Health Q&ANutritionEvidence based
In the companion article we walk through what “ultra-processed” actually means and why the category as a whole tracks with worse health. But the category is blunt — and when researchers follow its sub-groups separately, the risk concentrates in a handful of products.1 Here they are, ranked by the strength of the evidence against them, with the reason each one earns its place.
1. Processed meat — the only one with a formal verdict
Bacon, salami, ham, hot dogs, cured sausages. This is the only food on the list — indeed one of very few foods anywhere — that the WHO’s cancer agency IARC has classified as a group 1 carcinogen: the evidence that it causes colorectal cancer is considered established, with roughly 18% higher relative risk per 50 g eaten daily (about two slices of ham).2 Group 1 describes the certainty, not the potency — a daily ham sandwich is not smoking — but no other everyday food carries that label.
The mechanisms are unusually concrete: nitrite curing forms nitrosamines in the product and the gut, heme iron promotes the same nitroso chemistry, and high-temperature processing adds its own compounds. On top of the cancer file, processed meat is the sub-category that most consistently drives the ultra-processed association with heart disease and diabetes in cohort analyses.1 If you change one daily habit on this page, make it this one.
2. Sugar-sweetened drinks — the metabolic offender
Soda, energy drinks, sweetened iced teas and juice drinks deliver rapid sugar with zero satiety — the calories simply do not register the way food does. The epidemiology is correspondingly blunt: habitual consumption predicts type 2 diabetes even after adjusting for body weight, with one extra serving a day associated with roughly 13–18% higher risk in a BMJ meta-analysis.3 Liquid sugar also feeds liver fat, and “fruit drink” branding changes nothing about the chemistry. This is the largest single source of added sugar in many diets — and the easiest to cut, because a zero-sugar swap already removes most of the problem (with the sweetener caveats we cover in Sugars & sweeteners).
3. Industrial pastries & confectionery — engineered to overeat
Packaged cakes, cookies, chocolate bars and sweet snacks combine refined flour, sugar and cheap fats into the most energy-dense, hyper-palatable matrix the food industry knows how to build — the exact product profile that made participants in the NIH’s controlled trial passively eat ~500 kcal/day more on an ultra-processed diet.4 Sub-category analyses regularly flag sweet packaged snacks among the drivers of weight gain and cardiometabolic risk. Nothing here is acutely toxic; the harm is that the format quietly defeats your fullness signals, day after day.
4. Fried snacks — energy density with extras
Chips and other fried snack products stack the same energy-density problem (over 500 kcal per 100 g), aggressive salt loads, and — from high-temperature frying of starches — acrylamide, a probable carcinogen picked up as a bonus. They also occupy the “mindless eating” slot: engineered crunch and vanish-in-the-mouth texture that makes a serving size a fiction. A single bag on movie night is noise; the daily bag is a pattern the cohort data can see.5
5. Instant & ready meals — the displacement problem
Instant noodles, freezer pizzas and heat-and-eat meals are the most defensible entry on the list — and still earn their place. Their salt content routinely covers half a day’s limit in one serving, refined starches dominate, and their real damage is displacement: every ready-meal dinner is a dinner that isn’t vegetables, legumes, fish or whole grains. In the cohort data they contribute steadily, if less dramatically, to the ultra-processed risk signal.5 As an occasional Tuesday rescue they are fine; as the default dinner they reshape the whole diet.
Largely acquitted: bread, cereals — and your protein powder
Honesty requires the other half of the story. In the large EPIC analysis that followed ultra-processed sub-categories separately, ultra-processed breads and cereals showed no increased risk — if anything a slight trend the other way — and the overall association was carried by the processed meats and sweetened drinks above.1 The same logic clears whey protein and similar single-purpose products: technically NOVA group 4, nutritionally nothing like a hot dog. “Ultra-processed” tells you where to look; it does not deliver the verdict.
The swap list
- Daily salami/ham on bread → egg, mackerel, chicken, hummus or cheese most days; keep cured meat for weekends. This is the highest-value change on the page.
- Daily soda or energy drink → water, sparkling water, coffee or tea; a zero-sugar version as the stepping stone.
- The office cookie habit → fruit + a handful of nuts covers the same craving with an opposite evidence profile.
- Chips as TV default → popcorn you pop yourself, or nuts — demote chips to actual occasions.
- Ready meals as default dinner → a rotation of 15-minute real-food meals; our recipe collection is built for exactly this.
Read next
References
Adult nutrition summary for education, not a prescription.
- Cordova R, Viallon V, et al. Consumption of ultra-processed foods and risk of multimorbidity of cancer and cardiometabolic diseases: a multinational cohort study (EPIC). Lancet Reg Health Eur 2023;35:100771. pubmed.ncbi.nlm.nih.gov
- Bouvard V, Loomis D, et al. (IARC Monograph Working Group). Carcinogenicity of consumption of red and processed meat. Lancet Oncol 2015;16(16):1599–1600 — a short announcement without an abstract; the freely readable summary is the WHO Q&A on red and processed meat. pubmed.ncbi.nlm.nih.gov
- Imamura F, O’Connor L, et al. Consumption of sugar sweetened beverages, artificially sweetened beverages, and fruit juice and incidence of type 2 diabetes. BMJ 2015;351:h3576. pubmed.ncbi.nlm.nih.gov
- Hall KD, Ayuketah A, et al. Ultra-processed diets cause excess calorie intake and weight gain: an inpatient randomized controlled trial. Cell Metab 2019;30(1):67–77. pubmed.ncbi.nlm.nih.gov
- Lane MM, Gamage E, et al. Ultra-processed food exposure and adverse health outcomes: umbrella review of epidemiological meta-analyses. BMJ 2024;384:e077310. pubmed.ncbi.nlm.nih.gov
This article summarises nutrition science for education. It is not a substitute for individual medical advice.