Skip to content
ContentLora

    Tip: press / anywhere to search.

    Analysis

    Ultra-processed foods: what the evidence shows

    Large reviews of observational studies consistently link diets high in ultra-processed foods with obesity, type 2 diabetes, cardiovascular death and other outcomes, and one small inpatient trial found people ate about 500 more calories a day on an ultra-processed diet.[1][2][3] How much of this is cause and effect is disputed: most of the evidence is observational and rated low or very low quality, there is no agreed definition, and UK advisers say the links may reflect the sugar, salt and fat these foods contain.[4][5][6] The 2025-2030 US Dietary Guidelines nonetheless tell Americans to avoid highly processed foods, and in 2025 FDA and USDA sought information to develop a federal definition.[7][5]

    Editor reviewedStrict sourcingUpdated Analysis12 claims · 100% primaryFood and farming, Health and medicine
    Show:

    This analysis of ultra-processed foods separates what has been measured, by type of evidence, from how different expert bodies interpret it.

    This page is not medical or dietary advice. Anyone thinking about changing their diet because of a health condition should talk to a clinician.

    The definition problem

    There is no agreed definition of ultra-processed foods. In July 2025 FDA and USDA asked for information to help develop a uniform US definition for research and policy.[5] A 2023 review by the UK Scientific Advisory Committee on Nutrition (SACN) found that most of the studies it looked at used the NOVA classification, and that NOVA was the only system that met its screening criteria, but said its categories are very broad and group foods with different nutritional profiles.[8] Of 12 studies SACN found that applied NOVA to UK survey data, 9 included authors from the group that developed NOVA.[9]

    The Scientific Foundation behind the 2025-2030 Dietary Guidelines uses its own term, “highly processed”, for foods made mainly from substances extracted from foods or containing industrial chemical additives. On that definition it estimates highly processed foods supply about two-thirds of US youths’ calories and 60% of adults’.[10]

    Trial evidence

    The main randomized trial is small. In 2019 NIH researchers admitted 20 weight-stable adults to a clinical centre and fed them an ultra-processed diet and an unprocessed diet for two weeks each. The meals were matched for presented calories, energy density, macronutrients, sugar, sodium and fiber, and participants ate as much as they wanted.[11] They ate about 508 more calories a day on the ultra-processed diet. They gained about 0.9 kg during that fortnight and lost about 0.9 kg on the unprocessed diet.[3]

    Observational evidence

    A 2024 umbrella review in the BMJ pooled 45 meta-analyses covering almost 9.9 million people. It found direct associations between ultra-processed food exposure and 32 of the 45 health outcomes.[1] It rated the evidence convincing for links with cardiovascular death, type 2 diabetes, anxiety and common mental disorders. Under the GRADE system, though, 22 of the pooled analyses were low quality, 19 very low and only four moderate.[4]

    The US Scientific Foundation’s own umbrella review of 27 meta-analyses, mostly of large cohort studies, found higher intake associated with higher risk for nearly every outcome studied. Relative risks ranged from 1.12 for cancer to 1.55 for obesity, and the review rated the type 2 diabetes evidence higher certainty.[2]

    What the evidence cannot yet show

    SACN said the studies are almost exclusively observational. It said confounders such as energy intake, body mass index, smoking and socioeconomic status may not be adequately accounted for.[12]

    SACN’s reading is that it is unclear whether the associations come from processing itself or from the high energy density, salt, saturated fat and free sugars typical of many ultra-processed foods. It adds that eating a lot of these foods may be a marker of other unhealthy diet and lifestyle patterns.[6] The trial result fits a mechanism of passive overeating, but it involved 20 people over four weeks, so it cannot show long-term disease effects.[11][3]

    How policy has moved

    The 2025-2030 Dietary Guidelines advise avoiding highly processed packaged foods that are salty or sweet. They also advise limiting foods with artificial flavors, petroleum-based dyes, artificial preservatives and low-calorie sweeteners, and avoiding sugar-sweetened drinks.[7] How those guidelines were set is covered in the Dietary Guidelines explainer, and US dye rules in the food dyes and additives explainer.

    FDA and USDA have said they want a uniform definition of ultra-processed foods for research and policy, which would give US studies and any future rules a common basis.[5]

    Competing views

    The evidence is strong enough to act

    Dozens of meta-analyses covering millions of people point the same way, a randomized trial shows a plausible mechanism through overeating, and US federal guidance now tells people to avoid highly processed food.[1][2][3][7]

    Consistent associations, uncertain causes

    The studies are almost all observational, most are rated low or very low quality, and it is unclear whether processing itself matters or whether these foods simply tend to be high in calories, salt, saturated fat and sugar.[4][12][6]

    Fix the definition first

    The NOVA system used in most of the research groups together foods with very different nutrition, and there is no agreed definition, so results are hard to compare or turn into rules until one is set.[8][9][5]

    Next in the crash course

    Go deeper

    Questions readers ask

    What counts as an ultra-processed food?

    There is no agreed definition. Most studies reviewed by UK advisers used the NOVA classification, and in July 2025 FDA and USDA asked for information to develop a uniform US definition. The Scientific Foundation behind the 2025-2030 Dietary Guidelines defines highly processed foods as those made mainly from substances extracted from foods or containing industrial chemical additives.[5][8][10]

    Is there a randomized trial on ultra-processed food?

    Yes, a small one. In a 2019 NIH inpatient trial, 20 adults ate about 508 more calories a day on an ultra-processed diet than on an unprocessed diet matched for presented calories and nutrients, and gained about 0.9 kg in two weeks.[11][3]

    How strong is the link with disease?

    A 2024 BMJ umbrella review rated the evidence convincing for links with cardiovascular death, type 2 diabetes, anxiety and common mental disorders, but under GRADE most of the pooled analyses were low or very low quality.[4]

    Why are some scientists cautious?

    The UK's Scientific Advisory Committee on Nutrition said the studies are almost all observational, may not fully account for confounders such as smoking and income, and cannot show whether processing matters beyond the nutrients these foods contain.[12][6]

    How much of the US diet is highly processed?

    The Scientific Foundation estimates highly processed foods supply about two-thirds of the calories eaten by US youth and 60% of those eaten by adults.[10]

    Sources

    · 12 claims from 4 publishers

    Each numbered claim is a statement we checked against the sources listed with it. Tap a number in the text to see its sources in place.

    1. [1]

      A 2024 BMJ umbrella review pooled 45 meta-analyses covering almost 9.9 million people and found direct associations between ultra-processed food exposure and 32 of the 45 health measures. confirmedas of 2024-02-28

    2. [2]

      The Scientific Foundation's umbrella review of 27 meta-analyses, mostly of large cohort studies, found highly processed food intake associated with higher risk of nearly all outcomes studied, with relative risks from 1.12 for cancer to 1.55 for obesity, and rated the type 2 diabetes evidence higher certainty. confirmedas of 2026-01-07

    3. [3]

      In that trial participants ate about 508 more calories a day on the ultra-processed diet, gaining about 0.9 kg on it and losing about 0.9 kg on the unprocessed diet. confirmedas of 2019-05-16

    4. [4]

      The BMJ review rated the evidence convincing for links with cardiovascular disease death, type 2 diabetes, anxiety and common mental disorders, but under the GRADE system most pooled analyses were low or very low quality and only four were moderate. confirmedas of 2024-02-28

    5. [5]

      There is no agreed definition of ultra-processed foods; in July 2025 FDA and USDA asked for information to develop a uniform US definition for research and policy. confirmedas of 2025-07-25

    6. [6]

      SACN said it is unclear whether the associations are due to, or independent of, the high energy density, salt, saturated fat or free sugars typical of many ultra-processed foods, and that such foods may be an indicator of other unhealthy dietary patterns. confirmedas of 2023-07-11· interpretation

    7. [7]

      The 2025-2030 guidelines tell Americans to avoid highly processed packaged foods that are salty or sweet, to limit foods with artificial flavors, petroleum-based dyes, artificial preservatives and low-calorie non-nutritive sweeteners, and to avoid sugar-sweetened beverages. confirmedas of 2026-01-07

    8. [8]

      SACN found that most of the studies it reviewed used the NOVA classification, and that NOVA was the only classification meeting its screening criteria, but said its categories are very broad, group foods with differing nutritional attributes, and classify some foods discordantly with nutrition-based systems. confirmedas of 2023-07-11

    9. [9]

      Of 12 studies SACN found that applied NOVA to UK survey data, 9 included authors from the research group that developed NOVA. confirmedas of 2023-07-11

    10. [10]

      The Scientific Foundation behind the 2025-2030 Dietary Guidelines defines highly processed foods as those made mainly from substances extracted from foods or containing industrial chemical additives, and estimates they supply about two-thirds of youths' and 60% of adults' calories. confirmedas of 2026-01-07

    11. [11]

      A 2019 NIH inpatient randomized trial gave 20 weight-stable adults an ultra-processed diet and an unprocessed diet for 2 weeks each, with meals matched for presented calories, energy density, macronutrients, sugar, sodium and fiber, and let them eat as much as they wanted. confirmedas of 2019-05-16

    12. [12]

      The UK Scientific Advisory Committee on Nutrition said in 2023 that studies linking processed foods to poor health are almost exclusively observational and may not adequately account for confounders such as energy intake, BMI, smoking and socioeconomic status. confirmedas of 2023-07-11

    Revision history (1)
    1. Page created.

    Created Nov 8, 2026. Last reviewed by an editor on Nov 8, 2026. Next scheduled review: Feb 8, 2027.

    Cite this page

    "Ultra-processed foods: what the evidence shows." ContentLora, updated Nov 8, 2026. https://contentlora.com/analysis/ultra-processed-foods-evidence-debate

    Spotted an error? Suggest a correction or emailcorrections@contentlora.com.