Ultra-Processed Foods and Weight: What the Evidence Shows

By the Healthy Weight Loss 4 U Clinical Content Team · Published October 19, 2026 · Updated October 19, 2026 · 9 min read

Key takeaways

• In the NIH inpatient crossover trial, 20 adults ate about 500 kcal/day more on the ultra-processed diet and gained about 0.9 kg (2 lb) over two weeks, while losing a similar amount on the minimally processed diet ^[1].
• In the 2025 UPDATE trial (50 adults analyzed, 8 weeks per diet), weight fell 2.06% on a minimally processed diet versus 1.05% on an ultra-processed one, a 1.01 percentage point difference (95% CI 0.14 to 1.87) ^[3].
• A 2024 BMJ umbrella review of 14 meta-analyses found obesity was associated with higher ultra-processed intake (OR 1.55, low certainty), but every included study was observational and none were pooled trials ^[7].
• The 2025 Lancet series reviewed 104 prospective studies, 92 of which reported increased risk, and its authors say most of the health evidence is observational ^[4].
• The 2025-2030 Dietary Guidelines for Americans, released in January 2026, say to avoid highly processed packaged foods and set a protein serving goal of 1.2 to 1.6 g per kg of body weight per day ^[9][10].

What counts as an ultra-processed food, and what is NOVA?

Ultra-processed foods are the fourth group in NOVA, a classification that sorts foods by the purpose and extent of industrial processing. The 2025 Lancet series describes them as branded commercial products made mostly from cheap, industrially derived ingredients and additives, with little or no whole food ^[4].

NOVA has four groups: unprocessed or minimally processed foods, processed culinary ingredients, processed foods, and ultra-processed foods. Group 4 is identified by ingredients and additives not typically used in home cooking ^[4]. The U.S. dietary guidelines give chips, cookies and candy as examples of highly processed foods but do not give a formal definition ^[9].

Ultra-processed foods make up a large share of diets in some countries. Across 36 countries they ranged from 9% of energy intake in Iran to 60% in the United States ^[4]. There is not yet a single U.S. government definition. The FDA states that no uniform definition exists, and the FDA and USDA issued a request for information on July 24, 2025 ^[11]. On August 10, 2026, HHS announced that it and USDA had submitted for final review the federal government's first proposed definition of ultra-processed foods; the release does not give the definition text ^[12].

What did the NIH inpatient trial find?

The NIH trial found that people ate about 500 more calories a day on an ultra-processed diet and gained weight, but it was small and short. Kevin Hall's team at the National Institute of Diabetes and Digestive and Kidney Diseases enrolled 20 healthy adults (10 men, 10 women) for a 28-day stay at the NIH Clinical Center. Each person ate an ultra-processed diet for two weeks and a minimally processed diet for two weeks, in random order ^[1][2].

The meals were matched for calories, sugar, fiber, fat and carbohydrate as presented, and participants could eat as much or as little as they wanted. On the ultra-processed diet they ate about 500 kcal/day more, mostly from carbohydrate and fat, ate faster, and gained about 0.9 kg (2 lb) on average. On the minimally processed diet they lost a similar amount ^[1][2]. The study was published in Cell Metabolism in 2019 and funded by NIDDK ^[2].

The NIH named its own limits: a small sample, only two weeks per diet, and no way to say which feature of ultra-processed food drove the effect. Slight protein differences between the diets could account for up to half of the calorie-intake difference. Hall also cautioned that less-processed foods take more time and money to prepare, so advice to eat that way may not work for everyone without better access to healthy food ^[1][2].

What did the 2025 UPDATE trial in the UK find?

The UPDATE trial found slightly more weight loss on a minimally processed diet than on an ultra-processed diet, even though both diets followed UK healthy-eating guidance and both led to weight loss. Dicken and colleagues published it in Nature Medicine in 2025. It was a single-center, community-based crossover trial in England of adults with a BMI of 25 to under 40 who habitually got at least 50% of their energy from ultra-processed foods ^[3].

Of 55 people randomized, 50 were in the intention-to-treat analysis (90.9% female, mean age 43.2). Each person ate two 8-week diets, separated by a 4-week washout, with all food delivered to the home. Both diets followed the UK Eatwell Guide. The ultra-processed diet used nutritionally improved supermarket products rather than typical nutrient-poor ones ^[3].

Weight fell 2.06% (95% CI -2.99 to -1.13) on the minimally processed diet and 1.05% (95% CI -1.98 to -0.13) on the ultra-processed diet. The difference was 1.01 percentage points (95% CI -1.87 to -0.14; P = 0.024), or about 0.96 kg ^[3]. Fat mass fell 0.98 kg more on the minimally processed diet, and savory craving and ability to resist cravings improved more on it. LDL cholesterol fell more on the ultra-processed diet, and constipation, reflux and fatigue were more frequent on it ^[3].

Adherence and acceptability cut the other way. Self-reported adherence was 84.5% on the minimally processed diet and 91.2% on the ultra-processed diet, withdrawals were more frequent on the minimally processed diet, and participants rated it lower for flavor and for delivery and preparation. There was also a significant order effect, with less weight loss on the second diet in both arms ^[3]. The authors listed limits: energy intake and eating rate were not measured directly, body composition used bioelectrical impedance, adherence could not be closely monitored, and results may not apply to people with low ultra-processed intake. The trial reported no industry or commercial influence, and the authors caution that it does not show that moving from a low to a high ultra-processed diet is neutral ^[3].

What do large observational studies show, and what can't they show?

Observational studies link higher ultra-processed intake to higher weight and worse health, but they cannot prove that the foods caused it. A 2024 BMJ umbrella review covered 14 meta-analyses and 45 pooled analyses. Obesity was highly suggestive in the non-dose-response analysis (OR 1.55, 95% CI 1.36 to 1.77, low certainty) and was based on seven cross-sectional studies, which cannot establish what came first ^[7].

A European cohort looked forward in time. In the EPIC study of 348,748 adults aged 25 to 70 from nine countries, each 1 standard deviation increase in ultra-processed intake was linked to 0.12 kg more weight gain over a median of about 5 years (95% CI 0.09 to 0.15). Comparing highest and lowest quintiles, the risk of becoming overweight or obese was 15% higher in normal-weight participants ^[8]. That is a small average difference per person.

The BMJ review's own limits are the standard limits of this field: all included studies were observational, most certainty ratings were low or very low, diet was measured with questionnaires or recalls, and the authors found no pooled randomized trials ^[7]. The Lancet series reviewed 104 prospective studies, 92 of which reported increased risk, and judged that the evidence meets seven of nine Bradford Hill criteria for causality. Its authors also acknowledge the limits of observational data, including residual confounding, and several of them helped develop NOVA ^[4].

What do critics say about NOVA?

Critics say the definition is inconsistent and that it adds little beyond looking at nutrients. Michael Gibney, in Current Developments in Nutrition, argued that the definition of ultra-processed food changed over 2009 to 2017 and that the example lists are inconsistent, which may lead to subjective coding of food databases. He reported that in national data from the US, UK, France, Brazil and Canada higher intake tracked with sugars and fiber but not with total fat, saturated fat or sodium, and that two UK and French surveys showed no link with body mass index. He disclosed consulting work for Nestle and other food-industry funding ties ^[5].

In a 2022 American Society for Nutrition debate, Arne Astrup argued that NOVA adds little to existing nutrient profiling systems, labels some healthful plant-based and dairy alternatives as ultra-processed, and can classify fast-food and homemade versions of similar foods differently. Carlos Monteiro, who helped develop NOVA, argued that nationally representative studies show negative effects and that the evidence is strong enough to discourage ultra-processed foods in guidelines ^[6]. Monteiro acknowledged that the mechanisms linking these foods to health are not fully understood ^[6].

That is the honest state of the field. The two trials show that diets differing in processing can move weight in different directions over a few weeks. What remains unsettled is which feature drives that effect, and whether a processing-based label is the best way to guide food choice ^[1][3][6].

What do the 2025 Lancet series and the new Dietary Guidelines say?

The Lancet series argues for action now, and the U.S. Dietary Guidelines for the first time tell Americans to avoid highly processed foods. The first Lancet paper, published November 18, 2025, concluded that the totality of evidence supports the thesis that high ultra-processed intake harms health, that meta-analyses found significant associations for 12 of 15 outcomes, and that the need for more evidence should not delay public health action ^[4].

The Lancet authors cite two short trials, a 2-week US inpatient study and a 1-week Tokyo study, as showing weight and fat gain on ultra-processed diets with higher energy intake ^[4]. They also disclose that the work was funded by Bloomberg Philanthropies, which they state had no role in the content, and that only one of the 104 included studies reported industry funding ^[4].

The Dietary Guidelines for Americans, 2025-2030, were released on January 7, 2026 ^[10]. They say to avoid highly processed packaged, prepared, ready-to-eat or other foods that are salty or sweet, and to avoid sugar-sweetened beverages ^[10]. The guideline document gives a protein serving goal of 1.2 to 1.6 grams per kilogram of body weight per day, adjusted for individual caloric requirements, and says to prioritize fiber-rich whole grains and to eat whole vegetables and fruits in their original form ^[9]. It also says no amount of added sugars or non-nutritive sweeteners is recommended and sets a limit of 10 grams of added sugars per meal ^[9]. The document does not formally define highly processed foods ^[9].

What this means in practice

• Aim for a simple pattern instead of a perfect one: the federal guidelines say to prioritize whole, nutrient-dense foods and to eat whole vegetables and fruits in their original form ^[9].
• Put protein in every meal; the 2025-2030 guidelines set a serving goal of 1.2 to 1.6 g per kg of body weight per day, adjusted to individual needs, and the NIH team noted that small protein differences could explain up to half of the intake gap in its trial ^[1][9].
• Look for fiber: the guidelines say to prioritize fiber-rich whole grains, and fiber content was matched between the diets as presented in the NIH trial, so that trial did not test fiber itself ^[1][9].
• Consider cutting back on sugar-sweetened drinks, since the guidelines say to avoid them and that no amount of added sugars is recommended ^[9][10].
• Do not treat it as all or nothing: in the UPDATE trial, a guideline-based diet built from improved ultra-processed products still led to weight loss, though less than the minimally processed diet ^[3].
• Plan for time and cost, because the NIH team noted that less-processed foods take more time and money to prepare, so look for changes that fit your budget and schedule rather than an overhaul ^[1][2].
• Remember that nothing here predicts your own result: the trials averaged groups of 20 to 50 people, and no diet label replaces a clinician's review of your health ^[1][3].

When to talk to a clinician

Talk with a licensed clinician before making big changes to how you eat if you have diabetes or take glucose-lowering medicine, have another chronic medical condition, are pregnant, take a GLP-1 medicine, or have ever had an eating disorder. A clinician can review your medicines, lab results and history, and decide which food changes, including protein targets, are appropriate for you. This article summarizes research and is not a prediction of what any food pattern will do for an individual.

At Healthy Weight Loss 4 U, every program begins with a clinician evaluation, and that evaluation is the place to raise questions about ultra-processed foods, protein and your daily eating routine. Your clinician decides what, if anything, is right for you after that evaluation, and no outcome is promised.

Frequently asked questions

Sources

1. NIH study finds heavily processed foods cause overeating and weight gain. National Institutes of Health (NIDDK news release, May 16, 2019), 2019. www.nih.gov/news-events/news-releases/nih-study-finds-heavily-processed-foods-cause-overeating-weight-gain
2. Eating highly processed foods linked to weight gain (citing Hall et al., Cell Metabolism 2019, doi 10.1016/j.cmet.2019.05.008). NIH Research Matters, 2019. www.nih.gov/news-events/nih-research-matters/eating-highly-processed-foods-linked-weight-gain
3. Ultraprocessed or minimally processed diets following healthy dietary guidelines on weight and cardiometabolic health: a randomized, crossover trial. Nature Medicine, 2025. doi.org/10.1038/s41591-025-03842-0
4. Ultra-processed foods and human health: the main thesis and the evidence. The Lancet (first paper of the Series), 2025. doi.org/10.1016/S0140-6736(25)01565-X
5. Ultra-Processed Foods: Definitions and Policy Issues. Current Developments in Nutrition (PubMed Central), 2018. pmc.ncbi.nlm.nih.gov/articles/PMC6389637/
6. Are Ultra-Processed Foods Harmful? Experts Weigh the Evidence (Monteiro and Astrup debate). American Society for Nutrition, 2022. nutrition.org/are-ultra-processed-foods-harmful-experts-weigh-the-evidence/
7. Ultra-processed food exposure and adverse health outcomes: umbrella review of epidemiological meta-analyses. BMJ, 2024. bmj.com/content/384/bmj-2023-077310
8. Consumption of ultra-processed foods associated with weight gain and obesity in adults: A multi-national cohort study. Clinical Nutrition, 2021. dspace.library.uu.nl/handle/1874/443574
9. Dietary Guidelines for Americans, 2025-2030. U.S. Department of Health and Human Services and U.S. Department of Agriculture, 2026. cdn.realfood.gov/DGA.pdf
10. Fact Sheet: Trump Administration Resets U.S. Nutrition Policy, Puts Real Food Back at the Center of Health (January 7, 2026). U.S. Department of Health and Human Services, 2026. www.hhs.gov/press-room/fact-sheet-historic-reset-federal-nutrition-policy.html
11. Ultra-Processed Foods. U.S. Food and Drug Administration, 2025. www.fda.gov/food/nutrition-food-labeling-and-critical-foods/ultra-processed-foods
12. Secretary Kennedy Announces Landmark Food Policy Reforms to Advance President Trump's MAHA Agenda (August 10, 2026). U.S. Department of Health and Human Services, 2026. www.hhs.gov/press-room/hhs-announces-ultra-processed-foods-gras-reforms.html

Keep reading