BETHESDA, Maryland
It has been the villain of countless headlines: the ultra-processed food, blamed for everything from obesity and diabetes to dementia and cancer. Supermarket aisles have been called a health hazard. Campaigners have demanded warning labels. Now a huge new study from US government scientists is complicating the story.
Researchers at the National Cancer Institute followed more than 461,000 Americans for almost two decades and found no overall link between how much ultra-processed food (UPF) they ate and their chances of developing colorectal cancer. The findings, published on 15 October in the journal PLOS Medicine, landed in the news cycle this week and have lit up the nutrition world.
But before you reach for the family-size bag of crisps in celebration, read on. The study did not give ultra-processed food a clean bill of health. It found something more nuanced — and potentially more useful.
The study: 461,682 people, 18.8 years
The research, titled “Associations of food processing and dietary risk factors with colorectal cancer incidence in the NIH-AARP Diet and Health Study: A cohort study,” was led by Erikka Loftfield of the National Cancer Institute, with colleagues including Leila Abar, Caitlin O’Connell and Hyokyoung Hong. It was funded by the NCI’s intramural research programme at the National Institutes of Health, which had no role in the study’s design, analysis or decision to publish.
The team analysed 461,682 Americans aged 50 to 71 who joined the NIH-AARP Diet and Health Study in 1995 and 1996. At the start, participants filled in a food frequency questionnaire describing what they typically ate. Researchers then tracked them through 2018, a median follow-up of 18.8 years, counting new cases of colorectal cancer.
The headline result is simple. “We did not observe an association between total ultra-processed food (UPF) intake and colorectal cancer incidence,” Loftfield and colleagues reported. People who ate the most ultra-processed food overall were not more likely to develop the disease than those who ate the least.
What counts as “ultra-processed”?
The term comes from classification systems that sort food by how much industrial processing it undergoes, rather than by its nutrients. Ultra-processed foods are typically formulations made largely from refined ingredients and additives, such as packaged snacks, sugary cereals, soft drinks, instant noodles, and reconstituted meat products.
That is the catch. The category is enormous and wildly varied. A fizzy drink and a sliced wholegrain loaf from a supermarket might both qualify. So might a flavoured yoghurt, a fish finger and a sausage. Critics have long argued that lumping all of that together can hide as much as it reveals, and this study appears to bear that out.
The twist: not all UPFs are equal
When the scientists looked inside the category, the picture changed. Whole grains, fibre, dairy and calcium that came from ultra-processed sources were associated with a lower incidence of colorectal cancer. In other words, a fortified breakfast cereal high in whole grains or a processed dairy product may not carry the same risk as a processed sausage.
On the other side of the ledger, meat from ultra-processed sources was associated with a higher incidence of colorectal cancer. That fits with decades of earlier evidence on processed meat. The World Health Organization’s cancer agency classifies processed meat as carcinogenic to humans, a finding that predates the modern UPF debate.
The authors also ran a “swap” analysis. Replacing 10 per cent of ultra-processed intake, by weight, with minimally processed foods was associated with a 2 per cent lower incidence of colorectal cancer. That is a small effect, and the researchers stress it describes an association rather than a guarantee, but it points in a clear direction.
Why this matters now
Colorectal cancer is one of the most common cancers worldwide, and it is rising in younger adults in many countries, which has driven intense interest in diet as a possible culprit. Ultra-processed food has been a prime suspect because Western diets have shifted heavily toward it over recent decades.
Large observational studies have linked high UPF consumption with a range of poor outcomes, and some earlier work has tied UPFs to bowel cancer risk, particularly in men. This new analysis does not erase that research, but it adds weight to the argument that the “total UPF” number is too blunt a tool. What matters, the data suggest, is which products make up the total.
For public health messaging, that is a headache. A simple rule such as “avoid anything ultra-processed” is easy to remember but may sweep up foods that offer genuine nutritional value, such as fortified grains and dairy. A rule such as “cut back on processed meat” is more precise and has stronger evidence behind it.
The fine print: what the study can’t tell us
Nutrition science is notoriously messy, and the authors are upfront about limitations. Most participants were older and white, so the findings may not apply to younger people or to more diverse populations. That matters because the sharpest rise in bowel cancer is in younger adults, who were not the focus of this cohort.
Diet was measured only once, at the start, through a questionnaire. People’s eating habits change over 19 years, and the range and availability of ultra-processed products grew substantially during the study period, so the baseline snapshot may well underestimate what participants ate later in life. Food frequency questionnaires also depend on memory and honesty.
Finally, the study is observational. It can show that certain patterns of eating go with higher or lower cancer rates, but it cannot prove that a particular food causes or prevents cancer. People who eat more whole grains may also exercise more, smoke less and see their doctor more often, and statistical adjustment can only go so far.
How does this fit with the wider UPF debate?
The study does not exist in a vacuum. Researchers have linked heavy UPF consumption to weight gain, type 2 diabetes, heart disease and early death in various studies, and the debate over mechanism is unresolved. Some scientists point to the high levels of sugar, salt and saturated fat; others to additives, emulsifiers or the way processed foods are engineered to be easy to overeat; still others to the foods UPFs displace, such as vegetables, fruit and pulses.
This work suggests that for colorectal cancer in particular, the specific ingredients — meat on one hand, fibre, whole grains and calcium on the other — may carry more explanatory power than the overall level of processing. That is a useful nudge for researchers designing the next generation of studies, and for governments weighing how to write food labels and dietary guidelines.
There is also a practical lesson about how we talk about food. Public conversations about diet often swing between extremes: one year a nutrient is a miracle, the next a poison. A study of this size, run by government scientists with no commercial stake in the outcome, is a useful corrective. It shows that the evidence on any single food group tends to be more tangled than a headline allows, and that careful researchers are increasingly asking not just how processed a food is, but what exactly is in it and what it replaces on the plate.
Dietitians have long made a similar point to patients: a ready meal built on vegetables and legumes is not the same as a packet of fried snacks, even if both are factory-made. Studies like this one give that everyday intuition a firmer scientific footing, while still leaving room for the possibility that other health outcomes, from heart disease to weight gain, may behave differently from bowel cancer.
It would be a mistake, however, to read one study as the final word. Science advances by accumulation, and results from this American cohort will need to be tested against data from other countries, younger groups and more recent diets.
What this means for you
So what should you do with your next shopping trip? Experts and the study’s own findings point to a few sensible takeaways. First, processed meat is still the clearest red flag. Bacon, sausages, ham and similar products are the items to limit, and the study’s finding on UPF-sourced meat is consistent with that advice.
Second, do not fear every packet. Foods that deliver fibre, whole grains and calcium may be part of a healthy pattern even when they are industrially produced. Check the label for whole grain content and fibre rather than judging purely on how processed a product sounds.
Third, small swaps may add up. The 2 per cent figure is modest, but it came from replacing just a tenth of ultra-processed intake with minimally processed foods like fruit, vegetables, plain yoghurt, nuts and cooked-from-scratch meals. Over a lifetime and a population, small shifts matter.
Fourth, and crucially, do not let a reassuring headline stop you from getting screened. Colorectal cancer is highly treatable when caught early. If you are due a screening test, or notice persistent changes in bowel habits, blood in your stool, unexplained weight loss or ongoing abdominal pain, speak to your doctor.
The bottom line
The ultra-processed food debate just got more interesting, and less black-and-white. This enormous study found no overall link between total UPF intake and colorectal cancer, but it did find that the type of food inside the packet matters: processed meat appears to raise risk, while whole grains, fibre, dairy and calcium appear to lower it. In an era of scary headlines, that is a refreshing reminder that nutrition rarely comes down to a single villain.
The study’s authors are careful not to overreach, and so should we. The healthiest approach remains the least exciting one: eat more plants and minimally processed foods, go easy on processed meat, stay active and get screened on schedule.
