The short version: the claim that people eating a varied whole-food plant-based diet need to fear “plant toxins” or “anti-nutrients” does not survive contact with the evidence. These compounds are dosed for insects and microbes, humans encounter them at a tiny fraction of the relevant per-kilogram exposure, and at real dietary intakes many are actively health-promoting, while higher plant intakes consistently track with lower mortality in large cohorts, backed by randomized trials showing fewer cardiovascular events and better risk factors. The legitimate caveats are narrow and conditional. The fix is variety, a few simple kitchen habits, and individualized support for the minority who need it.

Two different arguments wearing one costume

Spend ten minutes in the carnivore corner of the internet and you’ll meet a recurring performance: broccoli is “trying to kill you,” spinach is “loaded with oxalates,” beans are “full of lectins,” grains are “anti-nutrient bombs.” It sounds like one coherent indictment. It is actually two separate claims mashed together, and keeping them apart is the first step to seeing why neither works.

“Plant toxins” refers to defense chemicals, compounds a plant synthesizes to deter things that eat it: alkaloids, glucosinolates (and their breakdown products like sulforaphane and goitrin), polyphenols, salicylates, and caffeine. The argument here is “these are literally pesticides, so they must be poisoning you too.”

“Anti-nutrients” refers to compounds that interfere with the absorption of nutrients: phytate (which binds minerals), lectins (which bind carbohydrate structures), oxalates (which bind calcium), tannins (which bind iron and protein), and goitrogens (which interfere with iodine uptake). The argument here is “these steal your minerals, so plant foods leave you deficient.”

Different mechanisms, different evidence, different rebuttals. Blurring them is rhetorically convenient because it lets a scary word from one bucket (“toxin”) borrow menace for a compound in the other. Let’s refuse the blur.

Part 1. “Plant toxins”: dosed for bugs, not for you

Are plant defense chemicals actually dangerous at human body size?

Plants can’t run. Their defense is chemistry, and that chemistry was shaped by an evolutionary arms race with the organisms that actually threaten a plant: insects, mites, nematodes, fungi, and bacteria. The dose that matters, biologically, is the dose per unit body mass of the thing being deterred, and here the numbers are decisive.

An aphid weighs on the order of 1 milligram. A caterpillar, around 1 gram. A human, about 70,000 grams. That’s the plant’s target audience versus you: a difference of four to nearly eight orders of magnitude in body mass. Now factor in how much each eats. A caterpillar routinely consumes multiples of its own body weight in leaf tissue every day; its entire existence is eating the very tissue where defense compounds concentrate. A human eating a generous 80-gram serving of broccoli is consuming roughly one one-thousandth of body weight of that plant, once, as part of a mixed meal. Stack the mass difference on top of the intake difference and the per-kilogram chemical exposure a human experiences from a serving of vegetables is thousands to millions of times lower than what the compound was “designed” to deliver to a chewing insect. “The plant is trying to poison bugs” is true and completely beside the point at human scale.

It gets better. Many of the most potent botanical insecticides act on targets humans don’t have or barely share:

• Insect-specific ion channels and receptors. Caffeine and pyrethrin-type compounds disrupt invertebrate neural signaling. Insect sodium channels and receptor subtypes differ enough from ours that the selectivity is real, and mammals carry esterases that hydrolyze pyrethrins before they can do much. It is why pyrethroid insecticides are relatively low-toxicity to us.
• Molting and developmental hormones. Numerous plant compounds mimic or antagonize insect molting hormones (ecdysteroids) and juvenile hormone. Azadirachtin, the active principle in neem, is the classic case: it disrupts molting, and humans do not molt. These are endocrine systems we simply do not possess.
• Alkaline insect midguts. Many plant protease inhibitors and lectins are tuned to the alkaline gut of insects. The sharpest illustration is Bt toxin, which needs both an alkaline midgut and specific receptors to work, and humans have an acidic stomach and lack the receptors entirely. A number of these compounds are degraded by that acidity or by cooking, though some, like raw kidney-bean lectin, survive gentle heat and have to be destroyed by thorough boiling (more on that below).

And humans are not passive. We evolved as omnivores eating a chemically diverse plant diet, and we carry the toolkit to prove it: a large family of cytochrome P450 enzymes and glutathione-S-transferases that metabolize and conjugate xenobiotics for excretion; 25 functional bitter taste receptors (the TAS2R family) that let us detect and modulate intake of concentrated bitter compounds, and a liver-and-kidney clearance system that handles the trickle of plant chemistry a normal diet delivers.

On top of biology sits history. Ten thousand years of crop domestication has been, in large part, an unbroken program of selecting against toxicity and bitterness. Wild almonds carry enough amygdalin to kill; domesticated sweet almonds don’t. Wild potatoes and their glycoalkaloids, wild cucurbits and their cucurbitacins, wild lettuce and its bitter latex, all bred down. The vegetables in your kitchen are the tamed descendants of far nastier ancestors.

What is hormesis, and does it make plant compounds good for you?

Here is where the carnivore argument doesn’t just fail. It inverts. At the low doses humans actually consume, many plant defense compounds act as hormetins: they impose a brief, mild, fully recoverable stress that trips the cell’s adaptive machinery, and the adaptation leaves you better off than baseline. This is not fringe theory; it’s mainstream cell biology. Dietary hormetins such as sulforaphane, quercetin, and green-tea catechins react with sensor proteins like Keap1, releasing the transcription factor Nrf2 to switch on antioxidant response elements that induce hundreds of cytoprotective genes, including superoxide dismutase and glutathione peroxidase. The same low-dose stress recruits glutathione synthesis, heat-shock proteins, autophagy, and the AMPK and sirtuin pathways.

In other words: the “toxin” gives your cells a small workout, and your cells respond by upgrading their defenses. The net effect at dietary doses is protective. This is the identical logic behind exercise, sauna, and fasting, a point we’ll return to.

Sulforaphane is the strongest worked example, and it’s human data. In Qidong, China (a region with high aflatoxin and air-pollution exposure and correspondingly high liver-cancer rates), Johns Hopkins-led teams ran randomized trials of broccoli-sprout beverages. The first, in 200 adults (Kensler et al., Cancer Epidemiology, Biomarkers & Prevention 2005), was negative on its main comparison: urinary markers of aflatoxin-DNA damage and of the combustion carcinogen phenanthrene did not differ significantly between the sprout and placebo arms. But sulforaphane absorption varied enormously from person to person, and among those drinking the sprout infusion, the more sulforaphane a person absorbed, the less of both markers they excreted. A later randomized trial of 291 participants (Egner et al., Cancer Prevention Research 2014) found the broccoli-sprout beverage rapidly and durably raised excretion of the benzene detoxification product, by about 61%, and of acrolein by about 23%. Mechanistically, sulforaphane is a potent inducer of Nrf2-driven phase II detoxification enzymes (NQO1, glutathione-S-transferases), inhibits histone deacetylase, and, in a randomized human and mouse study (Yanaka et al., Cancer Prevention Research 2009), suppresses Helicobacter pylori colonization and gastritis, though it reduces rather than eradicates the infection. It’s worth being precise about what these trials measured: detoxification biomarkers and bacterial colonization, not cancer incidence or cure. That’s still a defense compound demonstrably helping human physiology, which is the opposite of the scare.

Other defense compounds have suggestive-to-strong evidence: curcumin, resveratrol, EGCG, and quercetin are all documented Nrf2 activators, though much of their disease-outcome evidence remains preclinical and is complicated by bioavailability problems. The honest framing: sulforaphane is the best-substantiated, and it is reasonable to expect (though not yet proven at the same tier for each) that other dietary hormetins share in these benefits.

Is the benefit of eating plants all just hormesis?

Here is the move the “plants are low-dose poison” framing depends on you not noticing. It collapses several unrelated mechanisms into one. Hormesis is real, and it explains sulforaphane and the polyphenols well. It explains almost nothing else on your plate.

• Nitrate is the usual example, with beetroot as its showcase. Pooled across 22 randomized trials, beetroot juice lowered systolic blood pressure by about 3.5 mmHg. The textbook explanation is that the body converts nitrate to nitrite and then to nitric oxide, a vasodilator, but this meta-analysis found the effect did not scale with nitrate content, so other compounds in beetroot likely contribute too. Either way, none of it is a stress response. Your body is being handed raw materials, not provoked into adapting.
• Beta-glucan, the soluble fiber in oats and barley, forms a gel that traps bile acids so they are excreted rather than reabsorbed, and the liver then pulls cholesterol out of circulation to build replacement bile acids. Around 3 g a day lowers LDL cholesterol by about 0.25 mmol/L in pooled randomized trials, an effect solid enough that both the FDA in 1997 and EFSA in 2010 authorized the health claim. That is physical chemistry happening in your intestine.
• Folate. After the United States mandated folic acid fortification of enriched grains in 1998, CDC surveillance recorded a 27% decline in neural-tube-defect-affected pregnancies, from an estimated 4,130 a year to 3,020, with spina bifida alone falling from about 2,490 to 1,640. Nobody got tougher from the deficiency. Babies got spina bifida.
• Vitamin C. Our lineage carries a broken GULO gene and cannot synthesize it at all. That is not a plant attacking us. It is evidence that our ancestors outsourced a metabolic function because the dietary supply was so dependable that maintaining the enzyme stopped being worth its cost. That is the exact inverse of chemical warfare.
• Fiber feeds the bacteria that produce butyrate, which the next section takes up in full.

Bile acid sequestration, the blood-vessel effects of beetroot, bacterial fermentation, and vitamins acting as enzyme cofactors are not variations on a single theme. They are four different mechanisms, and not one of them is a stress response. The argument only works if you pretend there is only one.

If caffeine is a natural insecticide, why is coffee good for you?

If you want a single example that detonates the “it’s a pesticide, therefore it’s bad for you” syllogism, it’s caffeine. Caffeine works as an insecticide at the concentrations plants actually contain. James Nathanson showed in Science in 1984 that caffeine and related methylxanthines inhibit insect feeding, stunt growth, and at the levels found in coffee beans and tea leaves kill insect larvae, which led him to propose them as naturally occurring pesticides.

And yet coffee, a warm caffeine-and-polyphenol extraction of a defended seed, is among the most reliably health-associated foods studied. The largest synthesis is Poole et al.’s 2017 BMJ umbrella review of 201 observational meta-analyses covering 67 outcomes, which found the largest reductions at three-to-four cups a day versus none: about 17% lower all-cause mortality (relative risk 0.83, 95% CI 0.79–0.88) and roughly 19% lower cardiovascular mortality, while high-versus-low consumption was associated with an 18% lower risk of incident cancer (0.82, 0.74–0.89), plus lower risk of type 2 diabetes, Parkinson’s disease, and liver disease. The authors are candid about the limits: most of this evidence is observational and of low or very low quality, and they single out pregnancy and women at higher risk of fracture as exceptions to the reassuring picture.

A genuine insecticide, consumed by choice by billions, tracks with living longer. A compound’s status as a “plant toxin” tells you essentially nothing about its effect on a 70-kilogram omnivore with a functioning liver.

Part 2. The microbiome: your 100-trillion-cell answer to “anti-nutrients”

The carnivore framing treats the human gut as a naive tube that either absorbs a nutrient or gets robbed of it. Real digestion is a three-way negotiation between food, host, and roughly 100 trillion microbes, and that third party changes everything.

Fiber becomes short-chain fatty acids. Fiber and resistant starch that humans can’t digest are fermented by colonic bacteria into acetate, propionate, and butyrate. Butyrate is widely regarded as the primary fuel for colonocytes, the cells lining your colon, though a 2023 review cautions that it can’t meet all their needs on its own. It gets shuttled into their mitochondria and run through beta-oxidation into acetyl-CoA, then burned in the citric acid cycle for ATP, clean chemistry your own cells are built to run. Butyrate also inhibits histone deacetylase, strengthens the gut barrier, and exerts anti-inflammatory and anti-carcinogenic effects. A meat-only diet delivers essentially none of this substrate.

Unabsorbed protein becomes something else entirely. Not all dietary protein makes it past the small intestine. An isotope tracer study gave ten healthy volunteers 25 grams of labeled egg protein and measured what escaped digestion: about 6% of cooked egg protein reached the colon undigested, and that jumped to roughly 35% for raw egg protein. Once protein is in the colon, it isn’t fermented the clean way fiber is. It’s putrefied: colonic bacteria break the unabsorbed amino acids down into thiols, phenols, ammonia, indoles and amines, which the researchers describe as potentially toxic. Where fiber fermentation runs the same mitochondrial chemistry that feeds your colon cells, protein putrefaction produces a chemical signature closer to what a septic system does. This is one proposed reason colorectal cancer risk tracks with red and processed meat.

You cannot supplement your way out of it. This isn’t theoretical. Black South Africans, rural and urban, eat below-recommended amounts of fiber yet have colon cancer rates under 1 per 100,000, among the lowest recorded, and the researchers linked that rarity to low animal-product consumption rather than fiber, with resistant starch from their maize staple a likely contributor. A controlled crossover trial then tested whether resistant starch could offset meat directly: eight healthy men ate a low-meat diet, a high-meat diet, and a high-meat diet with resistant starch added, each for nineteen days. The high-meat diet tripled fecal N-nitroso compounds, which can damage DNA in the bowel lining, and doubled fecal ammonia compared with the low-meat diet, and adding resistant starch had no significant effect on either. These compounds formed in the bowel, not on the plate: the study diets deliberately contained no foods with meaningful preformed N-nitroso compounds or nitrite, and fresh meat alone still tripled their output. Buying nitrite-free or uncured meat doesn’t sidestep the chemistry. Fiber and resistant starch measurably help in other ways. They do not cancel out a diet built on animal protein, and the African data points the same way: the protection tracked with eating little meat, not with fiber intake.

Polyphenols become postbiotics. Much of a polyphenol’s benefit is actually delivered by bacteria that transform it. The cleanest example is urolithin A: gut microbes convert ellagitannins from pomegranates, walnuts, and berries into urolithin A, and human trials show it induces mitophagy signatures in skeletal muscle. A randomized, placebo-controlled trial in middle-aged adults (Singh et al., Cell Reports Medicine 2022) reported roughly 12% improvements in muscle strength alongside meaningful gains in aerobic endurance and six-minute walk performance; though in fairness, the study’s pre-specified primary endpoint, peak power output, did not reach statistical significance, and the trial was sponsored by Amazentis, with six of its authors employed by the company. The mechanism is sound and the direction is encouraging; the “plant chemical” isn’t the finished product, your microbiome finishes it.

“Anti-nutrients” are microbial food. Phytate and resistant starch aren’t just mineral-binders, they’re fermentable substrates that feed beneficial microbes. And a healthy microbiome directly neutralizes the compounds the carnivore camp fears: Oxalobacter formigenes, a gut bacterium that lives on oxalate, was found in 38% of healthy controls but only 17% of people with recurrent calcium-oxalate kidney stones, and carrying it was associated with about 70% lower odds of being a recurrent stone-former. The mechanism is less settled than the association: surprisingly, urinary oxalate did not differ with colonization in that study. What predicts losing the bacterium is telling. In a study of 240 healthy adults, colonization was 55% among people who had never used certain antibiotics, but 17% among those who had used them within the past year. The main threat to this protection is antibiotic exposure, not eating plants.

Diversity beats labels. In the observational, self-selected American Gut Project (McDonald et al., mSystems 2018), spanning 15,096 samples across the US, UK, and Australia, people eating more than 30 different plant types per week had more diverse gut microbiomes and fewer antibiotic-resistance genes than those eating 10 or fewer, though that head-to-head comparison involved only about 85 people drawn from the larger project. Strikingly, plant diversity tracked microbiome diversity more closely than the coarse vegan-versus-omnivore label did. It’s association rather than proof, but it’s the single best real-world signal for the variety prescription this article keeps returning to.

The contrast case. Low-fiber, animal-heavy patterns push the microbiome the other direction, toward bile-tolerant organisms and away from fiber fermenters, and load the diet with the precursors (carnitine from red meat, choline from eggs and meat) that gut microbes convert to trimethylamine, which the liver oxidizes to TMAO. TMAO is a biomarker associated with cardiovascular risk in humans and accelerated atherosclerosis in mice, from work by Stanley Hazen’s group (Nature Medicine 2013 and after); whether it is causal in people is still debated, but the diet–microbiome link is robust. That same group found that after a steak-based carnitine challenge, five vegans produced markedly less TMAO than five omnivores, through a gut-bacteria-dependent mechanism. The investigators also hold patents on cardiovascular diagnostics, one more reason to treat TMAO as a marker rather than a verdict. The microbiome you build determines the metabolites you make.

Part 3. “Anti-nutrients” are often protective

What is chelation, and why does it explain most anti-nutrient claims?

Before going through these compounds one at a time, it helps to name the single mechanism underneath most of them, because once you see it the whole category stops looking mysterious. (Lectins and goitrogens work differently, and get their own treatment below.)

Chelation takes its name from the Greek chele, meaning crab’s claw. A chelating molecule wraps around a mineral or a macronutrient and locks it into a ring, gripping it at several points at once. Once bound, the resulting complex is frequently too large, too stable, or too insoluble for the intestine to pull through the gut wall. That one image accounts for phytate, oxalate, tannins and polyphenols simultaneously. They are not four unrelated threats. They are four molecules doing the same thing.

Tannins make the mechanism unusually literal. The word comes from tanning animal hides into leather, a process that works precisely because these compounds bind protein fibers so tightly that the resulting material resists decay. That etymology is not a quirk of language, it is the chemistry. Tannins bind dietary protein into insoluble complexes, and they also bind the digestive enzymes sent to break that protein down, trypsin among them, which is a double hit on digestibility.

There is one human trial that measured this directly, and it is worth reporting carefully because it is both real and very small. Kashyap and colleagues used a dual isotope tracer method to test what happened when healthy Indian adults drank black tea alongside a meal of whole boiled egg with ghee rice and tomato curry. True indispensable amino acid digestibility of the egg protein fell by 17% when tea was co-ingested. The sample was five people, so treat the precise figure as indicative rather than settled.

Two design details matter. The no-tea comparison came from the same five people measured in an earlier experiment, not a side-by-side run. And spirulina was not a test food: it was the method’s reference protein, checked with and without tea in just three people, and the authors say an effect of tea on it cannot be ruled out. So this trial shows tea can blunt the digestibility of a high-quality protein eaten alongside it. It does not show plant protein is spared, and the authors’ own concern was populations already at risk of protein inadequacy rather than the average tea drinker. For anyone drinking tea with dinner, the practical reading is unchanged: a real but modest effect, trivially offset by drinking tea between meals instead of with them, and no reason to give up tea.

Now the part the panic depends on you not noticing. The anti-nutrient framing treats binding as inherently harmful, as though a molecule gripping a mineral were self-evidently an act of theft. Yet the supplement industry chelates minerals on purpose and sells it as a premium feature. That is what magnesium glycinate and calcium bisglycinate are: the mineral pre-bound to an amino acid, specifically so it travels through the stomach in a protected form and reaches the intestine intact. Identical chemistry, opposite intent, opposite marketing. One version gets called an anti-nutrient and the other gets a higher price tag.

Which tells you the honest answer. Chelation is not good or bad in itself. Whether it helps or hurts depends entirely on what is being bound, where in the digestive tract it happens, and what else is on the plate at the time. And it explains why the strongest argument in this whole section is the one about phytate below: phytate’s mineral binding and its antioxidant and anti-cancer activity are not two separate properties to be weighed against each other. They are the same molecular behavior, described by people with different agendas. The compound grips iron. Look at the iron that grip keeps you from absorbing and a critic calls it an anti-nutrient. Look at the iron that same grip keeps from generating hydroxyl radicals and a researcher calls it an antioxidant. Both happen in the same place, your digestive tract. Meanwhile soaking, sprouting, fermenting and leavening reduce the gut-side cost anyway, which is why traditional food preparation solved this problem long before anyone had a word for it.

Is phytate (phytic acid) harmful, or is it actually protective?

Phytate (inositol hexaphosphate, or IP6) is the marquee “anti-nutrient.” Yes, it chelates iron, zinc, and calcium and can reduce their absorption from a single meal. But that same iron-chelating property also makes it an antioxidant: by locking up free iron, IP6 blocks the iron-catalyzed formation of hydroxyl radicals, a mechanism with an important qualifier about where it works, covered in the next section. And the anti-cancer literature is genuinely striking, though only at the preclinical level. IP6 reduces proliferation, induces apoptosis and differentiation in cancer cell lines, inhibits colorectal cancer in carcinogen-treated rodents, and reduces liver metastasis in mouse models; epidemiologically, IP6-rich diets track with lower colon-cancer risk.

Here I’ll be more careful than most write-ups, because the evidence deserves it. The robust IP6 anti-cancer evidence is preclinical. The human clinical evidence is a single small pilot, Bačić et al. (Journal of Experimental & Clinical Cancer Research, 2010), 14 breast-cancer patients randomized to IP6 plus inositol versus placebo alongside chemotherapy. It found significantly better quality of life and functional status and protection against chemotherapy-induced drops in white cells and platelets, but tumor markers were explicitly unchanged in both groups. So the honest claim is: IP6 has broad, consistent preclinical anti-cancer activity and one small human trial showing it eased chemotherapy toxicity. That’s a compound plausibly working for you, the opposite of the scare, while stopping short of proven human anti-tumor efficacy.

There’s also a phytate paradox worth noting: despite binding calcium, higher phytate intake is associated with better bone outcomes, not worse. In a study of 157 postmenopausal women, those with low urinary phytate lost about 3.1% of lumbar-spine bone over a year, against about 0.4% in those with high levels. The proposed reason is that phytate adsorbs to the surface of bone crystals and slows their breakdown, the same crystal-binding chemistry that lets it inhibit calcification in blood vessels, as a 2023 review describes. One disclosure matters here: most of this work comes from a single Spanish research group, and in that bone study two authors were inventors on a pending patent based on the work, and two were employees of Laboratoris Sanifit, one also a stockholder. The mineral-absorption concern, meanwhile, is real only for people whose diets are already marginal in iron or zinc and who don’t use the trivially simple mitigations below. The adaptation evidence is worth stating precisely, because it cuts both ways. Phytate’s grip on the iron in any single meal does not seem to weaken with habit: in a 1989 study of 22 long-term vegetarians and vegans, bran cut iron absorption from a test meal by 92%, the same as in controls, although the vegetarians’ iron status was satisfactory despite years of high phytate intake. What does adjust is the body’s total uptake. When 31 men ate controlled diets with high or low iron bioavailability for 12 weeks, nonheme iron absorption partially adapted, shrinking the difference between the diets from eightfold to fourfold, and the researchers concluded that short-term measurements overestimate those differences. And in a small randomized trial of young women with low iron stores, eight weeks on a high-phytate diet improved their blood iron response to a high-phytate test meal by 41%, while a low-phytate group showed no such change. So single-meal tests, the kind the “mineral thief” argument leans on, capture the inhibitor correctly but overstate what it does to your iron over time.

One adjacent claim is worth handling carefully, because it circulates in plant-based circles and it is only half right. You will see it said that phytate chelates heavy metals and therefore shields you from them. Rodent studies do show dietary phytate reducing lead accumulation in bone. But the same literature shows no corresponding reduction in tissue cadmium, and in at least one study phytic acid increased liver cadmium while the enzyme that destroys phytate lowered it. There is no human data either way. Phytate is a mineral-binder with a useful antioxidant side, not a detox agent, and overstating it hands critics an easy correction.

Is phytic acid an antioxidant, and how does it actually work?

Yes, but not in the way most people picture an antioxidant, and the distinction matters. There are two broad ways a compound can protect against oxidative damage. The first is scavenging: the phenolic groups in polyphenols can donate a hydrogen atom and neutralize the unpaired electron on a free radical that already exists. The second is prevention: stopping radicals from forming at all, chiefly by chelating transition metals such as iron so they cannot drive the Fenton reaction, in which iron reacts with hydrogen peroxide to produce the hydroxyl radical. Polyphenols can do both. Phytate does only the second. It is not an electron donor mopping up radicals; it is a chelator that keeps iron from making them.

The chemistry behind phytate’s version is more interesting than “it binds iron.” Many iron-binding compounds actually make oxidation worse, because the iron they hold can still reach hydrogen peroxide. The 1987 paper that first described phytic acid as a natural antioxidant found that phytate’s iron complex behaves differently: it blocks iron-driven hydroxyl radical generation and suppresses lipid peroxidation. The authors proposed that this is one of phytate’s jobs in the plant, protecting seeds from oxidative damage during dormancy, and noted that at high concentrations it also slows the browning and spoilage of fruits and vegetables.

Where it works is the part worth being precise about. The same researchers, in a 1990 review, placed the dietary effect in the digestive tract, proposing that phytate may lower colon cancer risk and protect against inflammatory bowel disease by the same mechanism, a hypothesis rather than a demonstrated outcome. A 1998 rat study then tested the systemic question directly: phytic acid was a potent inhibitor of iron-catalyzed hydroxyl radical formation in the test tube, yet feeding it to growing rats had no significant effect on their liver’s oxidant or antioxidant status. Part of the reason is that iron in the blood travels tightly bound to its carrier protein, transferrin, rather than loose. So the honest summary is that phytate is a preventive antioxidant whose clearest action is chemical, and whose likely stage in the body is the gut, where it meets dietary iron directly. That is the chelation point from the top of this part, now with the chemistry filled in.

Reviews of phytate go further, crediting it with inhibiting calcium-salt crystallization in blood vessels and kidneys, reducing protein glycation in diabetes, and protecting neurons. Treat those as promising rather than settled. A 2023 review of the human evidence noted that the clinical trials share limitations such as small numbers of patients and short intervention times, and much of the rest comes from laboratory and animal work.

Polyphenols are the more complicated case. A 2023 review of their chemistry describes them as ambivalent, able to act as antioxidants or pro-oxidants depending on conditions, and argues that this back-and-forth is part of how they switch on the body’s own NRF2 antioxidant defenses, the same system described in the hormesis section above. Neither compound fits the cartoon of an antioxidant as a molecule that simply soaks up damage, and neither fits the cartoon of an “anti-nutrient” either.

Does chewing or soaking almonds reduce their phytate content?

Almonds are a genuinely useful test case, because two separate habits get recommended for them, chewing them thoroughly and soaking them overnight, and neither does what people think it does.

Start with chewing. Phytic acid is a small molecule with six phosphate groups locked onto an inositol ring, and breaking that ring apart takes either the enzyme phytase or fairly strong acid hydrolysis. Chewing does neither. Turning an almond into a slurry increases surface area and mixes it with salivary amylase, which digests starch, a completely different class of bond. Mechanical grinding does not touch the phosphate-ester bonds holding phytate together. If anything, the increased surface area from thorough chewing could let phytate mix more freely with the minerals in the rest of your meal, which cuts the other way from the intended benefit, though this effect has not been measured directly and is almost certainly trivial at normal intakes.

Soaking is the more interesting case, because it is well studied and the result surprises people. As part of a randomized crossover trial in 76 adults comparing raw and soaked almonds, the researchers measured phytate in the nuts themselves by HPLC. Soaking did not lower phytate. It significantly raised it, from 531 mg per 100 g unsoaked to 563 mg per 100 g soaked, and sliced soaked almonds showed no meaningful difference from whole unsoaked ones. The same research group followed up with a wider study across almonds, hazelnuts, peanuts and walnuts, testing several soaking regimens with HPLC and ICP-MS. Changes in phytate ranged from a 12% decrease to a 10% increase depending on the nut and the method, and minerals leached out into the soak water along with it, so soaking did not improve the phytate-to-mineral ratio in any nut tested. The claim that soaking activates a phytase enzyme and drains phytic acid out of nuts the way it does in legumes and grains does not hold up when someone actually measures it.

That contrast with legumes and grains is worth being precise about, since this post cites soaking and sprouting favorably elsewhere. Grains and legumes have an active endogenous phytase that switches on during soaking and germination, which is why the fermentation and sprouting data earlier in this piece hold up. Almonds are a mature, dormant seed with a hard testa, and the same biochemistry simply is not running in the same way. One rule does not fit every plant food, and the honest version of this post’s argument has to say so.

None of that makes almonds worth worrying about. Their phytate content varies considerably across studies and cultivars, but the actual measured content is nowhere near dangerous at a normal handful-a-day intake, and the same iron-chelating chemistry that makes phytate an “anti-nutrient” is the antioxidant property described earlier in this section. The practical takeaway is simply that neither chewing almonds to liquid nor soaking them is doing what the internet claims it does. Eat them however you like them best.

Do anti-nutrients weaken your bones?

The deeper worry behind the whole anti-nutrient argument is that a diet full of mineral-binding compounds must eventually cost you bone. It has been tested directly, and it doesn’t hold up. The old acid-ash hypothesis claimed that protein acidifies the body and that bone dissolves to buffer it. Calcium isotope tracer work overturned this: in a tightly controlled study of 13 women, about 93% of the extra urinary calcium on a high-protein diet came from increased intestinal absorption, with no measurable effect on bone balance, at least in the short term. A 2009 meta-analysis of the five best-designed calcium balance studies reached the same place: a higher dietary acid load raised urinary calcium but did not change calcium balance or a marker of bone breakdown. Separately, a meta-analysis of seven randomized trials comparing soy protein against animal protein found bone mineral density differences near zero. An Austrian study of long-term vegans makes the same point from the other direction: their dietary acid load was less than half the omnivores’, which the acid-ash theory says should protect bone, yet the vegans who did no resistance training still had weaker bone structure. Acid load is not the lever.

Where a real signal does exist, it deserves precision. In EPIC-Oxford, vegans do show higher hip fracture risk, based on 41 vegan hip fractures among nearly 55,000 people followed for 17.6 years. A 2025 meta-analysis put vegans at 75% higher risk but graded that estimate “very low certainty,” with a confidence interval running from a 17% to a 163% increase. In absolute terms the gap is modest: in UK Biobank, the vegetarian excess worked out to about 3 extra hip fractures per 1,000 people over 10 years.

What drives it is the more useful question, and anti-nutrients are not on the list. An early EPIC-Oxford analysis found the vegan excess in total fractures disappeared among those eating at least 525 mg of calcium a day, though that rested on just 31 vegan fractures. A later, larger analysis of the same cohort was less reassuring: adjusting for calcium closed roughly three times more of the gap than adjusting for protein, yet vegans eating 700 mg or more a day still had about 2.4 times the hip fracture risk. The Adventist Health Study-2 adds a more encouraging data point: vegan women taking neither calcium nor vitamin D supplements had close to triple the hip fracture risk, while those taking both had no detectable excess. In UK Biobank, lower body weight explained roughly a quarter of the vegetarian excess, though that estimate is imprecise.

So the practical levers are calcium, vitamin D, body weight and strength training, and they work best together: roughly 700–1,000 mg of calcium a day from fortified plant milk, calcium-set tofu and low-oxalate greens such as bok choy, kale and broccoli, adequate vitamin D, a healthy weight, and regular resistance work. VeganLinked has a full dedicated piece working through the fracture data, the uncertainty in it, and exactly how to build a bone-protective plate.

Are lectins actually dangerous, or is it just raw kidney beans?

The lectin panic leans hard on one true fact and then overgeneralizes wildly. The true fact: raw and undercooked red kidney beans contain phytohaemagglutinin, and eating as few as four or five raw beans can cause acute vomiting and diarrhea. Per the FDA’s own accounting, raw kidney beans carry 20,000–70,000 hemagglutinating units; proper cooking drops that to 200–400 units, a reduction of roughly 99%. Lab work found that 10 minutes at a full boil destroys the toxin, but the FDA advises boiling for at least 30 minutes to be safe, and it warns against slow cookers, which often reach only 75°C or less. The entire dietary-lectin hazard is a preparation problem that human cooking solved millennia ago.

What about lectins as anti-cancer agents? Again, honesty matters. Purified lectins are legitimate laboratory anti-cancer candidates: mushroom (Agaricus bisporus) lectin reversibly inhibits colon-cancer cell proliferation in vitro (Yu et al., Cancer Research 1993), and many plant lectins show antiproliferative activity in cells and mice. But this is preclinical, and it involves purified, often injected lectins, not dietary ones. Mistletoe extract standardized to its lectin content, given by injection as an add-on cancer therapy, actually failed its randomized trial (Steuer-Vogt et al., European Journal of Cancer 2001, 477 head-and-neck cancer patients: no reduction in recurrence, metastasis, survival, or quality of life). If a standardized, dose-controlled lectin preparation can’t move the needle, dietary lectins in cooked beans are not the cancer engine the scare implies. And the robust epidemiology showing legumes lower colorectal-cancer risk (Zhu et al., Scientific Reports 2015, 14 cohorts) points to compounds like fiber and isoflavones and doesn’t mention lectins at all.

The legume evidence itself is strong. Higher legume intake is associated with lower all-cause mortality: a 2023 dose–response meta-analysis of 32 cohorts and over 1.1 million people (Zargarzadeh et al., Advances in Nutrition) found about 6% lower all-cause mortality per 50 g/day. Worth being precise about what that particular analysis did and didn’t find: it looked at mortality outcomes, and its cancer-mortality estimate did not reach statistical significance. The cancer signal rests on other work. The Zhu meta-analysis cited above found lower colorectal cancer risk with higher legume intake, and a PREDIMED sub-analysis (Papandreou et al., Clinical Nutrition 2019) found the highest total-legume intake associated with 49% lower cancer mortality and the highest lentil intake with 37% lower cancer mortality. That same PREDIMED paper also observed higher cardiovascular mortality with high legume and dry-bean intake, a reminder that single-cohort subgroup findings are hypothesis-generating and best read alongside the larger meta-analysis. The signal holds up in older adults across very different cuisines, too. In a study of 785 people aged 70 and over in Japan, Sweden, Greece and Australia, legumes were the only food group that consistently predicted survival, with about 8% lower mortality for every 20 g a day. If lectins were the menace advertised, the most lectin-rich staple on the planet would not keep turning up as a marker of longer life.

So the fair conclusion is a two-parter that still defeats the carnivore claim: dietary lectins in properly prepared food are not a meaningful hazard, and legumes are among the foods most consistently associated with longer life, even if the specific “lectins cure cancer” version is not established in humans.

Do oxalates, tannins, goitrogens and saponins actually matter?

• Oxalates have one everyday effect and one real risk. The everyday effect is that oxalate locks up the calcium in its own food, which is why spinach is a poor calcium source despite its label (see the calcium section in Part 5); that is a reason to count spinach’s calcium at a discount, not to fear spinach. The real risk is narrower: it matters for calcium-oxalate stone-formers and people with specific predisposing conditions, and at one behavior, extreme intake. The literature carnivore influencers cite is almost entirely case reports of mono-eating: a 65-year-old woman who developed oxalate nephropathy after a green-smoothie cleanse (she also had prior gastric bypass and recent antibiotics, both predisposing); juice-diet and high-dose-vitamin-C nephropathy cases. These are stories about kilograms of a single high-oxalate food, usually in someone already predisposed. The practical fix is already built into a normal varied diet: rotate your greens instead of basing every smoothie on spinach or chard, and see Part 4 for why variety closes this door on its own.
• Tannins and polyphenols inhibit non-heme iron absorption when consumed with a meal, which is exactly why the practical fix is to keep tea and coffee away from meals. Otherwise they’re antioxidants associated with benefit.
• Goitrogens from raw brassicas become a real thyroid concern at enormous raw intakes, and the risk is generally considered higher when iodine intake is low. The famous case: an 88-year-old woman who developed myxedema coma after eating 1–1.5 kg of raw bok choy daily for months (New England Journal of Medicine, 2010). Cooking largely deactivates the myrosinase enzyme responsible, as the case authors note, and keeping iodine intake adequate is sensible insurance. This is not an argument against kale, bok choy, or any cooked brassica.
• Saponins, the compounds you rinse off quinoa, are bitter surfactants with largely neutral-to-beneficial profiles at dietary levels, including cholesterol-binding and some anti-inflammatory activity.

The pattern is unmistakable: every genuine harm is a story of extreme mono-consumption or a specific predisposition, and every one is solved by variety, cooking, and, rarely, individualized adjustment.

Part 4. Mono-mealing versus variety: the real dividing line

The honest concession, the one that makes this argument credible rather than cheerleading, is that a few compounds can cause trouble if you build your diet around a single high-compound food. Liters of raw spinach-and-chard smoothies every day can overload oxalate. Star fruit can harm people with kidney disease. Kilograms of raw brassicas with an iodine-poor diet can suppress the thyroid. These are real.

But look at what they have in common. They are all failures of mono-eating, and the solution is precisely what a whole-food plant-based diet is supposed to be: variety across legumes, greens, grains, fruits, vegetables, mushrooms, nuts, seeds, seaweed, herbs and spices. Spread your intake across that spectrum and no single compound accumulates to a problematic dose. No oxalate overload. No goitrogen problem. No mineral-binding crisis.

Meanwhile diversity maximizes the upside: more distinct fibers feeding more microbial species, more hormetins hitting more adaptive pathways, more of the phytochemical spectrum the American Gut Project linked to a healthier gut. Variety isn’t just the safety mechanism; it’s the benefit multiplier.

And notice the irony: the carnivore diet is the actual mono-diet. It is the eating pattern most vulnerable to the “too much of one thing” failure mode it projects onto plants.

Part 5. The absorption hacks (this part is genuinely easy)

The mineral-absorption objection is the most technically legitimate part of the anti-nutrient argument, and it is largely defused by a handful of habits, most of which take nothing more than ordinary cooking. Calcium is the one place where food choice and fortified foods matter too, covered at the end of this part.

• Pair vitamin C with meals. Vitamin C is the best-documented enhancer of plant iron absorption. Hallberg and colleagues describe its role as generally accepted and working two ways: it stops iron forming insoluble compounds with inhibitors such as phytates and tannins, and it keeps iron in the ferrous form intestinal cells take up; they suggested about 50 mg with each main meal. The effect can be large. In 63 men, adding vitamin C to a test meal raised iron absorption in proportion to the dose, from 1.65-fold at 25 mg to more than ninefold at 1,000 mg, though less when the meal contained meat. In a study of 199 people, 30 mg overcame the inhibitory effect of maize-bran phytate on iron absorption from a meal. The honest caveat: those are single-meal measurements, and when researchers labeled a complete diet over several days, the effect of vitamin C was far less pronounced. Squeeze lemon on your lentils, have fruit with your oatmeal, add bell peppers to the bean chili. It is a real effect and a free habit, just not a magic multiplier on your whole day’s iron.
• Add a little acid. Acetic and other organic acids improve mineral solubility and absorption. A splash of vinegar (with the mother, if you like) or a squeeze of citrus on greens and grains helps.
• Keep tea and coffee away from meals. Their tannins inhibit non-heme iron uptake, so enjoy them an hour or so on either side of iron-rich meals rather than with them.
• Soak, sprout, ferment, and cook. These traditional techniques slash anti-nutrient loads. A 2023 review puts the reduction from soaking at roughly 17–80% and from sprouting at more than 60%, depending on the food and conditions, and notes that your gut bacteria carry on the job during digestion, since they produce the phytate-degrading enzyme phytase that humans lack. Fermentation is more powerful still: one maize study combining soaking, germination, and lactic fermentation (Nsabimana et al., Frontiers in Nutrition 2024) cut phytate by about 86% and lowered the phytate-to-iron and phytate-to-zinc ratios by roughly 85% and 81%, substantially improving estimated iron and zinc bioavailability. Cooking destroys lectins and reduces oxalate: boiling cut soluble oxalate by 30–87% across the vegetables tested, against 5–53% for steaming, and the lost oxalate ends up in the cooking water, so drain it.

For accessible, well-referenced walkthroughs of these exact hacks, NutritionFacts.org is an excellent free resource. Search it for phytates, lectins, oxalates, iron absorption, and anti-nutrients. Dr. Greger translates the primary literature into practical kitchen steps better than almost anyone.

Which leafy greens give you the most absorbable calcium?

Calcium is where oxalate, the other major mineral binder, matters most, and the differences between greens are larger than most people expect. A 1999 review by calcium researchers at Purdue and Creighton compared estimated absorbable calcium per serving. Per half-cup serving, about 85 g:

• Bok choy delivered 79 mg of calcium with 53.8% absorbed, about 42 mg actually taken up.
• Kale delivered 61 mg with 49.3% absorbed, about 30 mg.
• Broccoli had the highest absorption of all at 61.3%, though it carries less calcium, about 21 mg absorbed.
• Spinach looks impressive on a label at 115 mg, but only 5.1% was absorbed, under 6 mg.

For comparison, the calcium in milk was absorbed at 32.1%, so these low-oxalate greens absorb their calcium better than milk does; they simply carry less of it per serving. It took 2.3 half-cups of bok choy, or 3.2 of kale, to match the absorbable calcium in one cup of milk. The authors attribute the pattern to oxalate, with calcium absorption falling as oxalate rises, and they note that the fiber in vegetables does not inhibit calcium absorption. Soy is a notable exception to the binder story: soybeans are rich in both oxalate and phytate, yet soy foods deliver calcium well, and a serving of calcium-set tofu provided about 80 mg of absorbable calcium, close to a cup of milk. Common dried beans, also high in phytate, delivered considerably less.

Bok choy deserves more attention than it gets. In that comparison it edged out kale on both calcium content and absorption, and it is mild enough to disappear into a stir-fry, soup or noodle bowl. And yes, it is the same vegetable from the thyroid case report in the goitrogens section, but that case involved eating 1–1.5 kg of it raw every day for months. A cooked side of bok choy is a different exposure entirely.

The same authors are candid about the limit: because most vegetables carry modest amounts of calcium, the quantity needed makes an exclusively vegetable-based calcium intake impractical for most people without fortified foods or supplements. That is why fortified plant milk and calcium-set tofu belong alongside the greens, not instead of them. Aim toward the upper end of the 700–1,000 mg range: in the calcium tracer study discussed in the bones section, women eating 800 mg a day stayed in negative calcium balance at both protein levels. Our bone density post has the full calcium playbook, including the two label checks that decide whether these foods deliver: tofu set with calcium sulfate or calcium chloride rather than magnesium, and fortified plant milk shaken before every pour, because in tests of fortified soy and rice drinks most of the added calcium sat in a separable, particulate form rather than dissolved.

Part 6. When people need individualized help

A subset of people run into real friction when they shift toward a plant-rich diet, and it’s worth naming honestly, because the carnivore camp weaponizes these experiences as proof plants are toxic. Usually the truth is the reverse: the change reveals a pre-existing condition that was previously masked.

• Undiagnosed SIBO or IBS can flare when fiber increases suddenly. The fix is usually gradual ramp-up, cooking, and sometimes temporary low-FODMAP structuring, not fiber elimination.
• A pre-existing kidney-stone tendency calls for oxalate-aware choices, hydration, adequate dietary calcium (which binds oxalate in the gut), and rebuilding oxalate-degrading gut flora.
• Thyroid and iodine status should be squared away with iodized salt or seaweed before anyone eats large amounts of raw brassicas.
• Histamine intolerance can react to fermented foods specifically.

These are conditions to manage, not reasons to abandon the healthiest dietary pattern. People with them may simply need help learning to diversify their plates for their individual needs, so they can acclimate and make this way of eating simple and second nature.

If that’s you, be encouraged: all great things take a little effort up front, and more for some people than others. Your best bet is a physician, registered dietitian, or nutritionist who practices lifestyle medicine, someone who can give evidence-based guidance that respects both your values and your health goals. The American College of Lifestyle Medicine and similar directories are a good place to start looking.

Part 7. The epidemiological reality check

Here’s the test that settles it. If plant foods were net-toxic, higher plant intake should predict worse health outcomes. It predicts the opposite, and the observational and trial evidence point the same way:

• Fiber and whole grains. Reynolds et al.’s 2019 Lancet series pooled 185 prospective studies (roughly 135 million person-years) and 58 clinical trials. In the observational data, the highest fiber eaters had 15–30% lower all-cause and cardiovascular mortality and 15–30% lower incidence of coronary heart disease, stroke, type 2 diabetes and colorectal cancer than the lowest. The trials, which measured risk factors rather than disease, showed significantly lower body weight, systolic blood pressure and total cholesterol at higher fiber intakes. Benefit was greatest at 25–29 g/day, with dose-response curves suggesting more benefit above that.
• Dietary patterns. Randomized trials of plant-forward patterns point the same way, though none tested a vegan diet. In PREDIMED, a Mediterranean diet supplemented with olive oil or nuts cut major cardiovascular events by roughly 30% compared with a low-fat control, without a detectable difference in overall deaths. In the Lyon Diet Heart Study, a Mediterranean-style diet after a first heart attack cut cardiac deaths plus nonfatal heart attacks from 44 to 14, and all-cause deaths from 24 to 14. And in DASH, a diet rich in fruits, vegetables and low-fat dairy lowered blood pressure by 5.5/3.0 mmHg within eight weeks.
• Legumes and longevity. Higher legume intake tracks with lower mortality across large meta-analyses, and in a cross-cultural study of adults over 70 legumes were the most consistent dietary predictor of survival.
• Fruits and vegetables. A dose-response meta-analysis of 95 prospective studies found about 10% lower all-cause mortality for every additional 200 g a day of fruit and vegetables, with risk continuing to fall up to about 800 g a day.

Note the asymmetry, because it’s the whole ballgame. The carnivore case against plants rests on mechanistic speculation (“this compound is a pesticide,” “this binds that mineral”) extrapolated from insects and petri dishes. The case for plant-rich eating rests on outcome data: actual humans, living longer and getting sick less in large cohorts, with randomized trials confirming fewer cardiovascular events and better risk factors. When mechanism and outcomes disagree, outcomes win.

Part 8. The argument that eats itself

The deepest problem with “these compounds activate stress pathways, therefore they’re poisons” is that the people making it don’t believe it. You can tell, because they endorse a long list of other things that work the exact same way.

• Exercise works through oxidative stress, mechanical muscle damage, and transient inflammation, followed by adaptation. It’s hormesis, activating the same Nrf2 and AMPK pathways as dietary hormetins.
• Sauna, cold plunge, and fasting, staples of the same influencer ecosystem, are textbook hormetic stressors, deliberately imposed for the adaptive rebound.
• Meanwhile the carnivore plate isn’t chemically pristine. Cooked and charred meat generates heterocyclic amines and polycyclic aromatic hydrocarbons (established mutagens) and advanced glycation end-products, and heme iron is a pro-oxidant, one of several suspects in the link between red meat and colorectal cancer, though a 2026 analysis of three large US cohorts, 211,851 people followed for up to 36 years, found no association between heme iron intake and colorectal cancer; our bowel cancer article walks through the candidates.

So the standard is applied to exactly one food group. If “it activates a stress response” condemns broccoli, it condemns your workout, your sauna, and your fast. And if, as the evidence actually shows, a mild and recoverable stress that triggers adaptation makes you stronger, then that is a point in favor of eating a varied plant diet, not against it.

The argument doesn’t just fail. It reverses on the person making it.

Part 9. The clichés, answered in one place

If you’ve argued about this online, you’ve met these lines. They recycle endlessly because they’re short, confident, and sound mechanistic. Here’s each one against what the evidence actually shows.

Do plants not want to be eaten?

Some don’t. Many actively do. Fruit is a seed-dispersal bribe, a plant paying an animal in sugar to move its offspring. Capsaicin is the giveaway: it deters mammals, which grind seeds, but birds lack the receptor, and birds disperse seeds intact. That’s not warfare, it’s negotiation. And “want” is doing a lot of work in a sentence about an organism with no intentions.

Is broccoli trying to kill you?

Broccoli’s defense chemistry is dosed for a caterpillar that eats multiples of its body weight in leaf tissue daily. You are tens of thousands of times heavier and eat roughly a thousandth of your body weight in it, occasionally. The main compound in question, sulforaphane, has randomized human trials showing increased clearance of benzene and acrolein and suppression of H. pylori, an infection strongly associated with stomach cancer.

Are lectins the new gluten?

The entire hazard is raw or undercooked red kidney beans, which carry 20,000–70,000 hemagglutinating units. Proper boiling drops that to 200–400, about a 99% reduction. Meanwhile legumes track with roughly 6% lower all-cause mortality per 50 g/day and lower colorectal cancer risk. This is a cooking instruction masquerading as a food-group indictment.

Will spinach give you kidney stones?

The cited cases are almost uniformly people drinking liters of raw greens daily, often with prior gastric bypass or recent antibiotics. Boiling cuts soluble oxalate by roughly a third to almost 90%, dietary calcium binds oxalate in the gut before it reaches your kidneys, and rotating greens keeps any single load low. If you’re a known stone-former, that’s a reason for oxalate-aware choices, not for deleting vegetables.

Do phytates steal minerals from your food?

Soaking, sprouting, and fermenting cut phytate dramatically (fermentation studies show reductions up to roughly 86%), and vitamin C at meals helps keep iron in an absorbable form. Phytate is also an iron-chelating antioxidant, working mainly in the gut, with broad preclinical anti-cancer activity and observational links to higher bone density and less vascular calcification. The “thief” may be working for you. The honest caveat is that vegetarians do run lower iron stores. A meta-analysis of 24 cross-sectional studies found their ferritin averaged about 30 µg/L below non-vegetarians, a gap much larger in men than in women, and vegetarians more often fell below the cut-offs used to define iron deficiency. That is why menstruating women, pregnant women and endurance athletes should have ferritin checked. The same authors add the other half: high iron stores are themselves a risk factor for conditions such as type 2 diabetes, so they recommend everyone, not only vegetarians, keep an eye on iron status. Our nutrient conversion post walks through that evidence.

Do kale and soy harm your thyroid?

The signature case in the literature is an 88-year-old woman eating 1–1.5 kg of raw bok choy daily for several months. Cooking largely deactivates the enzyme responsible. For soy specifically, a meta-analysis of 18 randomized trials found no effect on thyroid hormones and only a very modest rise in TSH of unclear clinical significance.

Is fiber unnecessary if it isn’t an essential nutrient?

“Essential” is a technical term meaning your body can’t synthesize it, not a synonym for “beneficial.” By that standard exercise is unnecessary too. Reynolds et al.’s 2019 Lancet analysis of 185 prospective studies found 15–30% lower all-cause and cardiovascular mortality at the highest fiber intakes in observational data, with dose-response data suggesting more benefit above 30 g/day, and its 58 clinical trials showed lower body weight, blood pressure and cholesterol. Butyrate, the primary fuel for your colon cells, is made by bacteria fermenting the fiber you were told you don’t need.

Are nutrients from plants poorly absorbed?

Bioavailability is a solvable engineering problem, and the solutions are a lemon wedge, a splash of vinegar, an overnight soak, and drinking your tea between meals. Vitamin C can multiply the iron you absorb from a single meal, though its effect across a whole diet is much smaller. Note also that non-heme iron’s regulated absorption is arguably a feature: the body absorbs more when stores are low and less when they are full, while heme iron’s absorption barely adjusts to need. Our nutrient conversion post explains why that difference matters.

Is plant protein incomplete?

Every whole plant food contains all nine essential amino acids, in varying ratios. The protein-combining rule was popularized by Frances Moore Lappé in 1971, and she walked it back herself in the 1981 tenth-anniversary edition, writing that in fighting the myth that only meat supplies quality protein she had created a new one. Her point was that a varied plant diet with enough calories almost always supplies enough protein. More on this in our nutrition deep dive.

Are humans apex predators built to eat meat?

Our lineage deleted the GULO gene and lost the ability to synthesize vitamin C, an outsourcing that only makes sense with a reliably plant-rich diet. We also carry expanded salivary amylase gene copies for starch digestion, a colon adapted to fermenting plant fiber, and 25 bitter taste receptors. Obligate carnivores have none of that profile.

Can a carnivore diet cure autoimmune disease?

Elimination diets frequently produce short-term improvement; they remove ultra-processed food, alcohol, most additives, and usually a lot of calories, all at once. That tells you something was removed that mattered; it doesn’t identify which thing, and it isn’t outcome data. And there are, as far as we can find, no long-term controlled studies of morbidity or mortality on carnivore diets at any evidence tier. The most cited human data is one social media survey of 2,029 self-selected adults who had followed the diet for at least six months. It had no control group, and everything in it, from health conditions to lab results, was self-reported by the participants. Its senior author discloses royalties from books that recommend a carbohydrate-modified diet.

Is nutritional epidemiology junk science?

This argument disarms the person making it. The plant-forward case has randomized controlled trials sitting on top of the cohort data: PREDIMED and the Lyon Diet Heart Study on hard outcomes, and DASH plus the 58 fiber trials in the Reynolds meta-analysis on blood pressure, cholesterol and body weight. The carnivore case has mechanistic speculation and testimonials. If observational evidence is worthless, the carnivore position has nothing left at all.

Were antioxidants debunked?

What failed were isolated, high-dose supplements. In the ATBC trial of 29,133 Finnish male smokers, beta-carotene supplementation produced an 18% excess in lung cancer incidence and an 8% excess in overall mortality; CARET was stopped early in 1996 after showing 28% more lung cancer and 17% higher overall mortality in the active arm. That result supports the hormesis framing rather than undermining it: mega-dose antioxidants can blunt the very adaptive signaling that makes whole foods and exercise work, an effect Ristow and colleagues suggested in 2009 (PNAS) when vitamin C and E supplementation appeared to blunt exercise-induced insulin sensitivity gains. Whole foods deliver a low-dose, matrixed signal. Pills deliver a sledgehammer.

Do nightshades cause inflammation?

There is no human trial evidence for this. Glycoalkaloid levels in ripe tomatoes, peppers, eggplant, and potatoes are far below any threshold of concern.

Is fruit bad for you because of its sugar?

Whole fruit arrives packaged in fiber, water, and polyphenols, which is why it behaves nothing like refined sugar metabolically. Muraki et al. (BMJ 2013) found whole fruit intake associated with lower type 2 diabetes risk while fruit juice was associated with higher risk. Same sugar, different matrix.

Does soy feminize men?

Testosterone is the claim, and testosterone is where the evidence is cleanest. Reed et al.’s 2021 meta-analysis in Reproductive Toxicology (41 studies, with testosterone measured in 1,753 men) found no effect on total or free testosterone. Worth disclosing: one of its authors consults for companies that sell soy products, which is why the next analysis matters. An independent 2025 dose–response meta-analysis of randomized trials (Rajaie et al., Food Frontiers), whose authors report no conflicts, found no significant pooled effect on testosterone or estradiol. It did find a nonlinear signal at higher doses: estradiol rose by about 6.5 pg/mL around 72 mg/day of isoflavones, and SHBG by about 9 nmol/L around 120 mg/day, both well above the 30–50 mg/day typical of older Japanese adults. Worth knowing. But “a small hormone shift at supplemental doses” is a different sentence from “soy feminizes men.”

Is liver nature’s multivitamin?

It’s genuinely nutrient-dense, and dense enough in preformed vitamin A (roughly 5,000 µg RAE per 100 g raw, about 550% of the Daily Value) that a single modest serving can exceed the adult tolerable upper limit of 3,000 µg, so frequent large servings risk hypervitaminosis A. The plant route has a brake. In experiments in mice and human cell lines, researchers showed that vitamin A itself switches on an intestinal “gatekeeper” protein, ISX, that throttles both the absorption of beta-carotene and its conversion to vitamin A. (Two of the authors were affiliated with DSM Nutritional Products.) Our nutrient conversion post explains the mechanism. More to the point, “dense in some nutrients” doesn’t address the fiber, resistant starch, and phytochemical spectrum that are simply absent from it: the substrates your microbiome runs on.

Are plants full of heavy metals?

This one has a real kernel, so it gets a real answer. Cadmium does accumulate in cereal grains, leafy greens, cacao, and seeds, and plant-based eaters do carry somewhat more of it. In a small 2026 German cross-sectional study (Koeder et al., British Journal of Nutrition) comparing 38 vegans, 37 lacto-ovo-vegetarians and 40 omnivores, vegans had 42% higher urinary cadmium and 48% lower mercury in the standard analysis. That is the concession, and it is worth making plainly.

But when the researchers corrected for urine concentration using creatinine, the usual way to compare urine samples fairly, both the cadmium and the mercury differences disappeared. The gap that held, and widened, was arsenic: 59% lower in vegans and 63% lower in lacto-ovo-vegetarians. A carnivore accounting that counts the cadmium and skips the arsenic is not an accounting.

What settles it is whether any of this produces disease, and here the chain simply breaks. Cadmium’s primary target organ is the kidney. If dietary cadmium were doing meaningful damage, plant-based eaters should show more kidney disease. They show less. In UK Biobank, following 416,584 adults for 12.9 years with 23,084 incident cases, vegetarians had the lowest risk of any dietary group, hazard ratio 0.81. That analysis was underpowered for vegans specifically, so it’s worth adding three cross-sectional Taiwanese studies that weren’t: among patients with hyperuricemia, vegans had a CKD odds ratio of 0.62; among diabetic patients, 0.68; and in a general screening population, 0.87. Note what those first two populations are. People with gout-range uric acid and people with diabetes are the most renally vulnerable groups there are, the ones where any additional kidney insult should surface first. Cross-sectional data can’t show which came first, but vegans in both still came out ahead. All-cause mortality points the same direction: in the Adventist Health Study-2, vegetarians as a group had about 12% lower all-cause mortality than non-vegetarians. And we found no study documenting a heavy-metal-attributable illness in a vegetarian or vegan population.

This also isn’t a hidden explanation for the fracture signal discussed earlier. The researchers studying it point to calcium, vitamin D, body weight, protein and possibly muscle strength and falls, and the fracture pattern fits that: the vegan excess sits in hips and legs, with no significant excess at the wrist or ankle. Diet quality points the same way. Among postmenopausal US women in the Nurses’ Health Study, long-term adherence to a healthful plant-based pattern, rich in whole grains, legumes, nuts and seeds, showed no increase in hip fracture risk, and the most recent intake tracked with about 21% lower risk, though the authors note that finding could partly reflect reverse causation. And in a small study, vegans who did resistance training had bone microarchitecture similar to omnivores who trained. Lifting does not chelate anything.

A worked example may help, since the foods people ask about most tend to cluster in one breakfast bowl: oats, wheat germ, flaxseed, cocoa or cacao powder, beet powder, and sometimes sweet potato or carrot. Here is the honest ranking. Cocoa is the standout. Across 72 cocoa and dark chocolate products tested in the US, cadmium averaged about 0.4 micrograms per gram, so a tablespoon of cocoa powder delivers roughly two micrograms, and products labeled organic were significantly more likely to run high. Origin matters too: a 2019 review summarizing earlier bean measurements reported South American beans averaging about 15 times the cadmium of West African ones. Flaxseed is next, notable enough that the EU sets a higher permitted limit for it (0.5 mg/kg) than for ordinary cereals, so a couple of tablespoons is fine but a half-cup daily is not the place to be casual. Wheat germ and bran run higher in cadmium, which is why the EU allows 0.20 mg/kg in wheat germ and 0.15 in wheat bran against 0.050 in barley and rye, so keep them a topping rather than a base. Oats are low-to-moderate, and the form, rolled, steel-cut or groat, makes no difference to metal content. Beet powder is used by the teaspoon and contributes little; sweet potato and carrot are low-accumulating root vegetables and are not a concern at any normal intake. For a 70 kg adult, EFSA’s tolerable weekly intake of 2.5 µg per kilogram of body weight works out to roughly 25 µg of cadmium a day. EFSA is candid that average European adults already sit close to that line, and that vegetarians, because they eat more cereals, nuts, oilseeds and pulses, can reach roughly double it. A bowl built on cocoa, flax and wheat germ uses a real share of that budget, which is exactly why variety, sourcing and good iron and zinc status matter more than avoidance. This is worth a fuller treatment than a rebuttal post can give it, and one is coming.

Two final points. Smoking roughly doubles cadmium body burden, which dwarfs any dietary difference in the general population. And animal foods are not clean: liver and kidney concentrate cadmium enough that the EU sets separate legal limits for offal, and fish carries the mercury. If you want to lower your own exposure, the highest-leverage moves are keeping iron and zinc status good (zinc, iron and calcium compete with cadmium for absorption, and the iron transporter that ramps up when iron stores run low also carries cadmium; in one Swedish study of women, low iron stores tracked closely with higher blood cadmium), varying your grains instead of eating rice daily, and, if you use cacao daily, preferring West African origin, which runs far lower in cadmium than South American. For rice in particular, cooking it in excess water and draining it removes roughly half the inorganic arsenic, which is a free fix for the one exposure where plant-based eaters genuinely run higher.

Is oxalate dumping real?

This is not a recognized clinical phenomenon. It has no diagnostic criteria, no measurable signature, and no presence in the medical literature. It exists as a post-hoc explanation for why people feel unwell after eliminating plants.

Aren’t plants sprayed with pesticides anyway?

A separate argument, and one that goes badly for the carnivore position once you account for the plants fed to livestock. We handled it in the pesticide argument the carnivore movement can’t answer.

The through-line: every one of these takes a real fact (a real compound, a real case report, a real limitation of epidemiology) and inflates it past what it can carry. None of them survive contact with dose, preparation, or outcome data.

What to actually do

If you’re a healthy person worried by carnivore content: stop worrying and eat the rainbow. Concretely:

1. Aim for variety. Work toward roughly 30 different plants a week across legumes, greens, grains, fruits, vegetables, mushrooms, nuts, seeds, seaweed, herbs and spices. Diversity is both your safety margin against any single compound and your biggest lever on microbiome health.
2. Cook your beans properly: boil dried kidney beans for at least 30 minutes rather than slow-cooking them from raw, and lean on soaking, sprouting, and fermenting for grains and legumes.
3. Deploy the iron hacks: a vitamin C source at meals, a splash of acid on greens and grains, and tea or coffee kept about an hour away from iron-rich meals.
4. Build in calcium: roughly 700–1,000 mg a day from fortified plant milk, calcium-set tofu and low-oxalate greens such as bok choy, kale and broccoli, alongside adequate vitamin D.
5. Cover the genuine gaps that have nothing to do with anti-nutrients: reliable B12, adequate iodine from iodized salt or seaweed, vitamin D as needed, and an omega-3 source.

If you have a specific condition (stone history, thyroid or iodine issues, IBS or SIBO, histamine intolerance), these are conditional adjustments, not elimination orders. Ramp fiber gradually, hydrate, get iodine status right before heavy raw-brassica intake, be oxalate-aware if you form stones, and get individualized help from a lifestyle-medicine clinician.

The thresholds that would change this advice: if you are eating kilograms of a single raw high-oxalate or high-goitrogen food daily, or building your entire diet around one plant, stop. That’s mono-eating, and it’s the one failure mode that’s real. If you have primary hyperoxaluria, advanced kidney disease, or a diagnosed absorptive disorder, your intake genuinely needs clinical tailoring. Short of those, the plant-toxin and anti-nutrient arguments give you no reason to remove a single food group.

Where I’m keeping the claims honest

• Evidence tiers, stated plainly. Sulforaphane’s effects on detoxification biomarkers have randomized human trial support, though not yet on cancer outcomes. IP6’s anti-cancer activity is strong in cells and rodents but rests on a single small human pilot that improved quality of life and reduced chemo toxicity without changing tumor markers. The “lectins fight cancer” idea is preclinical and largely involves purified or injected lectins; the one lectin with a large human trial failed it. The population-level benefit of legumes is real but is credited to fiber and isoflavones, not specifically to lectins. None of these caveats rescue the carnivore argument; they just keep our claims calibrated.
• The legitimate conditional risks are narrow and named: oxalate for stone-formers and extreme intakes; goitrogens with massive raw-brassica intake, especially when iodine is low; phytohaemagglutinin in improperly cooked kidney beans; single-meal mineral binding, which the hacks address. These call for adjustment, not elimination.
• Mechanism labels. Grouping nitrate, beta-glucan, folate, and fiber as “not hormesis” is a claim about mechanism, not a claim that hormesis is unimportant. Sulforaphane and the polyphenols really do work through adaptive stress pathways. The point is only that they are one mechanism among several, not the master explanation.
• Observational versus causal. Much of the outcome evidence (coffee, legumes, fiber and mortality) is observational and carries residual confounding. The strength of the plant case comes from consilience across mechanism, cohorts, and randomized trials all pointing the same way, plus the striking asymmetry versus the purely mechanistic carnivore claims.
• Individual variation is real. Microbiome composition (whether you make urolithin A, whether you carry Oxalobacter) and genetics both modulate these effects, which is exactly why some people transitioning need personalized support.

References

Every study named in this post, with links to the primary source. Where a claim rests on consensus science rather than a single paper, no citation is listed.

Sulforaphane, hormesis, and caffeine

1. Kensler TW, Chen JG, Egner PA, et al. Effects of glucosinolate-rich broccoli sprouts on urinary levels of aflatoxin-DNA adducts and phenanthrene tetraols in a randomized clinical trial in He Zuo township, Qidong, People’s Republic of China. Cancer Epidemiology, Biomarkers & Prevention. 2005;14(11):2605–2613. View source
2. Egner PA, Chen JG, Zarth AT, et al. Rapid and sustainable detoxication of airborne pollutants by broccoli sprout beverage: results of a randomized clinical trial in China. Cancer Prevention Research. 2014;7(8):813–823. View source
3. Andrés CMC, Pérez de la Lastra JM, Juan CA, Plou FJ, Pérez-Lebeña E. Polyphenols as antioxidant/pro-oxidant compounds and donors of reducing species: relationship with human antioxidant metabolism. Processes. 2023;11(9):2771. View source
4. Yanaka A, Fahey JW, Fukumoto A, et al. Dietary sulforaphane-rich broccoli sprouts reduce colonization and attenuate gastritis in Helicobacter pylori-infected mice and humans. Cancer Prevention Research. 2009;2(4):353–360. View source
5. Nathanson JA. Caffeine and related methylxanthines: possible naturally occurring pesticides. Science. 1984;226(4671):184–187. View source
6. Poole R, Kennedy OJ, Roderick P, et al. Coffee consumption and health: umbrella review of meta-analyses of multiple health outcomes. BMJ. 2017;359:j5024. View source
7. Poole R, Kennedy OJ, Roderick P, et al. Correction: Coffee consumption and health: umbrella review of meta-analyses of multiple health outcomes. BMJ. 2018;360:k194. View source

Microbiome

1. Kaufman DW, Kelly JP, Curhan GC, et al. Oxalobacter formigenes may reduce the risk of calcium oxalate kidney stones. Journal of the American Society of Nephrology. 2008;19(6):1197–1203. View source
2. Kelly JP, Curhan GC, Cave DR, Anderson TE, Kaufman DW. Factors related to colonization with Oxalobacter formigenes in U.S. adults. Journal of Endourology. 2011;25(4):673–679. (Abstract only.) View source
3. Shi W, Wang P, Giovannucci EL. Heme iron intake and the risk of colorectal cancer: a prospective analysis of the modifying role of calcium. Cancer Causes & Control. 2026;37:132. (Abstract only.) View source
4. Singh A, D’Amico D, Andreux PA, et al. Urolithin A improves muscle strength, exercise performance, and biomarkers of mitochondrial health in a randomized trial in middle-aged adults. Cell Reports Medicine. 2022;3(5):100633. View source
5. McDonald D, Hyde E, Debelius JW, et al. American Gut: an open platform for citizen science microbiome research. mSystems. 2018;3(3):e00031-18. View source
6. Pereira FC, Ge X, Kostopoulos I, et al. Beyond butyrate: microbial fiber metabolism supporting colonic epithelial homeostasis. Trends in Microbiology. 2023. View source
7. 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-Gastrointestinal and Liver Physiology. 1999;277(5):G935–G943. View source
8. Silvester KR, Bingham SA, Pollock JR, Cummings JH, O’Neill IK. Effect of meat and resistant starch on fecal excretion of apparent N-nitroso compounds and ammonia from the human large bowel. Nutrition and Cancer. 1997;29(1):13–23. View source
9. O’Keefe SJ, Kidd M, Espitalier-Noel G, Owira P. Rarity of colon cancer in Africans is associated with low animal product consumption, not fiber. American Journal of Gastroenterology. 1999;94(5):1373–1380. View source
10. O’Keefe SJ, Chung D, Mahmoud N, et al. Why do African Americans get more colon cancer than Native Africans? Journal of Nutrition. 2007;137(1 Suppl):175S–182S. View source
11. 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. View source

Phytate, lectins, and legumes

1. López-González AA, Grases F, Monroy N, et al. Protective effect of myo-inositol hexaphosphate (phytate) on bone mass loss in postmenopausal women. European Journal of Nutrition. 2013;52(2):717–726. Authors disclose a pending patent and employment at Laboratoris Sanifit. View source
2. 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. View source
3. Bačić I, Družijanić N, Karlo R, et al. Efficacy of IP6 + inositol in the treatment of breast cancer patients receiving chemotherapy: prospective, randomized, pilot clinical study. Journal of Experimental & Clinical Cancer Research. 2010;29:12. View source
4. Yu L, Fernig DG, Smith JA, Milton JD, Rhodes JM. Reversible inhibition of proliferation of epithelial cell lines by Agaricus bisporus (edible mushroom) lectin. Cancer Research. 1993;53(19):4627–4632. View source
5. Steuer-Vogt MK, Bonkowsky V, Ambrosch P, et al. The effect of an adjuvant mistletoe treatment programme in resected head and neck cancer patients: a randomised controlled clinical trial. European Journal of Cancer. 2001;37(1):23–31. View source
6. Zhu B, Sun Y, Qi L, Zhong R, Miao X. Dietary legume consumption reduces risk of colorectal cancer: evidence from a meta-analysis of cohort studies. Scientific Reports. 2015;5:8797. View source
7. Zargarzadeh N, Mousavi SM, Santos HO, et al. Legume consumption and risk of all-cause and cause-specific mortality: a systematic review and dose-response meta-analysis of prospective studies. Advances in Nutrition. 2023;14(1):64–76. View source
8. Papandreou C, Becerra-Tomás N, Bulló M, et al. Legume consumption and risk of all-cause, cardiovascular, and cancer mortality in the PREDIMED study. Clinical Nutrition. 2019;38(1):348–356. View source
9. 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. View source
10. Orlich MJ, Singh PN, Sabaté J, Jaceldo-Siegl K, Fan J, Knutsen S, Beeson WL, Fraser GE. Vegetarian dietary patterns and mortality in Adventist Health Study 2. JAMA Internal Medicine. 2013;173(13):1230–1238. View source
11. 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. View source
12. Thorpe DL, Beeson WL, Knutsen R, Fraser GE, Knutsen SF. Dietary patterns and hip fracture in the Adventist Health Study 2: combined vitamin D and calcium supplementation mitigate increased hip fracture risk among vegans. American Journal of Clinical Nutrition. 2021;114(2):488–495. View source
13. 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. View source
14. Webster J, Greenwood DC, Cade JE. Risk of hip fracture in meat-eaters, pescatarians, and vegetarians: a prospective cohort study of 413,914 UK Biobank participants. BMC Medicine. 2023;21:278. View source
15. Sotos-Prieto M, Struijk EA, Fung TT, et al. Plant-based diets and risk of hip fracture in postmenopausal women. JAMA Network Open. 2024;7(2):e241107. View source
16. Kashyap S, Shivakumar N, Varkey A, Preston T, Devi S, Kurpad AV. Co-ingestion of black tea reduces the indispensable amino acid digestibility of hens’ egg in Indian adults. The Journal of Nutrition. 2019;149(8):1363–1368. View source
17. US Food and Drug Administration. Phytohaemagglutinin. Bad Bug Book: Foodborne Pathogenic Microorganisms and Natural Toxins Handbook, 2nd ed. View source

Oxalate and goitrogen case reports

1. Chu M, Seltzer TF. Myxedema coma induced by ingestion of raw bok choy. New England Journal of Medicine. 2010;362(20):1945–1946. View source
2. Makkapati S, D’Agati VD, Balsam L. “Green smoothie cleanse” causing acute oxalate nephropathy. American Journal of Kidney Diseases. 2018;71(2):281–286. View source

Absorption and food preparation

1. Cook JD, Monsen ER. Vitamin C, the common cold, and iron absorption. American Journal of Clinical Nutrition. 1977;30(2):235–241. View source
2. Hallberg L, Brune M, Rossander L. Effect of ascorbic acid on iron absorption from different types of meals. Human Nutrition: Applied Nutrition. 1986;40(2):97–113. (Abstract only.) View source
3. Siegenberg D, Baynes RD, Bothwell TH, et al. Ascorbic acid prevents the dose-dependent inhibitory effects of polyphenols and phytates on nonheme-iron absorption. American Journal of Clinical Nutrition. 1991;53(2):537–541. View source
4. Cook JD, Reddy MB. Effect of ascorbic acid intake on nonheme-iron absorption from a complete diet. American Journal of Clinical Nutrition. 2001;73(1):93–98. View source
5. Brune M, Rossander L, Hallberg L. Iron absorption: no intestinal adaptation to a high-phytate diet. American Journal of Clinical Nutrition. 1989;49(3):542–545. View source
6. Hunt JR, Roughead ZK. Adaptation of iron absorption in men consuming diets with high or low iron bioavailability. American Journal of Clinical Nutrition. 2000;71. View source
7. Haider LM, Schwingshackl L, Hoffmann G, Ekmekcioglu C. The effect of vegetarian diets on iron status in adults: a systematic review and meta-analysis. Critical Reviews in Food Science and Nutrition. 2018;58(8):1359–1374. View source
8. Lobo GP, Hessel S, Eichinger A, et al. ISX is a retinoic acid-sensitive gatekeeper that controls intestinal β,β-carotene absorption and vitamin A production. FASEB Journal. 2010;24:1656–1666. View source
9. Chai W, Liebman M. Effect of different cooking methods on vegetable oxalate content. Journal of Agricultural and Food Chemistry. 2005;53:3027–3030. View source
10. Heaney RP, et al. Not all calcium-fortified beverages are equal. Nutrition Today. 2005;40(1). (Abstract only.) View source
11. Armah SM, Boy E, Chen D, Candal P, Reddy MB. Regular consumption of a high-phytate diet reduces the inhibitory effect of phytate on nonheme-iron absorption in women with suboptimal iron stores. Journal of Nutrition. 2015;145(8):1735–1739. View source
12. Weaver CM, Proulx WR, Heaney R. Choices for achieving adequate dietary calcium with a vegetarian diet. American Journal of Clinical Nutrition. 1999;70(3 Suppl):543S–548S. View source
13. Nsabimana S, Ismail T, Lazarte CE. Enhancing iron and zinc bioavailability in maize (Zea mays) through phytate reduction: the impact of fermentation alone and in combination with soaking and germination. Frontiers in Nutrition. 2024;11:1478155. View source
14. Hallberg L, Brune M, Rossander L. The role of vitamin C in iron absorption. International Journal for Vitamin and Nutrition Research Supplement. 1989;30:103–108. View source

Fiber, nitrate, beta-glucan, and folate

1. 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. View source
2. Reynolds A, Mann J, Cummings J, Winter N, Mete E, Te Morenga L. Carbohydrate quality and human health: a series of systematic reviews and meta-analyses. Lancet. 2019;393(10170):434–445. View source
3. 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. (Republished after retraction of the 2013 report.) View source
4. de Lorgeril M, Salen P, Martin JL, Monjaud I, Delaye J, Mamelle N. Mediterranean diet, traditional risk factors, and the rate of cardiovascular complications after myocardial infarction: final report of the Lyon Diet Heart Study. Circulation. 1999;99(6):779–785. View source
5. Appel LJ, Moore TJ, Obarzanek E, et al. A clinical trial of the effects of dietary patterns on blood pressure. New England Journal of Medicine. 1997;336(16):1117–1124. View source
6. Bahadoran Z, Mirmiran P, Kabir A, Azizi F, Ghasemi A. The nitrate-independent blood pressure-lowering effect of beetroot juice: a systematic review and meta-analysis. Advances in Nutrition. 2017;8(6):830–838. View source
7. Whitehead A, Beck EJ, Tosh S, Wolever TM. Cholesterol-lowering effects of oat β-glucan: a meta-analysis of randomized controlled trials. American Journal of Clinical Nutrition. 2014;100(6):1413–1421. View source
8. Centers for Disease Control and Prevention. Spina bifida and anencephaly before and after folic acid mandate, United States, 1995–1996 and 1999–2000. MMWR. 2004;53(17):362–365. View source

Antioxidant supplement trials and hormones

1. The Alpha-Tocopherol, Beta Carotene Cancer Prevention Study Group. The effect of vitamin E and beta carotene on the incidence of lung cancer and other cancers in male smokers. New England Journal of Medicine. 1994;330(15):1029–1035. View source
2. Omenn GS, Goodman GE, Thornquist MD, et al. Risk factors for lung cancer and for intervention effects in CARET, the Beta-Carotene and Retinol Efficacy Trial. Journal of the National Cancer Institute. 1996;88(21):1550–1559. View source
3. Ristow M, Zarse K, Oberbach A, et al. Antioxidants prevent health-promoting effects of physical exercise in humans. PNAS. 2009;106(21):8665–8670. View source
4. Reed KE, Camargo J, Hamilton-Reeves J, Kurzer M, Messina M. Neither soy nor isoflavone intake affects male reproductive hormones: an expanded and updated meta-analysis of clinical studies. Reproductive Toxicology. 2021;100:60–67. View source
5. Rajaie S, et al. Effects of soy isoflavones on male reproductive hormones: a dose-response meta-analysis of randomized controlled trials. Food Frontiers. 2025;6(6):3166–3179. View source
6. Lennerz BS, Mey JT, Henn OH, Ludwig DS. Behavioral characteristics and self-reported health status among 2,029 adults consuming a carnivore diet. Current Developments in Nutrition. 2021;5(12):nzab133. View source
7. Otun J, Sahebkar A, Östlundh L, Atkin SL, Sathyapalan T. Systematic review and meta-analysis on the effect of soy on thyroid function. Scientific Reports. 2019;9:3964. View source
8. Muraki I, Imamura F, Manson JE, et al. Fruit consumption and risk of type 2 diabetes: results from three prospective longitudinal cohort studies. BMJ. 2013;347:f5001. View source

Heavy metals

1. Koeder C, Michels KB, Bauer A, Niederreiter L, Kowarschik S, Hannibal L, Huber R, Storz MA. Urinary trace element levels in vegans compared with lacto-ovo-vegetarians and omnivores. British Journal of Nutrition. 2026 Jul 20:1–12. View source
2. Candussi CJ, et al. Vegetarian diet is associated with a lower risk of chronic kidney disease in a population-based study. Scientific Reports. 2026;16:23849. View source
3. Liu HW, Tsai WH, Liu JS, Kuo KL. Association of vegetarian diet with chronic kidney disease. Nutrients. 2019;11(2):279. View source
4. Wu CL, et al. Vegan diet is associated with a lower risk of chronic kidney disease in patients with hyperuricemia. Nutrients. 2023;15(6):1444. View source
5. Hou YC, et al. Vegetarian diet was associated with a lower risk of chronic kidney disease in diabetic patients. Frontiers in Nutrition. 2022;9:843357. View source
6. Carey M, Jiujin X, Farias JG, Meharg AA. Rethinking rice preparation for highly efficient removal of inorganic arsenic using percolating cooking water. PLoS One. 2015;10(7):e0131608. View source
7. 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 & Metabolism. 2022;107(10):2900–2911. View source
8. Kerstetter JE, O’Brien KO, Caseria DM, Wall DE, 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. View source
9. Fenton TR, Lyon AW, Eliasziw M, Tough SC, Hanley DA. Meta-analysis of the effect of the acid-ash hypothesis of osteoporosis on calcium balance. Journal of Bone and Mineral Research. 2009;24(11):1835–1840. View source
10. Taylor H, Webster K, Gray AR, Tey SL, Chisholm A, Bailey KB, Kumari S, Brown RC. The effects of “activating” almonds on consumer acceptance and gastrointestinal tolerance. European Journal of Nutrition. 2018;57(8):2771–2783. View source
11. Kumari S, Gray AR, Webster K, Bailey K, Reid M, Kelvin KAH, Tey SL, Chisholm A, Brown RC. Does “activating” nuts affect nutrient bioavailability? Food Chemistry. 2020;319:126529. View source
12. Rose HE, Quarterman J. Effects of dietary phytic acid on lead and cadmium uptake and depletion in rats. Environmental Research. 1984;35(2):482–489. View source
13. Rimbach G, Pallauf J, Brandt K, Most E. Effect of phytic acid and microbial phytase on Cd accumulation, Zn status, and apparent absorption of Ca, P, Mg, Fe, Zn, Cu, and Mn in growing rats. Annals of Nutrition and Metabolism. 1995;39:361–370.
14. Graf E, Empson KL, Eaton JW. Phytic acid: a natural antioxidant. Journal of Biological Chemistry. 1987;262(24):11647–11650.
15. Graf E, Eaton JW. Antioxidant functions of phytic acid. Free Radical Biology & Medicine. 1990;8:61–69.
16. Rimbach G, Pallauf J. Phytic acid inhibits free radical formation in vitro but does not affect liver oxidant or antioxidant status in growing rats. Journal of Nutrition. 1998;128(11):1950–1955. View source
17. Pujol A, Sanchis P, Grases F, Masmiquel L. Phytate intake, health and disease: “Let thy food be thy medicine and medicine be thy food.” Antioxidants. 2023;12(1):146. View source
18. Pires SMG, Reis RS, Cardoso SM, et al. Phytates as a natural source for health promotion: a critical evaluation of clinical trials. Frontiers in Chemistry. 2023;11:1174109. View source
19. Hands JM, Anderson ML, Cooperman T, Balsky JE, Frame LA. A multi-year heavy metal analysis of 72 dark chocolate and cocoa products in the USA. Frontiers in Nutrition. 2024;11:1366231. View source
20. Fechner C, Greiner M, Heseker H, Lindtner O. Dietary exposure assessment of aluminium and cadmium from cocoa in relation to cocoa origin. PLOS ONE. 2019;e0217990. View source
21. Agency for Toxic Substances and Disease Registry. Toxicological Profile for Cadmium. US Department of Health and Human Services; 2012. View source
22. EFSA Panel on Contaminants in the Food Chain. Scientific opinion on cadmium in food. The EFSA Journal. 2009;980:1–139.
23. Commission Regulation (EU) 2023/915 of 25 April 2023 on maximum levels for certain contaminants in food. Official Journal of the European Union. 2023;L 119:103.
24. Berglund M, Åkesson A, Nermell B, Vahter M. Intestinal absorption of dietary cadmium in women depends on body iron stores and fiber intake. Environmental Health Perspectives. 1994;102:1058–1066.
25. Shams-White MM, Chung M, Fu Z, Insogna KL, Karlsen MC, LeBoff MS, et al. Animal versus plant protein and adult bone health: a systematic review and meta-analysis from the National Osteoporosis Foundation. PLOS ONE. 2018;13(2):e0192459. View source

Books and primary documents

1. Lappé FM. Diet for a Small Planet, 10th anniversary edition. New York: Ballantine Books; 1981:162.

Further reading

• Do we need animals for any nutrient?
• Raw versus cooked: why cooking is a tool, not the enemy
• What the longest-living people actually eat
• The Adventist Health Study
• The seed oil question
• The pesticide argument the carnivore movement can’t answer
• Cholesterol
• B12
• Vegan bone density: the honest fracture data and how to protect against it
• Preformed nutrients versus plant precursors: why animal nutrients look better on paper
• Diet and bowel cancer

Last updated: 2026-10-06 12:52 UTC