Coffee and Liver Disease: What a 354,957-Person Study Found, Including the Decaf Result
followed for a median of 13 years, people drinking five or more cups of coffee a day had a
32% lower risk of cirrhosis, 47% lower risk of liver cancer and 42% lower risk of liver-related
death than non-drinkers. Benefits appeared from one to two cups and were described as strongest
around three to four. The result held for decaffeinated coffee too, which means caffeine
is not the explanation. This is an observational study, and the researchers state plainly that it does not
prove coffee prevents liver disease.
Coffee and liver health has been a recurring finding in nutrition research for years, usually in
studies too small or too short to move anyone’s opinion. This one is neither.
Researchers at Cedars-Sinai, with Hyunseok Kim, MD, MPH, PhD as corresponding author, used the UK
Biobank to follow 354,957 people who had no cirrhosis or liver cancer at the start, tracking
new diagnoses through linked health records over a median of 13 years. The work was published in
Clinical Gastroenterology and Hepatology (DOI: 10.1016/j.cgh.2026.04.035).
Two things make it worth reading past the headline: the study looked inside the body as well as at
diagnoses, and it tested decaf.
The numbers
Comparing people drinking five or more cups a day against people drinking none:
- Cirrhosis: 32% lower risk
- Liver cancer: 47% lower risk
- Liver-related death: 42% lower risk
Those are the figures being quoted everywhere, and they are the top of the dose range. The more useful
part of the finding sits lower down: associations were present at one to two cups a day, and
Cedars-Sinai describes the benefit as appearing strongest around three to four cups per day.
That is worth noticing, because the headline invites people to read this as an instruction to drink five
cups. The institution’s own summary does not say that. If you already drink two or three cups, this study
does not tell you to drink more.
The decaf result changes what the study means
Here is the finding most coverage buries.
Cedars-Sinai reports that similar protective associations were observed for caffeinated and
decaffeinated coffee. That single line does more work than any of the percentages.
If decaf tracks with the same outcomes, then caffeine is not the mechanism. Whatever is
going on involves something else in the bean, and coffee contains a great many other compounds: chlorogenic
acids, diterpenes such as cafestol and kahweol, melanoidins formed during roasting, and more.
Practically, it also removes the usual trade-off. People who avoid caffeine for blood pressure, anxiety,
reflux, pregnancy or sleep have generally read findings like this as not applying to them. On this evidence,
they may.
What the scans and blood work showed
Most large cohort studies can only tell you who was diagnosed with what. This one also looked at the
organ itself.
Coffee drinkers showed lower liver fat, lower liver iron, less fibrosis and less inflammation on
MRI. On blood proteins, they had higher levels of proteins tied to healthy liver function and lower
levels of proteins linked to scarring and inflammation.
This matters for interpretation. A diagnosis is a late event, arriving years after the biology has moved.
Finding the same direction of effect in imaging and in circulating proteins, in the same people, makes it
less likely that the diagnosis pattern is purely an artefact of who ends up in a hepatology clinic.
It is still not proof of cause. It is a coherent picture rather than a single number.
Why this is not a recommendation, and the researchers say so
The study is observational. Nobody was assigned to drink coffee. Researchers recorded what
people already did and followed what happened.
Cedars-Sinai states directly that the study does not prove that coffee itself prevents liver disease,
and that coffee should complement, not replace, established strategies. That is an unusually
clean caveat from an institution promoting its own research, and it deserves to travel with the percentages.
The specific worry with any coffee-and-health cohort is reverse causation. People with early
liver problems often feel unwell, lose their taste for coffee, or are told to cut back. If quietly declining
health reduces coffee intake, then low intake will look like a cause of disease when it is partly a
consequence of it. Long follow-up reduces this problem but does not remove it.
There is also the familiar confounding question. Coffee drinking travels with other habits, and heavy coffee
drinkers differ from non-drinkers in ways that are not all measurable.
What this does and does not support
Supported: A consistent association across diagnoses, imaging and blood proteins, in a very
large cohort, over a median 13 years, that does not depend on caffeine.
Not supported: That drinking coffee will protect your liver. That five cups is a target.
That coffee offsets alcohol, obesity or viral hepatitis, which are the things that actually drive the diseases
measured here. That anyone who dislikes coffee should start.
The established strategies the researchers refer to are unglamorous and well known: limiting alcohol,
managing weight and metabolic health, vaccination against hepatitis B, treatment for hepatitis C, and caution
with medications and supplements that stress the liver. Nothing in this study displaces any of them.
A note on how to read the percentages
A 47% lower risk of liver cancer sounds enormous, and in relative terms it is. But liver cancer is uncommon
in the general population, so a large relative reduction applies to a small absolute risk. Cutting a small
number by half still leaves a small number.
This is not a criticism of the study, which reports what it found. It is a caution about the arithmetic that
disappears when a finding becomes a headline. Relative risk reductions in cohort studies almost always look
more dramatic than the change they represent for one person.
Frequently Asked Questions
How much coffee was linked to the biggest benefit?
The 32%, 47% and 42% figures come from comparing people drinking five or more cups a day against
non-drinkers. However, Cedars-Sinai describes the benefit as appearing strongest around three to four cups per
day, and associations were already visible at one to two cups. The honest summary is that more was not clearly
better past a moderate intake, and the study gives no reason for a two-cup drinker to increase.
Does decaf count?
According to this study, yes. Similar protective associations were observed for both caffeinated and
decaffeinated coffee. That is the most practically useful line in the paper, because it means the association
does not depend on caffeine and may therefore apply to people who avoid caffeine for blood pressure, anxiety,
reflux, pregnancy or sleep reasons.
Does this prove coffee protects the liver?
No, and the researchers say so. This is an observational study, and Cedars-Sinai states it does not prove
that coffee itself prevents liver disease. The main alternative explanations are reverse causation, where early
undiagnosed liver disease reduces coffee intake rather than the other way round, and confounding by other
habits that travel with coffee drinking.
What did the liver scans show?
Coffee drinkers had lower liver fat, lower liver iron, less fibrosis and less inflammation on MRI. Their
blood also showed higher levels of proteins tied to healthy liver function and lower levels of proteins linked
to scarring and inflammation. Finding the same direction in imaging, blood proteins and diagnoses makes the
overall pattern more coherent than a diagnosis count alone would be.
Should I start drinking coffee for my liver?
This study does not support that. It observed people who already drank coffee; it did not test what happens
when a non-drinker starts. If you dislike coffee, or avoid it for a medical reason, nothing here is a reason to
change that. If you already drink it and enjoy it, this is reassurance rather than instruction.
Does coffee cancel out alcohol or a poor diet?
Nothing in this study supports that idea. Alcohol, obesity, metabolic dysfunction and viral hepatitis are the
principal drivers of the outcomes measured here, and the researchers explicitly frame coffee as complementing
rather than replacing established strategies. Treating a cup of coffee as an offset against the things that
actually damage the liver misreads the finding completely.
How big is the study, really?
Large: 354,957 participants with no cirrhosis or liver cancer at baseline, followed for a median of 13 years
using linked health records. Size is a genuine strength here because the outcomes are relatively rare and a small
cohort would not accumulate enough cases to measure anything. Size does not, however, fix the limitation of
observational design; a very large observational study is still observational.
What is a cup, in a study like this?
Self-reported intake in a cohort of this kind reflects what participants said they drank, not a standardised
measured volume, and a home mug is not a café serving. This is one reason to treat the dose figures as broad
categories rather than precise thresholds, and another reason not to read five or more cups as a target.
Sources
- Cedars-Sinai Health Sciences University, Study: New Biological Clues Behind Coffee’s Benefit to Liver
Health, newsroom summary of research published in Clinical Gastroenterology and Hepatology
(DOI: 10.1016/j.cgh.2026.04.035), corresponding author Hyunseok Kim, MD, MPH, PhD —
Institutional release - UK Biobank, the cohort resource used for the analysis —
ukbiobank.ac.uk
Participant number, median follow-up, the risk reductions by intake,
the imaging and blood protein findings, the decaffeinated coffee result and the causation caveat are all taken
from the Cedars-Sinai institutional summary of the published study, checked August 2026.
