Blood Type Diet: Is There Any Scientific Evidence?

Blood type diet evidence — at a glance, here are the most important points to walk away with before you read the deep dive below.

• D'Adamo's evolutionary timeline for blood types is contradicted by genetics — all four ABO types predate Homo sapiens and appear in our primate relatives. • A 2013 American Journal of Clinical Nutrition review found only one study ever directly tested the theory, and it found no support for it. • A 2014 University of Toronto study of 1,455 people found each blood-type diet's benefits held equally regardless of the person's actual blood type. • People who feel better on blood type diets are likely responding to eating more whole foods and less sugar and processed food, not blood type matching. • Dietary lectins are mostly destroyed by cooking, soaking, and fermenting — blanket lectin avoidance eliminates legumes and whole grains linked to lower disease risk. • Legitimate personalised nutrition is driven by genetics, gut microbiome, metabolic health, and individual food sensitivities — not blood type.

D'Adamo's theory rests on several interconnected claims. First, that the four blood types evolved at different stages of human history: Type O is supposedly the oldest, emerging when humans were primarily hunter-gatherers; Type A arose with the development of agriculture; Type B with nomadic pastoral societies; and Type AB is the most recent hybrid. Second, that lectins — proteins found in many foods — interact differently with the surface antigens of different blood types, causing agglutination (clumping) of blood cells and leading to health problems when the wrong lectins are consumed.

Third, that each blood type has different optimal levels of stomach acid, intestinal enzymes, and stress hormones, making some foods easier to digest for certain types. The theory is internally consistent and intuitively appealing, which partly explains its popularity. However, scientific theories must be evaluated on evidence, not intuitive appeal. The evolutionary timeline D'Adamo proposes for blood type emergence is not supported by genetic evidence — all four ABO blood types appear to be millions of years old, predating Homo sapiens entirely and existing in our primate relatives. The lectin-agglutination mechanism, while theoretically possible, has not been demonstrated to occur at dietary concentrations in living humans.

A systematic review published in the American Journal of Clinical Nutrition in 2013 searched the entire medical literature for studies testing the blood type diet hypothesis. Of the 1,415 references identified, only one study directly tested the theory, and it did not support the blood type diet's claims. A subsequent large study published in PLOS ONE in 2014 by researchers at the University of Toronto examined 1,455 participants, analysed their diets, blood types, and cardiometabolic markers, and found no evidence that blood type modified the relationship between diet and health outcomes.

Specifically, the Toronto study found that following a Type A diet (plant-based emphasis) was associated with lower BMI, blood pressure, and cholesterol regardless of whether the person actually had Type A blood. Similarly, following a Type O diet (higher protein, lower carbohydrate) was associated with lower triglycerides in all blood types equally. The health effects of each dietary pattern were explained entirely by the nutritional composition of the diet itself, not by any interaction with blood type. A 2018 study in the Journal of the Academy of Nutrition and Dietetics reached the same conclusion. As of 2026, no published peer-reviewed study has found evidence supporting blood type-specific dietary recommendations.

Despite the lack of supporting evidence, many people report genuine improvements after adopting their blood type diet. This is not surprising and does not validate the theory. The blood type diet, regardless of which type you follow, replaces processed foods with whole foods, increases vegetable intake, reduces sugar and refined grains, and encourages mindful eating. These changes would make almost anyone feel better, regardless of blood type.

The placebo effect also plays a role. When people believe a dietary change will help them, they often experience subjective improvements in energy, digestion, and wellbeing. This psychological effect is real and measurable but does not indicate a blood type-specific mechanism. Additionally, the restriction of certain food groups may inadvertently remove a food that a person is actually sensitive to — for example, a Type O person who removes wheat and dairy might feel better because they happen to have undiagnosed lactose intolerance, not because of their blood type. Confirmation bias then reinforces the belief that the blood type diet works: improvements are attributed to the diet while failures are attributed to imperfect adherence. This is a common pattern with unfalsifiable health claims.

Lectins are a real class of proteins found in many plant foods, and some lectins are genuinely harmful in large quantities — raw kidney bean lectin (phytohaemagglutinin) can cause severe food poisoning, which is why kidney beans must be thoroughly cooked. However, D'Adamo's specific claim that dietary lectins interact with blood type antigens in a clinically meaningful way has not been substantiated. Most dietary lectins are destroyed by cooking, and those that survive are present in quantities too small to cause the widespread agglutination that the theory requires.

The lectin conversation has been further confused by the popularity of Steven Gundry's Plant Paradox, which argues that lectins are a primary cause of modern disease. While certain lectins can be problematic for certain individuals (particularly those with inflammatory bowel conditions), blanket lectin avoidance eliminates some of the healthiest food groups — legumes, whole grains, tomatoes, peppers — that are consistently associated with reduced disease risk in large population studies. The longest-lived populations on Earth (Blue Zones) consume lectin-rich foods daily. Cooking, soaking, sprouting, and fermenting all substantially reduce lectin content in foods, making raw lectin exposure largely a non-issue in a diet that includes cooked foods.

The desire for personalised dietary advice is legitimate — we are not all identical, and individual responses to food do vary. However, the variables that meaningfully differ between individuals are not blood type but rather genetics (including nutrient metabolism genes like MTHFR, and taste receptor variants), gut microbiome composition, metabolic health (insulin sensitivity, liver function), physical activity level, age, sex, and individual food sensitivities. Emerging fields like nutrigenomics and microbiome-based nutrition are beginning to provide genuinely personalised dietary recommendations based on measurable biological variables.

For most people, the most impactful dietary personalisation is identifying individual food sensitivities through elimination protocols, understanding personal satiety and hunger cues, adapting macronutrient ratios to activity levels and metabolic health, and accounting for cultural food preferences and practical constraints. A plant-rich, whole-food diet with adequate protein, healthy fats, and minimal ultra-processed food improves health markers across all blood types, all genetic backgrounds, and all microbiome profiles. This universal foundation, adjusted for individual tolerances and preferences, is far more evidence-based than blood type categorisation.

The blood type diet is one of many personalised-nutrition narratives that promise to translate a single biomarker into a comprehensive diet plan. Body-type diets (ectomorph/mesomorph/endomorph), metabolic-type diets, and many at-home microbiome test kits make broadly similar promises. A few practical filters help separate genuinely useful claims from marketing dressed in scientific language. First, ask whether the biomarker has been shown in randomised trials to actually predict differential response to food — not just correlate with anything. For blood type, no such trial has succeeded. Second, ask whether the dietary recommendation depends on the biomarker, or whether it is the same generic whole-food advice repackaged. If removing the biomarker produces sensible nutrition advice, the biomarker is decorative. Third, check for unfalsifiability: claims that adapt to every counterexample (you did not feel better because you did not follow it strictly enough) are warning signs. Fourth, ask whether reputable bodies — Academy of Nutrition and Dietetics, British Dietetic Association, Cochrane reviews — endorse the approach. Finally, look at opportunity cost. Following an unsupported personalisation framework means you are not following one that does have evidence, like a [Mediterranean-style pattern](/blog/mediterranean-diet-beginners-guide/) or a [DASH approach for blood pressure](/blog/dash-diet-blood-pressure-guide/). The cost is invisible because nothing dramatic goes wrong — but better outcomes were available.

The field of genuinely evidence-based personalised nutrition has moved on substantially from blood type theorising. Continuous glucose monitoring has made it possible to identify individual postprandial glucose responses, with research like the Israeli PREDICT studies showing that the same food can produce very different blood sugar responses in different people — and that this variability is partly predictable from gut microbiome composition, anthropometrics, and dietary history rather than from blood type. Validated genetic tests can identify variants relevant to caffeine metabolism (CYP1A2), lactose tolerance (LCT), alcohol metabolism (ADH1B, ALDH2), and folate metabolism (MTHFR), with modest but real implications for diet. Food sensitivity testing using elimination-and-reintroduction protocols remains the gold standard for identifying personal trigger foods — far more reliable than the IgG food panels marketed online, which most professional allergy societies explicitly reject. Microbiome stool testing is improving but still produces variable, hard-to-act-on results in routine settings. For most people, the highest-value personalisation in 2026 is still the simplest: keep a 14-day food and symptom diary, use a CGM for two weeks if you can access one, and review the data with a registered dietitian. That combination identifies more genuinely useful, actionable personalisation than any blood-type chart ever could.

If you found this guide useful, the following deeper reads expand on neighbouring topics and will help you put the principles into practice across the rest of your kitchen routine: Type 2 Diabetes Diet Plan: Evidence-Based Strategies for Blood Sugar Control, Type 2 Diabetes Diet: Evidence-Based Foods That Help Control Blood Sugar, Suboptimal magnesium status in the United States: are the health consequences underestimated?, Primary Prevention of Cardiovascular Disease with a Mediterranean Diet Supplemented with Extra-Virgin Olive Oil or Nuts. Each of these has been written to stand alone, so dip in wherever the topic feels most relevant to what you are working on this week — together they form a connected library of practical, evidence-based home-cooking knowledge that grows more useful the more of it you read.

The guidance in this article draws on peer-reviewed nutrition and food-science literature as well as guidance from major public-health bodies. Key reference sources we have consulted while writing and updating this piece include:

• Harvard T.H. Chan School of Public Health, *The Nutrition Source*, 2024. • U.S. National Institutes of Health (NIH), Office of Dietary Supplements, fact sheets, 2024. • World Health Organization (WHO), Healthy Diet fact sheet, 2024. • Cochrane Database of Systematic Reviews — relevant systematic reviews, 2020–2024. • British Dietetic Association (BDA) Food Fact Sheets, 2024.

These references are provided so that motivated readers can verify claims and explore the underlying evidence directly. Where a specific trial, meta-analysis, or named author is referenced in the body of the article, that citation takes precedence over the general sources listed here. The article is reviewed periodically against newly published evidence and updated when meaningful new findings emerge.