There’s a book making the rounds in carnivore circles called Why Vegans Have Smaller Brains and How Cows Can Reverse Climate Change, by Anita Tagore, Alison Morgan and David Ellis. Anthony Chaffee gave it a two-part interview. Shawn Baker had them on. Judy Cho hosted all three. It’s been passed around social media with confident captions about brain atrophy and B12 and DHA.

So let’s go to the source.

The title comes from a single 2008 study out of Oxford. Here is Anita Tagore describing it, twice, to two different hosts:

“What this study shows is that they’ve followed a group of middle-aged and elderly people over five years, they’ve noted what they eat, they’ve done blood tests to look at various markers, and they’ve done brain scans at the beginning and end. And what they found is that the less animal food that you have in your diet, the more your brain shrinks with age. So the vegans in this study had the most brain atrophy, the vegetarians next, and the meat eaters the least.”

That study is Vogiatzoglou et al., Vitamin B12 status and rate of brain volume loss in community-dwelling elderly, published in Neurology in 2008 (PMID 18779510).

It followed 107 volunteers aged 61–87 in Oxford for about five years with annual MRI. At baseline it measured plasma total B12, holotranscobalamin, methylmalonic acid, homocysteine and serum folate. It found that people in the lowest tertile of B12 status lost brain volume fastest, with an odds ratio of 6.17 (95% CI 1.25–30.47).

It did not assess diet. There were no vegans in it. There were no vegetarians in it. There were no meat-eaters in it. Not as groups, not as a variable, not at all. Participants were sorted by a blood marker, not by what was on their plate. The paper has no dietary exposure variable at all.

The famous “six times” figure that gets repeated by Valerie Smith and attributed to Ben Bikman (“those who ate a meat-free diet were found to be suffering from 6X the rate of brain shrinkage”) is the odds ratio comparing the lowest to highest B12 tertile. It has nothing to do with diet. There is no Oxford study showing meat-free eaters had six times the brain shrinkage, because nobody in that study was asked what they ate.

That’s the whole trick. A B12 biochemistry paper gets a vegan costume put on it, and the costume becomes a book title.

Before we go further: what the evidence actually says about vegans

If you eat a well-planned vegan diet, nothing in this book applies to you. And the professional position isn’t grudging.

The Academy of Nutrition and Dietetics’ current position paper (approved January 2025, in effect until 2032, built on systematic reviews) holds that “in adults, appropriately planned vegetarian and vegan dietary patterns can be nutritionally adequate and can offer long-term health benefits such as improving several health outcomes associated with cardiometabolic diseases.” Its 2016 predecessor was blunter about which ones: vegetarians and vegans are “at reduced risk of certain health conditions, including ischemic heart disease, type 2 diabetes, hypertension, certain types of cancer, and obesity.”

That’s the baseline. Not tolerable if you’re careful. Not adequate with effort. Reduced risk of the diseases that actually kill people in wealthy countries. The large cohorts behind that (EPIC-Oxford and Adventist Health Study-2) are observational, with the usual confounding caveats, and I’ll apply that same skepticism to the book’s evidence later. But it’s a substantial, consistent body of it, and it points the opposite direction from the cover of this book. We’ve covered it in more depth on the Adventist studies and what the longest-living populations actually eat.

So: a person eating a varied plant-based diet, taking B12, and eating enough food has no reason to be alarmed by any of this. That’s the conclusion, stated at the top rather than buried at the bottom.

What does “well-planned” mean for a vegan diet?

Throughout this post, “a vegan diet” means reliable B12, genuine variety across whole plant foods, and enough calories.

That isn’t a special exemption carved out for veganism. It’s the baseline every way of eating gets. Nobody evaluates omnivorous eating by pointing at a man living on hot dogs and beer, and nobody should. Every diet has nuance and every diet has failure modes. The Institute of Medicine already tells omnivores over fifty to take supplemental B12. The authors of this book cannot say what the nutrient requirements of their own recommended diet are. More on that below, in their own words.

And I’m not going to soften the other half, because softening it would be both dishonest and dangerous: an unsupplemented vegan diet is genuinely risky, and in pregnancy and infancy it can cause real and sometimes permanent harm. That is true, it is in this post, and anyone who tells you otherwise is not doing you a favour.

But notice that this is where the book’s argument quietly changes shape. Its evidence is about deficient people: B12-deficient elderly in the 2008 study, B12-deficient mothers in the case reports. Its title is about vegans as a category. Those are not the same set, and the argument only functions if you don’t catch the substitution. A vegan who supplements is not in the data the book cites. Neither, for that matter, is a vegan of any kind.

Read every concession below against that. “Unsupplemented vegans are at risk” is not a concession about veganism. It’s a statement about deficiency, which is the only thing anyone here has actually measured.

The tell: when pressed, they retreat to babies

Here’s what makes this hard to write off as an honest mistake.

In her long solo interview with the dentist who hosts “Dr Abs,” Tagore describes the same study accurately: B12 tertiles, brain scans, no diet groups. She knows what it says. The false version appears in the soundbites, where it does the rhetorical work; the accurate version appears when she’s speaking carefully at length.

And when Judy Cho asked her directly whether vegans actually have smaller brains, and why the book carries that title, watch where she goes:

“There is plenty of evidence that diet does affect the size of your brain. The best evidence is in babies who’ve been breastfed exclusively by vegan or vegetarian mothers.”

That’s the concession. The “best evidence” isn’t the adult study the title is built on. It’s a set of infant case reports. And she adds, unprompted: “we’ve also chosen the title because it’s controversial and gets attention.”

On Chaffee’s show she was even more explicit about the origin: a friend on the phone suggested it, she thought “vegetarians” would offend too many people, so she swapped in “vegans,” “because we decided we wanted the attention.”

So let’s take the babies seriously, because that part is real.

The infant B12 cases: real, serious, and not about veganism

There is a documented, published literature on severe brain atrophy in exclusively breastfed infants of B12-deficient mothers. Case reports come from Turkey, India, Switzerland, Japan and elsewhere. The atrophy can be dramatic on MRI. And here is the important part: it is often completely reversible with B12 treatment if caught early.

Tagore describes this accurately. This is the strongest nutritional point in the entire book, and it deserves to be stated plainly rather than waved away: an unsupplemented vegan mother can cause serious, sometimes permanent neurological harm to her breastfed infant. That is not carnivore propaganda. That is real.

But notice what the cases actually show. They are about maternal B12 deficiency, not about veganism. The same picture appears in infants of mothers with pernicious anaemia, with malabsorption, after bariatric surgery, and in undernourished non-vegan populations in low-income settings. B12 deficiency is the cause. Veganism is one of several routes to it, and the only route that comes with a free, cheap, universally recommended fix.

Every major dietetic body (the Academy of Nutrition and Dietetics, the British Dietetic Association, the Canadian and Australian equivalents) treats reliable B12 as mandatory for vegans, and especially in pregnancy and lactation. A vegan mother who follows standard guidance is not the mother in these case reports.

Which means the honest version of the argument is: unsupplemented vegan mothers can harm their infants’ brains. That’s true. It’s also a sentence with an obvious solution in it, which is presumably why it didn’t make the cover.

For the full picture on B12, including sources, testing, dosing, and why “it’s not in plants” is a weaker argument than it sounds, see our B12 article.

The B12 argument, done properly

Two things collapse the book’s B12 case as an argument against veganism specifically.

First: the intervention that actually worked in trials was a supplement, not meat.

The Oxford group didn’t stop at observation. The VITACOG trials (Smith et al. 2010, PLoS One; de Jager et al. 2012; Douaud et al. 2013, PNAS) randomised people with mild cognitive impairment to B-vitamin supplements and found slowed brain atrophy in those with elevated homocysteine. Tagore even cites the follow-up finding correctly: Jernerén et al. 2015 in the American Journal of Clinical Nutrition showed the B-vitamin benefit appeared mainly in people with higher omega-3 status.

She’s right about the study. But look at what it means. The thing that slowed brain atrophy in a randomised trial was a pill. Not a ribeye. Not organ meat. Not the Sapiens Diet. If your evidence that animal foods protect the brain is a trial where the intervention was a B-vitamin capsule, you have accidentally made the case for supplementation.

Second: modern supplemented vegans do not have low B12.

The book’s model of a vegan is someone in 1985 who has never heard of a supplement. Unsupplemented vegans genuinely are at high risk, and the Pawlak review they cite is real. But in the German RBVD cohort, 97% of vegans supplemented (92% specifically with B12), and their B12 status was fine. A 2023 study from that same group is titled almost on the nose around adequate B12 status in supplemented vegans. Supplemented vegans routinely match or exceed omnivores.

And there’s a slightly awkward corollary the book never addresses: if low B12 shrinks brains and supplemented vegans often have higher B12 than meat-eaters, then by the book’s own logic, the arrow points the other way.

The double standard hiding inside “well-planned”

Every version of this argument runs through a phrase that sounds like a fair caveat and functions as a loaded one: a vegan diet is fine if well-planned.

The qualifier is accurate. The asymmetry in how it gets applied is the problem. Every way of eating requires planning, and B12 in particular is not a vegan problem. It’s an absorption-and-intake problem that happens to have one dietary route into it.

Start with the recommendation almost nobody in this debate quotes. The Institute of Medicine’s Dietary Reference Intakes advise that adults aged 51 and over should obtain most of their B12 from fortified foods or supplements, because a substantial share of older adults (the IOM’s 1998 DRI report puts it at 10–30%) cannot release B12 bound to food protein. The NIH Office of Dietary Supplements, which notes that deficiency affects between 3% and 43% of older adults, puts it to consumers without hedging: “People over 50 should get most of their vitamin B12 from fortified foods or dietary supplements.”

Read that again. The US national nutrition authority tells every adult over fifty, whatever they eat, to get B12 by precisely the mechanism vegans use. For a large slice of the population meat is the unreliable source and crystalline B12 is the reliable one. Stomach acid declines with age, B12 bound to animal protein never gets released, and the supplemental form doesn’t need releasing.

The prevalence figures follow from that. Deficiency runs around 25% in over-60s. In the Trinity-Ulster cohort of 3,299 adults aged 60 and over who were not taking B12 supplements, atrophic gastritis was present in 15%, and 38% of that group were B12 deficient, against 21% of proton-pump-inhibitor users and 15% of controls. Thirty-seven percent of the whole cohort had used PPIs for six months or longer. Then add long-term metformin, pernicious anaemia, bariatric surgery, H. pylori, and heavy alcohol use.

And here is the finding that ought to end the argument: in participants with atrophic gastritis, regular consumption of fortified foods was inadequate to restore normal B12 status. Eating more steak was never going to.

None of this means vegan diets carry no B12 risk. Unsupplemented, the risk is high, and I said so above. The point is narrower and worse for the book: B12 deficiency is a population-wide problem with a single universal solution, and the book has repackaged it as a vegan behaviour problem. Dean Sherzai, who treats dementia for a living, disposes of it in a word: deficiency is “actually ubiquitous.”

Now apply the standard symmetrically. In four long interviews, nobody once asks whether the Sapiens Diet is well-planned. Seventy percent of calories from animal fat and near-zero plants: what’s the vitamin C position? Folate? Potassium? Magnesium? Fibre?

Tagore has an answer, and it’s the most revealing thing she says in any of these conversations. On low-carbohydrate diets, she argues, requirements for vitamin C and iodine are lower than the guidelines state, because the guidelines “are all in the context of a standard western diet, which is high carb.” And then:

“So we don’t know what the actual amounts that we need on a low carb diet.”

That is the author of Why Vegans Have Smaller Brains conceding that the nutrient requirements of her own recommended diet are unknown, in the same conversation where she tells vegans theirs shrinks brains. One diet gets “we don’t know.” The other gets a book title.

DHA: where the book overclaims, and where honestly nobody knows

DHA deserves its own article and I’ll write one, because the evidence here is genuinely messier than either side admits. But three of the specific claims made in these interviews are checkable right now.

Claim 1: DHA isn’t only in fish; it’s in beef and the organ meats too.

This is quantitatively wrong, and not by a little.

Analyses of US beef find total long-chain omega-3 (EPA + DPA + DHA combined) running roughly 2–19 mg per 100 g in grain-finished beef and 5–33 mg per 100 g in grass-fed, and the research notes these are “primarily as EPA + DPA,” with only trace amounts of DHA specifically. The reason is rumen biohydrogenation: cattle destroy most unsaturated fatty acids before they reach the tissue.

Compare oily fish: salmon delivers roughly 0.9–1.4 g of DHA per 140 g portion. That’s on the order of a hundredfold difference, and beef’s small omega-3 contribution isn’t even mostly DHA.

Beef is not a DHA source. Saying it is doesn’t become true because you farm cattle.

As for her “the best source is the brain of ruminant animals”: evolutionarily interesting, practically irrelevant, and worth noting that bovine brain and spinal cord are classed as specified risk material and banned from the food chain across the UK and EU following BSE. Tagore is British and runs a cattle farm. This is not a food she can recommend.

Claim 2: “The conversion from ALA can grind to a halt completely with high omega-6.”

The conversion is genuinely poor: low single-digit percentages at best, and omega-6 intake competes for the same desaturase enzymes. That much is fair. “Grinds to a halt completely” is an overstatement of a real limitation.

But it’s also beside the point, because DHA isn’t animal-exclusive. Algal oil is where fish get their DHA in the first place. Algal DHA supplements raise blood DHA comparably to fish oil in randomised trials. You can skip the fish and go to the source.

Claim 3, and here’s the honest part: does any of this matter for cognition?

Vegans do have lower plasma and red-cell DHA. That’s not in dispute. What is in dispute, and what neither the book nor its loudest critics will tell you, is whether that translates into anything.

Because when you actually test it, the trials come back empty. The Cochrane review of omega-3 supplementation for cognitive decline and dementia prevention (Sydenham, Dangour & Lim, 2012) pooled three high-quality RCTs in 3,536 people aged 60+ and found no benefit to cognitive function over 6 to 40 months. The authors’ own conclusion: “The available trials showed no benefit of omega-3 PUFA supplementation on cognitive function in cognitively healthy older people.” Later reviews in MCI and Alzheimer’s are mixed at best: some signal for hippocampal volume in early-stage disease, negligible effects in others.

So the state of play is: lower DHA in vegans, uncertain significance, and an RCT literature that has repeatedly failed to show that raising DHA improves cognition. That cuts against the book’s confident causal story at least as hard as it cuts against anyone else’s.

I’m not going to tell you DHA doesn’t matter. I don’t know that, and neither does anyone else. What I’ll tell you is that anyone claiming certainty here, in either direction, is ahead of the evidence, and that the cheap, sensible move for a plant-based eater is an algal DHA supplement rather than a dietary philosophy.

What actual brain doctors say

The book’s medical authority is a retired GP with a one-year MSc in nutrition, who, on the Dr Abs podcast, said this about why she did it:

“I have to admit, part of the reason why I did the masters was just to give myself credibility to write a book on nutrition. It wasn’t to actually learn what was on the course.”

Fine. Let’s compare that to people who actually treat brain disease for a living.

What do neurologists Dean and Ayesha Sherzai say about plant-based diets and the brain?

Dean Sherzai, MD, PhD is a behavioural neurologist and neuroscientist, trained in neurology at Georgetown with fellowships in neurodegenerative disease and dementia at the NIH and UC San Diego, plus a PhD in healthcare leadership. Ayesha Sherzai, MD is a vascular neurologist with dual residencies in preventive medicine and neurology at Loma Linda, a fellowship in vascular neurology and epidemiology at Columbia, and a Master of Advanced Sciences in clinical research from UCSD; she received the American Heart Association’s Trudy Bush Fellowship in 2015 for research on cardiovascular disease in women.

They co-directed the Brain Health and Alzheimer’s Prevention Program at Loma Linda University, then the equivalent programme at Cedars-Sinai, and are now at Charles R. Drew University of Medicine and Science, where Dean is Executive Director of Clinical Research and Professor of Neurology and Neuroscience. They wrote The Alzheimer’s Solution and The 30-Day Alzheimer’s Solution, built the NEURO framework (Nutrition, Exercise, Unwind, Restore, Optimize), and appear as expert commentators in Netflix’s You Are What You Eat: A Twin Experiment.

They are, in other words, exactly the specialists this book is trying to talk over. They eat a whole-food plant-based diet themselves, and notably, per Viva!’s profile of them, not for ethical or environmental reasons but because they read the evidence as pointing there for brain health. They are not advocates who became neurologists. They are neurologists whose reading of the literature made them plant-based.

Here’s Dean Sherzai on the two nutrients the book builds its entire case on:

“B12 deficiency, yes, there’s a lot of it in plant-based but it’s actually ubiquitous… You just have to be aware of it, make sure that you eat a complex diet… and be aware of your levels.”

“Omega-3, be aware of it, eat foods that are rich in omega-3s like chia and flax seeds… But even then, if you’re worried, especially in certain times of life like developing brain, children, like pregnancy, or aging brain, take a supplement, algae-based.”

And on the specific question this book aims at pregnant women, Ayesha Sherzai has spoken about her own history:

“There was a time when I was pregnant and when I was having my kids, I supplemented with omega-3 fatty acids, but right now we’re taking vitamin B12.”

They are a long-term whole-food plant-based family, by their own account one that moved from lacto-ovo vegetarian toward whole-food plant-based over time, so I won’t claim more about the pregnancies themselves than she says. What she says is enough: she took the omega-3 supplement then, the family takes B12 now. That is exactly the protocol she and her husband give patients, and exactly the protocol Why Vegans Have Smaller Brains exists to tell prospective mothers is inadequate.

That’s the position two dementia neurologists hold. Not “plants shrink your brain.” Not “you must eat ruminant fat.” Just: know your levels, supplement B12, take algal omega-3 at the life stages where it matters most. The book’s central claim is that this is impossible. Two people who run a dementia prevention programme live it and recommend it, and, as covered above, they classify B12 deficiency as a general clinical problem rather than a vegan one.

What did Dean Ornish’s Alzheimer’s trial find?

Ornish ran the trial that carnivore advocates most need to explain away: the first randomised controlled trial of intensive lifestyle change in mild cognitive impairment and early Alzheimer’s disease, published in Alzheimer’s Research & Therapy in June 2024 (PMID 38849944, co-authored with Rudolph Tanzi, Steven Arnold, Miia Kivipelto, Rob Knight and Dorene Rentz; this is not a fringe author list).

Fifty-one participants aged 45–90, randomised 1:1, 20 weeks. The intervention: a whole-food, minimally processed plant-based diet low in refined carbohydrates, with selected supplements including B12, plus moderate exercise, stress management and support groups.

Results after 20 weeks: statistically significant differences favouring the intervention on CGIC (p=0.001), CDR-SB (p=0.032) and CDR Global (p=0.037), with ADAS-Cog borderline (p=0.053). Plasma Aβ42/40 ratio improved in the intervention group and decreased in controls. Extended 40-week results have since been published.

Now the limitations, because they’re real and I’m not going to hide them: 51 people is small. Twenty weeks is short. The intervention was multimodal, so you cannot isolate diet from exercise, stress management or social support. It was unblinded. Ornish and colleagues have financial interests in Ornish Lifestyle Medicine, which is disclosed in the paper. This is a phase 2 signal, not a settled result.

But notice what it is: a randomised controlled trial where a plant-based diet plus B12 supplementation was associated with improved cognition and function in early Alzheimer’s, while the usual-care controls declined. Whatever you think of its size, it is more than the carnivore side has for brain health, which is zero randomised trials and a pile of testimonials.

The affirmative case, which the book never has to answer

Everything so far has been defensive: this study didn’t say that, that citation drops its caveat. But the positive case is the more interesting half, and it runs through mechanisms the book’s own argument depends on.

Do vegans have less inflammation?

Chronic low-grade inflammation is increasingly treated as a driver of neurodegeneration, and plant-based eaters have less of it.

A systematic review and meta-analysis in Scientific Reports (Menzel et al. 2020, PMC7730154) pooled 21 studies and found vegan diets associated with lower C-reactive protein than omnivorous diets, with a mean difference of −0.54 mg/L (95% CI −0.79 to −0.28, p < 0.0001). Those were cross-sectional, so a 2026 meta-analysis in Nutrition, Metabolism and Cardiovascular Diseases went looking for randomised evidence and found seven RCTs covering 541 participants; plant-based dietary patterns significantly lowered CRP versus omnivorous diets. The authors rate the certainty as low given heterogeneity and bias risk, and I’ll flag that rather than bury it, but the observational and randomised evidence point the same way.

The plausible drivers are the ones the carnivore literature treats as liabilities: fibre, polyphenols, flavonoids and carotenoids, which is a topic we’ve covered separately. Anthony Chaffee’s “plants are trying to kill you” and the actual inflammatory biomarker data are not compatible.

Is what is good for the heart good for the brain?

This is the mechanism most people underrate. What damages arteries damages the brain: the vasculature feeding it is the same vasculature. Midlife hypertension and elevated LDL are among the most firmly established modifiable dementia risk factors, and vascular contributions to cognitive impairment are a major share of late-life decline, including in people who also have amyloid pathology.

So the cholesterol chapter isn’t a separate argument from the brain chapter. It’s the same argument. A dietary pattern that raises ApoB and blood pressure is not neutral for the brain, whatever it does for your triglyceride-to-HDL ratio, and this is the thread running through our cholesterol article. It is also, not coincidentally, why Ornish’s intervention and the Sherzais’ NEURO framework both spend as much attention on blood pressure and lipids as on any single nutrient.

Does the methionine in animal protein affect homocysteine and amyloid?

Here’s where it gets awkward for them.

The entire brain-atrophy argument in this book runs through homocysteine: low B12 → elevated homocysteine → accelerated brain volume loss. Fine. Now ask where homocysteine comes from. It’s a metabolite of methionine, and methionine comes disproportionately from animal protein, and plant proteins are lower in it, which is why studies comparing the groups have found vegans with measurably lower methionine intake.

In animal models the downstream effects are not subtle. Feeding mice a 2% high-methionine diet for nine weeks produced learning and memory impairment, hippocampal and cortical neuronal damage, elevated serum homocysteine, and increased brain Aβ1-40 and Aβ1-42, along with raised APP and BACE1 and reduced amyloid-clearing enzymes (Pi et al. 2021, Behavioural Neurology, PMC8046555). Diet-induced high homocysteine in 3xTg mice worsened all three hallmarks at once: memory deficits, amyloid, and tau. And a 2026 paper in Alzheimer’s & Dementia found late-life methionine restriction reduced amyloid deposition and neuroinflammation in aged Alzheimer’s mice.

Two honesty notes. First, the mechanism isn’t that methionine “becomes” amyloid. It’s methionine → homocysteine → amyloid and tau pathology, an indirect route. Second, it isn’t uniform: in one tauopathy model (rTg4510), methionine restriction increased tau and inflammatory markers. And all of this is rodent and mechanistic work, not human outcome data. I’m not going to do to this literature what the book did to Vogiatzoglou.

But the shape of it stands. The book selected homocysteine as its causal pathway and then declined to follow it upstream to the amino acid that generates it.

Do vegans have higher homocysteine?

Unsupplemented vegans do have higher homocysteine. That’s consistent across meta-analyses, including Niklewicz et al. (2024, Nutrition Bulletin), and I’m not going to pretend otherwise.

But homocysteine is not a fixed property of eating plants. It’s a modifiable number, and B12 is the modifier. Niklewicz’s own subgroup analysis found vegan supplement users had higher B12 and holotranscobalamin and lower MMA and homocysteine than non-supplementing vegans. Haddad et al. (1999) found supplementing vegans at 8.0 µmol/L against omnivores at 7.9, which is statistically indistinguishable. Crane et al. (1998) gave vegans 500 µg/day of B12 for two months and measured homocysteine below 5 µmol/L, which is low by any standard. In Adventist Health Study-2, serum B12, holoTC, MMA and homocysteine did not differ across diet groups on average.

I won’t overclaim this: “supplemented vegans have lower homocysteine than meat-eaters” is not an established general finding. What is established is narrower and sufficient. The single variable this book identifies as the driver of brain atrophy is one that vegans control directly, cheaply, and, when they supplement, well. The book presents that variable as an indictment of a diet. It’s an argument for a pill, and the pill costs a few cents a day.

“Meat made our brains big” and the Lamarckian error hiding inside it

The book leans hard on evolution. Dave Ellis’s 24-hour-clock metaphor: 23 of 24 hours as carnivores, agriculture in the last hour, dietary guidelines in the last ten seconds, and brain size shrinking ever since.

Three problems.

First, the meat hypothesis is not the consensus explanation. It’s one candidate among several: cooking (Wrangham), tubers and dietary carbohydrate (Hardy et al. 2015 in the Quarterly Review of Biology), fat specifically rather than muscle meat, social and ecological drivers. It is a live debate, not a settled finding.

Second, the hypothesis’s own author says the book is overreading it. The Expensive Tissue Hypothesis is Leslie Aiello’s. In 2008 she delivered the George Peabody Founder’s Lecture at Harvard, revisiting it after fifteen years of scrutiny. The Harvard Gazette’s report is worth quoting exactly. In the Gazette’s words: “Our human ancestors were not wholly carnivores” (“that would be silly,” said Aiello), and she “does not argue that meat-eating caused bigger brains,” arguing “just that it made bigger brains possible.”

The same article notes that ETH now “has theoretical competitors,” that some researchers point to bipedalism or reduced muscle mass instead, and that “ETH doesn’t hold true for all animals, including birds and bats.” Navarrete, van Schaik and Isler found no brain–gut size correlation across mammals in Nature in 2011.

The originator of the hypothesis says our ancestors weren’t carnivores, calls the idea silly, and explicitly declines to claim causation. The book claims causation and builds a hypercarnivore diet on it.

Third, and this is the load-bearing error nobody names: the argument is Lamarckian.

Even granting everything: meat enabled brain expansion over roughly two million years. That is Darwinian selection acting on populations. It means individuals with certain gut-brain configurations out-reproduced those without, given an energy-dense diet.

It does not mean that you, eating a steak in 2026, grow a bigger brain. It does not mean your children will have bigger brains. You will pass on the genes you have. The only way to shift brain-size distribution in a population is differential reproduction, with some lineages having fewer descendants than others.

“Meat made our brains big over two million years, therefore you personally need meat for your brain today” is a category error dressed as ancestral wisdom. It’s the same error as “our ancestors evolved lactase persistence, therefore drink milk.” Evolutionary history explains why a trait exists. It does not prescribe your dinner.

And on the brain-shrinkage-since-agriculture point: it’s real but contested. DeSilva et al. (2021) dated the decline to ~3,000 years ago; Villmoare and Grabowski rebutted the timing and dataset in 2022. Leading explanations include reduced body size, self-domestication and externalised cognition, and in several analyses the decline predates agriculture, which is inconvenient for a story blaming grain. Within the normal human range, brain volume correlates with cognition only weakly (r ≈ 0.24–0.4). Bigger is not straightforwardly better.

There’s also positive evidence for plant eating in hominins that the book skips: starch granules in Neanderthal dental calculus from Shanidar and El Sidrón, the AMY1 amylase copy-number expansion, and recent isotope work on the Taforalt Iberomaurusians in Morocco indicating a heavily plant-based pre-agricultural diet.

The cholesterol chapter

On Judy Cho’s show, Tagore and Ellis run through the standard set. Worth taking one at a time.

“Cholesterol is the fire engine at the fire, not the cause of the fire.” This analogy has been circulating for years and it fails at the same place every time: fire engines don’t show up before the fire. LDL does. The causal evidence for LDL and ApoB in atherosclerosis runs through Mendelian randomisation (people genetically randomised to lifelong low LDL get less heart disease), through familial hypercholesterolaemia (lifelong high LDL, early heart disease, no other explanation), through a dose-response gradient across dozens of LDL-lowering RCTs with different drug mechanisms, and through the European Atherosclerosis Society consensus statement (Ference et al. 2017). It’s about as settled as nutrition-adjacent science gets. Full treatment in our cholesterol article.

“Triglyceride-to-HDL ratio matters more than LDL.” Trig:HDL is a decent proxy for insulin resistance and worth knowing. It does not erase ApoB. These are not competing claims: you can improve insulin sensitivity and raise your atherogenic particle count, which is precisely what happens to a lot of people on this diet.

“In older people, higher cholesterol is associated with lower all-cause mortality.” A real observation with a well-understood explanation: reverse causation. Frailty, cancer, and undiagnosed illness lower cholesterol. Sick people have low cholesterol because they’re sick.

“It’s glucose that damages cholesterol, and carbs are the new smoking.” The comparison to smoking is theatre. And it flattens the distinction between refined and whole carbohydrate that does most of the actual work. The cleanest counterexample is the Tsimane of Bolivia: Kaplan et al. in The Lancet (2017) CT-scanned 705 adults and found the lowest coronary artery disease of any population ever measured, roughly five times less than the US, on a diet that’s about 72% carbohydrate.

On animal fat specifically, see our article on fat and animal fat versus coconut oil. The “just stop vegetable oil” chapter gets its own treatment in our seed oils piece.

Fiber, the colon, and the breastfed baby

Tagore makes specific, checkable claims here, and this is a case where she cites something real and then strips off the caveat.

On colon cancer, she describes “a Cochrane review of controlled trials where they increased fiber for 2 to 8 years, and that didn’t lower the risk of colon cancer, and paradoxically after four years there was a slight increase.”

She’s referring to the Cochrane review on dietary fibre and recurrent colorectal adenomas (Asano & McLeod 2002, updated by Yao et al. 2017, CD003430). And the numbers are as she says: no significant difference for at least one adenoma (5 RCTs, n=3,641, RR 1.04), and for colorectal cancer diagnoses the results “favoured the control group over the dietary fibre group” (2 RCTs, n=2,794, RR 2.70, 95% CI 1.07–6.85).

What she leaves out is the sentence Cochrane wrote immediately afterward:

“However, these results may be unreliable and should be interpreted cautiously, not only because of the high rate of loss to follow-up, but also because adenomatous polyp is a surrogate [outcome].”

Low-quality evidence. Two trials. A confidence interval running from 1.07 to 6.85, which is another way of saying “we have almost no idea.” And a surrogate endpoint measured over three to four years for a cancer that develops over decades, in trials that mostly tested fibre supplements rather than dietary patterns.

Set against that: Reynolds et al. in The Lancet (2019), covering ~135 million person-years across 185 prospective studies and 58 clinical trials, found the highest fibre consumers had a 15–30% lower all-cause and cardiovascular mortality and 16–24% lower incidence of coronary heart disease, stroke, type 2 diabetes and colorectal cancer. WCRF/AICR grade fibre-rich whole grains as protective against colorectal cancer. Meanwhile IARC classifies processed meat as a Group 1 carcinogen and red meat as Group 2A, a fact that goes unmentioned in a book warning about colon cancer. More on that in our piece on the meat and bowel cancer link.

On constipation, both Tagore and Judy Cho lean on a Singapore study. That’s Ho et al. 2012 in the World Journal of Gastroenterology (PMID 22969234): 63 patients with idiopathic constipation, all placed on a no-fibre diet for two weeks, then allowed to self-select their intake. No proper control group. Non-randomised. A highly selected population of people already constipated.

Some constipated people improve on less fibre. That’s clinically useful and I’ve no quarrel with it. It does not make fibre “a natural junk food” for everyone else.

And then there’s the baby argument, which both of them treat as a knockdown. Judy Cho: “he never had fiber in that one year and he massively grew… there’s no fiber in that.” Tagore: “breastfed babies have zero fiber in their diet, they have no problems with constipation, they don’t need fiber at all.”

Human milk oligosaccharides are the third most abundant solid component of human breast milk, after lactose and lipids. More than 200 distinct HMO structures have been identified. They run roughly 5–20 g/L in mature milk and 20–25 g/L in colostrum. A breastfed infant consumes up to around 10 g per day of them.

They are indigestible by the infant’s own enzymes. They pass intact through the upper GI tract and reach the colon, where they act as selective substrates for beneficial bacteria, principally Bifidobacterium.

That is fibre. It is functionally, definitionally fibre: indigestible carbohydrate that reaches the colon and feeds the microbiome. Breast milk is loaded with it, at metabolic expense to the mother, which in evolutionary terms is about as strong a signal of importance as you can get.

And the follow-on argument makes it worse for them. Tagore points out that breastfed and formula-fed infants have different short-chain fatty acid profiles. Correct, and SCFAs are produced by bacterial fermentation of exactly these indigestible carbohydrates. She has cited the mechanism by which fibre produces its benefits as evidence that fibre is unnecessary.

One more: human milk is roughly 40–45% of calories from carbohydrate, as lactose. For a book arguing we evolved to be ketogenic, the food evolution designed specifically for the fastest-growing human brain is a high-carbohydrate, high-fibre liquid.

On lectins, oxalates and the broader “plants are trying to kill you” framework, including Chaffee’s “Brussels sprouts contain 130 carcinogens,” see our article on plant toxins and antinutrients. On the beta-carotene conversion argument they raise, see here.

The clinical claims

Tagore says she’d recommend this diet “to absolutely everybody.” Chaffee claims MS reversal and tumour regression. Baker’s guests describe reversed autoimmune disease.

There is no randomised controlled trial of a carnivore or near-carnivore diet with hard clinical endpoints. Not one.

The largest dataset is Lennerz et al. (2021), a self-reported online survey of 2,029 people recruited from carnivore-diet social media. Self-selected. Self-reported outcomes. No control group. And structurally incapable of seeing the people who tried it, felt terrible, and left, which is the entire population you’d need to hear from.

The Hungarian Paleomedicina clinic on which the Sapiens Diet is explicitly based publishes mainly uncontrolled case reports and small case series, often in low-tier venues, claiming reversal of cancer and inflammatory bowel disease from single cases. That is the weakest possible evidence for the strongest possible claims.

Low-carbohydrate diets can drive type 2 diabetes remission, and that’s real. But so can other routes: DiRECT achieved 46% remission at one year using a low-calorie liquid formula diet. Weight loss is doing much of the work, and there are many ways to lose weight. More in our weight loss article.

On “leaky gut”: intestinal permeability is real and measurable. “Leaky gut syndrome” as a unified cause of autoimmune disease is not established. Tagore’s own evidence is a PEG-400 test she ran on herself before and after four weeks, an n=1, unblinded, uncontrolled self-experiment using an assay that is no longer commercially available. Commercial zonulin ELISAs, meanwhile, have been shown not to actually measure zonulin.

Two small things worth checking

Tagore states that brain atrophy runs about 1% per year from the mid-20s, accelerating to 2% per year in the 50s. The figures in the neuroimaging literature are considerably lower for midlife: whole-brain volume loss is generally placed nearer 0.2–0.5% a year through middle age, reaching and exceeding 0.5% only in later decades. I’d treat her numbers as inflated, though I’d encourage you to check the primary sources rather than take my word for it.

And “the brain is 60% fat,” repeated by everyone in these interviews as though it settles something, refers to dry weight. The brain is roughly three-quarters water. More to the point, the brain synthesises its own cholesterol de novo behind the blood-brain barrier, and dietary cholesterol doesn’t cross it. “Your brain is made of fat, therefore eat fat” has the same logical structure as “your bones are made of calcium, therefore eat chalk.”

The environmental half, briefly

The subtitle claims cows reverse climate change. Three facts, then I’ll point you elsewhere.

The ice-core record shows atmospheric methane stable at roughly 700 ppb for millennia and now around 1,900 ppb, unprecedented in 800,000 years. That alone refutes Alison Morgan’s claim that we have ruminants today “in similar numbers to the numbers we have today, wild and domesticated.” Bar-On, Phillips & Milo (2018, PNAS) found livestock biomass vastly exceeds wild mammal biomass; humans and livestock together are ~96% of mammalian biomass on Earth. And Poore & Nemecek (2018, Science), using consistent life-cycle boundaries across ~38,700 farms (which answers their complaint about inconsistent accounting), still finds beef roughly 25 times more emissions-intensive per gram of protein than tofu.

One new claim from the Judy Cho interview deserves flagging because it’s the most dubious thing either of them said: Morgan asserts that 45% of UK soy imports go to vegan and vegetarian meat alternatives, and that this rather than beef drives deforestation. This is not consistent with UK soy footprint data: the overwhelming majority of soy imported into the UK goes into animal feed, primarily for poultry and pigs. I’d treat that figure as unsupported pending a source.

We’ve handled the rest at length elsewhere: regenerative farming, rewilding versus regenerative, crop deaths, the pesticide argument, and the “livestock turn food we can’t eat into protein” claim.

The true statements in the book, and why not one of them reaches the title

A book can contain fifty accurate claims and one false title, and the fifty do not repair the one.

That’s card stacking, and it’s the standard architecture of this genre. You assemble real citations, real mechanisms, real concerns (none of them individually objectionable), and let the accumulated weight of “well, that part’s true” carry a conclusion that none of it supports. The reader comes away thinking the case was cumulative when it was only long.

So I’m not interested in grading the book. What follows isn’t credit to the authors; it’s the handful of things in the underlying science that a reader deserves to have straight no matter who’s saying them, each with the reason it doesn’t get you within a mile of the cover.

Unsupplemented vegans are genuinely at risk of B12 deficiency, and the infant cases are real. Plant-based advocates should say this more loudly, not less. But the title isn’t “unsupplemented vegans have smaller brains.” The condition is prevented by a supplement every dietetic body already mandates, and the same deficiency arises from pernicious anaemia, malabsorption, bariatric surgery, metformin and PPIs.

The pregnancy claim is the book’s most aggressive, and it’s the one where professional guidance is least ambiguous. A well-planned vegan pregnancy is endorsed as safe by the 2016 Academy of Nutrition and Dietetics position paper, by British Dietetic Association and NHS-aligned midwifery guidance (which puts it plainly that “well-planned vegan and plant-based diets are safe during pregnancy and lactation”), and by the expert panel of the Scientific Society for Vegetarian Nutrition, whose review concluded a completely plant-based diet “is suitable during pregnancy, lactation, infancy, and childhood, provided that it is well-planned.” A 2024 analysis in the Journal of the Academy of Nutrition and Dietetics went further and modelled it directly, finding that vegan, lacto-vegetarian, ovo-vegetarian and pescatarian adaptations of the USDA Healthy Vegetarian Dietary Pattern can all be nutritionally adequate for pregnancy with supplementation.

And the claim has been tested where it counts. An analysis from the Avon Longitudinal Study of Parents and Children compared 78 vegetarian and 2,144 omnivorous mothers in early pregnancy, and 91 versus 2,552 in late pregnancy, then followed the children. The vegetarian mothers had lower blood concentrations of docosahexaenoic acid, arachidonic acid and cobalamin in both early and late pregnancy, precisely the exposure this book says damages developing brains. The paper’s title is its finding: Vegetarian Diet during Pregnancy Is Not Associated with Poorer Cognitive Performance in Children at Age 6–7 Years. After controlling for confounders, there was no association with poorer neurocognitive development.

That’s vegetarian rather than vegan, and it’s one study. But it is the closest thing anyone has to a direct test of the book’s central claim about mothers and children’s brains, it ran the exposure the book nominates, and it came back null.

Well-planned here means a slightly longer checklist than B12 alone: reliable B12, iodine, vitamin D, iron, choline, zinc, algal DHA, and enough calories.

Before anyone reads that list as an indictment, look at what every omnivorous pregnant woman is handed. Folic acid, from before conception through the first trimester. Universal, non-negotiable, and a supplement. Vitamin D daily, per NHS guidance, for every pregnant woman in Britain regardless of diet. Iron at 27 mg a day, the pregnancy RDA, for everyone. Iodine, where deficiency has been documented in pregnant women in the south-east of England eating ordinary mixed diets. Choline, which most pregnant women fall short of on any dietary pattern. DHA, where omnivorous women get an advisory telling them which fish to avoid because of mercury.

There is no such thing as an unplanned pregnancy diet that anyone recommends. Every pregnant woman on earth is told to take supplements, watch specific nutrients, and see her doctor. Nobody calls that regime “well-planned omnivorous eating,” because the qualifier only ever gets attached to the diet somebody wants to make sound precarious. The vegan list differs by exactly one item that moves from optional to mandatory: B12. Everything else on it is the same list, handed to the same women, by the same clinicians.

So the advice is identical in shape whatever you eat. Cover your bases, take the prenatal, get your bloods checked, and talk to your doctor or midwife. A pregnant vegan taking B12 is not the woman in this book’s case reports, and never was.

Now the birth weight data, which is real but far weaker than the use they put it to. A 2024 meta-analysis found strict vegetarian diets associated with small-for-gestational-age births at a pooled odds ratio of 2.71, with a 95% confidence interval running from 1.24 to 5.95. A 2020 Journal of Perinatology analysis reported RR 5.9, CI 1.2 to 21.8. Intervals that wide are the statistical equivalent of a shrug. The 2024 Nutrients systematic review concluded the available studies had insufficient power and serious limitations.

More to the point, the specialist literature already separates the two populations the meta-analyses blur together. Baroni and colleagues, writing on vegan nutrition for mothers and children, state it directly: “The average birthweight of infants born to vegan mothers does not differ significantly from that of infants of omnivorous mothers. Macrobiotic vegan women, whose diets can be highly restricted in calories and nutrients, in contrast to well-planned vegan diets, give birth to infants whose weights are significantly lower than expected.” Much of the older data comes from exactly those restricted communities. Layer on the fact that vegans are leaner with lower gestational weight gain (both strong independent determinants of birth weight), and a smaller baby born to a smaller mother gets coded SGA without anything having gone wrong. The same authors note that plant-rich eating in pregnancy may be protective against pre-eclampsia.

On energy specifically: plant foods are less energy-dense, and in populations where most adults are now overweight or obese (72.4% of US adults aged 20 and over, per CDC/NCHS) that is overwhelmingly an asset. It’s one of the mechanisms by which plant-based eating helps with weight. Pregnancy is the narrow window where it cuts the other way and intake needs deliberate attention. A planning point for one life stage, not a defect in plants.

And even granting the signal in full: birth weight is not brain size, and brain size is not cognition. And as above, the one study that actually followed the children of mothers with lower DHA and B12 through to cognitive testing found no deficit. Tagore’s closing message to prospective mothers, that without animal foods their children “won’t fulfill their cognitive potential,” is not a finding. It is the book’s thesis restated as a warning to pregnant women, aimed at the group least able to shrug it off.

The Cochrane adenoma finding is real. RR 2.70 for colorectal cancer diagnoses favouring control. But Cochrane’s own next sentence calls it unreliable, on two low-quality trials with heavy loss to follow-up and a surrogate endpoint. Citing the number while dropping the caveat is not reporting a finding; it’s using one.

The Jernerén DHA–B vitamin interaction is real and cited correctly. But the intervention was a B-vitamin capsule. It is evidence for supplementation, which is the thing the book exists to argue against.

Nutritional epidemiology has real weaknesses. Food frequency questionnaires are bad, confounding is rampant, and the under-reporting research they cite is legitimate. But this one boomerangs hardest of all, because the entire book rests on an observational study its lead author described inaccurately on two podcasts.

The elderly cholesterol association is a real observation. And reverse causation explains it. Sick people have low cholesterol because they’re sick.

Run the list again and notice what’s absent. Not one item is evidence that vegans have smaller brains. Every one of them is either a solved problem, a surrogate endpoint, an argument for supplements, or a standard of evidence the authors decline to apply to themselves.

Where our side overreaches

Precision is the whole point of this post, so:

Mic the Vegan’s response video, which is right about the central claim, describes the Ornish trial as running “5 years.” It ran 20 weeks. His methane “airborne fraction” argument misunderstands how GWP is calculated (the CO₂ term already incorporates the atmospheric impulse response, including ocean and land uptake). And he sourced ruminant population figures via ChatGPT, on air, and said so. If we’re going to hold Tagore to the standard of describing studies accurately, we hold our own to it too.

More broadly: vegan diets require planning. They are not automatically optimal. EPIC-Oxford found higher haemorrhagic stroke risk in vegetarians alongside lower ischaemic heart disease. Anyone telling you a plant-based diet is bulletproof is selling something, same as the people telling you it shrinks your brain.

One study you’ll see cited at you is worth describing accurately, because almost nobody has.

Norgren et al. (JAMA Network Open, March 2026) analysed 2,157 adults from the SNAC-K cohort in central Stockholm, mean age 71.2, followed up to 15 years, and reported that higher unprocessed meat intake tracked with slower cognitive decline and lower dementia risk in carriers of APOE ε3/ε4 and ε4/ε4. Carnivore media ran with it, some under RFK Jr. headlines.

What the headlines left out. The association was not observed in the study population as a whole, and expert commentary gathered by the Science Media Centre described it as only marginally significant within the genotype subgroups. Across every genotype, a higher proportion of processed meat was associated with increased dementia risk. And the cohort (Swedes born roughly 1920–1941) contains no meaningful vegan population. This compares lower-meat omnivores to higher-meat omnivores. Nobody in it was following a planned plant-based diet or supplementing B12.

Credit where due: the model adjusted for education, occupation type, living arrangements, physical activity, smoking, alcohol, total energy intake and the Alternative Healthy Eating Index, so socioeconomic and diet-quality confounding was partly addressed. Reverse causation is harder to dismiss: people in early undiagnosed decline eat differently and recall diet worse, over fifteen years in a cohort aged 60+.

The mechanism the authors reached for is the interesting part. Post-hoc analyses using a proxy for B12 absorption suggested APOE ε4 carriers took up more vitamin B12 per unit of meat than other genotypes. Norgren calls this a “food matrix” effect and explicitly labels it hypothesis-generating: “we can only speculate.” Note what that implies. If the operative variable is B12 delivery from the food matrix, a crystalline B12 supplement bypasses the food matrix entirely, which is exactly why the Institute of Medicine recommends the supplemental form to everyone over fifty. Hussein Yassine’s lab has published a careful critique of the paper worth reading if you want the full methodological picture.

So: real study, serious group, genuinely open question for APOE4 carriers eating little meat and supplementing nothing. It says nothing whatsoever about a supplementing vegan, because it did not contain one. Which is, once again, the entire problem with this genre.

Set against it, the largest recent signal runs the other way: a 2025 Neurology analysis from Harvard T.H. Chan and Mass General Brigham found a quarter-serving of processed red meat daily associated with 13% higher dementia risk (HR 1.13, 95% CI 1.08–1.19) and roughly 1.6 years of additional cognitive ageing per daily serving, with substitution modelling suggesting 19% lower risk swapping in nuts and legumes and 28% swapping in fish. Critics have fairly noted that unprocessed red meat showed no significant association with dementia or objective cognitive measures there, only with subjective cognitive decline. It is also an observational cohort of predominantly white US health professionals. Both can be true. Neither gets you to the cover of this book.

On the population-level evidence, see our work on the longest-living populations, Adventist Health Studies, and both sides of the Blue Zones argument.

What to actually do

The honest version of the concern this book is exploiting is short, and it looks nothing like the Sapiens Diet.

• Take B12. Reliably, forever, and get your levels checked. Non-negotiable, and doubly so in pregnancy and lactation.
• Consider algal DHA/EPA. It’s the Sherzais’ recommendation, especially during pregnancy, in childhood, and in older age. The evidence that it improves cognition is weak. The cost of covering the base is trivial.
• Cover iodine, vitamin D, iron, zinc and choline. See our nutrition overview.
• Do the things with the strongest dementia evidence, which are mostly not on your plate: exercise, sleep, blood pressure control, hearing correction, social connection, cognitive engagement. Both Ornish’s trial and the Sherzais’ NEURO framework are multimodal for a reason.
• Eat plants. Fibre, polyphenols, and vegetable intake track with better outcomes across essentially every large dataset we have.

That’s it. That’s the whole brain-health protocol, and it’s compatible with a fully plant-based diet in a way that two practising dementia neurologists have been recommending publicly for years.

One note on the professional guidance, since it’s being misrepresented. The Academy of Nutrition and Dietetics’ 2016 position paper stated that appropriately planned vegetarian and vegan diets are “appropriate for all stages of the life cycle, including pregnancy, lactation, infancy, childhood, adolescence, older adulthood, and for athletes.” That paper expired at the end of 2021 and was replaced in January 2025 by Vegetarian Dietary Patterns for Adults, which reaffirms nutritional adequacy and long-term benefit in adults, and does not address pregnancy or children because AND judged those to need separate guidance. That is a scope decision, not a reversal, and anyone telling you AND now warns against vegan diets for children or pregnancy is misreading a table of contents. For pregnancy and lactation specifically, British Dietetic Association and NHS-aligned midwifery guidance, Dietitians of Canada, and the Italian Scientific Society for Vegetarian Nutrition all hold that well-planned vegan diets are appropriate, and a 2024 Journal of the Academy of Nutrition and Dietetics modelling study found vegan adaptations of the USDA Healthy Vegetarian Dietary Pattern nutritionally adequate for pregnancy with supplementation. And the American Academy of Pediatrics’ Pediatric Nutrition Handbook, in its 5th edition (2004), stated that “children exhibit good growth and thrive on most lacto-ovo vegetarian and vegan diets when they are well planned and supplemented appropriately,” with the 6th edition (2008) taking the same line on adequacy with fortified foods or supplements as needed.

One honest complication, since this post is about describing evidence accurately. The dietetic and obstetric bodies are supportive; some paediatric bodies are more cautious about infancy and childhood specifically. In August 2025 the ESPGHAN Nutrition Committee published a position paper (Verduci et al., Journal of Pediatric Gastroenterology and Nutrition), the first based on a systematic search rather than narrative review, which found only 10 qualifying studies covering roughly 1,500 vegan children, concluded that supplementation “including vitamin B12, is essential during paediatric age when following a strict vegan diet,” and called for high-quality trials to close the evidence gaps. The Canadian Paediatric Society is likewise more comfortable when dairy and eggs are included.

That’s a finding of insufficient evidence and a requirement for proper supervision, not a finding of harm, and its emphasis on B12 is the same one made throughout this post. But it’s real, it’s recent, and it deserves stating rather than skipping, particularly since it makes the same criticism of expert-opinion position papers that this post makes of the book. It also concerns weaning and childhood rather than pregnancy, and it says nothing about brain atrophy in adults, which is what the book on the table actually claims.

On raising children this way

The book’s closing message is aimed at prospective mothers, so it’s worth ending where it does.

Start with what the leading bodies actually say. Vegetarians and vegans are at reduced risk of ischaemic heart disease, type 2 diabetes, hypertension, several cancers and obesity. That’s the Academy of Nutrition and Dietetics, not an advocacy group, and its 2025 position paper reaffirms long-term benefit for cardiometabolic outcomes in adults. Those are, between them, most of what kills people early in wealthy countries and most of what steals the last decade of a life.

Now add the timing. Atherosclerosis doesn’t wait for adulthood. The Bogalusa Heart Study and PDAY found fatty streaks and early lesions in children and adolescents, tracking with LDL. The disease process the adult data is measuring begins in childhood. And dietary patterns established early tend to persist. If a way of eating lowers the risk of the diseases in question, the window in which it operates opens young.

Now the objection, and why it’s weaker than it sounds. No cohort has followed vegan-raised children from conception into late adulthood and measured chronic disease outcomes. True. It has also never been done for any other diet. Nobody has run that study on omnivorous children, and nobody demands it before handing out the standard advice, which is attached to a dietary pattern that has never cleared the bar being set here and that produces the population outcomes this post has spent several thousand words discussing.

That standard, applied consistently, would forbid all paediatric nutrition guidance. The reason children are told to eat vegetables and limit sugar is not a lifetime randomised trial. It is adult outcome data, plus mechanistic plausibility, plus short-term paediatric evidence, the identical structure of inference at work here. Demanding more of veganism specifically is the same asymmetry as “well-planned,” wearing a lab coat.

So the honest formulation isn’t “we don’t know.” It’s this: given that the best available evidence says this way of eating lowers the risk of the diseases most likely to kill your child in sixty years, and given that the disease process starts in childhood, raising them that way follows. Not proven in the sense that no dietary recommendation for children is proven. But reasoned, and reasoned from a stronger evidence base than the default it’s being measured against.

Which is what the parents I’ve interviewed are actually doing when they say, and they say it often, that raising their children vegan is the best gift they could give them. That isn’t a testimonial about how someone felt after a diet change. It’s a conclusion drawn from the same evidence the professional bodies publish, by people who then took the B12 seriously, had the bloods done, and got on with it. They’ve also factored in something the risk calculations don’t capture: the animals, and the planet those children inherit.

A values position resting on a defensible evidence base. Which is a different and more honest object than the book on the table: a values position resting on a study that never studied a vegan.

The book’s title is a marketing decision its own authors admitted to on two podcasts. The study underneath it never looked at a vegan. And when you press on the “best evidence,” what you find is a solved problem with a two-cent solution, wearing a much more dramatic hat.

If you spot an error above, tell me and I’ll correct it. That’s supposed to be how this works.

Last updated: October 6, 2026, 12:56 PM EDT