Kurt G. Harris MD

PāNu means paleonutrition. The "paleo" here signifies "old" and not necessarily paleolithic. The PāNu approach to nutrition is grounded on clinical medicine and basic sciences disciplined by knowledge of evolutionary biology and paleoanthropology. The best evidence from multiple disciplines supports eating a pastoral (animal-based) diet rather than a grain-based agricultural one, while avoiding what I call the neolithic agents of disease - wheat, excess fructose and excess linoleic acid.

Support PāNu

PāNu is ad-free, completely independent and has no outside sponsorship. If you value PāNu, now you can support it. Read this for more information.

In addition to buying from the book list, you can also support PāNu by making all of your Amazon purchases for any item through the Amazon Portal below

Amazon Portal

« Thoughts on Ketosis - I | Main | FODMAPs »

No Such thing as a macronutrient part II - Carbohydrates (revised)

Note: I've added significantly to this part of the 3 part essay, so to make sure you see it all, I've re-posted.

Now we can have some fun, as I think I can demonstrate to you that this is by far the least metabolically useful macronutrient category.


Here is plain old glucose:


I like these stereo models – black is carbon, red is oxygen and white is hydrogen. First is the unstable open form and next is the six carbon ring.

Much of the confusion in the nutrition wars is engendered by the dual roles that molecules have – as both food components and as internal fuel molecules. Dean Ornish and Dr. Davis think the palmitic acid our bodies use for fuel while we sleep is poison if we eat it. Zero-Carbers like Charles Washington think the oldest fuel in our evolutionary history – glucose - used by organisms a billion years ago and without which the brains of modern mammals cannot survive for more than a few minutes – is an unnatural toxin if you eat it.

Both views ignore basic facts of medical physiology and defy evolutionary history.

Hypergylcemia is bad, but hyperglycemia is caused by pathologic loss of glucoregulatory control, not simply by eating a food that has glucose in it.

High serum free fatty acids (NEFA or non-esterified fatty acids) in the context of metabolic syndrome is a marker for metabolic syndrome and the NEFA in this context may even mediate some of the pathology. But that does not mean it is necessary to avoid foods with saturated fat in them due to their “lipotoxicity”.

Both of these ideas confuse the effects of molecules as internal metabolites with their suitability as food. Our cells do not “see” glucose in the form of glycemic index or load – they only see hyperglycemia in the pathologic state where insulin fails to control release of glucose from the liver or clear it from the blood.

Our cells do not “see” the long chain fatty acids that we eat, they are deconstructed at the gut, packaged into chylomicrons and then distributed for storage as triglycerides or for fuel use as NEFA or free fatty acids. Absent a pathologic state, free fatty acids will be in the blood based on the demand for them as fuel, and this will be controlled in concert with the availability of glucose (The Randle cycle) but independently of whether we just ate steak or butter. Our Paleolithic ancestors did not have to be careful to avoid too much starch or too much saturated fat at one sitting. To suggest so would require that normal humans eating real food are in grave danger without access to fitday’s calculator. Only focusing on test-tube science to the exclusion of common sense can lead to such ideas.

I refuse to believe our bodies could be that stupid.

Glucose is a necessary internal fuel source and metabolite and it is also a food and the building block of foods that have the longest evolutionary history of any food that mammals use. It is a fact that we do not require glucose in the diet, and that we can make it from amino acids if we don't eat it. However, rather than viewing this as evidence that glucose is not important, we should view this as evidence that glucose is so metabolically important that we have evolved way to make sure we always have it. Fatty acids as fuel came later, and required the evolution of eukaryotic cells, which have the the mitochondria that probably started out as symbionts (helpful parasites) but now are the cellular furnaces that are the only way we can burn fatty acids as fuel.

Humans make and secrete amylase, the enzyme that allows digestion of starch, both in the saliva and in the exocrine pancreas. Amylase copies show some variability, with some human groups having perhaps 50% more activity than others. However, all humans have at least three times more activity than chimpanzees. It would seem a stretch to claim that any group of humans has “lost” the ability to eat starch, as that is the sole purpose of amylase – to hydrolyse starch, which is a polymer of glucose molecules, into glucose, which our small bowel has no difficulty absorbing. If we are not broken, glucose or starch that is consumed is absorbed and rapidly cleared by the liver and muscle cells. It can then be burned as fuel (preferentially) stored for later as glycogen (the body’s version of starch – another glucose polymer) or in some cases turned into fat – known as de novo lipogenesis. All this partitioning is done based on metabolic needs in a normal person. Arguments that humans are not “supposed” to eat any starch are, in my opinion, inconsistent with ethology (study of human societies), established paleoanthropology and medical science.

Glucose and starch in the DIET are not poisons in a healthy human.

In a healthy person, it makes little difference metabolically how fast the glucose is delivered. I do not think there is any pathology caused by the “glycemic index” of foods based on their content of glucose or its polymers.

So the first two carbohydrates in our scheme are versions of the phylogenetically oldest fuels that mammals use, fuels which are an absolute requirement of our brains. There is no good evidence that it is unnatural for humans to have foods that contain them as some portion of their diet.

Glucose is also found in shorter molecules, and paired with other simple sugars, or monosaccharides, to make disaccharides and longer molecules called oligosaccharides. Some of these other simple sugars are only different from glucose metabolically in minor ways. Lactose is glucose paired with galactose. If you have lactase to cleave the lactose dimer (disaccharide), glucose is glucose and galactose is converted to glucose after absorption.


One of the simple sugars glucose is often paired with is Fructose.

Fructose is special. I would argue that fructose is so special that even calling it a “carbohydrate” is misleading. Later on I’ll discuss cellulose, which like starch is a polymer of glucose, but which, as the main component of indigestible sawdust, could not be metabolically more different than starch. So we can have polymers made of the very same lego-like building blocks, but because they are attached to each other in a different way, it is really a completely different substance.

In the case of fructose, we have a monosaccharide that has the same chemical formula and a caloric content equivalent to glucose, but is treated quite differently by the body because it has a different 3- dimensional structure.


1)   Fructose is found in plants foods only, and especially in their fruit. Plants use fructose to attract animals like us to the fruit, so we will eat it and spread their digestion-resistant seeds about via our feces. The plant is not thinking of us when it does this. Fructose is fructose and is tolerated in reasonable amounts (whether you get it from table sugar, honey or high fructose corn syrup), but because fructose is tasty we have bred fruit and cultivated plants in order to increase its availability dramatically. Fructose has easily become an order of magnitude more abundant in our diets in the past few hundred years than it was at any time in the preceding several million years of human evolution. If fructose were as benign as saturated fat or starch, this would be no problem, but I am pretty sure it is not.

2) Like glucose, there is no dietary requirement for fructose, but unlike glucose, we do not require fructose for use as an internal fuel. There is no organ like the brain that has an absolute fructose requirement. In fact, our body has mechanisms that evolved specifically to keep most cells from being exposed to too much of it.

3)   Because fructose spends more time than glucose in the unstable and reactive open configuration, it can react with proteins in a chemical reaction known as the maillard reaction. This results in glycation – attachment of a sugar – to other molecules, especially proteins. As proteins can be important structurally or as enzymes, this can have pathologic consequences. These glycated compounds are known as advanced glycosylation end products – AGEs.

4)   Fructose absorption in the gut is most efficient when paired with equimolar (one-for-one equivalent) amounts of glucose.

5)   When there is fructose in excess of glucose, or even when there is a large amount of fructose with glucose, there is often malabsorption in the small bowel – this can lead to rapid fermentation by bacteria in the colon, or abnormal overgrowth of bacteria in the distal small bowel. I speculate that fructose malabsorption is actually a defense mechanism to keep the liver from being overwhelmed by this metabolic poison, and the fact that we have not evolved a mechanism to handle big-gulp doses of fructose to the small bowel indicates modern quantities are likely outside of our evolutionary experience – the EM2.

6)   When fructose is absorbed, it goes via the portal vein directly to the liver, and the liver attempts to clear it completely so it cannot get into the general circulation. This is good, as fructose seems to be about 10 times more likely to cause glycation than glucose. Even small amounts of it can wreak havoc.

7)   To keep fructose out of the general circulation, it must be immediately burned or stored as fat. Fructose is related to the spectrum of serious diseases known as NAFLD (non-alcoholic liver disease), including fatty liver and cirrhosis.

8)   Excess fructose, chiefly via the liver volunteering to “taking one for the team” causes a variety of negative effects that are linked to pathologic insulin resistance, metabolic syndrome, a general inflammatory state, and of course obesity.

9)   Finally, fructose has no immediate effect on insulin release, but is linked to pathologic hyperinsulinemia via it’s effects on the liver. This is the exact opposite of glucose, which requires insulin to partition it when eaten, but for which there is no good evidence to relate it to chronic pathologic hyperinsulinemia.

(Note: This does not mean eating glucose is harmless once you have metabolic syndrome. You also have to be careful of large boluses of fat once your gallbladder is diseased. This doesn’t mean eating fat caused your gallstones, though - quite the opposite in fact.)

So fructose is a “carbohydrate” that has the same chemical formula as glucose, but unlike glucose, is very highly reactive with other molecules, is obligately metabolized by the liver, is malabsorbed by the majority of the normal population, can lead to NAFLD including fatty liver and cirrhosis, and in my view, is thus partly accountable for the current epidemic of metabolic syndrome, obesity and all the related diseases of civilization, including coronary disease and epithelial tumors.

Why do we lump harmless starch and possibly toxic fructose together and say they are equivalent macronutrients? They seem to have very little in common metabolically. Who cares about the paper chemical formula?

How many human diet trials or animal trials have you seen that lump them together? How many that treat them as totally separate variables like they should?


So we’ve seen how two “carbohydrates” that have the same numerical chemical formulae, glucose and fructose, have little else in common metabolically. I already hinted at how the way we attach glucose monomers to each other can result in either an easy-to-digest source of fuel (starch) or something you can build ships and houses out of (cellulose) that you cannot really absorb at all. (Some of your colonic bacteria may hydrolyse and consume limited amounts of less lignified cellulose and turn it in into fatty acids you can then absorb – perhaps more on that in future posts.)

So what happens if you take what is arguably a dose-dependent toxin like fructose and make a polymer out of it – long chain molecules made of almost nothing but fructose? (There is often a terminal glucose molecule attached to inulins and shorter fructose polymers called fructans.)

You get a fashionable “prebiotic”, a “soluble fiber” called INULIN.

Now there is no time for a long digression on the merits of pro- or pre-biotics. For now I’ll say that there are theoretical reasons to think small amounts of some types of fiber may be beneficial and are well within the realm of our evolutionary history, but the idea that fiber supplements or the large amounts that come with a vegetarian diet are necessary to optimum health is at best unproven. Massive amounts of fiber are not likely to be any more a requirement than massive amounts of starch or saturated fat. Good? Maybe. The more the better? Very implausible. We're not gorillas.

Inulin is considered a soluble fiber, because it cannot be hydrolysed and digested in our small bowel, but the bacteria in our colon can eat it – and the result is fermentation into short-chain fatty acids and lactate. Oh, and I almost forgot. Gas. Hydrogen and CO2 are produced. If you have Irritable bowel syndrome (IBS) and really want to suffer, just eat a whole onion or a bunch of Jerusalem artichokes. You might want to warn anyone you cohabit with, though.


Resistant starch is a digestible starch that survives to make it to the colon. When you eat cold leftover baked potatoes or cold pasta, some of the starch granules re-aggregate into a form that makes them less accessible to digestion, similar to their form before cooking. The starch that makes it to the colon is then eaten by colonic bacteria similar to the fate of inulin.


FODMAPS are a group of carbohydrates that have in common that they are poorly absorbed in the small bowel, but can serve as a feedstock for our colonic bacteria. FODMAP stands for fermentable – oligo- di- and monosaccharides and polyols. And fermentation means that bacteria or fungi can anaerobically digest it in our colons.

What FODMAPs have in common is that they make it to the colon without being hydrolysed and absorbed in the small bowel the way our fuel-source starches and regular sugars are. They can all be considered “prebiotics”. However, unlike longer molecules like inulin that are also fermented in the colon, the FODMAPs tend to be rapidly fermented in the proximal parts of the colon, rather than slowly fermented more distally. This rapid proximal fermentation can even lead to increased growth of bacteria in the terminal ileum – the distal part of the small bowel that attaches to the proximal colon – then you have SIBO (small intestinal bacterial overgrowth).

So what are they?

Oligosaccharides (OS) are short polymers of simple sugars. Short polymers of glucose are just glucose oligosaccharides – they are hydrolysed and absorbed just like starch. To be a FODMAP, the OS must have a bond or bonds that we have no enzyme to hydrolyse (break apart). One example is Raffinose, which is Galactose (milk sugar) joined to the fructose sugar of sucrose by an alpha bond. Galactose – Fructose –Glucose. Alpha galactosidase (a-gal) is the enzyme used to break the alpha bond. Humans and hindgut fermenting mammals don’t make it, but bacteria do. The best illustration of the effects of eating a lot of raffinose remains the campfire scene in Blazing Saddles. Raffinose is one of the main OS found in beans. The A-gal enzyme is the active ingredient in “Beano”.

Fructans and fructo-oligosaccharides (FOS) are FODMAPs found in wheat and may account for much of the GI distress encountered with high-wheat diets, and conversely for much of the relief obtained by eliminating wheat.

Fructose is a monosaccharide FODMAP functionally. By this, I mean that if you absorb it all in your small bowel, it is not a FODMAP. But if you get a big dose of free fructose  - fructose in excess of glucose in the same meal - or if you get a very large bolus even with the same amount of glucose - two coca-colas in a row, a binge on glazed donuts, a bunch of Christmas cookies at the office party – it is likely that some of the fructose, or even a lot of it, will escape digestion and absorption in the small bowel and make it to the colon.

Examples of polyol FODMAPs include xylitol and sorbitol. These are sugar alcohols. Sugar alcohols are not broken down in the small bowel, so they do not cause significant blood glucose elevation when eaten. They do get fermented in the colon, though, and it is possible to exacerbate irritable bowel syndrome or get osmotic diarrhea by, say, compulsively chewing sugarless gum or eating “diabetic” chocolates.

Inulin, polyols and FOS are all very common additives found in the junk marketed as “sugar-free”, “low carb”, or “diabetic”. Many are not aware of how much digestive upset they can cause. Another good reason to stick with real food and not eat anything that comes in a box, whether it says “low fat” or “low carb” or “sugar free”.


Rounding out the spectrum from most-to-least digestible, we can add Cellulose. Cellulose joined with lignin is wood. Even though it’s made of the same glucose as in starch, by having more chemically resistant bonds between the individual sugars you can build ships and houses and furniture out of it. Cellulose in plants without much lignin (iceberg lettuce) can be partially digested, but is best considered “indigestible fiber” with no caloric value and no health benefit, unless you believe the “scrubbing your colonic mucosa with a wire brush is good” theory.

So let’s summarize our paradigmatic carbohydrates in a "nutritionist" taxonomy:

Glucose and Starch - the oldest animal fuel

Fructose -a special carbohydrate that we have evolved to tolerate but that is toxic in industrial quantities

Soluble Fiber/RS - Inulin and other carbohydrates slowly digested in the colon. Overlaps with FODMAPs.

FODMAPs - carbohydrates indigestible in the small bowel that are rapidly fermented in the colon. May be problematic in excess.

Insoluble Fiber - Cellulose or other fibers that are not fermented significantly. The bran of grains or wood dust added to "heart healthy" breads are good examples.

Can we now apply these ideas to a taxonomy of carbohydrate sources -  a taxonomy of plants as foods?

Using some evolutionary reasoning, as well as some common sense, I would like to suggest a heirarchy of plant food desirability. I take it as axiomatic that whatever might be gained by eating plants, some key ideas should guide us.

1) A human can do with relatively little plant food, but cannot live at all without animals and the essential fatty acids and fat soluble vitamins they provide. With modern knowledge and care plants can provide essential amino acids when combined, but almost any animal can be eaten with monotony and provide all essential amino acids.

2) Most micronutrients necessary on an animal-based diet can be obtained by eating high quality animal sources, including offal. After all, the animal already ate the plants for you and either was replete in the diet or synthesized what was needed.

3) "Favor Food That is Defenseless when Dead". With rare exceptions, eating animals offers few trade-offs in terms of exposure to antinutrients or poisons in order to obtain fuel or nutrients.

4) Regardless of persistent claims regarding the uniqueness of particular plants and their abilities to provide x special compound that some observational or rodent study suggests protects against y, the idea that humans evolved to require any particular plant is implausible. The heterogeneity of the human diet in regard to plants, and the ability to live on very few animal species is an argument against the essentiality of particular plants. It is not a good argument for the necessity of consumption of a large variety of plants to obtain critical magical compounds from each one.

5) The frequently observed and logical behavior of modern hunter-gatherers, horticulturalists, primitive agriculturalists and wild animals suggests that if a plant is favored or sought, it ought to be the one that supplies the most energy density - fuel. The ONE thing that plants can do for us that animals can't is to relieve us of the burden of diverting amino acids from their critical structural uses to their inefficient conversion into mere fuel.

6) Plants consumed should therefore provide as much fuel as possible with exposure to as few antinutrients as possible. They should be close to the kinds of plants we evolved eating - as free of GMO modifications and artificial industrial processing as reasonably achievable.

So the top of the plant heirarchy will be:

Starchy Plant Organs - tubers and root vegetables. Sweet potatoes, cassava, root vegetables, white potatoes, etc.

Then, in descending order:

Green Vegetables - The "Atkins type" vegetables - leafy greens, etc.

Non-gluten Grains - white rice, corn - acceptable starch sources when not made a staple food



Gluten Grains

I think these kind of food categories, as well as the the more "nutritionist" categories preceding them, would be far more meaningful variables in a food frequency questionnaire or dietary trial than categories like "percentage of carbohydrate" where fructose and potato starch are lumped together, or "servings of fruits and vegetables" as if green beans and watermelon were remotely the same in dietary value.


I'm not going to list the constituents of every category here or argue about the exact order. I have suggested many times that the most benefit can be obtained by eliminating the last category of gluten grains. Gluten grains are so much worse than legumes that legume elimination may not even be that important. Let the lack of detail emphasize the non-essentiality of any plant category. You very likely don't need anything on this list beyond green vegetables, so you shouldn't be making a shopping list from it.

The point of these categories is to emphasize that my plant heirarchy is the opposite of the usual low-carb idea - the "it's OK, there are no net carbs in it" one.

I think eating 15-20% of calories as starch is healthier than being VLC (very low carbohydrate) on only green veggies and meat. Starchy Plant Organs give you the most useful plant vitamins and minerals, spare your body the work of making your own glucose via gluconeogenesis, keep you out of ketosis, and keep your glycogen stores topped off, all while avoiding the antinutrients in grains. If you are fairly active, you might notice a big difference between 20% starch and 5%.

So this makes PaNu very highly aligned with Kwasniewski on both ketosis and potatoes.

Ketosis is useful for weight loss, but it is not likely to be optimal to live in ketosis most of the time if you don't have to.