{"schema_version":"1.0","service":"Publicasta","type":"article","id":1122,"slug":"blood_signatures_ultra_processed_foods_heart_diabetes_evidence_check","title":"Do Blood Signatures Prove Ultra-Processed Foods Cause Heart Disease? What the New Study Actually Shows","excerpt":"A new 20-year cohort study links blood-metabolite patterns associated with ultra-processed and minimally processed diets to cardiovascular disease and type 2 diabetes. The findings strengthen the biological case, but they do not turn an observational study into proof of causation.","language":"en","default_language":"en","canonical_url":"https://publicasta.com/health_myths_checked/blood_signatures_ultra_processed_foods_heart_diabetes_evidence_check?lang=en","image":{"url":"https://publicasta.com/storage/projects/38/pages/1122/2026/11/ec80e1fe-421f-442b-91e8-35def219d482.webp","alt":"Editorial still life contrasting ultra-processed and minimally processed foods with blood-sample tubes and a blurred metabolomics laboratory background."},"publisher":{"id":38,"slug":"health_myths_checked","name":"Health Myths Checked","url":"https://publicasta.com/health_myths_checked"},"author":{"name":"Anton Es"},"published_at":"2026-11-04T07:51:10+00:00","updated_at":"2026-11-04T07:51:10+00:00","content_markdown":"A new study has given the ultra-processed-food debate a more biological-looking layer: instead of relying only on what people remember eating, researchers identified patterns of molecules in the blood that were associated with diets higher in ultra-processed foods and diets higher in minimally processed foods. Those patterns were then linked to later cardiovascular disease and type 2 diabetes.\n\n ![Editorial still life contrasting ultra-processed and minimally processed foods with blood-sample tubes and a blurred metabolomics laboratory background.](https://publicasta.com/storage/projects/38/pages/1122/2026/11/ec80e1fe-421f-442b-91e8-35def219d482.webp)\n\n That is an important step in nutritional research. It is also easy to overstate. The study does not show that one packaged food caused a heart attack, that every food classified as ultra-processed is equally harmful, or that a blood test can currently tell an individual what to eat. It shows that diet patterns, blood chemistry and long-term disease risk appear to be connected in a large observational cohort.\n\n ## What the new study examined\n\n The research used data from 6,010 adults in the Whitehall II cohort, a long-running British study of health and aging. Participants provided dietary information and blood samples. Researchers measured 233 metabolites—small molecules involved in digestion, energy use and other biological processes—and used statistical modelling to identify combinations associated with reported intake of ultra-processed foods and minimally processed foods.\n\n The participants were then followed for roughly two decades for cardiovascular disease and type 2 diabetes. During follow-up, the cohort recorded 1,094 cardiovascular disease events and 853 cases of type 2 diabetes. The investigators were not simply asking whether a person ate one particular snack or drink. They were constructing broader metabolic signatures that tracked with dietary patterns.\n\n The ultra-processed-food signature included a mix of metabolites associated with a less favourable cardiometabolic profile, including patterns involving inflammation, atherogenic lipoproteins, omega-3 fatty acids and high-density lipoprotein-related measures. The minimally processed-food signature showed broadly opposing associations.\n\n When people in the highest third of the ultra-processed-food signature were compared with those in the lowest third, the higher signature was associated with a 19% higher relative risk of cardiovascular disease and a 20% higher relative risk of type 2 diabetes. A higher minimally processed-food signature was associated with lower risks: about 19% lower for cardiovascular disease and 40% lower for type 2 diabetes in the study’s statistical comparisons.\n\n Those numbers describe differences between groups. They do not mean that an individual’s personal risk rises by exactly 19% after eating a particular amount of ultra-processed food, nor that changing one item would reduce personal risk by a fixed percentage.\n\n ## What does “ultra-processed” mean here?\n\n The study used the NOVA framework, which classifies foods according to the extent and purpose of industrial processing. Ultra-processed products are generally formulations made with several ingredients and industrial techniques, often including additives, flavourings, colours, sweeteners or other components that are uncommon in ordinary home cooking. Examples can include some packaged snacks, sugary drinks, reconstituted meat products, sweetened breakfast products and ready-to-eat meals.\n\n The category is broad. It can contain products with very different nutritional profiles, portion sizes, prices, uses and social roles. A sweetened drink, a packaged whole-grain product and a frozen meal may all be placed in the same broad group even though their nutrient content and practical value differ.\n\n That breadth is one reason the research question remains contested. The label may capture several things at once: energy density, added sugar, sodium, low fibre, food texture, convenience, marketing, eating speed, portion size, replacement of fresh foods and the wider circumstances in which people eat. A statistical association with the category does not automatically identify which of those features is responsible.\n\n The new study is useful because it looks for biological pathways that might sit between diet reports and disease outcomes. But it does not solve the classification problem. Its blood signatures may reflect a mixture of food composition, overall dietary pattern, lifestyle and underlying health rather than a single effect of “processing.”\n\n ## What the blood findings add\n\n Diet studies often depend on food-frequency questionnaires or recalls. These tools are valuable, but they are imperfect. People may forget snacks, misjudge portions, simplify recipes or change their diet after developing health concerns. A metabolite pattern offers an additional source of information because it reflects what the body has recently processed and how several systems are functioning.\n\n The researchers also performed mediation analyses. In simple terms, they estimated how much of the association between dietary processing categories and later disease was statistically accounted for by the identified metabolite signatures. The estimates ranged from about 43% to 51%, depending on the dietary pattern and outcome.\n\n This is the part most likely to be reported as “blood chemistry explains half the damage.” That wording would be too strong. Mediation analysis in an observational study is not a direct measurement of a causal mechanism. It depends on assumptions about the variables included in the model, the timing of measurements and the absence of important unmeasured confounders. The percentage is a model-based estimate of statistical explanation, not a laboratory demonstration that half of the disease process has been discovered.\n\n The findings do, however, make a purely behavioural interpretation less satisfactory. If people who report different diets also show coherent differences in circulating metabolites, and those signatures are associated with later disease, researchers have a plausible biological trail to investigate. The next question is whether changing the dietary exposure changes the metabolites and whether changing the metabolites changes clinical outcomes.\n\n ## Why this still is not proof that ultra-processing causes disease\n\n The study is a prospective cohort study. That means diet and blood data were collected before many of the recorded outcomes, which is stronger than taking a snapshot after disease has already developed. It helps establish a time sequence. It does not, by itself, establish causation.\n\n People who eat more ultra-processed food may differ from people who eat less in many other ways. They may have different incomes, work schedules, sleep patterns, physical activity, smoking histories, access to food stores, stress levels, existing medical conditions or use of health services. Statistical adjustment can account for measured factors, but not for every relevant difference—and measurement errors can leave residual confounding.\n\n Reverse causation is another concern, even in a prospective design. People with early, undiagnosed metabolic problems may change what they eat, choose more convenient foods because of fatigue or mobility limits, or report their diets differently. The disease diagnosis may occur later, but the process leading to it may already have influenced the dietary data.\n\n There is also uncertainty about the exposure itself. A food-frequency questionnaire estimates usual intake; it does not provide a chemically precise record of every meal. The blood sample captures a biological state at a particular time. Neither measure perfectly represents a person’s long-term exposure.\n\n Finally, relative risks need context. A 20% relative difference can correspond to a much smaller absolute difference for someone whose baseline risk is low and a larger absolute difference for someone with other risk factors. The study did not provide a personal risk calculator, and its results should not be used as one.\n\n ## How this fits with controlled diet studies\n\n Observational studies are not the only evidence available. A 2025 randomized crossover trial published in *Nature Medicine* compared diets that followed healthy dietary guidance but differed in the amount of ultra-processed food. Participants spent periods eating each diet, allowing researchers to compare people with themselves rather than comparing entirely different populations.\n\n That trial found greater weight loss during the minimally processed diet period, despite the diets being designed to meet similar nutritional guidance. A randomized design gives the result more protection against some forms of confounding. It still had limits, including a relatively small sample and a short duration compared with the development of cardiovascular disease or diabetes. Weight change is also not the same as a diagnosis of long-term disease.\n\n Together, the two types of evidence point in the same general direction but answer different questions. The randomized trial asks what happened when participants were assigned to different dietary patterns over a limited period. The Whitehall II analysis asks whether long-term differences in diet-related blood signatures tracked with later disease across thousands of people. Neither one alone establishes that the processing label itself is the sole cause of harm.\n\n Longer and larger randomized studies would be difficult and expensive, especially if the outcome is heart disease or diabetes. Researchers therefore need to combine trials, cohorts, mechanistic studies and better dietary measurement rather than treating any single design as definitive.\n\n ## What the study cannot tell you about individual foods\n\n The results do not justify a simple list of foods that must be eliminated. The study did not test whether a particular brand, ingredient, additive or package causes disease. It did not compare every category of ultra-processed food with an equivalent minimally processed alternative. It did not establish a threshold at which a product changes from harmless to harmful.\n\n Nor does the classification mean that all processing is bad. Freezing, pasteurising, fermenting, drying, milling and canning can improve safety, shelf life, affordability and access. Some processed foods can contribute useful nutrients or make it easier for people to eat. The question is not whether food has ever been processed; most food has. The more relevant questions include what the product contains, how it fits into the overall diet, what it replaces and how often it is eaten.\n\n The same caution applies to minimally processed foods. A diet rich in minimally processed foods is not automatically adequate or suitable for every person. Nutrition depends on total dietary pattern, energy needs, cultural context, food access and medical circumstances. A category label cannot replace an assessment of the whole picture.\n\n ## Why social-media conclusions run ahead of the evidence\n\n The headline version of this research is tempting: “Scientists found the blood pathway by which ultra-processed food causes disease.” That sentence compresses several steps into one claim. The study found associations between diet-related metabolite signatures and disease outcomes. It estimated that the signatures statistically mediated part of the relationship. It did not experimentally manipulate the metabolites or prove that processing was the initiating cause.\n\n Another common leap is to treat “associated with higher risk” as “equally dangerous for everyone.” Risk is shaped by many factors, and the study’s group comparisons are not instructions for predicting an individual outcome. A third leap is to interpret a biological marker as a clinical test. The metabolite signatures were research tools derived from this cohort; they are not established diagnostic tests for personal diet quality or future cardiovascular disease.\n\n A useful evidence check therefore asks four separate questions:\n\n - Was the study observational or randomized?\n- What exactly was measured: a food, a dietary pattern, a category or a biomarker?\n- Does the result describe relative risk, absolute risk or a model-based association?\n- Was the outcome a short-term marker, a change in weight or a diagnosed disease?\n\n Keeping those questions separate makes the finding less dramatic but more informative.\n\n ## The most defensible interpretation\n\n The new analysis strengthens the case that dietary patterns high in ultra-processed foods are linked to measurable differences in blood chemistry and to higher long-term rates of cardiovascular disease and type 2 diabetes. It also suggests that the relationship is not explained only by a single nutrient or by one obvious ingredient.\n\n At the same time, it does not show that ultra-processing is a single, isolated biological cause. The category combines many products and behaviours. The study cannot remove all confounding, cannot establish a personal risk percentage and cannot identify a universal food rule. Its mediation estimates are clues about possible pathways, not proof that the pathways have been confirmed.\n\n For readers, the practical lesson is about proportion rather than purity. The evidence supports paying attention to the overall pattern of food intake, especially when packaged products regularly displace foods with more fibre, protein, vitamins and minerals. It does not support fear of every packaged item, guilt about occasional convenience foods or the idea that a blood-signature study has created a new medical test.\n\n Future research should test whether the identified metabolites change when people are assigned to different diets, whether changes persist over time and which product characteristics matter most. It should also examine affordability, cultural food practices and the effects of reducing particular subgroups rather than treating all ultra-processed foods as one uniform exposure.\n\n This article is for general information and does not provide a diagnosis, treatment plan or personalized dietary advice. Individual concerns about cardiovascular risk, blood sugar, pregnancy, chronic disease, medications or major dietary changes should be discussed with a qualified clinician or registered dietitian. It does not replace professional medical consultation.","available_translations":[{"language":"ar","title":"هل تثبت البصمات الدموية أن الأطعمة فائقة المعالجة تسبب أمراض القلب؟ ما الذي توضحه الدراسة الجديدة فعلاً","html_url":"https://publicasta.com/health_myths_checked/blood_signatures_ultra_processed_foods_heart_diabetes_evidence_check?lang=ar","markdown_url":"https://publicasta.com/health_myths_checked/blood_signatures_ultra_processed_foods_heart_diabetes_evidence_check.md?lang=ar","json_url":"https://publicasta.com/health_myths_checked/blood_signatures_ultra_processed_foods_heart_diabetes_evidence_check.json?lang=ar","api_url":"https://publicasta.com/api/public/v1/channels/health_myths_checked/articles/blood_signatures_ultra_processed_foods_heart_diabetes_evidence_check?lang=ar"},{"language":"de","title":"Beweisen Blut-Signaturen, dass hochverarbeitete Lebensmittel Herzkrankheiten verursachen? Was die neue Studie tatsächlich zeigt","html_url":"https://publicasta.com/health_myths_checked/blood_signatures_ultra_processed_foods_heart_diabetes_evidence_check?lang=de","markdown_url":"https://publicasta.com/health_myths_checked/blood_signatures_ultra_processed_foods_heart_diabetes_evidence_check.md?lang=de","json_url":"https://publicasta.com/health_myths_checked/blood_signatures_ultra_processed_foods_heart_diabetes_evidence_check.json?lang=de","api_url":"https://publicasta.com/api/public/v1/channels/health_myths_checked/articles/blood_signatures_ultra_processed_foods_heart_diabetes_evidence_check?lang=de"},{"language":"en","title":"Do Blood Signatures Prove Ultra-Processed Foods Cause Heart Disease? What the New Study Actually Shows","html_url":"https://publicasta.com/health_myths_checked/blood_signatures_ultra_processed_foods_heart_diabetes_evidence_check?lang=en","markdown_url":"https://publicasta.com/health_myths_checked/blood_signatures_ultra_processed_foods_heart_diabetes_evidence_check.md?lang=en","json_url":"https://publicasta.com/health_myths_checked/blood_signatures_ultra_processed_foods_heart_diabetes_evidence_check.json?lang=en","api_url":"https://publicasta.com/api/public/v1/channels/health_myths_checked/articles/blood_signatures_ultra_processed_foods_heart_diabetes_evidence_check?lang=en"},{"language":"es","title":"¿Las firmas sanguíneas demuestran que los ultraprocesados causan enfermedades cardíacas? Lo que realmente muestra el nuevo estudio","html_url":"https://publicasta.com/health_myths_checked/blood_signatures_ultra_processed_foods_heart_diabetes_evidence_check?lang=es","markdown_url":"https://publicasta.com/health_myths_checked/blood_signatures_ultra_processed_foods_heart_diabetes_evidence_check.md?lang=es","json_url":"https://publicasta.com/health_myths_checked/blood_signatures_ultra_processed_foods_heart_diabetes_evidence_check.json?lang=es","api_url":"https://publicasta.com/api/public/v1/channels/health_myths_checked/articles/blood_signatures_ultra_processed_foods_heart_diabetes_evidence_check?lang=es"},{"language":"fr","title":"Les signatures sanguines prouvent-elles que les aliments ultra-transformés causent les maladies cardiaques ? Ce que montre réellement la nouvelle étude","html_url":"https://publicasta.com/health_myths_checked/blood_signatures_ultra_processed_foods_heart_diabetes_evidence_check?lang=fr","markdown_url":"https://publicasta.com/health_myths_checked/blood_signatures_ultra_processed_foods_heart_diabetes_evidence_check.md?lang=fr","json_url":"https://publicasta.com/health_myths_checked/blood_signatures_ultra_processed_foods_heart_diabetes_evidence_check.json?lang=fr","api_url":"https://publicasta.com/api/public/v1/channels/health_myths_checked/articles/blood_signatures_ultra_processed_foods_heart_diabetes_evidence_check?lang=fr"},{"language":"pl","title":"Czy sygnatury krwi dowodzą, że żywność ultraprzetworzona powoduje choroby serca? Co naprawdę pokazuje nowe badanie","html_url":"https://publicasta.com/health_myths_checked/blood_signatures_ultra_processed_foods_heart_diabetes_evidence_check?lang=pl","markdown_url":"https://publicasta.com/health_myths_checked/blood_signatures_ultra_processed_foods_heart_diabetes_evidence_check.md?lang=pl","json_url":"https://publicasta.com/health_myths_checked/blood_signatures_ultra_processed_foods_heart_diabetes_evidence_check.json?lang=pl","api_url":"https://publicasta.com/api/public/v1/channels/health_myths_checked/articles/blood_signatures_ultra_processed_foods_heart_diabetes_evidence_check?lang=pl"},{"language":"ru","title":"Доказывают ли биомаркеры крови, что ультраобработанные продукты вызывают болезни сердца? Что на самом деле показало новое исследование","html_url":"https://publicasta.com/health_myths_checked/blood_signatures_ultra_processed_foods_heart_diabetes_evidence_check?lang=ru","markdown_url":"https://publicasta.com/health_myths_checked/blood_signatures_ultra_processed_foods_heart_diabetes_evidence_check.md?lang=ru","json_url":"https://publicasta.com/health_myths_checked/blood_signatures_ultra_processed_foods_heart_diabetes_evidence_check.json?lang=ru","api_url":"https://publicasta.com/api/public/v1/channels/health_myths_checked/articles/blood_signatures_ultra_processed_foods_heart_diabetes_evidence_check?lang=ru"},{"language":"zh","title":"血液特征能证明超加工食品会导致心脏病吗？新研究究竟说明了什么","html_url":"https://publicasta.com/health_myths_checked/blood_signatures_ultra_processed_foods_heart_diabetes_evidence_check?lang=zh","markdown_url":"https://publicasta.com/health_myths_checked/blood_signatures_ultra_processed_foods_heart_diabetes_evidence_check.md?lang=zh","json_url":"https://publicasta.com/health_myths_checked/blood_signatures_ultra_processed_foods_heart_diabetes_evidence_check.json?lang=zh","api_url":"https://publicasta.com/api/public/v1/channels/health_myths_checked/articles/blood_signatures_ultra_processed_foods_heart_diabetes_evidence_check?lang=zh"}],"_links":{"self":"https://publicasta.com/health_myths_checked/blood_signatures_ultra_processed_foods_heart_diabetes_evidence_check.json?lang=en","api":"https://publicasta.com/api/public/v1/channels/health_myths_checked/articles/blood_signatures_ultra_processed_foods_heart_diabetes_evidence_check?lang=en","html":"https://publicasta.com/health_myths_checked/blood_signatures_ultra_processed_foods_heart_diabetes_evidence_check?lang=en","canonical":"https://publicasta.com/health_myths_checked/blood_signatures_ultra_processed_foods_heart_diabetes_evidence_check?lang=en","markdown":"https://publicasta.com/health_myths_checked/blood_signatures_ultra_processed_foods_heart_diabetes_evidence_check.md?lang=en","json":"https://publicasta.com/health_myths_checked/blood_signatures_ultra_processed_foods_heart_diabetes_evidence_check.json?lang=en","channel":"https://publicasta.com/api/public/v1/channels/health_myths_checked","channel_articles":"https://publicasta.com/api/public/v1/channels/health_myths_checked/articles","search":"https://publicasta.com/api/public/v1/search","documentation":"https://publicasta.com/api-docs#reading-publicasta","openapi":"https://publicasta.com/api-docs/openapi.json","llms":"https://publicasta.com/llms.txt"}}