By Guest Author 09/08/2016

Guest post by Grant Schofield, Professor of Public Health at Auckland University of Technology and Director of the Human Potential Centre.

Another week, another warning that the paleo diet is not really where its at, especially for diabetes.

The paleo diet – the idea that we should be guided in human nutrition/public health nutrition by evolutionary history is steeped with controversy. Health experts and authorities are seemingly going well out of their way to make sure people are warned off such ways of eating.

Proponents are often mystified by this, because the idea of using human evolutionary history to understand human function is common in human biology. In fact its a guiding principle. As well, in the midst of a chronic disease epidemic, including diabetes and obesity which are potentially improved by this approach, you’d think approaches which are based on whole food eating, and appeal to at least some of the population would be welcomed.

I find it curious that other approaches such as vegetarianism, which are often based not around science, but religion and other beliefs are welcome in public health nutrition advice. Yet the paleo approach is not.

Yes, people who are follow this way of eating are restricted to eating much less processed food and often lower carbohydrate diets. Neither of these approaches are known to be anything but beneficial for human health, especially in the context of diabetes.

Anti-paleo attacks

Paleo foods
Credit: Flickr / Ines Hegedus-Garcia

As the author of a recent review of low-carbohydrate diets for diabetes in the New Zealand Medical Journal, we appreciate that Professor Sof Andrikopoulos’s article on Paleolithic (“Paleo”) diets for diabetes in the Australian Medical Journal was a fairer summary of the limited evidence, and his references to low carbohydrate diets for diabetes were more positive and consistent with the evidence than his comments in the mouse study in Nature Nutrition and Diabetes which he coauthored earlier this year. In his media activity after the mouse study he extrapolated wildly from the mouse feeding study to the paleo diet in humans. This wasn’t helpful for public health.

He does continue his anti-paleo attacks, more or less evidence free, on a couple of fronts which do need to be exposed. First, he contends that people on paleo diets are at risk nutrient deficiencies (specifically fibre and calcium) by not eating whole grains.  A claim which has absolutely no basis in reality or even theory. Nutrient profiles of people on paleo diets and similarly constituted low carb diets show no evidence of producing nutrient deficiencies. We are unaware of any condition ever recorded because of avoiding grains and eating whole foods from plants and animals instead.

He claims that “And high-fat, zero-carb diets promoted by some celebrities make this worse, as they can lead to rapid weight gain, as well as increase your risk of heart disease.” This is startlingly incorrect. Low carb diets are known to be useful in reducing weight and hanging diabetes.

Though we lack a universally agreed definition for the Paleo diet, the Paleo diet for diabetes, as used in practice, and as promoted in Australia by Pete Evans is merely a low-carbohydrate diet with some additional restrictions. Some of these restrictions are, in our view, unnecessary for most people with diabetes, but they are unlikely to prevent the Paleo diet from having the same benefits as other whole-food diets equally those restricted in carbohydrate.

Diabetes and the paleo diet

The defining feature of diabetes is high blood glucose; the liver should stop releasing glucose when we eat carbohydrate, but in diabetes it does not, due to insulin resistance or, in type 1 diabetes, the absence of insulin. The resulting combination of high blood glucose, combined with excessive insulin concentrations in the insulin resistant, causes the complications of diabetes, and can also play a role in conditions such as cardiovascular disease in those without a diabetes diagnosis.

The defining feature of diabetes is high blood glucose. Credit: Flickr / Alisha Vargas.

When the diet of someone with type 2 diabetes is low in carbohydrate, the liver no longer releases glucose from glycogen after meals. The reduced glucose concentration and insulin response, the drop of elevated blood pressure due to increased excretion of sodium, and the weight loss that is usual when carbohydrate calories are restricted all result in an improvement in the clinical picture, shown by a reduction in medication requirements. This is a simple and logical solution to diabetes management on first principles. Glucose from the diet puts of blood sugar and further stimulates extra production of it by the liver. So don’t eat it in the first place!

So we end up with this continuing determination to go well beyond the evidence in claiming a lack of evidence. In my world that’s the pot calling the kettle black. What we do need are a range of dietary practices which are effective in helping people manage their health and well being. For some, the paleo template makes sense and they will find their way to good health. I ask again, why paleo eaters are a target, and not the commercial food companies promoting highly processed food, the health system that is dominated by big pharma, and the health system that does sickness not heath?

Surely we have bigger fish to fry?

0 Responses to “Anti-paleo diet attacks miss the point”

  • I reviewed Andrikpoulos’ reference list and noticed he omitted a study that is pertinent to diabetes and the paleo diet:
    Ryberg, et al. doi: 10.1111/joim.12048
    Diminished fasting serum glucose in obese non-diabetic postmenopausal women on paleo diet.

    I’m also aware of other studies that show superior weight loss in non-diabetics on a paleo diet compared to control diet, which should help overweight type 2 diabetics (e.g., Mellberg doi: 10.1038/ejcn.2013.290)

    Thanks to Prof. Schofield for this post. I agree with it.

  • Actually eating carbs causes insulin production which causes glucose to be stored by the liver for later use, not released after a meal as stated.

    • Tricia, the storage of glucose as glycogen is a effect of insulin in normal metabolism, but there are several experiments showing that, in people with type 2 diabetes, there is a reduction in the formation of glycogen and an increase in the release of glucose from glycogen following a carbohydrate meal. If you feed these people a meal without carbohydrate, glycogen formation (non-oxidative glucose disposal) is increased and glycogenolysis is decreased.
      This is probably a result of the glucagon response to carbohydrate, which, in people with type 2 diabetes, is not sufficiently corrected by insulin due to insulin resistance.
      In type 1 diabetes an injection of glucagon is used to correct severe hypoglycaemia, and has the immediate effect of increasing glycogenolysis, releasing glucose from the liver’s glycogen stores. Diabetes is the inability of insulin to correct the effects of glucagon due to resistance or deficiency.

  • I wonder what percentage of people medicated for heart disease have either diabetes or insulin resistance that has not been diagnosed.
    I also wonder how many doctors have been surprised to see patients with diabetes, insulin resistance, hypertension, GERD and heart disease, get measurable health improvements and decreased need for medication, simply by following a self-prescribed low-carb version of Paleo diet.
    Finally I wonder when hospitals will stop feeding patients sugary high-carb meals. It’s like forcing patients to have a hazardous glucose tolerance test 3 times a day.

  • I’d like to make a few points about the differences between the experimental Paleo diets cited by Prof Adrikopoulos and the Paleo diet used by those with diabetes and that recommended by Pete Evans and others in Australia.

    The experimental or academic Paleo diets (which are not low carbohydrate diets) use canola oil, not usually considered a Paleo food. These diets are based on research into hunter-gatherer diets in Africa which showed game meat was high in omega-6 and omega-3 fatty acids found in canola. But is a refined vegetable oil really the same as the fat from wild game? In animal experiments diets high in polyunsaturated fat from refined oils consistently aggravate fatty liver disease, both alcoholic and non-alcoholic (NAFLD).
    This build-up of fat in the liver is thought to be a primary cause of insulin resistance and in humans NAFLD has been treated by a diet restricted in PUFA.
    Dairy fat is associated with reduced incidence of diabetes in epidemiological studies, olive oil has a beneficial effect on insulin sensitivity, coconut oil has various health benefits in experiments, and the use of these fats pre-dates by centuries the modern increase in diabetes and obesity. So paleo diets, outside of the academic field, usually recommend butter, extra virgin olive oil, and coconut oil be used in food.

    Whereas the experimental or academic Paleo diet avoids salt, the loss of sodium on a very low carbohydrate diet actually means you need salt in the diet, and this should include iodised salt.

    Similarly, there is no shortage of evidence that fibrous legumes and pulses are beneficial for people with diabetes, and some of these are relatively low in carbohydrate (frozen beans and peas, green beans, endamame). It makes no sense to avoid these pulses in treating diabetes, yet eat other carbohydrate-rich foods.
    Restricting pulses as well as grains in a Paleo diet context is something that often helps with gut and autoimmune diseases, but the evidence for diabetes say that if you are following a diet with a moderate or high amount of carbs then pulses are one of the least problematic carbs to eat.

    While there is evidence that adding milk to the diet increases weight slightly (less so with full-fat milk), and some theoretical concern about limited and controversial evidence linking A1 casein to pancreatic damage, in practice a little cheese and cream goes a long way to making a low carb diet both palatable and nutritious.

  • I am in favor of Paleo diet for controlling blood sugar, especially in prediabetics and individuals with insulin resistance. Avoiding the foods that cause high glycemic responses makes total sense, but seems to me that it is not solving the problem – it is merely trying to a worsening of the resistance. We now know that the fermentable resistant starch (an insoluble dietary fiber) helps to reverse insulin resistance and significantly improves metabolism. So, in my opinion, Paleo diets are a good start, but the next generation improvement in blood sugar management is fermentable fiber.

    It is commonly overlooked that the change in the digestibility of starch in processed carbohydrate-rich foods. People throw around “whole grains” as a good carbohydrate without bothering to distinguish that unprocessed whole grains are digested very very differently in the body than refined whole grains. Refined whole grains are as bad for you as other processed white grains.

    Intact carb-rich foods contain resistant starch, or starch that resists digestion and reaches the large intestine, where it is fermented by the resident bacteria. There are now 8 good clinical studies showing that resistant starch significantly improves insulin sensitivity, especially in individuals with prediabetes or insulin resistance. One study found 50-70% improved insulin sensitivity in overweight, insulin resistant men simply by supplementing their diet with 3-6 tablespoons of natural resistant starch. (Maki, Kevin C., Christine L. Pelkman, E. Terry Finocchiaro, Kathleen M. Kelley, Andrea L. Lawless, Arianne L. Schild and Tia M. Rains. “Resistant Starch from High-Amylose Maize Increases Insulin Sensitivity in Overweight and Obese Men.” The Journal of Nutrition 142, no. 4 (2012): 717-723.,​jn.111.152975. That’s even without exercising or losing weight! The US Food and Drug Administration is expected to rule on a health claim petition this summer that resistant starch reduces the risk of type 2 diabetes, based upon these published studies. See for more information. I believe that it is as much the lack of fermentable fibers in our diet that increases the risk of prediabetes and thus type 2 diabetes as too many high glycemic carbs.

    Whole grain foods that maintain the kernel or seed or shell of the grain still contain resistant starch, while whole grains that have been processed into a fine white flour do not contain any resistant starch. Now that 37% of American adults are prediabetic and insulin resistant, it is not surprising that they respond well to a low carbohydrate or Paleo diet. Yes, avoiding the high glycemic carbs so common in our diets is a great start, but it seems to me that a better approach would be to make sure that they increase resistant starch and other fermentable fibers as a way of reversing insulin resistance itself. More like a 1-2 punch instead of merely stepping out of the way of an oncoming hit.

    • Hi Rhonda,

      I believe that you are right, that diabetes is not caused by high-carbohydrate diets in which grains are unprocessed, for example the dark bread of Italy or the Pumpernickel rye bread of Northern Europe.
      Our modern bread-making techniques produce a very different product. Even the milled rice of southern Asia did not produce a high rate of diabetes in the context of the Chinese diet, where it accompanied dishes of fresh vegetables, fermented foods, and nose-to-tail animal foods, until refined highly-polyunsaturated oils and sugar entered the diet in larger amounts. These are the two foods we can identify as causal in fatty liver disease through animal experiments and human interventions, and the intake of which has accompanied the increase in obesity and diabetes.
      It is not usually true that the low carb diet in type 2 diabetes is merely an adaptation to insulin resistance that allows normal life without improving the underlying disease, though this may be its benefit in more difficult cases.
      Type 2 diabetes is improved by weight loss, and LCHF diets are more likely to result in weight loss than diets that restrict fat and calories.
      But also, the LCHF diet can improve glycaemic control without weight loss, which is probably a unique advantage.
      And can reduce hepatic fat, improving insulin sensitivity.
      Weight loss, decreased glycaemia, and hepatic fat mobilization (which we expect to be matched by pancreatic fat mobilisation, but to be honest we are still awaiting pancreatic MRI results to confirm these hepatic fat results) are all expected to improve insulin sensitivity, which is of course consistent with the effects of LCHF diets in clinical practice.
      The question then is when, and whether, people who have reversed type 2 diabetes on LCHf diets should eat foods such as wholegrains again.

      While resistant starch can improve glycaemic control, the paper you linked to only showed an effect in men. There were twice as many women in the study, and the product did not benefit them. It used a concentrated resistant starch source HAM-RS2, not the cooled pasta or cooled potato or even whole grains we read about in the media.
      Low carbohydrate nuts, vegetables and fruits supply on average more fibre than wholegrains per 100 grams, as do legumes, and also supply more fibre in proportion to carbohydrate. By limiting high carbohydrate foods like grains and potatoes, there is more room for fibre of all kinds in the diet, if there is not, a supplement such as HAM-RS2 or indeed purified potato resistant starch, which can be bought in some healthfood stores, may well be valuable.
      It may be that women did not respond to resistant starch in the HAM-RS2 experiment because women perhaps eat more fibre than men overall to begin with, are more keen on salads, Special K, and wholegrain bread. The study, which did not seem to record baseline fibre intake, may show this in table 2 where the short-chain fatty acid levels changed in the men, but not in the women, showing no change in fibre fermentation in the women.

      Thus it may well be that enough fibre is enough, whether or not it is resistant starch.

      In regard to Prof Adrikopoulos’s review, the Paleo diet was always as good as or better than the control diet, and the control diet was based on current diabetes guidelines. For example, the Nordic diet control featured plenty of dark rye bread, but did not out-perform the Paleolithic diet. So he might as well admit that there is no great evidence that current guidelines are effective, if he is going to draw this conclusion about the Paleo diet from these trials.

  • I do not believe you are correct that fiber is fiber and enough fiber would work. I think that it takes fermentable fiber to change the gene expressions within the large intestine that help control the mechanisms, and that bulking fibers like cellulose or viscous fibers like psyllium or beta-glucan won’t do it. Non-fermentable fibers or fibers that are minimally fermented cannot change the intestinal gene expression in the same way. Different types of fibers have different benefits and they are not interchangeable. The modern diet lacks fermentable fibers which can quickly change the intestinal microbiome. While glycemic reduction is also important, as you have stated, the metabolic changes that occur through the well-fed microbiome are becoming more recognized. The data on resistant starch is confirming the powerful benefits of this approach.

    The data is pretty clear that resistant starch (RS) improves insulin sensitivity directly, without weight loss or exercise. You are right about the Maki study not showing an effect in women. However, a more recent study more closely examined this issue in depth and did demonstrate improvement in insulin resistant women – See It showed that insulin resistant women were most responsive to resistant starch, while it had no effect in women with more normal insulin sensitivity. It also took higher levels of supplemental RS to see the effect as well. The Maki study showed 50% improvement in men at 15 grams of RS/day but the Gower study needed 30 grams of RS/day to see 16-23% improvement in insulin resistant women. One other factor to consider is that the Gower study delivered the RS baked into cookies and snacks, wheareas the Maki study fed RS supplement as a raw starch. It was very encouraging that the RS worked in the women that most needed it – the most insulin resistant ones.

    I do not believe that cooked and cooled starchy foods like pasta and potatoes can deliver these benefits – the quantity of RS is too low. The clinical studies demonstrate improved insulin sensitivity at 15-30 grams of RS/day, whereas 1 cup of cooked and cooled pasta was measured to have only 1.2 grams of RS. Maybe if someone were eating lower quantities of RS in foods all their life, they would not develop prediabetes or insulin resistance in the first place, but reversing it definitely takes higher doses. See for confirmation.

    You stated: “But also, the LCHF diet can improve glycaemic control without weight loss, which is probably a unique advantage.“. This study is a small, proof of concept study but when high protein was substituted for high glycemic starch, insulin levels were increased, even if glycemic response was reduced. Isn’t high insulin also damaging within the body? Wasn’t there a study published showing that tightly controlling blood sugar levels in T2D using frequent insulin treatments was more harmful such that the study was stopped early? I always thought that looking only at blood sugar levels was short-sighted and that reducing insulin levels was also quite important.

    Another study have found that high doses of supplemental RS from high amylose corn fed also changed fat storage within adipose tissue directly.

    • Hi Rhonda,
      . Tightly controlling blood glucose with insulin in T2DM is harmful for two reasons, neither of which applies to the insulin response to protein. Firstly, there’s a risk of hypoglycaemia, which is acutely dangerous. Secondly, the combination of excess glucose and excess insulin has a number of consequences such as increased lipogenesis and glycosis which is chronically pathogenic, better to lower both by keeping carbs low. Protein is essential (however high the protein in a diet, 0.66g/kg/day of it could not have been avoided anyway), it is not highly lipogenic, its metabolism has a high energy cost, and it is unlikely to elevate glucose even at higher intakes.
      In the LoBag20 part of the Gannon and Nuttall 2006 study, when both protein (30%) and fat (50%) were high, 5 weeks of the higher protein, low carb diet had this effect on insulin: (see figure 9)
      “The net insulin area response, using the fasting insulin concentration as baseline, was decreased by 40%. The total insulin area response, using zero as baseline, was decreased by 25% following 5 weeks on the LoBAG20 diet. Both insulin area decreases were statistically significant.”
      The Gannon and Nuttal LoBag 2006 LoBaG paper is not a pilot study but a report on a series of experiments designed to test those effect of diet composition, independent of weight loss, on type 2 diabetes. Its conclusions are supported by clinical findings such as

      I agree that not all fibre is equal, but I do not think that the proposition that other fibre that is digestible to SCFA in the human gut has similar benefits is being tested in the resistant starch experiments. It seems to me that resistant starch might give better results after being fed for a period of time and allowing the gut microbiota to adapt. For example, metformin improves insulin sensitivity and dose is often gradually increased due to GI side effects perhaps caused by an increase in fermentation and LCFA production.
      Vinegar which provides the SCFA acetate in low concentrations (5%) has measurable effects on glycaemia, including hepatic glucose production
      And vegetables, nuts, and legumes which supply fibre other than resistant starch also increase SCFAs; though not a diabetes treatment in this paper, the Mediterranean diet gives fairly good results in T2DM studies.
      This 12-month study compared a resistant starch dietary intervention with a normal high fibre diet intervention (both 50% carb 32% fat 15% protein) in obese prediabetic adults (with high insulin). Neither dietary intervention helped much with insulin and fasting BG but the RS diet was not superior.

      If someone is going to the trouble of using purified resistant starch as a supplement to or instead of eating a diet naturally high in fibre, what is the purpose? People do not have to expose themselves to the glucose concentration of a tolerance test on a daily basis to be well-nourished. People with diabetes can live and eat well without doing this, and will generally be advised to avoid refined carbohydrates in these concentrations. Further, an important action of SCFAs involves the activation of AMPK, and this is also triggered by carbohydrate restriction and fasting, so that fibre may not be the only way to deliver the desired benefit. But resistant starch is certainly a tool people can experiment with to understand the effect that fibre and the microbiota can have on glycaemia, appetite, and so on.

  • Disappointed to see this pseudo-science here on Sci-Blogs to be honest.

    • Hi Mike,

      Do you mean this article, or the article it was written in response to, or both?
      What do you think science has to say about this issue, in regard to the clinical outcomes that are of interest here?

      • I mean the article, the “moveable feast” that is the paleo diet, seems as hard to nail down as the one true Scotsman.