Does Coffee Lower Your Risk of Death? Here Is What the Research Found
Multiple large-scale studies show that drinking 3–5 cups of coffee per day is linked to a 12–15% lower risk of death from all causes. Here is what the data actually says.
There is a cafe around the corner from where I live that I walk to most mornings. It is nothing special — plastic chairs, a slightly sticky counter, a woman behind the till who has never once smiled at me but makes the best flat white within a two-mile radius. I go because the coffee is good and because the walk gets me outside before 8am. I have never thought of it as a health intervention.
But apparently it might be.
Over the last two decades, more than a dozen large-scale studies — tracking hundreds of thousands of people over decades — have converged on a finding that surprised even the researchers running them: moderate coffee drinkers live longer than people who drink no coffee at all. Not by a trivial margin. By a measurable, statistically significant, dose-dependent margin. And the effect holds even after controlling for smoking, exercise, diet, alcohol, and body weight.
While you are reading this, people are dying from preventable causes — you can watch the toll climb in real time on our live death counter. Some of those deaths are from diseases that coffee, of all things, appears to offer a small degree of protection against.
The headline number
Across the largest and most rigorous studies, drinking 3 to 5 cups of coffee per day is associated with a 12 to 15% lower risk of death from all causes compared to drinking no coffee at all.
That is not a cure. It is not a miracle. But a 12-15% reduction in all-cause mortality is a genuinely significant effect — comparable to the benefit of eating fish twice a week or maintaining a healthy body weight. Our article on habits that add the most years to your life covers the full ranking, and coffee would sit comfortably in the middle of that list.
Here is a summary of the key findings across the major studies:
| Study | Participants | Duration | Optimal dose | All-cause mortality reduction |
|---|---|---|---|---|
| NIH-AARP Diet and Health Study (2012) | 402,260 | 13 years | 4-5 cups/day | 12% (men), 16% (women) |
| Nurses' Health Study / HPFS (2015) | 208,501 | 30 years | 3-5 cups/day | 15% |
| European EPIC Study (2017) | 521,330 | 16 years | 3+ cups/day | 12% (men), 7% (women) |
| UK Biobank (2018) | 498,134 | 10 years | 6-7 cups/day | 16% |
| Annals of Internal Medicine (2022) | 171,616 | 7 years | 2.5-4.5 cups/day | 29% (with sugar), 16-21% (without) |
| Korean NHIS Study (2021) | 190,000+ | 12 years | 3-4 cups/day | 12% |
These are not small studies. The combined participant count across the major coffee-mortality investigations exceeds two million people. This is not observational noise. This is a signal.
The dose-response curve
One of the most interesting findings is that the relationship between coffee and mortality is not linear. It follows a U-shaped curve — some coffee is better than none, but the benefit plateaus and may slightly diminish at very high consumption.
| Daily cups | All-cause mortality reduction (average across studies) |
|---|---|
| 0 cups | Baseline |
| 1 cup | 5-8% lower |
| 2 cups | 8-12% lower |
| 3 cups | 12-14% lower |
| 4-5 cups | 12-15% lower (optimal range) |
| 6-7 cups | 10-16% lower |
| 8+ cups | Benefit diminishes, some studies show no effect |
The sweet spot appears to be 3 to 5 cups. Beyond that, the benefit flattens rather than reverses — meaning even heavy coffee drinkers are not at increased risk in most studies. The UK Biobank study actually found the lowest mortality risk at 6-7 cups, though most other data points cluster around 3-5.
Worth noting: a "cup" in these studies is typically 8 ounces (about 237ml) of drip-brewed coffee containing roughly 95mg of caffeine. A large Starbucks is 20 ounces — so your single "cup" might count as two or three in research terms.
What coffee protects against
The mortality reduction is not evenly spread across all causes of death. Coffee appears to offer the strongest protection against specific diseases.
| Cause of death | Reduction at 3-5 cups/day | Key studies |
|---|---|---|
| Cardiovascular disease | 15-20% | EPIC, NIH-AARP, Korean NHIS |
| Type 2 diabetes | 25-30% (risk of developing) | Harvard meta-analysis of 28 studies |
| Liver disease (cirrhosis, liver cancer) | 40-65% | World Cancer Research Fund |
| Parkinson's disease | 25-30% | Pooled analysis of 13 studies |
| Colorectal cancer | 15-20% | EPIC, Nurses' Health Study |
| Stroke | 10-15% | NIH-AARP, Swedish Women's Study |
| Depression / suicide | 20-45% | Harvard School of Public Health |
| Alzheimer's and dementia | 20-65% (wide range) | CAIDE study, meta-analyses |
Cardiovascular disease
This is where the coffee data gets particularly interesting, because for decades doctors told heart patients to avoid coffee. The data says the opposite. Three to five cups per day is consistently associated with a 15-20% reduction in cardiovascular mortality. The European EPIC study — the largest of its kind — found that coffee drinkers had lower rates of heart disease, stroke, and heart failure than non-drinkers. Our deep dive on cardiovascular disease covers why it is the world's number one killer, and our heart attack data article shows the daily toll. The fact that something as simple as coffee appears to offer meaningful protection is worth paying attention to.
Diabetes
A 2014 Harvard meta-analysis of 28 prospective studies found that each additional cup of coffee per day was associated with a 7% lower risk of developing type 2 diabetes. Three cups cut the risk by roughly 21%. Five cups by 30%. The mechanism appears to involve improved insulin sensitivity and glucose metabolism. Diabetes has quietly become one of the world's top killers — our article on how diabetes became a top 10 cause of death covers the trajectory. Coffee will not cure diabetes, but the data on prevention is striking.
Liver disease
The liver data is the strongest of all. Coffee drinkers have dramatically lower rates of liver cirrhosis, fatty liver disease, and hepatocellular carcinoma (the most common form of liver cancer). A 2017 review in the BMJ found that drinking three cups per day was associated with a 40% lower risk of liver cancer compared to none. Four or more cups showed up to a 65% reduction. The liver, it turns out, really likes coffee.
Neurological diseases
Coffee drinkers have significantly lower rates of Parkinson's disease and appear to have some protection against Alzheimer's and other forms of dementia. The Parkinson's data is particularly robust — a pooled analysis of 13 studies found a 25-30% lower risk among regular coffee drinkers. Our article on dementia deaths doubling over 20 years covers why this matters. With no cure for either condition, any modifiable risk factor deserves attention.
Cancer
The cancer data is mixed but generally positive. Coffee is associated with lower rates of colorectal, liver, endometrial, and oral cancers. It does not appear to increase the risk of any cancer. The World Health Organization removed coffee from its list of possible carcinogens in 2016 after reviewing the accumulated evidence. You can see the full scale of daily cancer deaths on our dedicated article.
Decaf vs regular
Here is the part that genuinely surprised me. Decaf coffee shows nearly identical mortality benefits to regular coffee.
The NIH-AARP study found that decaf drinkers had a similar reduction in all-cause mortality as regular coffee drinkers. The 2017 EPIC study confirmed the same pattern. The 2022 Annals of Internal Medicine study found that both caffeinated and decaffeinated coffee were associated with lower mortality.
| Coffee type | All-cause mortality reduction (3+ cups/day) |
|---|---|
| Regular (caffeinated) | 12-15% |
| Decaffeinated | 11-14% |
| Instant | 11% |
This tells us something important: the benefit is not primarily about caffeine. It is about the other compounds in coffee — and there are over 1,000 of them.
Why coffee might protect you: the biological mechanisms
Coffee is not just caffeine and hot water. A single cup contains over 1,000 bioactive compounds, many of which have documented health effects.
| Compound | What it does | Amount per cup |
|---|---|---|
| Chlorogenic acid | Powerful antioxidant, anti-inflammatory, improves insulin sensitivity | 70-350mg |
| Caffeic acid | Antioxidant, may inhibit cancer cell growth | 15-50mg |
| Trigonelline | Neuroprotective, antibacterial | 40-110mg |
| Melanoidins | Antioxidant, prebiotic (feed gut bacteria) | Varies |
| Caffeine | Stimulates CNS, may protect dopamine neurons | 95mg (average) |
| Diterpenes (cafestol, kahweol) | Anti-inflammatory, may protect liver | 0.2-18mg (brew-dependent) |
| Magnesium | Supports heart rhythm, glucose metabolism | 7mg |
The chlorogenic acid content alone makes coffee one of the most antioxidant-rich beverages in the Western diet. For many people, coffee is their single largest source of antioxidants — not because it contains more than fruits and vegetables per serving, but because they drink it every single day.
The anti-inflammatory effects are particularly relevant. Chronic low-grade inflammation is now understood to be a root driver of cardiovascular disease, diabetes, cancer, and neurodegenerative disease. Coffee appears to reduce markers of systemic inflammation including C-reactive protein, interleukin-6, and TNF-alpha. Our article on how stress kills covers the inflammation pathway in detail — coffee may counteract some of the same mechanisms.
Who should NOT drink coffee
The data is not universally positive. There are groups for whom coffee carries genuine risks.
| Group | Concern | Recommendation |
|---|---|---|
| Pregnant women | Associated with increased risk of miscarriage and low birth weight above 200mg caffeine/day | Limit to 1 cup or switch to decaf |
| People with anxiety disorders | Caffeine can worsen panic attacks, insomnia, and generalised anxiety | Avoid or use decaf |
| People with uncontrolled high blood pressure | Acute caffeine intake can temporarily spike blood pressure | Discuss with GP |
| People with acid reflux / GERD | Coffee relaxes the lower oesophageal sphincter | May need to reduce or eliminate |
| People on certain medications | Caffeine interacts with lithium, some antibiotics, and heart medications | Check with pharmacist |
| People with sleep disorders | Even afternoon coffee can disrupt sleep architecture | Stop by 2pm or switch to decaf |
Sleep is worth emphasising. Coffee has a half-life of about 5 to 6 hours, meaning half the caffeine from a 3pm coffee is still in your system at 9pm. Poor sleep is independently associated with higher mortality — our sleep deprivation and death article covers the data. If coffee is helping your heart but wrecking your sleep, the net effect may not be positive. Timing matters.
Coffee consumption by country
Not every population drinks the same amount, and the variation is enormous.
| Country | Average cups per person per day | Notable |
|---|---|---|
| Finland | 4.0 | Highest per capita consumption on earth |
| Norway | 3.5 | Nordic pattern: filter coffee, black |
| Netherlands | 3.3 | Strong filter tradition |
| Sweden | 3.2 | "Fika" culture (coffee breaks) |
| Denmark | 3.0 | |
| Germany | 2.5 | |
| Italy | 2.4 | Espresso-based, smaller volumes |
| United States | 2.0 | Mix of drip and espresso |
| United Kingdom | 1.7 | Tea still dominant |
| Japan | 1.5 | Growing rapidly, canned coffee culture |
The Nordic countries dominate global coffee consumption. Finland, where the average adult drinks four cups a day, also has relatively high life expectancy and low cardiovascular mortality. Correlation is not causation — the Finns also have universal healthcare, low pollution, and high physical activity. But the pattern is consistent with the study data.
Japan's coffee culture is worth noting alongside our article on Japan's secret to living past 100 and their emphasis on green tea. The Japanese consume moderate amounts of both tea and coffee, which are the two most antioxidant-rich beverages humans drink.
Does it matter how you brew it?
Yes, actually. The brewing method affects both the chemical profile and the health impact.
| Brewing method | Diterpene content | Cholesterol effect | Notes |
|---|---|---|---|
| Paper-filtered (drip, pour-over) | Very low | Neutral | Removes cafestol, the compound that raises LDL |
| Espresso | Moderate | Slight increase | Short contact time limits extraction |
| French press / cafetiere | High | Can raise LDL by 6-8% | No paper filter, diterpenes pass through |
| Boiled (Turkish, Scandinavian) | Very high | Can raise LDL by 8-10% | Longest contact time |
| Cold brew | Moderate | Slight increase | Lower acidity, similar antioxidants |
| Instant | Very low | Neutral | Processing removes most diterpenes |
A 2020 study in the European Journal of Preventive Cardiology followed 508,747 Norwegians for 20 years and found that filtered coffee was associated with the lowest mortality — 15% lower than no coffee at all. Unfiltered methods (French press, boiled) were still associated with lower mortality than no coffee, but the benefit was smaller.
The mechanism is straightforward: paper filters trap cafestol and kahweol, two diterpene compounds that raise LDL cholesterol. Removing them keeps the antioxidant and anti-inflammatory benefits while eliminating the cholesterol concern. If you are watching your cholesterol levels — and obesity's link to the top causes of death makes this relevant for many people — filtered coffee is the better choice.
One more thing: adding sugar. The 2022 Annals of Internal Medicine study found that people who drank coffee with up to one teaspoon of sugar still had lower mortality than non-drinkers. But loading your coffee with syrups, whipped cream, and flavoured sauces is a different proposition. A Starbucks Frappuccino can contain 50-70 grams of sugar — more than a can of Coca-Cola. At that point you are drinking a milkshake, not a coffee, and the metabolic harm from the sugar likely cancels out whatever the coffee is doing.
What the data does and does not tell us
Almost all of the coffee-mortality research is observational. Researchers track large groups of people, ask them how much coffee they drink, then follow them for years and see who dies. This approach can identify associations but cannot prove causation.
It is possible — some sceptics argue likely — that people who drink moderate amounts of coffee are simply healthier in other ways that the studies did not fully capture. They may be wealthier. They may have more structured routines. They may exercise more. The studies control for known confounders, but there is always the possibility of residual confounding.
That said, the consistency of the findings across dozens of studies, in different populations, over decades, using different methodologies, makes it difficult to dismiss as coincidence. The biological mechanisms are plausible. The dose-response is logical. And the decaf data — showing similar benefits without caffeine — suggests the effect is real and related to the non-caffeine compounds in coffee.
No one is suggesting you start drinking coffee purely as a health intervention. But if you already drink it, the data suggests you should not feel guilty about it. Three to five cups of filtered coffee per day, without excessive sugar, is associated with meaningful health benefits across virtually every large study ever conducted.
The people who live the longest — in Blue Zones from Sardinia to Okinawa to Nicoya — drink coffee, tea, or both. Our article on what the longest-living people eat covers their full dietary pattern. Coffee fits comfortably within it. Check how your overall lifestyle habits stack up against the data with our life expectancy tools.
While you have been reading this — roughly 11 minutes — approximately 660 people have died worldwide. One every second. Many of them from cardiovascular disease, cancer, and diabetes — the very conditions that coffee appears to offer some protection against. See the numbers climb in real time on our live death counter.
Data sources: NIH-AARP Diet and Health Study, Nurses' Health Study, Health Professionals Follow-Up Study, European Prospective Investigation into Cancer and Nutrition (EPIC), UK Biobank, Annals of Internal Medicine (2022), BMJ (2017 liver meta-analysis), European Journal of Preventive Cardiology (2020), Harvard T.H. Chan School of Public Health, World Health Organization. This article is for educational purposes only and does not constitute medical or dietary advice.
You might also like
How Marriage Affects Your Lifespan — The Data Is Surprisingly Clear
Married people live 7–15 years longer than single people on average. Here is what the research actually shows about marriage, mortality, and why the gap is so large.
Japan's Secret to Living Past 100
Japan has more centenarians than any country on earth — over 90,000 people aged 100 or older. Here is what the data reveals about why the Japanese live longer, and what the rest of the world can learn from it.
The 5 Habits That Add the Most Years to Your Life (Backed by Data)
Adopting five simple habits can add 12-14 years to your life expectancy, according to a landmark Harvard study. Here is the data on what works, how much each habit matters, and why it is never too late to start.