One clarification comes before any of the evidence, because the whole debate usually skips it: veganism is not a diet. It is a way of living that seeks to avoid animal exploitation as far as is possible and practicable, and by itself it says nothing about what is on the plate. Two people can both be vegan while one lives on fried food, refined flour, and soda and the other on lentils, greens, and whole grains. So a sentence like vegans in study X had outcome Y is only ever as meaningful as the diet those particular vegans happened to eat, which is why a good deal of the confusion in this area dissolves the moment the word is pinned down.

What this review examines is a well-planned vegan way of eating: legumes, greens, grains, fruits, vegetables, mushrooms, nuts, seeds, seaweeds such as nori, herbs and spices, with variety within each group, plus a reliable vitamin B12 supplement and vitamin D as a baseline that a person adjusts for their own circumstances. That is the exposure being tested here. When the evidence below draws on unsupplemented cohorts recruited decades ago, the gap between vegan and well-planned vegan turns out to do most of the explanatory work.

This is a review of 48 peer-reviewed studies on what happens to chronic disease risk as a person’s diet moves from heavy animal-product consumption toward a fully plant-based one. Every study was screened for peer review, journal quality, funding source, conflicts of interest, and retraction status, then sorted by the hierarchy of evidence, from umbrella reviews and randomized trials at the top to cross-sectional and modeling studies at the bottom. Every number below was checked against the source paper rather than an abstract or a press release. The question it answers is narrower and more useful than the usual shouting match: what does the best available evidence say about a well-planned, B12-supplemented vegan diet compared with eating animal products, and does risk climb step by step with the amount of animal food a person eats?

The piece is layered on purpose. The first section is the plain-English version anyone can follow. Each section after that adds detail, so a reader can stop whenever they have what they came for.

The short answer

Is a vegan diet healthier than eating meat?

Across the strongest evidence available, the answer points one way. Diets built on whole plant foods, with animal products reduced or removed, are consistently associated with less coronary heart disease, less type 2 diabetes, lower body weight, lower blood pressure, and a lower chance of dying early. This shows up in randomized trials where people are assigned diets and their blood is measured, in cohorts of hundreds of thousands of people followed for decades, and in the genetic evidence for the mechanism underneath it all. The mechanism is not mysterious: plant-based eating lowers LDL cholesterol, LDL cholesterol is an established cause of the artery-clogging process behind most heart attacks, and diets built on meat, butter, and cheese push LDL the other way.

Does eating fewer animal products lower chronic disease risk?

Yes, and the relationship is graded. In the studies that put every dietary pattern inside one dataset, risk tends to fall as animal-product intake falls. The clearest example is type 2 diabetes in the Adventist cohort, where new cases over two years fell steadily across the diet groups, from 2.12 percent of non-vegetarians to 1.29 percent of pescatarians, 1.08 percent of lacto-ovo vegetarians, and 0.54 percent of vegans (Tonstad 2013). Within low-carbohydrate diets, the version built on animal foods was tied to higher death rates and the version built on plant foods to lower ones, which shows the animal-versus-plant source matters even when the macronutrient label is held constant (Fung 2010; Seidelmann 2018).

Is there a safe amount of meat, or a sweet spot where a little animal food helps?

No study in this set found one. There is no identified point at which adding animal products back to a well-planned plant-based diet improves outcomes. For the foods with the steepest harm gradient, red and processed meat, the evidence is dose-dependent with no protective floor: less is proportionally better, and nothing in the data suggests the direction reverses at low intake. The reasonable way to say this is not that a single serving of meat will hurt you, any more than one dessert will, but that the evidence gives no reason to think any amount of it is doing you good, and considerable reason to think the direction of travel is toward zero. The usual reply here is evolutionary, that humans are built for meat and therefore need some, but that claim does not survive contact with the fossil, dental, and genetic record, which is the subject of what human evolution actually says about how we evolved to eat.

What does a well-planned vegan diet mean?

Throughout this review it means what it means across this site: a diet composed of legumes, greens, grains, fruits, vegetables, mushrooms, nuts, seeds, seaweeds, herbs, and spices, with variety within each of those groups, a reliable vitamin B12 supplement, and vitamin D as needed. Seaweed and iodized salt cover iodine; flax, chia, hemp, and walnuts cover plant omega-3, with an algal supplement removing the question entirely; legumes, greens, and fortified foods cover calcium and protein; and iron, zinc, and selenium are reachable from a varied whole-food diet, helped along by pairing iron-rich foods with vitamin C, by soaking, sprouting, fermenting and leavening, and by a couple of Brazil nuts for selenium. The one nutrient genuinely specific to this way of eating is B12, which is made by bacteria rather than by animals and is the single non-negotiable supplement. Vitamin D is the other common addition, and it is not a vegan issue at all, since it is low across every diet in low-sun conditions and is routinely supplemented for the whole population. It is worth being blunt about a false asymmetry here: a fortified plant milk and a fortified dairy product are the same move, both foods with nutrients added at the factory, so the vegan taking one is doing nothing the fortified-food-eating omnivore is not also doing. This diet does not mean a diet of refined flour, sugar, fried food, and mock meats that happens to contain no animal products. That distinction matters for this review, because several of the findings that get quoted against plant-based eating come from cohorts where most of the vegans were not supplementing at all, and those findings say more about missing B12 than about plants. The practical details live in The Great B12 Gamble, essential nutrients on a vegan diet, and the nutritional case for plants over animals.

This is health information, not medical advice. Anyone with questions or concerns about their own diet should consult a physician, registered dietitian, or nutritionist, ideally one who practices lifestyle medicine, which puts the evidence first and respects a person’s values.

How this review was built

Which studies were included?

Forty-eight documents made the cut, sorted into six tiers. Tier 1 holds umbrella reviews and meta-analyses of randomized trials, including a Cochrane review. Tier 2 holds individual randomized trials, including an NIH metabolic-ward feeding study and an identical-twin trial. Tier 3 holds meta-analyses of prospective cohorts. Tier 4 holds Mendelian randomization and other genetic causal evidence. Tier 5 holds the landmark cohorts: EPIC-Oxford, the Adventist Health Study-2, UK Biobank, and the Harvard Nurses’ Health and Health Professionals cohorts. Tier 6 holds cross-sectional and modeling studies, kept for context only.

How were conflicts of interest handled?

Every paper’s funding statement and disclosure section was read from the PDF, not summarized from memory. Forty-three of the 48 were funded by public bodies such as the NIH, the Medical Research Council, Wellcome, Cancer Research UK, or national research councils, with no material conflicts. Five carry a minor flag, each named where the study appears and listed together in the fact-check section near the end. Advocacy-affiliated trials are disclosed rather than excluded, on the principle that peer review and journal quality are the real filters and that affiliation alone is a cruder test than the work deserves. So trials run by the Physicians Committee for Responsible Medicine appear here, in peer-reviewed journals, with their affiliation named at the point of citation, and they are treated as corroboration rather than as anchors for any conclusion. One study is excluded on design grounds rather than funding: the Ornish Lifestyle Heart Trial bundled diet with exercise, stress management, and group support, so it cannot isolate the effect of the food.

Were any of these studies retracted?

None. Five have published corrections, all checked and described in that same section. The retraction history of the field is worth knowing anyway: the original 2013 PREDIMED Mediterranean-diet paper was retracted and republished in 2018 over randomization problems, and a 2025 study claiming a ketogenic diet did not promote arterial plaque was retracted in 2026 after its methods and undisclosed conflicts fell apart (Retraction Watch 2026). Neither is part of this evidence set, but both are reminders that the checking has to be done.

What is established beyond reasonable doubt

Do plant-based diets lower LDL cholesterol?

Yes, and this is causal, because it comes from randomized trials. Pooling 30 trials with about 2,400 participants, vegetarian and vegan diets lowered total cholesterol by 0.34 mmol/L, LDL cholesterol by 0.30 mmol/L, and apolipoprotein B, the single best marker of artery-clogging particles, by about 13 mg/dL compared with omnivorous diets (Koch 2023). In an identical-twin trial where one twin ate vegan and the other omnivore for eight weeks, with both diets designed to be healthy, the vegan twin’s LDL fell by about 14 mg/dL more, fasting insulin fell, and body weight fell by about 2 kg more (Landry 2023). The trial was funded by the Vogt Foundation and National Institutes of Health grants; one of its authors disclosed funding from Beyond Meat outside the submitted work, which is named here because critics raise it. A methylation follow-up of the same twins found the vegan arm showed significant decreases in epigenetic age acceleration across multiple clocks and organ-system scores while no clock moved in the omnivore arm; read it with two caveats, since several authors are employees of the company that sells the epigenetic tests used, and the vegan twins also ate fewer calories and lost more weight, so part of the shift may reflect weight loss rather than diet composition, and epigenetic clocks are surrogate markers, not disease outcomes (Dwaraka 2024). When a calorie-matched vegetarian diet was compared head to head with a Mediterranean diet in a randomized crossover, both produced similar weight loss, but the vegetarian arm lowered LDL more (Sofi 2018). For the full staircase of evidence on why LDL matters, see cholesterol, from the ground up.

Does LDL cholesterol cause heart disease?

Yes. This is not a diet study; it is the consensus of the genetic, epidemiological, and trial evidence assembled by the European Atherosclerosis Society, covering more than two million people. LDL causes atherosclerotic cardiovascular disease in a dose-dependent, cumulative way, with no threshold below which lowering it stops helping (Ference 2017). One disclosure belongs with it: the panel’s travel and meeting logistics were paid for by unrestricted educational grants from two pharmaceutical companies, MSD and Amgen, and the paper states that neither company attended the panel meetings or had any role in the manuscript’s design or content. The step-by-step mechanism, how an LDL particle enters the artery wall, gets trapped, and becomes plaque, is traced in how plaque forms. This is the bridge that connects the trial results above to the disease outcomes below. If a diet reliably lowers LDL, it should reliably lower heart disease, and the cohort data confirm that it does.

Do vegan diets cause weight loss and better blood sugar?

In randomized trials, yes. Eleven trials in overweight adults and people with type 2 diabetes found vegan diets produced about 4.1 kg more weight loss than control diets and a modest improvement in HbA1c (Termannsen 2022). An umbrella review rated the reductions in body weight, BMI, waist circumference, LDL, and HbA1c as moderate-certainty evidence (Selinger 2023). A separate meta-analysis of trials in people with diabetes, commissioned to inform European guidelines, found vegetarian patterns improved HbA1c, LDL, and weight, though several of its authors carry long food-industry disclosure lists, so it is included as corroboration rather than as an anchor (Viguiliouk 2019). A separate 16-week randomized trial in 244 overweight adults found a low-fat vegan diet produced significant weight loss, a 34 percent reduction in liver fat, improved insulin sensitivity, and an 18.7 percent rise in postprandial calorie burn; two caveats belong with it, since the control group was asked to make no dietary changes at all rather than follow an active comparison diet, which inflates the apparent effect, and the trial was run by employees of the Physicians Committee for Responsible Medicine, a plant-based advocacy organization, who were involved in its design, conduct, and analysis (Kahleova 2020). Its direction matches the independently funded trials above, so it corroborates rather than carries the finding. The mechanism of the weight effect is covered in the healthiest way to lose weight.

Does a plant-based diet make you eat less without trying?

In the tightest-controlled diet comparison in this entire set, it did. Twenty adults lived on an NIH metabolic ward and ate two diets in random order, two weeks each, with unlimited food: a plant-based, low-fat diet and an animal-based ketogenic diet, both minimally processed. On the plant-based diet they spontaneously ate 689 fewer calories per day, with no difference in hunger or enjoyment (Hall 2021). That is the strongest available evidence that the difference is not willpower. The body-composition results are the part usually left out of the retellings, and they need stating carefully. Participants lost weight on both diets, about 1.8 kg on the ketogenic diet and 1.1 kg on the plant-based one, a difference that was not statistically significant. Most of the weight lost on the ketogenic diet was fat-free mass rather than fat, and body fat did not change significantly by the end of it, while the plant-based low-fat diet produced a significant loss of body fat, about 0.7 kg over two weeks. The difference in total fat lost between the two diets did not reach statistical significance over so short a period, though the rate of fat loss was significantly faster on the plant-based diet. Weight on a scale and fat on a body are not the same measurement.

Does a keto diet raise cholesterol?

Yes, and with unusual certainty. An umbrella review of 17 meta-analyses covering 68 randomized trials found only four conclusions about ketogenic diets supported by high-quality evidence. One is that they raise LDL cholesterol (Patikorn 2023). They also lower triglycerides and reduce seizure frequency, and they produce short-term weight loss and lower HbA1c with moderate certainty. Set beside the LDL consensus above, this is the core of the case against animal-heavy low-carbohydrate eating: it raises the marker that causes heart disease, and no one has followed keto eaters long enough to see what happens next.

Does cutting saturated fat actually prevent heart attacks?

Yes, and this is randomized event data, not a marker. The Cochrane review of 15 trials with 56,675 participants found that reducing saturated fat cut cardiovascular events by 17 percent, with the benefit tracking how much LDL fell and what the fat was replaced with: swapping it for plant unsaturated fat worked, swapping it for refined starch and sugar did not (Hooper 2020). That last detail also explains the studies sometimes said to have exonerated saturated fat; they compared it against refined carbohydrate, a bad replacement, rather than against plant foods. Since animal products are the main source of saturated fat and account for essentially all the cholesterol in the diet, this is the trial-level link between eating fewer of them and having fewer heart attacks.

Does fiber lower the risk of dying, and where does it come from?

A synthesis commissioned for the World Health Organization, pooling 185 prospective studies and 58 trials, found the highest fiber consumers had 15 to 30 percent lower all-cause and cardiovascular mortality and 16 to 24 percent less heart disease, stroke, type 2 diabetes, and colorectal cancer, with the benefit greatest around 25 to 29 grams a day and still climbing beyond (Reynolds 2019). Fiber comes essentially only from plants; meat, fish, eggs, and dairy contain none to speak of. That gives the gradient in this review a second mechanism alongside LDL: each step away from animal foods and toward whole plants adds fiber, and an all-animal diet scores a structural zero on the single dietary factor most consistently linked to living longer. How fiber does this, from binding cholesterol in the gut to the butyrate that colon bacteria make from it, is covered in the carbohydrate deep dive.

What the large population studies show

Do vegetarians and vegans have less heart disease?

Consistently. Pooling the prospective cohorts, vegetarians had about 15 percent less cardiovascular disease and 21 percent less ischemic heart disease than non-vegetarians (Dybvik 2023). In EPIC-Oxford, the largest British cohort, vegetarians and vegans had 22 percent lower rates of ischemic heart disease over 18 years, a hazard ratio of 0.78 with a confidence interval of 0.70 to 0.87, which works out to roughly ten fewer cases per 1,000 people over a decade (Tong 2019). An older meta-analysis found 25 percent lower ischemic heart disease and 8 percent lower cancer incidence in vegetarians, and 15 percent lower cancer incidence in vegans (Dinu 2017). An umbrella review of 48 prior reviews reached the same conclusion on lipids, blood sugar, weight, inflammation, ischemic heart disease, and cancer, while rating the underlying evidence as average quality with large heterogeneity (Capodici 2024).

Do vegans live longer?

Probably, but the vegan-specific numbers are less precise than the heart-disease numbers, for one simple reason: strict vegans are a small slice of every cohort. In the Adventist Health Study-2, all vegetarians combined had 12 percent lower all-cause mortality than meat-eaters. Vegans specifically had 15 percent lower mortality, with a confidence interval of 0.73 to 1.01 that just touched the line of statistical significance; pescatarians had 19 percent lower, with an interval of 0.69 to 0.94 (Orlich 2013). Those two intervals overlap almost entirely. The 2024 follow-up of the same cohort, with more than twice the deaths, tightened the vegetarian estimate to 11 percent lower mortality, with large reductions in deaths from ischemic heart disease, diabetes, kidney failure, and infection (Abris 2024). The full accounting of what that cohort can and cannot show is in what the Adventist Health Studies actually tell us.

Outside Loma Linda, the picture is mixed. EPIC-Oxford found no significant difference in all-cause mortality between vegetarians and comparable non-vegetarians (Appleby 2016), and the pooled cohort estimate for vegans and cardiovascular disease was 8 percent lower with an interval that crossed 1.0 (Dybvik 2023). Read together, the heart-disease benefit is solid and the total-mortality benefit is probable but imprecise, and the reason is sample size, not a contrary signal.

Do vegans get less diabetes?

This is where the gradient is clearest. In the Adventist cohort, new diabetes cases over two years fell across the diet groups in order: 2.12 percent of non-vegetarians, 1.29 percent of pescatarians, 1.08 percent of lacto-ovo vegetarians, and 0.54 percent of vegans. In the adjusted models, which controlled for body weight along with age, sex, education, income, television watching, physical activity, sleep, alcohol, and smoking, three groups reached significance against non-vegetarians: vegans at an odds ratio of 0.38 (0.24 to 0.62), lacto-ovo vegetarians at 0.62 (0.50 to 0.76), and semi-vegetarians at 0.49 (0.31 to 0.76). Pescatarians did not reach significance. Two things are worth saying plainly about that. The semi-vegetarian result sits between the vegan and lacto-ovo figures rather than below them, so the adjusted ladder is not perfectly monotonic even in the cohort where the gradient is strongest. And the authors note the associations grew stronger when body weight was removed from the models, which means part of the benefit runs through lower body weight rather than around it (Tonstad 2013). Body weight ran the same staircase: average BMI was 23.6 in vegans, 25.7 in lacto-ovo vegetarians, 26.3 in pescatarians, and 28.8 in non-vegetarians (Tonstad 2009). In three large US cohorts, a healthful plant-based diet was associated with 34 percent lower diabetes incidence, and an unhealthful plant-based diet with 16 percent higher (Satija 2016). And from the meat side, a federated analysis of 1.97 million adults across 20 countries found a modest daily serving of processed meat linked to 15 percent higher type 2 diabetes risk and unprocessed red meat to 10 percent higher (Li 2024).

Do vegans get less cancer?

In the Adventist cohort, vegans had 16 percent lower overall cancer incidence than meat-eaters and 34 percent lower incidence of female-specific cancers (Tantamango-Bartley 2013). Across three Harvard cohorts covering nearly five million person-years, each ten-point rise in healthful plant-based eating was associated with 7 percent lower digestive-system cancer and 32 percent lower liver cancer, while unhealthful plant-based eating was tied to higher colorectal cancer (Kim 2023). A Mendelian randomization study, which uses genetic variants to sidestep confounding, linked genetically predicted processed-meat intake to roughly double the risk of colorectal cancer (Yun 2023). That converges with the formal carcinogen review: in 2015, a working group of 22 scientists from 10 countries convened by the WHO’s cancer agency classified processed meat as a Group 1 carcinogen, meaning the evidence for its link to colorectal cancer is as strong as evidence gets, and estimated each daily 50 gram portion raises colorectal cancer risk by 18 percent, with red meat classified one rung down as probably carcinogenic (IARC 2015). Group 1 describes the certainty of the evidence, not the size of the risk, which is far smaller than tobacco’s; the point is the direction, established at the highest level of review. Dairy deserves its own line: within the Adventist cohort, higher dairy milk intake was associated with 50 percent higher breast cancer risk, while soy milk was not (Fraser 2020). One honest complication belongs here rather than buried: a 2026 pooled analysis of 1.8 million people found vegans had about 40 percent higher colorectal cancer risk than meat-eaters, a finding that gets quoted as if it sinks the meat and bowel cancer link. It does not, and the reasons are the same ones this review keeps returning to. The vegan estimate rested on 93 cases against wide confidence intervals, the pescatarians in the very same pooling had 15 percent lower risk, and the excess points at what unplanned vegan diets can lack rather than at anything protective in meat. The full breakdown, including why the paper contradicts the inference drawn from it, is in why a 40 percent number in one subgroup did not overturn the meat and bowel cancer link.

Is a plant-based diet healthy if it is made of junk food?

No, and this is the single most consistent finding across the biggest cohorts. Every large study that scored plant-based diets by quality found the same split. In the Harvard cohorts, the healthiest plant-based pattern was associated with 25 percent lower coronary heart disease, while a plant-based pattern built on refined grains, fries, sweets, and sugary drinks was associated with 32 percent higher (Satija 2017); several of that paper’s authors disclose food- and drug-industry ties, detailed in the fact-check section. In 126,000 UK Biobank participants, the healthful pattern was tied to 16 percent lower mortality, 7 percent less cancer, 8 percent less cardiovascular disease, 14 percent fewer heart attacks, and 16 percent fewer ischemic strokes, and the unhealthful pattern to higher risk on the same outcomes (Thompson 2023). In 205,000 US health professionals followed for up to 32 years, the healthful plant-based index was associated with 17 percent lower combined risk of major cardiovascular disease, diabetes, and cancer (Wang 2023). In 13,000 older Chinese adults, the healthful pattern meant 19 percent lower mortality and the unhealthful pattern 17 percent higher (Chen 2022). In US adults, overall plant-based scores were not associated with mortality at all; only the healthful version was (Kim 2018). A junk-food vegan is not protected. That is not a point against veganism; it is the reason the word planned is in this review’s title.

Does it matter whether a low-carb diet is based on meat or plants?

Enormously, and this is the strongest evidence in the whole set for the animal-versus-plant question, because it holds the macronutrient label constant and changes only the source. In the Nurses’ Health Study and Health Professionals Follow-up Study, a low-carbohydrate score built on animal foods was associated with 23 percent higher all-cause mortality, while a low-carbohydrate score built on plant foods was associated with 20 percent lower all-cause mortality and 23 percent lower cardiovascular mortality (Fung 2010). In the ARIC cohort plus a meta-analysis of 432,000 people, replacing carbohydrate with animal fat or protein raised mortality by 18 percent and replacing it with plant fat or protein lowered mortality by 18 percent (Seidelmann 2018). In NHANES, a healthy low-carbohydrate diet was tied to 9 percent lower mortality and an unhealthy one to 7 percent higher (Shan 2020). The same split shows up in imaging rather than death certificates: in the CARDIA cohort, followed from young adulthood into middle age, a higher low-carbohydrate diet score was associated with greater progression of coronary artery calcium, the calcified plaque that marks advancing coronary disease, and that association was driven by the animal-based version of the pattern rather than the plant-based one (Gao 2021). Same diet name, opposite outcomes, and the thing that flipped the sign was whether the calories came from animals or plants.

Is plant protein healthier than animal protein?

Across 32 cohorts and 715,000 people, plant protein intake was associated with 8 percent lower all-cause mortality and 12 percent lower cardiovascular mortality, with each additional 3 percent of calories from plant protein tied to 5 percent lower risk of death; total and animal protein showed no such benefit (Naghshi 2020). In the Harvard cohorts, each 3 percent of calories from plant protein was associated with 10 percent lower mortality, while animal protein was not associated with all-cause mortality (hazard ratio 1.02 per 10 percent of energy, not significant) but was associated with higher cardiovascular mortality (1.08 per 10 percent of energy, 1.01 to 1.16). That cardiovascular signal was concentrated among people who also smoked, drank heavily, were inactive, or were overweight, but it is present in the main model and should not be described as absent (Song 2016). Among adults with type 2 diabetes, the highest plant-protein intake was associated with 27 percent lower mortality, and animal protein showed no association (Jayedi 2025). The NIH-AARP cohort of more than 400,000 adults found the same substitution effect at the same size: replacing 3 percent of energy from animal protein with plant protein was associated with about 10 percent lower all-cause mortality (Huang 2020). Replacing processed meat with nuts, legumes, or whole grains was associated with lower cardiovascular disease, diabetes, and mortality in a substitution meta-analysis (Neuenschwander 2023). And in six pooled US cohorts, each two servings a week of processed meat was tied to 7 percent higher cardiovascular disease, with smaller significant increases for unprocessed red meat and poultry (Zhong 2020). A mechanism runs underneath this: gut bacteria convert the carnitine concentrated in red meat into TMAO, a compound that accelerates artery disease in animal models, and omnivores produce far more of it than vegans given an identical carnitine dose, because habitual meat eating cultivates the bacteria that make it (Koeth 2013). The wider protein question, including amino acid completeness, whether protein must be combined at each meal, what the 2026 Dietary Guidelines protein target actually rests on, and whether plant eaters build muscle as well, is worked through in the full protein breakdown.

Is white meat better than red meat?

Less than almost everyone assumes. The APPROACH feeding trial compared red meat, white meat, and non-meat plant protein head to head under controlled conditions and found red and white meat raised LDL cholesterol about equally, while the plant protein sources beat both; independent of protein source, higher saturated fat raised LDL and apoB (Bergeron 2019). Swapping beef for chicken is a much smaller step than swapping either for beans, which is exactly what the gradient across the cohorts predicts.

Is there a stepwise increase in disease risk with more animal products?

What does the dose-response actually look like?

Yes, there is a gradient, and it is visible from several independent angles. Within one cohort, diabetes incidence and body weight both climb from vegan through lacto-ovo vegetarian and pescatarian to omnivore, with the semi-vegetarian group the one group that does not sit where a strict ladder would put it (Tonstad 2013; Tonstad 2009). Across protein sources, mortality falls as plant protein rises and does not fall as animal protein rises (Naghshi 2020; Song 2016). Within low-carbohydrate diets, the animal-based version raises mortality and the plant-based version lowers it (Fung 2010; Seidelmann 2018). Across food substitutions, every swap of an animal food for a whole plant food improves the outcome (Neuenschwander 2023). And at the level of mechanism, LDL tracks the animal-fat content of the diet and the disease tracks LDL with no floor (Ference 2017). Five different lines of evidence, one direction.

Is the gradient the same for every animal food?

The slope is steepest for processed and red meat, which show harm at every level of evidence from cohorts to Mendelian randomization (Zhong 2020; Yun 2023; Neuenschwander 2023). It is shallower and less certain for dairy, where one Danish genetic study found no effect of lifelong higher milk intake on mortality (Bergholdt 2018) and a Swedish study found the opposite, so that literature is genuinely unsettled, though the breast-cancer association above is a real signal. And for fish the cohort picture is complicated in a way that deserves its own section below. What the evidence does not show, anywhere, is a category of animal food whose addition to a well-planned plant-based diet improves a hard outcome.

Can a little meat undo a plant-based diet?

The honest framing is proportional. The harm associated with processed and red meat is dose-dependent, which means small amounts carry small risk, the way an occasional dessert does. No study shows a threshold below which the risk turns into a benefit, and no study shows one above which it suddenly appears. So the logic runs the way it does for any dose-dependent exposure: less is better, the benefit continues all the way down, and there is no evidence-based reason to keep a floor of animal food in the diet for health. Whatever reasons people have for keeping some, the science in this set does not supply one.

Do pescatarians outlive vegans? The fish question, handled honestly

Where does the claim that fish-eaters do better come from?

Two places. In the 2013 Adventist mortality paper, pescatarians had a point estimate of 19 percent lower mortality and vegans 15 percent lower, and the pescatarian estimate reached significance while the vegan one narrowly missed (Orlich 2013). In UK Biobank, eating fish rather than meat or poultry was associated with lower risk across several cardiovascular outcomes, while vegetarianism was associated only with lower cardiovascular incidence (Petermann-Rocha 2021). In EPIC-Oxford, fish-eaters had 13 percent lower ischemic heart disease and vegetarians 22 percent lower, so there the vegetarians did better (Tong 2019).

Does the data show fish-eaters outlive vegans?

No, for a statistical reason and two methodological ones. The Adventist vegan and pescatarian confidence intervals, 0.73 to 1.01 and 0.69 to 0.94, overlap almost entirely; each group’s estimate sits inside the other’s interval, and the study was never designed to compare the two head to head. Treating a four-point gap between overlapping estimates as a ranking is exactly the mistake confidence intervals exist to prevent. In UK Biobank the vegetarian group was small enough that its ischemic heart disease estimate rested on only 302 events, which is underpowered rather than null (Petermann-Rocha 2021). And in every one of these cohorts, the comparison is between fish-eaters and vegetarians or vegans who were, for the most part, not supplementing B12, not against the well-planned pattern this review is about.

Is there anything in fish that you cannot get from plants?

No. The nutrients usually named are DHA and EPA, vitamin B12, and iodine. Fish do not make DHA; they accumulate it from algae, and algal supplements deliver the same molecule in the same triglyceride form. Why the body converts plant omega-3 into DHA slowly, and whether that is a defect or a control system, is taken up in precursors versus preformed nutrients. Supplemental DHA does reach the brain: a 2026 randomized trial in 365 older adults at risk of dementia confirmed that a high dose raised the DHA level in cerebrospinal fluid, settling the question of whether an oral supplement crosses into the central nervous system. Two honest qualifications belong with that result, because it is often quoted without them. The trial used fish-derived DHA rather than algal DHA, so it establishes that the molecule gets in, not that a particular source is superior. And its cognitive results were null: participants taking DHA did no better on cognitive testing, and showed no benefit in hippocampal volume or cortical thickness, than those on placebo (Yassine 2026). That cuts against anyone claiming DHA supplements protect the aging brain, whichever source they come from. B12 is made by bacteria, not by fish, and the supplement is the direct route to the same bacterial product. Iodine comes from seaweed such as nori, from iodized salt, or from a supplement. There is no molecule in a fish that a well-planned vegan diet cannot supply, so any advantage fish-eaters show in a cohort has to be explained by something other than a unique nutrient. The evidence on whether vegans need to supplement DHA at all is reviewed in detail in Do Vegans Need DHA and EPA?; the short version is that the supplement trials for heart and brain outcomes have been null, plant omega-3 has its own favorable outcome data, and a small algal dose is cheap insurance rather than a proven benefit.

Why might pescatarians look healthier than they are?

Because of who chooses each diet. Research on dietary motivation finds a consistent pattern: the less restrictive a form of animal-product avoidance, the more likely it is driven by health rather than ethics. Pescatarians are largely health-motivated, consistent with decades of dietary guidance recommending fish, while vegans report stronger ethical motives concerning animals and the environment than health motives (Rosenfeld 2018; Rosenfeld and Tomiyama 2021; Hopwood 2020). Follow that through and the usual healthy-user objection inverts. If health-optimizers preferentially land at pescatarianism while ethically motivated people land at veganism, the health-conscious bias favors the fish-eating group, and an ethically motivated vegan who never thought about B12 gets counted in the vegan column. A related point from the omega-3 literature: through the decades these cohorts ran, people were told to swap red meat for fish, so a high fish or omega-3 marker often identifies the person who followed health advice generally, not the effect of fish itself. The motivation research comes from general populations rather than from Adventists, so the sorting is plausible in those cohorts rather than proven there, but it is the more likely direction of bias, not the less. A related ranking comes up often in this debate: the American Heart Association’s 2023 scoring of popular diets placed pescatarian eating in its top tier and vegan eating in its second. That scoring measured how closely each pattern aligns with the association’s dietary guidance, not health outcomes, and it cited the vegan pattern’s restrictiveness and B12 risk; the ranking and its reasons are covered in the keto section below.

Is fish safe to eat during pregnancy?

The official guidance in the United States and elsewhere tells women who are pregnant, breastfeeding, or planning a pregnancy to avoid the highest-mercury species entirely and to cap total fish intake, because methylmercury crosses the placenta and harms the developing nervous system. That is not a blanket instruction to avoid fish, and it should not be quoted as one, but it is a reminder that fish is the one food category with an official contaminant warning attached to it, and that the DHA it supplies is available from tank-grown algae with no mercury, dioxin, or microplastic load at all. For the developmental side of the question, see raising children vegan.

So what is the honest verdict on fish?

The cohort findings that favor fish-eaters are real but do not show what they are usually claimed to show. They compare fish-eaters against unsupplemented vegetarians, the differences are mostly within statistical noise, no study has tested fish against a well-planned vegan diet, and nothing in fish is nutritionally unique. What the fish data do establish is that fish-eaters do better than meat-eaters, which is one more step on the gradient this review describes, not an exception to it.

The stroke and dementia signals: what they actually show

Do vegetarians have a higher risk of stroke?

In one large cohort, yes. EPIC-Oxford found vegetarians had 20 percent higher total stroke rates, a hazard ratio of 1.20 with an interval of 1.02 to 1.40, mostly hemorrhagic stroke, working out to about three extra strokes per 1,000 people over ten years against ten fewer heart attacks (Tong 2019). This is the most-quoted finding against plant-based eating, and it is where the definition of well-planned earns its keep.

Was B12 status controlled in the stroke finding?

No. EPIC-Oxford recruited in the 1990s, before routine supplementation, and its own biomarker sub-study found 52 percent of the cohort’s vegans and 7 percent of its vegetarians were B12-deficient by a strict cutoff, against one omnivore in 226 (Gilsing 2010). Low B12 raises homocysteine, which is associated with stroke, and the study’s mortality and stroke models included neither serum B12 nor supplement use. A finding from a cohort where half the vegans were deficient in a nutrient linked to the outcome cannot be transferred to a supplemented diet; it is evidence about missing B12, not about plants. The stroke signal also did not replicate: the Taiwanese Tzu Chi cohort found vegetarians had roughly half the stroke rate of non-vegetarians (Chiu 2020), and a 2022 systematic review found no overall association between vegetarian diets and stroke. The full B12 story, including what deficiency does and how to prevent it for a few dollars a year, is in The Great B12 Gamble.

What about the Adventist dementia and stroke deaths?

The 2024 Adventist follow-up found that among participants aged 85 and older, vegetarians had higher rates of death from stroke, dementia, and Parkinson’s disease, with two of the three intervals touching or crossing 1.0 (Abris 2024). Three things put this in proportion. The vegetarian category in that analysis pooled everyone who did not eat meat, including pescatarians who eat fish and carry omega-3 levels similar to omnivores, and the signal appeared anyway, which argues against the idea that fish or DHA is what was missing. The Adventist calibration studies found about 15 percent of vegans fell below the B12 requirement, and the mortality models did not adjust for B12 status (Damayanti 2018). And the study’s own authors attributed the signal to possible B12 deficiency and low long-chain omega-3, that is, to nutrient gaps with identifiable remedies, not to anything plants do. A signal worth watching, and the strongest point the other side has, but one that points toward supplementation and monitoring rather than toward eating animals.

Do vegans break bones more easily?

In EPIC-Oxford, vegans had 2.31 times the hip-fracture risk of meat-eaters, about 15 extra hip fractures per 1,000 over ten years, a figure already adjusted for body weight (Tong 2020). The finding is real and it replicates, and it also deserves its precision context. This is a well-powered cohort, 54,898 people over 17.6 years with fractures taken from hospital records rather than self-report, and it adjusted for age, sex, body weight, smoking, alcohol, physical activity, socioeconomic status, and education while stratifying by calcium and protein, so it is not the uncontrolled study it is sometimes painted as. What it could not see matters just as much: it had no data on supplement use, no serum vitamin D at analysis, no measure of diet quality within the vegan group, and no measure of how long anyone had been vegan. And the vegans it did see were an unsupplemented 1990s population, averaging 591 mg of calcium a day against the meat-eaters’ roughly 1,000, drawn from a cohort in which a separate biomarker study found about half the vegans B12-deficient. The whole-cohort signal rests on 41 vegan hip fractures, and when the 2025 pooled analysis combined the available cohorts, it put the vegan estimate at 1.75 with an interval of 1.17 to 2.63 and graded it very low certainty, meaning the true effect is likely to differ substantially from the estimate (Ballarin 2025). The umbrella review likewise rated the fracture evidence low certainty (Selinger 2023).

What drives it is largely identified, and it is not the plants. In the earlier EPIC-Oxford analysis, vegans getting at least 525 mg of calcium a day had a fracture rate ratio of exactly 1.00 (Appleby 2007). In the Adventist cohort, vegan women taking both calcium and vitamin D had no excess hip-fracture risk at all, against nearly triple the risk for those taking neither, an independent replication of the calcium story on another continent (Thorpe 2021). Body weight mediated about a quarter of the excess in the cohorts that tested it, and protein turned out to be a minor factor: adjusting for calcium explained roughly three times as much of the excess as adjusting for protein, and vegans already eating adequate protein showed no lower risk (Tong 2020). One honest caveat remains: in the strongest single study, even the higher-calcium vegans retained some excess, so something beyond the measured nutrients is still unaccounted for, with the type of exercise a leading candidate, since vegans who did resistance training had bone microarchitecture indistinguishable from omnivores who trained while vegans who did not had measurably worse structure (Wakolbinger-Habel 2022). And diet quality flips the sign entirely: across 70,000 postmenopausal women followed up to 30 years, a healthful plant-based diet was associated with 21 percent lower hip-fracture risk and an unhealthful one with 28 percent higher (Sotos-Prieto 2024). The full study-by-study treatment, including what to actually do about calcium, vitamin D, body weight, and lifting, is in do vegans have weaker bones?. Note the pattern across every adverse finding in this section: they cluster around nutrients and habits a plant-exclusive diet supplies poorly unless someone plans for them, not around anything plants do to the body. That is the argument for planning, and it is the reason this review defined its subject the way it did.

Are there any other trade-offs?

A meta-analysis of oral-health studies found vegans and vegetarians had better periodontal health but more dental erosion than omnivores, a result its authors called counterintuitive and connected to acidic foods and drinks such as citrus fruits, fruit juices, and apples; among adults over 60, caries were more common in omnivores, while complete tooth loss was more common in vegetarians (Azzola 2023). A systematic review of four randomized trials found vegetarian and omnivorous athletes gained strength and muscle equally from training, with creatine supplementation helping both groups the same amount (Seimetz 2023), and a twelve-week controlled trial at matched protein intake found habitual vegans and omnivores built muscle and strength identically (Hevia-Larrain 2021). Neither finding is a reason to eat animals; both are reasons to know the erosive potential of acidic fruit and drinks, and to know what creatine does. The broader questions of plant toxins and antinutrients, raw diets, and the anti-inflammatory diet each have their own review on this site.

Keto, paleo, Atkins, and carnivore: what the evidence can and cannot say

Is there any long-term study of the carnivore diet?

No. The entire published human outcome literature on carnivore eating is one online survey of 2,029 self-selected adults who had been on the diet for at least six months. It had no control group, no blood tests, and no disease endpoints; it reported that participants said they felt well (Lennerz 2021). Its senior author discloses royalties from books recommending carbohydrate-restricted eating. That is the whole of it. Any claim that carnivore eating lowers or raises long-term chronic disease risk is an extrapolation from risk factors, and the risk factor that matters most, LDL, moves in the wrong direction. When a physician who once ate plant-based switches to red meat and eggs and presents that switch as evidence, the specific claims tend to be the same handful, on lipids, insulin, DHA, and diabetes reversal, each answered against the published evidence in the Ken Berry vegan-cardiologist fact-check.

Is there any long-term study of keto or paleo diets and chronic disease?

No. Every ketogenic-diet trial in the umbrella review was short, with a median of 13 weeks, and 76 percent of the meta-analyses were rated critically low on quality (Patikorn 2023). Stepping back from individual trials, when the American Heart Association scored ten popular dietary patterns against its own 2021 heart-health guidance in a formal scientific statement, the very low-carbohydrate/ketogenic and paleo patterns landed in the bottom tier (Gardner 2023). The top tier went to DASH, Mediterranean, pescatarian, and lacto-ovo vegetarian eating, and it is worth saying plainly that the vegan pattern did not make it: it scored 78 out of 100, in the second tier alongside low-fat diets. The association’s own summary credited the vegan pattern’s emphasis on fruits, vegetables, whole grains, legumes, and nuts, and attributed the lower score to how restrictive it can be to follow long term or when eating out, and to the risk of vitamin B12 deficiency (AHA News 2023). The B12 point is exactly why the definition used throughout this review treats a B12 supplement as non-negotiable; the adherence point is a real practical cost, and a different kind of claim from a health effect of the foods themselves. Paleo trials are likewise short and measure risk factors, where paleo improves weight, blood pressure, and lipids relative to standard diets in the short term; the meta-analysis had to be corrected after a methods critique, and the corrected version is the one used here (Ghaedi 2019). The one study that pitted a low-fat plant-based diet against an animal-based ketogenic diet under controlled conditions found the plant-based diet led to 689 fewer calories eaten per day (Hall 2021). The one head-to-head trial of a healthy low-fat diet against a healthy low-carbohydrate diet, both built on quality whole foods, found no difference in weight loss at 12 months, which shows once more that quality, not macronutrient label, is what moves the needle (Gardner 2018).

What did the retracted keto plaque study claim?

A 2025 paper in a cardiology journal claimed that people whose cholesterol rose on a ketogenic diet did not develop more arterial plaque over a year. The journal retracted it in 2026, stating that concerns had been raised about the methodology and that the errors identified were too great to fix with a correction. The fuller account of what went wrong, including selective reporting, contested statistics, and an undisclosed equity stake held by one of the authors, comes from the reporting that followed rather than from the retraction notice itself (JACC retraction notice; Retraction Watch 2026). It is mentioned here because it is still circulating, and because it is the clearest recent example of why conflict-of-interest screening was built into this review. The cholesterol-raising effect of these diets is established; the claim that it does not matter has never survived peer scrutiny. See also dietary fat, seed oils, and coconut oil versus animal fat.

Fact checks and corrections made while building this review

Which studies had published corrections?

Five, and none affects a conclusion drawn here. The DIETFITS trial was corrected twice in 2018: the lipid rows in its Table 3 were mislabeled mmol/L when the values were mg/dL, with the conversion instructions reversed, and a funding attribution was fixed; the primary weight-loss result was untouched (Gardner 2018). The identical-twin trial issued a December 2023 erratum correcting one sentence describing its secondary outcomes (Landry 2023). The paleo meta-analysis published a corrigendum re-running its analysis with a corrected method after a critique of how it had imputed standard deviations (Ghaedi 2019). The Harvard protein paper has a minor erratum (Song 2016). The Global Burden of Disease dietary-risk paper issued a 2021 correction adding a collaborator who had been left off its author list, with no change to any figure (Lancet Department of Error 2021).

Which conflicts of interest were found in the source documents?

Reading the actual disclosure sections changed three verdicts from the initial screen. The Harvard healthful-versus-unhealthful coronary disease paper, initially marked clean, discloses that one author was an AbbVie employee and that two senior authors had received support from the California Walnut Commission, Metagenics, and the Blueberry Highbush Council, so it is now marked with a minor flag (Satija 2017). The CARDIVEG trial, initially flagged for possible food-retailer funding, states plainly that it had no funding and no disclosures, so it is now marked clean (Sofi 2018). The vegetarianism genetics paper declares no competing interests despite a co-author’s affiliation with a genomics company, so it too is marked clean (Yaseen 2023). The Harvard protein paper was also read more carefully than the first summary of it: animal protein was not associated with mortality overall, and the harm was confined to people with other unhealthy habits, which is how it is reported above (Song 2016). With those verdicts settled, five of the 48 carry a minor flag, each also named where the study appears: the Harvard coronary paper above; the diabetes meta-analysis, several of whose authors list food-industry ties (Viguiliouk 2019); the carnivore survey, whose senior author discloses royalties from books recommending carbohydrate-restricted eating (Lennerz 2021); the European Atherosclerosis Society consensus on LDL, whose panel logistics were funded by unrestricted educational grants from MSD and Amgen (Ference 2017); and the identical-twin trial, one of whose authors disclosed funding from Beyond Meat outside the submitted work (Landry 2023). That researcher also led DIETFITS, which reported no conflicts of interest, and the Beyond Meat funding documented in his later work post-dates it, so DIETFITS is not counted; he also chaired the writing group for the American Heart Association statement cited in the keto section, which placed the vegan pattern below four others. Readers who weigh that tie should weigh it across all three. None of these flags changes a conclusion drawn here.

Which well-known studies were left out on purpose?

The Ornish program, for the design reason given earlier, and the industry-linked 2019 red-meat guideline from the NutriRECS group, which was later found to have had undisclosed funding ties. On the other side of the ledger, the Blue Zones literature was left out because it rests on demography rather than on prospective cohorts, with the exception of Loma Linda, which is included here through its cohort data. The demographic critique and its answers are handled in Saul Newman versus the Blue Zones, a fact-check and are the Blue Zones debunked?, and the broader longevity evidence in what the longest-living people actually eat.

What this evidence cannot tell you

Is nutrition science ever proven?

No science is, in the everyday sense of the word, and this review does not claim proof. What it claims is convergence: randomized trials, large cohorts, food-substitution analyses, and genetic causal evidence all pointing the same way, with a mechanism that explains why. That is as strong as nutrition evidence gets, and it is the same standard by which smoking, hypertension, and LDL itself were accepted as causes of disease.

Why is there no randomized trial of a vegan diet on heart attacks or deaths?

Because such a trial would need thousands of people to stay on an assigned diet for a decade or more, and nobody has funded or run one. The Cochrane review of vegan diets for cardiovascular prevention found no trial at all that reported cardiovascular events; every trial measured blood markers instead (Rees 2021). That is a limitation shared by every diet, including the ones marketed as alternatives, and it is why the LDL mechanism carries so much weight: it is the validated link between what the trials can measure and what the cohorts observe.

Could the cohort results be explained by healthier people choosing plant-based diets?

Partly, and the best studies work hard to rule it out. Vegetarians and vegans in Western cohorts smoke less, move more, weigh less, and earn more, and the studies adjust for all of it. What adjustment cannot do is eliminate residual confounding, and hazard ratios of 0.8 to 0.9 are in the range confounding can produce. Two points limit how far this objection reaches. The Adventist cohort was designed to hold lifestyle constant by comparing people inside one community with shared values, and the gradient appeared anyway. And the motivation research reviewed above suggests that, among the plant-forward groups, the health-optimizer bias runs toward the fish-eaters rather than the vegans. Note also that when cohorts adjust for body weight, the vegetarian heart-disease benefit shrinks by about a fifth (Dybvik 2023), which is sometimes read as weakening the case. It is arguably over-adjustment: lower body weight is one of the ways a plant-based diet does its work, and controlling for it controls away part of the benefit.

Why are the vegan-specific numbers less precise?

Because vegans are rare in every cohort, their membership is unstable, and they are classified less reliably than other groups. In the Adventist cohort, lifetime vegans were 1.1 percent of participants who kept the same diet across their lives, against 31 percent for lifetime lacto-ovo vegetarians; most vegans were recent converts, so a six-year follow-up largely measures the residue of a previous diet (Martins 2017). Measurement error of that kind blurs groups together and pulls estimates toward no effect, which means the true vegan advantage is more likely understated than overstated in these data.

What do the global numbers say about the biggest dietary risks?

The Global Burden of Disease analysis attributed 11 million deaths in 2017 to dietary risks, about a fifth of adult deaths, and found that too little whole grain, fruit, nuts, and vegetables accounted for more deaths than too much red and processed meat (GBD 2017 Diet Collaborators 2019). That is a modeling study and cannot rank diet patterns, but it is a useful corrective: the case for plant-based eating is as much about what people gain as about what they give up. A cross-sectional study of 6,700 Iranian adults found no association between plant-based eating and obesity (Asoudeh 2023), and the two biomarker snapshots in this set show the expected pattern, better insulin and inflammation markers with healthful plant-based eating and worse with unhealthful (Huang 2023), with the animal-protein and growth-hormone picture mixed rather than clean (Lee 2022). Not every study lands the same way, and a review that hides the ones that do not is not a review.

The bottom line

What can be said with confidence?

1. A diet built on healthy plant foods, with animal products reduced or removed, is associated with meaningfully lower risk of coronary heart disease, type 2 diabetes, and premature death, on the order of 15 to 35 percent for the best-versus-worst comparison depending on the outcome.
2. The mechanism is understood and causal: plant-based eating lowers LDL, and LDL drives atherosclerosis in a dose-dependent way with no floor.
3. Replacing animal protein and fat with plant protein and fat improves outcomes, and doing the reverse worsens them, even inside low-carbohydrate diets.
4. There is a gradient. For diabetes and body weight it runs downward from omnivore through pescatarian and vegetarian to vegan, with occasional groups out of order, and no study identifies a point where adding animal food back improves a hard outcome.
5. Nothing in any animal food is nutritionally unique. B12 comes from bacteria, DHA from algae, and iodine from seaweed or salt, and a well-planned vegan diet supplies all of them.

What should be said with care?

1. The vegan-specific mortality estimates are imprecise because vegans are few in every cohort, not because any cohort found them worse off; the plant-forward groups cluster together at lower mortality and cannot be ranked against each other with the data available.
2. The stroke, dementia, and fracture signals come from cohorts where many of the vegans were not supplementing, and they cluster around nutrients that planning supplies. They are evidence for planning, not against plants.
3. On guideline scoring, vegan eating did not rank first. The American Heart Association’s 2023 statement placed it in the second tier, below DASH, Mediterranean, pescatarian, and lacto-ovo vegetarian eating, crediting the foods it is built on while citing how restrictive it can be to follow and the risk of B12 deficiency. That scoring measures alignment with guidance, not outcomes.
4. Well-planned is doing real work in the phrase. Unhealthful plant-based eating is worse than a healthy omnivorous diet in every cohort that measured it.

What cannot be said, because the studies do not exist?

1. That keto, paleo, Atkins, or carnivore diets raise or lower long-term chronic disease risk. What can be said is that they raise the marker that causes heart disease and that no one has ever checked what happens next.
2. That fish adds anything to a well-planned vegan diet. No study has tested that comparison; the fish-eater findings compare against unsupplemented vegetarians and are mostly within statistical noise.

It is worth naming the actual trade on offer, because it is lopsided in a way the fear stories obscure. On one side sits a little extra planning that becomes second nature within a few weeks, one reliable supplement that is genuinely specific to this diet, B12, one more that nearly everyone on any diet should take anyway, vitamin D, and occasionally one or two others depending on a person’s circumstances. On the other side sits a measurably lower risk of the diseases most likely to kill you: heart disease, type 2 diabetes, and several cancers. A fortified plant milk is not a heavier lift than a fortified dairy product, and remembering one small tablet is not a heavier lift than the medication most people end up on for the conditions this way of eating helps avoid. Set honestly against the size of the benefit, the trouble of a well-planned vegan diet is close to a rounding error. And that is before weighing the thing this review deliberately brackets but will not pretend is weightless: that this way of eating asks no animal to suffer or die, and withholds support from one of the most invasive industries on the planet. That belongs to a separate argument, and the efficiency and land-use defenses of that industry are taken apart on their own terms in the “livestock eat food we can’t” argument, counted honestly, but it is a real weight on the same scale.

A word on the objection that the science is not settled, since it is the most common one this review will meet. No question at this scale ever is settled in the sense that critics usually mean by it. Smoking was not settled for decades: people who never smoked developed lung cancer, confounding muddied the early cohorts, and isolated findings even pointed the other way, yet the direction of the evidence was plain long before a warning label ever appeared, and it took thousands of studies to reach that plainness rather than a single decisive one. Nutrition is the same kind of problem, built from imperfect observational cohorts, a smaller number of trials, and genetic evidence that has to be triangulated rather than read off directly. What can be asked of such a field is not certainty but consistency, and across the hierarchy of evidence the consistency here runs one way. “Not fully settled” is not the same as “genuinely a toss-up,” and it is not a reason to default to the pattern the evidence consistently disfavors.

Put plainly: the evidence points at animal products and toward more chronic disease, and at whole plant foods and toward less, with no sweet spot in between where a little animal food turns the direction around. The less animal food a person eats, the better the evidence looks, all the way down to a well-planned vegan diet, which is where the best-measured outcomes in this set sit. Anyone who wants a health reason to keep eating animals will not find one in these 48 studies. What they will find is a reason to take a B12 tablet.

None of this replaces individual advice. Talk to your physician, dietitian, or nutritionist about any change to your own diet, ideally someone trained in lifestyle medicine.

How to read the numbers in this review

What is a hazard ratio?

Most findings here are hazard ratios or relative risks. A hazard ratio of 0.78 means 22 percent lower risk in that group compared with the reference group over the study period. The 95 percent confidence interval is the range the true value probably sits in; if it includes 1.0, the finding could be chance. So a vegan cardiovascular estimate of 0.92 with an interval of 0.79 to 1.06 means: the best guess is 8 percent lower risk, it could plausibly be anywhere from 21 percent lower to 6 percent higher, and chance cannot be ruled out. That single line is why this review says probable rather than proven for vegan-specific mortality, and why it can still say established for the LDL effect, where the intervals sit comfortably away from zero.

References

Links open in a new window. The 48 core studies were read in full, with funding and disclosure statements checked against the published PDF. Supporting sources cited for context were checked against the full text, abstract, or publisher page.

1. Tonstad S, et al. Vegetarian diets and incidence of diabetes in the Adventist Health Study-2. Nutrition, Metabolism and Cardiovascular Diseases. 2013;23(4):292–299. https://doi.org/10.1016/j.numecd.2011.07.004
2. Fung TT, et al. Low-carbohydrate diets and all-cause and cause-specific mortality: two cohort studies. Annals of Internal Medicine. 2010;153(5):289–298. https://doi.org/10.7326/0003-4819-153-5-201009070-00003
3. Seidelmann SB, et al. Dietary carbohydrate intake and mortality: a prospective cohort study and meta-analysis. Lancet Public Health. 2018;3(9):e419–e428. https://doi.org/10.1016/S2468-2667(18)30135-X
4. Retraction Watch. Widely criticized keto diet study retracted. May 22, 2026. https://retractionwatch.com/2026/05/22/widely-criticized-keto-diet-study-retracted/
5. Koch CA, Kjeldsen EW, Frikke-Schmidt R. Vegetarian or vegan diets and blood lipids: a meta-analysis of randomized trials. European Heart Journal. 2023;44(28):2609–2622. https://doi.org/10.1093/eurheartj/ehad211
6. Landry MJ, et al. Cardiometabolic effects of omnivorous vs vegan diets in identical twins: a randomized clinical trial. JAMA Network Open. 2023;6(11):e2344457 (erratum December 2023). https://doi.org/10.1001/jamanetworkopen.2023.44457
7. Dwaraka VB, Aronica L, Carreras-Gallo N, et al. Unveiling the epigenetic impact of vegan vs. omnivorous diets on aging: insights from the Twins Nutrition Study (TwiNS). BMC Medicine. 2024;22:301. https://doi.org/10.1186/s12916-024-03513-w
8. Sofi F, et al. Low-calorie vegetarian versus Mediterranean diets for reducing body weight and improving cardiovascular risk profile: CARDIVEG study. Circulation. 2018;137(11):1103–1113. https://doi.org/10.1161/CIRCULATIONAHA.117.030088
9. Ference BA, et al. Low-density lipoproteins cause atherosclerotic cardiovascular disease: evidence from genetic, epidemiologic, and clinical studies. A consensus statement from the European Atherosclerosis Society Consensus Panel. European Heart Journal. 2017;38(32):2459–2472. https://doi.org/10.1093/eurheartj/ehx144
10. Termannsen AD, et al. Effects of vegan diets on cardiometabolic health: a systematic review and meta-analysis of randomized controlled trials. Obesity Reviews. 2022;23(9):e13462. https://doi.org/10.1111/obr.13462
11. Selinger E, et al. Evidence of a vegan diet for health benefits and risks: an umbrella review of meta-analyses of observational and clinical studies. Critical Reviews in Food Science and Nutrition. 2023;63(29):9926–9936. https://doi.org/10.1080/10408398.2022.2075311
12. Viguiliouk E, et al. Effect of vegetarian dietary patterns on cardiometabolic risk factors in diabetes: a systematic review and meta-analysis of randomized controlled trials. Clinical Nutrition. 2019;38(3):1133–1145. https://doi.org/10.1016/j.clnu.2018.05.032
13. Kahleova H, Petersen KF, Shulman GI, et al. 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. 2020;3(11):e2025454 (Clarification of Role of Funder Statement published 2021). https://doi.org/10.1001/jamanetworkopen.2020.25454
14. Hall KD, et al. Effect of a plant-based, low-fat diet versus an animal-based, ketogenic diet on ad libitum energy intake. Nature Medicine. 2021;27(2):344–353. https://doi.org/10.1038/s41591-020-01209-1
15. Patikorn C, et al. Effects of ketogenic diet on health outcomes: an umbrella review of meta-analyses of randomized clinical trials. BMC Medicine. 2023;21:196. https://doi.org/10.1186/s12916-023-02874-y
16. Hooper L, Martin N, Jimoh OF, et al. Reduction in saturated fat intake for cardiovascular disease. Cochrane Database of Systematic Reviews. 2020;5:CD011737. https://doi.org/10.1002/14651858.CD011737.pub3
17. Reynolds A, Mann J, Cummings J, et al. Carbohydrate quality and human health: a series of systematic reviews and meta-analyses. Lancet. 2019;393(10170):434–445. https://doi.org/10.1016/S0140-6736(18)31809-9
18. Dybvik JS, Svendsen M, Aune D. Vegetarian and vegan diets and the risk of cardiovascular disease, ischemic heart disease and stroke: a systematic review and meta-analysis of prospective cohort studies. European Journal of Nutrition. 2023;62(1):51–69. https://doi.org/10.1007/s00394-022-02942-8
19. Tong TYN, et al. Risks of ischaemic heart disease and stroke in meat eaters, fish eaters, and vegetarians over 18 years of follow-up: results from the prospective EPIC-Oxford study. BMJ. 2019;366:l4897. https://doi.org/10.1136/bmj.l4897
20. Dinu M, et al. Vegetarian, vegan diets and multiple health outcomes: a systematic review with meta-analysis of observational studies. Critical Reviews in Food Science and Nutrition. 2017;57(17):3640–3649. https://doi.org/10.1080/10408398.2016.1138447
21. Capodici A, et al. Cardiovascular health and cancer risk associated with plant based diets: an umbrella review. PLOS ONE. 2024;19(5):e0300711. https://doi.org/10.1371/journal.pone.0300711
22. Orlich MJ, et al. Vegetarian dietary patterns and mortality in Adventist Health Study 2. JAMA Internal Medicine. 2013;173(13):1230–1238. https://doi.org/10.1001/jamainternmed.2013.6473
23. Abris GP, et al. Cause-specific and all-cause mortalities in vegetarian compared with those in nonvegetarian participants from the Adventist Health Study-2 cohort. American Journal of Clinical Nutrition. 2024;120(4):907–917. https://pubmed.ncbi.nlm.nih.gov/39098708/
24. Appleby PN, et al. Mortality in vegetarians and comparable nonvegetarians in the United Kingdom. American Journal of Clinical Nutrition. 2016;103(1):218–230. https://pubmed.ncbi.nlm.nih.gov/26657045/
25. Tonstad S, Butler T, Yan R, Fraser GE. Type of vegetarian diet, body weight, and prevalence of type 2 diabetes. Diabetes Care. 2009;32(5):791–796. https://pubmed.ncbi.nlm.nih.gov/19351712/
26. Satija A, et al. Plant-based dietary patterns and incidence of type 2 diabetes in US men and women: results from three prospective cohort studies. PLOS Medicine. 2016;13(6):e1002039. https://doi.org/10.1371/journal.pmed.1002039
27. Li C, et al. Meat consumption and incident type 2 diabetes: a federated meta-analysis of 1.97 million adults with 100,000 incident cases from 31 cohorts in 20 countries. Lancet Diabetes and Endocrinology. 2024;12(9):619–630. https://doi.org/10.1016/S2213-8587(24)00179-7
28. Tantamango-Bartley Y, et al. Vegetarian diets and the incidence of cancer in a low-risk population. Cancer Epidemiology, Biomarkers and Prevention. 2013;22(2):286–294. https://doi.org/10.1158/1055-9965.EPI-12-1060
29. Kim J, et al. Plant-based dietary patterns and the risk of digestive system cancers in 3 large prospective cohort studies. European Journal of Epidemiology. 2023;38:617–627. https://doi.org/10.1007/s10654-023-01007-2
30. Yun Z, et al. Processed meat, red meat, white meat, and digestive tract cancers: a two-sample Mendelian randomization study. Frontiers in Nutrition. 2023;10:1078963. https://doi.org/10.3389/fnut.2023.1078963
31. International Agency for Research on Cancer. IARC Monographs evaluate consumption of red meat and processed meat. Press release 240. 2015. https://www.iarc.who.int/wp-content/uploads/2018/07/pr240_E.pdf
32. Fraser GE, et al. Dairy, soy, and risk of breast cancer: those confounded milks. International Journal of Epidemiology. 2020;49(5):1526–1537. https://pubmed.ncbi.nlm.nih.gov/32095830/
33. Satija A, et al. Healthful and unhealthful plant-based diets and the risk of coronary heart disease in US adults. Journal of the American College of Cardiology. 2017;70(4):411–422. https://doi.org/10.1016/j.jacc.2017.05.047
34. Thompson AS, et al. Association of healthful plant-based diet adherence with risk of mortality and major chronic diseases among adults in the UK. JAMA Network Open. 2023;6(3):e234714. https://doi.org/10.1001/jamanetworkopen.2023.4714
35. Wang P, et al. Optimal dietary patterns for prevention of chronic disease. Nature Medicine. 2023. https://doi.org/10.1038/s41591-023-02235-5
36. Chen H, et al. Plant-based dietary patterns in relation to mortality among older adults in China. Nature Aging. 2022. https://doi.org/10.1038/s43587-022-00180-5
37. Kim H, Caulfield LE, Rebholz CM. Healthy plant-based diets are associated with lower risk of all-cause mortality in US adults. Journal of Nutrition. 2018;148(4):624–631. https://doi.org/10.1093/jn/nxy019
38. Shan Z, et al. Association of low-carbohydrate and low-fat diets with mortality among US adults. JAMA Internal Medicine. 2020;180(4):513–523. https://doi.org/10.1001/jamainternmed.2019.6980
39. Gao JW, Hao QY, Zhang HF, et al. Low-carbohydrate diet score and coronary artery calcium progression: results from the CARDIA study. Arterioscler Thromb Vasc Biol. 2021;41(1):491-500. https://pubmed.ncbi.nlm.nih.gov/33115269/
40. Naghshi S, Sadeghi O, Willett WC, Esmaillzadeh A. Dietary intake of total, animal, and plant proteins and risk of all cause, cardiovascular, and cancer mortality: systematic review and dose-response meta-analysis of prospective cohort studies. BMJ. 2020;370:m2412. https://doi.org/10.1136/bmj.m2412
41. Song M, et al. Association of animal and plant protein intake with all-cause and cause-specific mortality. JAMA Internal Medicine. 2016;176(10):1453–1463. https://doi.org/10.1001/jamainternmed.2016.4182
42. Jayedi A, et al. Intake of animal and plant proteins and risk of all-cause mortality in patients with type 2 diabetes: results from NHANES. European Journal of Clinical Nutrition. 2025;79:214–223. https://doi.org/10.1038/s41430-024-01535-4
43. Huang J, et al. Association between plant and animal protein intake and overall and cause-specific mortality. JAMA Internal Medicine. 2020;180(9):1173–1184. https://doi.org/10.1001/jamainternmed.2020.2790
44. Neuenschwander M, et al. Substitution of animal-based with plant-based foods on cardiometabolic health and all-cause mortality: a systematic review and meta-analysis of prospective studies. BMC Medicine. 2023;21:404. https://doi.org/10.1186/s12916-023-03093-1
45. Zhong VW, et al. Associations of processed meat, unprocessed red meat, poultry, or fish intake with incident cardiovascular disease and all-cause mortality. JAMA Internal Medicine. 2020;180(4):503–512. https://doi.org/10.1001/jamainternmed.2019.6969
46. Koeth RA, et al. Intestinal microbiota metabolism of L-carnitine, a nutrient in red meat, promotes atherosclerosis. Nature Medicine. 2013;19(5):576–585. https://doi.org/10.1038/nm.3145
47. Bergeron N, Chiu S, Williams PT, King SM, Krauss RM. Effects of red meat, white meat, and nonmeat protein sources on atherogenic lipoprotein measures: the APPROACH trial. American Journal of Clinical Nutrition. 2019;110(1):24–33. https://www.sciencedirect.com/science/article/pii/S0002916522011492
48. Bergholdt HKM, et al. Lactase persistence, milk intake, and mortality in the Danish general population: a Mendelian randomization study. European Journal of Epidemiology. 2018;33(2):171–181. https://doi.org/10.1007/s10654-017-0328-x
49. Petermann-Rocha F, et al. Vegetarians, fish, poultry, and meat-eaters: who has higher risk of cardiovascular disease incidence and mortality? A prospective study from UK Biobank. European Heart Journal. 2021;42(12):1136–1143. https://doi.org/10.1093/eurheartj/ehaa939
50. Yassine HN, et al. CNS target engagement of high-dose DHA supplementation in older adults at risk for dementia: a randomized clinical trial. eBioMedicine. 2026;106316 (NCT03613844). https://doi.org/10.1016/j.ebiom.2026.106316
51. Rosenfeld DL. The psychology of vegetarianism: recent advances and future directions. Appetite. 2018;131:125–138. https://pubmed.ncbi.nlm.nih.gov/30227184/
52. Rosenfeld DL, Tomiyama AJ. How proximal are pescatarians to vegetarians? An investigation of dietary identity, motivation, and attitudes toward animals. Journal of Health Psychology. 2021;26(5):713–727. https://pubmed.ncbi.nlm.nih.gov/30969133/
53. Hopwood CJ, et al. Health, environmental, and animal rights motives for vegetarian eating. PLOS ONE. 2020;15(4):e0230609. https://doi.org/10.1371/journal.pone.0230609
54. Gilsing AMJ, et al. Serum concentrations of vitamin B12 and folate in British male omnivores, vegetarians and vegans: results from a cross-sectional analysis of the EPIC-Oxford cohort study. European Journal of Clinical Nutrition. 2010;64(9):933–939. https://pubmed.ncbi.nlm.nih.gov/20648045/
55. Chiu THT, et al. Vegetarian diet and incidence of total, ischemic, and hemorrhagic stroke in two cohorts in Taiwan. Neurology. 2020. https://pubmed.ncbi.nlm.nih.gov/32102976/
56. Damayanti D, et al. Foods and supplements associated with vitamin B12 biomarkers among vegetarian and non-vegetarian participants of the Adventist Health Study-2 calibration study. Nutrients. 2018;10(6):722. https://pubmed.ncbi.nlm.nih.gov/29867046/
57. Tong TYN, et al. Vegetarian and vegan diets and risks of total and site-specific fractures: results from the prospective EPIC-Oxford study. BMC Medicine. 2020;18:353. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7682057/
58. Ballarin RS, et al. Vegetarian and vegan diets and the risk of hip fracture: a systematic review and meta-analysis of cohort studies. Nutrition Reviews. 2025;nuaf241. https://pubmed.ncbi.nlm.nih.gov/41275510/
59. Appleby P, Roddam A, Allen N, Key T. Comparative fracture risk in vegetarians and nonvegetarians in EPIC-Oxford. European Journal of Clinical Nutrition. 2007;61(12):1400-1406. https://pubmed.ncbi.nlm.nih.gov/17299475/
60. Thorpe DL, et al. Dietary patterns and hip fracture in the Adventist Health Study 2: combined vitamin D and calcium supplementation mitigate increased hip fracture risk among vegans. American Journal of Clinical Nutrition. 2021;114(2):488-495. https://pubmed.ncbi.nlm.nih.gov/33964850/
61. Wakolbinger-Habel R, et al. Self-reported resistance training is associated with better HR-pQCT-derived bone microarchitecture in vegan people. Journal of Clinical Endocrinology and Metabolism. 2022;107(10):2900-2911. https://pubmed.ncbi.nlm.nih.gov/35924941/
62. Sotos-Prieto M, et al. Plant-based diets and risk of hip fracture in postmenopausal women. JAMA Network Open. 2024;7(2):e241107. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10905300/
63. Azzola LG, Fankhauser N, Srinivasan M. Influence of the vegan, vegetarian and omnivore diet on the oral health status in adults: a systematic review and meta-analysis. Evidence-Based Dentistry. 2023. https://doi.org/10.1038/s41432-023-00853-z
64. Seimetz LG, Zanella PB. Effects of omnivorous and vegetarian diets in neuromuscular adaptations to physical exercise: a systematic review. Science and Sports. 2023. https://doi.org/10.1016/j.scispo.2022.07.017
65. Hevia-Larrain V, et al. High-protein plant-based diet versus a protein-matched omnivorous diet to support resistance training adaptations: a comparison between habitual vegans and omnivores. Sports Medicine. 2021;51(6):1317–1330. https://doi.org/10.1007/s40279-021-01434-9
66. Lennerz BS, Mey JT, Henn OH, Ludwig DS. Behavioral characteristics and self-reported health status among 2029 adults consuming a carnivore diet. Current Developments in Nutrition. 2021;5(12):nzab133. https://doi.org/10.1093/cdn/nzab133
67. Gardner CD, Vadiveloo MK, Petersen KS, et al. Popular dietary patterns: alignment with American Heart Association 2021 dietary guidance: a scientific statement from the American Heart Association. Circulation. 2023;147(22):1715-1730. https://pubmed.ncbi.nlm.nih.gov/37128940/
68. American Heart Association News. Here’s how 10 popular diets scored for heart health. April 27, 2023. https://www.heart.org/en/news/2023/04/27/heres-how-10-popular-diets-scored-for-heart-health
69. The Lancet. Department of Error (GBD 2017 Diet Collaborators author list). Lancet. 2021;397(10293):2466. https://pubmed.ncbi.nlm.nih.gov/34175088/
70. Ghaedi E, et al. Effects of a Paleolithic diet on cardiovascular disease risk factors: a systematic review and meta-analysis of randomized controlled trials. Advances in Nutrition. 2019;10(4):634–646 (with 2020 corrigendum). https://doi.org/10.1093/advances/nmz007
71. Gardner CD, et al. Effect of low-fat vs low-carbohydrate diet on 12-month weight loss in overweight adults and the association with genotype pattern or insulin secretion: the DIETFITS randomized clinical trial. JAMA. 2018;319(7):667–679 (corrections April 3 and April 24, 2018). https://doi.org/10.1001/jama.2018.0245
72. Retraction notice for the KETO-CTA lean mass hyper-responder study. JACC: Advances. 2026. https://www.jacc.org/doi/10.1016/j.jacadv.2025.101686
73. Yaseen NR, et al. Genetics of vegetarianism: a genome-wide association study. PLOS ONE. 2023;18(10):e0291305. https://doi.org/10.1371/journal.pone.0291305
74. Rees K, Al-Khudairy L, Takeda A, Stranges S. Vegan dietary pattern for the primary and secondary prevention of cardiovascular diseases. Cochrane Database of Systematic Reviews. 2021;2:CD013501. https://doi.org/10.1002/14651858.CD013501.pub2
75. Martins MCT, et al. A new approach to assess lifetime dietary patterns finds lower consumption of animal foods with aging in a longitudinal analysis of a health-oriented Adventist population. Nutrients. 2017;9(10):1118. https://www.mdpi.com/2072-6643/9/10/1118
76. GBD 2017 Diet Collaborators. Health effects of dietary risks in 195 countries, 1990–2017: a systematic analysis for the Global Burden of Disease Study 2017. Lancet. 2019;393(10184):1958–1972. https://doi.org/10.1016/S0140-6736(19)30041-8
77. Asoudeh F, et al. The association of plant-based dietary pattern with general and abdominal obesity: a large cross-sectional study. Journal of Diabetes and Metabolic Disorders. 2023;22:469–477. https://doi.org/10.1007/s40200-022-01166-1
78. Huang Y, et al. Associations of healthful and unhealthful plant-based diets with plasma markers of cardiometabolic risk. European Journal of Nutrition. 2023;62:2567–2579. https://doi.org/10.1007/s00394-023-03170-4
79. Lee DH, Tabung FK, Giovannucci EL. Association of animal and plant protein intakes with biomarkers of insulin and insulin-like growth factor axis. Clinical Nutrition. 2022;41:1272–1280. https://doi.org/10.1016/j.clnu.2022.04.003

Last updated: September 28, 2026, 10:32 PM ET