A deep dive into the science of diet and longevity, climbing the hierarchy of evidence from population patterns to identical-twin trials, and why every rung points the same direction.

What do the longest-living people actually eat?

Ask people what the world’s longest-living humans eat and you’ll get confident answers pointing in opposite directions. Some will tell you it’s fish. Some will swear it’s pork, citing Hong Kong. Others will say it doesn’t matter at all: it’s genes, or wine, or walking.

Does cholesterol actually cause heart disease?

The actual answer is more interesting, and more consistent, than the debate suggests. But to get there, we have to be honest about how we know what we know. Nutrition science gets criticized, sometimes fairly, for leaning on observational studies that can’t prove cause and effect. So this post is organized the way evidence-based medicine itself is: as a climb up the hierarchy of evidence. We’ll start on the bottom rung, population-level patterns and their famous failure modes, and ascend through prospective cohorts and meta-analyses to the tools that dodge confounding by design: randomized controlled trials, identical-twin experiments, and Mendelian randomization. The case will rest its weight only on rungs built to hold it. (A note for readers who’ve followed the recent controversies over Blue Zones demography: so have we. The Blue Zones appear near the end of this post, as illustration rather than foundation, nothing here stands on them.) When every independent rung points the same direction, that convergence is how science builds confidence about anything it can’t test in an 80-year randomized trial.

Spoiler: the direction is plants. The question worth exploring is how far that direction goes.

The ladder: how do you even study what makes people live to 100?

How do scientists study diet and longevity?

No one can randomize newborns to eat one way for a lifetime and count who’s still standing at 100. So researchers climb a ladder of evidence, and each rung has known strengths and known failure modes.

Rung one: ecological studies compare whole populations, country A eats X and lives to Y. They’re cheap, vivid, and the weakest rung on the ladder, because populations differ in a thousand ways besides diet. Drawing individual-level conclusions from group-level data even has a name: the ecological fallacy. Keep that term handy; we’ll need it shortly, and note that it disqualifies rung-one arguments no matter which side they seem to favor.

What is a prospective cohort study?

Rung two: prospective cohort studies: the workhorse of nutritional epidemiology, follow tens of thousands of individuals for decades, record what they eat, and track real endpoints: heart attacks, cancers, deaths. Studies like the Adventist Health Studies, EPIC-Oxford, the Nurses’ Health Study, and the Health Professionals Follow-up Study have collectively tracked millions of person-years [1–5]. Their strengths are enormous sample sizes, hard outcomes, and long time horizons. Their weaknesses are just as well documented: confounding (health-conscious people differ in many ways beyond diet), the healthy-user bias, measurement error in food-frequency questionnaires, and reverse causation (sick people sometimes change their diets). Good studies adjust statistically for smoking, exercise, weight, education, alcohol, and more, but adjustment is never perfect. Systematic reviews and meta-analyses pool these cohorts into steadier estimates, inheriting both their power and their blind spots.

What is the hierarchy of evidence in nutrition science?

The top rungs are the tools that dodge confounding by design: randomized controlled trials that change diets and watch disease itself, identical-twin experiments that hold genetics constant, and Mendelian randomization studies that use gene variants, dealt at conception, immune to lifestyle, as nature’s own randomizer. And the capstone is triangulation: does the association show a dose-response gradient? Does it replicate across populations with different confounding structures? Do the trials, the twins, and the genes move in the predicted direction? Do the biological mechanisms make sense? When the answer to all of those is yes, the case becomes hard to dismiss: this is essentially the framework epidemiologists have used since Austin Bradford Hill laid out his causal criteria in 1965, and it’s how we became confident smoking causes lung cancer without ever running a smoking RCT [6].

Hold onto that ladder. The rest of this post climbs it in order.

A tale of two questions: average lifespan vs. exceptional longevity

Doesn’t Hong Kong eat the most meat and live the longest?

Here’s where a popular counterargument usually enters the chat: “Hong Kong has the longest life expectancy in the world and eats enormous amounts of pork and meat. Checkmate, plant people.”

It’s a genuinely useful objection, because unpacking it teaches us exactly which questions population data can and can’t answer.

Average life expectancy at birth is a whole-society statistic. It is powerfully shaped by infant and child mortality, healthcare access, wealth, sanitation, road safety, and, above all, smoking rates. When researchers at the University of Hong Kong analyzed 263 million deaths across 18 high-income regions to figure out why Hong Kong tops the tables, their answer, published in The Lancet Public Health, wasn’t pork. It was the rare combination of rapid economic development with world-leading tobacco control, producing, in the authors’ words, “fewer diseases of poverty while suppressing the diseases of affluence” [7]. Hong Kong has pushed daily smoking below 9%, 8.5% as of 2024, among the lowest rates of any developed economy, and smoking alone accounted for roughly a third to a half of Hong Kong’s longevity edge over other rich regions in the study’s models [7, 8]. Add a public hospital system that’s essentially free at the point of care, among the world’s lowest infant mortality rates, a dense walkable city where daily movement is built into life, and strong family integration of elders, and you’ve explained the ranking without ever mentioning a barbecue pork bun [7, 9].

Why life expectancy doesn’t tell you what a country eats

There’s also a cohort effect hiding in plain sight. Life expectancy tables are pulled upward by today’s oldest residents, people in their 80s and 90s who spent their childhoods and young adulthoods in an era of postwar scarcity, eating predominantly rice and vegetables with meat as a garnish. Hong Kong’s famously high per-capita meat consumption is a phenomenon of recent decades of affluence. In other words, the generation responsible for the longevity statistics did not eat the modern Hong Kong diet for most of their lives, and the generations who have are the ones now driving Hong Kong’s substantial burden of colorectal cancer, which ranks among the city’s most common cancers, consistent with everything we know about processed and red meat [10]. Tellingly, Hong Kong’s own public health researchers point out that the city’s gains in lifespan have not been matched by gains in healthspan, years lived in good health, leaving a growing burden of late-life disability [9].

So the pork argument commits the ecological fallacy twice over: it attributes a whole-society statistic to one food, and it attributes today’s diet to yesterday’s survivors. Average life expectancy is a referendum on a society’s healthcare, wealth, and cigarettes far more than on its dinner plates.

And intellectual honesty demands that this blade cut both ways. If country-level pork statistics can’t indict meat, then region-level bean statistics can’t acquit it, which is why this post refuses to build on the most famous pro-plant geography story of all, the Blue Zones. They’re real places with a real dietary pattern, and we’ll visit them near the end as corroborating color; but rung one can’t bear weight in either direction, no matter whose argument it happens to flatter. The evidence that can bear weight starts one rung up: following individuals.

Yes, longevity has many parents, let’s name them all

Before going deeper on food, let’s be upfront about everything else that shapes a long life, because pretending diet is the only variable would be exactly the kind of sloppy reasoning we just criticized.

How much of lifespan is genetic?

Genetics matters, but far less than most people assume. The famous Danish twin studies estimated that only about 20–25% of the variation in human lifespan is heritable [16], and a 2018 analysis of 400 million historical family records concluded that true heritability may be well under 10% once you account for the fact that people tend to marry others with similar lifestyles [17]. Not smoking is enormous: it’s most of Hong Kong’s secret. Sleep, stress management, and physical activity all independently predict mortality. Socioeconomic status and access to healthcare shape everything from infant survival to cancer detection. Avoiding excess alcohol, drugs, and environmental toxins shows up in every longevity population.

What does the Harvard longevity study say actually matters?

And then there’s the one most longevity content skips, which may be the largest of all: relationships. The Harvard Study of Adult Development has tracked the same men since 1938, originally 724 of them, split between Harvard sophomores and boys from Boston’s poorest tenement neighborhoods, following them through work, marriage, illness, and old age for more than eight decades, with the study now extending into their children [105]. Its current director, psychiatrist Robert Waldinger, summarizes eighty-plus years of data in one sentence: good relationships keep us happier and healthier. Not wealth, not fame, not achievement. The study’s most arresting finding is a prediction test: gathering everything they knew about the men at age 50, the variable that best forecast who would be thriving at 80 wasn’t their midlife cholesterol: it was how satisfied they were in their relationships [105]. Loneliness, by contrast, behaves like a clinical risk factor; pooling 148 studies and more than 300,000 people, Holt-Lunstad’s team found strong social ties carried an odds ratio for survival of 1.50, a 50% greater likelihood of being alive at follow-up, an effect they rank alongside quitting smoking and ahead of obesity or inactivity [106]. And the quality matters more than the count, high-conflict marriages predict worse health outcomes than divorce, while feeling you can genuinely count on someone in your 80s tracks with sharper memory for longer [105].

Two things follow, and they cut in opposite directions from what you might expect. The first is that anyone building a longevity plan around supplements and biomarkers while neglecting the people in their life has the priorities backwards. The second is a caution about reading that cholesterol finding too fast: a single cohort of 724 men showing relationship satisfaction outpredicting cholesterol does not overturn the genetic and trial evidence that LDL causes heart disease: those are different questions on different rungs. Both are true. Relationships are a massive and underrated lever; LDL still clogs arteries.

Notice, too, that the world’s celebrated longevity hotspots, which we’ll visit near the end, are saturated with exactly these features: natural daily movement, strong community and purpose, low smoking, low-stress cultures [12]. So is Loma Linda. Which is precisely what makes the Adventist cohort such a useful natural experiment: it holds community, faith, purpose, and low smoking roughly constant across its members, then lets diet vary from meat-eating to vegan. The dietary gradient we’re about to examine appears on top of strong social connection, not instead of it.

Is diet really the biggest factor in longevity?

So no, diet is not the whole story, and this post has no interest in pretending otherwise. But here is why it deserves the spotlight rather than a footnote. First, it’s the single largest modifiable risk factor for death on Earth: the Global Burden of Disease collaborators pinned 11 million deaths in 2017 on dietary risks, about one in five deaths worldwide, more than tobacco [18]. Zoom into the United States and the picture sharpens further: Micha and colleagues traced 318,656 cardiometabolic deaths in a single year, 45.4% of all US deaths from heart disease, stroke, and type 2 diabetes, to suboptimal intake of just ten foods and nutrients, too much processed meat, red meat, and sodium; too few vegetables, fruits, nuts, and whole grains [74]. Second, unlike your genes, your childhood socioeconomic status, or your luck, what goes on your plate is under your direct control three times a day, starting today. And third, critically for this post, diet is the variable where the research lets us see a clean dose-response gradient. Which brings us back to the food.

The gradient: every step toward plants, measured

Do vegans have better health outcomes than meat eaters?

If plant-predominant eating drives the longevity signal, we’d predict something specific: within a single population, health outcomes should improve stepwise as diets move from meat-heavy to fully plant-based. Not a binary “vegetarians vs. everyone,” but a staircase. That is exactly what the best cohorts show.

What is the Adventist Health Study?

The Adventist Health Study-2 (AHS-2) is the crown jewel of rung two. It enrolled about 96,000 Seventh-day Adventists across North America, a famously long-lived community, centered on Loma Linda, California, whose members outlive their fellow Americans by up to a decade [15], where almost nobody smokes, almost nobody drinks, and everyone shares similar religious culture, community support, and health consciousness, yet whose members happen to span the full dietary spectrum from regular meat-eaters to vegans [1]. That internal variation is a gift to science: comparing vegan Adventists to meat-eating Adventists controls for the lifestyle confounders that plague comparisons between, say, vegans and the general public. It’s about as close to isolating the diet variable as observational research gets, and it’s the single most-argued-over dataset in nutrition, which is why it has its own companion piece: What the Adventist Health Studies Actually Tell Us About Vegan Diets, And What They Don’t.

The staircase appears on essentially every outcome measured. Average BMI climbs in orderly steps from 23.6 in vegans to 25.7 in lacto-ovo vegetarians to 26.3 in pescatarians to 28.8 in nonvegetarians [19]. Type 2 diabetes prevalence runs 2.9% in vegans versus 7.6% in nonvegetarians, with lacto-ovo, pesco, and semi-vegetarians falling neatly in between [19]. A separate, more rigorous analysis of the same cohort, measuring new cases rather than existing ones, tells the story with a sharper edge: over two years, incident diabetes ran 2.12% in nonvegetarians, 1.29% in pescatarians, 1.08% in lacto-ovo vegetarians, and 0.54% in vegans, an adjusted odds ratio of 0.38 for vegans against nonvegetarians [111]. Even here the ladder isn’t perfectly monotonic, semi-vegetarians land between lacto-ovo and vegan rather than below both, and the authors note the association strengthens once body weight is removed from the model, the same over-adjustment wrinkle we’ll meet again with the 2024 mortality update. Hypertension and high cholesterol show the same gradient [20]. Vegan Adventists had the lowest overall cancer incidence of any diet group, including meaningfully lower rates of female-specific cancers [21]. And on the outcome that matters most, the landmark 2013 mortality analysis in JAMA Internal Medicine found all vegetarian patterns combined had 12% lower all-cause mortality than nonvegetarians (HR 0.88), with pescatarians at 0.81, vegans at 0.85, lacto-ovo vegetarians at 0.91, and semi-vegetarians at 0.92, and vegan men doing best of all at 0.72 [1]. The 2024 update, following the cohort through 2015, with 12,515 deaths, sharpened rather than simply confirmed this picture, and honesty requires reporting what it found. In the minimally adjusted model, vegans still died at about 0.88 times the rate of their meat-eating co-religionists; but once the researchers layered in BMI, exercise, and prevalent disease, the all-vegan advantage shrank toward statistical noise, a hazard ratio around 0.97 to 0.99 [22]. The clearest surviving signal was in men: vegan men at 65 carried a hazard ratio of 0.72, yet by 85 that edge had vanished, the point estimate drifting just above 1.0 [22]. Two things are worth saying plainly about this. First, adjusting for BMI is arguably over-adjustment, since a lower body weight is one of the mechanisms by which a vegan diet extends life, controlling it away partly controls away the diet’s own benefit. Second, the authors flag a plausible culprit for higher stroke and neurodegenerative signals in some vegans: under-supplementation of B12 and omega-3s, not the plants themselves [22]. This is the honest shape of the data, a real advantage, strongest in midlife men, sensitive to how much you adjust for diet’s own downstream effects, and contingent on doing veganism competently. (We’ll dissect the vegan-vs-pescatarian nuance shortly: it’s important, and it cuts differently than people assume.)

Do vegetarians have a higher risk of stroke?

EPIC-Oxford, the UK’s counterpart cohort of about 48,000 meat-eaters, fish-eaters, and vegetarians/vegans followed for 18 years, found vegetarians and vegans had a 22% lower rate of ischemic heart disease than meat-eaters, translating to about 10 fewer cases of heart disease per 1,000 people per decade [2]. That echoes an earlier pooled analysis of five prospective studies finding 24% lower ischemic heart disease mortality in vegetarians [23]. EPIC-Oxford also handed us a finding any honest write-up must include: vegetarians showed a higher rate of hemorrhagic stroke, about 3 extra cases per 1,000 people per decade, against the 10 fewer heart disease cases [2]. Researchers suspect very low B12 status (this cohort largely predates routine supplementation) and possibly very low cholesterol in a subset may play roles; the paper itself notes the vegetarians’ fiber and saturated fat intakes weren’t even that different from the meat-eaters’, and the net cardiovascular math still favored the plant-based groups [2, 24]. The takeaway isn’t “eat meat”: it’s “supplement B12 and eat enough of the right plants,” which modern evidence-based vegan guidance already insists on. We’ll see this theme again.

What is the plant-based diet index?

The Harvard cohorts, Nurses’ Health Study and Health Professionals Follow-up Study, together tracking over 200,000 people, added a crucial refinement: the plant-based diet index, which scores everyone (not just self-identified vegetarians) on how plant-predominant and how healthfully plant-predominant their diet is. Higher healthful plant-based scores predicted 25% lower coronary heart disease and 34% lower type 2 diabetes; higher unhealthful plant scores (refined grains, fries, soda) predicted increased risk [3, 4]. This kills two birds: it shows the plant gradient holds in mainstream American populations, and it shows the claim was never “anything without meat is magic”: it’s whole plant foods doing the work. The same split shows up inside a completely different macronutrient argument: among low-carbohydrate dieters specifically, the version built on animal foods tracked with 23% higher all-cause mortality, while the version built on plant foods tracked with 20% lower, same carb percentage, opposite outcome, because the source of the calories was the variable that actually mattered [128, 129].

Does red meat increase mortality risk?

And the mirror image holds for animal foods. In the same Harvard cohorts, each additional daily serving of unprocessed red meat was associated with 13% higher mortality, and processed meat 20% higher [5]; people who increased red meat intake over an eight-year span had elevated death rates in the years that followed [25]; and statistically swapping a few percent of calories from animal protein to plant protein predicted lower mortality [26]. A pooled analysis of 32 cohorts and 715,000 people found plant protein intake associated with 8% lower all-cause and 12% lower cardiovascular mortality, with each additional 3% of calories from plant protein tied to 5% lower risk of death, while total and animal protein showed no such benefit [130]. The same cohorts tie red meat to sharply higher type 2 diabetes risk, with processed meat worst of all [76]; a meta-analysis in Circulation found each daily serving of processed meat associated with 42% higher coronary heart disease and 19% higher diabetes incidence [75]; and Europe’s own EPIC cohort of nearly half a million people replicated the mortality signal, estimating that over 3% of all deaths could be prevented if processed meat intake fell below 20 grams a day [77]. Even within AHS-2’s health-conscious, low-meat population, unprocessed red meat intake was associated with 18% higher all-cause mortality [27]. In 2015, the WHO’s International Agency for Research on Cancer, after reviewing over 800 studies, classified processed meat as a Group 1 carcinogen (the same certainty category as tobacco, though not the same magnitude) and red meat as Group 2A, probably carcinogenic [28].

Are dairy and eggs linked to cancer?

And it isn’t only meat: the evidence reaches dairy and eggs, which is what makes this a vegan question rather than merely a vegetarian one. A meta-analysis of cohort studies in the American Journal of Clinical Nutrition found higher intakes of total dairy, milk, cheese, and dairy-derived calcium each associated with increased prostate cancer risk [78], a link the World Cancer Research Fund independently grades as limited-but-suggestive [54]. Within AHS-2 itself, the community’s parallel milk habits set up a natural comparison with a striking result: the highest dairy-milk drinkers carried about a 50% higher breast-cancer hazard than the lowest, while soy milk showed no such association [81]. For eggs, a dose-response meta-analysis of prospective studies found no link with total prostate cancer but a 47% higher risk of fatal prostate cancer per five eggs a week [80], echoing the Harvard physicians’ cohort in which men eating 2.5 or more eggs weekly had an 81% higher risk of lethal prostate cancer than those eating fewer than half an egg [79], plausibly via choline, the same TMAO precursor we’ll meet again in the mechanisms section. None of these signals is as airtight as the processed-meat literature, and honest writing should say so; but they all lean the same direction, and they’re part of why the fully plant-exclusive pattern, not just the meatless one, keeps earning the top of the staircase.

Which foods are most strongly linked to a longer life?

Now flip to the affirmative half of the ledger, because every food on the winning side has a literature of its own. Legumes may be the single most consistent dietary predictor of survival ever measured: a multi-country study of older adults across four cultures found roughly 7–8% lower mortality risk for every 20-gram daily increment, in every population studied [89]. Daily nut eaters ran about 20% lower total mortality across three decades of Harvard follow-up [90], and adding nuts or extra-virgin olive oil to a plant-forward Mediterranean pattern cut cardiovascular events by roughly 30% in the randomized PREDIMED trial, a study worth citing carefully, because critics raise its history first: the 2013 original was retracted and republished in 2018 after a statistician showed roughly a fifth of participants hadn’t been properly randomized, yet the reanalysis reran the numbers honestly and the ~30% reduction held [99]. A trial that survives its own retraction is stronger evidence, not weaker. Whole grains show a clean dose-response, with about 90 grams a day associated with roughly one-fifth lower cardiovascular risk [91]; fruit and vegetable benefits scale all the way up to ten servings a day, an intake estimated to avert nearly eight million premature deaths worldwide each year [92]; and the fiber threading through all of them, per a Lancet analysis commissioned for WHO guidelines, delivers 15–30% reductions in all-cause mortality, heart disease, type 2 diabetes, and colorectal cancer at higher intakes [93]. Daily leafy-green eaters declined slowly enough that, on the study’s own cognitive slope, they tested like people some eleven years younger, an observational association, not a fountain of youth, but a large one [94]. Even the quieter corners of the plant kingdom hold up: a meta-analysis spanning over 600,000 people tied mushroom consumption to lower all-cause mortality [95]; in Japan’s large cohorts, daily seaweed eaters had substantially lower ischemic heart disease rates [96]; and a randomized controlled-feeding trial found that seasoning an ordinary American diet with about 1.3 teaspoons of mixed herbs and spices a day modestly lowered 24-hour blood pressure [97]. When researchers bundled plant components into a single head-to-head intervention: the “portfolio” diet of nuts, soy protein, viscous fiber, and plant sterols: it lowered LDL cholesterol nearly as much as a starting dose of statin in a randomized JAMA trial [98]. There is no counterpart literature on the other side: no cohort or trial showing red meat, processed meat, or eggs add protection that a whole-plant pattern lacks: their best showings are null results, and their worst are cataloged above. Dairy’s one genuine bright spot, a probable reduction in colorectal cancer risk largely attributed to calcium [54], is achievable from plant calcium sources without accepting the prostate and breast signals we just covered.

Step back and look at the shape of all this evidence. More plants, better outcomes; more animal foods, worse outcomes; the middle positions land in the middle; and the pattern replicates in California Adventists, British vegetarians, and Boston nurses alike, populations with completely different confounding structures. Pooled meta-analyses agree: vegetarians run roughly 25% lower ischemic heart disease incidence and mortality, and vegans about 15% lower total cancer incidence [82]. And the top of the hierarchy has only firmed up since: a 2024 umbrella review commissioned by the Academy of Nutrition and Dietetics pooled 21 systematic reviews and tied vegetarian and vegan patterns to lower cardiovascular disease incidence and mortality, while a second 2024 umbrella review in PLOS ONE, spanning dozens of meta-analyses, reached the same verdict on lipids, glycemic control, and cancer risk [109, 110]. Cochrane’s own dedicated review of vegan diets for cardiovascular prevention, the methodologically strictest tier of evidence synthesis there is, lands in the same place [127]. And the global INTERHEART study, spanning 52 countries, found that a short list of modifiable factors, daily fruit and vegetable intake protective among them, accounted for roughly 90% of first-heart-attack risk on every continent studied [72]. A monotonic dose-response gradient replicated across cohorts is one of Bradford Hill’s strongest signatures of causation [6]. So the natural question becomes: what happens at the top of the staircase?

Where do vegans fit? Reading the evidence like a scientist

Is veganism a diet or an ethical position?

Before reading the vegan data, we have to fix a category error that quietly distorts it. Veganism is not, at its root, a dietary pattern. As the movement has defined it since 1944, it’s an ethical commitment to avoiding animal exploitation “as far as is possible and practicable”, in food, clothing, and everything else [83]. The diet is a consequence of the ethic, not the point of it. Epidemiology, though, needs diet categories, so it borrows the word “vegan” to label everyone who excludes animal products, regardless of what they eat instead. Every other diet label in these studies describes what people eat; vegan alone describes what people refuse. Notice the measurement mismatch that creates: “pescatarian” and “Mediterranean” describe patterns people typically adopt for their health, so those labels arrive pre-bundled with health-seeking food choices. “Vegan” describes an exclusion people most often adopt for the animals, bundled with whatever plant foods they happen to like. Oreos, white bread, and french fries are all vegan.

Now hold that beside what nutrition science actually crowns. Ask which foods carry the strongest evidence for preventing chronic disease and extending healthy life, and the answer is strikingly consistent across the plant-based diet indices, the position papers, the cancer reports, and the global commissions: legumes, leafy greens and other vegetables, whole grains, fruits, mushrooms, nuts, seeds, seaweeds, herbs, and spices [3, 4, 13, 48, 54, 67]. Every single item on that list is a vegan food. The vegan pantry contains the entire evidence-backed roster, but nothing about the ethic guarantees an individual vegan eats from it. And in the early decades of any social movement, ethics recruits before health does. The vegans captured by EPIC-Oxford (enrolled 1993–2001) and AHS-2 (enrolled 2002–2007) largely predate mainstream whole-food plant-based medicine, universal B12 guidance, and even decent vegan options at the grocery store [1, 2]. This is why researchers studying “vegans” have to work so hard to control for diet quality, and why, as we’re about to see, the older cohorts mostly couldn’t.

Why is there so little research on vegans?

Here’s the awkward truth the epidemiology has to work with: there is no vegan country. No nation, no large population anywhere, has eaten fully plant-based for generations, so vegans enter these cohorts as small subgroups, roughly 5,500 in AHS-2, about 2,600 in EPIC-Oxford [1, 2]. Small numbers mean wide confidence intervals, and wide confidence intervals mean statistics can’t cleanly separate the groups at the top of the staircase. In the AHS-2 mortality data, the pescatarian point estimate (0.81) nosed ahead of the vegan one (0.85), but their confidence intervals overlapped heavily: the study cannot tell you pescatarians outlived vegans, only that both handily beat meat-eaters [1]. Statistically, a tie at the top.

But even that tie deserves scrutiny, because several confounders plausibly ran against the vegans:

Do pescatarians live longer than vegans?

Motivation shapes diet quality, and the psychology literature says the sorting runs opposite to what skeptics assume. The standard objection to favorable vegan findings is healthy-user bias: vegans, the story goes, are health-obsessed self-selectors whose numbers are inflated by unmeasured conscientiousness. The actual research on dietary motivation points the other way. Across studies, the less restrictive a form of animal-product avoidance, the more likely it is health-driven [84]. Pescatarians skew heavily health-motivated, theirs is the least restrictive pattern and the one mainstream guidelines actively bless with recommendations to eat fish regularly [85]. Vegans, by contrast, are distinguished by animal and environmental motives more than health motives [86], with the commitment running through a rejection of speciesism itself [87]. In other words, health-optimizers preferentially sort into pescatarianism, while ethics sorts people into veganism, where the only requirement is that animal products be absent, a bar Oreos famously clear. That supplies a concrete mechanism for why the vegan category harbors more nutritionally careless members despite owning the healthiest food base, why B12 gaps concentrate there, and, critically, why any residual healthy-user bias in these cohorts plausibly flatters the fish-eaters, not the vegans. The Harvard PDI data proved unhealthful plant-based eating carries real risk [4], yet AHS-2’s categories couldn’t distinguish a whole-food vegan from a junk-food vegan; nobody controlled for diet quality within groups. (One honest footnote: this motivation research comes from general populations. AHS-2’s participants all sit inside a church that promotes plant-forward eating on religious and health grounds, so the same sorting is plausible there but not directly demonstrated.)

Nobody measured B12. Vitamin B12 deficiency, entirely preventable with a supplement costing pennies, raises homocysteine and plausibly harms vessels and nerves. Many vegans in these older cohorts weren’t reliably supplementing (recall EPIC-Oxford’s stroke signal) [2, 24]. Unsupplemented veganism isn’t the diet any modern authority recommends, but it’s part of what these studies measured.

Nobody controlled for duration or consistency. A “vegan” at baseline may have adopted the diet six months ago, and food-frequency questionnaires misclassify people in both directions. Both problems dilute real effects.

Despite all that headwind, vegans still posted the lowest BMI, the lowest diabetes prevalence, the lowest hypertension, the lowest overall cancer incidence, and a mortality point estimate 15% below meat-eaters, with vegan men showing the largest survival advantage of any group in the study [1, 19–21]. If anything, the fair reading is that the epidemiology understates the potential of a well-planned vegan diet, because it mostly measured unplanned ones. And one symmetry to concede in the same breath: if the vegan category is too internally varied to lean on when a result looks unfavorable, it is equally varied when a result looks favorable. Group averages over a category defined by refusal are blunt instruments in both directions, which is exactly why this case doesn’t rest on them.

And this is where the logic of inference matters. When a dose-response curve improves monotonically at every measured step, one serving of red meat to none, meat to fish, fish to dairy-and-eggs, most-plants to all-plants on biomarkers: the burden of proof falls on anyone claiming the curve suddenly reverses at the final step. Nothing in the data suggests it does, provided the final step is taken with whole foods and basic supplementation.

Is eating a little meat okay for your health?

Which raises the question that actually divides plant-based clinicians: not whether the curve reverses at the end, but whether it flattens. Many doctors who recommend plant-forward eating tell patients that a small amount of animal food, often pegged around 5% of calories, is unlikely to matter much. That advice has a real basis. The longest-lived populations on record ate roughly 95% plants, not 100%, with meat appearing as a rarity: Blue Zones research puts it around five times a month in small portions, and Dan Buettner’s own recommendation is framed as a range, 95% to 100% plant-based [12, 13]. Notice what that range does. It isn’t a prescription for 5% animal food; it’s an upper bound with zero explicitly inside it. The plants are credited with the work, and the small animal share is treated as tolerated rather than required.

And the honest position is that at that intake, the health difference between 5% and 0% is probably small, possibly too small for any study we have to detect. But be precise about why we can’t detect it. It isn’t that researchers measured the endpoint and found the curve flat. It’s that the endpoint was never populated: with no vegan country and no large vegan population, the lowest-animal group anyone could observe is the 5% group. “Five percent is fine” is a statement about where the data stops, not a finding about where the benefit stops. Three practical problems follow. First, 5% of what, processed meat and occasional fish carry very different risk profiles, and a percentage target flattens a distinction the evidence says matters enormously. Second, a threshold is harder to hold than a rule; in practice “a little” drifts, and clinicians often end up recommending zero precisely because it’s the more stable instruction. Third, and most telling: the case for keeping that 5% is about palatability and adherence, not physiology. Nobody argues the animal share is doing nutritional work that plants plus a B12 supplement can’t do. So even granting the generous reading: that the last 5% is a wash for your health: the health argument has simply gone quiet at that margin. It hasn’t come out in favor of anything. And when the health ledger is neutral, the remaining question isn’t medical: if you don’t need it, why eat it?

To move beyond inference, though, we need the tools that don’t share epidemiology’s weaknesses. That’s next.

Lifespan vs. healthspan: aiming at the target you can actually hit

What is the difference between lifespan and healthspan?

A quick but essential reframe before the heavy evidence. Lifespan, your final number, is partly hostage to genetics and plain luck: the drunk driver, the rare mutation, the pandemic. As we saw, genes explain perhaps a fifth of lifespan variation, maybe far less [16, 17]. Healthspan: the years you live free of chronic disease, disability, and a pharmacy’s worth of prescriptions, is a different animal. It is overwhelmingly determined by the modifiable factors, diet chief among them, and it is where the gap between eating patterns yawns widest.

Stanford physician James Fries called the goal “compression of morbidity”: pushing the onset of chronic disease so late that the sick, dependent phase of life shrinks to a brief window at the very end [29]. This is precisely what the plant-based data show. Adventist vegetarians don’t just die later; along the way they take fewer medications, undergo fewer surgeries, and carry less diabetes, hypertension, and heart disease [15, 19, 20]. Contrast that with Hong Kong, where researchers openly worry that lifespan has outrun healthspan, leaving more years of late-life dependency [9]. Quality of life is not a consolation prize for the years column, and conveniently, the same choices that expand healthspan (preventing the heart attacks, strokes, cancers, and diabetes that kill most people) are the ones that feed back into lifespan anyway. Aim at healthspan; lifespan largely comes along for the ride, minus the luck you never controlled.

So: which diet maximizes healthspan? Epidemiology gave us the gradient. Now let’s stress-test it with every tool that doesn’t depend on observational data.

Triangulation: the evidence that doesn’t share epidemiology’s weaknesses

Can a plant-based diet reverse heart disease?

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

Randomized controlled trials. You can’t randomize lifespans, but you can randomize diets and watch disease itself. Dean Ornish’s Lifestyle Heart Trial randomized patients with coronary artery disease to a whole-food, very-low-fat vegetarian diet plus lifestyle changes or usual care: at one year, 82% of the intervention group showed measurable regression of their arterial blockages on angiography, and at five years the diet group had continued regression while controls progressed and suffered more than twice the cardiac events [30, 31]. The trial was small, a few dozen patients, and bundled diet with exercise and stress management, so it proves the package works rather than isolating the plants; but the direction is unambiguous. Caldwell Esselstyn’s case series at the Cleveland Clinic, not a randomized trial, and reliant on self-selected adherents, put 198 patients with established cardiovascular disease on whole-food plant-based nutrition: among the 89% who stuck with it, recurrent cardiac events hit 0.6%, versus 62% among those who abandoned the diet [32]. Take the exact numbers with the grain of salt any uncontrolled series deserves; the effect size is nonetheless extraordinary. The BROAD trial in New Zealand randomized patients to a whole-food plant-based diet with no calorie counting and recorded the largest BMI reduction of any comparable trial at 6 and 12 months [33]. Neal Barnard’s randomized trials found a low-fat vegan diet outperformed the American Diabetes Association’s own recommended diet for glycemic control in type 2 diabetes [34]. Even DASH: the diet that topped the AHA’s rankings, earned its reputation in randomized feeding trials where a pattern centered on fruits and vegetables lowered blood pressure within weeks [73]; it’s a plant-forward design that can be executed fully plant-based, pointing the same direction while stopping short of the destination. And in the tightest-controlled comparison anyone has run, the NIH put 20 adults on a metabolic ward and fed them two diets in random order, two weeks each, unlimited food, nobody counting anything: a minimally processed plant-based low-fat diet against a minimally processed animal-based ketogenic one. On the plant-based diet, people spontaneously ate 689 fewer calories a day, with no difference in hunger or enjoyment [112]. The retellings usually stop there, but the body-composition data is the real story: the scale favored the ketogenic diet, which produced more total weight loss, but participants lost body fat on the plant-based diet and did not significantly lose body fat on the ketogenic one, where what came off was largely fat-free mass and water despite higher protein intake. Weight on a scale and fat on a body are not the same measurement, and a diet can win the first while losing the second. These are causal demonstrations that moving all the way to plants changes disease trajectories, not just biomarkers.

What did the Stanford identical twin vegan study find?

Twin studies, genetics, controlled. In 2023, Stanford ran the study skeptics kept asking for: 22 pairs of identical twins, one twin randomized to a healthy vegan diet, the other to a healthy omnivorous diet, for eight weeks. Same genes, same upbringing, both arms eating well, isolating the animal-food variable about as cleanly as human research allows. The vegan twins ended with significantly lower LDL cholesterol (~14 mg/dL lower, from an already-healthy baseline), roughly 20% lower fasting insulin, and 4.2 pounds more weight loss than their genetically identical omnivore siblings [35]. A follow-up analysis of the same twins found the vegan arm reduced biological age acceleration across multiple epigenetic clocks: they got younger by molecular measures of aging [36]. Two caveats belong with that result: several of the paper’s authors are employees of the company that sells the epigenetic tests used to measure it, and the vegan twins also lost more weight than their siblings, so part of the shift may reflect weight loss rather than diet composition specifically. Epigenetic clocks are surrogate markers, not disease outcomes, suggestive, not proof. Senior author Christopher Gardner’s conclusion was that, thinking about longevity, most people would benefit from shifting further toward plants [35].

Mendelian randomization, nature’s own RCT. MR studies use gene variants, randomly assigned at conception and immune to lifestyle confounding, to test causality. MR has established beyond reasonable doubt that LDL cholesterol and ApoB-containing lipoproteins cause atherosclerotic cardiovascular disease, cumulatively, over a lifetime [37]. That matters here because randomized trials consistently show whole-food plant-based and vegan diets produce the largest dietary LDL reductions, a 2023 pooling of 30 RCTs found vegetarian and vegan diets lowered ApoB, the single best marker of artery-clogging particles, by about 13 mg/dL versus omnivorous diets, an effect the senior author sized at roughly a third of a statin’s [113], while the primary dietary drivers of LDL are saturated fat and cholesterol, found overwhelmingly in animal foods [30, 34, 35]. And this isn’t just biomarker movement: Cochrane’s own review of 15 trials and 56,675 participants found that reducing saturated fat cut cardiovascular events, not just cholesterol, by 17%, with the benefit tracking exactly how much LDL fell and what the fat was replaced with, plant unsaturated fat worked, refined starch and sugar didn’t, which is also why some studies claim to have “exonerated” saturated fat: they swapped it for the wrong comparator [114]. The causal chain is welded shut at both ends: diet moves LDL (trials), and LDL moves heart disease (genetics). Similarly, MR studies implicate higher IGF-1 levels in causing breast, prostate, and colorectal cancers [38], and animal protein intake raises IGF-1, with studies finding vegans have the lowest circulating levels [39, 40]. One cohort analysis found that adults 50–65 reporting high protein intake, predominantly animal-based, had a fourfold increase in cancer death risk while plant protein showed no such association: the authors themselves reached for a comparison to smoking, though the honest framing is narrower: it’s a striking signal from a modest NHANES sample, it reversed entirely after 65 (where higher protein turned protective), and it’s worth citing for the mechanism rather than the shock value [40].

How do plant-based diets work biologically?

Mechanisms: the biology all points the same way. Fiber, found only in plants, feeds gut bacteria that produce short-chain fatty acids like butyrate, which nourish the colon, tame inflammation, and improve metabolic health; a Harvard experiment showed that even a few days on an animal-based diet rapidly shifts the microbiome toward bile-tolerant, inflammation-associated species [41]. Carnitine and choline, concentrated in meat, eggs, and dairy, are converted by gut microbes into TMAO, a metabolite that promotes atherosclerosis, and vegans, lacking the microbes and the substrate, barely produce it even when fed carnitine [42, 43]. TMAO has real critics, worth naming rather than skipping: some argue it’s a marker of diet and gut ecology rather than a driver of disease, pointing out that fish carries preformed TMAO without an apparent penalty, a live scientific dispute, not a settled one. A meta-analysis found vegetarians and vegans have significantly lower C-reactive protein, the blood’s cardinal inflammation marker [44]. Ornish’s team showed comprehensive plant-based lifestyle change increased telomerase activity and, over five years, actually lengthened telomeres: the protective chromosome caps whose erosion tracks cellular aging, while controls’ telomeres shortened [45, 46]. Add heme iron’s oxidative and nitroso-compound chemistry, advanced glycation end-products, and the inflammatory sugar molecule Neu5Gc from red meat, though that last one deserves its own hedge: the human colorectal evidence is a single antibody study with an odds ratio of 2.98 whose confidence interval, 0.80 to 11.1, crosses 1.0, so treat it as a plausible mechanism rather than an established one [28, 47].

And notice what’s actually on offer isn’t protein against no protein. Animals cannot synthesize any of the nine essential amino acids either, no cow, no chicken, no fish, any more than a human can. Every essential amino acid in every steak on earth was assembled first by a plant or a nitrogen-fixing microbe; the animal harvests it, burns most of it simply staying alive, and stores the remainder in tissue. Eating the plant is going to the factory; eating the animal is buying secondhand, after the first buyer already spent most of the inventory. What arrives with the animal version is a toll the plant version doesn’t charge: between 5% and 35% of egg protein escapes absorption in the small intestine and reaches the colon, where bacteria putrefy it into hydrogen sulfide, ammonia, phenols, and indoles [115]. And measured across more than 3,100 foods, the plant kingdom’s antioxidant capacity outruns the animal kingdom’s by roughly ninefold at the conservative median and by far more on average, with most animal foods sitting near zero [116]. Fuel plus fire-suppression versus fuel alone is the trade on offer, once you look past the amino acid headline both sides agree on.

When mechanism, genetics, randomized trials, twins, and epidemiology all converge, that’s not a coincidence. That’s consilience: the strongest form of proof available for a question no ethics board would ever let us test directly.

So what about the Blue Zones?

Are the Blue Zones real or debunked?

You may have noticed that this post has now made its entire case, gradient, trials, twins, genetics, mechanisms, without once leaning on the most famous longevity story in popular culture. That was deliberate. With the load-bearing walls in place, the Blue Zones can be enjoyed for what they are: corroborating color.

The story: demographers Michel Poulain and Gianni Pes, later joined by author Dan Buettner, identified regions with unusual concentrations of very old people, Okinawa (Japan), Sardinia (Italy), Nicoya (Costa Rica), Ikaria (Greece), and Loma Linda, California [11, 12]. These places differ in geography, religion, and cuisine, but their plates rhyme: an analysis of 154 dietary surveys conducted across the five zones found that roughly 95% of what the longest-lived residents ate came from plants, beans above all, plus whole grains, vegetables, greens, tubers, fruits, and nuts, with meat treated as a rarity or celebratory food [13]. Traditional Okinawans are the starkest case: mid-century surveys show they derived about 69% of their calories from sweet potatoes alone, with well over 90% of the diet from plants, and Okinawa long held the world’s highest concentration of centenarians and the longest disability-free life expectancy [14]. Sardinia’s hotspot runs on minestrone, sourdough, and fava beans; Nicoya on the “three sisters” of beans, squash, and corn; Ikaria on greens, legumes, and olive oil [12, 13].

Now, the heat. Demographic critics, most prominently Saul Justin Newman, whose analysis of age-record errors and pension fraud in extreme-age data earned an Ig Nobel Prize, argue that some celebrated old-age clusters partly reflect bad paperwork rather than good habits [88]. The skeptic’s case deserves its strongest form, and Newman is right that most raw age claims at the extreme are unreliable. Two things temper him. First, his central paper remains a preprint, not peer-reviewed, while in 2025 the gerontologists Austad and Pes answered him in The Gerontologist, walking through the multi-source cross-validation (birth ledgers, church archives, censuses) behind each validated zone [107]. Second, his load-bearing examples don’t hold up claim by claim against the primary sources: Japan’s notorious “230,000 missing centenarians” came off the koseki family register, a genealogical record with no purge mechanism, while the life-expectancy figures are built from the resident registry, which was untouched; Greece’s 9,000 figure was a pension payment roll, while Ikaria was validated years earlier on separate records; and Costa Rica’s age exaggeration had already been caught, measured, and published by Luis Rosero-Bixby himself, the very researcher Newman accuses, who then rebuilt his estimates from birth-ledger dates to route around it. The full accounting is its own post, Saul Newman vs. the Blue Zones: A Fact-Check, with the broader ten-pillar version, covering Sardinia’s self-auditing verification and the Loma Linda zip-code fallacy, in Blue Zones: Debunked or Not?. The fairest reading leaves one honest casualty: Nicoya, whose longevity advantage Rosero-Bixby has since documented fading in later-born cohorts [107]. Individual supercentenarian claims warrant deep suspicion; the diet-longevity signal survives at the population level. And here’s the thing, anyway: this post doesn’t need to referee that fight, and its structure is the reason. Strike every Blue Zone from the record tomorrow and not one preceding section changes. The cohorts still show their gradient, the twins still show their LDL, the genes still show their causality.

And notice the neat closing of the loop. The one Blue Zone built on modern vital records with a formal prospective cohort attached is Loma Linda, and that cohort is AHS-2, the individually tracked, statistically adjusted dataset this post already spent its middle chapters mining [1, 15]. It’s also the one zone without the confounding problem the others carry: Adventists live in the same country as their neighbors, shop at the same supermarkets, use the same medical system, and sit in the same unremarkable California county, so what differs is mainly what they do. And the gradient shows up here too, Adventist men outlived other Californians by 7.3 years, while vegetarian Adventist men outlived them by 9.5 [15]. Same religion, same town, more of the behavior, more of the benefit. Critics who dismiss Loma Linda as the freeway-exit Blue Zone are attacking the geographic packaging; underneath it sits the cleanest natural experiment in the field. The single Blue Zone that was converted into rigorous rung-two science delivered the plant-based gradient in full. The Blue Zones are the illustration on the cover; the evidence above is the book.

The steelman corner: the best studies from the other side

What are the best arguments against a plant-based diet?

A case that only rehearses friendly evidence isn’t a case; it’s a pep rally. So here are the exhibits that do the heaviest lifting in arguments against plant-based eating, each taken seriously, and each placed honestly on the ladder we built at the start.

What did the PURE study really show?

Exhibit one: the PURE study. In 2017, the Prospective Urban Rural Epidemiology study of over 135,000 people across 18 countries reported that higher carbohydrate intake was associated with higher mortality, while higher fat intake, including saturated fat, was associated with lower mortality [100]. Cue a thousand butter headlines. But look at who was actually being compared. PURE’s high-carbohydrate participants were overwhelmingly low-income populations eating the bulk of their calories as white rice and refined grains, poverty diets, not whole-plant diets, and in those settings, meat and fat intake are markers of affluence, which travels with healthcare access, sanitation, and lower infectious-disease mortality; much of the excess death in the high-carb strata wasn’t cardiovascular at all. The study contained essentially no whole-food plant-based eaters to test. And here’s the part the butter headlines skipped: PURE’s own companion paper, published in the very same issue of The Lancet, found that higher intake of fruits, vegetables, and legumes predicted lower mortality, with legumes protective in their own right [101]. Where its saturated-fat association conflicts with hundreds of controlled feeding trials and the Mendelian randomization evidence on LDL [37], it is a single confounded multi-country food questionnaire arguing against rungs above it. Where PURE is methodologically strongest, whole plant foods: it agrees with this post.

Why did some scientists say to keep eating red meat?

Exhibit two: the 2019 “keep eating red meat” papers. The NutriRECS consortium’s guidelines in the Annals of Internal Medicine made global news by advising adults to simply continue their current red and processed meat consumption [102]. Read past the headline and something odd appears: the panel’s own meta-analyses found the same directional risk reductions everyone else finds for eating less meat. What changed was the grading, not the data. The authors applied GRADE, a certainty framework built for drug trials, which rates virtually all nutritional epidemiology “low certainty” because thirty-year randomized feeding trials don’t exist. By that yardstick, the evidence that cigarettes cause lung cancer is also “low certainty”, there were never any smoking RCTs either [6]. Major health bodies, including the AHA, the American Cancer Society, and Harvard’s nutrition department, publicly rejected the advice and kept their recommendations; subsequent reporting revealed undisclosed industry-linked funding ties for the lead author, prompting a correction to the paper’s disclosure statement. An older genre of null meta-analyses on meat and cancer, meanwhile, came from consultancy scientists with meat-industry groups among the funders [104]. None of this is new evidence for meat. It’s a re-grading exercise on the same evidence, and re-grading is not refutation.

Do studies show vegetarians are less healthy?

Exhibit three: the Austrian “unhealthy vegetarians” study. A 2014 cross-sectional survey found Austrian vegetarians self-reporting poorer health, more cancer, and more anxiety than meat-eaters [103]. Critics wave it constantly, and its ladder position settles it. Cross-sectional studies photograph a single moment, so they cannot say which came first, the diet or the disease, and reverse causation is the obvious reading, one the authors themselves flagged: people frequently adopt vegetarian diets after a cancer or digestive diagnosis, stacking the vegetarian category with the already-ill. Add a few hundred self-reporting participants and no diet-quality data, and you have a study sitting below even the cohorts this post spent pages stress-testing, a rung this post’s own side isn’t permitted to stand on either.

Do vegans have weaker bones or more hip fractures?

Exhibit four: vegans and hip fractures. Unlike the first three exhibits, this one doesn’t get defused: it gets contextualized, because the finding is real. EPIC-Oxford found vegans carry 2.31 times the hip-fracture hazard of meat-eaters after full adjustment for BMI and lifestyle, and it replicates: Adventist women came in at 1.55 times, and a 2025 pooled analysis across four cohorts and more than half a million people put vegans at 1.75 times [117–119]. Take that seriously before doing anything else with it.

Now look closely. The whole EPIC-Oxford signal rests on 41 vegan hip fractures, and the 2025 pooled estimate carries a confidence interval of 1.17 to 2.63, graded “very low certainty” by the reviewers’ own standard, meaning the true effect could be a 17% increase or a 163% one, and the data can’t say which [119]. In absolute terms, that’s roughly 15 extra hip fractures per 1,000 people over a decade. And when researchers traced the excess to its source, the sequential adjustment tells the whole story: moving from a base model through BMI, then calcium, then protein, calcium explains roughly three times as much of the excess as protein does, and vegans already eating adequate protein showed no reduction in risk at all: their hip-fracture hazard stayed as high as the unadjusted estimate [117]. That’s a direct hit on the old “acid-ash” theory that protein-poor plant eating is what’s driving this. The theory has a second problem: dual-isotope tracer studies find that the extra urinary calcium seen on higher-protein diets comes from increased intestinal absorption, not from bone breaking down to buffer acid: the opposite of what the hypothesis predicts [120].

Two independent cohorts, two continents, confirm what’s actually protective. EPIC-Oxford found vegans getting at least 525 mg of calcium a day had a fracture rate identical to meat-eaters; the Adventist cohort found 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 [118, 121]. Food choice within the diet does even more work: Adventists eating legumes daily had 64% lower hip-fracture rates, and those eating meat analogues daily had 49% lower rates, than infrequent eaters [122]. The same healthful-versus-unhealthful split that runs through every other outcome in this post shows up here too, a healthful plant-based pattern tracked with 21% lower hip-fracture risk across 70,000 postmenopausal women, an unhealthful one with 28% higher [123]. And an Austrian bone-microarchitecture study, using scans detailed enough to see trabecular and cortical structure rather than just overall density, found the remaining gap closes with resistance training: vegans who lifted were indistinguishable from omnivores who lifted; vegans who didn’t were measurably worse, and the difference wasn’t explained by nutrient intake [124]. None of this argues for eating animals, meat doesn’t reliably prevent fractures either, and the levers that do (calcium, vitamin D, resistance training) are fully available on a whole-food vegan diet. It argues for planning, which is the conclusion every other exhibit in this post keeps arriving at too.

Did a 2025 study prove high cholesterol on keto is safe?

Exhibit five: the study that claimed to disprove LDL, and the reason it no longer exists. In April 2025, JACC: Advances published data on 100 “lean mass hyper-responders”, people who develop sharply elevated LDL on a ketogenic diet, claiming a year of coronary CT scans showed no meaningful plaque progression, and the story spread fast as proof that cholesterol doesn’t matter after all if you’re lean and keto [125]. It didn’t survive scrutiny. The paper had quietly swapped its own pre-registered primary outcome for a friendlier secondary one; it ran with no comparator group; a year is short for plaque changes that typically take three to six to show up; and sample sizes shifted between figures depending on which told the better story. Then came the conflict: one author held an undisclosed equity stake in Cleerly, the company whose proprietary software read the scans the whole paper depended on. In March 2026, the journal retracted it outright, stating the errors were too extensive for a correction to fix [126]. This doesn’t touch the LDL-causation evidence itself, built from genetics and RCTs across millions of people: it’s simply the same pattern as the NutriRECS exhibit above, a year later and with better production values: a paper built to overturn established causation, promoted hard before anyone could check it, and gone within a year of publication.

One recurring exhibit deserves a final sentence: the U-shaped curves showing low cholesterol associated with higher mortality in some observational data. Same diagnosis, serious illness lowers cholesterol before it kills, and Mendelian randomization, which is immune to that trick, shows lifelong lower LDL means lifelong lower cardiovascular risk [37].

Notice the pattern. Most counter-exhibits sit on a lower rung than the evidence deployed against them, re-grade rather than refute, or collapse under exactly the scrutiny this post applied to its own side’s Blue Zones. One doesn’t collapse at all: the bone-fracture finding is real, and the honest move was to concede it and show what the concession actually means once you trace it to its source, the same treatment this post gave the EPIC-Oxford stroke signal and the AHS-2 2024 update. That’s the value of climbing the ladder in public: it flattens weak arguments no matter whose they are, and it holds the strong ones to the same light instead of hiding them.

What the world’s leading health authorities actually say

What do major health organizations say about vegan diets?

If the science above were fringe, you’d expect the institutions charged with reading all of it to disagree. They don’t.

The Academy of Nutrition and Dietetics: the world’s largest organization of food and nutrition professionals, with over 100,000 credentialed practitioners, published its updated position paper in 2025, concluding that appropriately planned vegetarian and vegan dietary patterns in adults can be nutritionally adequate and can offer long-term health benefits, particularly for cardiometabolic disease, and directing dietitians to actively support people adopting them [48]. (Its landmark 2016 position went further across the lifespan, deeming well-planned vegan diets appropriate for all life stages including pregnancy, infancy, and athletics; the 2025 update simply scoped itself to adults, explicitly leaving other life stages to dedicated future papers, not, as some meat advocates spun it, reversing anything [48, 49].)

The American Heart Association’s 2021 dietary guidance is built around plant-forward principles: plenty of vegetables, fruits, whole grains, and legumes, protein preferentially from plants, and minimal processed meat [50]. Even more telling, the AHA’s 2023 scientific statement graded ten popular diets against its criteria: the plant-forward patterns swept the top tier, DASH a perfect 100, pescatarian 92, Mediterranean 89, and vegetarian 86; the vegan pattern scored 78 in the second tier, docked not for disease outcomes but for practical adherence challenges and the need to manage B12; and the meat-centric paleo (53) and ketogenic (31) patterns landed dead last, flagged as contradicting heart-healthy guidance outright [51, 52].

The American Cancer Society’s guideline calls for eating patterns high in vegetables, fruits, legumes, and whole grains while limiting or eliminating red and processed meats [53], echoing the World Cancer Research Fund/AICR, whose recommendations are to eat a diet rich in whole plant foods, limit red meat, and eat little if any processed meat [54], and the WHO’s IARC carcinogen classifications noted earlier [28]. The American Diabetes Association’s consensus report lists vegetarian and vegan eating patterns among evidence-based options for diabetes management, with trials showing improved glycemia and cardiovascular risk factors [55]. The American Academy of Pediatrics’ own nutrition handbook recognizes that well-planned vegetarian and vegan diets can support normal growth and development in children, with appropriate attention to nutrients like B12, vitamin D, iron, and zinc [56].

What do official nutrition guidelines recommend?

Even the bodies that set the numerical rules tilt the same way. The National Academies’ Dietary Reference Intakes put the Acceptable Macronutrient Distribution Range for carbohydrate at 45–65% of calories: the highest range of any macronutrient, and a range that whole-plant diets naturally satisfy, versus 10–35% for protein and 20–35% for fat [57]. The same DRIs set the fiber recommendation at 14 grams per 1,000 calories (fiber exists only in plants, and roughly 19 in 20 Americans fall short) and concluded that trans fat and dietary cholesterol have no safe intake level, recommending consumption as low as possible within a nutritionally adequate diet, a bar only plant foods clear by default, since cholesterol comes exclusively from animal foods [57].

Why did the 2025 Dietary Guidelines ignore its own science panel?

Then there’s the revealing saga of the U.S. Dietary Guidelines. The independent 2025 Dietary Guidelines Advisory Committee, twenty of the country’s leading nutrition scientists, spent two years reviewing the evidence and delivered its most plant-forward scientific report ever: move beans, peas, and lentils into the protein group and list them first, ahead of nuts, seeds, and soy, then seafood, and meat, poultry, and eggs last; increase legume intake; reduce meat, poultry, and egg quantities; and replace saturated fat specifically with plant sources, while stating there is compelling evidence tying higher red and processed meat intake to harm [58, 59]. What happened next says more about politics than science: the final 2025–2030 guidelines, released January 7, 2026 under new HHS leadership, largely set that report aside, publishing a streamlined “real food” guide built around an inverted pyramid that foregrounds protein, full-fat dairy, and “healthy fats,” explicitly naming beef tallow and butter, while retaining the decades-old cap on saturated fat at under 10% of calories, a retention that sits awkwardly beside its promotion of the very foods richest in it. Harvard’s nutrition department, the American Heart Association, the Physicians Committee, and sidelined committee member Christopher Gardner all objected that the guidelines contradict the evidence their own advisory panel had assembled [60–62]. When the science panel and the political document diverge, remember which one read the evidence.

Which countries recommend plant-based diets?

The international trend needs no such asterisk. Canada’s 2019 Food Guide tells citizens to choose plant protein more often and eliminated the dairy food group entirely [63]. Denmark’s official dietary guidelines urge eating plant-rich, with legumes daily and far less meat [64]. The Nordic Nutrition Recommendations 2023, the science basis for five nations’ guidelines, prescribe a predominantly plant-based dietary pattern [65]. Germany’s nutrition society now recommends a diet of roughly three-quarters plant foods [66]. And the EAT-Lancet Commission, synthesizing global evidence on diets that maximize human health, first in 2019, updated as EAT-Lancet 2.0 in October 2025, landed on a planetary health plate that is overwhelmingly plant-based, with animal foods optional and minimal; the 2025 update estimates that global adoption is associated with roughly 27% lower risk of premature death, on the order of 15 million lives a year [67, 108]. Kaiser Permanente, one of the largest healthcare organizations in the US, published guidance in its own journal advising physicians to consider recommending plant-based diets to all their patients [68].

An entire plant-based medical ecosystem, with no counterpart on the other side

Which medical organizations are fully plant-based?

Beyond position papers, an institutional infrastructure has grown up around plant-predominant and fully plant-based nutrition, and its scale is worth appreciating. The American College of Lifestyle Medicine, among the fastest-growing medical professional societies in the country, now board-certifying physicians in lifestyle medicine, states as its official dietary position that an eating plan based predominantly on a variety of minimally processed, whole plant foods should be the foundation for treating and reversing chronic disease [69]. The Physicians Committee for Responsible Medicine counts more than 17,000 physicians among nearly a million members and supporters advocating plant-based prevention [13]. The Ornish Program for Reversing Heart Disease, built on a whole-food plant-based diet, became the first program ever covered by Medicare under its Intensive Cardiac Rehabilitation benefit in 2010, after CMS reviewed the peer-reviewed evidence that it improves and reverses coronary disease; major private insurers followed [70, 71]. (The other originally approved ICR program, Pritikin, is likewise plant-centered [71].) TrueNorth Health Center runs the largest medically supervised fasting facility in the world on an exclusively whole-food plant-based protocol; the National Health Association has promoted plant-exclusive eating since 1948; NutritionFacts.org provides a free, non-commercial video library translating the primary literature; Hippocrates Wellness has run plant-exclusive residential programs for decades. Whatever one thinks of any single organization, together they represent physicians, dietitians, researchers, insurers, and federal payers converging on the same dietary direction.

Is there a carnivore-diet equivalent of plant-based medicine?

Now run the comparison that ends the debate: where is the equivalent ecosystem for the opposite diet? There is no “American College of Carnivore Medicine” board-certifying physicians. No position paper from any national medical, cardiology, oncology, diabetes, dietetic, or pediatric organization anywhere on Earth endorses meat-maximal eating. No Medicare-covered carnivore program exists, because no body of peer-reviewed trials demonstrates disease reversal to submit to CMS. No cohort of 96,000 meat-maximizers shows a longevity gradient favoring more animal foods: the gradients all run the other way. The closest mainstream evaluation such diets have received is the AHA scoring keto and paleo at the bottom of its rankings as contradicting heart-healthy guidance [51, 52]. In science, when one hypothesis accumulates cohorts, trials, mechanisms, genetics, institutions, and insurers, and the rival hypothesis accumulates podcasts: the asymmetry is the answer.

The bottom line

What does all the evidence add up to?

Strip away the noise and the picture is coherent from every rung of the ladder. Within the best-studied longevity population on Earth, health improves stepwise with every rung toward fully plant-based, and vegans sit at or near the top on weight, diabetes, blood pressure, cancer, and survival despite every methodological headwind [1, 19–22]. Randomized trials show whole-food plant-based eating doesn’t just correlate with less disease: it reverses it [30–34]. Identical twins randomized to vegan eating beat their own genetic duplicates on the biomarkers that cause heart disease and even on molecular aging itself [35, 36]. The genetics, the microbiome, the inflammation data, and the telomeres all point the same way [37–46]. Even the world’s centenarian hotspots, taken as color, not foundation, rhyme with the same roughly 95%-plant, bean-anchored pattern [13]. And the institutions whose job is to weigh everything, from the Academy of Nutrition and Dietetics to the AHA to the world’s national food guides to Medicare’s own actuaries, keep landing on the same side [48–71].

Lifespan will always keep some cards hidden: genes, accidents, luck. Healthspan is the hand you play yourself, and the strongest cards in it are the people you keep close and what you put on your plate. The first of those this post can only point at; the second it can measure, and no eating pattern has more converging evidence behind it than one built entirely from whole plant foods, vegetables, fruits, legumes, whole grains, nuts, seeds, mushrooms, herbs, and spices, done properly, which means a reliable B12 supplement (and attention to vitamin D and omega-3s) as non-negotiable basics rather than afterthoughts. One distinction worth carrying out the door: veganism names the ethic; whole-food plant-based names the diet. The quiet convenience of our moment is that they meet on the same plate, every food group at the core of the evidence-backed pattern is vegan, so living the ethic and optimizing the diet call for the same grocery list. And all of that is tallying only the health ledger: the ethical and environmental columns are taken up in the companion post, Rewilding the Plate. You don’t need a vegan country to exist before acting on this; the staircase is visible, every step up improves the view, and nothing at the top suggests the pattern breaks. The longest-living people already showed us the direction. The science just keeps confirming how far it goes.

A companion document, an evidence table sorting all 130 references by their rung on the ladder, accompanies this post.

Related reading: What the Adventist Health Studies Actually Tell Us About Vegan Diets, And What They Don’t is the full deep-dive on the cohort at the heart of this post’s gradient. Rewilding the Plate tallies the environmental ledger, and The Burden of Proof Has Shifted asks what follows once the evidence lands this way. The Full Evidence Review is the systematically tiered, conflict-of-interest-screened version of everything in this post’s triangulation section, across 48 peer-reviewed studies. Do Vegans Have Weaker Bones? and The Full Protein Breakdown go further on their specific mechanisms than a single post has room for.

References

1. Orlich MJ, Singh PN, Sabaté J, et al. Vegetarian dietary patterns and mortality in Adventist Health Study 2. JAMA Internal Medicine. 2013;173(13):1230–1238.
2. Tong TYN, Appleby PN, Bradbury KE, 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. Link
3. Satija A, Bhupathiraju SN, Spiegelman D, 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.
4. Satija A, Bhupathiraju SN, Rimm EB, 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.
5. Pan A, Sun Q, Bernstein AM, et al. Red meat consumption and mortality: results from 2 prospective cohort studies. Archives of Internal Medicine. 2012;172(7):555–563.
6. Hill AB. The environment and disease: association or causation? Proceedings of the Royal Society of Medicine. 1965;58(5):295–300.
7. Ni MY, Canudas-Romo V, Shi J, et al. Understanding longevity in Hong Kong: a comparative study with long-living, high-income countries. The Lancet Public Health. 2021;6(12):e919–e931.
8. The University of Hong Kong. Tobacco control shown to be the key driver for Hong Kong’s world-leading longevity. HKUMed press release, 2021.
9. Woo J, et al. Non-communicable diseases, longevity, and health span: a Hong Kong perspective. Review, 2025 (PMC11942087).
10. Hong Kong Cancer Registry, Hospital Authority. Colorectal cancer statistics (leading cancer sites in Hong Kong).
11. Poulain M, Pes GM, Grasland C, et al. Identification of a geographic area characterized by extreme longevity in the Sardinia island: the AKEA study. Experimental Gerontology. 2004;39(9):1423–1429.
12. Buettner D, Skemp S. Blue Zones: lessons from the world’s longest lived. American Journal of Lifestyle Medicine. 2016;10(5):318–321.
13. Buettner D. The Blue Zones (analysis of 154 dietary surveys across the five zones); summarized in Institute of Medicine, Business Engagement in Building Healthy Communities. National Academies Press; 2015.
14. Willcox BJ, Willcox DC, Todoriki H, et al. Caloric restriction, the traditional Okinawan diet, and healthy aging. Annals of the New York Academy of Sciences. 2007;1114:434–455.
15. Fraser GE, Shavlik DJ. Ten years of life: is it a matter of choice? Archives of Internal Medicine. 2001;161(13):1645–1652.
16. Herskind AM, McGue M, Holm NV, et al. The heritability of human longevity: a population-based study of 2872 Danish twin pairs born 1870–1900. Human Genetics. 1996;97(3):319–323.
17. Ruby JG, Wright KM, Rand KA, et al. Estimates of the heritability of human longevity are substantially inflated due to assortative mating. Genetics. 2018;210(3):1109–1124.
18. GBD 2017 Diet Collaborators. Health effects of dietary risks in 195 countries, 1990–2017. The Lancet. 2019;393(10184):1958–1972.
19. 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.
20. Matsumoto S, et al. Association between vegetarian diets and cardiovascular risk factors in non-Hispanic white participants of the Adventist Health Study-2. Journal of Nutritional Science. 2019;8:e6.
21. Tantamango-Bartley Y, Jaceldo-Siegl K, Fan J, Fraser G. Vegetarian diets and the incidence of cancer in a low-risk population. Cancer Epidemiology, Biomarkers & Prevention. 2013;22(2):286–294.
22. Fraser GE, Cosgrove CM, Mashchak AD, et al. Cause-specific and all-cause mortalities in vegetarian compared with those in nonvegetarian participants from the Adventist Health Study-2 cohort. The American Journal of Clinical Nutrition. 2024. Link
23. Key TJ, Fraser GE, Thorogood M, et al. Mortality in vegetarians and nonvegetarians: detailed findings from a collaborative analysis of 5 prospective studies. The American Journal of Clinical Nutrition. 1999;70(3 Suppl):516S–524S.
24. Vegan Health (Norris J). Mortality and cardiovascular disease rates of vegetarians: discussion of EPIC-Oxford stroke findings, B12, and cholesterol. veganhealth.org.
25. Zheng Y, Li Y, Satija A, et al. Association of changes in red meat consumption with total and cause specific mortality among US women and men. BMJ. 2019;365:l2110.
26. Song M, Fung TT, Hu FB, et al. Association of animal and plant protein intake with all-cause and cause-specific mortality. JAMA Internal Medicine. 2016;176(10):1453–1463.
27. Alshahrani SM, Fraser GE, Sabaté J, et al. Red and processed meat and mortality in a low meat intake population. Nutrients. 2019;11(3):622.
28. Bouvard V, Loomis D, Guyton KZ, et al. Carcinogenicity of consumption of red and processed meat. The Lancet Oncology. 2015;16(16):1599–1600.
29. Fries JF. Aging, natural death, and the compression of morbidity. New England Journal of Medicine. 1980;303(3):130–135.
30. Ornish D, Brown SE, Scherwitz LW, et al. Can lifestyle changes reverse coronary heart disease? The Lifestyle Heart Trial. The Lancet. 1990;336(8708):129–133.
31. Ornish D, Scherwitz LW, Billings JH, et al. Intensive lifestyle changes for reversal of coronary heart disease. JAMA. 1998;280(23):2001–2007.
32. Esselstyn CB Jr, Gendy G, Doyle J, et al. A way to reverse CAD? Journal of Family Practice. 2014;63(7):356–364b.
33. Wright N, Wilson L, Smith M, et al. The BROAD study: a randomised controlled trial using a whole food plant-based diet in the community for obesity, ischaemic heart disease or diabetes. Nutrition & Diabetes. 2017;7(3):e256.
34. Barnard ND, Cohen J, Jenkins DJA, et al. A low-fat vegan diet improves glycemic control and cardiovascular risk factors in a randomized clinical trial in individuals with type 2 diabetes. Diabetes Care. 2006;29(8):1777–1783.
35. Landry MJ, Ward CP, Cunanan KM, et al. Cardiometabolic effects of omnivorous vs vegan diets in identical twins: a randomized clinical trial. JAMA Network Open. 2023;6(11):e2344457. Link
36. 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.
37. Ference BA, Ginsberg HN, Graham I, 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.
38. Larsson SC, Carter P, Vithayathil S, et al. Insulin-like growth factor-1 and site-specific cancers: a Mendelian randomization study. Cancer Medicine. 2020;9(18):6836–6842.
39. Allen NE, Appleby PN, Davey GK, et al. The associations of diet with serum insulin-like growth factor I and its main binding proteins in 292 women meat-eaters, vegetarians, and vegans. Cancer Epidemiology, Biomarkers & Prevention. 2002;11(11):1441–1448.
40. Levine ME, Suarez JA, Brandhorst S, et al. Low protein intake is associated with a major reduction in IGF-1, cancer, and overall mortality in the 65 and younger but not older population. Cell Metabolism. 2014;19(3):407–417.
41. David LA, Maurice CF, Carmody RN, et al. Diet rapidly and reproducibly alters the human gut microbiome. Nature. 2014;505(7484):559–563.
42. Koeth RA, Wang Z, Levison BS, et al. Intestinal microbiota metabolism of L-carnitine, a nutrient in red meat, promotes atherosclerosis. Nature Medicine. 2013;19(5):576–585.
43. Tang WHW, Wang Z, Levison BS, et al. Intestinal microbial metabolism of phosphatidylcholine and cardiovascular risk. New England Journal of Medicine. 2013;368(17):1575–1584.
44. Menzel J, Jabakhanji A, Biemann R, et al. Systematic review and meta-analysis of the associations of vegan and vegetarian diets with inflammatory biomarkers. Scientific Reports. 2020;10:21736.
45. Ornish D, Lin J, Daubenmier J, et al. Increased telomerase activity and comprehensive lifestyle changes: a pilot study. The Lancet Oncology. 2008;9(11):1048–1057.
46. Ornish D, Lin J, Chan JM, et al. Effect of comprehensive lifestyle changes on telomerase activity and telomere length in men with biopsy-proven low-risk prostate cancer: 5-year follow-up of a descriptive pilot study. The Lancet Oncology. 2013;14(11):1112–1120.
47. Samraj AN, Pearce OMT, Läubli H, et al. A red meat-derived glycan promotes inflammation and cancer progression. PNAS. 2015;112(2):542–547.
48. Raj S, Guest NS, Landry MJ, Mangels AR, Pawlak R, Rozga M. Vegetarian dietary patterns for adults: a position paper of the Academy of Nutrition and Dietetics. Journal of the Academy of Nutrition and Dietetics. 2025;125(6):831–846. Link
49. Melina V, Craig W, Levin S. Position of the Academy of Nutrition and Dietetics: vegetarian diets. Journal of the Academy of Nutrition and Dietetics. 2016;116(12):1970–1980.
50. Lichtenstein AH, Appel LJ, Vadiveloo M, et al. 2021 dietary guidance to improve cardiovascular health: a scientific statement from the American Heart Association. Circulation. 2021;144(23):e472–e487.
51. 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.
52. American Heart Association. 10 popular diets scored for heart-healthy elements. AHA Newsroom, April 27, 2023. Link
53. Rock CL, Thomson C, Gansler T, et al. American Cancer Society guideline for diet and physical activity for cancer prevention. CA: A Cancer Journal for Clinicians. 2020;70(4):245–271.
54. World Cancer Research Fund/American Institute for Cancer Research. Diet, Nutrition, Physical Activity and Cancer: A Global Perspective. Third Expert Report. 2018.
55. Evert AB, Dennison M, Gardner CD, et al. Nutrition therapy for adults with diabetes or prediabetes: a consensus report. Diabetes Care. 2019;42(5):731–754.
56. Kleinman RE, Greer FR, eds. Pediatric Nutrition, 8th ed. American Academy of Pediatrics; 2019 (chapter on vegetarian and vegan diets in children).
57. Institute of Medicine (National Academies). Dietary Reference Intakes for Energy, Carbohydrate, Fiber, Fat, Fatty Acids, Cholesterol, Protein, and Amino Acids. National Academies Press; 2005.
58. 2025 Dietary Guidelines Advisory Committee. Scientific Report of the 2025 Dietary Guidelines Advisory Committee. USDA/HHS; December 2024.
59. Physicians Committee for Responsible Medicine. Prioritizing plant-based protein in the Scientific Report of the 2025–2030 Dietary Guidelines is a step forward. December 2024.
60. U.S. Department of Health and Human Services and U.S. Department of Agriculture. Dietary Guidelines for Americans, 2025–2030. January 2026.
61. Harvard T.H. Chan School of Public Health, The Nutrition Source. Dietary Guidelines for Americans 2025–2030: progress on added sugar, protein hype, saturated fat contradictions. January 2026.
62. U.S. News & World Report. New Dietary Guidelines for Americans 2025–2030: reimagining the food pyramid (comments of DGAC member Christopher Gardner). January 2026.
63. Health Canada. Canada’s Food Guide. 2019.
64. Danish Veterinary and Food Administration. The Official Dietary Guidelines, Good for Health and Climate. 2021.
65. Blomhoff R, Andersen R, Arnesen EK, et al. Nordic Nutrition Recommendations 2023. Nordic Council of Ministers; 2023.
66. German Nutrition Society (DGE). Position statement on a more plant-based diet. 2024.
67. Willett W, Rockström J, Loken B, et al. Food in the Anthropocene: the EAT-Lancet Commission on healthy diets from sustainable food systems. The Lancet. 2019;393(10170):447–492. Link
68. Tuso PJ, Ismail MH, Ha BP, Bartolotto C. Nutritional update for physicians: plant-based diets. The Permanente Journal. 2013;17(2):61–66.
69. American College of Lifestyle Medicine. Dietary lifestyle position statement (whole-food, predominantly plant-based nutrition for treatment and reversal of chronic disease). 2022.
70. Centers for Medicare & Medicaid Services. National Coverage Determination 20.31.2: Ornish Program for Reversing Heart Disease (Intensive Cardiac Rehabilitation). Effective August 12, 2010.
71. Ornish Lifestyle Medicine. The Ornish Reversal Program: first program covered by Medicare under Intensive Cardiac Rehabilitation.
72. Yusuf S, Hawken S, Ôunpuu S, et al. Effect of potentially modifiable risk factors associated with myocardial infarction in 52 countries (the INTERHEART study): case-control study. The Lancet. 2004;364(9438):937–952.
73. Appel LJ, Moore TJ, Obarzanek E, et al. A clinical trial of the effects of dietary patterns on blood pressure. DASH Collaborative Research Group. New England Journal of Medicine. 1997;336(16):1117–1124.
74. Micha R, Peñalvo JL, Cudhea F, et al. Association between dietary factors and mortality from heart disease, stroke, and type 2 diabetes in the United States. JAMA. 2017;317(9):912–924.
75. Micha R, Wallace SK, Mozaffarian D. Red and processed meat consumption and risk of incident coronary heart disease, stroke, and diabetes mellitus: a systematic review and meta-analysis. Circulation. 2010;121(21):2271–2283.
76. Pan A, Sun Q, Bernstein AM, et al. Red meat consumption and risk of type 2 diabetes: 3 cohorts of US adults and an updated meta-analysis. The American Journal of Clinical Nutrition. 2011;94(4):1088–1096.
77. Rohrmann S, Overvad K, Bueno-de-Mesquita HB, et al. Meat consumption and mortality, results from the European Prospective Investigation into Cancer and Nutrition. BMC Medicine. 2013;11:63.
78. Aune D, Navarro Rosenblatt DA, Chan DSM, et al. Dairy products, calcium, and prostate cancer risk: a systematic review and meta-analysis of cohort studies. The American Journal of Clinical Nutrition. 2015;101(1):87–117.
79. Richman EL, Kenfield SA, Stampfer MJ, et al. Egg, red meat, and poultry intake and risk of lethal prostate cancer in the prostate-specific antigen-era: incidence and survival. Cancer Prevention Research. 2011;4(12):2110–2121.
80. Keum N, Lee DH, Marchand N, et al. Egg intake and cancers of the breast, ovary and prostate: a dose-response meta-analysis of prospective observational studies. British Journal of Nutrition. 2015;114(7):1099–1107.
81. Fraser GE, Jaceldo-Siegl K, Orlich M, et al. Dairy, soy, and risk of breast cancer: those confounded milks. International Journal of Epidemiology. 2020;49(5):1526–1537. Link
82. Dinu M, Abbate R, Gensini GF, Casini A, Sofi F. 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.
83. The Vegan Society. Definition of veganism (established 1944). vegansociety.com.
84. Rosenfeld DL. The psychology of vegetarianism: recent advances and future directions. Appetite. 2018;131:125–138.
85. 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;26:713–727.
86. Hopwood CJ, Bleidorn W, Schwaba T, Chen S. Health, environmental, and animal rights motives for vegetarian eating. PLOS ONE. 2020;15(4):e0230609.
87. Rosenfeld DL. Ethical motivation and vegetarian dieting: the underlying role of anti-speciesist attitudes. Anthrozoös. 2019;32(6):785–796.
88. Newman SJ. Supercentenarian and remarkable age records exhibit patterns indicative of clerical errors and pension fraud. bioRxiv preprint (first posted 2019; awarded the 2024 Ig Nobel Prize in Demography).
89. Darmadi-Blackberry I, Wahlqvist ML, Kouris-Blazos A, et al. Legumes: the most important dietary predictor of survival in older people of different ethnicities. Asia Pacific Journal of Clinical Nutrition. 2004;13(2):217–220.
90. Bao Y, Han J, Hu FB, et al. Association of nut consumption with total and cause-specific mortality. New England Journal of Medicine. 2013;369(21):2001–2011.
91. Aune D, Keum N, Giovannucci E, et al. Whole grain consumption and risk of cardiovascular disease, cancer, and all cause and cause specific mortality: systematic review and dose-response meta-analysis of prospective studies. BMJ. 2016;353:i2716.
92. Aune D, Giovannucci E, Boffetta P, et al. Fruit and vegetable intake and the risk of cardiovascular disease, total cancer and all-cause mortality, a systematic review and dose-response meta-analysis of prospective studies. International Journal of Epidemiology. 2017;46(3):1029–1056.
93. Reynolds A, Mann J, Cummings J, et al. Carbohydrate quality and human health: a series of systematic reviews and meta-analyses. The Lancet. 2019;393(10170):434–445.
94. Morris MC, Wang Y, Barnes LL, et al. Nutrients and bioactives in green leafy vegetables and cognitive decline. Neurology. 2018;90(3):e214–e222.
95. Ba DM, Gao X, Al-Shaar L, et al. Prospective study of dietary mushroom intake and risk of mortality: results from continuous NHANES 2003–2014 and a meta-analysis. Nutrition Journal. 2021;20:80.
96. Murai U, Yamagishi K, Sata M, et al. Seaweed intake and risk of cardiovascular disease: the Japan Public Health Center–based Prospective (JPHC) Study. The American Journal of Clinical Nutrition. 2019;110(6):1449–1455.
97. Petersen KS, Davis KM, Rogers CJ, Proctor DN, West SG, Kris-Etherton PM. Herbs and spices at a relatively high culinary dosage improves 24-hour ambulatory blood pressure in adults at risk of cardiometabolic diseases: a randomized, crossover, controlled-feeding study. The American Journal of Clinical Nutrition. 2021;114(6):1936–1948.
98. Jenkins DJA, Kendall CWC, Marchie A, et al. Effects of a dietary portfolio of cholesterol-lowering foods vs lovastatin on serum lipids and C-reactive protein. JAMA. 2003;290(4):502–510.
99. Estruch R, Ros E, Salas-Salvadó J, et al. Primary prevention of cardiovascular disease with a Mediterranean diet supplemented with extra-virgin olive oil or nuts. New England Journal of Medicine. 2018;378(25):e34. Link
100. Dehghan M, Mente A, Zhang X, et al. Associations of fats and carbohydrate intake with cardiovascular disease and mortality in 18 countries from five continents (PURE): a prospective cohort study. The Lancet. 2017;390(10107):2050–2062.
101. Miller V, Mente A, Dehghan M, et al. Fruit, vegetable, and legume intake, and cardiovascular disease and deaths in 18 countries (PURE): a prospective cohort study. The Lancet. 2017;390(10107):2037–2049.
102. Johnston BC, Zeraatkar D, Han MA, et al. Unprocessed red meat and processed meat consumption: dietary guideline recommendations from the Nutritional Recommendations (NutriRECS) Consortium. Annals of Internal Medicine. 2019;171(10):756–764.
103. Burkert NT, Muckenhuber J, Großschädl F, Rásky É, Freidl W. Nutrition and health: the association between eating behavior and various health parameters: a matched sample study. PLOS ONE. 2014;9(2):e88278.
104. Alexander DD, Mink PJ, Cushing CA, Sceurman B. A review and meta-analysis of prospective studies of red and processed meat intake and prostate cancer. Nutrition Journal. 2010;9:50.
105. Waldinger RJ. “What makes a good life? Lessons from the longest study on happiness.” TED, 2015. https://www.youtube.com/watch?v=8KkKuTCFvzI, reporting the Harvard Study of Adult Development, ongoing since 1938. See also Waldinger RJ, Schulz MS, The Good Life: Lessons from the World’s Longest Scientific Study of Happiness (Simon & Schuster, 2023). Link
106. Holt-Lunstad J, Smith TB, Layton JB. Social relationships and mortality risk: a meta-analytic review. PLoS Medicine. 2010;7(7):e1000316.
107. Austad SN, Pes GM. The validity of Blue Zones demography: a response to critiques. The Gerontologist. 2025;65(12):gnaf246. See also Rosero-Bixby L. The exceptional rural longevity of Nicoya, Costa Rica, and its later-cohort attenuation. Demographic Research. 2023;49:723–736; Saito Y, Yong V, Robine J-M. The mystery of Japan’s missing centenarians explained. Demographic Research. 2012;26:239–252. Link
108. EAT-Lancet Commission 2.0. Food, planet, health: the EAT-Lancet Commission on healthy, sustainable, and just food systems. The Lancet. 2025 (estimated ~27% lower premature mortality with adoption of the planetary health diet).
109. Landry MJ, Ward CP, Koh LM, et al. Vegetarian and vegan dietary patterns and cardiovascular disease and mortality: an umbrella review of systematic reviews (commissioned by the Academy of Nutrition and Dietetics). 2024.
110. Capodici A, Fantini MP, et al. Cardiovascular and metabolic benefits of plant-based diets: an umbrella review of 48 meta-analyses. PLOS ONE. 2024;19(5):e0300711.
111. Tonstad S, Stewart K, Oda K, Batech M, Herring RP, Fraser GE. Vegetarian diets and incidence of diabetes in the Adventist Health Study-2. Nutrition, Metabolism & Cardiovascular Diseases. 2013;23(4):292–299. Link
112. Hall KD, Guo J, Courville AB, et al. Effect of a plant-based, low-fat diet versus an animal-based, ketogenic diet on ad libitum energy intake. Nature Medicine. 2021;27:344–353. Link
113. Koch CA, et al. Vegetarian or vegan diets and blood lipids: a meta-analysis of randomized trials. European Heart Journal. 2023;44(28):2609–2622. Link
114. Hooper L, Martin N, Jimoh OF, et al. Reduction in saturated fat intake for cardiovascular disease. Cochrane Database of Systematic Reviews. 2020;5:CD011737. Link
115. Evenepoel P, Claus D, Geypens B, et al. Amount and fate of egg protein escaping assimilation in the small intestine of humans. American Journal of Physiology. 1999;277(5 Pt 1):G935–943. Link
116. Carlsen MH, Halvorsen BL, Holte K, et al. The total antioxidant content of more than 3,100 foods, beverages, spices, herbs and supplements used worldwide. Nutrition Journal. 2010;9:3. Link
117. Tong TYN, Appleby PN, Armstrong MEG, 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. Link
118. Thorpe DL, Knutsen SF, Beeson WL, et al. Dietary patterns and hip fracture in the Adventist Health Study 2. American Journal of Clinical Nutrition. 2021;114(2):488–495. Link
119. 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. Link
120. Kerstetter JE, O’Brien KO, Insogna KL. The impact of dietary protein on calcium absorption and kinetic measures of bone turnover in women. Journal of Clinical Endocrinology & Metabolism. 2005;90(1):26–31. Link
121. 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. Link
122. Lousuebsakul-Matthews V, Thorpe DL, Knutsen R, et al. Legumes and meat analogues consumption are associated with hip fracture risk in Adventist Health Study 2. Public Health Nutrition. 2014;17(10):2333–2343. Link
123. Sotos-Prieto M, et al. Plant-based diets and risk of hip fracture in postmenopausal women. JAMA Network Open. 2024;7(2):e241107. Link
124. 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. Link
125. Soto-Mota A, Norwitz NG, Manubolu VS, et al. Longitudinal data from the KETO-CTA study: plaque predicts plaque, ApoB does not. JACC: Advances. 2025 (retracted March 2026). Link
126. Retraction Watch. Widely criticized keto diet study retracted. May 22, 2026. Link
127. 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. Link
128. Fung TT, van Dam RM, Hankinson SE, Stampfer M, Willett WC, Hu FB. Low-carbohydrate diets and all-cause and cause-specific mortality. Annals of Internal Medicine. 2010;153(5):289–298. Link
129. Seidelmann SB, Claggett B, Cheng S, et al. Dietary carbohydrate intake and mortality: a prospective cohort study and meta-analysis. The Lancet Public Health. 2018;3(9):e419–e428.
130. 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. BMJ. 2020;370:m2412. Link