Ultra-Processed Foods: What a Study of 8.8 Million People Actually Found
Updated: 2 days ago
Written and medically reviewed by Dr. Baraa Alnahhal, MD · Last reviewed: September 2026
Ultra-processed foods are industrially made products built mostly from refined ingredients and additives rather than whole foods — packaged snacks, sweetened cereals, soft drinks, instant noodles, reconstituted meats and most ready meals. On 22 September 2026, researchers published the largest analysis of them ever assembled: 51 long-term studies covering 8,819,894 adults. It found that people eating the most ultra-processed food had higher rates of seven categories of chronic disease and of early death.
Key takeaways
Across 51 studies and nearly 8.8 million adults, the highest ultra-processed food intake was associated with 18% higher all-cause mortality and higher rates of heart disease, cancer, obesity, diabetes, depression and anxiety, and digestive disease (Liu et al., Family Medicine and Community Health, 2026).
The relationship was dose-dependent: every extra 100 g a day was linked to a 14% higher risk of cardiovascular events and a 3% higher risk of death from any cause.
This is association, not proof of cause. Under the formal GRADE system the researchers used, only one of the eight outcomes — obesity — was rated moderate certainty. The rest were rated low or very low.
Independent experts note that only 19 of the 51 studies adjusted for income or socioeconomic status — a gap that could explain part of the link.
The practical takeaway is not "eliminate," it is "shift the ratio." The dose–response pattern suggests eating somewhat less may still matter.
What counts as an ultra-processed food?
An ultra-processed food is one made largely or entirely from substances extracted from foods — refined oils, sugars, starches, protein isolates — combined with additives like preservatives, emulsifiers, sweeteners and colourings, with little or no whole food left in it. This is the fourth and highest group in the NOVA classification system that nutrition researchers use.
Mayo Clinic's patient guidance places foods on a spectrum, from minimally processed items such as fresh blueberries, cut vegetables and roasted nuts, through canned tomatoes and frozen vegetables, to ultra-processed products like sweetened breakfast cereals, soft drinks, chicken nuggets and hot dogs (Mayo Clinic Health System).
The category is broad, and that breadth is one of the honest criticisms of this research area: a diet cola and a packaged wholegrain bread can land in the same group despite having little in common nutritionally.
What did the new study find?
The analysis pooled 51 prospective cohort studies — research that tracks healthy people over years and records who develops disease. Comparing the highest ultra-processed food consumers with the lowest, the researchers reported these increases in risk:
Outcome | Increase in risk (highest vs lowest intake) | 95% confidence interval (risk ratio) |
|---|---|---|
Digestive diseases | 31% | 1.16–1.48 |
Depression and anxiety | 27% | 1.15–1.40 |
Cardiovascular events | 24% | 1.14–1.36 |
Metabolic syndrome / diabetes | 24% | 1.08–1.42 |
Obesity or overweight | 23% | 1.17–1.30 |
All-cause mortality | 18% | 1.10–1.26 |
Cancer | 12% | 1.04–1.21 |
Does eating a little less actually help?
This is the part of the study with the most practical value. The researchers ran a dose–response analysis, and the risk did not sit at a single cliff edge — it climbed gradually with intake.
Each additional 100 grams of ultra-processed food per day was associated with a 14% higher risk of cardiovascular events, a 4% higher risk of cancer, a 3% higher risk of death from any cause, and a 2% higher risk of metabolic syndrome or diabetes.
To put 100 grams in perspective: a small bag of crisps is roughly 30–50 g, a standard chocolate bar around 45 g, a can of soft drink about 330 g by weight. The gradual pattern is what supports a realistic goal — reducing, rather than eliminating.
Do ultra-processed foods cause disease?
No. Every study in this analysis was observational, which means it can show that two things occur together but cannot establish that one causes the other. The researchers graded the strength of their own findings using GRADE, a standard rating system, and the results were modest: only the obesity finding reached moderate certainty. The other seven outcomes — including cancer, heart disease and death — were rated low or very low certainty.
The core problem is confounding. People who eat the most ultra-processed food also tend to have lower incomes, less time to cook, more stressful work, more smoking and less exercise — and any of those could independently drive disease.
Dr Adam Jacobs, a biostatistician at Ergomed, made this point directly when the paper was published, noting that a large study "cannot compensate for biases or confounding in the individual studies" and that socioeconomic status is a "particularly worrisome" confounder here — because only 19 of the 51 studies adjusted for any kind of economic measure (Science Media Centre, 22 September 2026).
Professor Gunter Kuhnle, Professor of Nutrition and Food Science at the University of Reading, added that "the results can only be as good as the underlying research," pointing out that measuring what people actually eat is "notoriously difficult" and that randomised controlled trials "provide a much more nuanced picture." Prof Kuhnle has previously received research funding from Mars and consultancy payments from other commercial bodies, which he declares.
Both experts agree the finding is a signal worth taking seriously — and that it is not proof.
Why might processing itself matter?
Researchers have proposed several mechanisms, none of them settled. Ultra-processed foods tend to be energy-dense and low in fibre, which may lead people to eat more before feeling full. Soft textures mean faster eating. Some additives, particularly certain emulsifiers and preservatives, are being investigated for effects on the gut lining and blood pressure. And these foods often displace vegetables, legumes and whole grains from the plate.
It is also possible that processing is mostly a marker — that what matters is the salt, sugar, saturated fat and missing fibre, rather than the industrial processing itself. Current evidence cannot separate these explanations.
What can you actually do with this information?
The most defensible reading of this study is a shift in proportions, not a purge. A few changes that follow from the evidence:
Change the ratio, not the rulebook. Aim to displace one or two ultra-processed items a day with a minimally processed alternative. The dose–response pattern suggests partial change still counts.
Start with drinks. Sugar-sweetened beverages are high-volume and easy to swap, so they move the 100 g/day needle fastest.
Read the ingredient list, not the front of the pack. Mayo Clinic suggests watching for additions like high-fructose corn syrup, hydrogenated oils and food dyes.
Protect fibre first. Much of the apparent harm may run through what these foods crowd out. Adding vegetables, beans and whole grains is a positive action rather than a restriction.
Take one step at a time. Mayo Clinic's guidance explicitly recommends "baby steps by making one dietary change at a time" — consistent with the finding that risk scales gradually.
Convenience food is not a moral failing, and cost and time are real constraints. A realistic reduction you sustain is worth more than an elimination you abandon.
Frequently asked questions
Is all processed food bad for you?
No. Processing includes freezing, canning, pasteurising and cooking — all of which make food safer or more available. Frozen vegetables and tinned beans are processed and nutritious. The concern in this research is specifically ultra-processed products: industrial formulations built from refined ingredients and additives.
How much ultra-processed food is too much?
No safe threshold has been established, and this study does not set one. What it shows is a gradual gradient: each extra 100 g a day was linked to incrementally higher risk. That means there is no single number to hit — reducing intake at any starting point is the relevant goal.
Does this study mean ultra-processed food causes cancer or depression?
No. It found that higher intake occurs alongside higher rates of these conditions across large populations. Because the research is observational, other factors such as income, smoking and overall diet quality may explain part or all of the link. The researchers rated the certainty of most of these specific findings as low or very low.
Are food additives the problem, or the nutrition?
It is not yet known. Some additives are under active investigation for effects on the gut and blood pressure, but ultra-processed foods are also typically high in sugar, salt and saturated fat and low in fibre. Current studies cannot separate the effect of processing from the effect of nutrient content.
What is the single most useful change to make?
Replacing sugar-sweetened drinks with water or unsweetened alternatives. They are high in weight and volume, contribute nothing nutritionally, and are among the easiest ultra-processed items to substitute without changing how you cook or shop.
Health information, not medical advice. This article summarises published research for general information. It is not a diagnosis, a treatment plan, or personal medical advice, and it should not be used to start, stop or change any medication or diet prescribed to you. If you have a health condition, are pregnant, or take regular medication, speak with your doctor or a registered dietitian before making significant dietary changes.
Primary sources
Liu X, Luo S, Zeng Y, et al. Ultra-processed food consumption and risk of multiple chronic diseases among 8,819,894 adults from 51 prospective cohorts: a dose-response meta-analysis. Family Medicine and Community Health. 22 September 2026. https://doi.org/10.1136/fmch-2026-003925 (PROSPERO CRD42023421092)
Science Media Centre. Expert reaction to UPF consumption linked to higher risk of multiple chronic diseases. 22 September 2026. Link
Mayo Clinic Health System. Processed foods: What you should know. Link
About the author: Dr. Baraa Alnahhal, MD, is a medical doctor and the founder of Rinnit. Dr. Baraa Alnahhal reviews new medical research and translates it into clear, evidence-based health information for everyday readers.
Comments