What caffeine does to the heart
In people with high blood pressure, 200 to 300 mg of caffeine raised it by about 8/6 mmHg for at least 3 hours. In one trial, days with coffee brought more early beats of the ventricles.
Caffeine in dietary doses raises blood pressure.[1] In 5 trials in people with high blood pressure, 200 to 300 mg of caffeine raised it by about 8 mmHg systolic (the top number) and 6 mmHg diastolic (the bottom number). The rise began in the first hour and lasted at least 3 hours.[2] Only about half of regular caffeine users become tolerant to this rise.[1] In a trial of 100 adults, days with coffee brought more early beats of the ventricles (154 against 102 a day), but not significantly more early beats of the atria.[3]
A heart at rest
The heart has four chambers: two atria on top and two ventricles below. The left ventricle is the main pump. It sends blood through the aortic valve into the aorta and on to the body.[4] The sinus node, the heart's own pacemaker, sits where the superior vena cava joins the right atrium. It fires about 70 times a minute.[4]
One beat fills the left ventricle and pushes blood out
Stroke volume is the blood a ventricle pushes out in one beat. At rest it is 60 to 100 ml in healthy adults. Stroke volume times heart rate gives cardiac output, the blood the heart pumps in a minute.[5] In the model, the left ventricle fills to 126 ml and pushes out 74 ml. At 70 beats a minute, that is 5.2 litres a minute.[5]
Caffeine reaches the blood
The gut absorbs caffeine quickly and completely. The blood level peaks 15 to 120 minutes after a dose.[6] In healthy adults, half of it is gone after about 4 hours, with a range of 2 to 8 hours.[7]
Caffeine seems to act mainly in two ways. It blocks the receptors of adenosine, a substance that lowers the activity of nerve cells. It also blocks enzymes called phosphodiesterases.[6]
Blood pressure rises within the hour
In 5 trials in people with high blood pressure, 200 to 300 mg of caffeine raised systolic pressure by 8.1 mmHg and diastolic pressure by 5.7 mmHg on average. The rise began in the first hour and lasted at least 3 hours.[2] In the model, 200 mg in someone who rarely drinks coffee takes the pressure from 115/83 to about 122/90 mmHg at 1.5 hours. This is a model value, not a measurement. This rise of 6.4/7.0 mmHg lies within the 95% confidence intervals of these trials.[2]
An extra beat comes early
A premature ventricular contraction (PVC) is a beat that starts early, from a spot inside a ventricle. PVCs are common in the general population.[8] In the model, the extra beat comes before the left ventricle has filled: it holds only 60 ml and pushes out 19 ml. In 18 patients whose extra beats were triggered with a pacing wire, the pulse was weaker the earlier the extra beat came.[9]
A pause, then a larger beat
After the extra beat, the next normal beat comes after a pause. In the model, the left ventricle fills to 137 ml before this beat, against 126 ml in a normal beat. In 10 healthy people, the beat after the pause filled 8% more, and its ejection fraction, the share of blood the ventricle pushes out, was 14% higher. The longer filling was one reason for the stronger beat.[10]
How much does caffeine raise blood pressure?
In people with high blood pressure, a single dose of 200 to 300 mg raised it by about 8/6 mmHg for at least 3 hours.[2] The same review found 3 studies in which people drank coffee for 2 weeks. Their blood pressure did not rise, compared with a caffeine-free diet or with decaffeinated coffee.[2]
Drag on the chart to choose an hour. The daily drinker line is based on the half of daily caffeine users who did not become tolerant: their blood pressure stayed about 2.8/2.2 mmHg higher during waking hours. The model keeps the pulse at 70 a minute.
In 7 cohort studies of people with high blood pressure, habitual coffee drinking was not linked to a higher risk of heart and blood vessel disease.[2] A 2026 statement of the American Heart Association says that the short-term and the long-term effects of caffeine on the heart and vessels are often different.[11]
Does the rise fade in daily coffee drinkers?
In about half of them. In a 4-week trial, 85 men and women took capsules with 0, 300 or 600 mg of caffeine a day, one of these amounts each week.[1] In a 4-week trial of the same design by the same team, half of the people became tolerant to the rise in blood pressure. In the other half, caffeine still raised it. In these people, peripheral resistance, the resistance of the small arteries to blood flow, rose step by step as the daily dose went up.[12]
On the test day at the end of each week, people took three doses of 0 or 250 mg. In the group with low tolerance, the week of 300 mg a day ended with the 250 mg doses. Blood pressure during waking hours was then 2.8/2.2 mmHg higher than in the week with placebo only. During sleep, systolic pressure was 2.3 mmHg higher.[1]
Does coffee cause extra heartbeats?
In one trial, coffee brought more PVCs but did not significantly change early beats of the atria.[3] The trial, called CRAVE, gave 100 adults a heart monitor that recorded their ECG for 14 days. Each day a text message told them, at random, to drink caffeinated coffee or to avoid caffeine.[3]
The main outcome was premature atrial contractions (PACs), early beats of the atria: 58 a day on coffee days against 53 on days without caffeine. This difference was not significant (rate ratio 1.09, 95% CI 0.98 to 1.20). PVCs were not the main outcome. There were 154 a day with coffee against 102 without (rate ratio 1.51, 95% CI 1.18 to 1.94). On coffee days, people also walked 1058 more steps and slept 36 minutes less a night.[3]
Aorta Left ventricle, in mmHg. The coloured band marks the extra beat. Drag on the chart to move through the four beats.
Earlier studies did not find more extra beats or arrhythmias with caffeine or coffee:
- A 2016 analysis of 7 studies that gave people caffeine found no change in PVCs over 24 hours (relative risk 1.00, 95% CI 0.94 to 1.06).[13]
- In 1388 older adults (mean age 72) who wore a heart monitor for 24 hours, coffee, tea and chocolate habits were not linked to the number of PACs or PVCs.[14]
- In 386,258 adults in the UK Biobank, followed for 4.5 years on average, each extra daily cup of coffee was linked to a 3% lower risk of a new arrhythmia. This study observed habits and did not test coffee.[15]
The 2026 American Heart Association statement says that trials in regular coffee drinkers showed more PVCs with caffeinated coffee. Data on high doses, such as those in energy drinks, are limited and generally suggest harm to the heart and vessels.[11] Doctors look into PVCs when they cause symptoms or complications, such as heart muscle disease caused by PVCs.[8]
How much caffeine is considered safe?
For healthy adults, the European Food Safety Authority (EFSA) found no safety concerns with single doses up to 200 mg, about 3 mg per kg of body weight. It found none with up to 400 mg a day spread over the day.[16] In pregnancy, up to 200 mg a day raises no safety concerns for the unborn child. A single dose of 100 mg may change sleep in some adults, mainly close to bedtime.[16]
A 200 ml cup of filter coffee has about 90 mg of caffeine, a 60 ml espresso about 80 mg, and a 355 ml can of cola about 40 mg.[16]
The US Food and Drug Administration (FDA) gives 400 mg a day as an amount not generally linked to negative effects in most adults. People differ a lot in how sensitive they are to caffeine and how fast they clear it.[17] Signs of too much caffeine include a faster heart rate, palpitations and high blood pressure. About 1200 mg taken quickly can cause toxic effects such as seizures.[17]
The EFSA assessment did not cover people with a disease or medical condition, or caffeine taken with medicines.[7] The FDA advises talking to a health care provider about caffeine in case of a medical condition or a medicine. The same applies in pregnancy, when planning a pregnancy, during breastfeeding, and with signs of too much caffeine.[17]
In short
- In people with high blood pressure, 200 to 300 mg of caffeine raised blood pressure by about 8/6 mmHg, from the first hour for at least 3 hours.[2]
- In 3 studies in people with high blood pressure, drinking coffee for 2 weeks did not raise it.[2]
- Only about half of regular caffeine users become tolerant to the rise.[1]
- In the CRAVE trial, coffee days brought 154 PVCs a day against 102, and no significant change in PACs.[3]
- In 1388 older adults, coffee, tea and chocolate habits were not linked to more extra beats.[14]
- EFSA found no safety concerns for healthy adults with single doses up to 200 mg and up to 400 mg a day.[16]
What this figure leaves out
The heart and its vessels come from the Human Reference Atlas, version 1.3: the female heart and the female blood vessels, shared under the CC BY 4.0 licence.[18][19] The atlas has no conduction system. The sinus node label marks the wall of the right atrium, where the node sits; the node itself is not drawn.[4]
The motion and the numbers come from a small model of the heart and the circulation. A normal beat gives 115/83 mmHg in the aorta. The left ventricle fills to 126 ml and pushes out 74 ml, inside the normal range of 60 to 100 ml.[5] The pulse stays at 70 a minute, about the rate of the sinus node.[4] The model does not change the heart rate, although too much caffeine can raise it.[17]
With caffeine, the heart itself looks almost the same: fewer than 1 in 100 pixels of the picture change. Caffeine shows in the hours chart and in the numbers. In the model, caffeine raises the resistance of the small arteries, as peripheral resistance rose in people who did not become tolerant.[12] The blood level of caffeine peaks at 45 minutes, inside the 15 to 120 minutes seen in people,[6] and halves every 4 hours.[7]
For someone who rarely drinks coffee, the model gives about 122/90 mmHg at the peak, a rise of 6.4/7.0 mmHg. This is a model value. It lies within the 95% confidence intervals of the trials in people with high blood pressure: 5.7 to 10.6 mmHg systolic and 4.1 to 7.4 mmHg diastolic.[2] The line for daily drinkers uses the 2.8/2.2 mmHg that one trial measured over waking hours in people with low tolerance.[1]
The extra beat is a separate switch. In the model, caffeine does not cause it. It makes the third of every four beats a PVC that starts in the ventricle muscle.[8] The model leaves out early beats of the atria, any change in heart rate, and the differences between people, such as how fast they clear caffeine.[17] For education; not for diagnosis.
Sources (19)
- Caffeine tolerance is incomplete: persistent blood pressure responses in the ambulatory setting. Farag NH, Vincent AS, Sung BH, Whitsett TL, Wilson MF, Lovallo WR Am J Hypertens, 2005. PMID 15882556
- The effect of coffee on blood pressure and cardiovascular disease in hypertensive individuals: a systematic review and meta-analysis. Mesas AE, Leon-Muñoz LM, Rodriguez-Artalejo F, Lopez-Garcia E Am J Clin Nutr, 2011. PMID 21880846
- Acute Effects of Coffee Consumption on Health among Ambulatory Adults. Marcus GM, Rosenthal DG, Nah G, Vittinghoff E, Fang C, Ogomori K, et al. N Engl J Med, 2023. PMID 36947466
- Anatomy, Thorax, Heart. Rehman I, Rehman A StatPearls, 2023. PMID 29262022
- Physiology, Stroke Volume. Khalil B, Foster C, Launico MV StatPearls, 2026. PMID 31613466
- Pharmacology of Caffeine. Institute of Medicine (US) Committee on Military Nutrition Research NCBI Bookshelf: Caffeine for the Sustainment of Mental Task Performance, National Academies Press, 2001. www.ncbi.nlm.nih.gov
- EFSA explains risk assessment: Caffeine. European Food Safety Authority, 2015. www.efsa.europa.eu
- Premature ventricular complexes. Yap R, Calian D, Wright M, Rinaldi A, Whitaker J Clin Med (Lond), 2026. PMID 42342153
- Influence of left ventricular filling profile during preceding control beats on the occurrence of pulse deficit caused by ventricular premature contractions. Otsuji Y, Kisanuki A, Toda H, Kuroiwa R, Murayama T, Matsushita R, et al. Eur Heart J, 1993. PMID 7691601
- Mechanics of postextrasystolic potentiation in normal subjects and patients with valvular heart disease. Wisenbaugh T, Nissen S, DeMaria A Circulation, 1986. PMID 2423268
- Caffeine and Cardiovascular Disease: A Scientific Statement From the American Heart Association. Marcus GM, Hu FB, van Dam RM, Cornelis MC, Dewland TA, Kang J, et al. Circulation, 2026. PMID 42473796
- Hemodynamic mechanisms underlying the incomplete tolerance to caffeine's pressor effects. Farag NH, Vincent AS, McKey BS, Whitsett TL, Lovallo WR Am J Cardiol, 2005. PMID 15904654
- Effect of caffeine on ventricular arrhythmia: a systematic review and meta-analysis of experimental and clinical studies. Zuchinali P, Ribeiro PA, Pimentel M, da Rosa PR, Zimerman LI, Rohde LE Europace, 2016. PMID 26443445
- Consumption of Caffeinated Products and Cardiac Ectopy. Dixit S, Stein PK, Dewland TA, Dukes JW, Vittinghoff E, Heckbert SR, et al. J Am Heart Assoc, 2016. PMID 26813889
- Coffee Consumption and Incident Tachyarrhythmias: Reported Behavior, Mendelian Randomization, and Their Interactions. Kim EJ, Hoffmann TJ, Nah G, Vittinghoff E, Delling F, Marcus GM JAMA Intern Med, 2021. PMID 34279564
- Caffeine. European Food Safety Authority. www.efsa.europa.eu
- Spilling the Beans: How Much Caffeine is Too Much?. US Food and Drug Administration, 2024. www.fda.gov
- 3D Reference Organ for Heart, Female, v1.3. Human Reference Atlas, NIH HuBMAP (CC BY 4.0), 2026. lod.humanatlas.io
- 3D Reference Organ for Blood Vasculature, Female, v1.3. Human Reference Atlas, NIH HuBMAP (CC BY 4.0), 2026. lod.humanatlas.io
For education; not medical advice. Talk to a doctor before you start, stop or change a supplement.