Introduction: the U-shaped risk curve
Coffee is the most studied nutraceutical on the planet. More than 36 000 peer-reviewed papers have been published on it, including two umbrella reviews and the largest cohort meta-analysis involving 1.2 million people. The conclusion is consistently the same: the dose-risk relationship is U-shaped.
Both zero consumption and more than 6 cups per day are associated with worse cardiovascular outcomes. The optimum, at which all-cause mortality decreases the most, is 2–4 cups of filtered coffee per day, in the first half of the day (Poole R, BMJ 2017, PMID 29167102; Ding M, Circulation 2014, PMID 24201300).
Key idea of the md_pereligyn protocol: coffee is not a “bad habit” and not an unconditional benefit. It is a nutraceutical with dose-dependent cardioprotection and an upper threshold. The goal is to identify your personal optimum, stay within it, and remove situations in which even a moderate dose becomes a trigger.
In this article, I discuss mechanisms, the real dose-effect relationship, the AFib paradox, effects on blood pressure, the difference between decaf and regular coffee, timing of intake, and patient categories for whom restriction or discontinuation provides measurable benefit.
🌀
The effect is not only mediated by caffeine
When people say “caffeine is beneficial/harmful,” that is a reduction. A cup of coffee contains more than 1 000 biologically active molecules, and caffeine is only one of them.
▸Chlorogenic acids are the largest class of polyphenols in coffee. They reduce postprandial glucose, inhibit α-glucosidase, increase insulin sensitivity, and act as antioxidants on the endothelium. ▸Trigonelline is an alkaloid that is partially converted into nicotinic acid (niacin) during roasting. It has a neuroprotective profile and improves the lipid profile. ▸Melanoidins are products of the Maillard reaction during roasting. They are antioxidants, microbiota modulators, and may reduce the glycemic response. ▸Diterpenes (cafestol and kahweol) are found in unfiltered coffee (French press, espresso, Turkish coffee, Scandinavian boiled coffee). They raise LDL and total cholesterol by 5–18% with regular consumption. ▸Caffeine is an antagonist of A1/A2A adenosine receptors. It is a direct cardiostimulant, increases BP by 5–10 mmHg in the acute phase, and the effect weakens as tolerance develops.
Decaf retains chlorogenic acids, trigonelline, and melanoidins. Therefore, most of its cardioprotection remains, and in large cohorts decaf is not inferior to regular coffee in reducing the risk of T2D, stroke, and all-cause mortality (Grosso G, Eur J Epidemiol 2016, PMID 27680332).
🌀
Dose-effect in numbers
Meta-analyses consistently show the same curve shape. The numbers vary by ±2–3% between cohorts, but the direction does not change.
•0 cups/day — 8% higher all-cause mortality versus 2–3 cups/day. •1 cup/day — 5% lower CV mortality, 8% lower T2D risk. •2–4 cups/day — 15% lower CV mortality, 25% lower T2D risk, 20% lower stroke risk, 20% lower depression risk, 30% lower Parkinson’s disease risk. •5 cups/day — cardioprotection plateaus; the risk of AFib and sleep disturbances begins to rise in sensitive individuals. •>6 cups/day — higher BP, AFib triggering in predisposed individuals, sleep disorders, increased anxiety; protection against T2D and all-cause mortality is lost.
A cup is defined as 240 mL of coffee brewed from 8–10 g of ground beans, corresponding to ~80–100 mg of caffeine. Espresso 30 mL = 60–80 mg of caffeine. This is convenient for calculating the daily dose.
🌀
The AFib paradox: trigger and protector at the same time
For many decades, coffee was considered proarrhythmic, and patients with paroxysmal AFib were advised to stop it completely. Modern data require this view to be revised.
In a large population cohort from UK Biobank (Chieng D, Eur J Prev Cardiol 2022, PMID 36190433), moderate coffee consumption reduced AFib risk by 4–8% compared with zero consumption. The effect persisted for 1–3 cups/day and disappeared at ≥5 cups/day.
The paradox is explained as follows: at the population level, the anti-inflammatory and insulin-sensitizing effects of coffee outweigh the direct adrenergic stimulus. But in an individual patient with paroxysmal AFib, the individual trigger threshold may be low — 1–2 cups. Therefore, the general recommendation to “stop coffee for everyone with AFib” is outdated, but monitoring personal triggers is mandatory.
Practical approach: a patient with paroxysmal AFib keeps a diary of episodes and caffeine intake (including tea and dark chocolate) for at least 4 weeks. If a correlation is found — reduction or discontinuation. If not — moderate consumption is acceptable.
🌀
Effects on blood pressure
Caffeine acutely raises BP by 5–10 mmHg systolic and 3–7 mmHg diastolic for 1–3 hours after intake. Chronic consumers develop partial tolerance, and 24-hour average BP increases less — by 1–4 mmHg.
This is clinically significant for resistant hypertension (BP >140/90 while taking three or more antihypertensive medications). In these patients, a trial discontinuation of caffeine for 2 weeks with BP measurement before and after is reasonable. A reduction in average BP by 5+ mmHg is a basis for permanent restriction.
In patients with controlled hypertension on therapy and a normal response to coffee, restriction is not indicated. A ban “for all hypertensive patients” is an outdated template.
🌀
Decaf vs regular
Decaf is not “second class.” It is an alternative that preserves most of the benefits and removes unnecessary risks in certain categories.
•Decaf — reduces the risk of T2D, stroke, liver events, cirrhosis, and all-cause mortality. For these outcomes, it is not inferior to regular coffee. It is suitable for patients with anxiety disorder, sleep disturbances, paroxysmal AFib (after testing), and pregnant women (>200 mg caffeine/day is not recommended). •Regular coffee — the same effects + additional protection against AFib at the population level, stronger neuroprotection against Parkinson’s and Alzheimer’s disease (at the population level), and a more pronounced effect on alertness and cognitive function.
The decaffeination method matters. Swiss water process and CO2 process do not leave organic solvents. Methylene chloride decaf is an outdated method and is now rarely found on the packaging of reputable producers.
🌀
Filtered vs unfiltered
This is a critical but underestimated fork. Diterpenes (cafestol and kahweol) are retained by a paper filter by 80–95%. In unfiltered coffee, they pass into the cup and, with regular consumption, raise LDL and total cholesterol.
▸Filtered — drip coffee maker with a paper filter, V60, Chemex, AeroPress with a paper filter. Diterpenes are minimal. ▸Unfiltered — French press, espresso, Turkish coffee, Scandinavian boiled coffee, moka pot, AeroPress with a metal filter. Diterpenes pass through fully.
For a patient with dyslipidemia or a family history of CAD, switching from espresso to filtered coffee at the same caffeine dose can reduce LDL by 8–14 mg/dL over 6–8 weeks. This is comparable to the effect of 5 g of soluble fiber per day.
🌀
Holistic protocol: how to drink it
### 1. Timing of intake
▸90–120 minutes after waking — the natural morning cortisol peak occurs 30–60 minutes after waking. Caffeine at this time masks the physiological rhythm and reduces adenosine receptor sensitivity. If intake is shifted to 1.5–2 hours after waking, caffeine is more effective at the same dose. ▸No later than 14:00 if sensitive — the half-life of caffeine is 5–6 hours. By 22:00, 25% of the morning dose is still in the blood, which is enough to impair sleep architecture in slow CYP1A2 metabolizers. ▸Not on an empty stomach in gastritis/GERD — coffee stimulates HCl secretion and relaxes the lower esophageal sphincter. Tolerability is better with food.
### 2. Dose
▸2–4 cups of filtered coffee/day — the optimum for most adults. ▸1–2 cups — for slow caffeine metabolizers (CYP1A2 *1F/*1F genotype), older adults, and pregnant women (≤200 mg/day). ▸Decaf without restriction in the absence of GI disorders.
### 3. What to add and what not to add
▸Cinnamon, cardamom, cocoa — polyphenol synergy, additional insulin sensitization. ▸Milk or plant-based milk — reduces the bioavailability of chlorogenic acids by approximately 10%, but does not eliminate the benefit. If you prefer it, drink it with milk. ▸Sugar and syrups — the main factor that neutralizes the benefit. 2 spoons of sugar turn a cup of coffee into a source of 12 g of simple carbohydrates. If insulin resistance is present, avoid sugar. ▸MCT oil — a working option for a ketogenic diet, but not for a general diet.
### 4. Nutraceutical partners
▸L-theanine 100–200 mg — partially offsets caffeine-induced anxiety while preserving the cognitive boost. Green tea contains it naturally. ▸Magnesium (glycinate) 200–400 mg in the evening — compensates for the diuretic effect of caffeine. ▸B vitamins — caffeine accelerates their excretion with regular consumption.
🌀
What does NOT work
▸Complete discontinuation of coffee as “health improvement” — for most people, this means losing a working nutraceutical without replacement. If coffee interferes (anxiety, AFib, insomnia), switch to decaf rather than stopping completely. ▸Energy drinks as a substitute — taurine + sugar + 200 mg caffeine without polyphenols. There is no cardioprotection, and the risk of arrhythmias and BP elevation increases. ▸“Bulletproof” — coffee with butter and MCT in large doses — increases morning caloric intake without proven benefit for most people; acceptable for a keto diet. ▸Coffee on an empty stomach in GERD/gastritis — provokes exacerbation. Changing the timing or taking it with food solves 80% of cases. ▸Instant coffee as the basis of intake — contains fewer chlorogenic acids than bean coffee and often more acrylamides (a Maillard reaction product from over-roasting). Not forbidden, but not the preferred choice.
🌀
Who should limit or stop it
•Paroxysmal AFib — individual trigger threshold; diary for 4 weeks. •Pregnancy — ≤200 mg caffeine/day (about 2 cups of espresso or 1 cup of 240 mL coffee). •Resistant hypertension — trial discontinuation for 2 weeks for assessment. •Chronic anxiety disorder, panic attacks — reduction or decaf. •Insomnia, sleep architecture disturbances — last cup no later than 12:00. •Active GERD or peptic ulcer disease — with food, not on an empty stomach, consider decaf. •Familial polyposis/long QT syndrome — individually with a cardiologist. •Dyslipidemia with LDL >190 mg/dL — switch to filtered coffee.
I perform personalized assessment of coffee intake patterns, taking into account metabolism genetics (CYP1A2), cortisol level, sleep architecture, and cardiovascular profile.
🌀
Conclusion
Coffee is a nutraceutical with dose-dependent cardioprotection. It is not a “bad habit,” but not an unlimited benefit either. The goal is 2–4 cups of filtered coffee in the first half of the day, with individual adjustment according to genotype, sleep, and cardiovascular status.
For most people, complete discontinuation means losing a working tool without replacement. Decaf is a practical alternative for anxiety, AFib, or pregnancy. Unfiltered brewing methods should be reconsidered in dyslipidemia.
If coffee works for you, drink it correctly. If not, there is a way to adapt the regimen rather than giving it up completely.
🌀
Sources
▸Ding M, Bhupathiraju SN, Satija A, et al. Long-term coffee consumption and risk of cardiovascular disease. *Circulation* 2014;129:643–659. PMID 24201300 ▸Poole R, Kennedy OJ, Roderick P, et al. Coffee consumption and health: umbrella review of meta-analyses. *BMJ* 2017;359:j5024. PMID 29167102 ▸Chieng D, Canovas R, Segan L, et al. The impact of coffee subtypes on incident cardiovascular disease, arrhythmias, and mortality. *Eur J Prev Cardiol* 2022;29:2240–2249. PMID 36190433 ▸Grosso G, Godos J, Galvano F, et al. Coffee, caffeine, and health outcomes: an umbrella review. *Eur J Epidemiol* 2016;31:1191–1205. PMID 27680332 ▸Cornelis MC, El-Sohemy A. Coffee, CYP1A2 genotype, and risk of myocardial infarction. *JAMA* 2006;295:1135–1141. PMID 16522833
Related articles: [Endothelium: the foundation of vascular health](/blog/endothelium-foundation-vascular-health), [Cholesterol without statins](/blog/kholesterin-bez-statinov).
🌀
FAQ
How many cups per day is safe? For most adults — 2–4 cups of filtered coffee (≈200–400 mg caffeine). Pregnant women — up to 200 mg/day. Slow CYP1A2 metabolizers, patients with paroxysmal AFib and resistant hypertension — individually, usually 1–2 cups or decaf.
Does coffee really cause arrhythmia? At the population level, moderate consumption reduces AFib risk (UK Biobank, 2022). But in an individual patient with pre-existing paroxysmal AFib, the individual trigger threshold may be low. The solution is a 4-week diary of episodes and intake, followed by individualized adjustment.
Is decaf as beneficial as regular coffee? For T2D, stroke, liver events, and all-cause mortality — yes, decaf is not inferior. For neuroprotection (Parkinson’s, Alzheimer’s) and AFib protection, the effect is weaker because part of the benefit works through caffeine. It is suitable for anxiety, insomnia, pregnancy, and AFib.
When is the best time to drink coffee? 90–120 minutes after waking, no later than 14:00 with normal sensitivity. Not on an empty stomach in GERD/gastritis. Tolerability is better with food, and the bioavailability of chlorogenic acids is preserved.
Do I need to give up coffee if I have hypertension? With controlled hypertension on therapy and a normal response — no. With resistant hypertension (BP >140/90 on 3+ medications) — trial discontinuation for 2 weeks with measurement. A BP reduction of ≥5 mmHg is a basis for restriction.
*This article is for informational purposes only and is not a substitute for professional medical advice. Discuss any nutraceutical, medication adjustment, or diagnostic procedure with your treating physician before starting.*
