How the Heart Pumps, and What Happens in a Heart Attack

Press two fingers against the inside of your wrist, below the thumb, and wait. Each soft knock you feel is a wave of pressure that left your chest a fraction of a second earlier, when a fist-sized muscle squeezed about 70 milliliters of blood, a little under a third of a cup, into an artery already full of it. It will do this roughly a hundred thousand times today, two and a half billion times in a lifetime, and it never rests for longer than about half a second.

That muscle is two pumps in one body, with four one-way doors and its own electrical clock. This page takes it apart: how it fills and empties, what times each beat, and what goes wrong when the arteries that feed it clog. Here is a real-sized model, beating at a third of its resting speed. Turn it, and cut it open:

Inside are four hollow chambers: two thin-walled atria on top that collect blood, and two thick-walled ventricles below that push it out. The wall on the left of the picture, the patient's right, is thin. The wall on the patient's left is almost three times thicker. That difference is the first clue to how the whole thing works.

Two pumps#

Blood makes two journeys, and each has its own pump. The right side of the heart sends it through the lungs, where it drops carbon dioxide and picks up oxygen, and back to the left side, which sends it out to everything else and back to the right. Change the heart rate and watch how much blood moves each minute:

At rest the heart moves about 5 liters a minute, close to all the blood in an adult body, so a red cell gets back to where it started about once a minute [1].

Because the loop is closed, the pumps are in series: every drop the right side sends into the lungs has to come out through the left. Over any stretch of minutes they must move exactly the same amount. How exactly?

Nothing wires the two pumps together, yet they match to a fraction of a percent. The trick is that each pump pushes harder when it is filled more. Model the circulation as two reservoirs, the body's veins and the lungs, with a pump moving blood from each to the other. Weaken the left pump to and watch the lungs:

For a few seconds the weak left pump falls behind, and blood comes into the lungs faster than it leaves. Their pressure rises, which fills the weak pump more, so it pushes harder, and within seconds the outputs match again, at a lower flow and a higher lung pressure. Nothing measured anything. This is the Frank–Starling law, named for Otto Frank, who saw it in frog hearts, and Ernest Starling, whose group measured it in dog hearts in 1914 [2].

Push it down to 30% and the lung pressure needed to keep up passes about 20 mmHg, where fluid is forced out of the capillaries into the air sacs. That is the breathlessness of left heart failure, and it starts with nothing more than this arithmetic.

Four rooms, four doors#

The real heart packs both pumps into one muscle, sharing a wall. Turn it and point at anything to name it; each name here lights its part. The superior and inferior vena cava bring blood from the body into the right atrium, then the right ventricle sends it up the pulmonary trunk to the lungs. The pulmonary veins bring it back to the left atrium, and the left ventricle pushes it out through the aorta:

It isn't laid out left and right as you might draw it. The right ventricle wraps around the front of the left one like a crescent, and the left ventricle makes the left side and the pointed tip, the apex, which points down, forward, and to the left. That is why you feel the beat just left of the breastbone. Now cut it open and compare the two ventricles' walls:

Both sides move the same 70 mL per beat, but not against the same resistance. The lungs are close and their vessels wide, so the right ventricle only reaches about 25 mmHg. The left ventricle has to reach about 120, nearly five times as much [1]. Muscle grows to match its work, so the left wall is about a centimeter thick and the right only 3 to 5 mm.

The four valves are rings of tissue with two or three thin flaps, the leaflets. A valve has no muscle and no nerve. It opens when the blood behind it is at a higher pressure than the blood in front, and shuts the instant that reverses. Here is the mitral valve, between the left atrium and ventricle. Drag the pressure difference across it:

With the atrium even 2 or 3 mmHg higher, the leaflets swing open and blood pours through. Once the ventricle's pressure climbs past the atrium's, the same blood pushes them shut. Unheld, they would blow through like an umbrella turning inside out; the chordae tendineae, cords tied to the papillary muscles, stop them just as they meet.

Now tear the cords on one leaflet. It flips back, and every squeeze sends a jet the wrong way. A ruptured cord, one complication of a heart attack, can flood the lungs this way.

One heartbeat, slowed down#

Every beat runs through the same sequence, and the valves split it into parts. Here is the heart cut open again with a clock and the volume of the left ventricle. A resting beat lasts 0.8 seconds. Move into it and watch which valves are open:

The atria contract first, adding the last fifth or so to ventricles that are already mostly full. Then the ventricles squeeze, and for about 50 milliseconds all four valves are shut: the inlets because the ventricle's pressure is now above the atrium's, the outlets because it is still below the artery's. The ventricle squeezes a sealed bag, and its pressure climbs from about 10 to 80 mmHg with no change in volume. This is isovolumic contraction.

When the pressure passes the aorta's, the aortic valve opens and blood leaves for about 0.28 seconds: ejection, 70 of the 123 mL. Then the muscle relaxes with every valve shut until its pressure drops below the atrium's, the mitral valve opens, and the ventricle refills, fast at first, as the relaxing wall almost sucks blood in1The right side runs the same sequence at about a fifth of the pressure, its valves moving a few hundredths of a second out of step with the left's..

Draw the pressures on one time axis and the whole beat fits on one chart, a Wiggers diagram2Named after the American physiologist Carl Wiggers, who recorded these pressures in the early twentieth century.. First, a question about the moment the aortic valve shuts.

Every event on the chart is a crossing of two curves. The ventricle rising through the atrium shuts the mitral valve, and rising through the aorta opens the aortic valve; falling back reverses them. The atrium's pressure has two humps: its own squeeze, then a rise while the mitral valve is shut and the lungs keep pouring blood in.

The aorta's pressure only sags from about 120 to 80 mmHg between beats, with a small notch where the valve caught the backflow. And the ECG at the bottom runs ahead of the mechanics: the QRS comes a few hundredths of a second before the ventricle's pressure starts to rise.

Why the pressure never drops to zero#

The ventricle pushes for only a third of each beat, yet blood flows through your organs all the time. The reason is the aorta, an elastic tube thick with stretchy elastin. During ejection it balloons out and stores about half of the stroke volume, and between beats it recoils and pushes it on3Engineers call this a windkessel, German for "air chamber", after the air-filled domes on old fire pumps that smoothed the strokes into a steady jet [4].. Change how stretchy the arteries are, , and watch the two numbers of a blood pressure reading:

With young arteries the pressure barely swings. Make them stiff, as they become with age, and the same heart gives about 130 over 50: the aorta can't absorb the surge, so the top number rises, and it can't push between beats, so the bottom number falls. With rigid pipes the flow would nearly stop between beats. The bottom number matters: as we'll see, the heart's own muscle is fed mostly between beats.

A beat as a loop#

Plot the ventricle's pressure against its volume instead of against time, and the beat becomes a loop it goes around once per beat, counterclockwise. Watch the dot:

Along the bottom the ventricle fills at low pressure; up the right side it squeezes sealed; across the top it ejects; down the left side it relaxes, sealed again. The loop's width is the stroke volume, and divided by the volume at the start of the squeeze it is the ejection fraction, the number on most heart scan reports: 57% here, and usually between about 55% and 75% in healthy hearts [3]. Shade the loop: its area is the work done on the blood each beat, about one joule.

Three different things change the loop. Preload is how much the ventricle is filled before it squeezes. Afterload is what it pushes against, mostly the resistance of the small arteries. Contractility is how hard the muscle squeezes from a given filling, raised by adrenaline and lost when muscle dies. Fill harder , raise the resistance , or change the strength :

Fill it more and the loop stretches right: the ventricle starts bigger and ejects more. This is Frank–Starling from inside: stretched muscle fibers pull harder [2]. Raise the resistance and the loop grows taller and narrower. Adrenaline steepens the red line and the ventricle empties further.

Whatever the filling and resistance, the loop's top left corner lands on the same red line, and only the muscle's strength moves it. Suga and Sagawa found this in dog hearts in 1973 [5]: the ventricle behaves like an elastic chamber whose stiffness rises and falls through the beat. Every heart figure on this page runs on that idea.

Now weaken the muscle, as a large heart attack does, and the stroke volume and pressure fall by a third. The body answers by retaining salt and water, which raises the filling: weak but compensated. Some stroke volume comes back, at a filling pressure above 20 mmHg that the lungs pay for.

The spark that times each beat#

A heart lifted out of the body keeps beating for a while with no nerves attached. Its rhythm comes from a few thousand cells in the right atrium's wall, the sinus node. They never rest at a steady voltage: after each beat, positive ions leak in until the voltage creeps up to a threshold, the cell fires, and the creep starts again. Turn up adrenaline or the vagus nerve and watch the slope of the creep:

With no nerve input the node fires about 100 times a minute in a young adult4Measured by blocking both sets of nerves with drugs; the rate falls with age [7].. At rest the vagus nerve brakes it to 60 or 70; adrenaline steepens the creep and can push it past 180. The sinus node only sets the tempo.

Beneath it, a ventricular cell's action potential has a long flat plateau, about a quarter of a second, while calcium flows in and the cell contracts. Until it ends, the cell can't fire again, so heart muscle can't be held in a cramp: every squeeze is followed by a forced relaxation long enough to refill.

From the sinus node the wave must reach billions of muscle cells in the right order: the atria, a deliberate pause, then the ventricles from the inside out and from the tip up. Here it is on a section of the heart, slowed ten times. Move through the first :

The wave crosses both atria in about a tenth of a second5A fast band, Bachmann's bundle, carries it across to the left atrium.. Then it hits a ring of fibrous tissue that doesn't conduct at all. The only way through is the AV node, which is deliberately slow, about 5 centimeters a second [1]. The wave waits there for nearly a tenth of a second while the atria finish emptying.

Past the node it runs fast down the His bundle in the septum and through the Purkinje fibers lining both ventricles, at 2 to 4 meters a second, so the ventricles light from the inside of the tip and the wave crawls outward through the wall. In isolated human hearts the inner left ventricle lit at several spots within milliseconds, and the base near the valves came last [8]. That order squeezes blood up toward the exits, like toothpaste from the bottom of the tube.

Reading the wave from the skin#

An ECG picks up this wave from electrodes on the arms and legs. At any instant, the wavefront between fired cells and resting ones carries current, like a battery. Add up every little battery and you get one arrow, the heart vector, which swings as the wave moves; a pair of electrodes reads how much it points along the line between them. Move through the beat, , and watch it trace lead II, right arm to left leg:

The small first bump, the P wave, is the atria. The flat stretch is the wave held in the AV node, too few cells to register. The tall QRS complex is the ventricles, the arrow pointing down and left toward the thick left ventricle. During the plateau every cell is charged, there is no boundary, and the trace sits flat: the ST segment. Remember that flat line; it is where a heart attack shows itself.

Then the T wave, as the ventricles recover. It points the same way as the QRS because the cells activated last recover first: recovery runs back the other way, and two reversals cancel.

A standard ECG looks at the arrow from twelve directions, six of them in this section's plane. Choose one, :

From the left leg's side, lead II, the QRS is tall. From the right shoulder, aVR, everything is upside down. Read together, the leads tell which way the arrow swings, which is how an ECG can say roughly where a heart attack is happening.

Keeping up with a sprint#

A fit young adult running flat out needs three or four times the resting 5 liters a minute. Every lever pulls at once: adrenaline speeds the sinus node and strengthens the squeeze, working leg muscles pump blood back so the heart fills harder, and the muscles' arterioles open wide. Push the effort, :

Most of the gain comes from the rate, which nearly triples, while the stroke volume rises by a third: 5 to almost 18 liters a minute in this model. Endurance athletes go much further, mostly with bigger ventricles6Swedish champions in a classic study averaged 36 liters a minute at peak, with stroke volumes near 190 mL [9]., which is also why their resting rate is low.

The heart's own blood supply#

The heart is half a percent of body weight but uses about a tenth of its oxygen at rest, and it already pulls about three quarters of the oxygen out of the blood passing through it, so the only way to get more is more flow [10]. The wall is far too thick to be fed from the blood inside. It has its own arteries, the coronary arteries, which leave the aorta just above the aortic valve and run over its surface like a crown:

The left anterior descending artery feeds the front of the left ventricle and most of the septum, more muscle than any other. The circumflex wraps round the left side, and the right coronary artery round the right and the back. Each is 3 to 4 mm wide, and together they carry about 225 mL a minute at rest, 4 to 5% of the output [1].

There is a catch. When the ventricle squeezes, the pressure inside its wall is as high as in the arteries feeding it, highest in the inner layers, and the squeeze pinches the branches shut. So the left ventricle is fed mainly between beats. Speed up the heart, :

At rest the heart is relaxed nearly 60% of each beat, and about three quarters of the flow arrives then. At 170 beats a minute each squeeze is a third shorter, but the gaps shrink to under a third of their resting length. The heart works hardest exactly when its feeding window is shortest. A healthy artery copes by widening its small branches, raising flow four to five times [10]. A narrowed one can't.

How an artery clogs#

Atherosclerosis starts decades before anyone notices. Where an artery's lining is irritated, cholesterol-carrying particles slip into the wall and get stuck. White blood cells move in to clear them, gorge on the fat, and die there, and the debris collects into a soft core under a fibrous cap of scar tissue [11]. Grow a plaque, of the wall's area, in a 3.5 mm coronary artery:

For a long time nothing seems to happen. The wall grows outward around the plaque, so the channel keeps its size and an angiogram, which shows only the channel, looks normal. Glagov's group found the channel holds until the plaque fills about 40% of the area inside the wall [12]. Past that, the plaque takes the channel itself.

Flow through a tight spot loses pressure twice: to friction in the narrow channel, and to the turbulent jet beyond it. At rest the small vessels downstream open wider to make up for it. Narrow the artery, , and compare the flow at rest with the most it can deliver:

Resting flow holds until the artery is about 80% narrowed in this model, but the maximum starts falling at around 50%, and by 75% there is almost no reserve. Gould found the same pattern in dogs [13]. This is stable angina: chest pain on exertion, when the muscle asks for more than the artery can bring, fading within minutes of rest. A warning, not yet damage.

Most heart attacks don't come from the tightest narrowing slowly closing. They come from a plaque that breaks. Only the fibrous cap separates the blood from the fatty core, and inflammation can thin it from beneath; caps that ruptured were mostly under 65 micrometers, thinner than a hair [14]. Thin the cap, , and let the pulse pound on it:

A thick cap takes the pounding indefinitely. A thin one splits at the shoulder, where the stress is highest, and the blood meets the core, the body's strongest trigger for clotting. Platelets stick within seconds, fibrin knits them together, and the clot grows until it seals the channel. A plaque that narrowed the artery by a third has blocked it in minutes.

In men who died suddenly with a fresh coronary clot, about two thirds of the clots sat on a ruptured plaque and the rest on a plaque whose surface had worn away [15]. Aspirin makes platelets less sticky, so the clot grows more slowly and may not close the artery. That is why it is given at once for a suspected heart attack7In a trial of 17,187 patients, a month of aspirin started on the first day cut deaths from heart and vessel disease over five weeks from 11.8% to 9.4% [16]..

A heart attack, minute by minute#

When a clot seals a coronary artery, the muscle downstream stops squeezing within seconds and burns through its oxygen within a minute or so. For about the first 20 minutes the damage is reversible. After that cells start to die, and the death spreads as a wavefront, which Reimer and Jennings mapped by tying off coronary arteries in dogs for different times [17]. Choose when the artery is reopened, after it closed:

The inner layer dies first: it is squeezed hardest, sits at the end of the supply, and relies most on flow between beats. From there death advances outward and sideways over hours. At 40 minutes a little over a third of the muscle at risk was dead; at 3 hours, more than half; by a day, most of it. Cardiologists say it in three words: time is muscle.

The ECG sees this in the first minutes. Injured cells can't hold their resting voltage, so current leaks between them and healthy muscle, and the flat ST segment lifts off the baseline. Run the electrical model with an injured region at the tip of the left ventricle, and move through time since the artery closed, :

Within minutes the T wave grows tall and peaked. Then the ST segment lifts in the lead facing the injury. ST elevation of at least 1 mm in two neighboring leads, a little more in some chest leads, marks a STEMI: the artery is fully closed and every minute counts [18]. Over the next day the dead muscle falls silent, the R wave shrinks (in a real heart often leaving a deep Q wave), and the T wave inverts. Weeks later the ST segment is flat again, but the lost R wave stays.

Opening the artery#

So open the artery as early as possible. Clot-dissolving drugs, given into a vein, switch on the body's own enzyme for breaking up fibrin. Or a cardiologist threads a wire from the wrist or groin into the blocked artery and pushes the clot and plaque aside with a balloon8Andreas Grüntzig first did this in a coronary artery in 1977 [19]. Stents followed a decade later [20]., which today carries a stent, a metal mesh tube that stays as scaffolding. Deploy one, :

The balloon, at around ten atmospheres, doesn't remove the plaque: it crushes it into the wall, and the stent holds it there9Most stents release a drug for months that stops scar tissue growing back through the mesh. When several arteries are badly diseased, surgeons may instead bypass the blockages with a leg vein or a chest-wall artery..

How much timing matters shows in 22 trials of clot-dissolving drugs with about 50,000 patients [21]. Move the delay from the first symptom, :

The first hour is worth twice the second, and after about three hours the benefit levels off, because most of the muscle that could be saved is already dead. Where a catheter lab is near, a balloon beats drugs, with targets of 90 minutes from first medical contact in the United States [22] and 120 minutes from diagnosis in Europe [23]. The longest delay is usually the first: the time before someone calls for help.

What to take away#

The heart is two pumps kept in step by their own filling, with valves that obey pressure and a wave that pauses at the AV node, then lights the ventricles from the inside out. It feeds itself between beats through arteries that can collect plaque for decades unseen. Most heart attacks come when a thin cap tears and a clot seals the channel; then the muscle dies in a wave, and how much is saved depends on how fast the artery is opened.

So if you or someone near you has chest pain or pressure lasting more than a few minutes, especially with breathlessness, sweating, nausea, or pain spreading to the arm, jaw, or back, call your local emergency number at once. Don't drive yourself, and don't wait to see if it passes.

References#

1. J. E. Hall and M. E. Hall, Guyton and Hall Textbook of Medical Physiology, 14th ed., Elsevier (2021).
2. S. W. Patterson and E. H. Starling, “On the mechanical factors which determine the output of the ventricles”, Journal of Physiology 48 (1914), 357–379.
3. A. M. Maceira, S. K. Prasad, M. Khan, and D. J. Pennell, “Normalized left ventricular systolic and diastolic function by steady state free precession cardiovascular magnetic resonance”, Journal of Cardiovascular Magnetic Resonance 8 (2006), 417–426.
4. N. Westerhof, J.-W. Lankhaar, and B. E. Westerhof, “The arterial Windkessel”, Medical & Biological Engineering & Computing 47 (2009), 131–141.
5. H. Suga, K. Sagawa, and A. A. Shoukas, “Load independence of the instantaneous pressure-volume ratio of the canine left ventricle and effects of epinephrine and heart rate on the ratio”, Circulation Research 32 (1973), 314–322.
6. N. Stergiopulos, J.-J. Meister, and N. Westerhof, “Determinants of stroke volume and systolic and diastolic aortic pressure”, American Journal of Physiology 270 (1996), H2050–H2059.
7. A. D. Jose and D. Collison, “The normal range and determinants of the intrinsic heart rate in man”, Cardiovascular Research 4 (1970), 160–167.
8. D. Durrer, R. T. van Dam, G. E. Freud, M. J. Janse, F. L. Meijler, and R. C. Arzbaecher, “Total excitation of the isolated human heart”, Circulation 41 (1970), 899–912.
9. B. Ekblom and L. Hermansen, “Cardiac output in athletes”, Journal of Applied Physiology 25 (1968), 619–625.
10. D. J. Duncker and R. J. Bache, “Regulation of coronary blood flow during exercise”, Physiological Reviews 88 (2008), 1009–1086.
11. P. Libby, J. E. Buring, L. Badimon, et al., “Atherosclerosis”, Nature Reviews Disease Primers 5 (2019), 56.
12. S. Glagov, E. Weisenberg, C. K. Zarins, R. Stankunavicius, and G. J. Kolettis, “Compensatory enlargement of human atherosclerotic coronary arteries”, New England Journal of Medicine 316 (1987), 1371–1375.
13. K. L. Gould, K. Lipscomb, and G. W. Hamilton, “Physiologic basis for assessing critical coronary stenosis”, American Journal of Cardiology 33 (1974), 87–94.
14. R. Virmani, A. P. Burke, A. Farb, and F. D. Kolodgie, “Pathology of the vulnerable plaque”, Journal of the American College of Cardiology 47 (2006), C13–C18.
15. A. P. Burke, A. Farb, G. T. Malcom, Y. H. Liang, J. Smialek, and R. Virmani, “Coronary risk factors and plaque morphology in men with coronary disease who died suddenly”, New England Journal of Medicine 336 (1997), 1276–1282.
16. ISIS-2 Collaborative Group, “Randomised trial of intravenous streptokinase, oral aspirin, both, or neither among 17,187 cases of suspected acute myocardial infarction: ISIS-2”, Lancet 2 (1988), 349–360.
17. K. A. Reimer, J. E. Lowe, M. M. Rasmussen, and R. B. Jennings, “The wavefront phenomenon of ischemic cell death. 1. Myocardial infarct size vs duration of coronary occlusion in dogs”, Circulation 56 (1977), 786–794.
18. K. Thygesen, J. S. Alpert, A. S. Jaffe, et al., “Fourth universal definition of myocardial infarction (2018)”, European Heart Journal 40 (2019), 237–269.
19. A. R. Grüntzig, Å. Senning, and W. E. Siegenthaler, “Nonoperative dilatation of coronary-artery stenosis: percutaneous transluminal coronary angioplasty”, New England Journal of Medicine 301 (1979), 61–68.
20. U. Sigwart, J. Puel, V. Mirkovitch, F. Joffre, and L. Kappenberger, “Intravascular stents to prevent occlusion and restenosis after transluminal angioplasty”, New England Journal of Medicine 316 (1987), 701–706.
21. E. Boersma, A. C. P. Maas, J. W. Deckers, and M. L. Simoons, “Early thrombolytic treatment in acute myocardial infarction: reappraisal of the golden hour”, Lancet 348 (1996), 771–775.
22. P. T. O'Gara, F. G. Kushner, D. D. Ascheim, et al., “2013 ACCF/AHA guideline for the management of ST-elevation myocardial infarction”, Circulation 127 (2013), e362–e425.
23. R. A. Byrne, X. Rossello, J. J. Coughlan, et al., “2023 ESC Guidelines for the management of acute coronary syndromes”, European Heart Journal 44 (2023), 3720–3826.