How a Breath Actually Happens: The Lung's Two Circulations and the Pump That Fills It

The inversion: you make room, the sky does the rest

Air is not pulled into the lung. It is pushed in — by the atmosphere. Start with the only law that matters here: Boyle's law, which says that if you enlarge a closed container without adding gas, the pressure inside it falls. The thorax is that container, and the muscles of respiration are what enlarge it. When the diaphragm contracts and the ribs swing up and out, the volume of the chest rises. Because the lung is glued to the inside of the chest wall by the surface tension of the thin film of fluid in the pleural cavity, the lung is dragged open with it. Alveolar pressure — the pressure inside the air sacs — drops about one centimetre of water below atmospheric, and air flows down that gradient, through the nose, the larynx, and every branching generation of the bronchial tree, until pressures equalize. That is the whole of inspiration. Nothing sucked; a room was made and the sky filled it.

The numbers are worth carrying in your head, because every respiratory emergency is a story about them. At the end of a quiet breath out, everything is still: alveolar pressure equals atmospheric (call it zero), and intrapleural pressure sits at about minus five centimetres of water — permanently negative, because the elastic lung is always trying to shrink inwards while the springy chest wall is always trying to spring outwards, and the fluid seal between them keeps them married. During quiet inspiration intrapleural pressure falls further, to about minus eight, and alveolar pressure dips to minus one. That single centimetre of water is what moves half a litre of air. Puncture the seal — a knife, a broken rib end, a ruptured bulla — and the negative pressure vanishes, the lung recoils inward like a released balloon, and the patient has a pneumothorax. The lung never inflated itself; it was only ever being held open.

Two circulations through one organ

The lung is the only organ in the body with two entirely separate blood supplies — and they do completely different jobs. The first is the pulmonary circulation, and it does not feed the lung at all — it uses the lung. Blood leaves the right ventricle through the pulmonary trunk, which runs up and back for about five centimetres before dividing under the arch of the aorta, at the level of the sternal angle, into the right and left pulmonary arteries. The right pulmonary artery is longer and more horizontal, passing behind the ascending aorta and the superior vena cava to reach the right hilum; the shorter left pulmonary artery is tethered to the concavity of the arch by the ligamentum arteriosum, the fibrous remnant of the fetal ductus. Inside the lung each artery divides into lobar and then segmental arteries that travel alongside the bronchi, one artery shadowing each bronchopulmonary segment, all the way down to a capillary net so dense it wraps every alveolus in an almost continuous sheet of blood. Then the drainage reverses the rule: the pulmonary veins do not follow the bronchi at all — they run in the intersegmental septa, between the segments, which is precisely why a surgeon can shell out a single segment along its veins. Two veins emerge from each lung, and the four pulmonary veins — superior and inferior on each side — deliver oxygenated blood to the left atrium.

Two things about that system are extraordinary. First, it is the great exception to anatomical naming: here the arteries carry deoxygenated blood and the veins carry oxygenated blood, because "artery" means away from the heart and "vein" means towards it — never oxygen content. Second, it is a low-pressure system. The right ventricle generates roughly 25 over 8 millimetres of mercury, a mean of about fifteen, against a systemic mean near ninety — and yet it pushes exactly the same cardiac output, the entire five litres a minute, through the lungs as the left ventricle pushes through the whole body. It can do that because the pulmonary vessels are thin-walled, sparsely muscled and enormously compliant: when flow rises during exercise, they simply recruit collapsed capillaries and distend the open ones, so pressure barely climbs. As the plan of the cardiovascular system makes clear, a circuit that only has to cross one organ needs no head of pressure — and the right ventricle's thin wall is the honest confession that it never expected to work hard.

The second circulation is small, high-pressure, and it is the one that actually feeds the lung. The bronchial arteries are nutritive vessels: typically one on the right (often arising from the third posterior intercostal artery or a common trunk) and two on the left, springing directly from the descending thoracic aorta. They carry systemic-pressure oxygenated blood to the structures that pulmonary blood merely passes through without nourishing — the walls of the bronchi down to the respiratory bronchioles, the connective tissue and septa, the walls of the pulmonary vessels themselves, and the visceral pleura. Their drainage is the interesting half. The larger, more proximal bronchial veins drain conventionally into the azygos system on the right and the hemiazygos or accessory hemiazygos on the left. But the deep bronchial veins drain into the pulmonary veins — that is, oxygen-poor blood is dumped directly into the oxygen-rich stream on its way to the left atrium. Add the Thebesian veins of the heart doing the same thing, and you have the physiological shunt: the reason arterial oxygen tension is always a few millimetres of mercury lower than alveolar oxygen tension in a completely healthy person. The alveolar–arterial gradient is not a measurement error. It is anatomy.

The blood–air barrier: half a micrometre between you and the sky

Gas exchange happens across a membrane so thin it is difficult to believe it survives a lifetime. Three layers stand between the air in an alveolus and the haemoglobin in a red cell: the flat cytoplasm of a type I pneumocyte, a single fused basement membrane shared by the epithelium and the endothelium, and the capillary endothelial cell itself. Total thickness: about 0.2 to 0.6 micrometres — well under a thousandth of a millimetre. Type I pneumocytes are squamous, terminally differentiated cells that cannot divide; they make up only about forty per cent of the alveolar cells yet cover roughly ninety-five per cent of the surface, because each one spreads itself out like cling film. Multiply that over three hundred million alveoli and the total exchange surface approaches seventy square metres — a tennis court folded into a chest cavity, kept apart from the bloodstream by a sheet thinner than a soap bubble.

Scattered among them sit the cells that keep the whole structure from collapsing. Type II pneumocytes are plump, cuboidal cells — about sixty per cent of the alveolar cells but only five per cent of the surface — studded with lamellar bodies. They do two indispensable jobs. They are the stem cells of the alveolus, dividing to replace type I cells after injury. And they secrete surfactant, a phospholipid-and-protein film (dipalmitoylphosphatidylcholine is the key molecule) that spreads over the watery lining of every alveolus and lowers its surface tension. Why that matters is pure physics: Laplace's law says the pressure tending to collapse a bubble is proportional to its surface tension and inversely proportional to its radius, so a small alveolus would always empty itself into a big one. Surfactant defeats this by concentrating as an alveolus shrinks, dropping surface tension fastest exactly where the radius is smallest — it stabilizes the small sacs, prevents collapse at end-expiration, and dramatically reduces the work of breathing. Surfactant production begins around the twenty-fourth week of gestation and is not reliably adequate until about thirty-four. A baby born before that faces alveoli that collapse with every breath out and must be re-opened with every breath in — neonatal respiratory distress syndrome, the cruelest arithmetic in medicine. Guarding all of it is the third resident: the alveolar macrophage, or dust cell, crawling over the surface swallowing inhaled particles and bacteria, and in chronic heart failure gorging on leaked red cells until it is stuffed with haemosiderin and earns the name "heart failure cell".

Hypoxic pulmonary vasoconstriction: the lung's private reflex

Every other organ in the body responds to a fall in oxygen by dilating its arterioles — a starved tissue shouts for more blood, and gets it. The lung does the exact opposite. When the oxygen tension in an alveolus falls, the small pulmonary arteries supplying that region constrict. It sounds perverse until you remember what the lung is for. Blood arriving at an unventilated alveolus cannot be oxygenated there no matter how much of it turns up; sending more only wastes it. So the lung shuts the tap on the bad neighbourhood and diverts flow towards alveoli that are still being ventilated. Hypoxic pulmonary vasoconstriction is, in other words, a local matching device — the lung's way of keeping ventilation and perfusion pointed at the same alveoli.

The elegance has a price, and the price is paid when the hypoxia stops being local. In lobar pneumonia the reflex is a friend: one segment is flooded, its vessels clamp, and blood is redirected to healthy lung. But in chronic obstructive pulmonary disease, in diffuse fibrosis, in obstructive sleep apnoea, or simply at high altitude, every alveolus in both lungs is hypoxic at once. The reflex fires everywhere. The entire pulmonary vascular bed constricts and, over years, remodels — the thin-walled vessels thicken and muscularize permanently. Pulmonary vascular resistance rises, and the thin-walled right ventricle that was never built for pressure work must now generate it with every beat. It hypertrophies, dilates, and eventually fails: cor pulmonale — right heart failure caused by lung disease. The patient presents with raised jugular venous pressure, an enlarged tender liver and swollen ankles, and the swelling in the legs is, in the end, a consequence of what is happening in the alveoli.

The muscles of a quiet breath — and of a desperate one

Quiet breathing is almost entirely the work of one muscle, and the exhale costs nothing at all. In quiet inspiration the diaphragm descends about one and a half centimetres, and that descent alone accounts for roughly seventy-five per cent of the air moved. Its motor supply is the phrenic nerve, C3, C4 and C5 — "C3, 4, 5 keeps the diaphragm alive" — which is why a high cervical cord injury stops breathing outright while a lower one does not. The external intercostals do the rest, elevating the ribs in two distinct motions: the upper ribs swing like a pump handle, increasing the anteroposterior diameter of the chest, while the lower ribs swing like a bucket handle, increasing the transverse diameter. Together they enlarge the thorax in all three dimensions at once. Expiration at rest then requires no muscular work whatsoever: the muscles simply relax, and the stretched elastic tissue of the lung, the surface tension of the alveolar lining, and the recoil of the deformed chest wall drive the air back out. Breathing out is free. Every calorie of the respiratory cycle is spent on breathing in.

Push the system and a second tier of muscles is recruited. For forced inspiration the accessory muscles take over: sternocleidomastoid pulling up on the manubrium, the scalenes lifting the first and second ribs, and — when the arms are braced so that their usual origins become their new insertions — pectoralis minor pulling the third to fifth ribs upwards and serratus anterior hauling on the ribs from the side. For forced expiration, which is now an active, expensive event, the abdominal wall does the work: rectus abdominis, external and internal oblique and transversus abdominis contract, raise intra-abdominal pressure and drive the relaxed diaphragm upwards like a piston, while the internal and innermost intercostals depress the ribs. That is the anatomy of a cough, of a sneeze, of shouting across a street, and of every note a singer holds. It is also the anatomy of why a patient with a fractured rib cannot cough: the movement that clears the airway is precisely the movement that hurts.

Compliance and recoil: the two springs in tension

Compliance is simply how much volume you get for a given change in pressure — how easy the lung is to inflate. Elastic recoil is its opposite twin: how strongly the lung springs back. A healthy lung balances the two, and the whole pressure–volume relationship is a gentle S-shaped curve which is not the same going up as coming down, because it costs more pressure to open a collapsed alveolus than to keep an open one open. Disease pulls this balance apart in two opposite directions, and each direction produces its own kind of breathlessness. In pulmonary fibrosis the lung is stiff: compliance is low, recoil is high, and the patient must work hard to inhale but empties easily — small, fast, shallow breaths. In emphysema the elastin scaffolding has been destroyed: compliance is abnormally high, recoil is lost, and the lung inflates far too easily but cannot empty, because it is recoil that normally drives expiration. The airways collapse during exhalation, air is trapped, the chest becomes barrel-shaped, and the patient instinctively purses their lips to hold a back-pressure that splints those airways open. That pursed-lip breathing is not a habit; it is a person doing physics with their mouth. Both diseases, and the drugs aimed at them, are laid out in asthma versus COPD.

Matching air to blood: why the bases work hardest

Gas exchange is not one number but a ratio, repeated millions of times over: ventilation divided by perfusion, V/Q, in each individual alveolus. The ideal is about 0.8 overall, and the lung goes to considerable trouble to defend it. But gravity refuses to cooperate. Stand upright and both ventilation and perfusion are greater at the bases than at the apices — the bases are ventilated better because the weight of the lung above makes intrapleural pressure less negative down there, so the basal alveoli start out smaller, sit on the steep part of the compliance curve, and therefore expand more for the same effort; and they are perfused better because the pulmonary artery pressure is so low that a column of blood a few centimetres tall matters, and gravity simply drags blood downhill. The catch is that perfusion increases from apex to base more steeply than ventilation does. So the ratio itself varies: high at the apex (well ventilated relative to its meagre blood flow, V/Q around 3) and low at the base (richly perfused relative to its ventilation, V/Q around 0.6). Two classic facts fall out of this at once — apical tuberculosis favours the oxygen-rich, poorly perfused apex, while an inhaled foreign body or aspirated stomach contents settle in the lower lobes, as the lungs describes in full.

Push the ratio to either extreme and you get the two great failures of gas exchange. Ventilation without perfusion — V/Q approaching infinity — is dead space: air is delivered to alveoli that have no blood beside them, and the breath is simply wasted. Perfusion without ventilation — V/Q of zero — is a shunt: blood is delivered to alveoli that have no air in them, and it returns to the left atrium exactly as deoxygenated as it left. The clinical distinction between them is one of the most useful in medicine, because a shunt does not correct with supplemental oxygen. Turning up the oxygen enriches the air in alveoli that are being ventilated, but shunted blood never meets those alveoli at all — it bypasses them. Hypoxia that stubbornly refuses to improve on high-flow oxygen is a shunt until proven otherwise.