Ultra-processed food
Peer-reviewed research · 1 source · Last verified 2026-07-20
A "low-fat" or "healthy" label does not mean a food is good for you - it often signals more processing, not less. Ultra-processed foods are industrial formulations with little intact whole food left, and higher intake is linked to worse outcomes across a strikingly wide range of diseases. There is no single villain ingredient: the evidence points at the overall pattern, and at the whole foods these products crowd out. The fix is one swap a day.
The one honest fix
Swap one ultra-processed thing a day for a whole-food version — just one swap
Free · behavioral
What the evidence says
Higher ultra-processed food exposure is directly associated with 32 of 45 examined health outcomes across nearly ten million people
Limitations & how to read this number
umbrella review of observational meta-analyses; breadth finding is robust, individual estimates mostly low/very-low GRADE certainty
SourceLane MM et al. (2024) · BMJ · 10.1136/bmj-2023-077310
Ultra-processed food is associated with about a 21% higher risk of early death
+21% · RR 1.21, 95% CI 1.15-1.27
Limitations & how to read this number
Two things decide how much weight this figure can carry. The first is the certainty the reviewers themselves attached to it: they graded this particular association as highly suggestive rather than convincing, and rated its underlying quality LOW. Across all forty-five pooled analyses in the review, only four were rated moderate quality and none was rated high - so this is a large, consistent body of evidence made of individually weak parts. The second is what the ultra-processed category is doing here. It is defined by how food is made, not by any single ingredient, and it sweeps together things with little in common. People eating the most of it also tend to eat less fruit, fibre and fish, to have less money and less time, and to differ in ways no statistical adjustment fully removes. The review is a summary of other reviews, so it inherits every weakness of the studies underneath it, and none of them assigned anybody to eat this way.
SourceLane MM et al. (2024) · BMJ · 10.1136/bmj-2023-077310
How to actually do it
Don't try to eliminate a whole category overnight - pick one ultra-processed item you eat daily and swap it for a whole-food version: the packaged snack for fruit or nuts, the sugary drink for water, the white-bread sandwich for one with real ingredients. One swap a day gradually shifts the balance of your diet toward whole food, which is the actual lever.
Go deeper
Open what you care about
What "ultra-processed" actually means
Ultra-processed foods aren't just "processed" - they're industrial formulations made largely from substances extracted from foods (oils, starches, sugars, protein isolates) plus additives for colour, flavour and texture, with little intact whole food left. Think packaged snacks, sugary drinks, mass-produced bread, reconstituted meats and many "diet" or "low-fat" products. A label like "low-fat" often signals more processing, not less.
What the evidence shows
An umbrella review - a study of studies - pooled evidence across nearly 10 million people and found that higher ultra-processed food intake was directly associated with 32 of 45 examined health outcomes, spanning heart disease, type 2 diabetes, mental health and mortality. Overall, higher intake was linked to about 21% higher risk of early death. The breadth of the association is the robust headline; the individual estimates mostly carry low certainty, which we state plainly.
It's the pattern, not one ingredient
People often ask which additive is the culprit. The evidence points instead to the overall dietary pattern: the degree of industrial processing itself, and the way ultra-processed foods displace whole foods like vegetables, fruit, legumes and unprocessed meat and fish. That's why there's no single "bad" ingredient to remove - and why the fix is about shifting the balance of your diet, not hunting one villain.
How to read this honestly
This evidence is observational - people who eat more ultra-processed food often differ in income, activity and other habits, so some of the association reflects those things, and the certainty grades are mostly low. It is not proof that a specific packaged food will harm you. But the sheer consistency across dozens of outcomes and millions of people makes "eat less of it, and more whole food" a low-risk, sensible direction.
Where the evidence is strongest, and where it is not
The review did not treat its forty-five findings as equal. It graded them, and only four met its top bar of convincing evidence: death from cardiovascular disease, type 2 diabetes, anxiety, and common mental disorders taken together. A further group - including death from any cause, heart disease, depression, disturbed sleep, wheezing and obesity - was graded one step down, as highly suggestive. Of the rest, thirteen pooled analyses showed no evidence of a link at all. That last number belongs on this page as much as the first: a review that found something for everything it looked at would be the more worrying result.
FAQ
Is any of this proof that ultra-processed food causes these problems?
No. This is a review of reviews of observational studies, and it inherits their limits entirely. Nobody was assigned to eat this way. People who eat the most ultra-processed food differ from people who eat the least in income, time, cooking facilities, smoking, and much else, and adjustment for those is never complete. The reviewers were careful about this themselves - they graded most of what they found as low or very low quality, and called for mechanistic research precisely because the question of cause is open. What the evidence supports is a broad, consistent association across many outcomes, which is a reason to shift the balance of what you eat, not a demonstrated mechanism.
Does this mean I have to cook everything from scratch?
No, and nothing on this page asks you to buy anything or overhaul your kitchen. The category is broad and includes plenty of things nobody would call a treat, so trying to eliminate it entirely is both impractical and not what the evidence supports. The fix here is a single swap on a meal you already eat, which is the smallest change the evidence gives any reason to make. If cost or time is the constraint on how you eat, that is a real constraint and not a personal failing - this page is about one substitution, not a standard to live up to.
Does a "low-fat" label mean it is healthier?
Often the opposite. A low-fat label frequently signals more processing, not less - the fat is replaced with sugars, starches, and additives to keep the product palatable. The label describes one nutrient; it says nothing about how far the food has been engineered away from its original form.
Sources
- Lane MM et al. (2024). BMJ. 10.1136/bmj-2023-077310
Last verified 2026-07-20 · adversarial fact-check + founder science gate.
Corrections are published openly. Educational content, not medical advice — talk to a professional about your own situation.