Short version: “carbohydrate” is not a food. It is a chemical category that stretches from a lentil to a lollipop, and treating those two as the same thing is the single most expensive mistake in modern nutrition. In whole plant foods, carbohydrate never arrives alone. It comes packaged with fiber, water, antioxidants, nitrates, minerals, vitamins, all nine essential amino acids, and the essential fatty acids. That package is why the National Academies put carbohydrate at 45 to 65 percent of calories, why the longest-lived populations on earth build their plates on beans and grains, and why people have reversed type 2 diabetes while deliberately eating more carbohydrate. The problem the low-carb movement identified is real. They just misidentified the culprit.

This post walks the whole thing: what a carbohydrate actually is at the molecular level, what happens to it from the moment you think about food until your gut bacteria are done with it, why plants are less calorically dense, why fiber is the most underrated substance in your diet, what insulin resistance is really caused by, whether sugar feeds cancer, and how to shop for and cook this stuff without spending much money. Claims were checked against their sources, primary studies wherever possible, and where a figure rests on an abstract, a summary, or evidence thinner than the headline, the text says so.

One clarification first, because it prevents most of the confusion downstream. Everything below assumes a well-planned way of eating: legumes, greens, grains, fruits, vegetables, mushrooms, nuts, seeds, seaweeds, herbs and spices, with variety inside each group, plus a reliable vitamin B12 supplement and vitamin D as needed. A diet of refined flour, fried food and soda that happens to contain no animal products is a different thing entirely, and as you will see below, the research says so emphatically.

This is health information, not medical advice. Anyone with questions about their own diet should talk to a physician, registered dietitian, or nutritionist, ideally one who practices lifestyle medicine.

Quick links to sections

• The executive summary
• What a carbohydrate actually is
• The journey of a carbohydrate, from thought to colon
• Fiber, your bacteria, and the most underrated molecule in your body
• Why the whole food beats isolated stand-ins
• Why plants let you eat more and weigh less
• The protein misunderstanding, in reverse
• The 64x number, and what it actually means
• What the National Academies actually recommend
• How whole carbohydrates got demonized
• The natural experiment: animal low-carb versus plant low-carb
• The staples of the healthiest diets on earth
• Whole grains: one piece of the puzzle
• Legumes: another piece of the puzzle
• It is not the carbohydrates. It is the fat inside your cells.
• People have reversed type 2 diabetes eating more carbohydrate
• Why glycemic load matters at all
• What a continuous glucose monitor is actually telling you
• If you want to go low-carb, you can do it on plants
• Muscle is a glucose sink
• How someone actually becomes prediabetic
• Calories, saturated fat or sugar
• What “metabolically healthy” actually means
• Does sugar feed cancer?
• Carbohydrates are not “essential,” and it does not matter
• How much carbohydrate is actually in each food group
• A tour of the categories, and what is special about each
• Lectins, phytates, and the antinutrient panic
• The practical part: cheap, easy, and how to prepare it
• Green banana to monk fruit: what different sweet foods do in your blood
• If switching feels bad at first
• The goal is not to manage carbohydrates. It is to stop needing to.
• The bottom line
• How this post was checked

The executive summary

If you read nothing else, read this.

• Carbohydrate quality matters far more than carbohydrate quantity. Refined flour and added sugar behave nothing like intact beans, grains, fruits and vegetables, because the fiber matrix that governs absorption has been stripped out along with most of the nutrition.
• A major WHO-commissioned synthesis is literally titled “Carbohydrate quality and human health.” Pooling 185 prospective studies and 58 trials for the World Health Organization, it found the highest fiber consumers had 15 to 30 percent lower all-cause and cardiovascular mortality, with benefit greatest around 25 to 29 grams a day and dose-response curves suggesting more could be better.93
• Whole plant foods carry roughly 64 times the antioxidant content of animal foods by average, and about 9 times by the more conservative median.1 Animal foods contain no fiber whatsoever.
• The same low-carb diet gives opposite results depending on whether it is built on animals or plants. Animal-based low-carb scoring was tied to 23 percent higher mortality; plant-based low-carb scoring to 20 percent lower.94 The macronutrient label was held constant. Only the source changed.
• The National Academies set carbohydrate at 45 to 65 percent of calories. The floor of that range is higher than the ceiling of both the protein range and the fat range, which cap at 35 percent each.2
• Fat carries 9 calories per gram; carbohydrate carries 4, and fiber roughly 2.2 Add the water in plants and you can eat a much larger volume of food for far fewer calories.
• Insulin resistance is driven by fat inside your muscle and liver cells, not by the carbohydrate in whole foods. This was first shown nearly a century ago and has since been confirmed with modern imaging.2525125649 What puts that fat there is mainly a calorie surplus. Per calorie, saturated fat builds the most liver fat, added sugar comes next, and unsaturated fat the least.185
• People with type 2 diabetes have come off insulin entirely on diets that were roughly 70 to 73 percent carbohydrate. In one six-month intervention every participant eliminated insulin treatment; in a randomized trial, adding about 100 grams of carbohydrate a day beat a recommended diabetes diet on every glycemic measure.5657
• The whole food beats isolated stand-ins. When researchers compared whole black beans against meals matched for the same fiber and the same antioxidant capacity, only the beans significantly lowered the after-meal insulin response, an effect the matched fiber and antioxidants could only partly explain.3
• Carbohydrate quality is measurable, and it flips the sign. In the Harvard cohorts, a plant-based diet built on whole foods was associated with 25 percent lower coronary heart disease, while one built on refined grains, fries, sweets and sugary drinks was associated with 32 percent higher.106
• Whole grains are one of the best-supported foods, as part of the pattern. In a meta-analysis of 12 food groups, each daily serving of whole grains was associated with 8 percent lower mortality, and whole grains were the only group with high-quality evidence. Combining the optimal intakes of whole grains, vegetables, fruit, nuts, legumes and fish was estimated, in the authors’ own calculation, to lower the risk of death by 56 percent.211
• Legumes earn their place differently. In randomized trials they modestly lower LDL cholesterol and fasting glucose, and in cohorts higher intake is linked to 6 percent lower mortality, with the clearest benefit when they replace processed meat.231227220
• Variety is the strategy. Across legumes, grains, greens, fruits, vegetables, mushrooms, nuts, seeds, seaweeds, herbs and spices, there are tens of thousands of edible options, and almost nobody is using more than a few dozen of them.

What a carbohydrate actually is

What is a carbohydrate, in plain English?

A carbohydrate is a molecule built from carbon, hydrogen and oxygen, and at its simplest it is a sugar built on six carbon atoms, usually drawn as a ring. Everything else in the category is those rings linked together in longer and longer chains. That is the whole family, sorted by chain length:

• Monosaccharides (one ring). Glucose is the body’s main fuel and the unit almost everything else is built from. Fructose is the sugar in fruit, and it is handled mainly by the liver rather than being used directly by cells. Galactose rarely appears alone.
• Disaccharides (two rings). Sucrose is glucose plus fructose, and it is what you call table sugar. Lactose is glucose plus galactose, and it is the sugar in milk, which is the reason animal foods are not quite carbohydrate-free. Maltose is two glucose units and shows up when starch is broken down.
• Oligosaccharides (a few rings, roughly three to ten). This is where raffinose and stachyose live, the compounds in beans that your own enzymes cannot break down and your gut bacteria find delicious. They are the reason beans have a reputation, and also part of the reason beans are so good for you.
• Polysaccharides (hundreds or thousands of rings). Starch is how plants store glucose. Glycogen is how animals store it, including in your own liver and muscle. And cellulose, along with pectins, beta-glucans and the rest, is fiber.

Notice the last line, because it is the one people miss. Fiber is a carbohydrate. It is not a separate nutrient that happens to travel alongside carbohydrates. It is made of the same sugar rings as starch, assembled with a different bond. When someone says they are avoiding carbohydrates, fiber is one of the things they are avoiding, which is why low-carbohydrate diets and low-fiber diets are usually the same diet wearing different clothes.

This is a deliberately quick tour. The fuller treatment of how these molecules behave, alongside fats, proteins and micronutrients, belongs in the nutrition post.

Why can your body digest starch but not fiber?

It comes down to how the sugar rings are joined. Starch is glucose linked by bonds that alpha-amylase, the starch-cutting enzyme your body makes, is built to break. Starch itself comes in two forms. Amylose is a long, straight chain that coils into a tight helix, and it is digested slowly. Amylopectin is heavily branched, which gives amylase many places to cut at once, so it is digested very quickly.243 The fiber in plant cell walls uses different linkages. In wheat, for instance, the main fibers are cellulose, arabinoxylan and beta-glucan, and arabinoxylan and beta-glucan are built on beta-linked backbones.243 Fructans such as inulin, found in chicory and Jerusalem artichoke, are chains of fructose joined by beta bonds.245 Human enzymes cannot break these down, so they pass through the small intestine undigested, and some of them are then broken down by gut bacteria, which have the enzymes we lack.244

Two refinements keep this honest. First, the bond is not the whole story. Some ordinary starch escapes digestion because of its physical structure: amylose’s tight helix and its tendency to recrystallize after cooking and cooling make it harder for amylase to reach.243 That is the resistant starch covered in the practical section. Second, fiber is defined by what your body cannot do, not by a single chemical shape. The international consensus definition describes fiber as plant carbohydrates, plus small amounts of associated compounds such as lignin, that are not digested or absorbed in the human small intestine, with plant cell walls as the main source. The same consensus distinguishes intact cell-wall fiber from purified fiber, because the three-dimensional structure of the cell wall confers benefits beyond fiber isolates, which is the supplement trap described below.244

What is the difference between simple and complex carbs?

Chain length, technically. One or two rings is “simple,” many rings is “complex.” But that distinction, on its own, is close to useless, and this is where most nutrition writing goes wrong. White bread is a complex carbohydrate. So is an oat groat. They do almost nothing alike in your body.

The distinction that actually predicts what happens to you is intact versus stripped. In a whole plant food, the starch is locked inside cell walls, wrapped in fiber, and accompanied by everything else the plant made. Grind it into flour or refine it and you have removed the packaging, the brake, and most of the nutrition, leaving glucose that hits your bloodstream almost as fast as sugar water.

What makes fiber different from starch?

As the previous answer explains, the difference starts with the bond: your amylase can cut starch apart, but your own enzymes cannot touch fiber. So starch is absorbed in your small intestine, while fiber travels intact to your colon, where an ecosystem of bacteria is waiting for it.

Fiber splits into two functional types, and the difference between them is the difference between two different jobs.

Soluble fiber (pectins, beta-glucans, gums, inulin) dissolves in water, and the viscous kinds, such as beta-glucan and pectin, form a gel. It slows stomach emptying, slows glucose absorption, and binds bile acids. Critically, it is the readily fermentable kind, meaning it is what your gut bacteria actually eat.

Insoluble fiber (cellulose, lignin, most wheat bran) does not dissolve. It adds bulk, holds water, and speeds transit. Most of it passes through without being fermented, which is exactly why it works as a bulking agent.

This distinction gets flipped around constantly, so it is worth stating plainly: if you want to feed your bacteria, you want soluble and fermentable fiber, not insoluble fiber. Insoluble fiber is doing something valuable, but it is mechanical, not microbial. Most whole plant foods contain both, which is the convenient part. Oats, beans, barley, apples, and flax lean soluble. Wheat bran, leafy greens, and vegetable skins lean insoluble.

Then there is a third category that nobody knew existed until 1982: resistant starch. It is starch that behaves like fiber, escaping digestion in the small intestine and arriving in the colon to feed bacteria. It occurs naturally in grains, beans, seeds and some nuts, and beans are far and away the best source.13

If humans cannot digest fiber, why does it matter when we are not cows?

This objection comes up constantly and it contains one true premise and one enormous non sequitur.

The true part: you cannot digest fiber. You have no enzyme that cuts the beta bonds in fiber, and you never will. That is correct and nobody disputes it.

The non sequitur is the leap from “I cannot digest it” to “therefore it does nothing.” You do not digest fiber. You ferment it. Or rather, roughly 38 trillion bacteria living in your large intestine ferment it on your behalf, and you absorb what they produce. You are not the one doing the digesting, which is precisely the point.

The cow comparison actually undercuts the argument rather than supporting it. Yes, cattle are foregut fermenters with a four-chambered stomach, and humans are not. But humans are hindgut fermenters, like pigs, horses, and every other great ape. Your colon is a fermentation chamber. It is not a waste pipe with bacteria that happen to be squatting in it; it is a functioning digestive organ, and it is substantially larger relative to body size than you would expect in an animal that had no use for plant fiber.

The scale is not trivial either. Reviewing the evidence for carbohydrate fermentation in the human large intestine, McNeil estimated that this process supplies 5 to 10 percent of human energy requirements.140 You are getting a measurable fraction of your calories from a substance you are told you cannot digest.

And calories are the least interesting part of it. The real payoff is what the bacteria make while they eat, which is the subject of the fiber section, plus the bulk, the bile acid binding, and the slowed absorption. Judging fiber by whether your own enzymes can break it down is like judging a library by whether you can eat the books.

The journey of a carbohydrate, from thought to colon

We have walked this road before with fat, in Animal Fat, Plant Fat, and the Journey From Your Plate to Your Arteries, and with whole meals in the anti-inflammatory diet post. This is the carbohydrate version, and it starts before the food is even in your mouth.

Why does your mouth water before you eat?

Because your body starts digesting carbohydrate on anticipation. The sight, smell, or even the thought of food triggers what physiologists call the cephalic phase, and your salivary glands release amylase, an enzyme whose only job is to start cutting starch into shorter chains. Chew a piece of bread long enough without swallowing and it turns sweet in your mouth. That is amylase, working in real time.

This is not a trivial detail. Humans carry multiple copies of the salivary amylase gene, and populations with historically starch-heavy diets tend to carry more of them. We are, at the level of our genome, a starch-eating animal, which is one of the threads we pull in the post on human evolution.

What does chewing actually accomplish?

Mechanical demolition. Plant cells are enclosed in walls made of cellulose, and your teeth rupture them so the contents can be reached. What you swallow is a bolus, a compressed lump of ruptured plant tissue already mixed with enzymes. Chew badly and you leave more of that work undone. (There is a case report of a man who habitually ate without chewing and turned up with gastrointestinal bleeding from a swallowed plastic bread-bag clip, which is either a cautionary tale or an argument for slowing down.92)

What happens in the mouth with leafy greens?

Something genuinely strange and wonderful, and it is one of the best arguments for the package-deal thesis in this entire post.

Leafy greens and beets are loaded with nitrate. You swallow it, absorb it, and then your salivary glands actively concentrate it back into your saliva, which makes no sense until you learn what happens next. On the back of your tongue lives a specific community of bacteria that reduces nitrate to nitrite. You swallow the nitrite, your stomach acid and your blood vessels convert it into nitric oxide, and nitric oxide is the molecule that tells your arteries to relax and open.8

The result is lower blood pressure, better endothelial function, and reduced platelet stickiness.10 And here is the evidence that the bacteria are doing the work: rinse with antiseptic mouthwash, kill those bacteria, and the blood pressure benefit of dietary nitrate is abolished. That has been shown in controlled human studies, both with chlorhexidine rinsing alone and with beetroot juice given after mouthwash.107108 Separately, in an observational study of adults free of diabetes, those carrying more of these nitrate-reducing oral bacteria had lower insulin resistance and lower fasting glucose.9 You have a symbiotic relationship with microbes on your tongue that converts salad into blood pressure medication.

Worth noting the mirror image: the nitrite added to processed meat meets abundant amines and heme iron under high heat and forms nitrosamines instead, which is part of why processed meat is classified as a Group 1 carcinogen. Same nitrogen, opposite outcome, and the difference is the plant matrix wrapped around it. We covered that chemistry in detail in the cholesterol post.

What do the pancreas and gallbladder do with carbohydrates?

The stomach mostly churns and acidifies; amylase is inactivated by stomach acid, so starch digestion pauses there. In the small intestine, the pancreas releases a fresh flood of amylase, along with enzymes that finish cutting the shorter chains into single glucose units at the intestinal wall. Those glucose molecules cross into your bloodstream and head for the liver.

The gallbladder releases bile, which matters here for two reasons. First, whole plant foods that carry fat (nuts, seeds, avocados) need bile to absorb their fat-soluble antioxidants. Second, bile acids are made from cholesterol, and what happens to them downstream is one of the most useful things fiber does, which we get to next.

Fiber, your bacteria, and the most underrated molecule in your body

How does fiber lower cholesterol?

Your liver pours roughly a gram of cholesterol and a large load of bile acids into your gut every day, and normally most of it is reabsorbed and recycled.174 Soluble fiber binds those bile acids and escorts them out in the stool. The liver then has to pull cholesterol out of your bloodstream to manufacture replacements. This is not a folk theory; it is the exact mechanism of an entire prescription drug class, the bile acid sequestrants. Oatmeal and beans are a food-strength version of the same mechanism. We laid out the full picture, including the electrostatic trapping of ApoB particles in the artery wall, in the cholesterol post.

Do carbohydrates lower cholesterol?

The right ones do, and the evidence comes from randomized trials.

• Oats. Across 28 randomized trials, eating at least 3 grams a day of oat beta-glucan lowered LDL cholesterol by 0.25 mmol/L, about 10 mg/dL, and total cholesterol by 0.30 mmol/L, with no change in HDL or triglycerides. For scale, one trial in the analysis delivered 3.6 grams from 84 grams of oatmeal. The effect was larger in people who started with higher LDL and in people with type 2 diabetes, and the authors note it depends on the fiber staying viscous in the gut. The analysis was supported by DSM Nutritional Products, which is worth knowing.156
• Even a short burst of oats. A small 2026 randomized trial of a modernized version of a century-old oat cure lowered LDL cholesterol in just two days. The next answer covers what it found, including a less impressive six-week result.170
• Beans, lentils, chickpeas and peas. Across 26 randomized trials in 1,037 people, about one serving a day of pulses (a median of 130 grams) lowered LDL by 0.17 mmol/L, roughly 5 percent. Effects on apoB and non-HDL cholesterol were not statistically significant, most trials were short and of low quality, and results varied a great deal between studies. A modest, real effect rather than a dramatic one.157
• Added sugar goes the other way. As covered in the sugar section further down, higher free-sugar intake raised LDL, total cholesterol and triglycerides in randomized trials.152

So carbohydrates do not lower cholesterol as a category, any more than they raise it as a category. Intact, fiber-rich carbohydrates lower it; refined sugars push it up. For how LDL particles actually become plaque, see the cholesterol post and the plaque post.

Do oats lower cholesterol through more than their fiber?

Probably, and the newest evidence points at your gut bacteria. The classic explanation is beta-glucan, the viscous soluble fiber in oats, trapping bile acids as described above. A German team running two small randomized trials in people with metabolic syndrome, with 17 people per group, found a second route. Oats carry phenolic compounds, and after an oat diet, blood levels of ferulic acid rose; with the higher-dose diet, so did dihydroferulic acid, a phenolic metabolite produced with the help of gut bacteria. The authors concluded that these microbially produced phenolic metabolites are driving factors in the cholesterol-lowering effect of oats. Even two days of a high-dose oat diet, a modern version of the old oat cure described in the insulin section, lowered LDL cholesterol by 10 percent and total cholesterol by 8 percent, though that diet was also calorie-restricted. A separate six-week trial was more sobering: one oat meal a day added to a typical Western diet, without cutting calories, had only a mild effect, and cholesterol stayed largely stable.170 A single bowl added to an otherwise unchanged diet is not the same thing as a whole-food pattern.

Oats also reshape the microbiome. In a randomized trial of 210 Chinese adults with mildly high cholesterol, 80 grams of oats a day for 45 days lowered total cholesterol more than the same amount of rice, though cholesterol fell in both groups, and the oats increased specific bacteria including Akkermansia muciniphila and Roseburia. Several of that trial’s authors work for PepsiCo, which makes Quaker oats. Shifts in particular species like these are a narrower finding than the broad microbiome change that a Danish whole-grain trial looked for and did not find, as the whole grains section notes.248 Smaller trials point the same way. In a crossover trial in Thailand, 70 grams of oatmeal a day, supplying 3 grams of beta-glucan, lowered LDL cholesterol 10 percent from baseline over four weeks, more than rice porridge did.249 In a six-week trial of 166 Chinese adults, 100 grams of instant oatmeal a day lowered total and LDL cholesterol and waist size compared with wheat noodles.250 Both of those trials also had PepsiCo-affiliated authors, which is worth knowing, though their direction matches the independent two-day trial above.

The long-term association fits. In a Danish cohort of 54,871 adults aged 50 to 64, followed for 13.6 years, each extra 25 grams of oatmeal a day was associated with 16 percent lower risk of heart attack in men, though not significantly in women; rye bread was associated with lower risk in both.247 In three Harvard cohorts of about 205,000 adults followed for an average of 25.8 years, eating oatmeal at least twice a week was associated with 21 percent lower coronary heart disease than eating it less than once a month. The benefit leveled off at about half a serving a week, so the association does not require oatmeal every day, and popcorn was the only whole grain food in the study not linked to lower risk.251

What does fiber do in the large intestine?

It arrives essentially untouched, and it is not waste. It is dinner.

Your colonic bacteria ferment fiber and resistant starch into short-chain fatty acids: acetate, propionate and butyrate. In people examined at autopsy, total concentrations were about 131 millimoles per kilogram of colon contents at the start of the colon, falling to about 80 further along.175 The three have different destinations.

Butyrate mostly stays put and fuels the cells lining your colon. Propionate travels to the liver, where it influences glucose production and cholesterol synthesis and contributes to satiety signaling. Acetate reaches the general circulation and can be taken up by muscle, other peripheral tissues, and the brain, where it participates in appetite regulation. One substrate, three messengers, three organs.

Butyrate deserves its own paragraph, because what it does to a colon cell is remarkable. In plain terms, the cell takes butyrate into its tiny internal power plants and burns it for energy, the same way the rest of your body burns glucose. In the technical version: it is carried into the mitochondria and run through beta-oxidation into acetyl-CoA, which enters the citric acid cycle and drives oxidative phosphorylation to produce ATP. Butyrate is a major energy source for colon cells, though recent work shows they also draw on glucose and other microbial products, particularly for building new cells.28

Read that again, because it is genuinely strange. The cells lining your colon lean heavily on a fuel that bacteria manufacture out of the fiber you ate. This is why the relationship is properly called symbiosis rather than tolerance. If you stop eating fiber, you are starving the lining of your own intestine of its preferred fuel.

That energy supply has a consequence most people never hear about. Well-fed colonocytes consume large amounts of oxygen, which keeps the gut lining slightly oxygen-poor, and that low-oxygen environment is exactly what your beneficial anaerobic bacteria require while it suppresses oxygen-tolerant pathogens. Feed the bacteria, they fuel the cells, and the fueled cells maintain the conditions the good bacteria need. Take the fiber away and the loop runs backwards.

What else does butyrate do?

Three jobs from one molecule, which is a nice illustration of why reductionist nutrition keeps failing.

1. Fuel. As above.
2. Gene regulation. Butyrate is a histone deacetylase inhibitor, meaning it changes which genes get read. In cancer cell lines, that same activity induces apoptosis, programmed cell death.24
3. Barrier and immune signaling. It supports the tight junctions that keep your gut lining sealed, calms inflammation, and signals through free fatty acid receptors that influence appetite, systemic inflammation, and the gut-brain axis.27

Do you have to eat fiber every day?

The cell data suggest yes, and it is a striking experiment. Leave colon cancer cells alone and they grow. Expose them to the butyrate concentration your bacteria produce when you eat fiber, and growth stops. Withdraw the butyrate, and growth resumes.23 The protection is not banked; it is produced continuously, and it stops when the substrate stops.25 Fiber-rich whole plant foods are a daily practice, not a weekly gesture. We go deeper on this in the colon cancer post, and we take apart a widely misquoted 2026 pooled analysis in why a 40 percent number in one subgroup did not overturn the meat and bowel cancer link.

Are Americans getting enough fiber?

No, and the number is worse than most people guess. Americans get approximately half the recommended daily amount, and only about 5 percent of the population meets the recommendation.36 Fiber comes from plants and fungi, never from animal foods, so this is a direct readout of how little whole plant food the country eats.

Does fiber actually lower your risk of dying, and how much do you need?

Yes, and the best available answer comes from a synthesis commissioned for the World Health Organization that pooled 185 prospective studies and 58 clinical trials. The paper’s title is worth quoting because it is the thesis of this entire post: “Carbohydrate quality and human health.”93

The highest fiber consumers had 15 to 30 percent lower all-cause and cardiovascular mortality, and 16 to 24 percent less heart disease, stroke, type 2 diabetes and colorectal cancer. The benefit was greatest around 25 to 29 grams a day, and dose-response curves suggested higher intakes could confer even greater benefit.

Now hold that next to the fact that meat, fish, eggs and dairy contain no fiber to speak of. An all-animal diet scores a structural zero on one of the dietary factors most consistently associated with living longer. That is not an argument about macronutrient ratios. It is an argument about which foods contain the thing.

Can fiber help clear forever chemicals?

People ask constantly whether fiber pulls toxins out of the body. For heavy metals the honest answer is that the popular claim runs well ahead of the evidence, and this post covers it in the companion piece on plant toxins and antinutrients. But there is one class of toxin where fiber has a genuinely strong, mechanistically clean, human-tested effect, and it happens to be one of the most talked-about contaminants of the decade.

PFAS, the per- and polyfluoroalkyl substances known as “forever chemicals,” are in roughly a third of groundwater globally and are classified by the WHO’s cancer agency as carcinogenic. They earned the “forever” label because their half-lives in the human body run for years. The CDC’s range for PFOS alone spans roughly 3 to 27 years.

Why so long? Because of a loop. Your liver does dump PFAS into bile and sends it into your intestine. But intestinal transporters efficiently reabsorb it, so it goes straight back into circulation. The same enterohepatic recirculation that recycles your bile acids also recycles your forever chemicals.

Why would fiber interrupt that loop?

Because of a chemical coincidence that is genuinely remarkable. To your gut, a long-chain PFAS molecule looks almost identical to a bile acid: roughly the same size and weight, the same kind of electrical charge, the same water-and-fat-loving shape, and it even rides through the same doorways into your gut wall. As far as your intestine can tell, one is wearing the other’s uniform.

Which means the viscous gel-forming fiber that traps bile acids and carries them out, the mechanism described a few paragraphs above, should trap PFAS by the same action. Same net, different fish.

Does it actually work?

Three lines of evidence, in ascending order of relevance.

The drug proof of concept. Among 56,175 adults in the C8 Health Project, the 36 people taking the bile acid sequestrant drug cholestyramine had significantly lower serum PFAS, most notably PFOS.142 That supports the idea that interrupting bile acid recirculation lowers PFAS in humans.

The mouse study. Mice exposed to a seven-PFAS mixture in drinking water were fed either oat beta-glucan or inulin. The beta-glucan mice showed lower serum concentration trends for several PFAS, plus lower body fat ratios and lower liver and intestinal triglycerides.143 Two honest caveats: those were trends at p<0.1, not significance, and the beta-glucan mice drank more water and therefore received a larger PFAS dose. The comparator matters though: inulin is fermentable but does not form a gel, which points at viscosity rather than fermentation as the active property.

The human trial. In a randomized, double-blind, placebo-controlled trial in Toronto, 72 men with elevated cholesterol took 3 grams a day of oat beta-glucan or a rice powder control for four weeks, with 17 PFAS measured.144 The seven priority long-chain PFAS decreased significantly only in the beta-glucan group, driven mainly by PFOA and PFOS.

One correction worth making, because it is being reported incorrectly: total PFAS fell in both groups, and there was no significant difference between them. The beta-glucan advantage was specific to the long-chain priority compounds, which is exactly what the bile acid mechanism predicts, since chain length is what determines whether a PFAS molecule resembles a bile acid. The selective result is more scientifically interesting than the overstated one, but it is not the same claim.

Supporting this, an analysis of national survey data found that serum concentrations of several perfluoroalkyl acids appear to be reduced by dietary fiber intake.145

Wait, doesn’t fiber bind metals the same way?

This is worth a beat, because it prevents a natural confusion. Fiber does not have one mechanism, it has two, and they are unrelated.

• Ionic binding requires a negative electrical charge, which only certain fibers carry (pectin and alginate do; beta-glucan does not). This is the pathway people invoke for heavy metals, and as the plant toxins post covers, its human evidence is weak.
• Bile acid sequestration requires viscosity and gel formation, not charge at all. Beta-glucan is outstanding at it, which is exactly why it lowers LDL cholesterol, and it is the pathway that traps PFAS. This one has a randomized human trial behind it.

So beta-glucan is a poor metal-binder and an excellent forever-chemical trap, with no contradiction between those facts once you see that they run on different machinery. Psyllium and pectin are strong bile acid sequestrants too, so they belong on the PFAS list alongside oats and barley.

How much oatmeal is that?

A normal amount, which is the rare and pleasant part. The trial used 3 grams of beta-glucan a day; in an oat trial in Thailand, 70 grams of oatmeal supplied 3 grams.249 That is an ordinary breakfast rather than a research protocol. Barley contains close to twice as much beta-glucan as oats, and beta-glucans are also found in mushrooms, though the fungal form is structurally different and may not behave identically.

There is a larger implication buried in this evidence. If a compound with a multi-year half-life can be meaningfully reduced in weeks by interrupting its reabsorption, then part of what makes forever chemicals “forever” may be a low-fiber diet. On a diet with almost no viscous fiber, there is no net in the loop, so the chemical recirculates indefinitely. That reframes fiber as something closer to continuous, passive detoxification, running quietly every day, which is a far better description of what it does than any cleanse.

Can any foods help the body clear pollutants itself?

Yes, and it is not fiber at all. It is the sulfur compounds in cruciferous vegetables, working through a completely different route: rather than binding a toxin in the gut, they upregulate your own detoxification enzymes.

Sulforaphane, a compound your body makes when you chew broccoli, cabbage, kale or Brussels sprouts, does not grab toxins in the gut at all. Instead it flips a master switch inside your cells, called NRF2, that turns up your own built-in detox crew: a set of enzymes that tag toxins for disposal and escort them out. In a 12-week randomized trial of 291 people in a heavily polluted region of China, a broccoli sprout beverage produced rapid and sustained increases in urinary excretion of the carcinogen benzene (61 percent) and the irritant acrolein (23 percent) compared with placebo.141

That is a real randomized human trial with a measured excretion endpoint, which is more than the fiber-and-metals literature offers. But note precisely what it showed: airborne organic pollutants, not heavy metals. Metals are elements and cannot be metabolized or conjugated the way benzene is. Extending this result to lead or mercury is exactly the kind of leap this post is trying to avoid.

What is the fair summary?

• For heavy metals, the popular claim that fiber pulls them out of the body runs well ahead of the human evidence, as the plant toxins post explains.
• Cruciferous vegetables have the best randomized human evidence for enhanced excretion, but for organic pollutants rather than metals.
• For organic pollutants such as PFAS, the picture is much stronger: viscous fibers like beta-glucan interrupt enterohepatic recirculation, with a randomized human trial behind it.
• Eating fruit, seaweed, oats, barley, flax, chia and cruciferous vegetables is well justified for many reasons. “Heavy metal detox” specifically is not one of the well-established ones.

And on the related claim that fiber binds and removes microplastics: there is no credible human evidence. It may turn out to be true. Right now it is speculation, and saying so is more useful than repeating it.

If you are genuinely concerned about metal exposure, the useful move is testing and source removal with a clinician, not a grocery list. The strongest protection whole plant foods offer is probably not pulling metals out of you; it is not putting them in. Many heavy metals bioaccumulate up the food chain.

Why do some populations have low fiber intake and still get almost no colon cancer?

This puzzle is worth the detour, because solving it reframes the entire carbohydrate argument.

Colorectal cancer is the third most common cause of cancer death worldwide. Rural sub-Saharan African populations were historically protected by very high fiber intake, but the modern diet in parts of the region is centered on refined cornmeal. Low fiber, and still remarkably low colon cancer rates. So what is protecting them?

First clue: their stools are more acidic than comparison populations despite similar fiber intake.17 Acidic is the protective direction.

Second clue: low colonic pH is produced by short-chain fatty acids, which bacteria make from fermenting fiber. But they were not eating more fiber, and their bacteria were producing short-chain fatty acids in abundance anyway.18 Something else was feeding the microbes.

The answer is resistant starch. Cooking the cornmeal as a porridge and eating it cooled generates resistant starch, which acts as a prebiotic exactly as fiber does. Only about 5 percent of starch may reach the colon compared with 100 percent of fiber, but people eat up to ten times more starch than fiber, so the absolute amount arriving can be substantial. Countries that eat the most starch have among the lowest colon cancer rates.19 A starch-based diet low in animal foods, even one built on a refined grain, can deliver some of what a high-fiber diet does, through resistant starch.

Can you just add resistant starch to a meat-heavy diet and be fine?

No, and this is the crucial control that the story would be dishonest without.

The African protection was twofold: high resistant starch and low animal product intake.2021 When researchers randomized people to a low-meat diet, a high-meat diet, or a high-meat diet with lots of added resistant starch, the high-meat groups produced roughly three times more presumptive carcinogens (N-nitroso compounds) and twice the ammonia in their stool. Adding resistant starch did not rescue them. About 90 percent of those compounds were formed in the bowel itself, which means buying nitrite-free uncured meat does not help; the chemistry happens inside you.22

You cannot out-supplement, or even out-fiber, a diet built on animal products. It has to be both.

What happens to your gut bacteria on a low-carb diet?

They lose their food supply. When obese subjects reduced dietary carbohydrate, fecal butyrate concentrations fell and butyrate-producing bacteria declined measurably.26 This is a direct, mechanistic cost of carbohydrate restriction that rarely makes it into the marketing. You are not just removing glucose from a meal; you are removing the raw material your colon cells’ primary fuel is made from. Researchers have a name for what a diet short on fermentable carbohydrate does to the gut ecosystem: starving your microbial self.90

Does cutting out fiber cure constipation?

A 2012 study is often shared as proof that it does, so it is worth knowing exactly what it tested. The researchers enrolled 63 adults with chronic idiopathic constipation, meaning constipation with no cause found after a normal colonoscopy. Patients were told to stop eating fiber entirely for two weeks and then to eat as much or as little as they found acceptable. At six months, 41 were on no fiber, 16 on reduced fiber, and 6 had returned to a high-fiber diet for religious or personal reasons. In the no-fiber group, bowel movements went from about once every 3.75 days to about once a day; the group that stayed on high fiber saw no change.179

That is a real result for a specific group, but look at the design. There was no separate control group; each patient acted as their own control. Patients chose their own fiber level after the first two weeks, so the people who felt better on less fiber were the ones who stayed on it. Everyone knew what they were changing, the study had 63 people, and it lasted six months, with nothing measured about long-term health.179 That makes it a case series, which sits near the bottom of the standard hierarchy of medical evidence, below cohort studies, randomized trials and systematic reviews.180

The honest reading is narrow: some people who already have severe, unexplained constipation may feel better with less bulk moving through a colon that is not moving it well. That is a question for a gastroenterologist about one condition. It is not evidence that fiber causes constipation in healthy people, and it says nothing about fiber’s effects on heart disease, diabetes or colon cancer covered throughout this post.

What does the better evidence say about diet and constipation?

It points the other way. A 2025 study followed 95,917 middle-aged and older adults across three large US cohorts, with diet measured repeatedly over 25 to 30 years, and excluded anyone who already had chronic constipation at the start. Comparing the top fifth of each dietary pattern with the bottom fifth, a plant-based diet was associated with a 20 percent lower risk of developing chronic constipation (95 percent confidence interval 14 to 27 percent) and a Mediterranean-style diet with a 16 percent lower risk (9 to 22 percent). An inflammatory dietary pattern was associated with 24 percent higher risk and a Western diet with 22 percent higher risk, while a low-carbohydrate diet showed no significant change. The authors report these associations were independent of total fiber and ultra-processed food intake, and that vegetables and nuts were associated with lower risk.181 It is observational, so it shows association rather than proof, but it is exactly the kind of long-term, large-scale evidence the 2012 study cannot provide.

A smaller cross-sectional survey of 1,431 adults in Luxembourg points the same direction: higher grain intake was associated with lower constipation scores, and higher intake of sugary products with higher scores.183

Treatment guidelines agree too. A review of the European, French, Spanish and Korean guidelines for chronic idiopathic constipation found that all of them agree fiber is effective. Soluble fiber such as psyllium is preferred over insoluble fiber because it causes fewer side effects, and three of the guidelines found dried plums more effective than psyllium for mild to moderate constipation. Drinking more fluid helps mainly in people who are dehydrated or eating a high-fiber diet, and osmotic laxatives are the first-line medication when diet is not enough.182

So the practical advice is the same as the rest of this post: build meals around a variety of whole plant foods, increase fiber gradually as described in the transition section below, and drink enough water. Constipation that persists despite that deserves a medical evaluation, not a switch to a fiber-free diet on the strength of one small, uncontrolled study.

Why the whole food beats isolated stand-ins

If you take one section from this post into an argument, take this one.

Is a bean better than the fiber and antioxidants in the bean?

Yes, and somebody actually tested it. Researchers fed adults with metabolic syndrome a typical Western-style meal three ways: with black beans, with a fiber-matched version using cellulose and psyllium, and with an antioxidant-matched version using grape seed extract. Only the black bean meal significantly lowered the after-meal insulin response, and the authors concluded the effect could only partly be explained by fiber content or antioxidant capacity.3 Two precision notes, because they matter. On blood antioxidant capacity, the grape-seed meal actually came out ahead of the beans. And triglycerides and the inflammatory marker IL-6 rose after every meal and were not differentially improved by the beans. The clean finding is about insulin.

The package is not just a metaphor. When researchers tried to reproduce the bean’s insulin effect with fiber and antioxidants from other sources, they could not fully do it.

Which part of the package gets thrown away first?

Fiber, almost every time, and it is worth pausing on why that is the tell.

Fiber is the component with no calories to speak of, no flavor, and no shelf-life benefit, so it is the first thing industrial processing removes and the thing animal foods never had. Refining a grain strips it. Juicing a fruit strips it. A steak never contained any. It is, by a wide margin, the single nutrient most reliably absent from everything the modern diet is built on.

And look at what leaves with it. Fiber is not one ingredient in the package, it is the part that governs how the rest of the package behaves: it sets the speed of glucose absorption, carries bile acids and excess cholesterol out of the body, provides the bulk that moves everything along, and feeds the only organ in your body that runs on a fuel made by something other than you. Remove it and the remaining carbohydrate behaves like an entirely different substance, because functionally it is one.

This is the clearest possible illustration of the argument of this whole post. The difference between a lentil and a lollipop is not the glucose. Both deliver glucose. The difference is everything the lentil brings with it, and fiber is the piece doing most of the work.

Do resistant starch supplements prevent cancer?

They did not in two randomized trials of people with hereditary colorectal cancer syndromes. In young people with familial adenomatous polyposis, a condition carrying virtually a 100 percent cancer risk if untreated, resistant starch did not reduce polyps; in Lynch syndrome carriers, it showed no benefit for colorectal cancer.2930 One honest complication: a 10-year follow-up of the Lynch syndrome trial later found resistant starch roughly halved non-colorectal cancers in Lynch syndrome carriers, mostly upper gastrointestinal (hazard ratio 0.54, 95 percent confidence interval 0.33 to 0.86).109 So the supplement was not inert. It just did not do the job it was given. One plausible explanation is that resistant starch may need to reach the far end of the colon where most tumors form, which likely requires accompanying fiber to push it along. The investigators concluded that resistant starch supplementation does not emulate the apparently protective effect of diets rich in dietary fiber.

Do fiber supplements improve the gut microbiome?

Isolated fiber supplements failed to increase bacterial richness. Feeding people whole grain barley, brown rice, or both for a month did increase bacterial diversity, alongside a drop in systemic inflammation.31 Bacterial richness matters: people with low gut bacterial richness carry more body fat, more insulin resistance, higher triglycerides and more inflammation, and gain more weight over time.3233

Better still, barley and brown rice together outperformed either alone. A whole food’s compositional complexity supports a wider range of bacterial species than any single purified fiber can.

Who first noticed that it was the foods, not the fiber?

Hugh Trowell, who spent thirty years practicing medicine in Africa and first linked fiber intake to chronic disease. He insisted the protection came from the high-fiber foods, not the fiber molecule itself; fiber intake was simply a marker for plant food intake.34 His view got flattened into the reductionist “fiber hypothesis,” and then the reductionist version got tested, underperformed, and was used to argue fiber does not matter. The reductionism people are arguing against today was a distortion of the original finding.

Why plants let you eat more and weigh less

Why is fat more fattening, gram for gram?

Arithmetic you cannot argue with. Fat carries 9 calories per gram. Carbohydrate and protein carry about 4.2 Fiber carries roughly 2, the National Academies putting the range at 1.5 to 2.5,2 and much of it is never absorbed at all. Water carries zero.

Whole animal foods generally come with fat, including saturated fat, and contain no fiber whatsoever. (One precision note: it is a myth that animal foods contain zero carbohydrate, because dairy contains lactose. But they contain no fiber, full stop.) Whole plant foods flip the ratio: they are built largely from water, fiber and starch rather than concentrated fat.

What is calorie density and why does it matter?

Calorie density is calories per gram of food, and it is probably the most useful lever for eating less without feeling deprived. Because water and fiber add weight and volume with few or no calories, water-rich high-fiber foods let you eat larger, more satisfying portions for fewer calories. In controlled feeding studies, lowering the energy density of the diet caused people to spontaneously eat less without being told to.37

In practice this means you can eat until comfortably full and still run a calorie deficit, because the food does the accounting for you. In a multicenter workplace trial, people eating a low-fat plant-based diet with no calorie counting lost significant weight.38

Has anyone tested this under controlled conditions?

Yes, in the tightest-controlled diet comparison anyone has run. Twenty adults lived on an NIH metabolic ward and ate two diets in random order, two weeks each, with unlimited food: a minimally processed plant-based low-fat diet, and a minimally processed animal-based ketogenic diet.

On the plant-based diet, people spontaneously ate 689 fewer calories a day, with no difference in reported hunger or enjoyment.98 Nobody was counting anything. The food did it.

The part usually left out of the retellings is the body composition. The scale numerically favored the ketogenic diet, though the difference was not statistically significant. But the composition of that loss ran the other way: participants lost body fat on the plant-based low-fat diet and did not significantly lose body fat on the ketogenic one, where what came off was largely fat-free mass, water and lean tissue included, despite the higher protein intake. Weight on a scale and fat on a body are not the same measurement, and a diet can win the first while losing the second.

This is developed at length with the rest of the trial evidence in the weight loss post.

The protein misunderstanding, in reverse

Do plants have all nine essential amino acids?

Yes, every one of them, in every whole plant food. This surprises people, so it is worth being precise about why: plants build amino acids from scratch. They take nitrogen and carbon and manufacture the entire set. Animals cannot do this for the essential nine, which is why they are called essential. When you eat an animal, you are eating amino acids that a plant made, routed through a second organism at a considerable loss.

That includes cows, chickens and fish. Every essential amino acid in every steak on earth was assembled by a plant or a nitrogen-fixing microbe first, out of nitrogen pulled from the air or the soil. Eating the plant is going to the factory. Eating the animal is buying second-hand, after the animal has burned most of the inventory just staying alive.

Notice the symmetry with the carbohydrate mistake. People treat carbohydrate foods as if carbohydrate is all they are, exactly the way they treat animal foods as if protein is all they are. Both reductions are wrong, and for the same reason.

Is it hard to get enough protein on a vegan diet?

For people eating a reasonable variety across legumes, grains, greens, vegetables, nuts, seeds and the rest, no. The old advice about carefully combining proteins at every meal is outdated, as the next answer explains.

Some specifics worth knowing: soy protein scores essentially at the top of the standard quality scales, comparable to meat, eggs and dairy.39 A cup of cooked edamame delivers roughly 18 grams of complete protein. Buckwheat, which is not a grass at all but a fruit seed related to rhubarb, has a notably balanced amino acid profile that is high in lysine, the amino acid that limits true cereal grains, with an amino acid score reported as comparable to reference requirements for both children and adults.40 Its digestibility runs somewhat below animal protein because of tannins and protease inhibitors, which is a fair caveat to state.

Do grains and legumes need to be combined at the same meal?

No. A 2019 review of protein in vegetarian diets found no evidence that plant proteins need to be combined within the same meal, and concluded that a reasonable variety of sources over the course of the day appears to be appropriate. The body does buffer amino acids between meals, though only to a limited extent, partly through its pool of free amino acids; lysine in particular has a relatively large pool that lets the body recycle it between meals. For anyone eating even a slightly varied diet, the review calls strict attention to combining complementary proteins simply over-precautious.176

The kernel of truth is real but narrow: individual plant proteins can run low in lysine, and very occasionally in methionine.176 Eat a variety across the day, which anyone eating variety does automatically, and it is a non-issue.

Is there anything to actually watch?

One thing, and it is worth saying rather than cheerleading past. A 2025 study of 193 long-term vegans in New Zealand, which adjusted for how well each protein is digested rather than just counting grams, found that about three in four met their total protein requirement: 78.8 percent of men and 73 percent of women. But only about half met their requirements for lysine and leucine, making those the amino acids most likely to fall short.177 Three in four is good, and half is a real gap. It is not automatic.

The fix is not a steak. The fix is legumes, daily. In that same study, legumes and pulses were the biggest contributors to both protein and lysine intake.177 Which is a convenient thing to discover in a post arguing that beans belong on every plate.

Is the protein quality scoring system fair to plants?

This argument lands squarely on the theme of this post. The prevailing scoring systems, PDCAAS and DIAAS, measure amino acid profile and digestibility. That is a real measurement. But they do not measure fiber, saturated fat, heme iron, or anything else arriving in the same bite. Katz, Jenkins, Gardner and colleagues argued in 2019 that a definition of protein quality built on biochemistry alone has drifted away from what actually happens to the people eating the food.104

A system that rates a hot dog above a lentil on protein quality is measuring something genuine. It is not measuring health. That is the same reductionism, running in the opposite direction, that turns the word “carbohydrate” into a slur.

The full treatment, including protein per calorie comparisons where spinach beats salmon and kale beats eggs, lives in the protein post. The short version for our purposes: adequacy is a solved problem for people eating variety, and the panic about it is manufactured.

Shouldn’t people be asking where vegans get their fiber?

Every whole food, plant or animal, contains protein. Beans have it, grains have it, greens have it, meat has it. As the protein post lays out, almost nobody in a wealthy country is short on protein, which is why “where do you get your protein?” has been a settled question for decades. The question that actually separates plants from animals is a different one: where do you get your fiber?

Animal foods contain none. The National Academies define dietary fiber as nondigestible carbohydrate that is intrinsic and intact in plants, so by that definition it comes from plants; the only other source is fungi such as mushrooms.2 Only about 5 percent of Americans meet the fiber recommendation,36 and in the WHO-commissioned synthesis described earlier, the highest fiber consumers had 15 to 30 percent lower all-cause and cardiovascular mortality.93 Nobody asks where you get your fiber, and that is exactly the nutrient most people are missing.

If both have protein, what actually makes animal foods different?

What rides along with the protein. Three things stand out.

• Calorie density. Fat carries 9 calories per gram, against 4 for carbohydrate and protein.2 Fat-rich animal foods pack more energy into every bite, with no fiber to slow you down or fill you up first. The weight loss post covers what calorie density means for appetite and body weight.
• Saturated fat. According to the National Academies, the majority of dietary saturated fat comes from animal products such as meat and dairy, and animal fats are solid at room temperature precisely because they are high in saturated fatty acids. The main plant exceptions are oils like coconut and palm kernel.2 Saturated fat has no known beneficial role in preventing chronic disease and is not required in the diet,2 and in the overfeeding trial described in the prediabetes section, it was the saturated fat that drove the extra liver and visceral fat.133 That fat is the fuel for the insulin resistance story told in the insulin section below. The fat post and the cholesterol post follow saturated fat from the plate to the arteries, and the seed oils post covers the trial evidence on replacing it.
• The amino acid profile itself. Protein is not identical across sources. As the protein post reviews, animal proteins run higher in the sulfur-containing amino acid methionine and the branched-chain amino acids, a pattern observational research has linked to cardiovascular mortality. That post also covers how protein escaping digestion is fermented in the colon into compounds such as hydrogen sulfide, a thread the colon cancer post picks up.

So the useful question at the dinner table is not where you get your protein. Everyone eating enough food gets protein. It is where you get your fiber, and how much saturated fat came along with your protein.

The 64x number, and what it actually means

Do plants really have more antioxidants than animal foods?

Yes, and the definitive dataset measured more than 3,100 foods, beverages, spices, herbs and supplements from around the world using a standardized assay. The results, in millimoles per 100 grams:1

• Plant-based foods: mean 11.57, median 0.88
• Animal-based foods: mean 0.18, median 0.10

Divide the means and you get the widely quoted figure: plant foods average about 64 times the antioxidant content of animal foods.

Is the 64x figure honest?

Mostly, with one caveat that this post is going to state rather than bury. The paper itself notes that the high plant mean is driven by a minority of extreme outliers, mainly spices, herbs and traditional plant medicines. Clove is off the charts. Cinnamon is not a serving size.

The more conservative comparison is the median, and by that measure plant foods carry roughly 9 times the antioxidant content of animal foods. The paper also offers a category-level figure that avoids the outlier problem entirely: fruits, nuts, chocolate and berries run 5 to 33 times higher than the mean of meat products.

Either number makes the point. Nine or sixty-four, animal foods are an antioxidant desert, and this matters more than it sounds, because those antioxidants ride inside the very lipoprotein particles that drive atherosclerosis. We traced that whole journey in the nutrition post.

Why does your body need antioxidants at all?

Because making energy is dirty work. Think of your cells as tiny engines burning fuel, and like any engine they throw off exhaust. That exhaust is a stream of reactive oxygen species, or free radicals, unstable molecules that nick your DNA, proteins and fats as they fly past. This is not a malfunction; it is the unavoidable smoke of running the engine at all. (For the engineers: the leak happens at the mitochondrial electron transport chain during ATP production.)

Here is the elegant part. The whole plant foods that supply your glucose fuel arrive carrying the antioxidant defenses against the oxidative cost of burning it. Fuel and fire extinguisher in the same package. Refined carbohydrates deliver the fuel and none of the extinguisher. Animal foods deliver neither the fuel nor the extinguisher, because the animal already spent its antioxidants staying alive.

What the National Academies actually recommend

How much carbohydrate do the National Academies recommend?

The Institute of Medicine set Acceptable Macronutrient Distribution Ranges as follows:2

• Carbohydrate: 45 to 65 percent of calories
• Fat: 20 to 35 percent
• Protein: 10 to 35 percent

Look at the floor of the carbohydrate range. Forty-five percent is higher than the ceiling of both the fat range and the protein range. The body that defines which nutrients are “essential” puts carbohydrate at the center of the plate.

Why is the carbohydrate recommendation 130 grams a day?

Because that is roughly what the brain burns. The reference intake was set explicitly on the basis of glucose use by the brain.2 The same report sets a fiber target that only plant foods can meet, and concludes that saturated fat and dietary cholesterol have no known beneficial role in preventing chronic disease and are not required in the diet, and that trans fat provides no known benefit to human health.

We take the “no essential carbohydrate” argument apart in full in the essential nutrients post, and there is a summary of it further down this page.

How whole carbohydrates got demonized

Why do people think carbs are bad?

Because the thing they are actually reacting to is real. Americans eat roughly three times as many calories from refined grains and added sugars as from all whole plant foods combined. That is a genuine catastrophe, and the low-carb movement noticed it before most of the nutrition establishment did.

There is even trial evidence that vindicates part of their instinct. The Cochrane review of 15 randomized trials with 56,675 participants found that reducing saturated fat cut cardiovascular events by 17 percent. The review’s authors write that replacing saturated fat with polyunsaturated fat appeared somewhat more protective than replacing it with carbohydrate, though the difference between those two replacement groups did not reach statistical significance. The review’s authors also say they could not separate refined from whole-food carbohydrate as a replacement, because too few of the underlying trials reported that detail.103 So this is real evidence that swapping saturated fat for unsaturated fat helps. It is not, on its own, evidence about refined carbohydrate specifically. The case against refined carbohydrate rests on other evidence: the sugar and glycemic-load trials covered elsewhere in this post, and the large cohort studies reviewed in the seed oils post, where refined carbohydrate performed worst as a replacement for saturated fat.

Refined carbohydrate really is a poor replacement for saturated fat, and on that point the low-carb movement and the evidence agree. The error is in the diagnosis. Having correctly identified that the standard diet is drowning in refined starch and sugar, the conclusion drawn was that carbohydrate itself is the problem, which is like noticing that people are getting sick from cigarettes and concluding that plants are dangerous. The category “carbohydrate” contains both the lollipop and the lentil. Blaming the lentil for what the lollipop did is a category error, and it has cost an enormous number of people some of the most protective foods available to them.

Do grains cause Alzheimer’s disease?

The “grain brain” genre is one of the more influential forms of carb demonization, and the epidemiology points firmly the other way.

Dementia rates vary enormously worldwide: lowest in Africa, India and South Asia, highest in Western Europe and North America. Comparisons of populations with shared ancestry but different environments point toward environmental causes, though the investigators left genetic explanations open. Age-standardized annual dementia incidence was 1.35 percent among Yoruba in Ibadan versus 3.24 percent among African Americans in Indianapolis, with Alzheimer’s incidence 1.15 versus 2.52 percent, roughly half.79 Japanese Americans in the United States have rates closer to Americans than to Japanese.80

As the Japanese diet shifted from rice-based toward meat, dementia prevalence climbed, and the dietary factor most strongly associated with the rise was increased animal fat consumption; a parallel analysis in China reached the same conclusion.81 In a small matched Adventist substudy of 272 people, meat eaters were about twice as likely to develop dementia as vegetarians, a borderline result (relative risk 2.18, p = 0.065) that rose to about three times when past meat eating was counted (relative risk 2.99). A larger, unmatched substudy of 2,984 people found no significant difference, so this is a signal rather than a settled finding.82 And the lowest validated Alzheimer’s rates in the world are in rural India, on low-meat, high-grain, high-bean, high-carbohydrate diets.83 A Neurology editorial summarized it as: pass the grain, spare the brain.84

The honest caveat: these are ecological and observational data, and in Japan the decline in rice and the rise in meat happened together, so they are entangled. The pattern points away from grains and toward animal fat. It is not proof.

What are the highest-impact changes people can make to their diets?

According to the Global Burden of Disease analysis across 195 countries, diet is the leading cause of death worldwide, killing more people than tobacco. The five highest-impact changes: eat less salt, more nuts, more non-starchy vegetables, more fruit, and more whole grains.85

Four of the five are whole plant foods. One of them is explicitly a carbohydrate staple.

And there is a detail in that analysis that rarely gets quoted. Of the 11 million diet-attributable deaths a year, roughly a fifth of all adult deaths, too little whole grain, fruit, nuts and vegetables accounted for more deaths than too much red and processed meat.85 The deficits outweighed the excesses. This is a modeling study and cannot rank diet patterns, but it is a useful corrective in both directions: the case for eating this way is as much about what you add as about what you remove, and the single biggest thing most people are failing to add is whole carbohydrate food.

The natural experiment: animal low-carb versus plant low-carb

Here is the cleanest test anyone has run on whether carbohydrate is the problem, and it is not a lab study. It is what happens when you hold the macronutrient label constant and change only where the food came from.

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

Enormously, and the direction reverses. In the Nurses’ Health Study and the Health Professionals Follow-up Study, a low-carbohydrate score built on animal foods was associated with 23 percent higher all-cause mortality, while a low-carbohydrate score built on plant foods was associated with 20 percent lower all-cause mortality and 23 percent lower cardiovascular mortality.94

The ARIC cohort plus a meta-analysis of 432,000 people found the same split from the other side: replacing carbohydrate with animal fat or protein raised mortality by 18 percent, and replacing it with plant fat or protein lowered mortality by 18 percent.95 In NHANES, a healthy low-carbohydrate diet was tied to 9 percent lower mortality and an unhealthy one to 7 percent higher.96 And in the CARDIA cohort, followed from young adulthood into middle age with imaging rather than death certificates, a higher low-carbohydrate score was associated with greater progression of coronary artery calcium, driven by the animal-based version of the pattern.97 One honest complication from the ARIC analysis: overall carbohydrate intake showed a U-shaped relationship with mortality, with the lowest risk at 50 to 55 percent of calories and higher risk above 70 percent as well as below 40.95 Those were observational populations eating mostly typical, not whole-food, carbohydrate, but it is a reason to treat very high-carbohydrate intakes as something supported for whole plant foods specifically, not carbohydrate in general. A 2022 meta-analysis of 18 cohort studies found a similar J-shape for type 2 diabetes: lowest risk around 50 percent of calories from carbohydrate and higher risk at 70 percent (hazard ratio 1.18). The upper end of that curve came from Asian cohorts, where the authors point to refined carbohydrates such as white rice and bread, while in Western cohorts risk did not rise between 37 and 60 percent.210

Same diet name. Opposite outcomes. The variable that flipped the sign was not how much carbohydrate people ate. It was whether the calories came from animals or plants. If carbohydrate restriction itself were the active ingredient, this could not happen.

Does the quality of a plant-based diet matter too?

Yes, and the same Harvard cohorts showed it from the plant side. Across 209,298 nurses and health professionals followed for more than 4.8 million person-years, researchers scored diets two ways. A healthful plant-based index rewarded whole grains, fruits, vegetables, nuts, legumes, oils, tea and coffee. An unhealthful one rewarded juices and sweetened drinks, refined grains, potatoes and fries, and sweets. Comparing the highest and lowest scorers, the healthful pattern was associated with 25 percent lower coronary heart disease, and the unhealthful pattern with 32 percent higher.106 Both scores rewarded plant foods. What separated them was whether those plant foods were whole or refined. These are observational data, and the study was funded by the National Institutes of Health.

How do low-carb and ketogenic diets rank against other patterns?

When the American Heart Association formally scored ten popular dietary patterns against its own heart-health guidance, DASH, Mediterranean, pescatarian, and lacto-ovo vegetarian eating landed in the top tier, while very-low-carbohydrate and ketogenic patterns ranked at the bottom. The stated reason for ranking those last is worth noting: even when followed exactly as designed, they restrict foods central to cardiovascular health, specifically legumes and whole grains, while permitting high saturated fat intake. Vegan eating scored 78 out of 100, in the second tier, which the AHA attributed to how restrictive the pattern can be to follow long-term and to the risk of B12 deficiency, alongside crediting its emphasis on fruits, vegetables, whole grains, legumes, and nuts.100

An umbrella review of 17 meta-analyses covering 68 randomized trials found only four conclusions about ketogenic diets supported by high-quality evidence, one of which is that they raise LDL cholesterol. The trials had a median duration of 13 weeks, and 76 percent of the meta-analyses were rated critically low on quality.101

So is low-carb always worse?

No, and this post is not going to overreach. In DIETFITS, a head-to-head randomized trial of a healthy low-fat diet against a healthy low-carbohydrate diet, both built on quality whole foods, there was no significant difference in weight loss at twelve months.99

That result is often waved around as a defeat for one side or the other. Read properly, it makes the same point as everything else on this page: quality, not the macronutrient label, is what moves the needle. Strip the junk out of either diet and both improve. The fight over grams of carbohydrate is a distraction from the fight that matters, which is whole food against refined food, and plants against animals.

The full evidence review across 48 screened studies, sorted by the hierarchy of evidence, is in what the best available evidence actually shows.

The staples of the healthiest diets on earth

What do the longest-living populations actually eat?

Whole carbohydrate foods, in quantity, as the foundation of the plate.

The Adventist Health Study-2 enrolled 96,469 Seventh-day Adventists, and the vegetarian dietary patterns within it were associated with lower all-cause mortality.44 Earlier Adventist work found vegetarian men lived about 9.5 years longer and women about 6.1 years longer than the surrounding California population.45 Adventists eating legumes more than twice a week had a 47 percent lower risk of colon cancer (relative risk 0.53, 95 percent confidence interval 0.33 to 0.86).146

Across the Blue Zones, Loma Linda, Okinawa, Nicoya, Sardinia and Ikaria, the common thread, as described by the researchers who named them, is a plant slant: beans are the cornerstone of most centenarian diets, and meat is eaten on average only about five times a month.178 Across five cohorts of adults over 70 in Japan, Sweden, Greece and Australia, a study conducted under the auspices of the World Health Organization found that every 20-gram daily increase in legume intake was associated with a 7 to 8 percent lower mortality hazard.46

These are observational data, and they deserve the appropriate caution, which is exactly why we wrote the Adventist Health Studies post, what the longest-living people eat, the Blue Zones debunked-or-not post, and the Saul Newman fact-check. The critiques are worth taking seriously. The convergence across independent long-lived populations on beans and grains survives them.

Whole grains: one piece of the puzzle, and one of the best studied

The longest-lived populations build their plates on grains and beans, so it is worth asking how those two staples perform in the research. No single food covers everything a person needs, and grains are no exception. They are one piece of a whole-food pattern built on legumes, greens, grains, fruits, vegetables, mushrooms, nuts, seeds, seaweeds, herbs and spices. What grains bring to that pattern is distinctive: cereal fiber, which several of the studies below single out, and protein whose amino acid profile complements legumes, as the protein section explains. The question here is how well the research supports whole grains as part of that pattern. The answer is: very well, with a sharp line between whole and refined grains running through all of it.

How do whole grains compare with other foods?

The most useful comparison put 12 food groups side by side in one meta-analysis of prospective studies. Each daily serving of whole grains was associated with 8 percent lower all-cause mortality, alongside vegetables (4 percent), fruit (6 percent), fish (7 percent) and nuts (24 percent per serving). In terms of realistic servings, whole grains showed the largest reduction: about three servings a day, or 90 grams, was associated with 21 percent lower risk than none. Whole grains were also the only food group whose evidence was rated high quality.211

Then the authors made the point that matters most. Combining the optimal intakes of whole grains, vegetables, fruit, nuts, legumes and fish was estimated to lower the risk of death by 56 percent compared with eating none of them.211 That figure is a calculation that multiplies the separate associations together, not a measured result, but it captures the big picture: no single food gets you there, the pattern does. A 2025 umbrella review of 41 systematic reviews and meta-analyses found the same shape. Higher intakes of nuts, whole grains, fruits, vegetables and fish were associated with lower mortality, legumes looked favorable when comparing high with low intakes, and refined grains showed no clear association.212 Those comparisons ask what happens as people eat more of each food group, not whether fish beat the plant foods they might replace. What comes along with fish is covered in our fish post, and whether anyone needs long-chain omega-3s from food at all is the subject of our DHA post.

Do whole grains lower the risk of heart disease and early death?

In a separate pooled analysis of 45 prospective studies, each additional 90 grams of whole grains a day, about three servings, was associated with 19 percent lower coronary heart disease, 22 percent lower cardiovascular disease and 17 percent lower all-cause mortality. Refined grains, white rice and total grains showed little evidence of an association.213 An umbrella review of fiber meta-analyses rated the links between higher fiber intake and lower cardiovascular and all-cause mortality as convincing, its highest grade, and those were the only convincing findings built on prospective studies.214

Do whole grains protect against type 2 diabetes?

In three large Harvard cohorts totaling about 195,000 adults, people eating the most whole grains had a 29 percent lower rate of type 2 diabetes than those eating the least. The type of food mattered: whole-grain breakfast cereal and dark bread were associated with lower risk while popcorn was not (hazard ratio 1.08), and the benefit for cereal and dark bread leveled off at about half a serving a day.215 The fiber in grains appears to carry particular weight. Comparing high with low intakes, cereal fiber was associated with 23 percent lower diabetes risk, against 6 percent for fruit fiber.216

What do whole grains and cereal fiber do for the colon?

Each 10 grams a day of total fiber was associated with 10 percent lower colorectal cancer risk. Cereal fiber showed the same 10 percent, while fruit and vegetable fiber did not reach significance, and three extra servings of whole grains a day was associated with 17 percent lower risk.217 Diverticular disease showed a strikingly consistent signal, with no measurable disagreement between studies: risk fell 26 percent per 10 grams of fiber a day, reaching 41 percent lower at 30 grams compared with 7.5 grams.218 The association holds even after diagnosis. Among people with colorectal cancer, each additional 5 grams of cereal fiber a day was associated with 33 percent lower colorectal cancer mortality, and each additional 20 grams of whole grains with 28 percent lower.219

Does it matter what whole grains replace?

Swapping is where whole grains do some of their best work. In a meta-analysis of substitution analyses across 24 cohorts, replacing processed meat with whole grains was associated with 36 percent lower cardiovascular disease risk, a larger association than swapping it for nuts or legumes. Replacing red meat with whole grains was associated with 10 percent lower type 2 diabetes risk. Both findings were rated moderate certainty.220 Refined grains point the other way. In a separate PURE analysis of cereal grains across 21 countries, eating 350 grams or more of refined grains a day, about seven servings, was associated with 27 percent higher mortality than eating under 50 grams.225

What do randomized trials on whole grains show?

Here the evidence is thinner, and it is worth saying so plainly. A Cochrane review found nine randomized trials of whole-grain diets, none longer than 16 weeks, and none measured heart attacks, strokes or deaths. It found no difference in total cholesterol, rated that evidence low quality, and noted that some trials were funded by companies with commercial interests in cereals.222 Whole-grain trials have not shown weight loss either: across 2,060 participants, body weight did not change, though body fat percentage fell slightly.223

The trials that do show clear effects are telling. In adults with diabetes, increasing fiber lowered HbA1c, fasting glucose, LDL cholesterol and body weight across 42 trials, and in cohorts of people with diabetes, eating 35 grams of fiber a day rather than 19 was associated with 14 fewer deaths per 1,000.221 In a Danish crossover trial in adults at risk of metabolic syndrome, eight weeks of a whole-grain diet, compared with refined grains, lowered body weight and the inflammatory markers IL-6 and C-reactive protein, without changing glucose control or producing major changes in the gut microbiome.224 Oats lowering LDL cholesterol, covered in the fiber section, is the other well-established trial result.

How strong is the evidence for whole grains, honestly?

Strong in consistency, modest in size, and mostly observational. That PURE cereal-grain analysis is the main exception to the consistency: across 21 countries, it found no association between whole grains and mortality or cardiovascular disease in either direction.225 A burden-of-proof analysis, which deliberately takes the most conservative reading of the data, still found whole grains associated with lower risk of type 2 diabetes, colorectal cancer, stroke and ischemic heart disease. It rated all four links two stars out of five, the lowest rating that still indicates lower risk under that conservative reading.226 Observational studies also cannot fully separate whole-grain eaters from their other habits.

So where do grains fit in a healthy diet?

As one central piece, not the whole picture. The pattern that keeps showing up, from the 12-food-group analysis to the umbrella reviews to the substitution studies, is whole grains working alongside legumes, nuts, fruits, vegetables and the rest of a whole-food diet, and doing the most good when they replace refined grains and processed or red meat. Refined grains do not earn the same credit. Eaten whole and in variety, as part of that pattern, grains are among the best-supported foods in nutrition research. Which grains to try, and the angiography study on whole grains and arteries, are in the category tour further down.

Legumes: another piece of the puzzle, with a different kind of evidence

Beans, lentils, chickpeas, peas and soy foods are the other great carbohydrate staple of a whole-food plant diet, and they supply the lysine that complements grains, as the protein section explains. Their evidence looks different from the grain evidence, though. For whole grains, the large cohort studies are strongest and the trials thinner. For legumes it is closer to the reverse: randomized trials show consistent improvements in cholesterol and blood sugar, while long-term cohort results are smaller and more mixed. That is one more reason to judge foods as parts of a pattern rather than one at a time.

Do beans and lentils lower cholesterol and blood sugar?

Yes, modestly and consistently. A 2024 meta-analysis of 24 randomized trials, each lasting at least six weeks and providing an average of 86 grams of legumes a day to 1,938 people, found moderate-certainty reductions in total cholesterol (0.23 mmol/L), LDL cholesterol (0.16 mmol/L) and fasting blood glucose (0.18 mmol/L).231 An independent Nordic review restricted to whole legumes found similar reductions in LDL cholesterol (0.19 mmol/L) and fasting glucose, with high variation between trials.229 Both line up with the pulse and LDL meta-analysis in the fiber section.157 In 121 people with type 2 diabetes, a three-month trial that encouraged at least a cup of legumes a day as part of a low-glycemic diet lowered HbA1c by 0.5 percentage points, against 0.3 on a high wheat fiber diet, and lowered systolic blood pressure 4.5 mm Hg more. That trial was funded in part by a pulse growers’ association.232

Do legumes help with weight and blood pressure?

A little. Across 21 trials with 940 participants, diets including about one serving of pulses a day (a median of 132 grams) lowered body weight by 0.34 kilograms over a median of six weeks.234 Across eight trials with 554 participants, pulses lowered systolic blood pressure by 2.25 mm Hg, though none of those trials controlled participants’ meals in a metabolic ward.233 Pulses also blunt the rise in blood sugar after meals in people with and without type 2 diabetes.235

Do legumes help you live longer?

The cohort evidence says yes, by a modest amount. Pooling 32 cohorts with more than 1.1 million people, higher legume intake was associated with 6 percent lower all-cause mortality, and in the dose-response analysis each additional 50 grams a day was associated with 6 percent lower risk, at low to moderate certainty. Legumes were not significantly associated with cardiovascular, coronary or cancer mortality.227 That is smaller than the older estimate described earlier, which came from five cohorts of adults over 70. The Nordic guideline review rated the link with lower all-cause mortality as moderate evidence.230

Do legumes protect the heart and prevent diabetes?

Here the evidence is honestly mixed. A GRADE-rated review of cohort studies found pulses associated with 8 percent lower cardiovascular disease, 10 percent lower coronary heart disease, 9 percent lower hypertension and 13 percent lower obesity, but rated the certainty low for cardiovascular disease and very low for the rest, and found no association with stroke or diabetes.228 The independent Nordic review found associations with cardiovascular disease, coronary heart disease, stroke and type 2 diabetes that were suggestive of null, in populations with generally low legume consumption, and called the evidence limited.229 The Nordic guideline review points to that low consumption as the likely reason the cardiovascular picture is unclear.230

Does it matter what legumes replace?

Yes, and this is one of the stronger findings. Replacing processed meat with legumes was associated with 23 percent lower cardiovascular disease risk across eight cohorts, rated moderate certainty.220 Legumes also fit the food-group pattern described in the whole grains section: in the 2025 umbrella review, they were associated with lower mortality when comparing high with low intakes.212

Do legumes and soy affect cancer risk?

For colorectal cancer, each additional 100 grams of legumes a day was associated with 21 percent lower risk in a 2022 meta-analysis, but the protective association for high versus low intake appeared only in Asian studies, not American or European ones, and the authors called the evidence currently weak.241 The Adventist colon cancer finding described earlier is one of the stronger single studies in that literature.

Soy leans protective, with a regional caveat. In a meta-analysis of prospective studies, soy was associated with lower breast cancer risk in Asian populations (relative risk 0.76) but not Western ones (0.97).238 In the China Kadoorie Biobank, which followed more than 300,000 Chinese women for about 10 years, the highest soy intakes carried the same breast cancer risk as the lowest (relative risk 1.00), while the accompanying dose-response meta-analysis found 3 percent lower risk per 10 milligrams of isoflavones a day.239 Higher total soy food intake was associated with 29 percent lower prostate cancer risk across 16 studies.237 A 2026 umbrella review concluded that dietary soy foods are more beneficial than harmful, especially for cancer and cardiometabolic disease, but that evidence on fermented soy products is inconclusive, with some studies suggesting possible harm.240 Soy protein also lowers LDL cholesterol by about 3 to 4 percent.236

Do beans really cause gas?

Some, but less than their reputation. In three feeding studies adding half a cup of beans a day, fewer than half of participants reported more flatulence from pinto or baked beans in the first week, only 19 percent with black-eyed peas, and 3 to 11 percent reported more even on control diets without beans.242 The transition section covers how to ease in.

So where do legumes fit in a healthy diet?

As one of the workhorses of the pattern, not a miracle food. Their clearest benefits show up in trials on cholesterol and blood sugar, and when they replace processed meat, while the long-term cohort signal is real but modest. Most of this evidence is for whole or minimally processed legumes, such as cooked or canned beans, lentils, chickpeas, peas, tofu, tempeh and edamame; the soy cholesterol finding is the main exception, since it tested soy protein. Eaten alongside whole grains, greens, vegetables, fruits, nuts and seeds, they do their part. Which legumes to buy, and the studies on their antioxidants, bones and gut bacteria, are in the category tour further down.

It is not the carbohydrates. It is the fat inside your cells.

This is the heart of the post, and it starts with some history.

What did doctors feed people with diabetes before insulin?

The history is messier than either side of the carbohydrate debate usually admits. Egyptian papyri from around 1500 BC describe patients with excessive thirst and urination treated with plant extracts, although the translator of the Ebers papyrus judged its description of diabetes unsatisfactory and probably wrong.161 In the second century AD, Aretaeus of Cappadocia gave the first accurate description of the disease and, by one account, coined the term diabetes. Among the foods he thought could help were slimy vegetables, rice and white starch.161 Another popular history instead credits the name to Apollonius of Memphis around 250 BC.162

By the 1800s the usual approach had swung the other way, toward low-carbohydrate diets,162 and for several generations patients were told, in the words of the French physician Apollinaire Bouchardat, to eat as little as possible.161 Before insulin, type 1 diabetes was uniformly fatal, killing either through ketoacidosis or through slow wasting.167

Then, at the beginning of the 20th century, the German diabetologist Carl von Noorden tried something that seemed to break the rules: an oat cure.170 As a 1907 report described it, the recipe was 250 grams of oatmeal cooked into a soup with water and 300 grams of butter, plus added protein, fed every two or three hours.168 Keep the date in mind. This was well over a century ago, about fifteen years before insulin, when people with severe diabetes were passing large amounts of sugar in their urine and wasting away. The butter was there for calories: the report counts about 2,500 calories from fat, which it describes as adequate for an 80-kilogram patient even while losing 50 to 100 grams of sugar a day in the urine.168 So this was not a low-fat carbohydrate cure. It was a way to keep starving patients fed with a carbohydrate their bodies seemed to tolerate.

Some patients did remarkably well. A 1909 report describes a 37-year-old man whose urine carried 108 grams of sugar a day. A standard diabetic diet with two slices of bread brought it only to 98 grams, but within ten days of adding oatmeal the way von Noorden advised, the sugar fell to a small fraction of that. The same report says the oat diet reduced the acetone bodies that signal an impending diabetic coma, while warning that for some patients even oatmeal was harmful.169

The skeptics were on record too. The 1907 report summarized a physician named Lipetz who tested the cure: in one severe case, sugar in the urine went up on oats, and several patients vomited or refused the diet. But in a moderately severe case whose urine became nearly sugar-free on oats, the stool showed a marked rise in bacteria, and Lipetz concluded that the good results probably came from the oat carbohydrate being fermented by gut bacteria rather than absorbed.168 In 1907 that was a guess. Today it is a research field. A small 2026 randomized trial tested a modernized version of the cure, without the butter: people with metabolic syndrome ate three meals a day of 100 grams of rolled oats boiled in water for two days, at reduced calories, against a control diet matched for macronutrients. LDL cholesterol fell relative to the control group and tended to stay below baseline for six weeks after participants returned to their usual diet, and the researchers traced part of the effect to phenolic compounds made by gut bacteria from the oats; the fiber section has the numbers.170 With 17 people per group, that is a promising signal rather than a settled result.

In the years just before insulin, the standard was Frederick Allen’s strict fasting diet. Leonard Thompson, a 14-year-old who received insulin in Toronto in January 1922, had been on that diet and was in critical condition; after an improved second injection, his blood glucose fell from 520 to 120 mg/dL within 24 hours.161

So the history cuts both ways. Restricting carbohydrate is not a new idea that history overlooked; it was tried for a century, often as near-starvation, because without insulin there was little else to offer. And the oat cure, stripped of its hype and its butter, showed that some patients handled a large load of oat carbohydrate far better than the theory of the day predicted. What changed the picture was insulin, and then, in the 1930s, the finding described next: that a carbohydrate-rich diet improved the body’s handling of sugar rather than worsening it.

What did Himsworth discover in the 1930s?

Sir Harold Himsworth is the physician who separated diabetes into insulin-sensitive and insulin-insensitive forms, the distinction we now roughly call type 1 and type 2.52 Before that, he spent years asking what diet does to the body’s handling of sugar.

In a 1934 paper, he summarized his earlier findings in healthy men: diets that worsened glucose tolerance also slowed how fast insulin could bring blood sugar down. He then reproduced the effect in rabbits. A diet rich in fat decreased sugar tolerance and blunted insulin’s action on blood sugar, while a diet rich in carbohydrate improved both.51 He also noted it was already established that starvation and high-fat diets impair sugar tolerance, while carbohydrate-rich diets improve it.

He was not the first to see it. In 1927, a Dallas physician, J. Shirley Sweeney, gave young, healthy male medical students diets of only protein, only fat, rich carbohydrate, or no food at all for two days before a glucose tolerance test. Those who had starved and those on the fat diet, made up of olive oil, butter, mayonnaise made with egg yolk and cream, showed the most abnormal responses.252

In Himsworth’s experiments, carbohydrate tolerance was not being destroyed by carbohydrate. It was being impaired by fat. That was nearly a century ago, and everything since has filled in the mechanism.

How does fat actually block insulin?

Picture a muscle cell as a house with a locked door that glucose comes in through. It helps to be clear about who does what.

• Insulin is the doorbell, not the key. It rings from outside the house.
• Ringing the doorbell sets off a message passed hand to hand down a hallway inside: the receptor takes the signal, hands it to a relay called IRS-1, which hands it further along the chain.
• The last person in that chain has one job: go fetch the door. The door is a glucose transporter called GLUT4, and it sits parked inside the cell until the message arrives to move it into the wall.

Now load that cell with fat. Here is the part that gets stated wrong constantly: the fat does not sit on the doorbell. Insulin still rings, and the receptor still hears it perfectly. What the fat does is jam the hallway. It throws off byproducts that tackle one of the messengers partway down the chain, so the message never reaches the last person, nobody fetches the door, and glucose is left standing outside with the doorbell ringing and nothing opening.

That distinction matters, and not just for accuracy. “Fat clogs the insulin receptor” is the phrasing low-carb advocates love to attack, because it is genuinely wrong. The receptor is fine. The wiring behind it is what fat breaks. Say it that way and the point is airtight.

Here is the same thing in the language of the biology:

Skeletal muscle handles 80 to 90 percent of insulin-stimulated glucose disposal, so muscle is where insulin resistance mostly happens. Excess calories, particularly from saturated fat, load muscle cells with intramyocellular lipid, and that fat generates bioactive intermediates: diacylglycerol, ceramide, long-chain fatty acyl-CoA.

Those intermediates activate protein kinase C isoforms, which blunt the activity of IRS-1, a key relay in the insulin signaling chain. The signal degrades, GLUT4 transporters fail to move to the cell surface, and glucose cannot get in.50 The doorbell works; the fat jammed the hallway between it and the door.

The same process fills the liver, and visceral fat around the organs behaves like an inflamed gland pumping out signals that make everything worse. This is the machinery behind the “personal fat threshold” concept discussed in the weight loss post.

What is the Randle cycle, and does it explain insulin resistance?

In 1963, Philip Randle and colleagues at the University of Cambridge proposed what they called the glucose fatty-acid cycle, now usually called the Randle cycle. Muscle can burn either fat or glucose, and they argued that a high level of fatty acids in the blood slows the burning of glucose. They linked that high level to impaired insulin sensitivity across a list of conditions that included diabetes, starvation and carbohydrate deprivation.253

The fuel competition is real in people. When researchers kept blood fatty acids high during an insulin infusion in 10 healthy volunteers, the activity of pyruvate dehydrogenase, the enzyme at the gateway to burning glucose, fell by about half, and glucose’s share of the leg muscles’ oxygen use dropped from 76 to 53 percent. The biggest effect, though, was on glucose storage in the muscle, which fell by about 70 percent.254

Randle’s model also made a testable prediction: glucose should back up inside the muscle cell, so glucose-6-phosphate, the first form glucose takes once inside, should build up.255 When a Yale team measured it directly with magnetic resonance spectroscopy in nine healthy people, it fell instead, starting about an hour and a half into a lipid infusion and before muscle glycogen synthesis dropped by about half. The authors concluded that, in contrast to the originally postulated mechanism, fatty acids cause insulin resistance in humans by first blocking glucose transport into the cell.256 A follow-up found that glucose inside the muscle cells was lower too, pointing to the transport step,257 and a Vienna study found the same early block even at ordinary fasting levels of fatty acids.258 After six hours of infusion, as insulin-stimulated glucose uptake fell 43 percent, diacylglycerol and protein kinase C activity in the muscle rose about fourfold: the hallway-jamming machinery described above.259

So the Randle cycle is real physiology, a fuel switch, but not the main cause of insulin resistance. Two diabetes researchers put it plainly: the Randle cycle is “a valid physiological principle,” but “it may not explain insulin resistance in skeletal muscle.”260

One detail matters before anyone cites these experiments in an argument about saturated fat. The fat infused in them was not saturated fat: several used an emulsion of safflower and soybean oil, and the others used similar intravenous fat emulsions.259254258 What they show is what happens when the blood is flooded with fatty acids. The reverse works too: lowering blood fatty acids overnight with a drug more than doubled insulin-stimulated glucose uptake in obese people with and without diabetes.261 Which fats, eaten in which amounts, put fat into muscle and liver over time is a question for feeding trials, which the calories, fat and sugar section covers.

Why does insulin resistance in muscle matter for the heart?

Because it changes where carbohydrate goes. A Yale review argues that in lean, young, insulin-resistant people, muscle that cannot take up glucose and store it as glycogen sends the energy from carbohydrate to the liver instead, where it is turned into fat, promoting atherogenic dyslipidemia, the plaque-promoting pattern of blood fats, along with fatty liver.262 That is the chain in one line: fat jams the muscle, carbohydrate is rerouted, and the liver makes fat from it. The fix is to unjam the muscle, which is what this post keeps returning to: calories in balance, less saturated fat, and carbohydrate from whole foods.

Where does the fat inside muscle and liver cells come from?

Your body has a designated place to store surplus energy: the fat tissue under your skin. The trouble starts when that storage runs out of room. A 2025 review lays out the sequence. When energy intake is chronically greater than energy expenditure, the capacity of subcutaneous fat to store fat can be overwhelmed, and the excess spills over into places it does not belong, as ectopic fat in the liver, the pancreas and skeletal muscle. The review describes growing evidence that this ectopic fat directly causes insulin resistance and damages the insulin-producing cells of the pancreas, with a fat byproduct called diacylglycerol as the common thread across all three tissues.173

The same review describes the reverse. Creating a negative energy balance, through a very-low-calorie diet or bariatric surgery, drains ectopic fat from the liver and pancreas, lowers diacylglycerol inside the cells, and can restore insulin sensitivity and put type 2 diabetes into remission.173 That is the same mechanism behind the plant-based trial described below, which reduced fat inside liver and muscle cells.49 And the fat stored in muscle is not inert. In animal models of high-fat-induced obesity, it turns over rapidly and releases diacylglycerol and ceramides, the byproducts that impair insulin signaling.172

If fat inside muscle causes insulin resistance, why are athletes so insulin-sensitive?

This is called the athlete’s paradox, and it sharpens the picture rather than breaking it. Both trained endurance athletes and people with type 2 diabetes carry more fat inside their muscle cells than ordinary healthy people. The difference is what happens to it. In athletes, that fat sits in lipid droplets and is burned as fuel during exercise, in muscle with a high capacity to use it, and it does not harm insulin sensitivity.171 The problem is not stored fat as such. It is the harmful byproducts that build up when fat arrives faster than it can be burned.

The type of fat matters too, at least in the lab. When human muscle cells were treated with palmitate, a saturated fat, they produced more diacylglycerol and ceramide and became insulin resistant, while oleate, a monounsaturated fat, was mostly stored as harmless triglyceride.171 That is a cell-culture finding, not a human trial, but it matches the overfeeding trial described in the prediabetes section, where saturated fat drove far more liver and visceral fat than polyunsaturated fat at the same weight gain.133

Can a plant-based diet remove fat from inside muscle cells?

Yes, and it was measured directly. In a randomized trial, a low-fat vegan diet reduced body weight, improved insulin sensitivity, and cut fat inside liver cells by 34 percent and inside muscle cells by 10 percent, measured by magnetic resonance spectroscopy in a subset of 44 participants.49 The diet that people are told will spike their blood sugar drained the fat that was causing the spike.

Do low-carb diets fix insulin resistance or hide it?

They lower the number on the meter by removing the substrate. If you stop eating carbohydrate, you will not see a large glucose excursion after meals, because there is nothing to excurse. That is not the same as restoring the cell’s ability to take up glucose.

A glucose tolerance test exposes the difference. In a crossover study, nine healthy young men ate either a normal diet or a low-carbohydrate, high-fat diet (about 69 percent of calories from fat) for three days each, with similar total calories. After the low-carbohydrate days, their blood sugar during the test was significantly higher, and the early burst of insulin their bodies released was smaller.159 If three days is enough to make healthy people handle a glucose load worse, cutting carbohydrate is not repairing the underlying problem; it is avoiding the test. Meanwhile the elimination costs you the butyrate supply discussed above.26

Longer studies show the same switch. Six healthy men ate three diets for 11 days each, with the same calories and protein but 85, 44 or 2 percent of calories from carbohydrate. After the 2 percent diet, insulin-stimulated glucose burning was 1.7 micromoles per kilogram per minute, against 13.4 and 19.0 after the other two, and the liver kept releasing more glucose despite insulin; total glucose uptake showed no clear change.263 That is the Randle switch described above: a body running on fat burns little glucose when a load arrives. It cuts two ways. Cutting carbohydrate leaves the body worse at handling it, which is the opposite of a repair. And a glucose test taken right after a stretch of very-low-carbohydrate eating makes the carbohydrate itself look worse than it is, a problem a 1927 study already flagged when it recommended a standardized diet before the test.252

A metabolically healthy person eats a bowl of oatmeal and does not get a diabetic glycemic curve. That is the actual test.

People have reversed type 2 diabetes eating more carbohydrate

Can you reverse type 2 diabetes on a high-carb diet?

The Ma-Pi 2 studies are the clearest demonstration available, and they are worth walking through slowly.

The diet is strictly plant-based and centered on whole grains: brown rice, barley and millet, with vegetables, beans, sesame seeds and green tea. It runs roughly 70 to 73 percent of calories from carbohydrate.

In a six-month intervention in 16 adults with poorly controlled type 2 diabetes, everyone entering with glycated hemoglobin above 8.5 percent, the results were these: every participant was able to eliminate insulin treatment entirely, and only 25 percent remained on any medication at all, a single oral drug. Average HbA1 fell 54.5 percent. Total cholesterol fell 16.4 percent, LDL cholesterol 22.7 percent, and triglycerides 37.0 percent. The meals were prepared and served daily by macrobiotic specialists, which matters when judging how easily this translates to home cooking.5658

This was a small, uncontrolled study, and that caveat belongs here rather than in a footnote. What it establishes is not a precise effect size. It is that the direction is not what the carbohydrate story predicts.

Does diabetes risk fall step by step as animal foods drop?

It does, and this is worth seeing before the treatment trials, because it shows the same direction in prevention. In the Adventist cohort, new cases of type 2 diabetes over two years fell steadily across diet groups: 2.12 percent of non-vegetarians, 1.29 percent of pescatarians, 1.08 percent of lacto-ovo vegetarians, and 0.54 percent of vegans.102

The vegans, who by definition ate the most carbohydrate, had roughly a quarter of the diabetes incidence of the meat eaters. A graded relationship like that is one of the things epidemiologists look for when asking whether an association might be causal.

Has this been tested in a randomized trial?

Yes. The MADIAB trial randomized type 2 diabetics to the Ma-Pi 2 diet or to a recommended diabetes-association-style diet, which was itself a genuinely healthy comparison diet with vegetables, beans, whole wheat and fruit. The trial was isocaloric, meaning both arms ate the same number of calories.

The macrobiotic diet won on every measure of glycemic control, and did so on fewer drugs: five of seven participants taking oral hypoglycemics had to stop them to avoid hypoglycemia (no low-sugar episodes actually occurred), versus one in the control group. Fasting glucose fell from around 129 to around 95 in three weeks, and LDL cholesterol fell significantly more than on the control diet.57

Now the part worth sitting with. They added roughly 100 grams of carbohydrate per day, largely as grains, to the diets of type 2 diabetics, on a diet that was about 73 percent carbohydrate. Blood sugars did not skyrocket. They came down, within three weeks.

In a six-month follow-up of the MADIAB trial, the Ma-Pi group reached a median HbA1c under 5.7 percent, which is the non-diabetic range, and had a significantly greater reduction in HbA1c than the control diet: a median drop of 11.3 percent versus 5.9 percent. Forty of the original 51 participants took part in the follow-up.111

What about the oatmeal studies?

Short oatmeal interventions, typically two days of oatmeal at roughly 700 to 1,000 calories with no added fat and no animal protein, have repeatedly reduced insulin requirements. In an uncontrolled hospital pilot study of 14 patients, mean blood glucose fell from 158 to 118 mg/dL and insulin dose dropped 42.5 percent, from 145 to 83 units a day. The striking part is what happened next: four weeks later, in the 9 patients followed up after going home to a normal diet, insulin dose was still down at 83 units a day, with comparable glycemic control.596062

The most dramatic single case: a 67-year-old woman with type 2 diabetes admitted with pneumonia, who developed sepsis and, within 48 hours, severe insulin resistance requiring more than 200 units of insulin per day. She was given a 700-calorie plant-based oatmeal preparation through a nasogastric tube. Both her mean blood glucose and her total daily insulin requirement fell significantly during and after the intervention.61

Two days of low-calorie oatmeal with no added fat appears to reduce insulin resistance enough to coast on for weeks. The calorie restriction is part of that effect, so these are not clean tests of carbohydrate alone, and they are no substitute for a sustained whole-food plant-based diet. But as a demonstration that a large load of oat carbohydrate did not make insulin resistance worse, they are hard to beat.

Is diabetes reversal a new idea?

No, and the history is worth knowing. In his 1955 Banting Lecture, the Australian physician Ewen Downie described how Newburgh and Conn had shown in 1939 that, in many obese patients with diabetes, bringing body weight down to normal made the symptoms disappear and returned the glucose tolerance curve to normal, in some cases staying normal for years.158

Downie also described his own patients whose diabetes went into remission as they lost weight. Notice what they were eating: a mild carbohydrate allowance of 150 to 200 grams a day, raised to about 300 grams for the week before each glucose tolerance test.158 These remissions happened on plenty of carbohydrate.

He was also candid that several of those patients relapsed later, some with no change in weight, and in 1955 he still called reversibility of diabetes “a clinical dream.”158 That honesty is useful. Remission is real, but it depends on the underlying conditions staying changed, and some people relapse anyway, which is why follow-up with a clinician matters.

So can you reverse type 2 diabetes on a whole-food plant-based diet?

For many people, yes, in the sense doctors mean by remission: blood sugar back in the non-diabetic range, often on fewer medications or none. Pull together what this section has already shown:

• In the Ma-Pi 2 intervention, every participant came off insulin while eating a high-carbohydrate diet built on whole grains and vegetables.56
• In the six-month follow-up to the randomized MADIAB trial, the Ma-Pi group reached a median HbA1c in the non-diabetic range.111
• A low-fat vegan diet reduced the fat inside liver and muscle cells, the very fat that jams the insulin signal.49
• Prediabetes reverts to normal glucose at roughly the same rate it progresses to diabetes, which means the direction is not fixed.129

The mechanism is the one this post keeps returning to. Type 2 diabetes is driven by insulin resistance, and insulin resistance is driven by fat accumulating inside muscle and liver cells. Remove the conditions that put that fat there, and the machinery can recover. The carbohydrate was never the thing that broke it.

If insulin resistance is the problem, why do carbohydrates get the blame?

Because blood sugar is where the problem becomes visible. A glucose meter measures glucose, so the food that raises it looks guilty. But a high reading after a meal tells you the cells are not clearing glucose efficiently; it does not tell you why. The why is the jammed signal described in the insulin section.

There is a real kernel of truth in the blame, and it is about which carbohydrates. Where carbohydrate genuinely is part of the problem, it is the refined kind: sugary drinks, juice, sweets, and white-flour products with the fiber stripped out. Juice drove a larger insulin response and a blood sugar rebound that whole apples did not,150 added sugar raised blood lipids in randomized trials,152 and high-glycemic-index diets tracked with more cardiovascular events in the largest cohort available.119

That is a verdict on junk carbohydrates, not on carbohydrates. The complex, fiber-rich kind comes from a long list of foods, and variety across all of them is the point: legumes, greens, whole grains, fruits, vegetables, mushrooms, nuts, seeds, seaweeds like nori, and herbs and spices. None of those caused anyone’s diabetes.

Why do low-carb and carnivore diets seem to work at first?

Because they are elimination diets, and elimination diets change the numbers quickly. The strictest versions keep meat, eggs and dairy and drop nearly everything else; gentler versions keep a few plants. Take away most of the carbohydrate and the after-meal glucose spikes shrink, because there is less glucose to handle. As the insulin section explains, that lowers the reading without repairing the cell’s ability to take glucose up.

Many people lose weight at first too, which also helps blood sugar. But weight loss is not unique to cutting carbohydrate. In a year-long randomized trial, a healthy low-fat diet and a healthy low-carbohydrate diet produced no significant difference in weight loss.99 And on a metabolic ward, people eating a plant-based low-fat diet ate 689 fewer calories a day than on an animal-based ketogenic diet, without trying.98

So plants get blamed, while the drivers that matter stay in place. Fat carries 9 calories per gram against 4 for carbohydrate,2 which makes fatty animal foods easy to overeat. Most dietary saturated fat comes from animal products,2 and saturated fat is what drove the extra liver and visceral fat in the overfeeding trial.133 Ketogenic diets also reliably raise LDL cholesterol.101

Even dairy carries a carbohydrate: lactose, the sugar in milk. It has to be split by the enzyme lactase before it can be absorbed.2 In many people, lactase production drops after infancy, a trait called lactase non-persistence, and it is the genetic default in humans rather than a disease.160 When someone feels better after dropping dairy, that is often the lactose talking, not a verdict on carbohydrate in general.

The durable fix is the opposite of elimination: clear out what jammed the machinery, and eat the complex carbohydrates that arrive packaged with fiber.

Why does it matter to bring blood sugar down early?

Because the body keeps a record. The UK Prospective Diabetes Study began in 1977 and eventually enrolled 5,102 people with newly diagnosed type 2 diabetes.163 It compared early intensive blood sugar control using drugs against a conventional strategy that relied primarily on diet. The latest follow-up, extending to 24 years after the trial ended, with a median of 17.5 years of individual follow-up, found that people originally assigned to intensive treatment with sulfonylureas or insulin still had 17 percent fewer heart attacks, 26 percent fewer microvascular complications and 10 percent lower all-cause mortality. The metformin group had 31 percent fewer heart attacks and 20 percent lower mortality. Neither intensive group showed a significant benefit for stroke or peripheral vascular disease. The legacy effects showed no sign of waning.166 Those benefits persisted long after the groups’ HbA1c levels had converged, which is why commentators summarize the lesson as hitting early and hitting hard right after diagnosis.163

Type 1 diabetes shows the same pattern. The DCCT randomized 1,441 people with type 1 diabetes to intensive or conventional insulin therapy for an average of 6.5 years. Intensive control, at a median HbA1c of 7 percent versus 9 percent, reduced the early stages of microvascular disease by 35 to 76 percent, at the cost of a threefold higher risk of low blood sugar. In the follow-up study, the benefit persisted even after the two groups’ blood sugar converged, which researchers call metabolic memory, and it grew into fewer severe complications and less cardiovascular disease.167164

One honest note about the UKPDS: its conventional arm relied primarily on diet, and the intensive drug strategies did better.166 That is not a verdict against diet in general. The trial compared strategies for lowering blood sugar, and early, tighter control won; neither trial tested a whole-food plant-based diet built to reverse insulin resistance. What they establish is the principle underneath the whole reversal story: the earlier blood sugar comes down, and the more safely it stays down, the more lasting the protection. That is the case for acting at prediabetes rather than waiting.

Is diabetes reversal the same thing as the stem cell headlines?

No, and headlines blur the two. According to a 2024 letter summarizing the first such case, a woman with type 1 diabetes, the autoimmune form, began producing her own insulin about two and a half months after a transplant of insulin-producing cells grown from her own reprogrammed stem cells, and stayed off insulin injections for over a year. Experts quoted in the same letter caution that more people need to be tested and that years of sustained insulin production are needed before calling anyone cured.165 That is a cell-replacement procedure for autoimmune diabetes. The reversal in this post is about type 2 diabetes, where the central problem this post has traced is insulin resistance, and the tool is what you eat.

Is it safe to try this on your own?

No, not if you take medication for blood sugar. This needs stating plainly, because these interventions can work too well. If someone on 200 units of insulin a day suddenly becomes insulin-sensitive and nobody adjusts the dose, their blood sugar crashes and that is a medical emergency. A clinician has to be ready to deprescribe rapidly. Anyone on insulin or oral hypoglycemic medication must make dietary changes of this magnitude under supervision.

Additional honesty about this evidence: several of these studies are small, short, or single-arm, and some involve calorie restriction alongside the dietary change, so the carbohydrate increase is not perfectly isolated from the weight-loss effect. The MADIAB randomized trial is the strongest of the group because it was isocaloric with an active comparator. The DiRECT trial, a calorie-restriction study rather than a plant-based one, independently found that remission requires reducing fat in the liver and pancreas, which points at the same mechanism from a different direction.63 The broader plant-based diabetes literature runs the same direction: randomized trials of low-fat vegan diets have improved glycemic control and cardiovascular risk factors,53 lowered HbA1c more than a conventional diabetes diet over 74 weeks when measured before any medication changes, though not in the main intention-to-treat analysis,54 and improved beta-cell function and insulin resistance in 16 weeks.55 The American College of Lifestyle Medicine has a position statement on diabetes remission worth reading if this applies to you.

Why glycemic load matters at all

Everything so far argues that whole carbohydrates are not the problem. That is true, and it is not the same as saying blood sugar does not matter. It does. Here is the actual biology, because the reason to care is more specific than “spikes are bad.”

What does high blood sugar actually do to you?

It sticks to things. Think of glucose as slightly sticky, like sugar syrup, and think of the proteins in your body as machinery running all day. When there is a lot of sugar around, it starts gumming onto that machinery on its own, no enzyme required, a process called non-enzymatic glycation. At first the sugar is only loosely stuck and can come back off. Given time it locks on permanently, and the gummed-up, hardened result is called an advanced glycation end product, fittingly abbreviated AGE. (The chemistry runs from a reversible Schiff base to a more stable Amadori product to the final AGE.)

You already have a test for this and probably did not realize it. The standard diabetes blood test, HbA1c, measures exactly this sugar-gumming. It counts how much of the hemoglobin in your red blood cells has glucose stuck to it, and because red cells live two to three months, that fraction reflects your average blood sugar over that window. Worth sitting with: your A1c is not really a measurement of your blood sugar. It is a measurement of how much sugar has already glued itself to one protein, used as a stand-in for what is happening to all the rest.

AGEs then bind a receptor called RAGE, which activates NF-kB and drives inflammation and oxidative stress. This is the mechanistic bridge between chronically elevated glucose and the chronic diseases the rest of this post is about. Glycation does not stop at hemoglobin, either. It also happens to the cholesterol-carrying LDL particles themselves, changing how long they linger in the blood and how they behave in the artery wall, which the plaque post walks through step by step.

Are the AGEs in food the same problem?

Partly, and this is a place where popular writing overshoots. Uribarri and colleagues built a database of AGE content across 549 foods and found that dry heat raises AGE formation more than tenfold, and in some cases a hundredfold, above the uncooked state.116 The highest dietary AGEs are in animal foods that are high in fat and protein and cooked with dry heat: broiled, grilled, fried, roasted. The lowest are fresh fruits, vegetables, and carbohydrate-rich low-fat foods.

But keep the proportions straight. The relative contribution of dietary versus internally formed AGEs is still debated, though this paper argues dietary AGEs are absorbed and contribute significantly to the body’s AGE pool. Either way, it is another mark against high-heat animal cookery.

What does a single blood sugar spike do?

Ceriello and colleagues answered this directly. They held people’s blood sugar at controlled levels for a full day, some at a steady high level and some deliberately swinging up and down every few hours, and did it in both healthy people and people with type 2 diabetes.117 Glucose excursions produced endothelial dysfunction and oxidative stress in a concentration-dependent way, in healthy subjects as well as diabetic ones. Two findings are worth pulling out. Oscillating glucose was significantly more damaging than sustained high glucose. And infusing vitamin C reversed the endothelial impairment, which tells you the damage was running through oxidative stress.

This is the same redox biology the antioxidant section covered, arriving from the other direction: metabolism generates reactive oxygen species, and so does glucose swinging up and down. The full treatment of reactive oxygen species, antioxidants, and how they reach your organs is in the anti-inflammatory diet post, and the downstream consequence for arteries, where oxidized particles get trapped and become plaque, is in the cholesterol post and the fat post.

What does excess sugar do to your mitochondria?

Mitochondria, the power plants inside your cells, are not fixed shapes. They constantly split apart and fuse back together, a cycle called fission and fusion, and the balance between the two shapes how they work. Excess glucose tips that balance toward splitting.

In the study that established this, researchers bathed rat liver cells and rat muscle-derived cells in glucose at 25 to 50 millimoles per liter.148 Within 15 minutes, about 70 percent of the cells had fragmented mitochondria, along with a burst of reactive oxygen species. Blocking the machinery that splits mitochondria, or boosting the machinery that fuses them, prevented the burst. Glucose that the cells could not metabolize caused neither effect. And when the high glucose continued for many hours, a second, much longer burst of reactive oxygen species followed.

So the picture is fragmenting rather than blowing up, and the first round was rapid and reversible, back to normal within an hour. The authors themselves describe that early fragmentation as a physiological event rather than damage. The concern is what happens when high glucose never lets up.

Scale matters here. Those concentrations work out to roughly 450 to 900 mg/dL, while the healthy people wearing glucose monitors in the study described below averaged about 99 mg/dL.120 This is cell-culture work far above anything a healthy person experiences after a meal, so it explains a mechanism rather than predicting what ordinary eating does. The plaque post follows this thread into the artery wall, where the same oxidative stress damages trapped cholesterol particles.

Does glycemic variability itself cause harm?

This is genuinely contested and this post is not going to pretend otherwise. Monnier and colleagues found that in type 2 diabetes, the mean amplitude of glycemic excursions correlated with oxidative stress more strongly than sustained chronic hyperglycemia did.118 That is real mechanistic support.

What has not been established is that glucose variability predicts cardiovascular events independently of HbA1c. Trials that specifically targeted post-meal glucose have not delivered the expected benefit, and the variability indices themselves are not standardized. Treat variability as biologically plausible and clinically unproven.

Does carbohydrate quality predict actual heart attacks?

Yes, in the largest dataset available. One PURE analysis followed 137,851 people across 20 countries on five continents for a median of 9.5 years, recording 8,780 deaths and 8,252 major cardiovascular events. People in the highest glycemic-index quintile had roughly 25 percent higher risk of a major cardiovascular event or death compared with the lowest, in those with and without existing cardiovascular disease, and the association was most pronounced in people carrying excess weight.119

Caveat worth stating: this is observational and based on food frequency questionnaires, and some of the signal likely reflects fiber and whole-grain intake rather than glycemic index as such. But it is the best evidence available that carbohydrate quality, not quantity, tracks with hard outcomes.

What a continuous glucose monitor is actually telling you

Glucose monitors are now sold to people who do not have diabetes, and the results are being misread on a large scale. This section is here because the misreading almost always runs in one direction: people see a normal spike and conclude they are metabolically broken.

What does a normal blood sugar curve look like in a healthy person?

We have the data. Shah and colleagues put blinded continuous glucose monitors on 153 healthy non-diabetic people aged 7 to 80 and simply watched.120 The results:

• Mean glucose: 98 to 99 mg/dL for every age group except those over 60, who averaged 104.
• Median time between 70 and 140 mg/dL: 96 percent (interquartile range 93 to 98 percent).
• Median time above 140 mg/dL: 2.1 percent, which is about 30 minutes out of 24 hours.
• Within-person coefficient of variation: 17 percent.

Read that third line carefully. Healthy people do go above 140. They just do not stay there. Roughly half an hour a day above 140 is what normal metabolic health looks like on a sensor.

Are spikes bad?

Spikes are what a functioning insulin response looks like. You eat carbohydrate, glucose rises, insulin is released, the glucose is cleared, and you return to baseline. That is the system working.

The meaningful distinction is not whether you spike but how long you stay elevated. A sharp rise with a prompt return to baseline is the signature of intact insulin sensitivity. A modest rise that stays up for hours, with a slow drift back, is the signature of insulin resistance, because the glucose is not being cleared efficiently. Area matters more than peak. Someone comparing their peak heights to a screenshot on social media is measuring the wrong thing.

Should a healthy person wear a glucose monitor?

There is no good evidence that it helps, and several reasons for caution.

First, accuracy. Sensor accuracy is expressed as mean absolute relative difference, and the headline figures of roughly 8 to 9 percent come from validation in people with diabetes across a wide glucose range. In the narrow near-normal range where healthy people actually live, accuracy degrades substantially. Second, there is a physiological lag of roughly 5 to 10 minutes, and up to about 20 minutes during rapid changes, because the sensor measures interstitial fluid rather than blood. Third, the consensus targets that apps display, the familiar 70 to 180 mg/dL time-in-range goals, were developed and validated for people with diabetes and have no established meaning for people without it.121

Reviews of non-diabetic monitor use describe it as an area with essentially no outcome evidence, and even experienced clinicians disagree substantially when asked to interpret the same non-diabetic sensor report.122 No major professional body recommends routine monitoring in healthy people.

There is also a real risk that gets too little attention: watching a number after every meal can drive genuine food anxiety and restriction in people who were metabolically fine to begin with. If you are going to wear one, interpret it with a clinician who can put it in context, and do not let a graph talk you out of eating beans.

Is a blood sugar spike the disease or the symptom?

The symptom. A glucose monitor measures glucose, and glucose is downstream of everything that matters. When someone with insulin resistance eats carbohydrate, their curve rises higher and stays up longer because their muscle and liver cells cannot clear the glucose efficiently, for the reasons described in the insulin section: fat that has spilled into those cells jams the insulin signal. The carbohydrate did not create that problem. It revealed it, the way running water reveals a clogged drain.

That is why removing carbohydrate is not the same as fixing anything. A monitor will show a flatter line on a meat-and-eggs day, because there is little glucose to handle, while the fat inside the cells, the actual lesion, is still there. The durable fix is to remove that fat, which is what negative energy balance and a whole-food plant-based diet have been shown to do.17349

Can you judge oatmeal by testing it on your own glucose monitor?

Not the way it is usually suggested, because the test is easy to rig without meaning to. Four things commonly go wrong:

• Swapping the food mid-argument. A claim about plain steel-cut oats cannot be tested with oatmeal topped with banana and honey. Adding free sugar changes the test, which is exactly why whole apples and apple juice behave so differently.150
• Testing right after low-carb eating. Carbohydrate tolerance adapts to what you have been eating. In healthy young men, just three days of a low-carbohydrate, high-fat diet made blood sugar significantly higher during a glucose tolerance test.159 Someone who has been eating mostly meat and then tries a bowl of oats is almost guaranteed to see a worse curve, and that result says more about the preceding diet than about the oats.
• Treating a normal rise as damage. Healthy people spend about half an hour a day above 140 mg/dL.120 A rise that comes back down promptly is the system working.
• Measuring only glucose. A glucose meter cannot see LDL cholesterol, fiber, or the fat accumulating in your liver. Oats lower LDL across 28 randomized trials,156 and meat’s flat glucose line tells you nothing about its saturated fat. Comparing a carbohydrate food against a food with no carbohydrate, on an instrument that only measures carbohydrate’s effect, is a rigged comparison.

Why are glucose graphs so popular with diet influencers?

Because a spike is dramatic on video, easy to reproduce, and points to a simple villain. It is worth recognizing the pattern when you see it: a graph shown without a baseline, without the rest of the day, and without how long it took to come back down; a whole food blamed for a response that is really a sign of insulin resistance; a “test it yourself” challenge built on the problems above; and a call to action that ends at a paid community, a program, a course, or a book.

None of that proves a particular person wrong about everything, but a sales funnel is not evidence. The better question is what the claim looks like when it is tested properly: in randomized trials measuring outcomes such as HbA1c, body weight, LDL cholesterol, liver fat and heart disease, which is the kind of evidence this post is built on.

If you want to go low-carb, you can do it on plants

Some people arrive at this topic worried about a specific number: a fasting glucose creeping up, an A1c in the prediabetic range, a doctor’s comment. The honest answer is that you have options, and none of them require eating animals.

Can you eat low-carb on a vegan diet?

Comfortably. Look at the carbohydrate density table further down this page. Nuts, seeds, avocados, olives, tofu, tempeh, mushrooms, and non-starchy vegetables are all very low in net carbohydrate per calorie, several of them under 5 percent. A diet built on those foods plus greens is low-carbohydrate by any reasonable definition.

The best-studied version is the “Eco-Atkins” pattern tested by Jenkins and colleagues: a plant-based low-carbohydrate diet built on soy, gluten, nuts, vegetables, and vegetable oils. In a six-month randomized trial in 39 people with high cholesterol, the plant-based low-carbohydrate diet lowered LDL cholesterol by an additional 0.49 mmol/L (95 percent confidence interval 0.70 to 0.28, p<0.001) compared with a high-carbohydrate vegetarian control, along with greater reductions in triglycerides and weight.134

Note what that means. Going low-carbohydrate on plants lowered LDL. Going low-carbohydrate on animals characteristically raises it. The macronutrient label is not what determines the outcome, which is the same lesson as the animal-versus-plant cohort data above.

Is there ever a good reason to do a ketogenic diet?

Yes, and they are more specific than the internet suggests.

The strongest indication is drug-resistant epilepsy. The Cochrane review of 13 trials and 932 participants found children on ketogenic diets may be up to three times more likely to achieve seizure freedom and up to six times more likely to achieve a 50 percent or greater reduction in seizures compared with usual care, with up to 55 percent of children achieving seizure freedom on the classical 4:1 diet, though the review rates all of this evidence as low to very low certainty.135 Ketogenic diets are also the established treatment for GLUT1 deficiency syndrome and pyruvate dehydrogenase deficiency, where ketones supply an alternative fuel the brain can actually use.

For type 2 diabetes, the American Diabetes Association describes very-low-carbohydrate eating, which often targets 20 to 50 grams of carbohydrate a day, as a viable approach for select people not meeting glycemic targets or where reducing medication is a priority.136 It is framed for a specific subgroup, not as a general recommendation, and medication adjustment is essential.

How long should someone stay on it?

Not indefinitely, and almost certainly not by choice once the underlying problem is addressed. Several things are worth weighing.

As the umbrella review described earlier found, raising LDL cholesterol is one of the few ketogenic effects backed by high-quality evidence, and that evidence comes from trials with a median duration of only 13 weeks.101 Long-term adherence is poor. And by definition a ketogenic diet removes the foods that feed your gut bacteria, which is the butyrate problem covered earlier in this post.

You will also encounter the argument that dramatic LDL elevations on ketogenic diets are harmless in lean people, the “lean mass hyper-responder” phenotype. That claim is not established. It is a real observed pattern, and the interpretation that it is benign remains contested even by researchers describing it, one of whose own commentaries notes that such elevations deserve urgent clinical attention and further research. Do not bet your arteries on an unsettled hypothesis.

What is the actual recommendation here?

If you want to lower carbohydrate intake for a defined period while you address the underlying issue, it can be done well on whole plant foods, and the evidence says it will treat your cholesterol better than the animal version. Do it with a physician, registered dietitian, or nutritionist who practices lifestyle medicine, particularly if you take any medication for blood sugar, because these changes can require rapid deprescribing.

And hold onto the goal. Restriction is a tool for a period, not a destination. The aim is to repair insulin sensitivity so that you can eat a bowl of lentils without thinking about it, which is what the next few sections are about.

Muscle is a glucose sink

Here is the part of the blood sugar conversation that almost never makes it into the food fight, and it may be the highest-leverage thing in this entire post.

Where does the glucose in your blood actually go?

Mostly into muscle. Skeletal muscle accounts for the large majority of insulin-stimulated glucose disposal, which means your muscle mass is, quite literally, the size of the tank you have available to put glucose into. More muscle is a bigger sink.

The population data follow. In an analysis of 13,644 adults in a national US survey, relative skeletal muscle mass was inversely associated with insulin resistance and with prediabetes. After adjusting for age, sex, ethnicity, and both overall and abdominal obesity, people in the top quarter of muscle as a share of body weight had 26 percent lower insulin resistance and a 28 percent lower prevalence of prediabetes or diabetes than people in the bottom quarter. The relationship held across the whole range, not just at the low-muscle end.123

Three honest caveats. This was a snapshot survey, so it shows association rather than cause, and the authors say plainly they cannot definitively establish that more muscle protects. Muscle was estimated from electrical impedance rather than measured directly. And the link with diagnosed diabetes itself did not survive the adjustments. Still, the direction is consistent with the physiology: more muscle means a larger place to put glucose, which is why carrying muscle is a metabolic asset rather than a cosmetic one.

Why does walking lower blood sugar even if you are insulin resistant?

Because muscle has a second door, and insulin does not hold the key to it.

Glucose enters muscle through the GLUT4 transporter, which normally moves to the cell surface when insulin signals it to. But GLUT4 also translocates in response to muscle contraction itself, through a completely separate pathway involving AMPK, calcium, and nitric oxide signaling.124 This contraction pathway is insulin-independent.

The implication is enormous and underappreciated. Remember the jammed hallway: even when fat has broken the message between the doorbell and the door, contracting the muscle walks over and opens the door itself, with no doorbell and no hallway involved. This is the mechanistic reason exercise lowers blood sugar in people whose insulin is not working well. You are not fixing the wiring, you are opening the door by hand.

How much walking does it take?

Startlingly little, and the timing matters more than the duration. A meta-analysis found that light walking breaks reduced post-meal glucose by about 17 percent compared with prolonged sitting, with benefits appearing at durations as short as two to five minutes.125 Other work found that ten minutes of walking after each main meal improved glycemia more than a single 30-minute walk, with the post-dinner walk contributing most, since that is typically the largest meal at the time of day when insulin sensitivity is lowest.

Walking as soon as possible after eating beats walking later. A short walk after your biggest meal is probably the single cheapest metabolic intervention available to anybody.

Does building muscle help too?

Yes, and measurably. Meta-analyses of resistance training in type 2 diabetes find HbA1c reductions of roughly 0.4 to 0.5 percentage points, comparable to what some oral medications achieve.126 In one pooled analysis of 46 trials, HbA1c fell by 0.50 percentage points and fasting glucose by about 12 mg/dL.

So the full picture of “carbohydrate tolerance” is not a property of the food. It is a property of how much muscle you have, how recently you used it, and how much fat is sitting inside it.

How someone actually becomes prediabetic

Now the pieces assemble into a sequence, and notice how little of it is about carbohydrate.

What is the actual chain of events?

1. Chronic calorie surplus, most easily achieved on calorie-dense food. Fat carries 9 calories per gram against 4 for carbohydrate and protein, and animal foods carry no fiber to fill you up.
2. Fat accumulates past the personal fat threshold, the point beyond which subcutaneous tissue stops absorbing the surplus and it spills into the liver, the muscles, and the abdomen.
3. The type of fat matters, not just the amount. In a double-blind randomized overfeeding trial, young, normal-weight adults gaining similar amounts of weight on muffins made with saturated fat versus polyunsaturated fat had strikingly different outcomes: saturated fat markedly increased liver fat and caused roughly twice the increase in visceral fat.133
4. Intramyocellular lipid jams the insulin signal through the DAG, ceramide, PKC, and IRS-1 pathway described earlier. The receptor still binds insulin. The message just does not get through.
5. Glucose cannot get into the cell, so it stays in the blood. Fasting glucose drifts up. A1c drifts up.
6. The pancreas compensates by producing more insulin, which works for years until it does not.

At no point in that chain is carbohydrate from whole foods the cause. Added sugar eaten in a surplus is the exception, because the liver turns it into fat, as the next section shows. The carbohydrate is the thing that becomes visible when the machinery for handling it is broken, in the same way a leak becomes visible when you turn on the tap. Blaming the water for the leak is the error this entire post is about.

What are the numbers on prediabetes?

Prediabetes is defined by the American Diabetes Association as a fasting glucose of 100 to 125 mg/dL, an HbA1c of 5.7 to 6.4 percent, or a two-hour glucose tolerance result of 140 to 199 mg/dL.

The important context is that it is a risk state, not a sentence. Roughly 5 to 10 percent of people with prediabetes progress to diabetes each year, and a similar proportion revert to normal glucose.129 In older adults, reverting to normal is more common than progressing. Achieving that reversion produces durable long-term risk reduction.

Calories, saturated fat or sugar: what actually puts fat in your liver and muscles?

The chain above starts with a calorie surplus and ends with fat inside liver and muscle cells. That leaves a fair question, and it is the one low-carb and plant-based camps argue about most: is the problem simply too many calories, or does it matter whether those calories come from saturated fat, unsaturated fat or sugar? Controlled feeding trials, where researchers decide exactly what people eat, can answer this better than surveys can. Here is what they show, in rough order of how much each factor matters.

Is fatty liver just a matter of eating too many calories?

Mostly, yes. Weight change moved liver fat more than anything else in these studies. When 16 overweight adults were overfed simple sugars for three weeks, they gained 2 percent of their body weight and their liver fat rose 27 percent. Six months on a reduced-calorie diet then took off 4 percent of their weight and 25 percent of their liver fat.200 In adults with fatty liver disease, losing about 10 percent of body weight over six months cut liver fat from 19.4 percent to 4.5 percent and improved insulin sensitivity by 46 percent.201 The low-fat vegan trial described earlier fits the same pattern: participants lost 6.4 kilograms and fat inside their liver cells fell 34 percent.49

Does saturated fat build more liver fat than other calories?

Yes, when people are overeating. In a randomized trial, 38 overweight adults ate an extra 1,000 calories a day for three weeks, from saturated fat, unsaturated fat or simple sugars. Weight gain did not differ significantly between the groups (1.4, 0.9 and 1.4 kilograms). Liver fat rose 55 percent on saturated fat, 33 percent on sugar and 15 percent on unsaturated fat. Saturated fat produced the greatest rise in insulin resistance, a significant 23 percent on a fasting measure called HOMA-IR, and it was the only diet that raised total plasma ceramides, by 49 percent. Ceramides are the fat-derived molecules that interfere with insulin signaling.185 A double-blind Swedish trial in 61 overweight adults found the same direction over eight weeks: muffins made with palm oil raised liver fat by half, while sunflower-oil muffins did not raise it at all, despite similar weight gain.186 Together with the overfeeding trial in young normal-weight adults described above,133 that is three randomized trials pointing the same way.

Does saturated fat cause insulin resistance even without overeating?

Here the evidence is thinner. The best trial fed 162 healthy adults a controlled diet with matched calories for three months, high in either saturated or monounsaturated fat. Insulin sensitivity differed only when total fat stayed below 37 percent of calories: it was 12.5 percent lower on the saturated fat diet and 8.8 percent higher on the monounsaturated one. At higher total fat intakes, no significant effect was seen.187 Across 102 controlled feeding trials, replacing saturated fat with polyunsaturated fat lowered blood glucose, HbA1c and insulin resistance.191 But longer trials, mostly of fish-oil supplements, found little or no effect on new diabetes diagnoses,192 and large population studies have not linked total saturated fat intake to type 2 diabetes.188189 Blood measurements help explain the mixed picture, because saturated fats do not all behave alike. Palmitic acid in the blood tracked with higher diabetes risk, at low certainty, while two odd-chain saturated fats tracked with lower risk, at moderate certainty.190 In a six-week trial, healthy, slightly overweight men ate diets rich in saturated or polyunsaturated fat while their weight was held stable, and peripheral insulin sensitivity did not change with either fat.264 The fair summary: saturated fat reliably builds liver fat during overeating, while its effect on insulin sensitivity at stable weight is modest at most and depends on the rest of the diet.

Does sugar turn into fat in your body?

It can. The liver can build new fat from carbohydrate, a process called de novo lipogenesis, and much of what it makes is palmitic acid, the same saturated fat discussed above. The researchers behind the blood-marker analysis note that the body makes palmitic acid largely when sugar or alcohol is eaten in excess.190 In the three-week overfeeding trial, extra sugar raised this fat production by 98 percent, which is how sugar raised liver fat 33 percent.185 Fructose is the most potent form. In eight healthy men on a metabolic ward, a weight-maintaining diet with 25 percent of calories from fructose, compared with the same calories from complex carbohydrate, raised the share of newly made fat from 11 to 19 percent and raised liver fat by a median of 137 percent.197 Run the experiment the other way and the result reverses: when 41 children with obesity had their sugar swapped for starch at the same calories for nine days, liver fat fell from a median of 7.2 to 3.8 percent, and the drop held after accounting for weight change.198 This is the same fructose problem shown in the sweetened-drink trial in the sugar section below.153

One conflict is worth flagging rather than hiding. A pooled analysis of feeding trials found that fructose raised liver fat only when it added extra calories, and not when it replaced other carbohydrate.199 But in 86 percent of its “same calorie” trials, both groups were being overfed, and fructose was compared mainly with glucose rather than starch, so it does not settle the question for people eating at their needs.

Does your liver still turn carbs into fat when you eat less?

Barely, if at all. In a controlled study that set carbohydrate intake above or below people’s needs for five days at a time, the liver’s fat production rose more than tenfold on surplus carbohydrate and became unmeasurable when carbohydrate fell short. Even at a 50 percent carbohydrate surplus, it made fewer than 5 grams of fat a day.193 The form of the carbohydrate matters as much as the amount. On a calorie-matched liquid formula of glucose chains with only 10 percent fat, 44 percent of the fat in the blood’s triglyceride particles was newly made.194 In a follow-up, volunteers moved from that kind of formula to solid food with the same 10 percent fat, built from starch, sugars and fiber, and the newly made share fell from 30 to 54 percent down to 0 to 1 percent.195 A whole-food low-fat, high-carbohydrate diet in another study kept the liver’s contribution below 5 percent of fasting triglyceride.196 Weight loss pushes it lower still: in the 10 percent weight-loss study above, its contribution fell by two-thirds.201

So the “carbs turn into fat” argument is really about liquid and added sugar eaten in a calorie surplus. It does not describe beans, oats and potatoes eaten at or below your needs.

Can exercise cancel out overeating?

Partly, and more than many people assume. In a randomized trial, 26 active young men overate by 50 percent and cut their walking below 4,000 steps a day for a week. In the group that did no exercise, the insulin response to a glucose drink doubled. A second group added 45 minutes of daily treadmill running while given extra food to keep the same calorie surplus, and most of those changes were prevented.203 In eight healthy men on a weight-maintaining, high-fructose diet, two 30-minute cycling sessions a day prevented the rise in blood triglycerides that the fructose otherwise caused.204 That fits the athlete’s paradox described earlier: muscle that burns fat as fuel can hold more of it without jamming the signal. The limits are real, though. These were short studies in healthy young men, neither tested saturated fat, and neither measured liver fat, so whether exercise offsets months or years of surplus has not been tested in them.

Is saturated fat worse because it is easier to overeat?

Both mechanisms are at work, and our reading of the evidence is that overeating is the bigger one. On a metabolic ward, people offered ultra-processed food ate about 500 more calories a day than when offered unprocessed food matched for the calories, sugar, fat and fiber on the plate, and they gained 0.9 kilograms in two weeks.202 The same research group found people ate 689 fewer calories a day on a low-fat plant-based diet than on an animal-based ketogenic one.98 A gap that size, sustained, changes body weight, and body weight moved liver fat by 25 to 77 percent in the studies above. Per calorie, saturated fat adds a second push: at the same surplus, it built roughly 1.7 times the liver fat that sugar did and 3.7 times what unsaturated fat did.185 Food that is both calorie-dense and high in saturated fat gets you on both counts.

Do plant-based diets beat healthy diets that include meat?

For blood sugar, modestly, and mostly through weight. In identical twins assigned to a healthy vegan or a healthy omnivorous diet for eight weeks, the vegan twins’ fasting insulin fell 2.9 μIU/mL more than their siblings’, and they also lost 1.9 kilograms more.205 In a crossover trial, a low-fat vegan diet beat a Mediterranean diet on weight by 6 kilograms and improved one measure of insulin sensitivity more, while the difference in HOMA-IR was not significant and blood pressure fell more on the Mediterranean diet. All but two of its authors work for the Physicians Committee for Responsible Medicine, which promotes plant-based eating.206 Pooled across randomized trials, vegan diets lowered HbA1c by 0.18 percentage points overall207 and vegetarian or vegan diets lowered it by 0.40 points in people with type 2 diabetes.208

None of those trials can separate the diet from the weight loss that came with it, and one shows how much the weight carries. In the 74-week diabetes trial, the HbA1c advantage seen before medication changes was no longer significant once weight change was included in the analysis; each kilogram lost went with a 0.09-point drop in HbA1c.54 That is not a mark against the diet. Losing weight without counting calories is one of the main things a whole-food plant-based diet does. For preventing diabetes in the first place, the evidence is observational: across nine cohort studies of more than 300,000 people, closer adherence to a plant-based diet was linked to 23 percent lower risk, and 30 percent lower when the pattern emphasized healthy plant foods.209

So is it calories, saturated fat or sugar?

All three, in that order. First, the calorie balance that decides whether fat builds up in your liver and muscles at all. Second, the type of fat: during a surplus, saturated fat is the strongest amplifier of liver fat and raised insulin resistance the most. Third, added and liquid sugar, especially fructose, which the liver turns into saturated fat when you eat more than you need. Whole-food starch eaten at or below your needs produces almost none.

That is why a whole-food plant-based diet works for insulin sensitivity: it tends to lower calories without counting, it is low in saturated fat, and its carbohydrate arrives as starch and fiber rather than syrup. It is also why a plant-based diet built on juice, sweets and white flour does not earn the same credit.

What “metabolically healthy” actually means

This phrase gets used constantly without a definition attached. It has one.

What are the criteria?

Araújo and colleagues assessed five components in a national sample, requiring all five to be met without medication:127

• Waist circumference under 102 cm in men, 88 cm in women
• Glucose: fasting under 100 mg/dL and HbA1c under 5.7 percent
• Blood pressure under 120 systolic and under 80 diastolic
• Triglycerides under 150 mg/dL
• HDL cholesterol at or above 40 mg/dL in men, 50 mg/dL in women

How many people are metabolically healthy?

Almost nobody. 12.2 percent of American adults met all five criteria (95 percent confidence interval 10.9 to 13.6), about 27 million people. Among those with overweight it was 8.0 percent. Among those with obesity it was 0.5 percent. And fewer than one in three normal-weight adults qualified.127

That figure reframes the entire carbohydrate debate. When someone says carbohydrates cause problems for them, they are often right about the observation and wrong about the cause. Nearly nine in ten adults are carrying some degree of metabolic dysfunction. The food is revealing it, not creating it.

Can you be metabolically healthy and carry excess weight?

Temporarily, but it tends not to hold. In the MESA cohort, metabolically healthy obesity proved largely transitional: people converted to metabolically unhealthy status over time, and the excess cardiovascular risk was explained by that transition and by how long metabolic syndrome persisted.128 Similar patterns appear in long-running cohorts of women followed for two decades. Treat it as an early stage rather than a stable state.

Does sugar feed cancer?

What is the Warburg effect?

In the 1920s Otto Warburg observed that cancer cells consume glucose voraciously and produce lactate even when oxygen is plentiful, a pattern called aerobic glycolysis. He hypothesized that damaged mitochondria cause cancer. The observation was real and is still used clinically; PET scans work by finding tissue that takes up a lot of glucose.

So should cancer patients avoid carbohydrates?

The evidence does not support it, and the reason is that cancer is metabolically flexible.

Tumors also depend heavily on glutamine, an amino acid, and on fatty acids. When researchers genetically deleted the key glycolysis enzyme LDHA in a squamous cell carcinoma model, tumor formation was unaffected; the tumors simply switched to glutaminolysis, and when that was blocked too, they switched to consuming lactate.6465 Drugs designed to block the glucose pathway have not succeeded clinically.

More fundamentally, the Warburg effect describes a metabolic hallmark of established tumors. It does not explain cancer’s origin, which traces to oncogenic mutations. “Starve cancer of sugar” fails on both ends: it misidentifies the cause, and it targets a fuel the cancer can readily substitute. Meanwhile the diets built on that premise remove the fiber, the short-chain fatty acids, and the phytochemicals that the actual evidence associates with lower cancer risk.

Carbohydrates are not “essential,” and it does not matter

Is there such a thing as an essential carbohydrate?

No, and the chemistry is straightforward. A nutrient is “essential” for exactly one reason: your body cannot build it, so you must eat it. Your liver can build glucose out of other things entirely, spare parts from protein and fat, whenever it needs to. (The technical name for that build-it-yourself process is gluconeogenesis.) So carbohydrate fails the “you must eat it” test.

But “essential” was never a shopping list for a healthy diet. Fiber is not classified as essential either, and neither is any antioxidant or phytochemical, and those are among the most protective things you can eat. By the same standard, exercise is not essential. Nobody concludes that exercise is optional.

The full dismantling of this argument, including what the same National Academies report says three pages later, is in the essential nutrients post.

Why does the body work so hard to make glucose if it does not need it?

Because glucose sits at the center of how cells make energy, even though you do not have to eat it. It is the easiest thing to feed into the cell’s energy line, a conveyor belt that turns glucose into the usable energy currency your cells run on. The body guards its blood glucose level tightly precisely because so much depends on that conveyor belt staying fed. (The line, in order: glycolysis to pyruvate to acetyl-CoA into the citric acid cycle and oxidative phosphorylation.)

Two cases make this vivid. Red blood cells have no mitochondria at all, so they run exclusively on anaerobic glycolysis of glucose; they have no alternative. And the brain’s primary fuel is glucose, which is why the reference intake was set on brain glucose consumption. The brain can partially adapt to ketones during prolonged fasting, but glucose is the default.

During a fast, the order is predictable: liver glycogen is drawn down first, roughly within a day, and then the body shifts increasingly toward fat-derived ketones while gluconeogenesis covers the obligate glucose needs. That the body has an emergency generator does not mean you should run the house on it permanently.

How much carbohydrate is actually in each food group

Almost nobody presents plant foods this way, and it changes the conversation. Below are USDA values per 100 grams, with net carbohydrate meaning total carbohydrate minus fiber, and the final column showing what percentage of each food’s calories come from net carbohydrate.137 That last column is the one that matters, because it tells you what you are actually eating when you eat to a given number of calories.

Which plant foods are highest and lowest in carbohydrate?

All values below are per 100 grams of food.

Food	Cal	Carb	Fiber	Net carb	% cal
Macadamia nuts	718	13.8	8.6	5.2	3%
Pecans	691	13.9	9.6	4.3	2%
Walnuts	654	13.7	6.7	7.0	4%
Almonds	579	21.6	12.5	9.1	6%
Flaxseed	534	28.9	27.3	1.6	1%
Chia seeds	486	42.1	34.4	7.7	6%
Hemp seeds	553	8.7	4.0	4.7	3%
Pumpkin seeds	559	10.7	6.0	4.7	3%
Olives	145	3.8	3.3	0.5	2%
Avocado	160	8.5	6.7	1.8	5%
Tofu, firm	144	2.8	2.3	0.5	1%
Edamame	122	9.9	5.2	4.7	16%
Spinach, raw	23	3.6	2.2	1.4	25%
Romaine lettuce	17	3.3	2.1	1.2	28%
White mushrooms	22	3.3	1.0	2.3	41%
Kale, raw	49	8.8	3.6	5.2	42%
Broccoli, raw	34	6.6	2.6	4.0	48%
Cauliflower, raw	25	5.0	2.0	3.0	48%
Zucchini, raw	17	3.1	1.0	2.1	50%
Raspberries	52	11.9	6.5	5.4	42%
Lentils, cooked	116	20.1	7.9	12.2	42%
Black beans, cooked	132	23.7	8.7	15.0	45%
Chickpeas, cooked	164	27.4	7.6	19.8	48%
Oatmeal, cooked	71	12.0	1.7	10.3	58%
Quinoa, cooked	120	21.3	2.8	18.5	62%
Strawberries	32	7.7	2.0	5.7	71%
Sweet potato, baked	90	20.7	3.3	17.4	77%
Brown rice, cooked	123	25.6	1.6	24.0	78%
Potato, baked with skin	93	21.2	2.2	19.0	82%
Blueberries	57	14.5	2.4	12.1	85%
Apple	52	13.8	2.4	11.4	88%
Banana	89	22.8	2.6	20.2	91%

What does that table actually tell you?

Three things, and the second one surprises most people.

First, the plant kingdom spans nearly the entire possible range. From flaxseed at 1 percent of calories as net carbohydrate to banana at 91 percent, you can construct essentially any carbohydrate intake you want without eating an animal. The idea that “plant-based” and “high-carb” are synonyms is a failure of imagination, not a fact about plants.

Second, leafy greens are not low-carb per calorie. They are very low in carbohydrate per gram, which is why they feel like free food, and they are. But because they contain almost no calories at all, the small amount of carbohydrate they do contain represents a substantial share of those calories. Spinach is 25 percent carbohydrate by calories. Zucchini is 50 percent. This is a distinction worth understanding, because people routinely mix up “low-carb by weight” with “low-carb by calorie,” and the two lead to different diets.

Third, fiber changes everything about how you read a label. Chia seeds are 42 grams of carbohydrate per 100 grams, which sounds enormous, until you notice that 34 of those grams are fiber that you do not digest. Flaxseed is 29 grams of carbohydrate of which 27 are fiber. These are, functionally, among the lowest-carbohydrate foods in existence, and they look like the highest on a nutrition panel.

Which whole plant foods are genuinely low-carb?

If you want a short list to work from: nuts (macadamia, pecan, walnut, almond, hazelnut), seeds (flax, chia, hemp, pumpkin, sunflower), avocado, olives, soy foods (tofu, tempeh, edamame), mushrooms, and the whole range of non-starchy vegetables and leafy greens. Add berries, especially raspberries and blackberries, when you want fruit.

That is a satisfying way to eat lower in carbohydrate for a period, and nobody has to eat a ribeye to get there.

Which food groups have the most fiber?

Sort the same table by fiber and the order changes, and it changes again depending on whether you count per gram or per calorie.137 Per 100 grams, seeds lead by a wide margin (chia 34.4 grams, flaxseed 27.3), followed by nuts and cooked legumes (black beans 8.7, lentils 7.9, chickpeas 7.6). Cooked whole grains come in lower than their reputation suggests, at 1.6 to 2.8 grams, partly because cooking adds water. Most vegetables, fruits and greens sit between about 1 and 4 grams, with raspberries (6.5) and avocado (6.7) as standouts, and white mushrooms come in at 1.0. Mushrooms are fungi rather than plants, and their fiber reflects it: it includes chitin and glucans joined by beta bonds, a different chemistry from plant cell walls.246

Per calorie, the picture nearly flips. Raspberries and romaine lettuce deliver about 12 grams of fiber per 100 calories and spinach about 10, while broccoli, kale, chia, lentils and black beans land around 7. Cooked oatmeal and quinoa deliver about 2.4, brown rice about 1.3, and most nuts 1 to 2, because their calories come mostly from fat. That is why the most useful answer to “which foods have the most fiber” is not one food. Seeds and legumes are dense sources by weight, greens and berries are dense sources by calorie, and a plate that combines them covers both.

How can a food be high in fiber but low in carbohydrate?

Because fiber is counted inside total carbohydrate. Total carbohydrate on a label includes sugars, starch and fiber together, so net carbohydrate, total minus fiber, approximates the part your own enzymes can turn into blood glucose. When most of a food’s carbohydrate is fiber, total carbohydrate can look high while net carbohydrate stays low. Chia seeds show it best: 42.1 grams of total carbohydrate per 100 grams, of which 34.4 grams is fiber, leaving 7.7 grams net. Brown rice is the opposite case: 25.6 grams total, only 1.6 grams of fiber, and 24.0 grams net.137

Net carbohydrate is a useful number, but it is not a verdict. A low-net-carb food is not automatically healthier, and a high-net-carb whole food is not automatically a problem; the evidence on lentils, intact grains and blood sugar earlier in this post is about exactly those starchy whole foods. And the fiber you subtract is not nothing. It is the part that feeds your gut bacteria, as the fiber section explains.

A tour of the categories, and what is special about each

This is where the variety argument becomes concrete. Estimates of edible plant species vary widely with the definition used. The FAO commonly cites roughly 30,000 edible species, of which about 6,000 to 7,000 have been cultivated; the plant-science literature puts the figure as high as 250,000, with roughly 90 percent of the world’s calories coming from under one percent of them.87 Whichever number you take, wheat, corn and rice dominate and the rest is barely touched. The gap between what exists and what people eat is enormous, and closing even a little of it is where the interesting food is.

Which grains should you try beyond brown rice?

Start with oat groats, the intact whole oat before anything has been rolled or cut. Then kamut (khorasan wheat), farro (emmer), freekeh (young green wheat, smoked), einkorn (the oldest wheat), spelt, sorghum, teff (the tiniest grain, the base of injera), millet, barley, rye, and brown, black and red rice. Bulgur is parboiled cracked wheat and cooks in minutes. Quinoa, amaranth and buckwheat are pseudocereals, seeds from non-grass plants.

Are ancient grains actually better?

The evidence leans yes, and unusually for this topic it includes randomized human trials.

Einkorn and kamut carry more of the carotenoids lutein and zeaxanthin than modern bread wheat, partly because those yellow pigments were deliberately bred out to make white bread white.88 Modern wheat tends to have more vitamin E, so on straight vitamins and minerals it is close to a wash, but the primitive wheats show greater antioxidant capacity from higher polyphenol content. In human liver cells exposed to an inflammatory stimulus, digested bread made from ancient grains suppressed the inflammation better than bread from modern strains.89

The human trials all substituted kamut for modern wheat at equal amounts:

• Irritable bowel syndrome: less abdominal pain, less frequent pain, less bloating, better stool consistency, less interference with quality of life.67
• Non-alcoholic fatty liver disease: improved liver function.68
• Type 2 diabetes: better cholesterol and insulin sensitivity.69
• Acute coronary syndrome: better glycemic control and cholesterol.70
• Healthy adults: better arterial function.71

The honest caveat: the total number of human trials is small and several come from overlapping research groups, so you cannot conclude that ancient wheats beat every modern variety. The best available data lean that way.

Is quinoa a superfood?

It is nutritious, and the marketing is doing some work. One good randomized dose-response trial found that about a cup a day of cooked quinoa for twelve weeks produced a 36 percent drop in triglycerides, comparable to what you get from triglyceride-lowering drugs or high-dose fish oil.72 Red and black quinoa carry roughly double the antioxidant capacity of white.73

The counterweight: quinoa’s fiber content is only modest compared with barley or rye, and its higher protein is not especially useful given that protein is not what people are short on. Fiber is. Eat quinoa because it is good and interesting, not because a label called it a superfood. (Also, mention it to your gastroenterologist before a colonoscopy; undigested quinoa has been mistaken for parasites.91)

Do whole grains actually protect your arteries?

In postmenopausal women with existing coronary disease who had angiograms years apart, those eating less than one serving of whole grains a day had significantly more arterial narrowing than those eating one serving or more. The higher-whole-grain group’s disease still progressed, but significantly less, at a rate approaching what statins achieve for slowing progression.35

Which raises the obvious question: do you want to slow the rate at which you die of heart disease, or not die of it? Oatmeal with bacon and eggs beats bacon and eggs, but a whole-food plant-based pattern, without the bacon and eggs, is the better target.

Which legumes should you buy?

All of them, honestly, but here is the map. Lentils come in red, green, yellow, black (beluga) and brown, and the small ones cook in fifteen minutes without soaking. Beans: black, pinto, navy, cannellini, great northern, kidney, small red, adzuki, mung, lima, black-eyed peas. Split peas, green and yellow. Whole peas. Chickpeas, also called garbanzo beans or Bengal gram. Soybeans and their young green form, edamame.

Which beans have the most antioxidants?

This depends on the assay, which is a useful lesson in itself.

In the USDA’s ORAC dataset, small red beans ranked first among all foods tested on a per-serving basis, narrowly ahead of wild blueberries, with red kidney beans and pinto beans close behind.41 But in test-tube comparisons, the order shifts with the method. Among ten common legumes, lentils had the highest phenolic content and antioxidant activity, while baby lima beans and navy beans had the lowest.42 In another comparison of nine legumes, black beans, lentils, black soybeans and red kidney beans best protected LDL cholesterol from oxidation.43 One proposed explanation for lentils is surface area: they are small, and the antioxidants concentrate in the seed coat.

The practical answer is not to crown a winner. Lentils, small red beans, black beans, black soybeans and red kidney beans consistently cluster at the top, every legume tested is a strong source, and the variety is the point.

Is edamame really as high in protein as animal foods?

Close enough that the comparison is fair. A cup of cooked edamame carries roughly 18 grams of complete protein, and soy protein scores at or near the top of the standard protein quality scales, comparable to meat, eggs and dairy.39 It also arrives with fiber and isoflavones, which no chicken breast carries, along with folate, iron and calcium.

Do beans protect your bones?

The association is striking. In the Adventist Health Study-2, eating legumes once a day or more was associated with a 64 percent lower risk of hip fracture than eating them less than once a week (hazard ratio 0.36, 95 percent confidence interval 0.21 to 0.61).105 Meat eaten four or more times a week was also associated with lower risk (40 percent), as were meat analogues eaten daily (49 percent); the authors describe the legume association as the stronger of the two food findings. They point to lysine, which legumes supply more of than other plant proteins, as one possible reason.

This is observational, and bone health is a genuinely nuanced area where vegans have shown higher fracture risk in some cohorts, tied largely to calcium and vitamin D rather than to plants as such. The full picture is in do vegans have weaker bones. The narrow point here is that the bean keeps landing on the right side of outcomes that carbohydrate foods are supposed to harm.

What do chickpeas do to your gut in three weeks?

In a randomized crossover trial, 12 healthy adults added 200 grams of canned chickpeas a day for three weeks. Fewer of them carried bacterial groups that include pathogenic and putrefactive species, and a high-ammonia-producing bacterium turned up in 42 percent of people on chickpeas versus 83 percent on the control diet. Short-chain fatty acid levels did not change.78 Canned chickpeas. Three weeks.

What about greens, vegetables, fruits, mushrooms, nuts, seeds and seaweed?

Greens are the nitrate powerhouses driving the nitric oxide pathway above: arugula, spinach, chard, beet greens, lettuce, bok choy, collards, kale, mustard greens, watercress. They are also among the most nutrient-dense foods per calorie in existence.

Vegetables are where the carbohydrate quality argument gets most visible. Consider potatoes, the classic “bad carb.” Across varieties, glycemic index tracks polyphenol content, with purple flesh measuring lowest and white highest (purple 77, red 78, yellow 81, white 93). Worth stating plainly: in that small study the differences between varieties did not reach statistical significance, and it was the correlation with polyphenol content that did.74 The stronger evidence comes from isolating the pigments, where an anthocyanin-rich purple potato extract reduced the after-meal glucose and insulin response in healthy men.110 Pigmented potatoes also lowered oxidative and inflammatory damage in a controlled feeding study.75 Two weeks of purple potatoes improved arterial stiffness in healthy adults.76 Meanwhile, potato consumption in general is associated with increased type 2 diabetes risk in a dose-response meta-analysis of cohorts.77 Same vegetable; the pigments appear to make the difference.

Fruits deliver fiber, water and among the highest antioxidant densities of common foods. In test-tube work, berry polyphenols inhibited alpha-amylase, the enzyme that digests starch.66 Whether that translates into a meaningful effect in a meal has not been shown, but it is one more reason to put berries on oatmeal or into whole-grain porridge rather than reach for a sweetener.

Mushrooms bring beta-glucan fiber, selenium, ergothioneine, and vitamin D when they have been exposed to ultraviolet light. Nuts and seeds bring protein, minerals, essential fatty acids and high antioxidant content; Adventists eating nuts more than four times a week had roughly half the risk of heart attack and of fatal coronary heart disease (fully adjusted relative risks 0.49 and 0.52) compared with those eating them less than once a week.147 Seaweeds bring iodine, minerals and unique soluble fibers, and kombu earns its place in the bean pot below. Herbs and spices are, per gram, the most antioxidant-dense category of ordinary food in the global database; only herbal medicines and supplements score higher.1

Lectins, phytates, and the antinutrient panic

Are lectins and phytates dangerous?

In properly prepared whole foods, no, and several of them appear to be actively beneficial. The full treatment is in the plant toxins and antinutrients post; here is the carbohydrate-relevant summary.

Lectins in raw legumes are genuinely toxic. Phytohaemagglutinin in raw kidney beans will make you very sick. Soaking and boiling destroys it. This is a cooking instruction, not a reason to avoid beans, and legumes are a staple in several of the longest-lived populations studied.

Phytate, also called phytic acid or IP6, is the more interesting case. Its distinctive phosphate configuration lets it chelate iron so effectively that it blocks iron’s ability to catalyze hydroxyl radical formation, which makes it a potent antioxidant. That is its most widely accepted role.47

Does phytate have anticancer properties?

The preclinical evidence is substantial. IP6 induces cell cycle arrest, upregulates the tumor suppressors p21 and p27, promotes differentiation of cancer cells, and enters the intracellular inositol phosphate signaling pool, with inositol acting synergistically.4748 The research was motivated by epidemiology showing that high-IP6 diets, meaning diets rich in cereals and legumes, correlate inversely with colon cancer.

The caveat matters: this evidence is largely in vitro and in animals, plus epidemiology. It is promising, not a treatment claim.

Does phytate block mineral absorption?

Less than the panic suggests. Modern reviews conclude that lifetime animal studies show no negative effect on mineral bioavailability, and that any human effect is heavily context-dependent, meaning it depends on total mineral intake and overall diet.47 People eating varied whole-food plant diets are not turning up deficient because of phytate. The full treatment of phytate, lectins, oxalates and the rest, including the simple soaking, sprouting and vitamin-C tricks that neutralize them, is in the companion post on plant toxins and antinutrients.

The practical part: cheap, easy, and how to prepare it

How cheap are dry legumes?

They are among the least expensive sources of protein and calories available anywhere, they store for months to years, and they cook with almost no skill required. This is not a minor point. A secondary analysis of a randomized trial found that a low-fat vegan diet actually reduced food costs.86 The idea that eating this way is expensive comes from comparing it against the cheapest possible animal foods while shopping at the most expensive possible plant retailer.

Should you soak beans overnight?

It helps. Soaking and then discarding the soak water reduces the oligosaccharides (raffinose and stachyose) that cause gas, reduces phytate, and shortens cooking time. Boiling destroys lectins. A strip of kombu (kelp) or a splash of vinegar in the pot are traditional additions for digestibility; the mechanism is reasonable and the tradition is long, though controlled human trials on those specific additions are limited, so treat them as practical kitchen advice rather than proven clinical interventions.

Does sprouting do anything?

Yes. Sprouting legumes, grains and seeds increases free amino acids and the bioavailability of protein, raises vitamin C and several B vitamins, reduces phytate and enzyme inhibitors, and improves digestibility overall. It also costs nothing but a jar and a few days. Related territory is covered in the raw food post.

Is it better to eat grains and beans whole or ground up?

Whole, and there is a clean experiment on this. Two groups ate the same whole foods in the same amounts, differing only in particle size. One group got muesli, whole seeds and beans on salads. The other got the identical foods blended or milled: porridge, flour, hummus. Both groups were eating whole grains, not refined.

The intact-food group roughly doubled their stool output, and their stool pH dropped, meaning more short-chain fatty acid production.15 Apparently intact seeds recovered from stool turned out on inspection to be partly hollowed out; the bacteria had been feeding on the interiors.16 The fragments left after chewing ferry starch and nutrients down to the colon where blended food does not.

Whole grains are great. Intact whole grains may be better. Groats over flour, beans in the salad over hummus, at least some of the time.

Does cooling cooked starches really help?

Yes, modestly, and this is one where the internet has run ahead of the data.

Here is the mechanism in kitchen terms. Cooking loosens starch up; cooling lets it re-set into a tighter, more crystalline structure that your digestive enzymes can no longer pry apart, a bit like how gravy or cooled rice firms up as it sits. That firmed-up, digestion-resistant fraction is called type 3 resistant starch (the technical name for the re-setting is retrogradation), and it is a real, formally categorized thing, not simply “insoluble fiber.” Cooling cooked white rice for 24 hours and reheating it raised resistant starch about two and a half fold and significantly lowered the glycemic response.11 In people with type 1 diabetes, cooled and reheated rice cut the maximum glucose rise and the glycemic area under the curve substantially.12

But keep the magnitude in perspective. As a proportion of total starch, cooling typically moves resistant starch from roughly 3 percent to about 4 percent.13 Beans start higher: in one study, freshly cooked legumes averaged about 4 percent resistant starch by dry weight, against under 2 percent for cereals and tubers, and repeated cooking and cooling raised it further.14 Pasta salad genuinely beats hot pasta and potato salad genuinely beats a baked potato, but this is a bonus, not a strategy. If you want resistant starch, eat beans.

The glycemic index tables show the same effect in real food: boiled potatoes average a glycemic index of 73, while cooked potatoes refrigerated overnight average 49.155

Does eating your vegetables first actually work?

Remarkably well, and this is one of the easiest things on this page to do. When people with type 2 diabetes ate an identical meal but consumed the vegetables and protein before the carbohydrate, glucose was 28.6 percent lower at 30 minutes, 36.7 percent lower at 60 minutes, and the total glucose area under the curve was 73 percent lower, with lower insulin as well.112

It works in people who are not diabetic too. In people with prediabetes, carbohydrate-last ordering attenuated glucose peaks by more than 40 percent and cut the glucose area under the curve by 38.8 percent.113 Continuous monitoring studies have reproduced the effect in people with entirely normal glucose tolerance.

Same food. Same calories. Same meal. Different order. Salad first, beans and grain after.

Does vinegar lower blood sugar?

Yes, modestly and reliably. A meta-analysis of clinical trials found vinegar taken with a meal significantly reduced post-meal glucose (standardized mean difference −0.60, 95 percent confidence interval −1.08 to −0.11) and insulin (standardized mean difference −1.30, 95 percent confidence interval −1.98 to −0.62), in healthy people as well as those with glucose problems.114 A later umbrella review put the pooled reduction around 14.6 mg/dL.

Typical doses were one to two tablespoons with the meal, and liquid vinegar worked better than tablets. A vinaigrette on the salad you are eating first is doing two jobs at once.

Are green bananas better than ripe ones?

For blood sugar, yes, and the difference is large. Unripe green bananas are packed with RS2 resistant starch, measured at roughly 31 to 49 percent of their starch depending on cultivar, which drops below about 5 percent as the fruit ripens and the starch converts to sugar.115 Total fiber measured by methods that capture resistant starch falls from around 18 grams per 100 grams in unripe fruit to 4 or 5 grams when ripe.

Two caveats. Resistant starch in green bananas is largely destroyed by cooking, so this applies to eating them raw or as green banana flour. And the figures in this paragraph are compositional; the human blood sugar data are in the sugar section below, where green-tipped bananas produced a clearly lower response than ripe ones.

What is the second meal effect?

One of the most useful and least known findings in nutrition. In 1982, David Jenkins fed people a lentil breakfast or a bread breakfast with identical carbohydrate content. The lentil breakfast produced a 71 percent reduction in blood glucose area, which is not surprising. What was surprising is what happened at lunch.

Four hours later, everyone ate the same standard bread lunch. The people who had eaten lentils for breakfast had a 38 percent flatter glucose response to that lunch.4 Jenkins called this “lente,” or slow-release, carbohydrate. Breakfast was still working at lunch.

Is there a third meal effect?

Something like it, through at least two mechanisms.

The first is gastric emptying. In half of the subjects in one study, lentils eaten four to four and a half hours earlier stretched the time it took to empty half of a second meal from roughly 1.5 to 2 hours to more than 3 hours.6 The effect varied between people, and in that study the blood sugar response to the second meal was not different, so this is a digestion-speed finding rather than a blood sugar one.

The second is overnight fermentation. Low-glycemic foods eaten at dinner improve the glycemic response to breakfast the next morning, mediated substantially by colonic fermentation of fiber and resistant starch while you sleep.57 Notably, processed oats and wholemeal bread do not produce this effect; intact barley and rye kernels and legumes do.

There is no established term “third meal effect,” so it is more accurate to say the effect carries from one meal to the next and even overnight to the following morning. Which is remarkable enough.

Green banana to monk fruit: what different sweet foods do in your blood

Carbohydrate quality stops being abstract the moment you line up a few sweet foods side by side. A slightly green banana, a spotty ripe one, a spoonful of maple syrup, a sugar-sweetened drink and a drink sweetened with monk fruit all taste sweet. They do not behave the same way once they are inside you, and the difference mostly comes down to one question: how much of the original food’s structure made it into your mouth.

Does a banana with a little green on the end act differently from a ripe one?

Yes, and by a surprising margin. As a banana ripens, its starch turns into sugar. In a Danish study, bananas described as yellow with some green on the tip had 91 percent of their carbohydrate as starch. Bananas ripened for a week until they showed small brown spots had only 13 percent, with the rest as free sugars.149

That change showed up in the blood. In portions carrying about 20 grams of carbohydrate, the green-tipped bananas scored a glycemic index of 43 and the spotty ripe ones 74, measured against white bread as the reference. Converted to the glucose scale most modern tables use, that is roughly 31 versus 53.149155 Both bananas still produced a smaller blood sugar response than the same amount of carbohydrate from white bread, and insulin responses to all three foods were similar.

Two caveats keep this honest. The participants were 10 people with type 2 diabetes, so healthy people may respond with different magnitudes. And a ripe banana is not a problem food: a glycemic index in the low 50s still counts as low, since the cutoff is 55.155 The international tables point the same way in healthy adults: a raw banana is listed at 47 and a raw overripe banana at 57, from the same Australian lab, though the overripe test used only nine people.155 The point is that ripeness is a dial you can turn. It is the resistant starch story from the green banana section above, now showing up as a measured difference in the blood.

Why does a whole fruit behave differently from its juice?

Because the fiber is doing structural work, not just nutritional work. In a classic experiment, ten healthy adults ate 60 grams of carbohydrate three ways: as whole apples, as apples blended into purée, and as apple juice.150

• Speed. The juice went down more than ten times faster than the whole apples, and the purée nearly three times faster.
• Fullness. Even when all three were consumed over the same length of time, the apples were the most filling, the purée less so, and the juice least of all.
• The rebound. Blood sugar rose to about the same height after all three. The difference came afterward: after juice and purée, blood sugar fell below fasting levels in the second and third hours. After whole apples, it did not.
• Insulin. Juice sipped slowly over the same time it took to eat the apples still produced an insulin peak almost twice as high.

That last pairing is the lesson. The spike was not the difference. The whole apple kept blood sugar steady afterward and asked for far less insulin to do it. Blend the fiber into mush and part of that protection disappears; remove it entirely and nearly all of it does. It is a small study from 1977, but it isolates the fiber matrix more cleanly than almost anything since.

Is maple syrup a better sugar?

Not in the way it is usually marketed. Maple syrup is mostly sucrose, the same molecule as table sugar. In an analysis of 80 pure maple syrup samples from producers in Canada and the United States, sucrose made up between about 52 and 76 percent of the syrup, with only small and variable amounts of free glucose and fructose.151

Maple syrup does test a bit lower than table sugar. The international tables list pure maple syrup at a glycemic index of 54, against an average of 66 across seven tests of plain sucrose.155 That is a real difference, but it is one test of one brand, and at the tables’ standard serving for sugars and syrups, 5 grams of carbohydrate, both carry the same small glycemic load of 3.155

A little maple syrup on your oatmeal is not a crisis, and it tastes wonderful. But it arrives with no fiber and no cellular structure, which puts it in the same category as the apple juice above, not the apple. Treat it as an added sugar, because that is what it is.

What does added sugar actually do?

Table sugar is one glucose unit bonded to one fructose unit, so every spoonful delivers both. The glucose goes into your bloodstream. The fructose is cleared rapidly by the liver, which is why fructose raises blood glucose far less than you would expect from a sugar.149 On a glycemic index chart, that looks like good news. The liver tells a different story.

In a 10-week trial, overweight and obese adults drank beverages sweetened with either glucose or fructose, supplying 25 percent of their daily energy.153 Both groups gained similar weight. But only the fructose group gained visceral fat, the kind packed around the organs, and only the fructose group showed increased fat production in the liver, higher fasting apoB, LDL, small dense LDL and oxidized LDL, and a drop in insulin sensitivity. The glucose group was not untouched either: its fasting triglycerides rose about 10 percent.

Keep the dose in perspective. A quarter of your calories as sweetened drinks is far beyond what anyone gets from eating fruit, where fructose arrives in modest amounts inside a fiber matrix. The study is a verdict on added sugar at high doses, not on apples.

The broader trial evidence points the same way. A meta-analysis of 39 randomized trials found that higher intakes of free sugars raised triglycerides by 0.11 mmol/L, total cholesterol by 0.16 mmol/L and LDL by 0.12 mmol/L compared with lower intakes, and the authors concluded the relationship was independent of changes in body weight.152

Does sugar cause inflammation, and does that mean you should cut carbohydrates?

A 2022 review titled “Excessive intake of sugar: an accomplice of inflammation” sometimes gets shared as a case against carbohydrates, or even for a carnivore diet. It is worth reading what it actually covers. Its subject is excessive sugar, mainly fructose, glucose and high-fructose corn syrup, with sugar-sweetened beverages as a central concern, as a hallmark of the Western diet. It cites the World Health Organization’s recommendation to keep sugar below 10 percent of energy intake and the stricter 5 percent upper limit recommended in England.184

Much of its mechanistic evidence comes from animals, such as rats fed high-fructose diets and mice engineered to lack an enzyme that metabolizes fructose. The human trials it discusses are mixed. One found potentially harmful effects of high-fructose corn syrup on LDL particles, fasting glucose and CRP. But in another, adults who drank beverages sweetened with fructose, glucose or high-fructose corn syrup for three eight-day periods showed no significant change in inflammatory markers, and a 12-week trial of several high-fructose beverages found no significant differences in CRP, IL-1 beta, IL-6 or TNF-alpha.184

What the review does not discuss at all is starch, fiber, whole grains, legumes, or low-carbohydrate, ketogenic or carnivore diets. So it is evidence about too much added sugar, especially in drinks, which is exactly what this section has been saying. It is not evidence against oats, beans or whole fruit, and it offers nothing in favor of replacing them with meat, whose saturated fat has its own role in the insulin resistance story told earlier in this post.

Is monk fruit a free pass?

It is better understood as a neutral pass. Monk fruit extract has no sugar to absorb, and in a randomized crossover trial it behaved exactly that way at first: 30 healthy young men drank a monk fruit-sweetened beverage and their blood sugar barely moved over the next hour, while the same drink sweetened with 65 grams of sucrose produced a large spike.154

Then lunch arrived. After eating, blood sugar climbed higher on the days with the non-sugar sweeteners, monk fruit included, peaking at about twice the rise seen on the sugar day. Across the full three hours, total blood sugar exposure ended up no different (p = 0.960). The men also ate more at lunch after the sweet but empty drink, and made up the missing calories so completely that total daily energy intake did not differ (p = 0.831).

So monk fruit does not raise blood sugar by itself, and swapping it in for sugar means you are not delivering added sugar. But in this study it did not lower total blood sugar or total calories over the day either. It is a reasonable tool for cutting added sugar, not a metabolic advantage in its own right. The caveats: it was a single day, in lean young men only.

What is the pattern across the lineup?

Line them up by how much of the food’s original structure survived:

• Green-tipped banana: starch locked inside an intact fruit and digested slowly. The lowest response.
• Ripe banana: starch converted to sugar, but still inside the fruit’s fiber and cell walls. Still a low glycemic index.
• Maple syrup and table sugar: sugar with no structure at all, nothing to slow absorption, and a fructose load the liver has to process.
• Monk fruit: no sugar and no spike, but no sign of a daily blood sugar or calorie benefit either.

This is why fruit as a category averages a low glycemic index of 51 in the international tables, while sugars and syrups average 58 and range widely.155 The sugar molecule is not the whole story. The package it arrives in is.

If switching feels bad at first

Why does a high-fiber diet cause gas and bloating at first?

Because you are rebuilding an ecosystem. If your microbiome has spent years without much fermentable substrate, the bacterial populations that handle fiber are diminished, and suddenly delivering a large load of it produces gas and discomfort while those populations expand. This is an adaptation period, not a verdict.

What makes it worse: jumping in too fast, not knowing how to prepare legumes (unsoaked, undercooked beans are a common culprit), not drinking enough water, and having no idea what to actually cook.

How should you transition?

Slowly and deliberately. Add fiber gradually rather than all at once. Soak and cook legumes properly and discard the soak water. Start with the easier legumes: red lentils and split peas break down readily and tend to be better tolerated than whole kidney beans. Drink water. Give it a few weeks.

If you are struggling beyond that, or if you have an existing gut condition, this is worth professional help rather than internet guesswork. Look for a physician, registered dietitian, or nutritionist who practices lifestyle medicine, meaning someone who will help you get the diet working rather than telling you to eliminate the foods causing the adaptation.

And to say the important thing again: if you take insulin or oral medication for blood sugar, changes of this magnitude need medical supervision, because they can work fast enough to make your medication dangerous.

Is there anything a plant-based diet does need supplemented?

Yes, and it should be said plainly in any post recommending this way of eating. Vitamin B12 is non-negotiable. It is made by bacteria rather than by plants or animals, and a reliable supplement or reliably fortified foods are required. Vitamin D is the other common addition, and that is not a vegan issue at all, since it runs low across every diet in low-sun conditions. Details are in the B12 post and essential nutrients on a vegan diet.

None of that is a carbohydrate problem. It is worth stating because “well-planned” is doing real work in every claim on this page, and a diet of refined flour, fried food and soda that happens to contain no animal products is a different thing entirely from what is being described here.

Does it get easier?

It becomes invisible. There is a period of learning what to buy, how to cook it, and what you actually like, and then it is simply how you eat. The pressure cooker is worth owning. A bag of lentils thrown in with oat groats makes a breakfast in twenty minutes. Beans keep for years in the pantry.

What comes with it, once it is second nature, is the thing all of this has been about: reduced risk of the chronic diseases that kill most people in wealthy countries, supported at every level of the evidence hierarchy from mechanism through randomized trials to decades-long cohorts.

The goal is not to manage carbohydrates. It is to stop needing to.

Everything in this post points at one destination: becoming someone for whom the carbohydrate question simply does not arise. Not because you avoid them, but because your body handles them the way it was built to.

Diet is the largest single lever, and it is not the only one.

Does sleep affect blood sugar?

Dramatically, and faster than most people expect. In a randomized crossover study, four nights of 4.5 hours in bed reduced insulin sensitivity in fat cells by about 30 percent, measured directly in biopsied tissue, with total-body insulin sensitivity down about 16 percent compared with 8.5 hours in bed.130 Four nights. You can manufacture meaningful insulin resistance in under a week with a bad sleep schedule, and then blame it on bread.

Does stress raise blood sugar?

Yes, through a well-understood route. Cortisol drives the liver to manufacture glucose through gluconeogenesis and simultaneously antagonizes insulin’s action at the tissues. That combination raises blood glucose, which is adaptive during an actual emergency and corrosive when the emergency is your inbox.

Do environmental exposures matter?

More than most nutrition writing admits. Serum levels of persistent organic pollutants show a striking dose-response relationship with diabetes prevalence.138 And the Global Burden of Disease analysis concluded that in 2019, roughly a fifth of the global burden of type 2 diabetes was attributable to fine particulate air pollution.139 These are observational and confounding is possible, but the signal is large and consistent.

How much does the whole package actually buy you?

This is the number to end on. Following 78,865 women and 44,354 men, researchers scored five low-risk factors: never smoking, a healthy body weight, at least 30 minutes of daily moderate activity, moderate alcohol intake, and a diet in the top 40 percent for quality. People with all five, compared with none, had:131

• All-cause mortality hazard ratio 0.26 (95 percent confidence interval 0.22 to 0.31)
• Cardiovascular mortality hazard ratio 0.18
• Cancer mortality hazard ratio 0.35
• 14.0 additional years of life expectancy at age 50 for women and 12.2 years for men

The population attributable risk of not following these habits was 60.7 percent for all-cause mortality and 71.7 percent for cardiovascular mortality.

And for diabetes specifically, the number is even starker. In 84,941 women followed for 16 years, 91 percent of type 2 diabetes cases were attributable to habits that could have been different (95 percent confidence interval 83 to 95), and the risk reduction was similar whether or not women had a family history of diabetes.132 The low-risk group had a relative risk of 0.09.

Nine out of ten cases. Not genetic destiny, and not the lentils.

So what does the finished picture look like?

A person eating a varied whole-food plant diet, who moves their body most days and especially after meals, who carries some muscle, who sleeps enough, who has some handle on stress, who has people in their life, and who is not marinating in avoidable toxins. That person can eat a bowl of oatmeal, or a plate of beans and rice, or a mango, and their body will do exactly what it evolved to do with it.

They do not count carbohydrates. They do not need to. That is the whole objective, and it is the opposite of what the carbohydrate panic sells, which is lifelong vigilance about a macronutrient whose whole-food forms were never the problem.

The bottom line

Carbohydrate is not a food, it is a chemistry. The category contains the most protective foods available to humans and some of the most harmful, and collapsing them into one word has done extraordinary damage.

In whole plant foods, the carbohydrate never comes alone. It comes with the fiber that governs its absorption and feeds the bacteria that fuel your colon. It comes with the antioxidants that neutralize the free radicals produced by burning it. It comes with the nitrates that open your arteries, the minerals, the vitamins, all nine essential amino acids, and the essential fatty acids. Try to recreate the benefit with isolated stand-ins, and the pieces do not work as well as the bean did.

The problem was never the carbohydrate. It was the refining that stripped the package, and the fat that a calorie surplus, especially one heavy in saturated fat and added sugar, packed inside our cells, where it jammed the signal between insulin and the door glucose was trying to walk through. Fix those two things, and a bowl of oatmeal goes back to being what it always was: food, not a threat.

How this post was checked

Figures in this post were checked against the papers they are cited to, using the full text wherever it was available. Where a figure rests on an abstract, a summary, or thinner evidence, the text or the reference entry says so. A few background statements in older sections are still being sourced, and two older references, on bean antioxidant rankings and on phytate and cancer, have not yet been rechecked against their full text.

Where a study was funded by industry or an advocacy group, or its authors disclosed ties to one, the reference entry says so. Those findings are included when the evidence supports them, and the disclosure lets you weigh them yourself.

Where good studies disagree, the post says so rather than choosing a side: for example, on whether whole grains change the gut microbiome, and on how large the mortality benefit of legumes is. Some summary figures are calculations by the study authors rather than measured results, and the text labels them that way.

If you spot an error, please let us know. Corrections are made in the text and the date below is updated.

References

Links open in a new window.

1. Carlsen MH, Halvorsen BL, Holte K, et al. The total antioxidant content of more than 3100 foods, beverages, spices, herbs and supplements used worldwide. Nutr J. 2010;9:3.
2. Institute of Medicine, Food and Nutrition Board. Dietary Reference Intakes for Energy, Carbohydrate, Fiber, Fat, Fatty Acids, Cholesterol, Protein, and Amino Acids. National Academies Press; 2005.
3. Reverri EJ, Randolph JM, Steinberg FM, et al. Black beans, fiber, and antioxidant capacity pilot study: examination of whole foods vs. functional components on postprandial metabolic, oxidative stress, and inflammation in adults with metabolic syndrome. Nutrients. 2015;7(8):6139-6154.
4. Jenkins DJ, Wolever TM, Taylor RH, et al. Slow release dietary carbohydrate improves second meal tolerance. Am J Clin Nutr. 1982;35(6):1339-1346.
5. Wolever TM, Jenkins DJ, Ocana AM, Rao VA, Collier GR. Second-meal effect: low-glycemic-index foods eaten at dinner improve subsequent breakfast glycemic response. Am J Clin Nutr. 1988;48(4):1041-1047.
6. Lin HC, Moller NA, Wolinsky MM, Kim BH, Doty JE, Meyer JH. Sustained slowing effect of lentils on gastric emptying of solids in humans and dogs. Gastroenterology. 1992;102(3):787-792.
7. Higgins JA. Whole grains, legumes, and the subsequent meal effect: implications for blood glucose control and the role of fermentation. J Nutr Metab. 2012;2012:829238.
8. Lundberg JO, Weitzberg E, Gladwin MT. The nitrate-nitrite-nitric oxide pathway in physiology and therapeutics. Nat Rev Drug Discov. 2008;7(2):156-167.
9. Goh CE, Trinh P, Colombo PC, et al. Association between nitrate-reducing oral bacteria and cardiometabolic outcomes: results from ORIGINS. J Am Heart Assoc. 2019;8(23):e013324.
10. Bondonno CP, Croft KD, Hodgson JM. Dietary nitrate, nitric oxide, and cardiovascular health. Crit Rev Food Sci Nutr. 2016;56(12):2036-2052.
11. Sonia S, Witjaksono F, Ridwan R. Effect of cooling of cooked white rice on resistant starch content and glycemic response. Asia Pac J Clin Nutr. 2015;24(4):620-625.
12. Strozyk S, Rogowicz-Frontczak A, Pilacinski S, et al. Influence of resistant starch resulting from the cooling of rice on postprandial glycemia in type 1 diabetes. Nutr Diabetes. 2022;12(1):21.
13. Raigond P, Ezekiel R, Raigond B. Resistant starch in food: a review. J Sci Food Agric. 2015;95(10):1968-1978.
14. Yadav BS, Sharma A, Yadav RB. Studies on effect of multiple heating/cooling cycles on the resistant starch formation in cereals, legumes and tubers. Int J Food Sci Nutr. 2009;60(Suppl 4):258-272.
15. Hovey AL, Jones GP, Devereux HM, Walker KZ. Whole cereal and legume seeds increase faecal short chain fatty acids compared to ground seeds. Asia Pac J Clin Nutr. 2003;12(4):477-482.
16. Birkett AM, Jones GP, de Silva AM, Young GP, Muir JG. Dietary intake and faecal excretion of carbohydrate by Australians: importance of achieving stool weights greater than 150 g. Eur J Clin Nutr. 1997;51(9):625-632.
17. Walker AR, Walker BF, Walker AJ. Faecal pH, dietary fibre intake, and proneness to colon cancer in four South African populations. Br J Cancer. 1986;53(4):489-495.
18. Ahmed R, Segal I, Hassan H. Fermentation of dietary starch in humans. Am J Gastroenterol. 2000;95(4):1017-1020.
19. Cassidy A, Bingham SA, Cummings JH. Starch intake and colorectal cancer risk: an international comparison. Br J Cancer. 1994;69(5):937-942.
20. 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. Am J Gastroenterol. 1999;94(5):1373-1380.
21. O’Keefe SJ, Chung D, Mahmoud N, et al. Why do African Americans get more colon cancer than Native Africans? J Nutr. 2007;137(1 Suppl):175S-182S.
22. 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. Nutr Cancer. 1997;29(1):13-23.
23. Archer S, Meng S, Wu J, Johnson J, Tang R, Hodin R. Butyrate inhibits colon carcinoma cell growth through two distinct pathways. Surgery. 1998;124(2):248-253.
24. McBain JA, Eastman A, Nobel CS, Mueller GC. Apoptotic death in adenocarcinoma cell lines induced by butyrate and other histone deacetylase inhibitors. Biochem Pharmacol. 1997;53(9):1357-1368.
25. Louis P, Hold GL, Flint HJ. The gut microbiota, bacterial metabolites and colorectal cancer. Nat Rev Microbiol. 2014;12(10):661-672.
26. Duncan SH, Belenguer A, Holtrop G, Johnstone AM, Flint HJ, Lobley GE. Reduced dietary intake of carbohydrates by obese subjects results in decreased concentrations of butyrate and butyrate-producing bacteria in feces. Appl Environ Microbiol. 2007;73(4):1073-1078.
27. O’Riordan KJ, Collins MK, Moloney GM, et al. Short chain fatty acids: microbial metabolites for gut-brain axis signalling. Mol Cell Endocrinol. 2022;546:111572.
28. Fagundes RR, Belt SC, Bakker BM, Dijkstra G, Harmsen HJM, Faber KN. Beyond butyrate: microbial fiber metabolism supporting colonic epithelial homeostasis. Trends Microbiol. 2024;32(2):178-189.
29. Burn J, Bishop DT, Chapman PD, et al. A randomized placebo-controlled prevention trial of aspirin and/or resistant starch in young people with familial adenomatous polyposis. Cancer Prev Res. 2011;4(5):655-665.
30. Mathers JC, Movahedi M, Macrae F, et al. Long-term effect of resistant starch on cancer risk in carriers of hereditary colorectal cancer: an analysis from the CAPP2 randomised controlled trial. Lancet Oncol. 2012;13(12):1242-1249.
31. Martínez I, Lattimer JM, Hubach KL, et al. Gut microbiome composition is linked to whole grain-induced immunological improvements. ISME J. 2013;7(2):269-280.
32. Cotillard A, Kennedy SP, Kong LC, et al. Dietary intervention impact on gut microbial gene richness. Nature. 2013;500(7464):585-588.
33. Le Chatelier E, Nielsen T, Qin J, et al. Richness of human gut microbiome correlates with metabolic markers. Nature. 2013;500(7464):541-546.
34. Trowell H. Ischemic heart disease and dietary fiber. Am J Clin Nutr. 1972;25(9):926-932.
35. Erkkilä AT, Herrington DM, Mozaffarian D, Lichtenstein AH. Cereal fiber and whole-grain intake are associated with reduced progression of coronary-artery atherosclerosis in postmenopausal women with coronary artery disease. Am Heart J. 2005;150(1):94-101.
36. Mobley AR, Jones JM, Rodriguez J, Slavin J, Zelman KM. Identifying practical solutions to meet America’s fiber needs: proceedings from the Food & Fiber Summit. Nutrients. 2014;6(7):2540-2551.
37. Ello-Martin JA, Roe LS, Ledikwe JH, Beach AM, Rolls BJ. Dietary energy density in the treatment of obesity: a year-long trial comparing 2 weight-loss diets. Am J Clin Nutr. 2007;85(6):1465-1477.
38. Mishra S, Xu J, Agarwal U, Gonzales J, Levin S, Barnard ND. A multicenter randomized controlled trial of a plant-based nutrition program to reduce body weight and cardiovascular risk in the corporate setting: the GEICO study. Eur J Clin Nutr. 2013;67(7):718-724.
39. Herreman L, Nommensen P, Pennings B, Laus MC. Comprehensive overview of the quality of plant- and animal-sourced proteins based on the digestible indispensable amino acid score. Food Sci Nutr. 2020;8(10):5379-5391.
40. Sofi SA, Ahmed N, Farooq A, et al. Nutritional and bioactive characteristics of buckwheat, and its potential for developing gluten-free products: an updated overview. Food Sci Nutr. 2023;11(5):2256-2276.
41. Wu X, Beecher GR, Holden JM, Haytowitz DB, Gebhardt SE, Prior RL. Lipophilic and hydrophilic antioxidant capacities of common foods in the United States. J Agric Food Chem. 2004;52(12):4026-4037.
42. Zhao Y, Du SK, Wang H, Cai M. In vitro antioxidant activity of extracts from common legumes. Food Chem. 2014;152:462-466.
43. Xu BJ, Yuan SH, Chang SK. Comparative studies on the antioxidant activities of nine common food legumes against copper-induced human low-density lipoprotein oxidation in vitro. J Food Sci. 2007;72(7):S522-S527.
44. Orlich MJ, Singh PN, Sabaté J, et al. Vegetarian dietary patterns and mortality in Adventist Health Study 2. JAMA Intern Med. 2013;173(13):1230-1238.
45. Fraser GE, Shavlik DJ. Ten years of life: is it a matter of choice? Arch Intern Med. 2001;161(13):1645-1652.
46. 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 Pac J Clin Nutr. 2004;13(2):217-220.
47. Vucenik I, Shamsuddin AM. Protection against cancer by dietary IP6 and inositol. Nutr Cancer. 2006;55(2):109-125.
48. Shamsuddin AM. Anti-cancer function of phytic acid. Int J Food Sci Technol. 2002;37(7):769-782.
49. Kahleova H, Petersen KF, Shulman GI, et al. Effect of a low-fat vegan diet on body weight, insulin sensitivity, postprandial metabolism, and intramyocellular and hepatocellular lipid levels in overweight adults: a randomized clinical trial. JAMA Netw Open. 2020;3(11):e2025454. Corrected January, February and May 2021; figures here are from the corrected version. The corrected funder statement notes that the Physicians Committee for Responsible Medicine employees were involved in the study’s design, conduct and analysis.
50. Kitessa SM, Abeywardena MY. Lipid-induced insulin resistance in skeletal muscle: the chase for the culprit goes from total intramuscular fat to lipid intermediates, and finally to species of lipid intermediates. Nutrients. 2016;8(8):466.
51. Himsworth HP. Dietetic factors influencing the glucose tolerance and the activity of insulin. J Physiol. 1934;81(1):29-48.
52. Himsworth HP. Diabetes mellitus: its differentiation into insulin-sensitive and insulin-insensitive types. Lancet. 1936.
53. Barnard ND, Cohen J, Jenkins DJ, et al. A low-fat vegan diet improves glycemic control and cardiovascular risk factors in a randomized clinical trial in individuals with type 2 diabetes. Diabetes Care. 2006;29(8):1777-1783. Funded by the National Institute of Diabetes and Digestive and Kidney Diseases (R01 DK059362).
54. Barnard ND, Cohen J, Jenkins DJ, et al. A low-fat vegan diet and a conventional diabetes diet in the treatment of type 2 diabetes: a randomized, controlled, 74-wk clinical trial. Am J Clin Nutr. 2009;89(5):1588S-1596S. Funded by the National Institute of Diabetes and Digestive and Kidney Diseases (R01 DK059362).
55. Kahleova H, Tura A, Hill M, Holubkov R, Barnard ND. A plant-based dietary intervention improves beta-cell function and insulin resistance in overweight adults: a 16-week randomized clinical trial. Nutrients. 2018;10(2):189.
56. Porrata C, Sánchez J, Correa V, et al. Ma-Pi 2 macrobiotic diet intervention in adults with type 2 diabetes mellitus. MEDICC Rev. 2009;11(4):29-35.
57. Soare A, Khazrai YM, Del Toro R, et al. The effect of the macrobiotic Ma-Pi 2 diet vs. the recommended diet in the management of type 2 diabetes: the randomized controlled MADIAB trial. Nutr Metab (Lond). 2014;11:39.
58. Porrata-Maury C, Hernández-Triana M, Ruiz-Álvarez V, et al. Ma-Pi 2 macrobiotic diet and type 2 diabetes mellitus: pooled analysis of short-term intervention studies. Diabetes Metab Res Rev. 2014;30(Suppl 1):55-66.
59. Lammert A, Kratzsch J, Selhorst J, et al. Clinical benefit of a short term dietary oatmeal intervention in patients with type 2 diabetes and severe insulin resistance: a pilot study. Exp Clin Endocrinol Diabetes. 2008;116(2):132-134.
60. Delgado G, Kleber ME, Krämer BK, et al. Dietary intervention with oatmeal in patients with uncontrolled type 2 diabetes mellitus: a crossover study. Exp Clin Endocrinol Diabetes. 2019;127(9):623-629.
61. Storz MA, Helle P. Oatmeal interventions in severe insulin resistance on the intensive care unit: a case report. Complement Ther Med. 2019;46:69-72.
62. Storz MA, Küster O. Hypocaloric, plant-based oatmeal interventions in the treatment of poorly-controlled type 2 diabetes: a review. Nutr Health. 2019;25(4):281-290.
63. Lean ME, Leslie WS, Barnes AC, et al. Primary care-led weight management for remission of type 2 diabetes (DiRECT): an open-label, cluster-randomised trial. Lancet. 2018;391(10120):541-551.
64. Defining metabolic flexibility in hair follicle stem cell induced squamous cell carcinoma. Sci Adv. 2024;10(38):eadn2806.
65. Smith B, Schafer XL, Ambeskovic A, Spencer CM, Land H, Munger J. Addiction to coupling of the Warburg effect with glutamine catabolism in cancer cells. Cell Rep. 2016;17(3):821-836.
66. Grussu D, Stewart D, McDougall GJ. Berry polyphenols inhibit alpha-amylase in vitro: identifying active components in rowanberry and raspberry. J Agric Food Chem. 2011;59(6):2324-2331.
67. Sofi F, Whittaker A, Gori AM, et al. Effect of Triticum turgidum subsp. turanicum wheat on irritable bowel syndrome: a double-blinded randomised dietary intervention trial. Br J Nutr. 2014;111(11):1992-1999.
68. Dinu M, Whittaker A, Pagliai G, et al. A khorasan wheat-based replacement diet improves risk profile of patients with nonalcoholic fatty liver disease (NAFLD): a randomized clinical trial. J Am Coll Nutr. 2018;37(6):508-514.
69. Whittaker A, Dinu M, Cesari F, et al. A khorasan wheat-based replacement diet improves risk profile of patients with type 2 diabetes mellitus (T2DM): a randomized crossover trial. Eur J Nutr. 2017;56(3):1191-1200.
70. Whittaker A, Sofi F, Luisi ML, et al. An organic khorasan wheat-based replacement diet improves risk profile of patients with acute coronary syndrome: a randomized crossover trial. Nutrients. 2015;7(5):3401-3415.
71. Cicero AFG, Fogacci F, Veronesi M, et al. Short-term hemodynamic effects of modern wheat products substitution in diet with ancient wheat products: a cross-over, randomized clinical trial. Nutrients. 2018;10(11):1666.
72. Navarro-Perez D, Radcliffe J, Tierney A, Jois M. Quinoa seed lowers serum triglycerides in overweight and obese subjects: a dose-response randomized controlled clinical trial. Curr Dev Nutr. 2017;1(9):e001321.
73. Ballester-Sánchez J, Gil JV, Haros CM, Fernández-Espinar MT. Effect of incorporating white, red or black quinoa flours on free and bound polyphenol content, antioxidant activity and colour of bread. Plant Foods Hum Nutr. 2019;74(2):185-191.
74. Ramdath DD, Padhi E, Hawke A, Sivaramalingam T, Tsao R. The glycemic index of pigmented potatoes is related to their polyphenol content. Food Funct. 2014;5(5):909-915.
75. Kaspar KL, Park JS, Brown CR, Mathison BD, Navarre DA, Chew BP. Pigmented potato consumption alters oxidative stress and inflammatory damage in men. J Nutr. 2011;141(1):108-111.
76. Tsang C, Smail NF, Almoosawi S, McDougall GJM, Al-Dujaili EAS. Antioxidant rich potato improves arterial stiffness in healthy adults. Plant Foods Hum Nutr. 2018;73(3):203-208.
77. Bidel Z, Teymoori F, Davari SJ, Nazarzadeh M. Potato consumption and risk of type 2 diabetes: a dose-response meta-analysis of cohort studies. Clin Nutr ESPEN. 2018;27:86-91.
78. Fernando WM, Hill JE, Zello GA, Tyler RT, Dahl WJ, Van Kessel AG. Diets supplemented with chickpea or its main oligosaccharide component raffinose modify faecal microbial composition in healthy adults. Benef Microbes. 2010;1(2):197-207.
79. Hendrie HC, Ogunniyi A, Hall KS, et al. Incidence of dementia and Alzheimer disease in 2 communities: Yoruba residing in Ibadan, Nigeria, and African Americans residing in Indianapolis, Indiana. JAMA. 2001;285(6):739-747.
80. White L, Petrovitch H, Ross GW, et al. Prevalence of dementia in older Japanese-American men in Hawaii: the Honolulu-Asia Aging Study. JAMA. 1996;276(12):955-960.
81. Grant WB. Trends in diet and Alzheimer’s disease during the nutrition transition in Japan and developing countries. J Alzheimers Dis. 2014;38(3):611-620.
82. Giem P, Beeson WL, Fraser GE. The incidence of dementia and intake of animal products: preliminary findings from the Adventist Health Study. Neuroepidemiology. 1993;12(1):28-36.
83. Chandra V, Ganguli M, Pandav R, Johnston J, Belle S, DeKosky ST. Prevalence of Alzheimer’s disease and other dementias in rural India: the Indo-US study. Neurology. 1998;51(4):1000-1008.
84. Galvin JE. Pass the grain; spare the brain. Neurology. 2007;69(11):1072-1073.
85. GBD 2017 Diet Collaborators. Health effects of dietary risks in 195 countries, 1990-2017: a systematic analysis for the Global Burden of Disease Study 2017. Lancet. 2019;393(10184):1958-1972.
86. Kahleova H, Sutton M, Maracine C, et al. Vegan diet and food costs among adults with overweight: a secondary analysis of a randomized clinical trial. JAMA Netw Open. 2023;6(9):e2332106.
87. Cheng A. Shaping a sustainable food future by rediscovering long-forgotten ancient grains. Plant Sci. 2018;269:136-142.
88. Abdel-Aal el-SM, Young JC, Rabalski I, Hucl P, Fregeau-Reid J. Identification and quantification of seed carotenoids in selected wheat species. J Agric Food Chem. 2007;55(3):787-794.
89. Valli V, Taccari A, Di Nunzio M, Danesi F, Bordoni A. Health benefits of ancient grains: comparison among bread made with ancient, heritage and modern grain flours in human cultured cells. Food Res Int. 2018;107:206-215.
90. Sonnenburg ED, Sonnenburg JL. Starving our microbial self: the deleterious consequences of a diet deficient in microbiota-accessible carbohydrates. Cell Metab. 2014;20(5):779-786.
91. Norgan AP, Lewis JT, Faulk DL, Chandran R, Dharkar DD, Pritt BS. Pseudoparasitic appearance of undigested quinoa. Gastrointest Endosc. 2014;80(5):906-907.
92. Hamilton AER, Sebastian AA, Harris CA. Quick eating caused plastic bread bag clip ingestion and severe gastrointestinal bleed in young adult: a case report. ANZ J Surg. 2019;89(6):E256-E258.
93. 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.
94. Fung TT, van Dam RM, Hankinson SE, Stampfer M, Willett WC, Hu FB. Low-carbohydrate diets and all-cause and cause-specific mortality: two cohort studies. Ann Intern Med. 2010;153(5):289-298.
95. Seidelmann SB, Claggett B, Cheng S, et al. Dietary carbohydrate intake and mortality: a prospective cohort study and meta-analysis. Lancet Public Health. 2018;3(9):e419-e428.
96. Shan Z, Guo Y, Hu FB, Liu L, Qi Q. Association of low-carbohydrate and low-fat diets with mortality among US adults. JAMA Intern Med. 2020;180(4):513-523.
97. Gao JW, Hao QY, Zhang HF, et al. Low-carbohydrate diet score and coronary artery calcium progression: results from the CARDIA study. Arterioscler Thromb Vasc Biol. 2021;41(1):491-500.
98. Hall KD, Guo J, Courville AB, et al. Effect of a plant-based, low-fat diet versus an animal-based, ketogenic diet on ad libitum energy intake. Nat Med. 2021;27(2):344-353.
99. Gardner CD, Trepanowski JF, Del Gobbo LC, et al. Effect of low-fat vs low-carbohydrate diet on 12-month weight loss in overweight adults: the DIETFITS randomized clinical trial. JAMA. 2018;319(7):667-679.
100. Gardner CD, Vadiveloo MK, Petersen KS, et al. Popular dietary patterns: alignment with American Heart Association 2021 dietary guidance: a scientific statement from the American Heart Association. Circulation. 2023;147(22):1715-1730.
101. Patikorn C, Saidoung P, Pham T, et al. Effects of ketogenic diet on health outcomes: an umbrella review of meta-analyses of randomized clinical trials. BMC Med. 2023;21:196.
102. Tonstad S, Stewart K, Oda K, Batech M, Herring RP, Fraser GE. Vegetarian diets and incidence of diabetes in the Adventist Health Study-2. Nutr Metab Cardiovasc Dis. 2013;23(4):292-299.
103. Hooper L, Martin N, Jimoh OF, Kirk C, Foster E, Abdelhamid AS. Reduction in saturated fat intake for cardiovascular disease. Cochrane Database Syst Rev. 2020;(8):CD011737.
104. Katz DL, Doughty KN, Geagan K, Jenkins DA, Gardner CD. Perspective: the public health case for modernizing the definition of protein quality. Adv Nutr. 2019;10(5):755-764.
105. Lousuebsakul-Matthews V, Thorpe DL, Knutsen R, Beeson WL, Fraser GE, Knutsen SF. Legumes and meat analogues consumption are associated with hip fracture risk independently of meat intake among Caucasian men and women: the Adventist Health Study-2. Public Health Nutr. 2014;17(10):2333-2343.
106. Satija A, Bhupathiraju SN, Spiegelman D, et al. Healthful and unhealthful plant-based diets and the risk of coronary heart disease in U.S. adults. J Am Coll Cardiol. 2017;70(4):411-422.
107. Kapil V, Haydar SMA, Pearl V, Lundberg JO, Weitzberg E, Ahluwalia A. Physiological role for nitrate-reducing oral bacteria in blood pressure control. Free Radic Biol Med. 2013;55:93-100.
108. Bondonno CP, Liu AH, Croft KD, et al. Antibacterial mouthwash blunts oral nitrate reduction and increases blood pressure in treated hypertensive men and women. Am J Hypertens. 2015;28(5):572-575.
109. Mathers JC, Elliott F, Macrae F, et al. Cancer prevention with resistant starch in Lynch syndrome patients in the CAPP2-randomized placebo controlled trial: planned 10-year follow-up. Cancer Prev Res. 2022;15(9):623-634.
110. Jokioja J, Percival SS, Yang B, Linderborg KM. Anthocyanin-rich extract from purple potatoes decreases postprandial glycemic response and affects inflammation markers in healthy men. Food Chem. 2020;310:125797.
111. Soare A, Del Toro R, Khazrai YM, et al. A 6-month follow-up study of the randomized controlled Ma-Pi macrobiotic dietary intervention (MADIAB trial) in type 2 diabetes. Nutr Diabetes. 2016;6(8):e222.
112. Shukla AP, Iliescu RG, Thomas CE, Aronne LJ. Food order has a significant impact on postprandial glucose and insulin levels. Diabetes Care. 2015;38(7):e98-e99.
113. Shukla AP, Dickison M, Coughlin N, et al. The impact of food order on postprandial glycaemic excursions in prediabetes. Diabetes Obes Metab. 2019;21(2):377-381.
114. Shishehbor F, Mansoori A, Shirani F. Vinegar consumption can attenuate postprandial glucose and insulin responses: a systematic review and meta-analysis of clinical trials. Diabetes Res Clin Pract. 2017;127:1-9.
115. Bi Y, Zhang Y, Jiang H, et al. Bioaccessibility of phenolic compounds, resistant starch, and dietary fibers from Australian green banana during in vitro digestion and colonic fermentation. Molecules. 2024;29(7):1535.
116. Uribarri J, Woodruff S, Goodman S, et al. Advanced glycation end products in foods and a practical guide to their reduction in the diet. J Am Diet Assoc. 2010;110(6):911-916.
117. Ceriello A, Esposito K, Piconi L, et al. Oscillating glucose is more deleterious to endothelial function and oxidative stress than mean glucose in normal and type 2 diabetic patients. Diabetes. 2008;57(5):1349-1354.
118. Monnier L, Mas E, Ginet C, et al. Activation of oxidative stress by acute glucose fluctuations compared with sustained chronic hyperglycemia in patients with type 2 diabetes. JAMA. 2006;295(14):1681-1687.
119. Jenkins DJA, Dehghan M, Mente A, et al. Glycemic index, glycemic load, and cardiovascular disease and mortality. N Engl J Med. 2021;384(14):1312-1322.
120. Shah VN, DuBose SN, Li Z, et al. Continuous glucose monitoring profiles in healthy nondiabetic participants: a multicenter prospective study. J Clin Endocrinol Metab. 2019;104(10):4356-4364.
121. Battelino T, Danne T, Bergenstal RM, et al. Clinical targets for continuous glucose monitoring data interpretation: recommendations from the international consensus on time in range. Diabetes Care. 2019;42(8):1593-1603.
122. Klonoff DC, Wang J, Rodbard D, et al. A glycemia risk index (GRI) of hypoglycemia and hyperglycemia for continuous glucose monitoring validated by clinician ratings. J Diabetes Sci Technol. 2023;17(6):1686-1697.
123. Srikanthan P, Karlamangla AS. Relative muscle mass is inversely associated with insulin resistance and prediabetes: findings from the Third National Health and Nutrition Examination Survey. J Clin Endocrinol Metab. 2011;96(9):2898-2903. A correction was published in J Clin Endocrinol Metab. 2012;97(6):2203.
124. Richter EA, Hargreaves M. Exercise, GLUT4, and skeletal muscle glucose uptake. Physiol Rev. 2013;93(3):993-1017.
125. Buffey AJ, Herring MP, Langley CK, Donnelly AE, Carson BP. The acute effects of interrupting prolonged sitting time in adults with standing and light-intensity walking on biomarkers of cardiometabolic health: a systematic review and meta-analysis. Sports Med. 2022;52(8):1765-1787.
126. Jansson AK, Chan LX, Lubans DR, Duncan MJ, Plotnikoff RC. Effect of resistance training on HbA1c in adults with type 2 diabetes mellitus and the moderating effect of changes in muscular strength: a systematic review and meta-analysis. BMJ Open Diabetes Res Care. 2022;10(2):e002595.
127. Araujo J, Cai J, Stevens J. Prevalence of optimal metabolic health in American adults: National Health and Nutrition Examination Survey 2009-2016. Metab Syndr Relat Disord. 2019;17(1):46-52.
128. Mongraw-Chaffin M, Foster MC, Anderson CAM, et al. Metabolically healthy obesity, transition to metabolic syndrome, and cardiovascular risk. J Am Coll Cardiol. 2018;71(17):1857-1865.
129. Tabak AG, Herder C, Rathmann W, Brunner EJ, Kivimaki M. Prediabetes: a high-risk state for diabetes development. Lancet. 2012;379(9833):2279-2290.
130. Broussard JL, Ehrmann DA, Van Cauter E, Tasali E, Brady MJ. Impaired insulin signaling in human adipocytes after experimental sleep restriction: a randomized, crossover study. Ann Intern Med. 2012;157(8):549-557.
131. Li Y, Pan A, Wang DD, et al. Impact of healthy lifestyle factors on life expectancies in the US population. Circulation. 2018;138(4):345-355.
132. Hu FB, Manson JE, Stampfer MJ, et al. Diet, lifestyle, and the risk of type 2 diabetes mellitus in women. N Engl J Med. 2001;345(11):790-797.
133. Rosqvist F, Iggman D, Kullberg J, et al. Overfeeding polyunsaturated and saturated fat causes distinct effects on liver and visceral fat accumulation in humans. Diabetes. 2014;63(7):2356-2368.
134. Jenkins DJA, Wong JMW, Kendall CWC, et al. Effect of a 6-month vegan low-carbohydrate (“Eco-Atkins”) diet on cardiovascular risk factors and body weight in hyperlipidaemic adults: a randomised controlled trial. BMJ Open. 2014;4(2):e003505.
135. Martin-McGill KJ, Bresnahan R, Levy RG, Cooper PN. Ketogenic diets for drug-resistant epilepsy. Cochrane Database Syst Rev. 2020;6(6):CD001903.
136. Evert AB, Dennison M, Gardner CD, et al. Nutrition therapy for adults with diabetes or prediabetes: a consensus report. Diabetes Care. 2019;42(5):731-754.
137. U.S. Department of Agriculture, Agricultural Research Service. FoodData Central. Values are per 100 g from SR Legacy and Foundation Foods entries (for example, cooked lentils FDC 172421; avocado FDC 171705).
138. Lee DH, Lee IK, Song K, et al. A strong dose-response relation between serum concentrations of persistent organic pollutants and diabetes: results from the National Health and Examination Survey 1999-2002. Diabetes Care. 2006;29(7):1638-1644.
139. Burkart K, Causey K, Cohen AJ, et al. Estimates, trends, and drivers of the global burden of type 2 diabetes attributable to PM2.5 air pollution, 1990-2019. Lancet Planet Health. 2022;6(7):e586-e600.
140. McNeil NI. The contribution of the large intestine to energy supplies in man. Am J Clin Nutr. 1984;39(2):338-342.
141. 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 Prev Res. 2014;7(8):813-823.
142. Ducatman A, Luster M, Fletcher T. Perfluoroalkyl substance excretion: effects of organic anion-inhibiting and resin-binding drugs in a community setting. Environ Toxicol Pharmacol. 2021;85:103650.
143. Bello D, Schlezinger JJ, et al. An oat fiber intervention for reducing PFAS body burden: a pilot study in male C57Bl/6 J mice. Toxicol Appl Pharmacol. 2025;495:117188.
144. Schlezinger JJ, Wolever TMS, et al. Per- and poly-fluoroalkyl substances (PFAS) in circulation in a Canadian population: their association with serum-liver enzyme biomarkers and piloting a novel method to reduce serum-PFAS. Environ Health. 2025;24:10.
145. Dzierlenga MW, Keast DR, Longnecker MP. The concentration of several perfluoroalkyl acids in serum appears to be reduced by dietary fiber. Environ Int. 2021;146:106292.
146. Singh PN, Fraser GE. Dietary risk factors for colon cancer in a low-risk population. Am J Epidemiol. 1998;148(8):761-774.
147. Fraser GE, Sabaté J, Beeson WL, Strahan TM. A possible protective effect of nut consumption on risk of coronary heart disease: the Adventist Health Study. Arch Intern Med. 1992;152(7):1416-1424.
148. Yu T, Robotham JL, Yoon Y. Increased production of reactive oxygen species in hyperglycemic conditions requires dynamic change of mitochondrial morphology. Proc Natl Acad Sci U S A. 2006;103(8):2653-2658.
149. Hermansen K, Rasmussen O, Gregersen S, Larsen S. Influence of ripeness of banana on the blood glucose and insulin response in type 2 diabetic subjects. Diabet Med. 1992;9:739-743.
150. Haber GB, Heaton KW, Murphy D, Burroughs LF. Depletion and disruption of dietary fibre: effects on satiety, plasma-glucose, and serum-insulin. Lancet. 1977 Oct 1:679-682.
151. Stuckel JG, Low NH. The chemical composition of 80 pure maple syrup samples produced in North America. Food Res Int. 1996.
152. Te Morenga LA, Howatson AJ, Jones RM, Mann J. Dietary sugars and cardiometabolic risk: systematic review and meta-analyses of randomized controlled trials of the effects on blood pressure and lipids. Am J Clin Nutr. 2014;100:65-79.
153. Stanhope KL, Schwarz JM, Keim NL, et al. Consuming fructose-sweetened, not glucose-sweetened, beverages increases visceral adiposity and lipids and decreases insulin sensitivity in overweight/obese humans. J Clin Invest. 2009;119(5):1322-1334.
154. Tey SL, Salleh NB, Henry J, Forde CG. Effects of aspartame-, monk fruit-, stevia- and sucrose-sweetened beverages on postprandial glucose, insulin and energy intake. Int J Obes. 2017;41:450-457.
155. Atkinson FS, Brand-Miller JC, Foster-Powell K, Buyken AE, Goletzke J. International tables of glycemic index and glycemic load values 2021: a systematic review. Am J Clin Nutr. 2021;114:1625-1632. Values for individual foods from Online Supplemental Material, Supplemental Table 1.
156. Whitehead A, Beck EJ, Tosh S, Wolever TMS. Cholesterol-lowering effects of oat beta-glucan: a meta-analysis of randomized controlled trials. Am J Clin Nutr. 2014;100:1413-1421.
157. Ha V, Sievenpiper JL, de Souza RJ, et al. Effect of dietary pulse intake on established therapeutic lipid targets for cardiovascular risk reduction: a systematic review and meta-analysis of randomized controlled trials. CMAJ. 2014;186(8):E252-E262.
158. Downie E. The clinical approach to the problem of diabetes mellitus. Br Med J. 1955 Oct 8;(4944):863-867. Banting Lecture, University of Toronto, 1955.
159. Numao S, Kawano H, Endo N, et al. Short-term low carbohydrate/high-fat diet intake increases postprandial plasma glucose and glucagon-like peptide-1 levels during an oral glucose tolerance test in healthy men. Eur J Clin Nutr. 2012;66(8):926-931.
160. Misselwitz B, Butter M, Verbeke K, Fox MR. Update on lactose malabsorption and intolerance: pathogenesis, diagnosis and clinical management. Gut. 2019;68(11):2080-2091.
161. Karamanou M, Protogerou A, Tsoucalas G, Androutsos G, Poulakou-Rebelakou E. Milestones in the history of diabetes mellitus: the main contributors. World J Diabetes. 2016;7(1):1-7.
162. Diabetes Research Connection. The fascinating history of diabetes mellitus. July 21, 2023.
163. Bhattacharya S, Kalra S, Nagendra L, Dutta D. Forty-four years of the UK Prospective Diabetes Study: legacy effect and beyond. touchREVIEWS in Endocrinology. 2025;21(1):2-3. Summarizes Adler AI et al., Lancet. 2024;404:145-155 (UKPDS 91).
164. Murray P, Chune GW, Raghavan VA. Legacy effects from DCCT and UKPDS: what they mean and implications for future diabetes trials. Curr Atheroscler Rep. 2010;12:432-439.
165. Kumar D, Tanwar R. World’s first: stem cell therapy reverses diabetes. Stem Cell Res Ther. 2024;15:487. Letter to the editor summarizing a case first reported in Nature.
166. Adler AI, Coleman RL, Leal J, Whiteley WN, Clarke P, Holman RR. Post-trial monitoring of a randomised controlled trial of intensive glycaemic control in type 2 diabetes extended from 10 years to 24 years (UKPDS 91). Lancet. 2024;404(10448):145-155.
167. Nathan DM; DCCT/EDIC Research Group. The Diabetes Control and Complications Trial/Epidemiology of Diabetes Interventions and Complications Study at 30 years: overview. Diabetes Care. 2014;37(1):9-16.
168. The oatmeal cure in diabetes. Reprinted from Medical Standard, February 1907. PMC8999280.
169. Oatmeal diet for diabetics. November 6, 1909. PMC5202724.
170. Klümpen L, Mantri A, Philipps M, et al. Cholesterol-lowering effects of oats induced by microbially produced phenolic metabolites in metabolic syndrome: a randomized controlled trial. Nat Commun. 2026;17:598.
171. Li Y, Xu S, Zhang X, Yi Z, Cichello S. Skeletal intramyocellular lipid metabolism and insulin resistance. Biophys Rep. 2015;1(2):90-98.
172. Guo Z. Intramyocellular lipid kinetics and insulin resistance. Lipids Health Dis. 2007;6:18.
173. Janssen JAMJL. The pivotal role of the Western diet, hyperinsulinemia, ectopic fat, and diacylglycerol-mediated insulin resistance in type 2 diabetes. Int J Mol Sci. 2025;26(18):9191.
174. Lecerf JM, de Lorgeril M. Dietary cholesterol: from physiology to cardiovascular risk. Br J Nutr. 2011;106(1):6-14.
175. Cummings JH, Pomare EW, Branch WJ, Naylor CP, Macfarlane GT. Short chain fatty acids in human large intestine, portal, hepatic and venous blood. Gut. 1987;28(10):1221-1227.
176. Mariotti F, Gardner CD. Dietary protein and amino acids in vegetarian diets: a review. Nutrients. 2019;11(11):2661.
177. Soh BXP, et al. Evaluation of protein intake and protein quality in New Zealand vegans. PLOS One. 2025;e0314889.
178. Buettner D, Skemp S. Blue Zones: lessons from the world’s longest lived. Am J Lifestyle Med. 2016;10(5):318-321. The authors are affiliated with Blue Zones, LLC.
179. Ho KS, Tan CYM, Mohd Daud MA, Seow-Choen F. Stopping or reducing dietary fiber intake reduces constipation and its associated symptoms. World J Gastroenterol. 2012;18(33):4593-4596.
180. Vatkar A, Kale S, Shyam A, Srivastava S. Understanding the levels of evidence in medical research. J Orthop Case Rep. 2025;15(5):6-9.
181. Wang Y, Kuo B, Berschback M, Huttenhower C, Chan AT, Staller K. Dietary patterns and incident chronic constipation in three prospective cohorts of middle- and older-aged adults. Gastroenterology. 2025;169:1475-1488.
182. Bassotti G, Usai Satta P, Bellini M. Chronic idiopathic constipation in adults: a review on current guidelines and emerging treatment options. Clin Exp Gastroenterol. 2021;14.
183. Rollet M, Bohn T, Vahid F; ORISCAV Working Group. Association between dietary factors and constipation in adults living in Luxembourg and taking part in the ORISCAV-LUX 2 survey. Nutrients. 2022;14(1):122.
184. Ma X, Nan F, Liang H, Shu P, Fan X, Song X, Hou Y, Zhang D. Excessive intake of sugar: an accomplice of inflammation. Front Immunol. 2022;13:988481.
185. Luukkonen PK, Sädevirta S, Zhou Y, et al. Saturated fat is more metabolically harmful for the human liver than unsaturated fat or simple sugars. Diabetes Care. 2018;41(8):1732-1739.
186. Rosqvist F, Kullberg J, Ståhlman M, et al. Overeating saturated fat promotes fatty liver and ceramides compared with polyunsaturated fat: a randomized trial. J Clin Endocrinol Metab. 2019;104(12):6207-6219.
187. Vessby B, Uusitupa M, Hermansen K, et al. Substituting dietary saturated for monounsaturated fat impairs insulin sensitivity in healthy men and women: the KANWU study. Diabetologia. 2001;44(3):312-319. Study foods were supplied by food companies, including MD Foods, Svenska Nestlé and Van den Bergh Foods.
188. Neuenschwander M, Barbaresko J, Pischke CR, et al. Intake of dietary fats and fatty acids and the incidence of type 2 diabetes: a systematic review and dose-response meta-analysis of prospective observational studies. PLoS Med. 2020;17(12):e1003347.
189. Gaeini Z, Bahadoran Z, Mirmiran P. Saturated fatty acid intake and risk of type 2 diabetes: an updated systematic review and dose-response meta-analysis of cohort studies. Adv Nutr. 2022;13(6):2125-2135.
190. Schaefer E, Neuenschwander M, Schiemann T, et al. Fatty acid biomarkers and incidence of type 2 diabetes: a systematic review and dose-response meta-analysis of prospective observational studies. Adv Nutr. 2026;17:100565.
191. Imamura F, Micha R, Wu JHY, et al. Effects of saturated fat, polyunsaturated fat, monounsaturated fat, and carbohydrate on glucose-insulin homeostasis: a systematic review and meta-analysis of randomised controlled feeding trials. PLoS Med. 2016;13(7):e1002087. Two authors received research support from Unilever R&D; one reported consulting fees from food and supplement organizations.
192. Brown TJ, Brainard J, Song F, et al. Omega-3, omega-6, and total dietary polyunsaturated fat for prevention and treatment of type 2 diabetes mellitus: systematic review and meta-analysis of randomised controlled trials. BMJ. 2019;366:l4697.
193. Schwarz JM, Neese RA, Turner S, Dare D, Hellerstein MK. Short-term alterations in carbohydrate energy intake in humans: striking effects on hepatic glucose production, de novo lipogenesis, lipolysis, and whole-body fuel selection. J Clin Invest. 1995;96(6):2735-2743.
194. Hudgins LC, Hellerstein M, Seidman C, Neese R, Diakun J, Hirsch J. Human fatty acid synthesis is stimulated by a eucaloric low fat, high carbohydrate diet. J Clin Invest. 1996;97(9):2081-2091.
195. Hudgins LC, Seidman CE, Diakun J, Hirsch J. Human fatty acid synthesis is reduced after the substitution of dietary starch for sugar. Am J Clin Nutr. 1998;67(4):631-639.
196. Parks EJ, Krauss RM, Christiansen MP, Neese RA, Hellerstein MK. Effects of a low-fat, high-carbohydrate diet on VLDL-triglyceride assembly, production, and clearance. J Clin Invest. 1999;104(8):1087-1096.
197. Schwarz JM, Noworolski SM, Wen MJ, et al. Effect of a high-fructose weight-maintaining diet on lipogenesis and liver fat. J Clin Endocrinol Metab. 2015;100(6):2434-2442.
198. Schwarz JM, Noworolski SM, Erkin-Cakmak A, et al. Effects of dietary fructose restriction on liver fat, de novo lipogenesis, and insulin kinetics in children with obesity. Gastroenterology. 2017;153(3):743-752.
199. Chiu S, Sievenpiper JL, de Souza RJ, et al. Effect of fructose on markers of non-alcoholic fatty liver disease (NAFLD): a systematic review and meta-analysis of controlled feeding trials. Eur J Clin Nutr. 2014;68(4):416-423. One author disclosed research support from The Coca-Cola Company and the Calorie Control Council, among others.
200. Sevastianova K, Santos A, Kotronen A, et al. Effect of short-term carbohydrate overfeeding and long-term weight loss on liver fat in overweight humans. Am J Clin Nutr. 2012;96(4):727-734.
201. Lambert JE, Ramos-Roman MA, Valdez MJ, et al. Weight loss in MASLD restores the balance of liver fatty acid sources. J Clin Invest. 2025;135(9):e174233.
202. Hall KD, Ayuketah A, Brychta R, et al. Ultra-processed diets cause excess calorie intake and weight gain: an inpatient randomized controlled trial of ad libitum food intake. Cell Metab. 2019;30(1):67-77.e3. One author disclosed advisory and consortium ties to food companies.
203. Walhin JP, Richardson JD, Betts JA, Thompson D. Exercise counteracts the effects of short-term overfeeding and reduced physical activity independent of energy imbalance in healthy young men. J Physiol. 2013;591(24):6231-6243.
204. Egli L, Lecoultre V, Theytaz F, et al. Exercise prevents fructose-induced hypertriglyceridemia in healthy young subjects. Diabetes. 2013;62(7):2259-2265.
205. Landry MJ, Ward CP, Cunanan KM, et al. Cardiometabolic effects of omnivorous vs vegan diets in identical twins: a randomized clinical trial. JAMA Netw Open. 2023;6(11):e2344457.
206. Barnard ND, Alwarith J, Rembert E, et al. A Mediterranean diet and low-fat vegan diet to improve body weight and cardiometabolic risk factors: a randomized, cross-over trial. J Am Coll Nutr (now J Am Nutr Assoc). Published online 2021; 2022;41(2):127-139. All authors but two work for the Physicians Committee for Responsible Medicine.
207. Termannsen AD, Clemmensen KKB, Thomsen JM, et al. Effects of vegan diets on cardiometabolic health: a systematic review and meta-analysis of randomized controlled trials. Obes Rev. 2022;23(9):e13462. One author is employed by Novo Nordisk and another has received Novo Nordisk grants.
208. Guest NS, Raj S, Landry MJ, et al. Vegetarian and vegan dietary patterns to treat adult type 2 diabetes: a systematic review and meta-analysis of randomized controlled trials. Adv Nutr. 2024;15(10):100294. Funded by the Academy of Nutrition and Dietetics.
209. Qian F, Liu G, Hu FB, Bhupathiraju SN, Sun Q. Association between plant-based dietary patterns and risk of type 2 diabetes: a systematic review and meta-analysis. JAMA Intern Med. 2019;179(10):1335-1344.
210. Hosseini F, Jayedi A, Khan TA, Shab-Bidar S. Dietary carbohydrate and the risk of type 2 diabetes: an updated systematic review and dose-response meta-analysis of prospective cohort studies. Sci Rep. 2022;12:2491.
211. Schwingshackl L, Schwedhelm C, Hoffmann G, et al. Food groups and risk of all-cause mortality: a systematic review and meta-analysis of prospective studies. Am J Clin Nutr. 2017;105(6):1462-1473.
212. Onni AT, Balakrishna R, Perillo M, et al. Umbrella review of systematic reviews and meta-analyses on consumption of different food groups and risk of all-cause mortality. Adv Nutr. 2025;16:100393.
213. Aune D, Keum N, Giovannucci E, et al. Whole grain consumption and risk of cardiovascular disease, cancer, and all cause and cause specific mortality: systematic review and dose-response meta-analysis of prospective studies. BMJ. 2016;353:i2716.
214. Veronese N, Solmi M, Caruso MG, et al. Dietary fiber and health outcomes: an umbrella review of systematic reviews and meta-analyses. Am J Clin Nutr. 2018;107(3):436-444.
215. Hu Y, Ding M, Sampson L, et al. Intake of whole grain foods and risk of type 2 diabetes: results from three prospective cohort studies. BMJ. 2020;370:m2206.
216. Yao B, Fang H, Xu W, et al. Dietary fiber intake and risk of type 2 diabetes: a dose-response analysis of prospective studies. Eur J Epidemiol. 2014.
217. Aune D, Chan DSM, Lau R, et al. Dietary fibre, whole grains, and risk of colorectal cancer: systematic review and dose-response meta-analysis of prospective studies. BMJ. 2011;343:d6617.
218. Aune D, Sen A, Norat T, Riboli E. Dietary fibre intake and the risk of diverticular disease: a systematic review and meta-analysis of prospective studies. Eur J Nutr. 2020;59(2):421-432.
219. Song M, Wu K, Meyerhardt JA, et al. Fiber intake and survival after colorectal cancer diagnosis. JAMA Oncol. 2018;4(1):71-79. JAMA Oncology later published a notice of omitted conflict-of-interest disclosures covering this article.
220. Neuenschwander M, Stadelmaier J, Eble J, et al. Substitution of animal-based with plant-based foods on cardiometabolic health and all-cause mortality: a systematic review and meta-analysis of prospective studies. BMC Med. 2023;21:404.
221. Reynolds AN, Akerman AP, Mann J. Dietary fibre and whole grains in diabetes management: systematic review and meta-analyses. PLoS Med. 2020;17(3):e1003053.
222. Kelly SAM, Hartley L, Loveman E, et al. Whole grain cereals for the primary or secondary prevention of cardiovascular disease. Cochrane Database Syst Rev. 2017;8:CD005051.
223. Pol K, Christensen R, Bartels EM, et al. Whole grain and body weight changes in apparently healthy adults: a systematic review and meta-analysis of randomized controlled studies. Am J Clin Nutr. 2013;98:872-884.
224. Roager HM, Vogt JK, Kristensen M, et al. Whole grain-rich diet reduces body weight and systemic low-grade inflammation without inducing major changes of the gut microbiome: a randomised cross-over trial. Gut. 2019;68(1):83-93. One author was partly supported by an unrestricted grant from Cereal Partners Worldwide, a Nestlé and General Mills joint venture.
225. Swaminathan S, Dehghan M, Raj JM, et al. Associations of cereal grains intake with cardiovascular disease and mortality across 21 countries in Prospective Urban and Rural Epidemiology study: prospective cohort study. BMJ. 2021;372:m4948.
226. Liu H, Zhu J, Gao R, et al. Estimating effects of whole grain consumption on type 2 diabetes, colorectal cancer and cardiovascular disease: a burden of proof study. Nutr J. 2024;23:49.
227. 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. Adv Nutr. 2023;14(1):64-76.
228. Viguiliouk E, Glenn AJ, Nishi SK, et al. Associations between dietary pulses alone or with other legumes and cardiometabolic disease outcomes: an umbrella review and updated systematic review and meta-analysis of prospective cohort studies. Adv Nutr. 2019;10(Suppl 4):S308-S319. Author disclosures include funding from food industry organizations, among them Pulse Canada.
229. Thorisdottir B, Arnesen EK, Bärebring L, et al. Legume consumption in adults and risk of cardiovascular disease and type 2 diabetes: a systematic review and meta-analysis. Food Nutr Res. 2023;67:9541.
230. Torheim LE, Fadnes LT. Legumes and pulses: a scoping review for Nordic Nutrition Recommendations 2023. Food Nutr Res. 2024;68:10484.
231. Hossain MA, Soni K, Agyei D, Kebede B, Reynolds AN. Legume intakes on cardiometabolic profile and gut microbiome function: systematic review and meta-analyses of randomised controlled trials. Eur J Nutr. 2025;64:60. Figures checked against the published abstract only.
232. Jenkins DJA, Kendall CWC, Augustin LSA, et al. Effect of legumes as part of a low glycemic index diet on glycemic control and cardiovascular risk factors in type 2 diabetes mellitus: a randomized controlled trial. Arch Intern Med. 2012;172(21):1653-1660. Supported by ABIP through PURENet and the Saskatchewan Pulse Growers.
233. Jayalath VH, de Souza RJ, Sievenpiper JL, et al. Effect of dietary pulses on blood pressure: a systematic review and meta-analysis of controlled feeding trials. Am J Hypertens. 2014;27(1):56-64.
234. Kim SJ, de Souza RJ, Choo VL, et al. Effects of dietary pulse consumption on body weight: a systematic review and meta-analysis of randomized controlled trials. Am J Clin Nutr. 2016;103(5):1213-1223.
235. Hafiz MS, Boesch C, Campbell MD, et al. Pulse consumption improves indices of glycemic control in adults with and without type 2 diabetes: a systematic review and meta-analysis of acute and long-term randomized controlled trials. Eur J Nutr. 2022;61:809-824.
236. Blanco Mejia S, Messina M, Li SS, et al. A meta-analysis of 46 studies identified by the FDA demonstrates that soy protein decreases circulating LDL and total cholesterol concentrations in adults. J Nutr. 2019;149(6):968-981.
237. Applegate CC, Rowles JL, Ranard KM, Jeon S, Erdman JW Jr. Soy consumption and the risk of prostate cancer: an updated systematic review and meta-analysis. Nutrients. 2018;10(1):40.
238. Dong JY, Qin LQ. Soy isoflavones consumption and risk of breast cancer incidence or recurrence: a meta-analysis of prospective studies. Breast Cancer Res Treat. 2011;125:315-323.
239. Wei Y, Lv J, Guo Y, et al. Soy intake and breast cancer risk: a prospective study of 300,000 Chinese women and a dose-response meta-analysis. Eur J Epidemiol. 2020;35:567-578.
240. Li N, et al. Soy foods consumption and multiple health outcomes: an umbrella review of meta-analyses. Mol Nutr Food Res. 2026;70:e70255. Conclusions checked against the published abstract only.
241. Jin S, Je Y. Nuts and legumes consumption and risk of colorectal cancer: a systematic review and meta-analysis. Eur J Epidemiol. 2022;37:569-585.
242. Winham DM, Hutchins AM. Perceptions of flatulence from bean consumption among adults in 3 feeding studies. Nutr J. 2011;10:128.
243. Lafiandra D, Riccardi G, Shewry PR. Improving cereal grain carbohydrates for diet and health. J Cereal Sci. 2014;59:312-326.
244. Augustin LSA, Aas AM, Astrup A, et al. Dietary fibre consensus from the International Carbohydrate Quality Consortium (ICQC). Nutrients. 2020;12(9):2553.
245. Stribling P, Ibrahim F. Dietary fibre definition revisited: the case of low molecular weight carbohydrates. Clin Nutr ESPEN. 2023;55:340-356.
246. Singh A, Saini RK, Kumar A, Chawla P, et al. Mushrooms as nutritional powerhouses: a review of their bioactive compounds, health benefits, and value-added products. Foods. 2025;14(5):741.
247. Helnæs A, Kyrø C, Andersen I, et al. Intake of whole grains is associated with lower risk of myocardial infarction: the Danish Diet, Cancer and Health Cohort. Am J Clin Nutr. 2016;103(4):999-1007.
248. Xu D, Feng M, Chu Y, et al. The prebiotic effects of oats on blood lipids, gut microbiota, and short-chain fatty acids in mildly hypercholesterolemic subjects compared with rice: a randomized, controlled trial. Front Immunol. 2021;12:787797. Several authors are employees of PepsiCo.
249. Thongoun P, Pavadhgul P, Bumrungpert A, et al. Effect of oat consumption on lipid profiles in hypercholesterolemic adults. J Med Assoc Thai. 2013;96(Suppl 5):S25-S32. Two authors are affiliated with PepsiCo (PepsiCo Asia Services and PepsiCo R&D).
250. Zhang J, Li L, Song P, et al. Randomized controlled trial of oatmeal consumption versus noodle consumption on blood lipids of urban Chinese adults with hypercholesterolemia. Nutr J. 2012;11:54. The oatmeal was Quaker Instant Oatmeal; two authors are PepsiCo employees.
251. Hu Y, Willett WC, Manson JE, et al. Intake of whole grain foods and risk of coronary heart disease in US men and women. BMC Med. 2022;20:192. Funded by the National Institutes of Health; author disclosures include research support from the California Walnut Commission.
252. Sweeney JS. Dietary factors that influence the dextrose tolerance test: a preliminary study. Arch Intern Med. 1927.
253. Randle PJ, Garland PB, Hales CN, Newsholme EA. The glucose fatty-acid cycle: its role in insulin sensitivity and the metabolic disturbances of diabetes mellitus. Lancet. 1963.
254. Kelley DE, Mokan M, Simoneau JA, Mandarino LJ. Interaction between glucose and free fatty acid metabolism in human skeletal muscle. J Clin Invest. 1993;92(1):91-98.
255. Samuel VT, Petersen KF, Shulman GI. Lipid-induced insulin resistance: unravelling the mechanism. Lancet. 2010;375(9733):2267-2277.
256. Roden M, Price TB, Perseghin G, et al. Mechanism of free fatty acid-induced insulin resistance in humans. J Clin Invest. 1996;97(12):2859-2865.
257. Dresner A, Laurent D, Marcucci M, et al. Effects of free fatty acids on glucose transport and IRS-1-associated phosphatidylinositol 3-kinase activity. J Clin Invest. 1999;103(2):253-259.
258. Roden M, Krssak M, Stingl H, et al. Rapid impairment of skeletal muscle glucose transport/phosphorylation by free fatty acids in humans. Diabetes. 1999;48(2):358-364.
259. Itani SI, Ruderman NB, Schmieder F, Boden G. Lipid-induced insulin resistance in human muscle is associated with changes in diacylglycerol, protein kinase C, and IκB-α. Diabetes. 2002;51(7):2005-2011.
260. Kelley DE, Mandarino LJ. Fuel selection in human skeletal muscle in insulin resistance: a reexamination. Diabetes. 2000;49(5):677-683.
261. Santomauro AT, Boden G, Silva ME, et al. Overnight lowering of free fatty acids with Acipimox improves insulin resistance and glucose tolerance in obese diabetic and nondiabetic subjects. Diabetes. 1999;48(9):1836-1841.
262. Jornayvaz FR, Samuel VT, Shulman GI. The role of muscle insulin resistance in the pathogenesis of atherogenic dyslipidemia and nonalcoholic fatty liver disease associated with the metabolic syndrome. Annu Rev Nutr. 2010;30:273-290.
263. Bisschop PH, de Metz J, Ackermans MT, et al. Dietary fat content alters insulin-mediated glucose metabolism in healthy men. Am J Clin Nutr. 2001;73(3):554-559.
264. Lundsgaard AM, Holm JB, Sjøberg KA, et al. Mechanisms preserving insulin action during high dietary fat intake. Cell Metab. 2019;29(1):50-63. Supported in part by the Novo Nordisk Foundation; the authors declare no competing interests.

Last edited: October 6, 2026 at 8:02 PM ET