A large US study of more than 460,000 adults, published in October 2026, found that eating more ultra-processed food overall was not linked to a higher risk of colorectal (bowel) cancer. But that headline hides the real finding: some ultra-processed foods, like processed meat, were linked to higher risk, while others, like whole-grain bread, breakfast cereals and dairy, were linked to lower risk, and the two roughly cancelled out.
So no, this doesn’t make packaged snacks and soft drinks “healthy”. Here’s what the study found, how it fits with earlier research and a separate brain study from the same month, and what it means for a typical Indian diet.
What the 460,000-person study found
The study was led by Erikka Loftfield of the US National Cancer Institute and published in PLOS Medicine. The key details:
- Who: 461,682 Americans in the NIH-AARP Diet and Health Study, aged 50 to 71 when they joined in 1995 and 1996, and free of cancer at the start.
- How: they filled in one food-frequency questionnaire at the start. Researchers sorted each food by how processed it was, then followed people for a median of 18.8 years. During that time, 10,075 people developed colorectal cancer.
- How much ultra-processed food: a median of 43% of daily calories.
- Main result: people in the top fifth for ultra-processed food intake had about the same colorectal cancer rate as those in the bottom fifth (hazard ratio 0.97, 95% CI 0.91 to 1.03). Results were similar for the right side of the colon, the left side and the rectum.
The interesting part is what happened when the researchers split ultra-processed food into types. Comparing the highest and lowest intakes:
| Ultra-processed food group | Link with colorectal cancer |
|---|---|
| Breakfast cereals | 16% lower risk (HR 0.84) |
| Frozen desserts and milk drinks | 18% lower risk (HR 0.82) |
| Bread | 9% lower risk (HR 0.91) |
| Reconstituted meat and fish (sausages, ham, nuggets) | 10% higher risk (HR 1.10) |
Why would packaged bread and cereal look protective? In this group of Americans, ultra-processed foods were where most of their whole grain (82.8%) came from, along with about a third of their fibre and calcium. Current bowel cancer prevention advice already recommends all of these. Meanwhile, 74.3% of their processed meat came from ultra-processed sources. The authors concluded that their results support existing advice: eat whole grains, fibre, dairy and calcium, and avoid processed meat.
They also estimated that swapping 10% of ultra-processed food (by weight) for minimally processed food was linked to about 2% lower colorectal cancer risk. That’s small, but it points in the same direction as earlier research.
What does “ultra-processed” actually mean?
Most researchers use the NOVA system, which groups food by the extent and purpose of processing, not by nutrients. A 2019 paper in Public Health Nutrition by the system’s developers describes four groups:
| NOVA group | What it is | Indian examples |
|---|---|---|
| 1. Unprocessed or minimally processed | Whole foods that are cleaned, dried, frozen or pasteurised | Fresh vegetables, fruit, dal, rice, atta, milk, eggs, plain curd |
| 2. Processed culinary ingredients | Extracted from foods and used in cooking | Oil, ghee, sugar, salt |
| 3. Processed foods | Group 1 foods with group 2 ingredients added | Pickles, paneer, simple bread made of flour, water, salt and yeast |
| 4. Ultra-processed foods | Industrial formulations, often with additives | Packaged namkeen and chips, biscuits, instant noodles, cold drinks, sweetened breakfast cereals, sausages |
The authors suggest a simple test: read the ingredient list. If it has items you’d never use in a home kitchen (such as hydrolysed protein, high-fructose corn syrup or hydrogenated oil) or additives like flavours, colours, emulsifiers, sweeteners and thickeners, it’s probably ultra-processed.
The catch is that “ultra-processed” is a very wide group. It covers both a cola and a packaged multigrain loaf, which have very different effects on health.
What earlier studies found
The new result doesn’t come out of nowhere, but it does differ from some earlier work:
- BMJ 2022: a study of three large US cohorts of health professionals, with more than 200,000 people, found that men with the highest ultra-processed food intake had a 29% higher colorectal cancer risk (HR 1.29). There was no link in women overall. Ready-to-eat meat products and sugary drinks were linked to higher risk in men, while yogurt and dairy desserts were linked to lower risk in women.
- BMJ 2024 umbrella review: this review of 45 pooled analyses found “convincing” evidence linking more ultra-processed food to cardiovascular death (risk ratio 1.50), type 2 diabetes, anxiety and common mental disorders. For cancer, the evidence was much weaker. Colorectal cancer links were graded only “suggestive” or “weak”, and most evidence overall was rated low or very low quality.
In other words, the strongest case against ultra-processed food has always been about heart disease, diabetes, weight and mental health. The cancer link was the shakiest part, and this new study fits that picture.
Why do the studies disagree?
- Different foods inside the same label. If one group’s “ultra-processed” intake is mostly whole-grain bread and yogurt, and another’s is mostly hot dogs and soft drinks, you’d expect different results.
- One questionnaire, decades ago. The NIH-AARP study measured diet once, in the mid-1990s. People’s diets change, and ultra-processed food has grown since then.
- Who was studied. Both big US studies were mostly older, white Americans, which limits how well they apply to Indians.
- Observational design. These studies show links, not cause and effect. People who eat a lot of packaged food may also differ in exercise, smoking, alcohol and income.
The separate brain study
The same month, a Canadian study in JAMA Network Open looked at 8,783 adults aged 35 to 69 without heart disease. They got an average of 42.2% of their calories from ultra-processed food.
- Those eating the most had slightly more body fat (1.7 percentage points) and more visceral (deep belly) fat.
- On MRI, 11.3% of the highest group showed signs of small-vessel damage in the brain (“vascular brain injury”, such as silent strokes), compared with 8.1% of the lowest group. That’s about 40% higher odds.
- The link appeared in women but not clearly in men.
- There was no link with scores on memory and thinking tests.
This was a snapshot taken at one point in time, so it can’t show which came first. The authors also noted the finding changed depending on how ultra-processed intake was defined. It’s a reason for more research, not proof that biscuits damage your brain.
How Indian diets fit in
Most Indian home meals (dal, sabzi, roti, rice, curd) fall in NOVA groups 1 to 3. But packaged food is growing fast. A 2023 WHO and ICRIER report found India’s ultra-processed food sector grew at 13.37% a year in retail sales value between 2011 and 2021. The five biggest categories were chocolates and sugar confectionery, salty snacks, beverages, ready-made and convenience foods, and breakfast cereals.
That’s a different mix from the American study. In the US, ultra-processed food supplied most people’s whole grains. In India, the commonest ultra-processed foods are fried namkeen and chips, cream biscuits, instant noodles made from refined flour, and sugary drinks, which bring salt, sugar and refined oil without much fibre. So the protective effect seen for US breads and cereals isn’t likely to apply to a typical Indian snack plate.
Processed meat is where the evidence is firmest. The WHO’s cancer agency has classified processed meat as Group 1, carcinogenic to humans, which includes ham, sausages, hot dogs and canned meat. If your family eats salami or sausages often, that’s the change most worth making.
What to actually do
- Judge foods by what’s in them, rather than by the word “processed”. Plain curd, frozen peas, packaged atta or a simple whole-wheat bread aren’t the problem.
- Cut processed meat first. Keep sausages, salami and ham to occasional use.
- Swap sugary drinks for water, chaas or nimbu pani with little or no sugar.
- Treat packaged namkeen, biscuits and instant noodles as occasional, not daily. Our guide to healthier Diwali snacks has easy home swaps like roasted makhana and chana.
- Get your whole grains and fibre from home food: whole-wheat roti, millets, dal, rajma, chana, vegetables and fruit.
- Read the label. A long ingredient list full of things you wouldn’t cook with is a sign to have less.
Home cooking isn’t automatically healthy either. Deep-fried snacks and heavy sweets made at home still carry plenty of fat and sugar, as our pieces on how much ghee per day and Diwali sweet portions explain.
When to see a doctor
Diet is only one part of bowel cancer risk, and symptoms matter more than any single food. The NHS advises seeing a doctor if you notice:
- A change in bowel habits that’s unusual for you, such as looser stools, diarrhoea or constipation
- Blood in your stool, which may look red or black, or bleeding from the bottom
- Often feeling you need to pass stool, even just after going
- Tummy pain, a lump in the tummy, or bloating
- Weight loss you can’t explain, or tiredness and breathlessness (possible signs of anaemia)
Get urgent help if your stool is black or dark red or you have bloody diarrhoea. Call 112 or go to the emergency department if bleeding from the bottom won’t stop, or you pass a lot of blood or large clots.
Frequently asked questions
Does this study mean ultra-processed food is safe?
No. It found no overall link with colorectal cancer in older Americans. Ultra-processed food is still linked with heart disease, type 2 diabetes, obesity and mental health problems in many other studies.
Are all packaged foods ultra-processed?
No. Frozen vegetables, plain packaged curd, atta, dal and paneer are minimally processed or processed. Ultra-processed foods are industrial recipes with ingredients and additives you wouldn’t use at home.
Is bread bad for you?
It depends on the bread. In the US study, bread was linked to slightly lower bowel cancer risk, probably because much of it was whole grain. White bread made with refined flour, sugar and additives is a different product.
Are instant noodles ultra-processed?
Yes. The NOVA developers list instant noodles and instant soups as examples. An occasional packet won’t hurt, but they shouldn’t replace regular meals.
What is the single best change to lower bowel cancer risk through diet?
Cutting processed meat and eating more whole grains and fibre are the changes this study and earlier guidance agree on. Diet isn’t everything, though, so keep up with any screening your doctor recommends.
This article is for general information and is not a substitute for advice from a doctor. If you have symptoms or a health condition, please consult a qualified healthcare professional.
