Sunday, November 13, 2016

Thoughts on Spreadbury's "Cellular Carbohydrate" Hypothesis

Thanks to Jamie Scott and Melchior Meijer (PolderPaleo) for an interesting discussion about this hypothesis. This embedded tweet should get to you to the entire discussion, if you're interested.
Ian Spreadbury published this paper in 2012:

"Comparison with ancestral diets suggests dense acellular carbohydrates promote an inflammatory microbiota, and may be the primary dietary cause of leptin resistance and obesity" (Spreadbury, 2012)

It's an interesting read, and well-reasoned. At first I skimmed it, but then went back and reread thoroughly.  Spreadbury addresses many of my initial criticisms himself in the paper, so I won't go back and reiterate them. Read the whole thing, it's interesting.

My criticism of it is therefore only two-fold, although the first point is the key one.

First, the basic hypothesis is that "cellular" carbohydrates, that is, starches that are bound up in plant cell walls, are more healthy than "acellular" carbohydrates. An example of an acellular carbohydrate would be wheat flour or refined sugar, and a cellular carbohydrate would be a carrot, raw or cooked.

Plant starches are bound inside of cells, and these cells have largely indigestible (by humans) walls made of cellulose and pectin, basically. Fiber, in other words. In order to access the glucose in the starches for food, we must break down the cell walls.

We can do this in two ways: grinding with our teeth or with some machinery, like a mill; or cooking.

Spreadbury states:

"Tubers, fruits, or functional plant parts such as leaves and stems store their carbohydrates in organelles as part of fiber-walled living cells. These are thought to remain largely intact during cooking, which instead mostly breaks cell-to-cell adhesion."

He provides two links to back this up, one of which looks at carrots, the other at something which has nothing to do with cooking (footnotes 92 (Ng & Waldron, 1997) and 93 (Parker et al., 2001) in his paper).

First, a little background. There's a simple way to measure whether the carbohydrates in a food are cellular or acellular.  It's the glycemic index (GI).

GI measures how quickly glucose from food enters the bloodstream.  Since we can't access the glucose inside a cell wall very well, the GI of a food measures how acellular it is: cellular carbohydrates will have a low GI, as our digestive system struggles with the fiber and is unable to access the glucose within.  Acellular carbohydrates will have a high GI, as the cell wall won't impede digestion.

This is demonstrated in the following study, which compares raw and cooked potatoes to raw and cooked carrots:

"The Effect of Cooking upon the Blood Glucose Response to Ingested Carrots and Potatoes" (Vaaler et al., 1984)

I'll save you the numbers, here's the summary: raw potatoes and carrots have a similar effect on blood glucose: low.  The glucose is therefore cellular, and generally not available to us or our gut microbiome (I presume they were chewing the potatoes and carrots, so there is some availability). Cooked carrots have little change in GI: cooking may soften the walls, but doesn't break them down enough to allow our digestive tract to easily access the glucose.

Cooked potatoes, on the other hand, have almost the same GI as pure glucose: the cell walls are almost entirely broken down.  This holds in a similar fashion for every tuber I looked at except for carrots.  

Since Spreadbury discusses the Kitivans, I looked at some of their favorite tubers, like taro: same effect.  Cell walls are almost entirely broken down: they're high GI.

Spreadbury's hypothesis is that the high availability of acellular carbohydrates affects the gut microbiome, not blood sugar. But since our gut bacteria have a first shot at high GI foods, any effect on our blood sugar must also represent availability to our gut microbiome. Since clearly these tubers are no longer acellular after cooking, as reflected by the high GI, there must be some other explanation.

Some of these tubers have a higher GI than a hamburger bun, for instance, which is surely an acellular carbohydrate.

So he got that fact wrong.  

Cooked tubers like a potato or what the Kitivans eat are clearly acellular carbohydrates, just as acellular as a hamburger bun.  If you want acellular cooked tubers, you'd best stick to carrots, which are not on the Kitivan menu.

I agree with Spreadbury that low-GI foods are better for you than high-GI foods.  But if the Kitivans aren't obese and leptin resistant (and don't have any other symptoms of the metabolic syndrome commonly blamed on carbohydrates) a high-GI, high-carb, acellular diet cannot be the cause.

Second, Spreadbury notes that these tubers are basically composed of three ingredients: water, fiber, and carbohydrates.  This is correct, so let's examine this avenue as a "weak" version of Spreadbury's theory.  Spreadbury alludes to this here:
"This cellular storage appears to mandate a maximum density of around 23% non-fibrous carbohydrate by mass, the bulk of the cellular weight being made up of water. The acellular carbohydrates of flour,94 sugar and processed plant-starch products are considerably more dense." (Spreadbury, 2012)
Perhaps it's the density?  The fact that the carbohydrates are consumed with fiber and water?  The problem here is that this reading of the hypothesis is not novel.

It was first proposed by Dr. Denis Burkitt in 1966 (Cummings & Engineer, 2018).  Burkitt's "fiber hypothesis" is the reason so many people are today advised to eat fiber.  Since Spreadbury does not mention Burkitt in his paper, I assume he's unaware of him. [P.S. He's very aware of him. See his response below.]

Burkitt, like Spreadbury, was quite clear that his was a hypothesis, but Burkitt's was tested extensively over the years.  While people are still writing papers on it, and still recommending it, as far as I'm aware every solid test of fiber on disease has been negative, or such a small effect that it couldn't be said to be causal.  No disease that Burkitt noted has ever been cured by adding back fiber in any form to the diet.  Spreadbury even notes this:
"The proposed importance of the cellularity and low carbohydrate density of fruit and vegetables in maintaining an evolutionarily appropriate microbiota might explain the failure of supplementary fiber, vitamins, or antioxidants to replicate the health effects of a diet of fruit and vegetables when taken with a Western diet."
I came to Burkitt after being afflicted by diverticulosis, which is one of the diseases he attributes to the replacement of fiber-containing "natural" carbohydrates with refined carbohydrates, those in which the fiber has been removed.

Burkitt's theory is a better theory than Spreadbury's, as Spreadbury's relies on a misunderstanding of what cooking does to the cell walls in tubers.  Neither really had an explanation for a mechanism, but you don't need to know the mechanism for a treatment to be effective, so that's fine.

Based on Burkitt's fiber hypothesis, doctors recommended a high-fiber diet to people like me with diverticulosis for decades.  I'll focus on diverticulosis, since Burkitt did, and since I know it well.

As Burkitt explains in an article contained in Western Diseases, their Emergence and Prevention:
"The occurrence of diverticular disease is believed to depend on a long period of exposure to low fibre diets..." (Trowell & Burkitt, 1981)
In 2012, the same year Spreadbury's paper was published, someone finally got around to testing Burkitt's fiber hypothesis as it relates to diverticulosis.  It was published in the journal Gastroenterology:

"The complications of diverticulosis cause considerable morbidity in the United States; health care expenditures for this disorder are estimated to be $2.5 billion per year. Many physicians and patients believe that a high-fiber diet and frequent bowel movements prevent the development of diverticulosis. Evidence for these associations is poor. We sought to determine whether low-fiber or high-fat diets, diets that include large quantities of red meat, constipation, or physical inactivity increase risk for asymptomatic diverticulosis."
Their findings? 
"A high-fiber diet and increased frequency of bowel movements are associated with greater, rather than lower, prevalence of diverticulosis. Hypotheses regarding risk factors for asymptomatic diverticulosis should be reconsidered."
Burkitt himself should have figured this out, as the Maasai warriors in Africa, where he practiced, ate a low-fiber diet for decades, and didn't suffer any of the ill-effects that he describes.

But the important thing to note is that the Africans Burkitt encountered had a "high-fiber diet and increased frequency of bowel movements", but did not get diverticulosis.  So something else must be causative.

Spreadbury's article is a very interesting one, and he contains many observations with which I agree. I will also note that replacing the Standard American Diet with almost anything else is an improvement, a view with which I suspect Spreadbury would agree.  But I see no convincing evidence that this is because the replacement diet contains fiber, cellular or otherwise.  In fact, a diet comprised wholly of cellular foods has been shown to be very insufficient, given the human digestive tract.

Unfortunately his central thesis depends on a misunderstanding of the effects of cooking on cellular structures in food, and even, as he himself notes, the weak explanation of his theory has never been shown to work.

After first finding Burkitt's book six or so years ago, I've continued to study the matter of what causes what he terms Western Diseases.  There are better explanations, which better fit the universe of facts we have, and for which we have at least some understanding of the mechanisms.

That's one of the missions of this blog, after all.

P.S.  Here are Spreadbury's two comments on this topic.  First, to the twitter discussion:
.@TuckerGoodrich @PolderPaleo - Response from Dr Ian Spreadbury on our previous Twitter thread re: cellular vs acelluar CHO
"It's true we don't know enough about what different forms of cooking are doing to cellular or subcellular structure. There were a few papers I cited in 2012, which indicated cells were breaking off their neighbours, but were staying mostly intact. Cellularity is only one theoretical difference of life-derived minimally processed foods though, whatever is going on in poi after a long period of cooking likely doesn't make a difference to health, as that's how I believe a lot of the Kitavan root veg was being prepared. Exactly where on the slippery slope / grey area of cooking and more advanced processing things start becoming a problem is something that needs a lot more study. I don't feel too worried about whether it's whole cells knocking around, or complex molecular chaperoning, the patterns of microbial gene activation by a complex substrate, or just a reduced density of the rapidly usable stuff - at the end of the day it's nutrient broth versus life, the role of the food - bacterial - neuroimmune - metabolic interaction still seems a decent explanation however the details shake out.

"I'm thinking much more about the parallels with periodontal disease these days - bacterial infiltration into cells and all sorts starting to be unraveled. This is so much more complex than LPS, and the periodontal chaps are still trying to work out what's going on in as accessible an area as the mouth.

Second, to this post. The second has an Easter egg in it that refutes one of my presumptions above. See if you can pick it out. ;)
.@TuckerGoodrich @PolderPaleo A speedy reply from Ian Spreadbury to Tucker's post.
"He's right that we just don't know the factors that are making the difference, but it's something of a side argument. I take a stab at proposing some mechanisms by which bacteria might behave differently in the presence of 'real food'. However, it's ultimately Cooper's microbiota profiling from dental tartar and the broader dental archaeology that strongly suggest that somehow a flour / sugar interaction is what is going on, at least in the mouth. ‎How it works is going to be interesting for sure, but I'm not basing the argument on those micro-mechanisms. ‎Our mouth's microbial friends are somehow changed by modern foods, and a similar change is what I'm arguing then drives the health problems one swallow away in the gut.‎

"This is why Cleave ducked the argument , he was a smarter man than I ;)

'"there is no need to discuss here the biological mechanisms involved in this resistance; what matters is that such resistance is not present in inert foods".

[P.S. July 2022 The easter egg mentioned above is Spreadbury's reference to Cleave. Burkitt's work was inspired by Cleave's, and it demonstrates that, prior to my assumptions that Spreadbury had never heard of Burkitt, he clearly has. So once must conclude that his omission of references to the precursor work was intentional...]

Cummings, J. H., & Engineer, A. (2018). Denis Burkitt and the origins of the dietary fibre hypothesis. Nutrition Research Reviews, 31(1), 1–15.
Ng, A., & Waldron, K. W. (1997). Effect of Cooking and Pre-Cooking on Cell-Wall Chemistry in Relation to Firmness of Carrot Tissues. Journal of the Science of Food and Agriculture, 73(4), 503–512.<503::AID-JSFA762>3.0.CO;2-Z
Parker, C. C., Parker, M. L., Smith, A. C., & Waldron, K. W. (2001). Pectin Distribution at the Surface of Potato Parenchyma Cells in Relation to Cell−Cell Adhesion. Journal of Agricultural and Food Chemistry, 49(9), 4364–4371.
Peery, A. F., Barrett, P. R., Park, D., Rogers, A. J., Galanko, J. A., Martin, C. F., & Sandler, R. S. (2012). A High-Fiber Diet Does Not Protect Against Asymptomatic Diverticulosis. Gastroenterology, 142(2), 266-72.e1.
Spreadbury, I. (2012). Comparison with ancestral diets suggests dense acellular carbohydrates promote an inflammatory microbiota, and may be the primary dietary cause of leptin resistance and obesity. Diabetes, Metabolic Syndrome and Obesity: Targets and Therapy, 5, 175–189.
Trowell, H. C., & Burkitt, D. P. (1981). Western Diseases, Their Emergence and Prevention. Harvard University Press.
Vaaler, S., Hanssen, K. F., & Aagenæs, Ø. (1984). The Effect of Cooking upon the Blood Glucose Response to Ingested Carrots and Potatoes. Diabetes Care, 7(3), 221–223.


  1. Hi Tucker,

    Thanks a lot for this thorough analysis.

    I have to admit that the distinction between cellular and acellular falls apart. But in my mind this has never been the core of the concept. The core of the concept - as I see it - is that 'ancestral' foods, even if cooked, have such a low starch density per swallowed unit that they are not able to serve as a substrate for the microbiome in the upper small intestine (SI). Even if the walls of a tuber are distroyed by cooking, which they most ofte are as you pointed out, the absolute amount of starch reaching the SI will be lower after each swallow, than it would be after consuming, say, bread or pasta or any grain product, which are very dense. The GI of a cooked tuber is high, but since the amount of starch is limited, it will be almost immediately hydrolized and absorbed into the circulation. This is critcal in my view. At any given moment there will not be enough substrate in the upper SI to create or sustain a toxic microbiome. As soon as the amount of starch per bolus of chyme exceeds the capacity of the enzymes to hydrolyze it into glucose, there will be a time window in which the starch can be substrate for the flora. As to pure glucose, only small amounts, such as found in whole fruit, are safe. Glucose becomes a problem if the concentration is higher than can be immediately absorbed into the circulation, for the very same reason that in that case it becomes substrate for the flora.

    In this respect Spreadbury's hypothesis is entirely different than the falsified fiber hypothesis by Burkitt and Trowell. Adding fiber to very dense starch sources like grain products will do almost nothing to the starch load the upper SI will see. If anything, the fiber will serve as an extra substrate to feed into an already toxic upper intestinal soup, especially when transit time is altered.

    In a way I see Spreadbury's hypothesis as a neat explanation for why fiber free (all meat) diets turn out to be so tremendously effective against a plethora of ailments. Any diet that doesn't deliver a high load of substrate to the SI-microbiome, will produce health, I think. I'm not at all saying that tubers are necessary for or even contributing to good health, only that the absolute amount of starch they dump into our gut might be just low enough to not create problems. This would explain a lot of seemingly paradoxical observations.

    So far my first thoughts. Thanks again for creating this opportunity to discuss this concept!

    Melchior. (PolderPaleo :-) )

  2. 1. I can't help wondering how this cooking of tubers to acellularity is related to resistant versus non-resistant starch, for example when a potato is cooked and then cooled which supposedly makes (some of) the starch resistant again. And then there's soluble versus insoluble fiber. Are these all different names for the same thing?

    2. Here's a really dumb question: why is it so hard to determine whether cooking breaks down cell walls—can't this be seen under a microscope? Looks like they may have done it here, although this was a study on repeated cooking and cooling cycles involving resistant starch:

    3. Another way of processing food is fermentation, as in fermented vegetables such as sauerkraut. I'm getting ready to ferment some raw potatoes, for example.

    1. Regarding 2: the best way to determine if cell walls are being broken in vivo is blood glucose. It tells you if in fact the glucose behind the cell wall is making it into the blood, which is what we care about.


Please don't comment here. Go to

Note: Only a member of this blog may post a comment.