The study is real, large and interesting. It also found no significant association between yogurt and colorectal cancer overall — which is the sentence almost every headline left out.
Read the finding precisely. Published in Gut Microbes in 2025, the study drew on the Nurses’ Health Study and the Health Professionals Follow-up Study, covering 132,056 participants with 3,079 colorectal cancers during follow-up. People habitually eating two or more servings of yogurt a week had a 20% lower rate of Bifidobacterium-positive proximal colon cancer.
The crucial qualifier: that association appeared only in tumours that were Bifidobacterium-positive — about 31% of cases. The other 69% showed nothing. And there was no significant association between yogurt consumption and colorectal cancer incidence overall. 🥣
- What the study was
- The actual finding
- The sentence the headlines dropped
- Why subgroup findings need caution
- Why this is still good science
- Why proximal colon cancer specifically
- What observational means here
- Should you eat more yogurt?
- The thing that actually saves lives
- What genuinely changes your risk
- Symptoms that need a doctor now
- Frequently asked questions
- Your checklist
What the study was
This is a serious piece of work from two of the best-known cohorts in nutritional epidemiology.
| Element | Detail |
|---|---|
| Published | Gut Microbes, 2025;17(1):2452237 |
| Title | Long-term yogurt intake and colorectal cancer incidence subclassified by Bifidobacterium abundance in tumor |
| Data sources | Nurses’ Health Study and Health Professionals Follow-up Study |
| Participants | 132,056 |
| Colorectal cancers documented | 3,079 |
| Tumours analysed for | Bifidobacterium abundance |
| Design | Observational, prospective cohort |
What makes this design unusual and clever: rather than just asking “does yogurt relate to bowel cancer,” the researchers went back to the tumour tissue itself and classified cancers by whether Bifidobacterium — a bacterial genus found in yogurt — was present. That’s a genuine attempt to test a mechanism rather than just fish for a correlation.
The actual finding
| Outcome | Result |
|---|---|
| Colorectal cancer overall | No significant association with yogurt intake |
| Bifidobacterium-positive proximal colon cancer | 20% lower rate with ≥2 servings/week |
| Bifidobacterium-negative tumours | No association |
| Share of cases that were positive | About 31% |
| Share that were negative | About 69% |
So: a 20% reduction, in a subgroup representing roughly three in ten colorectal cancers, with nothing detectable in the other seven — and nothing detectable when you look at colorectal cancer as a whole.
The sentence the headlines dropped
“Researchers found no significant association between yogurt consumption and overall colorectal cancer incidence.”
That’s the finding a reader most needs, and it’s the one that vanished from most coverage — because “yogurt linked to 20% lower cancer risk” is a headline and “yogurt not linked to cancer risk overall, but possibly to one molecular subtype” is not.
Both statements come from the same paper. One of them describes what happens if you eat more yogurt; the other describes an interesting biological hypothesis. They are not the same claim and shouldn’t be read as one.
Ask three questions. Was the overall result significant, or only a subgroup? How big was the subgroup? Was the subgroup defined in advance, or found afterwards? If a headline reports a percentage without telling you the answer to the first question, you don’t yet have enough to act on. It doesn’t mean the study is bad — this one isn’t — it means the headline compressed away the part that determines what it means for you.
Why subgroup findings need caution
Subgroup analyses are a well-known source of results that don’t replicate, for reasons worth understanding:
- More comparisons, more chances of a false positive. Split a dataset enough ways and something will look significant by chance.
- Smaller numbers, wider uncertainty. A subgroup with 31% of cases has far less statistical precision than the whole.
- The direction can reverse on replication. Subgroup effects frequently shrink or vanish in the next study.
- Post-hoc versus pre-specified matters enormously. A hypothesis stated before looking is far stronger than one found afterwards.
None of that means this finding is wrong. It means it’s a hypothesis worth testing further, which is exactly how the authors framed it — not a dietary instruction.
Why this is still good science
It would be easy to over-correct into dismissiveness, and that would be wrong. Several things make this a better-than-average nutritional study:
- Enormous, long-running cohorts with detailed, repeated dietary data — far better than a single food-frequency questionnaire.
- A pre-existing biological hypothesis. Bifidobacterium is in yogurt; testing whether tumours containing it behave differently is a mechanistic question, not a fishing expedition.
- Molecular tumour classification. Going back to the tissue is expensive and rigorous.
- Honest reporting. The authors reported the null overall result rather than burying it — which is precisely why we can tell you about it.
The problem here isn’t the science. It’s the distance between the paper and the headline.
Why proximal colon cancer specifically
The association was in proximal colon cancer — the right-hand side, furthest along from the rectum. That specificity is biologically plausible rather than arbitrary.
The proximal colon has a different microbial environment from the distal colon and rectum, so a diet–microbiome effect showing up there rather than uniformly is consistent with the proposed mechanism.
It also matters clinically. Proximal tumours tend to be found later — they’re further from the reach of a sigmoidoscopy, and they often produce vaguer symptoms than left-sided cancers, which more often cause visible bleeding or obvious changes in bowel habit. Anything relevant to right-sided disease is therefore worth understanding, which is part of why this study attracted attention.
What observational means here
This is a cohort study, so it shows association, not causation. The standard caveat, but with a specific reason to take it seriously in this case.
People who habitually eat two or more servings of yogurt a week differ from those who don’t in many ways: they tend to eat more fibre, eat less processed meat, exercise more, smoke less, and engage more with preventive healthcare including screening. Good cohort studies adjust for these, and adjustment is never complete.
Residual confounding is a live possibility for a finding of this size in a subgroup. A randomised trial would settle it — and nobody is going to run a decades-long randomised trial of yogurt.
Should you eat more yogurt?
Here’s the practical answer, which is more relaxed than either the headline or the caveats suggest.
Yogurt is a perfectly good food. It provides protein, calcium and, in live-culture versions, beneficial bacteria. There’s no reason not to eat it and several ordinary reasons to.
But don’t eat it as cancer prevention, because that isn’t what the study supports. Eat it because it’s a decent food you enjoy.
| If you’re choosing yogurt | Look for |
|---|---|
| Live cultures | “Live and active cultures” on the label |
| Added sugar | Flavoured yogurts can carry a lot — check the panel |
| Plain plus fruit | Usually better than pre-sweetened |
| Protein and calcium | Genuine, uncontroversial benefits |
| Cancer prevention | Not what this study established |
The thing that actually saves lives
If this article changes one behaviour, let it be this one rather than a breakfast choice.
Colorectal cancer screening works. It detects cancers early, when treatment is far more effective, and — uniquely — it can find and remove precancerous polyps, preventing cancer from developing at all. Very few interventions in medicine do that.
The evidence gap between screening and any dietary factor isn’t close. And uptake is persistently poor, usually because people find the topic unpleasant rather than because they’ve weighed it up.
At-Home FIT Screening Kit
Screening detects cancer early and removes precancerous polyps — a far larger effect than any food in any study
A faecal immunochemical test detects hidden blood in stool, is done at home in a few minutes, and is the standard first-line screening method in many national programmes. If you’re in the eligible age range and haven’t been screened, this is the single highest-value thing in this article. Two important caveats. First, check whether you’re already covered by a national or insurance-funded programme — in many countries screening is free and posted to you, so buying a kit is unnecessary. Second, a FIT is a screening test, not a diagnosis: a positive result needs a colonoscopy, and a negative result does not rule out cancer — if you have symptoms, see a doctor regardless of what a home test says. Follow the collection instructions exactly, since technique affects reliability.
- Check for a free national programme first — don’t pay if you don’t need to
- Ask your doctor which test and interval suits your risk
- A positive result requires colonoscopy — arrange it, don’t repeat the test
- A negative result with symptoms still needs a doctor
- Higher-risk people (family history, IBD, genetic syndromes) need colonoscopy, not FIT
- Follow the collection instructions exactly
What genuinely changes your risk
Ranked roughly by the strength of evidence behind them:
- Screening at the recommended age and interval. Nothing else in this list can remove a precancerous polyp.
- Not smoking.
- Limiting alcohol.
- Limiting processed meat, and eating less red meat.
- Physical activity — consistently associated with lower risk.
- Maintaining a healthy weight.
- Dietary fibre, from wholegrains, pulses, fruit and vegetables.
- Knowing your family history, which may mean earlier or more frequent screening.
Yogurt doesn’t appear on that list, and that’s not a criticism of yogurt. Our related piece on recognising colon cancer symptoms covers what to watch for.
Symptoms that need a doctor now
Blood in your stool or rectal bleeding. Never assume it’s haemorrhoids, even if you have them.
A persistent change in bowel habit — looser, more frequent, or constipated — lasting more than a few weeks.
Unexplained weight loss.
Persistent abdominal pain, bloating or discomfort.
A feeling that your bowel doesn’t empty completely.
Unexplained tiredness or breathlessness — which can indicate iron deficiency anaemia from slow, invisible bleeding. This is the classic presentation of right-sided (proximal) colon cancer, the very type this study concerned, and it is frequently attributed to being busy or getting older.
Symptoms need assessment regardless of your age, and regardless of a negative home test. Colorectal cancer is rising in younger adults, and being under the screening age is not reassurance.
Frequently asked questions
Does yogurt prevent colon cancer?
Not on the evidence of this study, read properly. The Gut Microbes paper found no significant association between yogurt consumption and overall colorectal cancer incidence. What it did find was a 20% lower rate of Bifidobacterium-positive proximal colon cancer among people eating two or more servings a week — a subgroup representing about 31% of cases, with no association in the other 69%. That’s an interesting hypothesis about diet and the gut microbiome, not a demonstration that yogurt prevents bowel cancer.
What does “Bifidobacterium-positive” mean?
The researchers examined the tumour tissue itself and classified cancers by whether Bifidobacterium — a bacterial genus found in yogurt — was present in it. About 31% of the 3,079 cancers were positive; 69% were negative. The yogurt association appeared only in the positive group. That’s a genuinely clever design, because it tests a specific biological mechanism rather than just looking for a correlation — but it also means the finding applies to a minority of cases.
How big was the study?
Large: 132,056 participants from the Nurses’ Health Study and the Health Professionals Follow-up Study, with 3,079 colorectal cancers documented during follow-up. These are among the best-run cohorts in nutritional epidemiology, with detailed repeated dietary data rather than a single questionnaire. Size and quality aren’t the limitation here — the limitation is that the result is a subgroup finding in an observational study, and the overall result was null.
Why should I be cautious about subgroup results?
Because they’re a known source of findings that don’t replicate. Splitting a dataset many ways increases the chance something looks significant by accident; smaller subgroups carry wider uncertainty; and effects found in subgroups frequently shrink or disappear in the next study. It also matters a great deal whether the subgroup was specified in advance or identified afterwards. None of that makes this finding wrong — it makes it a hypothesis worth testing further, which is how the authors framed it.
Should I start eating more yogurt?
Eat it if you like it — it’s a good source of protein and calcium, and live-culture versions contain beneficial bacteria. Just don’t eat it as cancer prevention, because that isn’t what the study supports. If you’re choosing one, look for live and active cultures and check the sugar content, since flavoured yogurts can carry a lot; plain yogurt with fruit added is usually a better option than pre-sweetened.
What actually reduces colorectal cancer risk?
Screening, above everything else — it detects cancer early and removes precancerous polyps, preventing cancer entirely, which no dietary factor can do. After that: not smoking, limiting alcohol, limiting processed and red meat, staying physically active, maintaining a healthy weight, eating enough fibre, and knowing your family history, which may mean earlier or more frequent screening. If you’re in the eligible age range and haven’t been screened, that’s the change worth making.
Your checklist
- Know the headline finding: 20% lower rate in one subgroup
- Know the omitted finding: no significant association overall
- Only 31% of tumours were Bifidobacterium-positive
- Treat subgroup results as hypotheses, not instructions
- Remember this is observational — association, not causation
- Eat yogurt because it’s a good food, not as prevention
- Choose live cultures; check added sugar
- Get screened at the recommended age and interval
- Check whether a free national programme covers you
- A positive FIT needs a colonoscopy, not a repeat test
- A negative test doesn’t rule out cancer if you have symptoms
- Know that anaemia and fatigue can be right-sided colon cancer
- Never dismiss rectal bleeding as haemorrhoids
Medical disclaimer. This article is general information, not medical advice, and nothing here should be used in place of colorectal cancer screening. Speak to your doctor about which screening test and interval are right for you, particularly if you have a family history of bowel cancer, inflammatory bowel disease, or a known genetic syndrome — those situations usually require colonoscopy rather than a stool test. See a doctor promptly for rectal bleeding, blood in the stool, a persistent change in bowel habit, unexplained weight loss, persistent abdominal pain, or unexplained fatigue or breathlessness — the last of these can indicate anaemia from a right-sided colon cancer. Symptoms require assessment at any age and regardless of a negative home test result.
Affiliate disclosure. Some links on this page are affiliate links. As an Amazon Associate we earn from qualifying purchases. Note that our first piece of advice about the product we link to is to check whether you can get it free from a national screening programme instead.
