Blood Type Diet

Diet Reference Archive: a commercial diet built on a single claim

Blood Type Diet

A diet that gives each person a food list according to their ABO blood type, from the books of naturopathic doctor Peter D’Adamo. It is promoted as a way to improve digestion, energy and weight and to lower the risk of chronic disease.

Family: Fad, extreme, and debunked dietsPromoted by: Peter D’Adamo, ND, through his books, his website and the supplement and testing company that sells under his nameFirst appeared: 1996, with the book Eat Right 4 Your TypeGRADE: Very low for whether a closer match between usual eating and the Type A plan goes with a larger difference in systolic blood pressure in people with blood type A than in people of other blood types, young adults, one cross-sectional study; this page’s own judgment (Wang et al., 2014)

ON THIS PAGE

1. In brief
2. Overview
3. In practice
4. For the coach
5. The evidence (opens on request)
6. Mechanisms (opens on request)
7. Risks, contraindications and myths (opens on request)
8. Key terms
9. Review questions
10. References
11. Evidence rating and review record

The summary and sections 1 to 3 are the short version and are open when the page loads. Sections 4 to 6 hold the research detail, the biology, and the cautions. Each of those opens when you select its heading, and you can read the page without opening them. Back to the Diet Reference Archive index.

In brief

What is it? A plan from the 1996 book Eat Right 4 Your Type by Peter D’Adamo, a naturopathic doctor, that gives each of the 4 ABO blood types its own list of foods to eat, foods that are neutral and foods to avoid. The plan’s own summary is that type O does best on a lean, high-protein diet, type A on a mainly vegetarian diet, type B on a mixed diet of meat, fish and dairy, and type AB on a modified vegetarian diet (D’Adamo, n.d.-c).

Does it work? The diet’s central claim is that a person does better on the list matched to their own blood type. A 2013 systematic review found no study of the health effects of ABO blood type diets (Cusack et al., 2013). Since then, 2 Canadian studies have scored people’s usual eating against the type plans, and an analysis inside a trial of a low-fat vegan diet compared people with type A, the type told to eat a mainly vegetarian diet, with people of other blood types (Barnard et al., 2021; Wang et al., 2014, 2018). The available studies have not established an advantage from matching food lists to blood type. Several reported associations with eating patterns were similar across blood types, but these observational analyses do not establish equivalence or the effects of following the complete book plan. The certainty of this evidence is very low.

Is it safe? The searches for this page did not identify a study that measured harm from following the plan. The cautions come from what the lists remove: the type O plan limits dairy foods and wheat, and the type A plan limits meat and dairy (Deutsche Gesellschaft für Ernährung, 2013), so a person may need other sources of nutrients such as calcium (National Institutes of Health Office of Dietary Supplements, 2023). The company that sells under the author’s name also sells supplements and home tests, and its own pages state that their statements have not been evaluated by the U.S. Food and Drug Administration (D’Adamo Personalized Nutrition, n.d.-f).

Who should not do it without talking to a doctor or dietitian first? Anyone taking insulin or a sulfonylurea for diabetes, taking warfarin, or living with chronic kidney disease; anyone pregnant or breastfeeding; children and teenagers; anyone with a current or past eating disorder; and anyone hoping to use the plan or its supplements to treat a diagnosed condition.

Related pages: Vegetarian Diet, Vegan Diet, Paleo Diet, Elimination Diets. Full list on the archive index.

1. Overview

What you eat

The plan sorts foods for each blood type into 3 groups: foods to eat, foods that are neutral, and foods to avoid (Wang et al., 2014). The lists are long and specific, naming individual meats, fish, dairy foods, beans, grains, vegetables, fruits, oils, spices and drinks. This page did not open the book itself. It describes the plans from the diet’s own website and from published accounts of the food lists.

Type O. The website’s summary is “Blood Type O thrives on a lean, high protein diet” (D’Adamo, n.d.-c). Its type O page recommends lean, organic meats, vegetables and fruits, and avoiding wheat and dairy (D’Adamo, n.d.-b). Researchers who scored the food lists describe the type O plan as promoting meat and the avoidance of grain products (Wang et al., 2014).

Type A. The summary is that type A “thrives on a primarily vegetarian diet” (D’Adamo, n.d.-c). The type A page tells readers used to meat to move toward soy proteins, grains and vegetables, eaten “as natural a state as possible: pure, fresh and organic” (D’Adamo, n.d.-a). The same researchers describe the type A plan as high in grains, fruits and vegetables (Wang et al., 2014).

Types B and AB. Type B is told to eat “a mixed diet of meat, fish and dairy,” and type AB “a modified vegetarian diet” (D’Adamo, n.d.-c). In the researchers’ description, the type B plan recommends plenty of dairy foods and moderate amounts of other food groups, and the type AB plan is similar but more restrictive, with eggs and fish as its recommended animal protein foods (Wang et al., 2014).

Later books add a second layer. Live Right 4 Your Type adds secretor status, whether a person’s blood type substances also appear in saliva and other body fluids, as a refinement of the lists (D’Adamo, n.d.-c; D’Adamo, n.d.-e). The plan also assigns exercise by type: brisk aerobic exercise for type O and calming exercise such as hatha yoga, tai chi and meditation for type A (D’Adamo, n.d.-a, n.d.-b).

What you avoid or limit

Each type has its own list of foods to avoid. The German Nutrition Society, in a 2013 press release, summarized the lists this way: for type O, some legumes, wheat, milk and milk products are said to suit the person less well; for type A, meat and milk products, along with some beans and wheat; for type B, almost nothing is excluded apart from a few foods such as chicken; and for type AB, the plan favors little meat with dairy foods, grains and eggs (Deutsche Gesellschaft für Ernährung, 2013). The website presents the avoid lists as the heart of the plan, and it also advises starting gradually, for example by replacing at least 2 of the foods listed as weight-gaining for one’s type, such as wheat bread and pasta for types O and A (D’Adamo, n.d.-g).

Where it came from

Peter D’Adamo is a naturopathic doctor. His website gives his degree as a naturopathic doctorate from Bastyr College in Seattle in 1982, and says his work builds on a theory his father advanced decades earlier (D’Adamo, n.d.-d, n.d.-g). Eat Right 4 Your Type, written with Catherine Whitney, was published by G. P. Putnam’s Sons in 1996 (Open Library, n.d.-b), and a revised and updated edition marking its 20th anniversary followed in 2016 (D’Adamo Personalized Nutrition, n.d.-b; Open Library, n.d.-c).

The theory, as researchers have summarized it, is that each blood type arose at a different point in human history and is suited to the food of that time. Type O is cast as the ancestral hunter, suited to the high-animal-protein eating of hunter-gatherers; type A as the farmer, suited to a vegetarian diet; type B as the nomad, suited to dairy; and type AB as a mix of A and B. A second strand concerns lectins, sugar-binding proteins in foods, which the theory says can make blood cells clump if they do not suit a person’s blood type (Wang et al., 2014).

The book sold widely, though the company’s own figures vary from page to page. One page on the diet’s website says Eat Right 4 Your Type has “more than 5 million copies in print in over 50 languages,” and adds that the author’s books have sold over 10 million copies in more than 65 languages (D’Adamo, n.d.-g). Another says the book has “more than 10 million hardcover copies in print, translated into 70 languages” (D’Adamo, n.d.-c). Researchers writing in 2014 cited 7 million copies in over 60 languages (Wang et al., 2014). These are sales claims, not measures of how many people follow the plan.

What it claims to do

The website states that “your blood type is the single most important determinant in how your body interacts with the foods you eat and whether those foods cause inflammation, digestive problems, or weight gain for you” (D’Adamo, n.d.-g). For type A it says the plan’s mix of lifestyle advice, exercise and dietary guidelines will “decrease your natural risk factors for cancer, diabetes and cardiovascular disease” (D’Adamo, n.d.-a). It describes weight loss as “a side benefit” rather than the purpose (D’Adamo, n.d.-a).

The website also reports its own visitor results: of 6,617 people who reported following the plan for a month or more, it says 71 to 78 percent had significant improvement in a variety of health conditions, most often weight loss (D’Adamo, n.d.-g). Those are reports that visitors chose to submit, with no comparison group. The company’s product pages carry the statement “Our products and services are not intended to diagnose, treat, cure or prevent any disease” (D’Adamo Personalized Nutrition, n.d.-f). These are the promoter’s claims, reported as claims; section 4 reports how far research supports them.

What it gets right

Several of the plans push in directions that have good evidence of their own. The type A plan’s emphasis on vegetables, fruit, grains and plant protein, and the call for food in a natural state, overlap with features of a heart-healthy eating pattern in the American Heart Association’s 2026 dietary guidance: “eating plenty of vegetables and fruits and choosing a wide variety,” choosing healthy sources of protein, and choosing minimally processed foods instead of ultraprocessed foods (Lichtenstein et al., 2026). The same guidance favors whole grains over refined grains, which the type O plan limits.

The Toronto researchers who scored people’s usual eating against the 4 plans found that eating more like the type A plan went with somewhat better blood pressure, cholesterol and insulin measures, and they noted that the plan resembles patterns health agencies recommend. They reported these associations in people of each blood type and did not detect a clear difference between blood types for most measures (Wang et al., 2014). That is the useful distinction for a coach: a client may be drawn to the plan for reasons that line up with good evidence, such as eating more vegetables, even though the blood type part of the plan has not been shown to add anything. The website’s advice to type O followers to eat seated at a table and to chew slowly (D’Adamo, n.d.-b) also overlaps with practices described on the Mindful Eating page.

How clients describe it, and what it gets confused with

Clients say “I’m eating for my blood type,” “I’m type O, so I need meat,” “I’m an A, I’m supposed to be vegetarian,” “Eat Right 4 Your Type,” “the blood type diet,” or “I’m a non-secretor.” Some will have a home blood typing card or a secretor test result, or will mention the company’s software or supplements. In a 2026 survey of 580 female university students in Türkiye, about 39 percent had heard of a blood group diet and 7 said they had followed one (Kahve & Bayrak, 2026).

The plan is confused with several other things. The Paleo Diet also builds a food list from an account of ancestral eating, though not by blood type. The type A plan overlaps with the Vegetarian Diet and, at its strictest, the Vegan Diet, and the type O plan’s emphasis on meat and avoidance of grains resembles low-carbohydrate eating (Wang et al., 2014), a topic on the archive index as Low-Carbohydrate Diets. Removing foods to find out what causes symptoms is a clinical method covered on the Elimination Diets page; it does not depend on blood type. Related topics include the Alkaline Diet and Clean Eating.

The plan is also sometimes taken for genetic or personalized nutrition testing. In a 2022 focus group study of dietitians in the Philippines, several participants equated gene-based nutrition advice with recommending the blood type diet (Nacis et al., 2022). A blood type is one inherited trait; this page reports what the evidence shows about using it to choose food.

2. In practice

A typical day of eating

What a day looks like depends on the person’s blood type, so 2 people at the same table can be following different lists. The sketches below are built from the general guidance on the diet’s website; the book’s own menus were not opened for this page.

A person following the type O plan would build meals from lean meat, vegetables and fruit, with no wheat bread or pasta and no milk, yogurt or cheese; where bread is wanted, the website suggests spelt in place of wheat (D’Adamo, n.d.-b, n.d.-g). A person following the type A plan would build meals from grains, soy foods such as tofu, vegetables and fruit, with no meat (D’Adamo, n.d.-a; Wang et al., 2014). Which particular meats, grains, vegetables and fruits are allowed is set by each type’s detailed lists, which matter to followers more than these broad outlines suggest.

Cost, products, and who profits

The searches for this page found no study of what the plan’s food costs compared with usual eating. Cost depends on the type: a plan built around meat costs something different from a mainly vegetarian one. The products sold under the diet’s name are priced on the company’s own site. On November 8, 2026, the official store listed, at the same prices as on October 21, 2026, the revised edition of Eat Right 4 Your Type at $19.95, a home blood typing kit at $11.95, the 2 together at $24.95, and the official app at $3.99 (D’Adamo Personalized Nutrition, n.d.-b, n.d.-c, n.d.-d, n.d.-f).

The same store sold more personalized products at higher prices: the SWAMI Xpress diet software at $69.95, a secretor status collection kit at $109.95, and a bundle of the secretor kit, a genotyping kit and the software at $173.95 (D’Adamo Personalized Nutrition, n.d.-e, n.d.-f). It also sold supplements formulated for each blood type. One, a “lectin blocker” for type O, was listed at $51.95; the company says it “blocks harmful lectins based on your blood type” (D’Adamo Personalized Nutrition, n.d.-a). The searches for this page did not identify a trial of that product. Teas labeled for types A and B were listed at $18.95 each, and a book titled Eat Right for Your Baby at $13.95 (D’Adamo Personalized Nutrition, n.d.-b). Prices are a snapshot and change.

The company’s pages identify the seller as D’Adamo Personalized Nutrition and North American Pharmacal, Inc., and state: “The statements made on our websites have not been evaluated by the FDA (U.S. Food & Drug Administration)” (D’Adamo Personalized Nutrition, n.d.-f). The books and the software carry the author’s name and are sold in the same store as the supplements. The home kit page notes that the test is “not for screening purposes prior to transfusion” (D’Adamo Personalized Nutrition, n.d.-d). A kit is not the only way to learn a blood type: the American Red Cross, for example, invites blood donors to “donate and find out” (American Red Cross, n.d.).

Money runs through the research as well. The senior author of the 2014 Toronto study disclosed shares in a company that sells genetic testing for personalized nutrition (Wang et al., 2014), and the diet’s website points to that tie in its response to the research (D’Adamo, n.d.-f). The trial of a vegan diet described in section 4 was funded mainly by the Physicians Committee for Responsible Medicine, a nonprofit organization that provides nutrition education and research, whose employees took part in running and reporting it (Kahleova et al., 2020; “Clarification of Role,” 2021). A financial tie is a reason for a reader to look closely at how a study was done; it is not by itself evidence that a result is wrong.

For someone with limited money or time, the plan’s demands are its long lists, specialty items such as spelt products, and, on the type O plan, a meat-centered diet. None of the products is needed to eat more vegetables and fruit, which is the part of the plans that lines up with good evidence.

Environmental footprint

The searches for this page found no study of the plan’s environmental footprint. The footprint depends mostly on which type’s list a person follows, since the type O plan centers on meat and the type A plan is mainly vegetarian; the Vegetarian Diet page describes the research on vegetarian eating.

Cultural fit

This is a description of the plan’s design, not a research finding. The plan assigns food by blood type rather than by a person’s food traditions, so a family that shares meals may hold 2 or more different lists, and a person may be told to give up foods central to their culture’s cooking, such as wheat breads for type O or meat for type A. Blood type frequencies also differ by ancestry: the American Red Cross reports that about 45 percent of white people, 51 percent of African American people and 57 percent of Hispanic people in the United States are type O (American Red Cross, n.d.), so the meat-centered list falls more often on some groups than others. A client can keep any part of the plan that suits their own kitchen without adopting the rest.

What makes people stop

The searches for this page did not identify a study of how long people follow the plan or why they leave it. The researchers who scored usual eating against the plans could measure only how closely people’s eating resembled each list, not who followed one strictly (Wang et al., 2014), and the diet’s website, criticizing that study, says its participants were following only 13.7 percent of the diet (D’Adamo, n.d.-f). What follows is observation rather than data: long and specific food lists, family meals with different lists, eating out, the cost of specialty foods and products, and the absence of a promised change are the reasons a person is likely to stop.

What happens when you stop

The searches for this page did not identify any study that followed people after they stopped the plan. The 2 Canadian studies covered a single point in time and 6 months, and the vegan diet trial lasted 16 weeks (Barnard et al., 2021; Wang et al., 2014, 2018), so there is no evidence on which changes, if any, last beyond that.

Common variations

Named versions and add-ons a client may mention include Live Right 4 Your Type, which adds secretor status (D’Adamo, n.d.-c); condition-specific books in an Eat Right 4 Your Type health library, such as a volume on diabetes (Open Library, n.d.-a); a book titled Eat Right for Your Baby (D’Adamo Personalized Nutrition, n.d.-b); the SWAMI software, which the company says builds a plan from blood type, secretor status, genetics and other personal markers (D’Adamo Personalized Nutrition, n.d.-e); and type-specific supplements, teas and probiotics. The website also links blood type to personality and stress (D’Adamo, n.d.-a); this page covers only the diet.

3. For the coach

SCOPE OF PRACTICE

A health and wellness coach does not prescribe, adjust, endorse or discourage the Blood Type Diet or any other way of eating, and does not diagnose or treat. The coach helps the client understand what they have chosen, notice how it is going, and decide what to do next.

A coach does not order, request or interpret blood tests, including blood typing and secretor status tests, and does not suggest a home kit. A client’s blood type is the client’s own information. The coach does not need it, does not record it, and does not tell a client what it means; if a client shares it, the coaching conversation does not depend on it. A client who wants to use the plan or its supplements to treat a diagnosed condition takes that to the physician and a registered dietitian.

The cautions found for this page fall on particular groups rather than on everyone who follows the plan. For these clients, the coach makes sure the right professional is involved before a substantial change in eating:

• Taking insulin or a sulfonylurea for diabetes: the client keeps the prescriber informed before the change, because a plan that removes grains or other carbohydrate foods can lower blood glucose.
• Taking warfarin: the client keeps the prescriber informed before the change, because the amount of vitamin K in the diet needs to stay consistent.
• Chronic kidney disease: the kidney care team and a registered dietitian, because protein and mineral needs depend on the stage of disease, lab results and dialysis.
• Pregnant or breastfeeding: the obstetric clinician and a registered dietitian.
• A child or teenager: the pediatric clinician and a registered dietitian.
• A current or past eating disorder or disordered eating: the client’s clinician, through the coach’s own referral process or an existing care plan.
• A diagnosed condition the client hopes the plan or its supplements will treat: the physician who manages that condition.

Who does what

The physician

Decides whether a substantial change in eating is medically appropriate for this person, given their conditions and medicines. Adjusts diabetes medicine or warfarin if the client’s eating changes, and evaluates any symptom the client reports. If the client wants to know their blood type for a medical reason, such as pregnancy or an upcoming surgery, orders and explains that test. Reviews any supplement the client is taking, with the pharmacist where needed.

The registered dietitian

Checks whether the list a client has chosen meets their nutrient needs, for example calcium on a plan that limits dairy foods or vitamin B12 on a mainly vegetarian plan, and plans replacements. Adapts eating to a diagnosed condition such as diabetes or kidney disease, and plans eating in pregnancy, childhood or alongside a history of disordered eating.

The health and wellness coach

Finds out what drew the client to the plan and what they hope it will change. With the client’s permission, shares what the research has and has not found, and explores how the client wants to use it. Helps the client separate the parts of the plan that matter to them, such as eating more vegetables or cooking at home, from the parts they feel pressured by, and helps them build habits that fit their real life and household. Notices how things are going and holds the referral line.

What a client bringing this diet needs

A client may come to the plan because it promises something personal: a diet made for them rather than for everyone. Some have tried other diets and are looking for an explanation of why those did not suit them. Some feel better on the plan and want to keep going; others feel anxious about the foods on their avoid list, or are struggling to cook for a family with different lists. The coach does not take on a weight target as the goal of the work.

What a coaching conversation offers that a search engine cannot is room to look at the client’s own reasons and experience. The coach does not evaluate the plan or argue with it. With permission, the coach can share that the available studies have not established an advantage from matching food lists to blood type, and that some of the plans include foods with good evidence of their own, and then ask what the client makes of that. The decision stays with the client, within the referral boundaries above.

What to ask

• “What drew you to eating for your blood type?”
• “What have you noticed since you started, and what do you think is making the difference?”
• “Which parts of the plan feel worth keeping, and which feel like a burden?”
• “How does the plan work with the people you eat with?”
• “Is there anything about your health or your medicines that your doctor should know about before you change how you eat?”

Referral red flags

A client who takes insulin or a sulfonylurea and plans to drop grains and other carbohydrate foods needs the prescriber’s input before the change, because eating too little carbohydrate, or skipping or delaying meals, can drop blood glucose too low (National Institute of Diabetes and Digestive and Kidney Diseases, n.d.-b). If such a client reports shakiness, sweating or confusion, they may be having a low blood sugar episode right now: the first step is the one in their own diabetes plan for treating it, with emergency help if the symptoms are severe or the person cannot manage them, and telling the prescriber is the separate step that follows.

A client taking warfarin who is about to change how much leafy green food they eat, a client with chronic kidney disease who is about to change how much protein they eat, and a client taking prescription medicine who is adding the plan’s supplements should talk with the prescriber, pharmacist or kidney care team first (American Society of Health-System Pharmacists, 2026; National Institute of Diabetes and Digestive and Kidney Diseases, n.d.-a; National Institutes of Health Office of Dietary Supplements, 2023). A client who wants to stop or change a prescribed treatment in favor of the plan, or who has symptoms no clinician has looked at, belongs with the physician.

A client who speaks of avoid-list foods with fear, keeps adding foods to the list, skips meals with others to stay on it, or feels distress after eating a listed food is worth asking about it directly. On its own a behavior tells a coach little, and none of this is a diagnosis. The questions are what the behavior is for, whether it causes distress, whether it is narrowing what the person eats, and whether it is getting in the way of daily life. Where the answers raise a real concern about disordered eating, the coach follows their own referral process, or takes the concern to the client’s clinician under an existing care plan.

Special populations

Pregnancy and breastfeeding

The searches run for this page in October 2026, which screened a limited number of records from one database, did not identify a study of the Blood Type Diet in pregnancy or breastfeeding. A plan that limits dairy foods or meat changes where a person’s nutrients come from. Any substantial change in eating during pregnancy or breastfeeding is planned with the obstetric clinician and a registered dietitian.

Children and adolescents

The same searches did not identify a study of the plan in children or adolescents. The company sells a book titled Eat Right for Your Baby (D’Adamo Personalized Nutrition, n.d.-b), which was not opened for this page. A child’s eating belongs to the family, the pediatric clinician and a registered dietitian; a coach works with the adult in front of them.

Older adults

The searches did not identify a study designed for older adults. The 2 Canadian studies enrolled young adults aged 20 to 29 and overweight adults with an average age of about 45 (Wang et al., 2014, 2018), and the vegan diet trial enrolled adults aged 25 to 75 (Kahleova et al., 2020). An older client with health conditions or several medicines involves their physician before a substantial change.

Athletes and people with high activity

The searches did not identify a study of the plan in athletes. The website assigns types of exercise by blood type (D’Adamo, n.d.-a, n.d.-b); no study found for this page tested that advice. Whether a chosen list meets training needs is a question for a sports dietitian.

People taking medication for diabetes, blood pressure, or blood thinning

A switch to the type O plan can cut carbohydrate foods, which matters for anyone taking insulin, a sulfonylurea or a meglitinide (National Institute of Diabetes and Digestive and Kidney Diseases, n.d.-b). People taking warfarin are advised to eat consistent amounts of vitamin K-containing food from week to week and to talk to their doctor before changing their diet (American Society of Health-System Pharmacists, 2026). The standing rule applies: the client keeps the prescriber informed before a substantial change in eating.

People with a history of disordered eating

The searches did not identify a study of whether the plan raises the risk of disordered eating. The plan is built on lists of foods to avoid, and for a person with this history a new set of food rules can matter more than the foods themselves. A concern goes to the client’s clinician under the existing care plan, or through the coach’s own referral process where there is none.

Client misconceptions

Three misconceptions are worth knowing. The first is that research has shown people do better on the list for their own blood type. The available studies have not established an advantage from matching food lists to blood type. Several reported associations with eating patterns were similar across blood types, but these observational analyses do not establish equivalence or the effects of following the complete book plan (Barnard et al., 2021; Wang et al., 2014, 2018). A systematic review found no study of the health effects of following an ABO blood type diet (Cusack et al., 2013).

The second is that feeling better on the plan proves the blood type theory. People who ate more like the type A plan had lower blood pressure, cholesterol and insulin measures, and the study did not detect a clear difference in these associations by blood type for most measures (Wang et al., 2014), so the foods themselves, or simply paying attention to eating, may explain how someone feels. The third is that type O is the oldest blood type, so type O people should eat like hunters. Genetic studies of the ABO gene describe the A form as the ancestral one (Saitou & Yamamoto, 1997), and in any case the history of a gene does not show how a person responds to food. Each is taken up, with its sources, in section 6.

🤝 COACHING CONNECTION

Marcus tells his coach he started eating for his blood type 2 months ago. He is type O, so he has cut out bread and dairy and eats meat at every meal. He feels more energetic and is thinking of buying the type O supplement pack. The coach asks what he has noticed and what he thinks is making the difference. Marcus says he is eating more vegetables, cooking at home and no longer buying pastries on the way to work. The coach reflects that several changes happened at once and asks which of them feel worth keeping. Marcus mentions that his cholesterol was high at his last checkup and decides to ask his doctor about the meat before he buys anything. Nothing was recommended, and the plan was never evaluated.

4. The evidence

(select the heading to open or close this section)

Three different questions get mixed together when people talk about this diet, and this section keeps them apart. The first is the diet’s own claim: does a person do better on the food list matched to their blood type than on another list, or than a person of a different blood type on the same list? The second is whether the plans, as ways of eating, relate to health whatever a person’s blood type. The third is whether blood type itself is linked to disease. Only the first tests the diet. The certainty ratings for each outcome are in the Evidence rating and review record at the end of the page.

The research on the first question is small: a systematic review that found no qualifying study, 2 observational analyses from the same Canadian research group, and one analysis inside a vegan diet trial. All 4 have now been read in full: the first Toronto analysis from its open full text, and the systematic review, the second Toronto analysis and the vegan trial analysis through a university library during this page’s outside review.

What a fair test would look like

A 2013 systematic review asked: “In humans grouped according to blood type, does adherence to a specific diet improve health and/or decrease risk of disease compared with nonadherence to the diet?” Its authors searched the Cochrane Library, MEDLINE and Embase, screened 1,415 references, and found 16 articles, of which only one met their criteria; that study looked at how people of different MNS blood types, a blood group system other than ABO, responded to a low-fat diet, and did not answer the question directly. The review identified no study showing the health effects of ABO blood type diets, and concluded that “no evidence currently exists to validate the purported health benefits of blood type diets” (Cusack et al., 2013).

The full report was read during this page’s outside review. Using GRADE, its authors judged the evidence from that one indirect MNS study low to very low; that published judgment is about the MNS study and is not carried over to the ABO plans.

The same review described the study that would be needed: health outcomes compared, within one blood type, between people who follow the plan for their type and people who continue a standard diet (Cusack et al., 2013). The searches for this page did not identify a trial of that design. The studies below come at the question from the side, by comparing people of different blood types who ate in the same way.

Evidence from randomized controlled trials

A trial assigns people to a diet by chance and compares them with people assigned to something else, which is the strongest way to test whether the diet itself causes a result. No trial found for this page assigned anyone to a blood type plan. One trial of a different diet has been analyzed by blood type, and it is the closest the trial evidence comes to the diet’s claim.

In that trial, 244 adults with overweight in Washington, DC, were randomly assigned, 122 to a low-fat vegan diet with weekly classes and 122 to make no diet changes, for 16 weeks. Five people in the vegan group and 16 in the control group dropped out; the report first gave 222 completers, and a 2021 correction changed the number to 223 (Kahleova et al., 2020; “Error in the Results,” 2021).

The trial was funded by the Physicians Committee for Responsible Medicine, which the report describes as a nonprofit organization that provides nutrition education and research, with grants from the Yale Diabetes Center for 2 of the authors. Several authors worked for the Physicians Committee, and a 2021 correction to the funding statement says its employees took part in designing, running, analyzing and reporting the study (Kahleova et al., 2020; “Clarification of Role,” 2021). The Vegan Diet page covers the effects of the eating pattern itself; this page uses the trial only to examine the blood-type comparison.

A later analysis compared blood types within the vegan group: people with type A, the type the plan tells to eat a mainly vegetarian diet, with everyone else, and people with type O, the type told to eat meat, with everyone else (Barnard et al., 2021). Blood typing began after a protocol amendment in the second of four study replications. The analysis included 68 vegan-arm completers with blood typing: 20 with type A and 48 with other types. Blood type was not randomized. The timing of the subset and baseline differences, particularly in lipids, limit the unadjusted comparison. The full report, read during this page’s outside review, gives an interval for each group’s own change but not for the difference between blood types. No statistically clear between-blood-type difference was detected.

Cardiometabolic outcomes

In the same analysis, total cholesterol fell by 17.2 mg/dL in people with type A (95 percent interval for that group’s own change, a fall of 0.6 to 33.8 mg/dL) and by 18.3 mg/dL in everyone else (a fall of 11.3 to 25.2 mg/dL), and by 17.4 mg/dL in people with type O and 18.4 mg/dL in everyone else. These intervals are for each group’s change, not for the difference between blood types. Fat mass, visceral fat, blood lipids, fasting glucose and HbA1c were also compared, and no statistically clear between-blood-type difference was detected (Barnard et al., 2021). The type A group and the others started with markedly different lipid levels, which limits an unadjusted comparison of changes. Blood type was not assigned by chance, and the 68 people analyzed were the completers blood typed after the protocol amendment, so this is an observational comparison carried out inside a trial.

Cancer

The searches for this page did not identify a trial of the Blood Type Diet, or a trial analyzed by blood type, that measured cancer. The studies reviewed do not establish any effect of the plan on cancer risk or outcomes.

Body weight, as measured

This archive reports body weight as something a study measured, in that study’s own numbers, and not as a goal or as a measure of whether a diet worked; see the standing rule in the review record at the foot of this page. The one exception is weight a person loses without meaning to, which can signal a health problem and is a reason to involve a clinician, never a sign that a diet worked.

In the vegan trial analysis, measured body weight fell by an average of 5.7 kilograms in people with type A (95 percent interval for that group’s own change, 4.8 to 6.7 kilograms) and 7.0 kilograms in everyone else (5.8 to 8.2 kilograms), P = 0.09, and by 7.1 kilograms in people with type O and 6.2 kilograms in everyone else (P = 0.33) (Barnard et al., 2021). The blood-type comparison was observational. The reported P value does not establish equivalence, and the numerical changes in measured weight are not treated here as a measure of whether a diet succeeded. The report gives no interval for the difference between blood types. This page’s approximation from the 2 group intervals, not a published result, gives a measured decrease 1.3 kilograms smaller in the type A group, with an approximate interval from 0.2 kilograms larger to 2.8 kilograms smaller, which includes no difference.

Mortality

The searches for this page did not identify a trial of the plan that recorded deaths. The studies found were too small and too short to estimate any effect on survival.

Evidence from long-term prospective cohort studies

A cohort study follows people over time and records what happens to them. The searches for this page did not identify a cohort study that followed people who eat by blood type and compared their health with people who do not. Large cohort studies of blood type itself exist; they are described below under blood type as a risk factor, because they answer a different question.

Evidence from cross-sectional and other observational studies

The main test of the diet’s idea comes from the Toronto Nutrigenomics and Health Study, a single-time-point study of young adults aged 20 to 29. Of 1,639 participants, 1,455 remained after people with implausible food records or missing data were excluded. Each completed a food questionnaire covering the previous month, and the researchers gave every participant a score for each of the 4 plans, a point for each serving of a recommended food and a point off for each serving of a food to avoid. Blood type was found by testing 2 points in the ABO gene, and the analysis was adjusted for age, sex, ethnocultural group and calories (Wang et al., 2014). This page read the full text.

Whatever their blood type, people whose eating most resembled the type A plan had lower body mass index, waist size, blood pressure, total cholesterol, triglycerides and insulin measures than those whose eating resembled it least. For example, systolic blood pressure averaged 115.4 against 117.6 mm Hg in the highest and lowest thirds of the type A score. Eating more like the type AB plan went with several of the same differences, and eating more like the type O plan with lower triglycerides; the type B score showed no clear association (Wang et al., 2014).

The diet’s prediction is that these associations should be stronger in people eating the plan for their own type. The researchers tested whether blood type changed each association and did not detect a statistically clear difference for most measures. The few exceptions, for glucose and insulin measures with the type A score and glucose with the type AB score, did not show a pattern of greater benefit in the matched group, and the authors judged they may have been due to chance (Wang et al., 2014). The authors also noted that their score measured relative resemblance to each plan rather than strict adherence (Wang et al., 2014).

Two worked examples from the published tables show the size of the numbers involved. They are this page’s own approximations from the tables’ adjusted group averages and standard errors, treating the 4 averages in each example as independent; they are not results the authors reported, and averages adjusted in one statistical model can be correlated, which would make the true intervals narrower or wider.

Moving from the lowest to the highest third of the type A score, systolic blood pressure was 1.3 mm Hg lower among people with type A and 2.7 mm Hg lower among everyone else. The matched group’s difference was 1.4 mm Hg smaller, with an approximate interval from 1.2 mm Hg larger to 4.0 mm Hg smaller. Body mass index was 0.1 lower among people with type A and 0.9 lower among everyone else, a difference of 0.8 in the same direction, with an approximate interval from 0.2 larger to 1.8 smaller (Wang et al., 2014). Comparative imprecision is unresolved; the intervals shown are conditional approximations from adjusted cell means, not reported interaction intervals. In both examples the averages pointed away from a larger difference in the matched group, but the approximations do not establish that any extra difference is small.

The diet’s website rejects this study, saying its participants were eating foods such as potato chips, pizza and French fries while following only 13.7 percent of the diet (D’Adamo, n.d.-f). The concern behind that criticism is a fair one in part: a score of how much usual eating resembles a list is not the same as following the plan. The study was designed to measure that resemblance, and within that design it did not detect a clear difference by blood type in most links between eating and measurements.

A second analysis by the same group used data from the Toronto Healthy Diet Study, a 6-month dietary intervention study in adults with overweight, average age about 45. There were 973 participants in the baseline analyses and 576 in the 6-month analyses after exclusions. The study’s own assigned diets were not the blood type plans; the researchers scored each participant’s food questionnaire against all 4 plans before and after. At the start, higher scores for some plans went with lower diastolic blood pressure or waist size. Over 6 months, people whose eating moved toward the type A, type B or type O plans had larger decreases in measured body mass index or waist size (Wang et al., 2018).

The authors tested whether blood type changed these associations and, because many tests were run, set P < 0.005 as their threshold. No interaction met that threshold. Two type A interactions, for body mass index and for diastolic blood pressure, had nominal P values of 0.01, which do not pass it. The main report gives no estimate and interval for the interaction, so it does not show how large a difference by blood type the data rule out, and the absence of an interaction meeting the threshold does not establish that matching makes no difference (Wang et al., 2018). Blood-type matching and adherence to the plans were not assigned. The full report was read during this page’s outside review.

Blood type as a risk factor, a separate question

Blood type itself is linked to some diseases, and the diet’s promoters cite such links. In 2 United States cohorts of 62,073 women and 27,428 men followed for more than 20 years, people with blood types A, B or AB had a somewhat higher risk of coronary heart disease than people with type O after adjustment for other risk factors: hazard ratios of 1.06 (0.99 to 1.15), 1.15 (1.04 to 1.26) and 1.23 (1.11 to 1.36). The authors estimated that 6.27 percent of the heart disease cases were attributable to having a blood type other than O (He et al., 2012). This page read the authors’ accepted manuscript.

A genetic study using UK Biobank summary data linked the gene variant that marks type O with lower rates of circulatory diseases, especially deep vein thrombosis, and with lower total and LDL cholesterol (Li & Schooling, 2020). These studies are about blood type, not about diet. None found for this page tested whether eating by blood type changes these risks, and a higher risk in one blood type does not show that a particular food list lowers it.

Highlighted studies

These 4 carry the most weight for this page: the 2 Toronto analyses, the vegan trial analysis and the systematic review. Each is described so that a reader can judge it without opening the paper.

On a narrow screen, scroll the table sideways to see every column.

Highlighted studies for the Blood Type Diet

Study	Design	Sample size	Duration	Finding	The limitation that matters
Wang et al., 2014, Toronto Nutrigenomics and Health Study	Cross-sectional study	1,455 young adults analyzed, of 1,639	A single time point; food questionnaire covering the previous month	Eating more like the type A, AB or O plans went with some better risk factors in every blood type; matching the plan to blood type did not change the associations for most measures	Usual eating scored against the lists, not people assigned to follow a plan; many comparisons tested
Wang et al., 2018, Toronto Healthy Diet Study	Secondary analysis of a dietary intervention study, observational for the blood type plans	973 adults with overweight at baseline; 576 at 6 months	6 months	Moving toward the type A, B or O plans went with larger decreases in measured body mass index or waist size; no interaction with blood type met the study’s threshold of P < 0.005	Adherence to the plans was not assigned; 576 of 973 in the 6-month analyses; no interaction estimate with an interval
Barnard et al., 2021, within a randomized trial of a low-fat vegan diet	Comparison of blood types within the vegan group of a randomized trial	68 blood-typed vegan-arm completers (20 type A, 48 other types) of 122 assigned to the vegan diet	16 weeks	No statistically clear between-blood-type difference was detected; measured weight fell by 5.7 kilograms with type A and 7.0 kilograms with other types	Blood type was not assigned by chance; blood typing began after a protocol amendment; baseline lipids differed; intervals are given for each group’s change, not for the difference
Cusack et al., 2013	Systematic review	1,415 references screened, 16 articles, 1 eligible	Searches of 3 databases	No study of the health effects of ABO blood type diets was identified	Its one eligible study concerned MNS, not ABO, blood types; does not cover studies published after 2013

Head to head

The searches for this page did not identify a trial that compared a blood type plan with another diet. The nearest comparisons are indirect. In the vegan trial analysis, people with type A, whose plan is mainly vegetarian, and people with type O, whose plan centers on meat, were compared on the same low-fat vegan diet, and no statistically clear between-blood-type difference was detected (Barnard et al., 2021). In the Toronto study, eating more like the type A plan went with better measurements than eating more like the type O plan, in people of every blood type (Wang et al., 2014). Neither is a test of following one plan against another, and neither establishes that the plans are equivalent.

Where the evidence is absent

The searches for this page did not identify a trial assigning people of one blood type to their type’s plan or to a standard diet, which is the design a systematic review described as needed (Cusack et al., 2013). They did not identify a study of the plan that measured heart attacks, strokes, cancer or deaths, or one that followed people for longer than 6 months. The studies that compared blood types were observational, came from 2 research groups, and enrolled mostly adults in Canada and the United States.

The searches did not identify a study of the plan in pregnancy, children, older adults or athletes (see Special populations in section 3), a study of eating behavior, quality of life or disordered eating among followers, a study of secretor-based versions of the plan, or a trial of the supplements sold for each blood type. Not finding a study is a statement about these searches, not proof that no study exists.

5. Mechanisms

(optional reading; select the heading to open or close this section)

This section explains what blood type is, what lectins are, and how the diet is thought to work. You can skip it. Nothing in the coaching sections depends on it.

Established mechanisms

Blood type describes substances, called antigens, on the surface of red blood cells. There are 4 major ABO groups, set by the presence or absence of the A and B antigens, and the Rh factor adds a plus or minus to make 8 common types. Blood type is inherited from both parents, and matching it matters for safe transfusion (American Red Cross, n.d.). The same blood group substances are found on other tissues as well, and blood group is linked to differences in clotting factors, cholesterol and the risk of some cardiovascular diseases (He et al., 2012; Li & Schooling, 2020).

Lectins are proteins that bind to carbohydrates. They are found in all plants and are highest in raw legumes and whole grains. Raw or undercooked kidney beans contain a lectin that can make red blood cells clump together and can cause nausea, vomiting and diarrhea; soaking and boiling inactivate most lectins, which is why cooked and canned beans are low in them (Harvard T.H. Chan School of Public Health, n.d.). That effect does not depend on blood type. The measured differences in the Toronto study are explained more simply by the foods themselves: the type A plan resembles eating patterns linked with lower cardiovascular risk, and the type O plan resembles low-carbohydrate eating, which has been linked with lower triglycerides (Wang et al., 2014).

Proposed mechanisms

The diet’s central proposal is that food lectins react with blood type substances, so that a food that suits one type causes clumping, inflammation and weight gain in another (D’Adamo, n.d.-g). The Toronto researchers wrote that the studies available to them suggested no blood-type-specific clumping by food lectins (Wang et al., 2014). The website also proposes that blood types differ in stomach acid and digestive enzymes, for example that type O secretes more of an intestinal enzyme that helps it handle protein and fat (D’Adamo, n.d.-b); the searches for this page did not identify a study testing whether such differences change how people respond to a diet.

The evolutionary story, that type O came first among hunters and type A arose with farming, has also been questioned. Comparisons of human and other primate versions of the ABO gene describe the A form as the ancestral one from which B arose several times (Saitou & Yamamoto, 1997). The diet’s website points to a later analysis that found the working A form may have been lost and then re-created in human ancestors by genetic recombination less than 300,000 years ago (D’Adamo, n.d.-f; Kitano et al., 2012). Either way, the history of a gene is not a test of how a person responds to food. A proposed mechanism is a reason to study an idea, not evidence that it works.

6. Risks, contraindications and myths

(select the heading to open or close this section)

Warnings from professional or regulatory bodies

The German Nutrition Society, a scientific society for nutrition, issued a press release in 2013 whose title means “blood group diets without proven benefit.” It stated that there is no scientific evidence for the health benefits claimed for blood group diets, said that the promoted health effects are pure theory and not scientifically proven, and noted that it had issued an earlier statement on the diet in 2000. It described the claim that lectins incompatible with a person’s blood type clump red blood cells and cause disease as a hypothesis for which the promised evidence had not appeared in the scientific literature (Deutsche Gesellschaft für Ernährung, 2013). The press release is in German, and this page reports it in paraphrase rather than translating it as a quotation.

Australian Red Cross Lifeblood, a blood donation service that writes that it “might not be experts in nutrition,” published a short myth-busting article on the diet. It states that “research has not shown that following a diet based on your blood type leads to better health outcomes,” that the diet “could be too restrictive,” and that most lectins react with all blood types and are usually destroyed by cooking (Australian Red Cross Lifeblood, n.d.).

The searches for this page, last run November 8, 2026, did not identify a warning about the diet from a United States regulatory agency, or a warning letter to the company. The company’s own pages state that their statements have not been evaluated by the U.S. Food and Drug Administration and that its products are not intended to diagnose, treat, cure or prevent any disease (D’Adamo Personalized Nutrition, n.d.-f).

Nutrient gaps

What a person risks missing depends on which list they follow. The type O plan limits milk and milk products (Deutsche Gesellschaft für Ernährung, 2013; D’Adamo, n.d.-b), and milk, yogurt and cheese are the main food sources of calcium for most people in the United States. People who do not eat dairy foods need other sources, such as canned fish with bones, certain vegetables such as kale and broccoli, or foods with added calcium (National Institutes of Health Office of Dietary Supplements, 2023).

The type A plan is mainly vegetarian and also limits dairy foods (Deutsche Gesellschaft für Ernährung, 2013; D’Adamo, n.d.-c), which raises the nutrient questions described on the Vegetarian Diet and Vegan Diet pages, such as vitamin B12 and calcium. The type O plan also limits wheat and other grains (D’Adamo, n.d.-b), which are sources of fiber, and the American Heart Association’s 2026 guidance favors whole grains over refined grains (Lichtenstein et al., 2026). Whether a particular list meets a particular person’s needs is a question for a registered dietitian.

Who should not follow it without medical supervision

People taking insulin, a sulfonylurea or a meglitinide for diabetes, because cutting carbohydrate foods can lower blood glucose (National Institute of Diabetes and Digestive and Kidney Diseases, n.d.-b). People taking warfarin, who need a consistent intake of vitamin K (American Society of Health-System Pharmacists, 2026). People with chronic kidney disease, whose protein, potassium and phosphorus needs change as the disease advances and depend on their lab results and whether they receive dialysis, so the plan is set with their care team and dietitian (National Institute of Diabetes and Digestive and Kidney Diseases, n.d.-a).

People who are pregnant or breastfeeding, and children and teenagers, for whom the searches for this page found no study of the plan. People with a current or past eating disorder, for whom a new list of forbidden foods can matter more than the foods themselves. And anyone hoping to treat a diagnosed condition with the plan or its supplements, which the company itself says are not intended to diagnose, treat, cure or prevent any disease (D’Adamo Personalized Nutrition, n.d.-f).

Medication interactions

A coach never adjusts medication. The point of knowing these is to know when to send a client back to the prescriber or pharmacist. Insulin, sulfonylureas and meglitinides can cause low blood glucose, and not eating enough carbohydrate, or skipping or delaying meals, makes it more likely (National Institute of Diabetes and Digestive and Kidney Diseases, n.d.-b). People taking warfarin are advised to eat consistent amounts of vitamin K-containing food from week to week, not to eat large amounts of leafy green vegetables, and to talk to their doctor before changing their diet (American Society of Health-System Pharmacists, 2026).

The plan is sold with supplements for each blood type. Supplements can interact with medicines; the federal fact sheet on calcium, for example, notes that calcium supplements can interact with certain medicines (National Institutes of Health Office of Dietary Supplements, 2023). A client taking prescription medicine who wants to add any of the plan’s supplements asks the prescriber or pharmacist first.

Signs it is going wrong

Physical signs a coach or a family member might notice include shakiness, sweating, dizziness or confusion in someone taking diabetes medicine (National Institute of Diabetes and Digestive and Kidney Diseases, n.d.-b); easy bruising or bleeding that will not stop in someone taking warfarin (American Society of Health-System Pharmacists, 2026); new swelling in someone with kidney disease (National Institute of Diabetes and Digestive and Kidney Diseases, n.d.-a); and weight loss the person did not intend. Bleeding that will not stop, or severe or rapidly worsening symptoms, needs emergency help. Otherwise, each is a reason to involve a clinician, not a diagnosis; the coach notices and encourages the client to tell their clinician, and working out the cause is the clinician’s job.

A possible low blood sugar episode is treated first under the client’s own plan, with emergency help if symptoms are severe or the person cannot manage them, and the prescriber is told afterward.

Eating behavior matters as much, and has to be read with care. Following a food list can have ordinary reasons, such as curiosity, cooking for a household or a family member’s enthusiasm, so on its own it tells a coach little. A growing list of foods to avoid, fear or guilt after eating a listed food, or turning down meals with others to stay on the list carry more meaning and are worth asking about directly. The questions are what the behavior is for, whether it causes distress, whether it is narrowing what the person eats, and whether it is getting in the way of daily life. None of this is a diagnosis. Where the answers raise a real concern, the coach follows their own referral process, or takes it to the client’s clinician under an existing care plan.

Myths and where they came from

Myth: research shows I do better on the foods for my own blood type

Where it came from: the books and website, which present the plans as matched to each type’s biology and cite visitors’ reports of improvement (D’Adamo, n.d.-g). What is true: the available studies have not established an advantage from matching food lists to blood type. Several reported associations with eating patterns were similar across blood types, but these observational analyses do not establish equivalence or the effects of following the complete book plan (Barnard et al., 2021; Wang et al., 2014, 2018). A systematic review found no study showing health effects of ABO blood type diets (Cusack et al., 2013).

Myth: I feel better on the plan, so the blood type theory must be right

Where it came from: personal experience and the website’s collected success stories (D’Adamo, n.d.-g). What is true: in the Toronto study, eating more like the type A plan went with lower blood pressure, cholesterol and insulin measures, with no clear difference by blood type for most measures (Wang et al., 2014). A person may feel better because of what they now eat, because they are cooking or paying more attention to food, or for other reasons; feeling better does not show that blood type was the cause.

Myth: type O is the original human blood type, so type O people should eat like hunters

Where it came from: the book’s account of human history, as summarized by researchers (Wang et al., 2014). What is true: genetic comparisons describe the A form of the ABO gene as the ancestral one (Saitou & Yamamoto, 1997), with later debate about its history in humans (Kitano et al., 2012), and no study found for this page showed that the history of a blood type predicts how a person responds to food.

Myth: the wrong foods make my blood clump

Where it came from: the lectin theory at the center of the books (D’Adamo, n.d.-g; Wang et al., 2014). What is true: some raw lectins, such as the one in uncooked kidney beans, can make red blood cells clump and cause illness, but cooking inactivates most lectins (Harvard T.H. Chan School of Public Health, n.d.), the Toronto researchers wrote that available studies suggested no blood-type-specific clumping (Wang et al., 2014), and the German Nutrition Society stated that the promised evidence for the theory had not appeared (Deutsche Gesellschaft für Ernährung, 2013).

Myth: my blood type puts me at risk, so eating for my type will protect me

Where it came from: real research linking blood type to some diseases, such as a slightly higher risk of coronary heart disease in people who are not type O (He et al., 2012). What is true: those studies are about blood type, not diet. No study found for this page tested whether eating by blood type changes those risks.

7. Key terms

ABO blood group

The system that sorts blood into types A, B, AB and O by whether the A and B antigens are present on red blood cells.

Antigen

A substance on a cell’s surface that the immune system can recognize; blood type antigens are what make a transfusion compatible or not.

Rh factor

A protein on red blood cells that is either present, written as positive, or absent, written as negative.

Secretor status

Whether a person’s blood type substances also appear in saliva, the gut lining and other body fluids, which a separate gene controls.

Lectin

A protein that binds to carbohydrates, found in many plants and highest in raw legumes and whole grains; cooking inactivates most lectins.

Agglutination

Clumping of cells, such as red blood cells, together.

Genotype

A person’s genetic makeup at a particular gene; the Toronto studies found blood type by testing the ABO gene directly.

Effect modification

When the link between an exposure, such as a way of eating, and an outcome differs from one group to another, such as between blood types. The diet’s claim is a claim of effect modification.

Food frequency questionnaire

A form on which a person reports how often they ate each listed food over a period, used to estimate usual eating.

Diet score

A number that summarizes how closely a person’s usual eating resembles a defined plan; it measures resemblance, not whether the person was following the plan.

Cross-sectional study

A study that measures exposure and outcome at the same point in time, so it can show an association but not which came first.

Secondary analysis

An analysis of data collected for a different main purpose, such as comparing blood types within one group of a trial designed to test something else.

Risk factor

A trait or exposure linked to a higher chance of a disease; it does not by itself show that changing something else, such as diet, will change the risk.

Systematic review

A review that searches for and appraises all studies answering a defined question, using a method set out in advance.

8. Review questions

1. Describe in your own words what the type O and type A plans ask a person to eat and avoid, and name 2 products sold alongside the diet.
2. This question draws on section 4. The Toronto study found that people whose eating looked more like the type A plan had lower blood pressure. Explain why that finding does not support the Blood Type Diet’s central claim, and what result would have supported it.
3. A client tells you he is type A, takes insulin, and plans to switch to the type A plan and buy the matching supplements. Write what you would say and ask, say who needs to be involved and when, and say what the client should do if he feels shaky and sweaty after the change.
4. Where did the idea that type O is the oldest blood type come from, and what do genetic studies of the ABO gene report? Write the 2 sentences you would use, with the client’s permission, if a client raised it.

9. References

American Red Cross. (n.d.). Blood types explained: A, B, AB and O. Red Cross Blood Services. redcrossblood.org/donate-blood/blood-types.html

American Society of Health-System Pharmacists. (2026, March 15). Warfarin. MedlinePlus, National Library of Medicine. medlineplus.gov/druginfo/meds/a682277.html

Australian Red Cross Lifeblood. (n.d.). Myth busting: A diet for your blood type? Retrieved November 8, 2026, from lifeblood.com.au/news-and-stories/vital-reads/myth-busting-diet-your-blood-type

Barnard, N. D., Rembert, E., Freeman, A., Bradshaw, M., Holubkov, R., & Kahleova, H. (2021). Blood type is not associated with changes in cardiometabolic outcomes in response to a plant-based dietary intervention. Journal of the Academy of Nutrition and Dietetics, 121(6), 1080-1086. https://doi.org/10.1016/j.jand.2020.08.079

Clarification of role of funder statement. (2021). JAMA Network Open, 4(1), Article e2035088. https://doi.org/10.1001/jamanetworkopen.2020.35088

Cusack, L., De Buck, E., Compernolle, V., & Vandekerckhove, P. (2013). Blood type diets lack supporting evidence: A systematic review. The American Journal of Clinical Nutrition, 98(1), 99-104. https://doi.org/10.3945/ajcn.113.058693

D’Adamo, P. (n.d.-a). Blood type A. dadamo.com. dadamo.com/dadamo2024/show_blog.pl?blood-type-a

D’Adamo, P. (n.d.-b). Blood type O. dadamo.com. dadamo.com/dadamo2024/show_blog.pl?blood-type-o

D’Adamo, P. (n.d.-c). Books and learning essentials. dadamo.com. dadamo.com/dadamo2024/show_blog.pl?books-and-resources

D’Adamo, P. (n.d.-d). Dr. Peter D’Adamo. dadamo.com. dadamo.com/dadamo2024/show_blog.pl?dr-peter-dadamo

D’Adamo, P. (n.d.-e). The importance of knowing your secretor status. dadamo.com. dadamo.com/dadamo2024/show_blog.pl?importance-of-secretor-status

D’Adamo, P. (n.d.-f). Response to a few critics. dadamo.com. dadamo.com/dadamo2024/show_blog.pl?response-to-critics

D’Adamo, P. (n.d.-g). Welcome to the diet that knows you best. dadamo.com. dadamo.com/dadamo2024/show_blog.pl?welcome-to-the-blood-type-diet

D’Adamo Personalized Nutrition. (n.d.-a). Deflect lectin blocker supplement: Lectin shield for blood type O. 4YourType. Retrieved November 8, 2026, from 4yourtype.com/deflect-o-lectin-blocker

D’Adamo Personalized Nutrition. (n.d.-b). Eat right for your blood type: The bestselling Blood Type Diet book. 4YourType. Retrieved November 8, 2026, from 4yourtype.com/eat-right-for-your-type-book

D’Adamo Personalized Nutrition. (n.d.-c). Eat right for your blood type with the official Blood Type Diet app. 4YourType. Retrieved November 8, 2026, from 4yourtype.com/the-blood-type-diet-app

D’Adamo Personalized Nutrition. (n.d.-d). Original home blood-typing kit. 4YourType. Retrieved November 8, 2026, from 4yourtype.com/original-home-blood-typing-kit

D’Adamo Personalized Nutrition. (n.d.-e). SWAMI software. 4YourType. Retrieved November 8, 2026, from 4yourtype.com/swami-software

D’Adamo Personalized Nutrition. (n.d.-f). Tests and kits for the Blood Type Diet. 4YourType. Retrieved November 8, 2026, from 4yourtype.com/tests-kits

Deutsche Gesellschaft für Ernährung. (2013, December 10). Blutgruppendiäten ohne bewiesenen Nutzen [Press release]. dge.de/fileadmin/dok/presse/meldungen/2011-2018/DGE-Pressemeldung-AdW-02-2013-Blutgruppendiaet.pdf

Error in the results and eTable 2 in Supplement 2. (2021). JAMA Network Open, 4(5), Article e2115510. https://doi.org/10.1001/jamanetworkopen.2021.15510

Harvard T.H. Chan School of Public Health. (n.d.). Lectins. The Nutrition Source. nutritionsource.hsph.harvard.edu/anti-nutrients/lectins

He, M., Wolpin, B., Rexrode, K., Manson, J. E., Rimm, E., Hu, F. B., & Qi, L. (2012). ABO blood group and risk of coronary heart disease in two prospective cohort studies. Arteriosclerosis, Thrombosis, and Vascular Biology, 32(9), 2314-2320. https://doi.org/10.1161/ATVBAHA.112.248757

Kahleova, H., Petersen, K. F., Shulman, G. I., Alwarith, J., Rembert, E., Tura, A., Hill, M., Holubkov, R., & Barnard, N. D. (2020). Effect of a low-fat vegan diet on body weight, insulin sensitivity, postprandial metabolism, and intramyocellular and hepatocellular lipid levels in overweight adults: A randomized clinical trial. JAMA Network Open, 3(11), Article e2025454. https://doi.org/10.1001/jamanetworkopen.2020.25454

Kahve, A. N., & Bayrak, E. (2026). Assessment of the relationship between awareness of popular diets and body image among young women: A cross-sectional study. Journal of Health, Population and Nutrition, 45(1), Article 211. https://doi.org/10.1186/s41043-026-01388-6

Kitano, T., Blancher, A., & Saitou, N. (2012). The functional A allele was resurrected via recombination in the human ABO blood group gene. Molecular Biology and Evolution, 29(7), 1791-1796. https://doi.org/10.1093/molbev/mss021

Li, S., & Schooling, C. M. (2020). A phenome-wide association study of ABO blood groups. BMC Medicine, 18(1), Article 334. https://doi.org/10.1186/s12916-020-01795-4

Lichtenstein, A. H., Khera, A., Anderson, C. A. M., Appel, L. J., DeSilva, D. M., Gardner, C., Hu, F. B., Jones, D. W., Petersen, K. S., & American Heart Association. (2026). 2026 dietary guidance to improve cardiovascular health: A scientific statement from the American Heart Association. Circulation, 153(18), e1285-e1295. https://doi.org/10.1161/CIR.0000000000001435

Nacis, J. S., Galang, M. R., Labrador, J. P. H., Gonzales, M. S., Dablo, A. M. F. D., Domalanta-Ronquillo, D. G. A., Alfonso, V. F. J., Glorioso, I. G., & Rodriguez, M. P. (2022). “Right diet for the right person”: A focus group study of nutritionist-dietitians’ perspectives on nutritional genomics and gene-based nutrition advice. Journal of Community Genetics, 13(1), 49-57. https://doi.org/10.1007/s12687-021-00560-1

National Institute of Diabetes and Digestive and Kidney Diseases. (n.d.-a). Healthy eating for adults with chronic kidney disease. National Institutes of Health. niddk.nih.gov/health-information/kidney-disease/chronic-kidney-disease-ckd/healthy-eating-adults-chronic-kidney-disease

National Institute of Diabetes and Digestive and Kidney Diseases. (n.d.-b). Low blood glucose (hypoglycemia). National Institutes of Health. niddk.nih.gov/health-information/diabetes/overview/preventing-problems/low-blood-glucose-hypoglycemia

National Institutes of Health Office of Dietary Supplements. (2023, September 14). Calcium: Fact sheet for consumers. ods.od.nih.gov/factsheets/Calcium-Consumer

Open Library. (n.d.-a). Diabetes (Dr. Peter J. D’Adamo’s Eat Right 4 Your Type health library), by Peter J. D’Adamo and Catherine Whitney, Putnam edition dated 2001. Internet Archive. openlibrary.org/books/OL9349903M

Open Library. (n.d.-b). Eat right 4 (for) your type: The individualized diet solution to staying healthy, living longer & achieving your ideal weight, by Peter D’Adamo, G. P. Putnam’s Sons edition dated 1996. Internet Archive. openlibrary.org/books/OL996837M

Open Library. (n.d.-c). Eat right for your type: The individualized Blood Type Diet solution, by Peter D’Adamo, revised and updated edition dated 2016. Internet Archive. openlibrary.org/books/OL27227550M

Saitou, N., & Yamamoto, F. (1997). Evolution of primate ABO blood group genes and their homologous genes. Molecular Biology and Evolution, 14(4), 399-411. https://doi.org/10.1093/oxfordjournals.molbev.a025776

Wang, J., García-Bailo, B., Nielsen, D. E., & El-Sohemy, A. (2014). ABO genotype, “blood-type” diet and cardiometabolic risk factors. PLOS ONE, 9(1), Article e84749. https://doi.org/10.1371/journal.pone.0084749

Wang, J., Jamnik, J., García-Bailo, B., Nielsen, D. E., Jenkins, D. J. A., & El-Sohemy, A. (2018). ABO genotype does not modify the association between the “blood-type” diet and biomarkers of cardiometabolic disease in overweight adults. The Journal of Nutrition, 148(4), 518-525. https://doi.org/10.1093/jn/nxx074

Evidence rating and review record

Each row identifies the method, source, outcome or recommendation, and comparison being rated. Ratings from different tools are not interchangeable. The source is either a named published review or this page’s own judgment. The rating in the banner is for this diet’s main claimed outcome or recommendation, and the banner names the tool that produced it. A high rating means the research is solid; it does not mean the diet is right for everyone, and it does not mean a coach should recommend it.

Rows rated with GRADE use the same 4-level scale as the clinical guidelines you will meet in practice. High means further research is very unlikely to change the conclusion. Moderate means further research may change it. Low means further research is likely to change it. Very low means the conclusion is uncertain. GRADE rates each outcome separately, so a diet can be high for one outcome and very low for another. A row marked Not rated is an abstention, not a certainty level.

Three things about this table matter before reading it. First, every rating is this page’s own judgment, assigned from the evidence with every GRADE domain given in its reasons, because the searches found no published certainty rating of a blood type diet outcome; the 2013 systematic review used GRADE to appraise studies but found none that answered its question (Cusack et al., 2013). Second, the rows that test the diet’s claim ask whether blood type changes a link between eating and a measurement, which is the claim the diet makes.

Third, the intervals in the first 3 rows and the last rated row are this page’s conditional approximations from published adjusted group averages, treating those averages as independent; they are not results the authors reported, and averages adjusted in one statistical model can be correlated. The intervals in the 2 vegan trial rows are this page’s approximations from the intervals the report gives for each blood type group’s own change; those groups are different people, but rounding in the published figures limits accuracy. The thresholds of 2 mm Hg in blood pressure, 0.5 in body mass index, 0.1 mmol/L in triglycerides, 1 kilogram in body weight and 5 mg/dL in total cholesterol are this page’s analytical conventions for a material measured difference, not published minimal important differences and not client goals.

On a narrow screen, scroll the table sideways to see every column.

Evidence ratings, their sources and reasons

Outcome	Certainty	Who rated it, and for which comparison	Why it is rated this way
Systolic blood pressure: whether the difference between the highest and lowest thirds of the type A plan score is larger in people with blood type A than in people of other blood types; young adults aged 20 to 29; a single time point (the banner rating)	Very low	This page’s own judgment, using GRADE. Wang et al., 2014: cross-sectional study, 544 people with type A and 911 with other blood types	Observational evidence starts at low. Risk of bias is serious, lowered 1 level: usual eating was scored from a food questionnaire against the plan’s lists, and random error in such a score could obscure or distort an interaction; and 184 of 1,639 participants were excluded for implausible records or missing data, which calls for attention to missing data, though exclusion alone does not show the direction of any bias. Inconsistency could not be assessed for a single study and is not a deduction. Indirectness is not serious for this question, which concerns usual eating that resembles the plan; it is not a test of following the book’s full plan. Comparative imprecision is unresolved; the interval shown is a conditional approximation from adjusted cell means, not a reported interaction interval. Treating the 4 adjusted averages as independent, this page’s approximation gives a fall 1.4 mm Hg smaller in the matched group, with an approximate interval from 1.2 mm Hg larger to 4.0 mm Hg smaller. Whether a matched-group fall 2 mm Hg or more larger can be excluded depends on how the adjusted averages are correlated, which the report does not give (sensitivity check in the claim ledger). Publication bias could not be assessed. No upgrade. Low minus 1 for serious risk of bias is Very low, the lowest level, so the unresolved imprecision cannot lower it further. The authors’ test of this interaction gave P = 0.48.
Body mass index: whether the difference between the highest and lowest thirds of the type A plan score is larger in people with blood type A than in people of other blood types; young adults; a single time point	Very low	This page’s own judgment, using GRADE. Wang et al., 2014: 544 people with type A and 911 with other blood types	Observational evidence starts at low. Risk of bias is serious, lowered 1 level, for the same reasons as the row above. Inconsistency could not be assessed for a single study. Indirectness is not serious for this question. Comparative imprecision is unresolved; the interval shown is a conditional approximation from adjusted cell means, not a reported interaction interval. Treating the 4 adjusted averages as independent, this page’s approximation gives a measured difference in body mass index 0.8 smaller in the matched group, with an approximate interval from 0.2 larger to 1.8 smaller. Whether a material measured difference of 0.5 in the matched group’s direction can be excluded depends on the correlation between the adjusted averages and on rounding in the published figures (sensitivity check in the claim ledger). The 0.5 threshold is an analytical convention for a material measured difference, not a goal or a benefit. Publication bias could not be assessed. No upgrade. Low minus 1 for serious risk of bias is Very low, the lowest level, so the unresolved imprecision cannot lower it further. The authors’ interaction test gave P = 0.15.
Triglycerides: whether the difference between the highest and lowest thirds of the type O plan score is larger in people with blood type O than in people of other blood types; young adults; a single time point	Very low	This page’s own judgment, using GRADE. Wang et al., 2014: 543 people with type O and 912 with other blood types	Observational evidence starts at low. Risk of bias is serious, lowered 1 level, for the reasons in the first row. Inconsistency could not be assessed for a single study. Indirectness is not serious. Imprecision is also serious: treating the 4 adjusted averages as independent, this page’s conditional approximation gives a fall 0.04 mmol/L larger in the matched group, with an approximate interval from 0.18 mmol/L larger to 0.10 mmol/L smaller, which includes both no difference and a matched-group fall more than 0.1 mmol/L larger. It is not a reported interaction interval, and it would exclude those values only if the adjusted averages were strongly positively correlated, so it cannot support equivalence. Publication bias could not be assessed. Low minus 1 for serious risk of bias reaches Very low, the lowest level, so the imprecision cannot lower it further. The authors’ interaction test gave P = 0.65.
Body mass index over 6 months: whether the association between rising type A plan scores and falling body mass index differs between people with blood type A and people of other blood types; adults with overweight in a dietary intervention study	Very low	This page’s own judgment, using GRADE. Wang et al., 2018: 973 adults in the baseline analyses and 576 in the 6-month analyses; full report read	Observational evidence starts at low. The full report was read. There were 973 baseline participants and 576 six-month analyses after exclusions. Blood-type matching and plan adherence were not randomized. Risk of bias is serious, lowered 1 level: food-score measurement, selective availability at follow-up and residual confounding limit the association. Inconsistency could not be assessed for a single study. Indirectness is not serious for this question, which names this exposure and population. No verified estimate and interval for the specified interaction have been extracted, so comparative imprecision remains unresolved. Publication bias could not be assessed. No upgrade. Low minus 1 for serious risk of bias gives Very low. The absence of an interaction meeting the study’s multiplicity threshold does not establish equivalence.
Body weight change at 16 weeks on a low-fat vegan diet, people with blood type A compared with people of all other blood types; adults with overweight in the vegan group of a randomized trial	Very low	This page’s own judgment, using GRADE. Barnard et al., 2021: 68 blood-typed vegan-arm completers, 20 with type A and 48 with other types, of 122 assigned to the vegan diet; full report read	The diet was assigned by chance but blood type was not, so this comparison between blood types is observational and starts at low. It does not inherit the certainty of the randomized diet comparison. Risk of bias is serious, lowered 1 level: blood typing began after a protocol amendment in the second of four study replications, so the 68 analyzed are the vegan-arm completers typed after that point; the blood type groups differed at baseline, particularly in lipids; and the groups were compared without adjustment for confounding. Inconsistency could not be assessed for a single study. Indirectness is not serious for this question, which names the vegan diet. Imprecision is serious: the report gives intervals for each group’s own change in measured weight (a fall of 4.8 to 6.7 kilograms with type A and 5.8 to 8.2 kilograms with other types), not for the difference. This page’s approximation from those intervals, using a t distribution for each group and an unequal-variance comparison of the 2 independent groups, gives a measured decrease 1.3 kilograms smaller in the type A group, with an approximate interval from 0.2 kilograms larger to 2.8 kilograms smaller. It includes no difference and differences beyond 1 kilogram, an analytical convention rather than a validated clinical cutoff or a goal, and rounding in the published figures limits its accuracy. Publication bias could not be assessed. No upgrade. Low minus 1 for serious risk of bias is Very low, and the imprecision cannot lower it further. P = 0.09 does not establish similarity.
Total cholesterol change at 16 weeks on a low-fat vegan diet, people with blood type A compared with people of all other blood types; the same participants	Very low	This page’s own judgment, using GRADE. Barnard et al., 2021: the same 68 completers; full report read	Starts at low for the reasons in the row above. Risk of bias is serious, lowered 1 level, for the same reasons; the baseline difference in lipid levels between the blood type groups makes an unadjusted comparison of cholesterol change particularly limited. Inconsistency could not be assessed. Indirectness is not serious. Imprecision is serious: the report gives intervals for each group’s own change (a fall of 0.6 to 33.8 mg/dL with type A and 11.3 to 25.2 mg/dL with other types), not for the difference. This page’s approximation by the same method gives a fall 1.1 mg/dL smaller in the type A group, with an approximate interval from 16.7 mg/dL larger to 18.9 mg/dL smaller, which includes no difference and differences well beyond 5 mg/dL, an analytical convention, in both directions. Publication bias could not be assessed. No upgrade. Low minus 1 for serious risk of bias is Very low, and the imprecision cannot lower it further. P = 0.90 does not establish similarity.
Systolic blood pressure, highest against lowest third of the type A plan score, all blood types together (the plan as a way of eating, whatever a person’s blood type); young adults; a single time point	Very low	This page’s own judgment, using GRADE. Wang et al., 2014: 1,455 young adults	Observational evidence starts at low. Risk of bias is serious, lowered 1 level: people whose eating resembles the plan may differ in other ways, and the analysis was adjusted only for age, sex, ethnocultural group and calories, with eating measured by questionnaire. Inconsistency could not be assessed for a single study. Indirectness is not serious for this question about usual eating. Imprecision is also serious: averages of 115.4 and 117.6 mm Hg give a difference of 2.2 mm Hg, with this page’s conditional approximate interval, treating the 2 adjusted averages as independent, from 0.8 to 3.6 mm Hg lower, which includes differences smaller and larger than 2 mm Hg. Under the zero-covariance approximation, the interval includes differences smaller and larger than this page’s 2 mm Hg convention. The covariance of the adjusted averages is not reported, so the actual model-based interval remains unresolved. It is not a reported interval. Publication bias could not be assessed. No upgrade. Low minus 1 for serious risk of bias reaches Very low, and the imprecision cannot lower it further. This is not matching evidence, not a randomized effect of adopting the plan, and not a rating of the type A plan as followed.
Any health outcome, people of one blood type assigned to the plan for their type compared with people of the same blood type assigned to a standard diet	Not rated	This page’s own judgment. No study of this design identified	A 2013 systematic review described this as the design needed and found no such study (Cusack et al., 2013), and the searches for this page did not identify one published since. Not rated is an abstention, not a finding that the plan does nothing.
Heart attacks, strokes, cancer or death in people following the plan, against any comparison	Not rated	This page’s own judgment. No study identified	The searches for this page did not identify a study of the plan that measured these outcomes. Associations between blood type and heart disease are a different question; see the last row.
Onset or worsening of disordered eating in people following the plan, against people not following it	Not rated	This page’s own judgment. No study identified	The searches for this page did not identify a study with this design. The scope box and section 6 keep the eating-behavior referral boundary regardless.
Digestive symptoms, type-specific lectin-blocking supplements against placebo or no supplement	Not rated	This page’s own judgment. No trial identified	The searches for this page did not identify a trial of the supplements sold for each blood type. The company’s description of what its products do is reported in section 2 as the company’s claim.
Coronary heart disease, people with blood types A, B or AB compared with type O (blood type itself, not the diet)	Not rated	This page’s own judgment. He et al., 2012: 2 prospective cohorts, 89,501 adults; reported for context	This page rates the diet, and a cohort study of blood type does not estimate the effect of eating by blood type. The association itself is reported in section 4: hazard ratios of 1.06, 1.15 and 1.23 for types A, B and AB against type O. It is not transferred to any row about the diet.

How this archive talks about weight. This page reports what studies measured, including body weight, and does not treat weight change as a goal or as a measure of a diet’s success. Weight a person loses without meaning to can be a sign of a health problem, and where it is, the page names it as a reason to see a clinician. People are described the way the study described them.

Last reviewed: Reviewed and approved by James Peth, MS, MPH, PhD, on November 14, 2026, after a combined final review of the sixth batch of 5 archive pages, the repairs it asked for and a targeted recheck. Written October 21, 2026, and completed November 8, 2026.

What changed: First version. The judgments in the table are this page’s own, assigned from the evidence. The systematic review, the second Toronto analysis and the vegan trial analysis were read in full during the outside review; a few other sources could be read only as abstracts or as an author manuscript, and the page says so where each is used. After that review, the summaries were narrowed so that not detecting a difference by blood type is not described as showing that matching makes no difference, the precision judgments in the first 3 rows were marked as resting on approximations, and the vegan trial rows were updated from the full report. The diet’s books were not opened; the plans are described from the diet’s own website and from published accounts of them. Prices are a snapshot from the company’s pages, rechecked November 8, 2026, and can change.

Next scheduled review: Each year, because the company’s products and prices change. Pages built on annual guidance are rechecked each year.

---

Diet Reference Archive | One James

Prepared by James Peth, MS, MPH, PhD

© 2026–2027 One James. All rights reserved.

This page is for education and coaching practice. It is not medical or dietary advice.