MVIIIHealth The MVIII app
← Back

Research · Food and drink

Ultra-processed foods

Two feeding trials found ultra-processed diets led to more eating and weight gain, or less weight loss, than minimally processed diets. Cohorts link high intake to more type 2 diabetes, heart disease, depression and death, mostly at low certainty. Some ultra-processed groups, such as cereals and yogurt, were linked to lower risk, and specialists sorted foods into NOVA groups inconsistently.

Add to Bio HackChecked 2026-11-13
Status
A food, not a regulated health product. Ultra-processed foods are defined within the NOVA classification, which groups foods by the extent and purpose of industrial processing. Its authors identify them by ingredients rarely used in kitchens, such as high-fructose corn syrup, hydrogenated oils and hydrolyzed proteins, or by additives such as flavors, colors, emulsifiers and sweeteners [1]. The Dietary Guidelines for Americans, 2025-2030 say to avoid highly processed packaged, prepared or ready-to-eat foods that are salty or sweet, such as chips, cookies and candy, and to limit foods with artificial flavors, petroleum-based dyes, artificial preservatives and low-calorie non-nutritive sweeteners [2].
Also called
ultra-processed food, ultraprocessed food, UPF, NOVA group 4, NOVA classification, highly processed food, processed food, junk food, packaged snacks, ready meals, ready-to-eat meals, fast food, soda, soft drinks
Studied for
convenience, quick meals, long shelf life, fortified breakfast cereals, protein bars and shakes

Strong several reviews agree · Moderate one review or consistent trials · Limited one or a few small studies · Untested no study in people, listed with what it rests on

Shown to help1

Reported risks6

What it was studied for, and what was found

Shown to help1

Moderate

Breakfast cereals, whole-grain breads and yogurt

Linked to lower death, heart disease or type 2 diabetes for some ultra-processed groups, such as breakfast cereals, whole-grain breads and yogurt, in large cohorts.

Compared with the lowest intake of each food group [3][4][5]

The numbers and sources
Effect
Breakfast cereals were linked to lower death from any cause (RR 0.85, 95% CI 0.79 to 0.92), while sugar-sweetened drinks, artificially sweetened drinks and processed meat were linked to higher death [3]. In three US cohorts, bread and cold cereals, yogurt and dairy desserts, and savory snacks were linked to lower cardiovascular disease, while sugar- or artificially sweetened drinks and processed meats were linked to higher risk [4]. In the same cohorts, cereals, dark and whole-grain breads, packaged sweet and savory snacks, fruit-based products, and yogurt and dairy desserts were linked to lower type 2 diabetes, while refined breads, sauces and spreads, sweetened drinks, animal-based products and ready-to-eat mixed dishes were linked to higher risk [5].
Population
40 cohorts with 5,750,133 people [3]. 206,957 US adults in three cohorts [4]. 198,636 US adults in the same three cohorts [5].
Sources
[3], [4], [5]
Caveats
Observational. The links for single food groups come mostly from the same three US cohorts, whose participants were over 94% white [4].

Amounts studied

Cohorts measured ultra-processed food as a share of calories, a share of grams eaten or servings a day, and compared the highest with the lowest intake [6], or modeled risk per 10% more of the diet [5]. The inpatient trial gave ultra-processed and unprocessed diets for 2 weeks each, with meals matched for the calories, energy density, macronutrients, sugar, sodium and fiber offered, and people ate as much as they wanted [7]. The UK trial gave two 8-week diets that both followed national dietary guidance to adults who usually got at least 50% of their calories from ultra-processed food [8]. Critics argue that the NOVA definition is too broad to build dietary guidelines on, and that linking nutrients to disease serves as well [9]. When food and nutrition specialists were asked to assign foods to NOVA groups, their agreement was low [10].

Risks

Reported risks5

ModerateHow often is not known

Eating more and weight gain in feeding trials

Shown in one inpatient trial to raise intake by 508 kcal a day and weight by 0.9 kg in 2 weeks against unprocessed meals. Another found less weight loss.

The numbers and sources
How common
How often is not known: the sources give no rate.
Effect
On the ultra-processed diet, people ate 508 kcal a day more (±106), from carbohydrate and fat and not protein, and gained 0.9 kg, while they lost 0.9 kg on the unprocessed diet [7]. Over 8 weeks of guideline-based diets, weight fell 2.06% on the minimally processed diet and 1.05% on the ultra-processed diet, a difference of 1.01 percentage points (95% CI -1.87 to -0.14) [8]. Pooled observational studies linked higher intake to obesity (pooled effect 1.26, 95% CI 1.13 to 1.41) [11] and (OR 1.55, 95% CI 1.36 to 1.77, low quality) [6].
Population
20 weight-stable adults, mean age 31, mean BMI 27, in a crossover trial at the NIH Clinical Center [7]. 55 adults in England with a BMI of 25 to 40, 50 analyzed, in a crossover trial [8]. 14 studies, 13 of them cross-sectional [11].
Sources
[7], [8], [11], [6]
Caveats
Two small trials. The UK trial blinded participants to the weight outcome, and mild digestive adverse events were common on both diets [8]. The observational studies were mostly cross-sectional [11].
StrongHow often is not known

Death from any cause

Linked to 21% to 29% higher death from any cause at the highest ultra-processed food intake than the lowest, in pooled cohorts, at low certainty.

The numbers and sources
How common
How often is not known: the sources give no rate.
Effect
An umbrella review found greater ultra-processed food exposure linked to higher death from any cause (RR 1.21, 95% CI 1.15 to 1.27), graded highly suggestive and low quality [6]. A meta-analysis found the highest against the lowest intake linked to higher death (RR 1.29, 95% CI 1.17 to 1.42) [3].
Population
45 pooled analyses covering 9,888,373 participants [6]. 40 cohorts with 5,750,133 people [3].
Sources
[6], [3]
Caveats
Observational. High intake goes with other habits and income that affect health (confounding), and illness can change diet (reverse causation). The authors note the lack of standard methods for classifying ultra-processed food [3].
StrongHow often is not known

Heart disease and stroke

Linked to 17% more cardiovascular disease and 23% more coronary heart disease at the highest ultra-processed food intake than the lowest, in pooled cohorts.

The numbers and sources
How common
How often is not known: the sources give no rate.
Effect
Across 22 prospective studies, the highest intake was linked to 17% (95% CI 11% to 24%) more cardiovascular disease, 23% (95% CI 12% to 34%) more coronary heart disease and 9% (95% CI 3% to 15%) more stroke. In three US cohorts the hazard ratios were 1.11 (95% CI 1.06 to 1.16), 1.16 (95% CI 1.09 to 1.24) and 1.04 (95% CI 0.96 to 1.12) [4]. An umbrella review graded the link with cardiovascular death convincing (RR 1.50, 95% CI 1.37 to 1.63) though very low in GRADE quality, and found higher heart disease death (HR 1.66, 95% CI 1.51 to 1.84) [6].
Population
22 prospective studies, and 206,957 US adults with 16,800 cardiovascular cases [4]. 45 pooled analyses [6].
Sources
[4], [6]
Caveats
Observational. Evidence quality was rated high for coronary heart disease, moderate for cardiovascular disease and low for stroke [4]. Different food groups in the category went in opposite directions (see Outcomes).
StrongHow often is not known

Type 2 diabetes

Linked to 12% more type 2 diabetes per 10% more of the diet from ultra-processed food, in pooled cohorts rated moderate to high quality.

The numbers and sources
How common
How often is not known: the sources give no rate.
Effect
Against no intake, moderate intake was linked to 12% more type 2 diabetes (RR 1.12, 95% CI 1.06 to 1.17) and high intake to 31% more (RR 1.31, 95% CI 1.21 to 1.42) [12]. In three US cohorts, the highest against the lowest fifth of intake was linked to higher risk (HR 1.46, 95% CI 1.39 to 1.54), and in a meta-analysis each 10% increment to 12% higher risk (95% CI 10% to 13%) [5]. An umbrella review graded the dose-response link convincing and of moderate quality (RR 1.12, 95% CI 1.11 to 1.13) [6].
Population
18 longitudinal studies with almost 1.1 million people [12]. 198,636 US adults with 19,503 cases, and 415,554 participants in the meta-analysis [5].
Sources
[12], [5], [6]
Caveats
Observational, with confounding and reverse causation as above. Some food groups in the category were linked to lower risk [5].
ModerateHow often is not known

Depression and anxiety

Linked to 22% more depression at higher ultra-processed food intake in pooled prospective studies, and to more anxiety symptoms in cross-sectional studies.

The numbers and sources
How common
How often is not known: the sources give no rate.
Effect
In 2 prospective studies, greater intake was linked to more later depression (HR 1.22, 95% CI 1.16 to 1.28). In cross-sectional studies it was linked to depressive symptoms (OR 1.44, 95% CI 1.14 to 1.82) and anxiety symptoms (OR 1.48, 95% CI 1.37 to 1.59) [13]. An umbrella review graded the anxiety and common mental disorder links convincing but low in GRADE quality [6].
Population
385,541 people in 17 observational studies, 15 cross-sectional and 2 prospective [13].
Sources
[13], [6]
Caveats
Mostly cross-sectional. Low mood can change what people eat (reverse causation). The umbrella review's mental health estimates come from the same pooled analysis [6][13].

Mixed results1

LimitedHow often is not known

Blood lipids and blood sugar in feeding trials

Mixed results in two feeding trials: triglycerides improved less and LDL fell more on the ultra-processed diet in one trial, and most other markers did not differ.

The numbers and sources
How common
How often is not known: the sources give no rate.
Effect
Over 8 weeks, triglycerides changed 0.25 mmol/L more favorably on the minimally processed diet (P = 0.004), and LDL cholesterol fell 0.25 mmol/L more on the ultra-processed diet (P = 0.016). HbA1c, fasting glucose, HDL, total and non-HDL cholesterol, C-reactive protein and liver tests did not differ between diets [8]. In the inpatient trial, HDL cholesterol was 55.0 mg/dL after the ultra-processed diet and 48.3 mg/dL after the unprocessed diet [7]. The later trial's authors report that the inpatient trial found no significant differences in triglycerides, hs-CRP, HbA1c, glucose or LDL between diets [8].
Population
50 adults over two 8-week diets [8]. 20 adults over two 2-week diets [7].
Sources
[8], [7]
Caveats
Two small trials of a few weeks. In the UK trial both diets followed national dietary guidance [8].

Claims tested and not found

  1. That food specialists sort foods into NOVA groups consistently. When 159 and 177 French food and nutrition specialists assigned foods to NOVA groups, agreement was low (Fleiss' kappa 0.32 for marketed foods and 0.34 for generic foods), even when ingredient lists were given [10].
  2. That every ultra-processed food carries the same risk. In large US cohorts, some groups, such as cereals, whole-grain breads and yogurt, were linked to lower heart disease and type 2 diabetes, while sweetened drinks and processed meats were linked to higher risk [4][5].
  3. That ultra-processed diets lead to more eating only because they offer more sugar, fat or salt. With meals matched for the calories, energy density, sugar, sodium, fat and fiber offered, people still ate 508 kcal a day more on the ultra-processed diet [7].

Related

Added sugarNon-sugar sweetenersRed meat and processed meatWhole grainsSalt

Sources

  1. Monteiro CA, Cannon G, Levy RB, et al. Ultra-processed foods: what they are and how to identify them. Public Health Nutrition. 2019. https://doi.org/10.1017/s1368980018003762
  2. US Department of Agriculture and US Department of Health and Human Services. Dietary Guidelines for Americans, 2025-2030. realfood.gov. 2026. https://cdn.realfood.gov/DGA.pdf
  3. Taneri PE, Wehrli F, Roa-Díaz ZM, et al. Association Between Ultra-Processed Food Intake and All-Cause Mortality: A Systematic Review and Meta-Analysis. American Journal of Epidemiology. 2022. https://doi.org/10.1093/aje/kwac039
  4. Mendoza K, Smith-Warner SA, Rossato SL, et al. Ultra-processed foods and cardiovascular disease: analysis of three large US prospective cohorts and a systematic review and meta-analysis of prospective cohort studies. The Lancet Regional Health, Americas. 2024. https://doi.org/10.1016/j.lana.2024.100859
  5. Chen Z, Khandpur N, Desjardins C, et al. Ultra-Processed Food Consumption and Risk of Type 2 Diabetes: Three Large Prospective U.S. Cohort Studies. Diabetes Care. 2023. https://doi.org/10.2337/dc22-1993
  6. Lane MM, Gamage E, Du S, et al. Ultra-processed food exposure and adverse health outcomes: umbrella review of epidemiological meta-analyses. BMJ. 2024. https://doi.org/10.1136/bmj-2023-077310
  7. Hall KD, Ayuketah A, Brychta R, et al. Ultra-Processed Diets Cause Excess Calorie Intake and Weight Gain: An Inpatient Randomized Controlled Trial of Ad Libitum Food Intake. Cell Metabolism. 2019. https://doi.org/10.1016/j.cmet.2019.05.008
  8. Dicken SJ, Jassil FC, Brown A, et al. Ultraprocessed or minimally processed diets following healthy dietary guidelines on weight and cardiometabolic health: a randomized, crossover trial. Nature Medicine. 2025. https://doi.org/10.1038/s41591-025-03842-0
  9. Gibney MJ, Forde CG, Mullally D, Gibney ER. Ultra-processed foods in human health: a critical appraisal. The American Journal of Clinical Nutrition. 2017. https://doi.org/10.3945/ajcn.117.160440
  10. Braesco V, Souchon I, Sauvant P, et al. Ultra-processed foods: how functional is the NOVA system? European Journal of Clinical Nutrition. 2022. https://doi.org/10.1038/s41430-022-01099-1
  11. Askari M, Heshmati J, Shahinfar H, et al. Ultra-processed food and the risk of overweight and obesity: a systematic review and meta-analysis of observational studies. International Journal of Obesity. 2020. https://doi.org/10.1038/s41366-020-00650-z
  12. Delpino FM, Figueiredo LM, Bielemann RM, et al. Ultra-processed food and risk of type 2 diabetes: a systematic review and meta-analysis of longitudinal studies. International Journal of Epidemiology. 2022. https://doi.org/10.1093/ije/dyab247
  13. Lane MM, Gamage E, Travica N, et al. Ultra-Processed Food Consumption and Mental Health: A Systematic Review and Meta-Analysis of Observational Studies. Nutrients. 2022. https://doi.org/10.3390/nu14132568

Everything in the library