Breathing is one of the few body jobs that runs on its own and also takes orders. The same muscles that keep you breathing in your sleep will speed up the second your brain decides something is wrong.

That is useful when you need to run, but when the threat is an inbox or a hard talk, your breathing still speeds up, and you end up moving more air than a body sitting in a chair can use.

Most people never notice the change as it starts. The breaths get a little faster and shift up into the chest, the belly stops moving much, and the neck and shoulders start helping to lift the ribs.

In a lab, a plain memory task was enough to raise how much air people breathed. When a cash bonus hung on the result, they breathed more still, with no running or lifting at all.

Here is the part that feels unfair: quick, shallow breaths bring in plenty of air, but each one feels unfinished, so your brain reads the pattern as not getting enough air.

Researchers call that feeling air hunger, the urge to take a bigger breath. In lab tests it felt worse to people than hard breathing work, even when both felt equally strong.

So you top up by yawning, sighing or taking a gulp, and one good sigh can help, but a lot of them in a row keep your breathing uneven and keep your attention fixed on your chest.

That is the loop worth breaking. Anxiety makes the brain watch breathing more closely, and the closer you watch, the stranger each breath feels, so calming the system works better than fighting for air.

Your breathing has two drivers. One is automatic and keeps the air moving all night without a thought. The other is the part of the brain that handles feelings and plans, and it can take over at any moment.

When you are calm, most of the work is done by the diaphragm, the dome of muscle under your lungs. Your belly rises gently on each breath in, and the chest and shoulders barely move.

Stress shifts the work upward. Breathing gets faster and higher, and muscles in the neck and upper chest join in to lift the ribs. That style is built for a short burst of effort, not for an afternoon at a desk.

In one group of 62 people with breathlessness nobody could explain, the feeling of being short of breath showed up mainly in those whose breathing sat high in the chest. That is a small study with no comparison group, but it fits the pattern.

You do not have to move to breathe harder. In a lab study of 44 people doing a memory test at a screen, the amount of air they breathed went up as soon as the test began, and their breathing got faster too.

Nothing physical changed, because they were sitting still and thinking, yet the researchers found a small amount of over-breathing, the same pattern many people notice before a meeting or during a tense phone call.

A second lab study used timed mental sums. People sighed more often during the stressful sums, and during a calm task that needed steady focus their breathing became more rigid, with sighs piling up afterwards.

That second finding matters for anyone who works at a screen. Long stretches of tight focus can flatten your breathing without any stress at all, and the gulps and sighs arrive once you finally look up.

The size of the stress matters. In the memory study, breathing rose further when a cash bonus depended on doing well, so higher stakes meant more air moved, even with the same task on the screen.

Over years the link shows up as well. In a Swedish survey that followed 515 adults for nine years, people whose anxiety symptoms began during that time were about three and a half times as likely to become breathless.

That study is small for that kind of question and its margin of error is wide, so take the exact number loosely. The direction is the useful part: anxiety came first, and breathlessness did not predict later anxiety.

You do not need a big panic for this to feel bad. A low hum of pressure through the day is enough to keep your breathing a little fast and a little high, and a little is enough to make every breath feel short.

Posture is the first quiet factor, and in a study of 70 adults slumped sitting lowered how much air people could move compared with sitting upright, and standing did best of all. Hours bent over a laptop add up.

Coffee is the second, and in a small trial of seven adults a large dose of caffeine, about six or seven cups at once, raised resting breathing by about 40%. A tense week with extra coffee pushes in the same direction as the stress.

Hormones play a part too, since in women breathing runs a little faster in the second half of the menstrual cycle, when progesterone is higher, which can make stress breathlessness easier to set off on those days.

Then sleep. Short nights raise next-day anxiety, even when the sleep loss is modest, and a new medicine, low fitness or extra weight can all make you breathless too. Stress is rarely the only thing on the list.

Breathing patterns are learned, and that is good news, because learned patterns can change, and the most useful evidence comes from short daily practice rather than from one big fix.

In a month-long trial with 108 adults, five minutes a day of cyclic sighing, two breaths in through the nose and a long slow breath out, lifted mood and slowed resting breathing more than the same time spent meditating.

In a trial of 41 people with panic, four weeks of breathing training reduced their symptoms about as much as a talking therapy did. Only the breathing group brought their over-breathing back to normal, so there is more than one road in.

Those are small trials, and the second one was in people with a diagnosed panic problem. Still, they point the same way: a few minutes of slower, longer breaths out each day is a realistic place to start.

One popular belief deserves retiring: that stress breathlessness means you are running low on oxygen, or that it is all caused by breathing out too much carbon dioxide. The evidence does not support either version.

In a careful trial of 115 people, most who felt their symptoms while over-breathing felt the same things on a fake test that kept carbon dioxide normal. In their real attacks, the gas level barely dropped, and only after the feeling began.

The treatment evidence is thinner than you might expect as well. A major review of breathing exercises for this pattern found only one small, poorly reported trial, and said there was no credible evidence either way.

Even in large surveys, known causes such as age, weight, smoking and lung tests explained only about an eighth of who felt breathless. Much of this symptom is still not well understood, and that is worth saying plainly.