with what we see, though. Ask somebody how they’re breathing and they’ll tell you it’s fine, put a hand on them and it plainly isn’t.

Why The Carbon Dioxide Matters

This is the mechanism, and it’s worth understanding because it changes what the exercise is actually for.

You stand up. Blood pools in your legs, less comes back to the heart, and less gets pushed up to your brain. What happens next in POTS is that your breathing goes up ^[12].

And the order of those two is the useful part. In young people with POTS whose main complaint was breathlessness on standing, the fall in brain blood flow came first, and the breathing followed it about seventeen seconds later ^[13]. So the overbreathing isn’t a bad habit somebody picked up, and it isn’t anxiety. It’s a response to the blood supply to the brain dropping. Which is worth saying out loud, as most people with POTS have been told at some point that they just need to calm down.

The problem is that the response makes the original problem worse. Breathing more blows off carbon dioxide, and carbon dioxide is what keeps the blood vessels in your brain open. Measured standing, carbon dioxide runs lower in POTS than it does in healthy people, and brain blood flow falls a great deal further ^[14]. Adjust that flow for the carbon dioxide and the difference between the two groups stops being significant ^[14], which is about as clean a way as you’ll get of saying the carbon dioxide is doing the work rather than just sitting alongside it.

There are also people carrying this entire picture without the tachycardia, whose heart rate never rises enough to earn them the diagnosis, and one reading of that is that they and POTS are the same thing with and without the racing heart ^[15]. That’s offered as a hypothesis rather than as settled, and it’s worth sitting with, as heart rate is the number everybody watches.

For this audience the most directly relevant work came out this year, and it asks the question in hypermobility specifically. Adults with hEDS who also had POTS were put in an MRI scanner while blood was pulled down into their legs with suction, which is standing up without the standing up. Under that stress the hEDS group started breathing faster and their carbon dioxide fell, and the people without hEDS did neither ^[16]. How hard it was to push blood through their brain went up by more than twice as much as it did in the controls, and blood pressure didn’t move in anybody ^[16]. It’s a small group, every one of them had POTS as well so you can’t pull the hEDS apart from it, and the people who ran it call their own subgroup work hypothesis generating. It’s still the closest thing there is to a mechanism written for the person reading this.

Two honest limits on all of that, though. Not everybody with POTS is running low on carbon dioxide, and if you tilt a room full of people with POTS then about half of them will be and half won’t ^[17]. It also comes by degrees rather than as a switch. So it’s a real mechanism in a meaningful subgroup rather than the explanation for POTS, and anybody telling you low carbon dioxide causes all of it is well ahead of the evidence.

What Happens If You Put The Carbon Dioxide Back

Everything above describes a problem. The reason we order things the way we do sits in what happens when somebody puts the carbon dioxide back in.

Tilt people upright, have them hyperventilate, then hand them a bag and let them breathe their own carbon dioxide back in, and the heart rate comes down by around fifteen beats while the blood pressure stays where it was ^[18]. That was a handful of people, on one afternoon, with a paper bag, so hold the size of it lightly.

The better version gives supplemental carbon dioxide through a mask near the end of a tilt, to the subgroup of people with POTS who were short of breath standing up. Brain blood flow went back to normal. Their breathing settled. And afterwards their heart rate, the amount of blood the heart was shifting, their blood pressure and how hard the body was squeezing its own vessels were all no longer different from people without POTS at all ^[19].

The caveats on that are serious. It’s one session, in a lab, in teenagers and young adults, in a subgroup rather than in everybody. A gas through a mask is not a breathing exercise, and nobody has shown that any breathing exercise reproduces it. Nobody has run it over weeks and measured whether a single person’s life improved.

What it does establish is narrower and still worth having. In that subgroup the low carbon dioxide is load bearing rather than incidental. Take it away and the physiology goes with it, and put it back and a good part of it comes back too.

Which is the whole argument for the order we use. If a chunk of your upright tachycardia is riding on your breathing, then being sent straight to upright cardio is being asked to do the hardest version of the task with the amplifier still turned up.

What We Do First, And It Isn’t Cardio

Breathing, before anything is loaded and before anything is upright.

The model we teach is a canister. The diaphragm is the lid, the pelvic floor is the floor, the abdominal wall and the obliques are the sides, and the lower back closes it off behind. Breathe in and the diaphragm drops, the pelvic floor descends with it, and the whole thing expands in every direction rather than just forwards, and that expansion is what creates the pressure that stabilises you. Breathe out and it all comes back in, the obliques cinch the waist and the abdominals flatten.

The brace is the breath. Your core is supposed to be an autonomous unit that responds to what you’re doing without you thinking about it, and the deep abdominal wall fires a few milliseconds before you lift your arm rather than after it. You are not supposed to be holding it in all day. If you’re hypermobile, you very possibly are.

Here’s the walkthrough of how the two are meant to work together, which is the version most people have never been shown.

What that looks like in practice, and there’s no rep count anywhere in it:

– Lie on the floor, feet flat, knees bent. Put a hand just below your belly button, or a light weight or a water bottle four fingers below it, and make the thing move with your breath rather than by sticking your stomach out.

– Everything above the breastbone stays quiet. No shoulders lifting, no neck working. If your neck is doing your breathing then you’re chest breathing, and that’s a big part of why the tension up there never leaves.

– On the exhale, draw the belly button in and down, at about ten to twenty per cent of what you could do. Not in and up, which hollows everything out and jams it all up under the ribs. In and down, so the obliques cinch and the ribs soften.

– For the sides, put the web of your hand between the bottom rib and the top of your hip bone and say “ha”, or cough. You’ll feel the pressure push against your hand. Hold that pressure for a second, then let it go. Most people need this to find their lateral expansion, and most don’t get it first time.

– Then stop doing it consciously. Just breathe, and let the exhale bring everything back in on its own.

Nose breathing at rest, for all of it. Your nose has considerably more airflow resistance than your mouth, which is exactly what makes it useful when you’re sitting still and exactly what makes it useless when you’re going flat out. At low and moderate effort, keep it in the nose.

What we’re not doing is breath holds. Breath holding in a group of people who faint is not something to be practising off the back of a blog post, and slow, deep, controlled breathing in a supported position is the version with something behind it, as done during a tilt it takes a few beats off the rise in heart rate and brings the standing blood pressure down a little as well ^[20]. The people who ran that call it a proof of concept, and they’re right to.

Give this a fortnight before you add anything to it. That two weeks is our number, out of the studios, not out of a trial.

Then Foot Arch Control

Once the breathing isn’t falling apart the second you concentrate on something else, the first loaded thing we do is foot arch control, rather than squats or a bike or a walk round the block.

The logic is straightforward. Control the arch of your foot and you can control what the knee does, and control the knee and you can control the hip, all the way up. It’s also about as low demand as loading gets, and that matters enormously here, as you can do it kneeling, or sitting in a chair, without going anywhere near being upright and symptomatic.

Three points of contact: the fat pad under the big toe, the fat pad under the little toe, and the heel. Your shin bone works like a drive shaft, and rotating it inwards drops the arch while rotating it out lifts it. The easiest way in is weight distribution through the heel. Push about sixty per cent of the weight to the outside of the heel and watch the arch lift, then push it to the inside and watch it drop without fully collapsing. Toes up makes it much easier to see what’s going on.

Keep all three points down. If you’re hypermobile you can get into ranges most people can’t, and the contact under the little toe lifting off the floor is the thing that goes wrong for nearly everybody. Keep it down and you stop where you’re mechanically supposed to stop.

There’s a checklist for standing on one leg as well, and it works like mirror, signal, manoeuvre. Feet together, stand tall. Lean and load, so your head goes over the leg you’re standing on. Spread the toes and find your three points. Weight slightly to the outside of the heel. Gently grip the floor with the big toe. That locks the ankle off, and then you can push through the leg instead of hiking your hip up with your obliques.

The longer version of all of this is in our foot arch work.

Now, this is our method and we’re not going to dress it up as anything else, as nobody has trialled arch control as a POTS intervention. What we can say is that the standard programmes start lying down because upright is the problem, and this is lower demand than that while still being loaded and still being progressive, and it means the first thing somebody does isn’t the thing that makes them dizzy.

Then Unloaded Knee Work

The next step up is banded knee work, and the one we usually reach for is the reversed version.

A band about four fingers above the kneecap, banded leg forward. You come up like a small calf raise, then focus on getting the heel back down to the floor and letting the knee travel backwards. The band pulls you back, so the hamstrings have to work to keep the knee soft.

Start genuinely light. Nobody’s knee wants yanking into hyperextension, and the whole point of it is finding neutral, which is neither bent nor locked out, so most of your weight sits on the unbanded leg to begin with.

Use a mirror, or prop your phone up and watch yourself. Your brain is working out where your limbs are from the information it’s getting, and if that information is poor then the guess is poor, so give it something better: a tactile cue where you want the attention, and visual feedback so it can check its own answer against something outside itself. That’s the same reasoning behind the tactile cues we use throughout, and it’s why everything here is slow.

There’s no set and rep target for any of it, and that’s deliberate rather than vague. What we’re watching is whether the breathing holds. If you’re holding your breath through the movement, or your shoulders have started lifting, or you’ve drifted into fast shallow chest breathing, that’s the end of the set, whether that was four reps or forty. The movement is the easy part, and doing the movement without the breathing falling over is the actual exercise.

Then Upright, And Only Then

This is where the trial protocols pick up, and they’re worth following, as this is the part somebody has actually tested. Recumbent or semi recumbent cardio first, built up gradually, seated strength work added, and upright work brought in around the end of the second month ^[2]. Worth knowing though that the upright step arrives on a calendar rather than when you’re ready for it, and the only things gated on how you’re feeling are jogging and the stair stepper ^[2].

Two things to have running alongside it. Compression isn’t a comfort measure, as it genuinely changes the numbers, and an abdominal garment took around fifteen beats off both the standing heart rate and the rise on standing in a group of women with POTS ^[21]. That was a single day with everybody acting as their own control, so read it as a demonstration that the garment does something rather than as a long term result. If you’ve got compression garments, wear them for the session rather than just on bad days. Fluid and salt sit in the same category, as useful groundwork rather than as the intervention.

What volume won’t do is buy you fitness. Put a litre of saline into somebody with POTS and their fitness test comes out the same as it did on the placebo day ^[22]. It helps how you feel standing about, and it doesn’t substitute for training.

Post Exertional Malaise, And Why Pacing Comes First

Before any of the above, the question to settle is what happens to you a day or two after you exert yourself.

If your symptoms reliably get worse twelve to seventy two hours later, if rest doesn’t restore you, and if pushing through consistently sets you back rather than building you up, then graded progression is the wrong tool, and it’ll keep being the wrong tool no matter how gently you apply it. Test people with ME/CFS two days running and they can’t reproduce their own results on the second day ^[1]. A deconditioned body repeats the test, and a body with post exertional malaise doesn’t.

That distinction sits underneath everything above it, and the long version of it is in our guide to pacing, including how to find a baseline and how to use heart rate instead of how you feel. If you’ve got post exertional malaise, read that one first and treat this article as what comes afterwards.

The Short Version

Breathing first, for a fortnight. Then the arch, kneeling or sitting. Then unloaded knee work, with the breathing still running. Then upright, and only then.

That ordering is ours. It comes out of doing this with a lot of people over a lot of years, it’s built on mechanisms that are reasonably well described, and nobody has tested it against anything. In our opinion it’s the right order, which is a different claim from saying it’s proven.

And in our opinion, being sent straight to upright cardio with nobody ever checking how you’re breathing is how a good many people end up in the dropout column. The dropout column itself isn’t in dispute. Nobody has ever gone looking for the breathing.

The Fibro Guy

References

[1] Trimble, K.Z., Switzer, J.N. and Blitshteyn, S. (2024) ‘Exercise in Postural Orthostatic Tachycardia Syndrome: Focus on Individualized Exercise Approach’, Journal of Clinical Medicine. https://doi.org/10.3390/jcm13226747

[2] Fu, Q. and Levine, B.D. (2018) ‘Exercise and non-pharmacological treatment of POTS’, Autonomic Neuroscience. https://doi.org/10.1016/j.autneu.2018.07.001

[3] Levine, M., Shapiro, D., Hayburn, A., Cantrell, C. and Wilson, R. (2025) ‘A Survey-Based Study Examining Exercise in Postural Orthostatic Tachycardia Syndrome (POTS) Patients’, Cureus. https://doi.org/10.7759/cureus.84458