--- topic: "Ultra-Processed Food" title: "Ultra-Processed Food: Harmful or Just Unhealthy?" order: 1 description: "What trials and large cohorts show about ultra-processed food, weight and disease, where the NOVA label falls short, and how to eat well without rigid rules." question: "Are ultra-processed foods harmful, or is it just calories and sugar?" answer: "Likely more than calories and sugar: in a trial people ate about 500 more calories a day on ultra-processed diets, though the category is broad." actions: - "Check the ingredient list: items you would not find in a home kitchen, such as high-fructose corn syrup, hydrogenated oils, flavour enhancers and emulsifiers, mark a product as ultra-processed [5]." - "Start with the biggest offenders: ready-to-eat processed meats and sugary or artificially sweetened drinks were the ultra-processed groups most tied to higher death rates in two US cohorts [8]." - "Aim for overall diet quality, not purity: in those cohorts, better diet quality went with lower mortality at every level of ultra-processed food intake [8]." cautions: - "Eating-disorder history, or anxiety about food? Avoid strict clean-eating rules and talk to a clinician, because extreme diets pursued for health can lead to malnutrition and impaired daily life [12]." - "If you rely on fortified shakes or soft-texture products for a medical reason, check with a dietitian before replacing them." confidence: 3 confidenceNote: "Two feeding trials found worse intake or weight outcomes on ultra-processed diets, but they were small and short, the cohort links are observational, and the category itself is debated." evidenceSummary: rctCount: 3 observationalCount: 3 metaAnalysisCount: 1 totalParticipants: "9,900,000+" replicated: true newestStudyYear: 2025 references: - id: "1" pmid: "31105044" title: "Ultra-Processed Diets Cause Excess Calorie Intake and Weight Gain: An Inpatient Randomized Controlled Trial of Ad Libitum Food Intake" authors: "Hall KD, Ayuketah A, Brychta R, et al." year: 2019 journal: "Cell Metab" - id: "2" pmid: "40760353" title: "Ultraprocessed or minimally processed diets following healthy dietary guidelines on weight and cardiometabolic health: a randomized, crossover trial" authors: "Dicken SJ, Jassil FC, Brown A, et al." year: 2025 journal: "Nat Med" - id: "3" pmid: "38418082" title: "Ultra-processed food exposure and adverse health outcomes: umbrella review of epidemiological meta-analyses" authors: "Lane MM, Gamage E, Du S, et al." year: 2024 journal: "BMJ" - id: "4" pmid: "31142457" title: "Ultra-processed food intake and risk of cardiovascular disease: prospective cohort study (NutriNet-Santé)" authors: "Srour B, Fezeu LK, Kesse-Guyot E, et al." year: 2019 journal: "BMJ" - id: "5" pmid: "30744710" title: "Ultra-processed foods: what they are and how to identify them" authors: "Monteiro CA, Cannon G, Levy RB, et al." year: 2019 journal: "Public Health Nutr" - id: "6" pmid: "26962035" title: "Ultra-processed foods and added sugars in the US diet: evidence from a nationally representative cross-sectional study" authors: "Martínez Steele E, Baraldi LG, Louzada ML, et al." year: 2016 journal: "BMJ Open" - id: "7" pmid: "34774538" title: "Randomized Controlled-Feeding Study of Dietary Emulsifier Carboxymethylcellulose Reveals Detrimental Impacts on the Gut Microbiota and Metabolome" authors: "Chassaing B, Compher C, Bonhomme B, et al." year: 2022 journal: "Gastroenterology" - id: "8" pmid: "38719536" title: "Association of ultra-processed food consumption with all cause and cause specific mortality: population based cohort study" authors: "Fang Z, Rossato SL, Hang D, et al." year: 2024 journal: "BMJ" - id: "9" pmid: "30820487" title: "Ultra-Processed Foods: Definitions and Policy Issues" authors: "Gibney MJ" year: 2019 journal: "Curr Dev Nutr" - id: "10" pmid: "36253965" title: "Does the concept of ultra-processed foods help inform dietary guidelines, beyond conventional classification systems? Debate consensus" authors: "Astrup A, Monteiro CA" year: 2022 journal: "Am J Clin Nutr" - id: "11" pmid: "37356502" title: "Dietary Guidelines Meet NOVA: Developing a Menu for A Healthy Dietary Pattern Using Ultra-Processed Foods" authors: "Hess JM, Comeau ME, Casperson S, et al." year: 2023 journal: "J Nutr" - id: "12" pmid: "26724459" title: "On orthorexia nervosa: A review of the literature and proposed diagnostic criteria" authors: "Dunn TM, Bratman S" year: 2016 journal: "Eat Behav" --- Probably more than the label numbers. In an inpatient trial, ultra-processed meals matched for calories offered, sugar, fibre and salt still led people to eat about 500 more calories a day and gain weight; they lost weight on unprocessed meals [1]. A longer trial found less weight loss on ultra-processed food even when both diets followed healthy guidelines [2]. Large cohorts link high intake to heart disease and early death, but the category is broad and cause is unproven [3][4]. ## How it works **What counts.** The NOVA system sorts foods by industrial processing. Ultra-processed foods are formulations built from ingredients rarely used in home kitchens (such as high-fructose corn syrup, hydrogenated or interesterified oils, hydrolysed proteins) or from additives that make them appealing (flavours, colours, emulsifiers, sweeteners, thickeners). A practical test is to look for one of those items in the ingredient list [5]. In the US, these foods supply about 58% of energy and about 90% of the energy from added sugars [6]. **Why they might do more than sugar and calories.** In the trial above, meals were matched for presented calories, energy density, macronutrients, sugar, sodium and fibre. People still ate about 508 more calories a day on the ultra-processed diet, from carbohydrate and fat but not protein, and the trial was not designed to show why [1]. Emulsifiers are one lead: in an 11-day feeding study of 16 healthy adults, 15 g a day of carboxymethylcellulose reduced gut microbiome diversity and caused modest abdominal discomfort [7]. See our [emulsifiers and thickeners article](/food-additives/emulsifiers-and-thickeners). **Not all of it is equal.** In two US cohorts, the link to mortality was strongest for ready-to-eat meat-based products and sugary or artificially sweetened drinks, and diet quality mattered at any intake level [8]. See our guides to [processed meat](/processed-meat/nitrates-and-cancer-risk) and [sugary drinks](/sugary-drinks/soda-juice-and-energy-drinks). **The label is contested.** Critics say the definition has shifted over time, that nutrient differences between categories are concentrated in sugars and fibre, and that national surveys in two countries found no BMI link [9]. In a published debate, both sides agreed processing affects health but disagreed on whether "ultra-processed" adds anything beyond nutrient profiling [10]. A seven-day menu with 91% of calories from ultra-processed foods still scored 86 out of 100 on a diet-quality index, though that was a paper exercise, not a trial [11]. **What to do.** Build most meals from foods that need little more than washing, cooking or freezing: vegetables, fruit, beans, eggs, fish, nuts, oats, plain yoghurt. Treat sweet drinks, processed meats and packaged snacks as the first things to cut back [8], not as a moral test. Aim for a good overall pattern; rigid purity rules can backfire [12]. ## The evidence in detail ### Feeding trials In the NIH inpatient trial, 20 weight-stable adults (mean age 31, BMI 27) were randomised to ultra-processed or unprocessed diets for 2 weeks, then switched for 2 weeks, eating as much or as little as they liked. Intake was 508 ± 106 kcal a day higher on the ultra-processed diet (P = 0.0001), from carbohydrate (+280 kcal) and fat (+230 kcal) but not protein. Participants gained 0.9 ± 0.3 kg on the ultra-processed diet and lost 0.9 ± 0.3 kg on the unprocessed one [1]. It was small and brief, so it shows intake and short-term weight change, not long-term outcomes. The UK trial randomised 55 adults with overweight or obesity who habitually got at least half their energy from ultra-processed food to two 8-week diets that both followed the Eatwell Guide. Both caused weight loss, but more on the minimally processed diet (-2.06%, 95% CI -2.99 to -1.13) than on the ultra-processed diet (-1.05%, -1.98 to -0.13); the difference was -1.01 percentage points (-1.87 to -0.14, P = 0.024). Mild digestive side effects were common on both diets [2]. ### Cohorts and pooled analyses An umbrella review of 45 pooled analyses (9,888,373 people) found direct associations with 32 health outcomes. Evidence was graded convincing for cardiovascular death (RR 1.50, 95% CI 1.37 to 1.63) and type 2 diabetes (dose-response RR 1.12, 1.11 to 1.13), and highly suggestive for all-cause mortality (RR 1.21, 1.15 to 1.27); most pooled analyses were rated low or very low quality by GRADE, and all were observational [3]. In NutriNet-Santé (105,159 adults, median follow-up 5.2 years), each 10-point rise in the share of ultra-processed food was linked to 12% more cardiovascular disease (HR 1.12, 1.05 to 1.20), unchanged after adjusting for saturated fat, sodium, sugar and fibre [4]. A longer US analysis (74,563 women and 39,501 men, 30 to 34 years, 48,193 deaths) found the highest versus lowest quarter had only 4% higher all-cause mortality (HR 1.04, 1.01 to 1.07), driven by non-cancer, non-cardiovascular causes, with no link to cancer or cardiovascular death. Meat-based ready-to-eat products carried hazard ratios of 1.06 to 1.43, sugary and artificially sweetened drinks 1.09, dairy desserts 1.07 and breakfast foods 1.04; better diet quality was linked to lower mortality within every quarter of ultra-processed intake [8]. ### Mechanisms and exposure In NHANES 2009-2010 (9,317 people), ultra-processed foods provided 57.9% of energy and 89.7% of energy from added sugars; added sugars rose from 7.5% of energy in the lowest quintile to 19.5% in the highest [6]. The emulsifier study compared 7 people on carboxymethylcellulose with 9 on an emulsifier-free diet; it lowered short-chain fatty acids, and 2 participants showed bacteria encroaching on the gut's inner mucus layer, but the sample was tiny [7]. ### Where the evidence disagrees The cohort effects are modest and inconsistent across outcomes (cardiovascular disease in one cohort, no cardiovascular mortality link in another) and rest on food-frequency or diet-record data that must be sorted into NOVA groups [4][8]. Critics argue the category is poorly defined, misclassifies foods and is confounded with obesity-prone lifestyles, whereas supporters note associations persist after nutrient adjustment [9][10]. Both feeding trials were short, and the second showed weight loss on both diets [2]. A healthy-pattern menu made mostly of ultra-processed foods was feasible on paper but lacked vitamin D, vitamin E and choline [11]. **Bottom line (confidence 3 of 5).** Ultra-processed diets probably drive overeating and slightly worse health, but "ultra-processed" lumps together very different foods. Eat mostly whole foods, cut the worst offenders first, and avoid turning food rules into a source of anxiety.