What mean arterial pressure is and why it beats systolic
Mean arterial pressure is the time-weighted average of the arterial pressure waveform over one complete cardiac cycle. It is not the midpoint between systolic and diastolic, and it is not the average of the two numbers on a cuff. It is the area under the pressure curve divided by the duration of the beat.
That distinction matters because organs are perfused by a pressure gradient sustained over time, not by a peak. Flow into the kidneys, the gut, the coronary bed during diastole, and the brain is driven by MAP minus the downstream pressure. Systolic pressure describes the highest instantaneous force the vessel wall sees — which is why it predicts stroke and aneurysm rupture well — but it says little about whether tissue is being perfused. Every vasopressor titration protocol in critical care is therefore written against MAP, not systolic.
Normal resting MAP in an adult is roughly 70 to 100 mmHg. Autoregulation keeps cerebral and renal blood flow nearly constant across a MAP band of about 60 to 150 mmHg in a healthy person; below that band, flow falls in direct proportion to pressure. In someone with chronic hypertension the whole autoregulatory range shifts upward, which is why a MAP of 65 that is entirely adequate for one patient can be ischaemic for another.
Where the one-third comes from
At a resting heart rate, systole occupies roughly one-third of each cardiac cycle and diastole the other two-thirds. If you approximate the pressure as sitting at its systolic value for that first third and at its diastolic value for the remaining two-thirds, the time-average is
MAP ≈ (1/3)·SBP + (2/3)·DBP = (SBP + 2·DBP) / 3
which rearranges to the familiar bedside form, DBP + PP/3. Both expressions are the same number; the second is easier to do in your head because you only divide the pulse pressure.
The weighting is the whole point. Diastolic pressure counts twice as much as systolic, so a 3 mmHg fall in diastole costs you 2 mmHg of MAP while a 3 mmHg fall in systole costs only 1. A patient whose diastolic is drifting down is losing perfusion pressure twice as fast as the systolic number suggests.
The approximation depends on the one-third/two-thirds split holding, and it does not hold when the heart is fast. As heart rate rises, diastole shortens far more than systole does, so systole takes up a growing share of the cycle and true MAP moves closer to systolic. Razminia and colleagues fitted an exponential form factor, 0.01·e(4.14 − 40.74/HR), to replace the fixed 1/3. Evaluate it at 60 bpm and you get 0.3185; at 100 bpm, 0.4179; at 140 bpm, 0.4695. The two methods cross at about 64 bpm — below that the exponential form returns a lower MAP than the one-third rule, above it a higher one.
Worked example: 120/80 with an ICP of 10 mmHg
Take a routine adult reading of 120/80 mmHg, heart rate 72, with an intracranial pressure monitor reading 10 mmHg.
- Pulse pressure. PP = 120 − 80 = 40 mmHg.
- One-third of it. 40 ÷ 3 = 13.33 mmHg.
- Mean arterial pressure. MAP = 80 + 13.33 = 93.33 mmHg. Check it the other way: (120 + 2 × 80) ÷ 3 = 280 ÷ 3 = 93.33. Same answer.
- Cerebral perfusion pressure. CPP = 93.33 − 10 = 83.33 mmHg, comfortably inside the 60–70 mmHg target band and above it.
- Proportional pulse pressure. 40 ÷ 120 = 33.3%, well above the 25% threshold.
- Rate correction. At 72 bpm the form factor is 0.01 × e(4.14 − 40.74/72) = 0.01 × e3.5742 = 0.3567. MAP = 80 + 0.3567 × 40 = 94.27 mmHg — less than one mmHg from the simple rule, because 72 bpm is close to the rate the one-third rule was built for.
Now repeat step 6 at a heart rate of 140. The form factor becomes 0.01 × e(4.14 − 0.291) = 0.4695, giving MAP = 80 + 0.4695 × 40 = 98.78 mmHg. The same cuff reading in a tachycardic patient carries about 5.5 mmHg more mean pressure than the one-third rule credits it with.
How to read the result
MAP 65 mmHg is the number to know. The Surviving Sepsis Campaign recommends an initial MAP target of 65 mmHg for adults with septic shock requiring vasopressors. That figure comes largely from the SEPSISPAM trial, which randomised septic shock patients to a 65–70 mmHg target or an 80–85 mmHg target and found no difference in 28-day mortality, with more atrial fibrillation in the higher-target group. The subgroup with chronic hypertension needed less renal replacement therapy at the higher target — evidence that the right MAP is patient-specific even where the protocol default is not.
CPP 60–70 mmHg is the corresponding number in neurocritical care. The Brain Trauma Foundation guidelines for severe traumatic brain injury recommend a CPP target in that band for survival and favourable outcomes, and specifically warn against pushing CPP above 70 mmHg with fluids and pressors because of the acute respiratory distress syndrome risk that carries. CPP is only as good as the ICP number feeding it, so treat the output as an estimate unless a ventricular drain or parenchymal monitor is in place.
Pulse pressure reads the stiffness of the arterial tree and the volume of the stroke. A resting adult pulse pressure of about 40 mmHg is typical. A wide pulse pressure — sustained above roughly 60 mmHg — is characteristic of stiff conduit arteries in older adults and of high-output states such as aortic regurgitation, thyrotoxicosis, anaemia and arteriovenous fistula. A narrow pulse pressure points the other way: low stroke volume from hypovolaemia, cardiac tamponade, severe aortic stenosis or cardiogenic shock.
Proportional pulse pressure normalises that judgement. Stevenson and Perloff described a proportional pulse pressure below 25% as a bedside marker of a cardiac index under 2.2 L/min/m² in advanced heart failure. It is a screening heuristic on a selected population, not a measurement — but it is a good reason to look harder at a patient whose numbers land below it.
MAP and pulse pressure for common blood pressure readings
| Reading | MAP (mmHg) | Pulse pressure | Proportional PP | Category |
|---|---|---|---|---|
| 70/40 | 50.0 | 30 | 42.9% | Shock range — MAP below target |
| 80/50 | 60.0 | 30 | 37.5% | MAP below the 65 mmHg target |
| 90/60 | 70.0 | 30 | 33.3% | Low-normal |
| 100/60 | 73.3 | 40 | 40.0% | Normal |
| 110/70 | 83.3 | 40 | 36.4% | Normal |
| 120/80 | 93.3 | 40 | 33.3% | Stage 1 (diastolic 80) |
| 130/80 | 96.7 | 50 | 38.5% | Stage 1 hypertension |
| 140/90 | 106.7 | 50 | 35.7% | Stage 2 hypertension |
| 160/100 | 120.0 | 60 | 37.5% | Stage 2 hypertension |
| 180/120 | 140.0 | 60 | 33.3% | Hypertensive crisis threshold |
Note the pattern in the first three rows: identical pulse pressures of 30 mmHg give very different MAPs, because MAP is anchored on diastole.
A cuff MAP and an arterial-line MAP are not the same measurement
An intra-arterial catheter integrates the pressure waveform and reports the true mean. An automated oscillometric cuff does the opposite of what most people assume: it detects the point of maximum oscillation, which is MAP, and then derives systolic and diastolic from that using proprietary algorithms. So the MAP shown on a monitor from a cuff is usually its most trustworthy number, while the MAP you calculate by hand from the displayed systolic and diastolic is a second-hand estimate of it.
Manual auscultation gives you no MAP at all — only Korotkoff endpoints — which is exactly the situation this formula was invented for.
Pitfalls and limits
- Averaging systolic and diastolic. The midpoint of 120 and 80 is 100; MAP is 93.3. The error grows with pulse pressure and always runs in the same direction — the midpoint overstates MAP whenever systolic exceeds diastolic.
- Using the one-third rule at high heart rates. The split between systole and diastole is not fixed. At 140 bpm the correction is worth several mmHg, and it always raises the estimate relative to the one-third rule above about 64 bpm.
- Reporting CPP without a real ICP. With no monitor you are subtracting an assumption. State it as such.
- Wrong cuff size. A cuff too small for the arm overestimates pressure; too large underestimates it. The bladder should encircle about 80% of the arm circumference.
- Treating 65 mmHg as universal. It is a starting target for septic shock in adults. Children, chronic hypertensives, patients with raised ICP and patients with spinal cord injury all have different appropriate targets.
- Ignoring the transducer level. An arterial line zeroed above or below the phlebostatic axis shifts every pressure reading by about 0.74 mmHg per centimetre of height error.
- Applying adult thresholds to children. Neonatal and paediatric MAP targets are age-dependent; a common neonatal rule of thumb sets the MAP floor near the gestational age in weeks.
Where MAP sits among the other bedside numbers
MAP is one term in the equation that actually governs flow. Rearranging Ohm's law for the circulation, MAP − CVP = CO × SVR: mean pressure is the product of cardiac output and systemic vascular resistance. A low MAP is therefore never a diagnosis on its own — it is either a pump problem, a volume problem or a tone problem, and the treatment differs completely. Measuring the output side directly is what the cardiac output Fick calculator does, and pairing it with MAP is how you separate vasodilatory shock from cardiogenic shock.
Perfusion pressure is only half of oxygen delivery; the other half is arterial oxygen content, which is where the PaO₂/FiO₂ ratio calculator and the A–a gradient calculator come in. When perfusion fails, the metabolic signature shows up as a raised lactate and a widened gap on the anion gap calculator, which is often the first objective sign that a MAP that looks acceptable is not adequate for that patient.
Two other resuscitation calculations lean on the same weight-based, target-driven logic: the Parkland burn fluid calculator for crystalloid volume, and the QTc interval calculator for the drug-safety side of the vasopressor and antiarrhythmic decisions that follow.
Finally, if you are working from a home cuff rather than a monitor, take at least two readings a minute apart after five minutes of quiet sitting and average them before calculating anything. A single reading carries enough beat-to-beat and white-coat variability to move MAP by more than the difference between any two rows of the table above.