Search Results for: selenium - Page 3

Around the Web; The Case of the Killer Vitamins

I’d like to thank Patrick Timpone for a very enjoyable interview on The Morning Show at One Radio Network. Here is the MP3; I’m on for the second half of the show. You can find a zip file at the archive for October 13. Patrick’s producer Sharon tells me that she’s already benefited from our book:

I was following The Primal Diet and since I read the book, I’ve been allowing myself potatoes and rice and doing very very well on them among doing some other things you recommend.

Also, I’d like to thank Jimmy Moore once more for hosting his highly entertaining “safe starch” symposium (Jimmy’s original post; my response, here and at Jimmy’s). It was great to get the opportunity to explain ourselves to so many people in the low-carb and Paleo movements.

Jimmy is planning to try our diet for a week in November, which will be a good occasion for us to publish a 7-day meal plan. We’ll invite anyone who’s curious to try the diet along with Jimmy, and compare notes.

[1] Interesting posts this week:

Angelo Coppola on Latest in Paleo wonders if Denmark’s saturated fat tax will apply to mother’s milk. If so, it’s bad news for unemployed infants! (He also discusses the “safe starch” debate.)

I once knew a French astronomer who died from snorting cocaine while observing at 14,500 feet. Emily Deans makes me wonder:  Did he have Crisco for dinner?

Stan the Heretic offers his mitochondrial dysfunction theory of diabetes. Peter Dobromylskyj and JS Stanton are also developing ideas along this line. Speaking of JS, his post this week has some great photos of Sierra wildflowers and reflections on the state of the Paleo community.

CarbSane partially confirms Dr. Ron Rosedale: eating carbs does raise leptin levels compared to eating fat, but it is a mild rise over an extended period of time, not a “spike.”

Beth Mazur explains why her bathroom door is always closed.

Chris Kresser discusses why chronic illness often generates a form of hypothyroidism, low T3 syndrome.

Joshua Newman knows how to flatter.

How solid is the case against Andrew Wakefield? Autism is certainly characterized by intestinal dysfunction, and Age of Autism notes that distinguished scientists are citing Wakefield’s work.

Richard Nikoley claims he doesn’t know the words to “Kumbayah.”

Seth Roberts points out that the Specific Carbohydrate Diet has been curing Crohn’s for 80 years, but still no clinical trial.

Jamie Scott, That Paleo Horse Doctor, asks: Why do horses get laminitis?

We’ve quoted vegetarian Dr. Michael Greger’s concerns about arsenic in eggs. I’m more concerned about soy protein in eggs.

Following Steve Jobs’s death, Tim asked for an opinion about the unconventional cancer therapies of Dr Mercola’s friend Nicholas Gonzalez. David Gorski, toward the end of a detailed examination of Jobs’s medical condition and treatment, links to his own claim that the Gonzalez protocol is “worse than useless.”

[2] Music to read by:

[3] Cute animal photo:

[4] Notable comments this week:

PeterC’s dad, who has diabetes, is doing well on our diet. Daniel’s stepdad had a similar experience.

Helen informs us that sweet potato intolerance may be due to raffinose.

Mario Iwakura gives us his infectious theory of diabetes. I think a lot of the cases of disrupted glucose regulation, where people get frequent hyperglycemic and hypoglycemic episodes, may be due to occult infections.

Dr Jacquie Kidd (who blogs at drjacs.com) has gotten some great advice from Jamie Scott.

Ellen tells us of cases of iodine supplementation controlling diabetes.

Ned is looking for grass-fed cowbells.

[5] Do Vitamins Kill?: An analysis of the Iowa Women’s Health Study came out this week, and it purported to show that nearly all supplements except calcium and vitamin D increased mortality, with iron being the worst. Oskar asked us to look into it, so we did.

The study followed a large number of women in Iowa, and queried them several times about supplement use. In 1986, the baseline, the women had an average age of 62 (range of 55 to 69) and 66% were taking supplements. By 2004, the surviving women had an average age of 82 and 85% were taking supplements.

Here is the data on overall mortality vs supplement use:

“Cases” are instances of someone dying. “HR” or hazard ratio is the likelihood of dying if you supplement divided by the likelihood of dying if you don’t. Note that all the hazard ratio estimates are “adjusted.”

Unadjusted Hazard Ratios

The left columns of the table give us death statistics and allow us to calculate raw hazard ratios, with no adjustment whatsoever. Seven of the supplements have unadjusted HRs below 1.00, eight have unadjusted HRs above 1.00. The 15 HRs average to 1.01. Without copper, which has an unadjusted HR of 1.17, they average to 0.998. In short, death rates among supplementers were almost identical to death rates among non-supplementers.

This is interesting because supplement usage rose rapidly with age. It was 66% at age 62 and 85% at age 82. Supplement users were, on average, older than non-supplement users. But mortality rises rapidly with age. So there should have been a lot more deaths among the supplement users, just because of their more advanced age.

The paper should have, but didn’t, report age-adjusted hazard ratios. Adjusting for age is very important, since mortality depends strongly on age, and so does supplement use. However, it’s obvious what the result of age-only adjustment would have been. Supplement usage would have shown a substantial reduction in the risk of dying.

Hazard Ratios Adjusted for Age and Energy Intake

The least-adjusted hazard ratios reported in the paper are adjusted for age and energy intake.

The energy intake adjustment is disappointing, because energy intake is affected by health: healthier people are more active and eat more, and obese people also eat more. Including indices of health as independent variables in a regression analysis will tend to mask the impact of the supplements on health, creating misleading results.

However, let’s go with what we have. Based on “Age and Energy Adjusted” hazard ratios, supplements generally decrease mortality. Nine of the fifteen supplements decreased mortality, five increased mortality. At the 95% confidence interval, five supplements decreased mortality, only one increased mortality.

Looking at the specific supplements, results are mostly consistent with our book analysis. Let’s start with the five that showed harm:

  • Folic acid and iron – two nutrients we regard as dangerous and recommend not supplementing – both elevate mortality, as we would expect. Iron is particularly harmful, and should generally be avoided by women once they have stopped menstruating.
  • Multivitamins slightly increase mortality, a result that has been found before and that we acknowledge in the book. This is probably due to (a) an excess of folic acid, (b) an excess of iron (if the women are taking iron-containing multis after menopause), (c) an excess of vitamin A (this is no longer the case – multi manufacturers have reduced the A content of vitamins in response to data – but in 1986-2004 most multis contained substantial amounts of A) which is harmful in women with vitamin D and/or K2 deficiencies (both extremely common, and D deficiency in this cohort is supported by the benefits of D and calcium in the study and the northerly latitude of Iowa) or (d) imbalances in other nutrients; for reasons of bulk multis tend to lack certain minerals, notably magnesium and calcium.
  • Vitamin B6 is an anomaly, as we wouldn’t expect B6 to be harmful in moderation. I’m guessing B6 would have been taken to reduce high homocysteine and for this purpose would often have been taken along with folic acid, a harmful supplement. Also, B6 should be balanced by vitamin B12 and biotin, and may not have been. Perhaps people with cancer were unaware that B6 promotes tumor growth; (UPDATE: See comments; I was misremembering studies, B12 and folic acid can promote tumor growth, but in other studies B6 looks protective against cancer) indeed, in the breakdown by cause of death in Table 3, B6 increases cancer mortality by 6%, but CVD mortality by only 1%. (Folic acid and vitamin A were other cancer-promoting supplements.) The harm from B6 was not statistically significant and I wouldn’t read much into it.
  • Copper is another anomalous result, but this was the least popular supplement, taken by only 229 women or 0.59%. Copper’s hazard ratios were dramatically affected by adjustment: in the raw data, mortality is only 17% higher among copper supplementers, but after age and energy adjustment it is 31% higher, and multivariable adjustment increases it substantially again. Clearly the effect of copper is highly sensitive to adjustment factors, indicating that copper was being taken by an unusual population. I think the hazard ratio for copper is impossible to interpret without knowing why these women were supplementing copper. If we knew their situation, there would probably be an appropriate adjustment that would make a huge difference in mortality. I would say the numbers are too small, the population too skewed, and the information too limited to draw any conclusion here.

Overall, I would interpret the nine that showed benefits as being highly supportive of micronutrient supplementation. The fact that vitamin A, vitamin B complex, vitamin C, vitamin D, vitamin E, calcium, magnesium, selenium, and zinc all reduced mortality suggests that a well-formulated multivitamin would likely have reduced mortality.

Hazard Ratios After Multivariable Adjustment

Now, what about the “Multivariable Adjusted” results, which were responsible for the headlines?

We have to keep in mind a famous aphorism from the mathematician John von Neumann:

With four parameters I can fit an elephant, and with five I can make him wiggle his trunk.

The multivariable adjustments use 11 parameters and 16 parameters respectively. Using so many parameters lets the investigators generate whatever results they want.

I don’t think it’s a coincidence that both multivariable adjustments substantially increased the hazard ratio of every single one of the 15 supplements. The 11-variable adjustment increased hazard ratios by an average of 7%, the 16-variable adjustment by an average of 8.2%.

Rest assured, it would have been easy enough to find multivariable adjustments that would have decreased hazard ratios for every single one of the 15 supplements.

I believe it verges on the unethical that the variables chosen include dangerous health conditions: diabetes, high blood pressure, and obesity. These three health conditions just happen to be conditions that are often improved by supplementation.

Anyone familiar with how regression analyses work will immediately recognize the problem. The adjustment variables serve as competing explanations for changes in mortality. If supplementation decreases diabetes, high blood pressure, and obesity, and through these changes decreases mortality, the supplements will not get credit for the mortality reduction; rather the decreased diabetes, blood pressure, and obesity will get the credit.

Imagine we had a magic pill that completely eliminated diabetes, obesity, and high blood pressure, and reduced mortality by 20%, with no negative health effects under any circumstances. But if regression analysis showed that non-diabetic, non-obese, and non-hypertensive people had 25% less mortality, then a multivariable adjusted analysis would show that the magic pill increased mortality. Why? Because the elimination of diabetes, obesity, and hypertension should have decreased mortality by 25% (the regression analysis predicts), but mortality was only decreased 20%, so adjusted for diabetes, obesity, and hypertension the magic pill must be credited with the additional 5% dead. The multivariable adjusted HR for the magic pill becomes 0.8/0.75 = 1.067.

Of course, what ordinary people want to know is: Will this magic pill improve my health? The answer to that would be yes.

What (too many) scientists want to know is: Which methodology for analyzing this magic pill data will get me grant money? That depends on whether the funding authorities are positively or negatively disposed toward the magic pill industry. Once you know that, you search for the 16-variable multivariable regression that generates the hazard ratios the authorities would like to see.

My take? Judging by the data in Table 2 plus corroborating evidence from clinical trials reviewed in our book, I would say that a well-formulated supplement program, begun at age 62, may increase the odds of survival to age 82 by something on the order of 5% to 10%. Perhaps not a magic pill; but worthwhile.

[6] Not the weekly video: An exceptional magic show:

[7] Shou-Ching’s Photo Art:

[8] Weekly video: A new tool for stroke recovery:

Jimmy Moore’s seminar on “safe starches”: My reply

UPDATE: This was cross-posted on Jimmy’s site, so discussion is occurring on both sites.

I’d like to thank Jimmy for organizing this discussion on the desirability of including starches in a low-carb diet. (See: Is There Any Such Thing as “Safe Starches” on a Low-Carb Diet?). Not many people could bring such a roundtable together, and it’s an honor for us to be part of it.

I think unfortunately the discussion began with a few misunderstandings. So let me start with a few clarifications:

  • We advocate a low-carb diet. “Low-carb” to us means eating less than the body’s actual glucose utilization, so that a glucose deficit has to be made up by gluconeogenesis.
  • The concept of “safe starches” has nothing to do with their glucose content. “Safe starch” is a term of our invention and refers to any starchy food which, after normal cooking, lacks toxins, chiefly protein toxins. We do not consider glucose to be a toxin, though it may become toxic in hyperglycemia. Thus wheat, which includes gluten and various inhibitors of digestion that survive cooking, is an unsafe starch, while white rice, in which the known toxins (possibly excepting a recently discovered miRNA) are destroyed in cooking, is a safe starch. To say that something is a “safe starch” is not to imply that it is a desirable food for, say, a Type I diabetic.
  • Our “regular” diet is not specifically directed at diabetic or metabolically damaged persons. We have a basic diet that is designed for healthy people (represented in the apple – food plate) and we recommend modified versions of the diet for various health conditions – including diabetes.
  • We agree that diseases of metabolic derangement may benefit from lower carb consumption than our regular diet. This is especially the case in diabetes, if beta cell loss has reduced basal insulin levels and excessive gluconeogenesis is occurring. In this case, replacing protein rather than providing dietary carbs may be a more helpful strategy.
  • We agree that there is no single prescription that is optimal for every person. We often say, borrowing from Tolstoy, that “every healthy person is biologically alike, every diseased person is unhealthy in his own way.” Obesity, for instance, is a heterogeneous disease and there is not a single prescription that will be optimal for every obese person. Diseases of metabolic derangement raise rather complex issues which we explore regularly on our blog. The science remains somewhat unsettled. But we do favor a low-carb approach.

After reading all the responses, it seems to me the debate boils down to two primary questions:

  1. On low-carb diets, is it better to eat 400 carb calories per day, as we argue, or some lower number of carb calories, say 100 calories per day?
  2. Are “safe starches” the best source of carb calories?

After answering these I’ll respond to individuals.

Part I: Why Are 400 Carb Calories Better Than 100?

This is a pivotal claim of our diet and apparently the core issue of debate. Allow me to discuss the biology in some detail.

Glucose Utilization of the Human Body

Brain and nerves typically consume about 480 calories per day of glucose. Ketones can displace up to perhaps 60% of this, but ketones do not diffuse well into cortical areas of the brain and the brain always requires some glucose.

After 3 days of fasting, when the brain’s glucose consumption has been roughly halved by ketosis and the rest of the body is conserving glucose, the body’s rate of glucose manufacture in liver and kidneys is about 600 calories per day. [1]

Two things to note:

  • Even in fasting, peripheral utilization of glucose exceeds the brain’s.
  • The fasting level of glucose utilization is likely to be suboptimal for health: fasting invokes glucose-and-protein-conservation measures which evolved to make us more likely to survive famine, but almost certainly have a cost in long-term health. (The logic is similar to Bruce Ames’s triage theory [2].)

This fasting level of glucose production of about 600 calories per day is a key number: the body must obtain glucose at at least this level, either through diet or endogenous production, if it is to avoid a glucose deficiency.

When not fasting, the body’s glucose utilization is somewhat higher – say, 800 to 1000 calories per day for a sedentary person. Glucose needs are slightly reduced by some endogenous sources of glucose, such as from glycerol released from lipolysis of triglycerides or phospholipids. So the body’s net glucose needs are on the order of 600 to 800 calories per day.

As I noted above, we consider Perfect Health Diet to be a low-carb diet because we favor eating fewer carbs than the body utilizes. For most people, we suggest 400 to 600 carb calories per day, about 200 less than the body utilizes. The remainder is made up by gluconeogenesis – manufacture of glucose from protein. We are a slightly or moderately low-carb diet.

The Human Glycome

Why is so much glucose consumed outside the brain? Immune function (which may utilize significant glucose in people with infections) and glycogen replacement (high utilization in athletes) are two reasons that can be significant in some persons, but in the vast majority of people the biggest reason for glucose utilization is the construction and maintenance of the human glycome.

There are about 20,000 human genes and, due to transcriptional variants and manufacture of proteins from multi-gene subunits, about 200,000 human proteins. However, these proteins are subject to various post-translational modifications, chief of which is glycosylation. Over half of all human proteins need to be glycosylated for proper function, and such is the variety of ways in which they can be glycosylated that there are an estimated 2,000,000 compounds in the human glycome.

These glycosylated proteins coat the plasma membrane of all cells. For many proteins, only glycosylated forms are allowed to leave the endoplasmic reticulum and Golgi complexes where they are formed; nonglycosylated forms are ubiquinated and destroyed.

Nearly every major extracellular molecule has significant carbohydrate content. Glycosaminoglycans such as hyaluronan and proteoglycan components such as heparan sulfate and chondroitin sulfate are important building blocks of the extracellular matrix. Proteoglycans in general mediate all intercellular interactions.

All the body’s lubricating molecules are rich in carbohydrate. Mucins, the most important molecules in mucus, tears, and saliva, are predominantly composed of carbohydrate. Mucin-2, the dominant mucin of the intestine, is 80% sugar by weight.

Production of hyaluronan alone consumes 5 gm, or 20 calories, of glucose per day. [3] I have been unable to find detailed measurements of daily mucin production, but if mucin constitutes 1.5% of the 400 g daily stool weight, then it consumes 5 gm of glucose per day. Since gut flora can break down and metabolize mucin sugars, this may be an underestimate.

So: whole body measurements indicate peripheral glucose utilization of around 100 to 150 g (400 to 600 calories) per day in normal humans, and a mere two of the 2,000,000 carbohydrate-containing compounds in the human body account for nearly 10% of that.

Glucose Deficiency Symptoms

Several responders argued that there cannot be such a thing as a human glucose deficiency on very low-carb diets because blood sugar levels do not leave the normal range.

However, this argument may prove a bit too much, because blood sugar levels don’t leave the normal range during human starvation either, and yet it still proves fatal. Why, if cells can run on glucose and blood glucose remains normal, do starving people die?

A clue is the fact that starving people develop a hacking cough in their final weeks of life. Despite blood glucose levels in the normal range, they cease producing mucus and their airways become dry and irritated.

The reality is this: peripheral glucose utilization is not determined by blood glucose levels, but is hormonally regulated. The brain may import glucose passively, driven by a concentration gradient, but not so the rest of the body. During times of glucose scarcity, blood glucose levels are maintained to sustain brain and nerve function, but hormonal patterns change to prevent peripheral tissues from using glucose to make compounds like hyaluronan and mucin.

What are the hormones that regulate glucose utilization? This is an understudied area of physiology, but the primary regulators seem to be thyroid hormones. During glucose deficiency, T3 thyroid hormone levels decrease and reverse-T3 levels increase. I discussed this in a recent blog post (Carbohydrates and the Thyroid, Aug 24, 2011).

Decreased production of molecules like hyaluronan and mucin and reduced levels of T3 thyroid hormone, then, are outcome of dietary glucose deficiency. Pathologies this may produce include dry eyes, dry mouth, constipation or hard stools, and slow healing of scratch wounds.

Do Glucose Deficiency Symptoms Actually Occur in Low-Carb Dieters?

Yes.

I discussed the reduced mucus production very low-carb dieters sometimes experience in an early blog post (Dangers of Zero-Carb Diets, II: Mucus Deficiency and Gastrointestinal Cancers, Nov 15, 2010). Since that was published, well over 50 low-carb Paleo dieters have reported to me that dry eyes and other mucin deficiency symptoms were cured by adding safe starches to their diet.

I have put up a “Results” page which has case studies drawn mainly from the comment section of my blog. This includes many cases of glucose deficiency symptoms that developed on very low-carb Paleo or GAPS diets and were cured on our diet. (GAPS is a very low-carb diet.) Here is a sampling.

Angie:

All four people in my family experienced a variety of new symptoms (seasonal allergies, constipation, worsening of heartburn, bladder spasms, dry eyes, increasing tiredness and low energy) when we did GAPS. These problems didn’t resolve until we luckily stumbled upon PHD and added back safe starches.

Susan:

I’ve instituted “Paleo” in our house since 1/1/11. Very strict about only plants and protein. About 4/1/11 I realized I was experiencing extremely dry eyes and mouth. I read your post about glucose deficiency and added rice and potatoes back into our diet. This cleared the problem up within 3 days and I was super grateful.

Melinda:

I had severe dry eyes while eating too low carb. Following Dr. Paul’s recommendations at “Perfect Health Diet”, I upped my carbs to his minimum of 50 grams of starch per day and the dry eyes went away.

There are many more cases; in addition to those on my “Results” page, many anecdotes can be found on PaleoHacks and in my comment thread.

What is the incidence of such deficiency symptoms on low-carb diets? At the Ancestral Health Symposium, two dozen people came up to Shou-Ching and I and told us their health had been improved by adding safe starches to their low-carb Paleo diets. As this was about 5% of conference attendance of 500, and not all people at the conference were low-carb and only a minority had tried our diet, I think it’s a safe bet that at least 20% of people who eat very low-carb diets will experience overt glucose deficiency symptoms.

Another Low-Carb Risk: Impaired Immunity

Low-carb diets generally improve immunity to bacteria and viruses, but not all is roses and gingerbread.

Low-carb diets, alas, impair immunity to fungal and protozoal infections. The immune defense against these infections is glucose-dependent (as it relies on production of reactive oxygen species using glucose) and thyroid hormone-dependent (as thyroid hormone drives not only glucose availability, but also the availability of iodine for the myeloperoxidase pathway). Thus, anti-fungal immunity is downregulated on very low-carb diets.

Moreover, eukaryotic pathogens such as fungi and protozoa can metabolize ketones. Thus, a ketogenic diet promotes growth and systemic invasion of these pathogens.

As the fungal infection case studies on our “Results” page illustrate, low-carb dieters often develop fungal infections, and these often go away with increased starch consumption.

Another issue is that mucus is essential for immunity at epithelial surfaces, and glycosylation is essential for the integrity of cellular junctions and tissue barriers such as the intestinal and blood-brain barriers. Thus, reduced production of mucus can impair intestinal immunity and promote gut dysbiosis or systemic infection by pathogens that enter through the gut.

Finally, a very low-carb diet is not entirely free of risks of gut dysbiosis, and not just from fungal infections. Bacteria can metabolize the amino acid glutamine as well as mucosal sugars, so it is not possible to completely starve gut bacteria with a low-carb diet. Nor is it desirable, as this would eliminate a protective layer against systemic infection by pathogens that enter the body through the gut. As our “Results” page shows, several people who had gut trouble on the very low-carb (and generally excellent) GAPS diet were cured on our diet.

The Possibility of Slow-Developing Problems Cannot Be Ruled Out

The majority of very low-carb dieters may experience no immediate ill effects. However, this does not guarantee that problems cannot develop over time.

Is it possible that peripheral downregulation of glucose utilization may increase the risk of some chronic diseases? There is too little experience with very low-carb diets to answer this question, but I think no biomedical scientist would exclude the possibility.

Biomedical researchers are gradually realizing the importance of glycosylation defects in leading diseases. I’ve mentioned previously that downregulation of glycosylation is an important part of the cancer phenotype (see An Anti-Cancer Diet, Sep 28, 2011; Dangers of Zero-Carb Diets, II: Mucus Deficiency and Gastrointestinal Cancers, Nov 15, 2010). A few papers:

  • N- and O-glycosylation of proteins in Golgi bodies is impaired in cancer cells. [4]
  • Cancer cells have systematically incomplete glycosylation, including deficient galactosylation of terminal beta-N-acetyl-D-glucosamine residues. [5]
  • Genetic defects in O-glycans production increase cancer susceptibility. [6]

A recent report in Nature Medicine found that a specific glycosylation defect may commonly underly Type 2 diabetes. [7] From the abstract:

[A] deficit of GnT-4a glycosyltransferase expression in beta cells … produced signs of metabolic disease, including hyperglycemia, impaired glucose tolerance, hyperinsulinemia, hepatic steatosis and diminished insulin action in muscle and adipose tissues. Protection from disease was conferred by enforced beta cell-specific GnT-4a protein glycosylation and involved the maintenance of glucose transporter expression and the preservation of glucose transport. We observed that this pathogenic process was active in human islet cells obtained from donors with type 2 diabetes … [7]

I report these papers, not because I think they tell us how many carbohydrates we should eat – they don’t – but to remind everyone of the complexity of biology.

We lack data on the long-term effects of very low-carb diets. On the Standard American Diet, many diet-induced diseases do not show up for 40 to 50 years. Very low-carb diets have become popular only in the last few years. We cannot be sure that there may not be negative health effects from severe carb restriction that will show up only after decades.

I am not saying that such insidious health effects exist. I am only saying that while I believe low-carb is good, I don’t believe that very low-carb is better, and I think everyone should acknowledge that very low-carb diets may have unexplored risks.

In conclusion: Moderation in carb-hostility is no vice.

Part 2: Are “Safe Starches” Healthy Carb Sources?

So far I’ve defended our recommendation of 400 carb calories per day. Now we reach the question of which plant foods should provide them.

The main choice is between starchy plants and sugary plants. Sugary plant foods typically provide a mix of glucose and fructose; starches digest entirely to glucose.

Loren Cordain, whose “Paleo Diet” recommends a carb intake similar to or larger than ours, favors sugary plants:

[A]nyone who advocates eating white rice and potatoes obviously is unaware of the concept of either glycemic index or glycemic load … Yams, sweet potatoes, plantains and berries are healthful carb sources that most people can eat without a problem.

Yams, sweet potatoes, plantains and berries – all, by the way, foods our diet recommends and that we eat ourselves – contain some sugars which digest to a mix of glucose and fructose, while rice and potatoes contain starches which digest to glucose alone.

Because the concepts of “glycemic index” and “glycemic load” refer to blood glucose levels, they are sensitive to the glucose content of food, not the fructose content. Pure fructose has a glycemic index of only 19, compared to 100 for glucose.

We favor starchy plants over sugary plants for several reasons:

  • Nutritional value. Glucose is more nutritious because, as noted above, it has structural uses throughout the human body. Fructose has no structural uses.
  • Toxicity. Glucose is less toxic than fructose for several reasons. First, it is less reactive, less likely to glycate (fructate) proteins or promote lipid peroxidation. Second, Paracelsus’s rule tells us that the “dose makes the poison.” Dietary glucose is distributed via the blood throughout the body, so that levels are low in any one location. Fructose, however, is concentrated in the liver.

The body’s evolved machinery for handling glucose and fructose is a good indicator of their relative healthfulness. Glucose is treated by our evolved physiology as a non-toxic nutrient: it is allowed free entry to the blood where it is accessible to all cells of the body. Fructose is treated by our evolved physiology as a toxin: it is shunted to the liver where it is rapidly disposed of.

The toxicity of fructose is well supported by a host of biochemical, biomedical, and epidemiological data. In general, the more fructose people consume, the worse their health. Dr. Robert Lustig spoke at the Ancestral Health Symposium on this topic.

While I think glucose should be favored over fructose, I don’t want to exaggerate the dangers of limited fructose consumption: fruits, berries, and other sugary plants are, in moderation, fine components of a healthful diet. But I see no obvious reason to tout them as superior to starchy plants.

Glycemic Index and Load in Dietary Context

I do not believe that “glycemic index” or “glycemic load” are sufficient indexes of the healthfulness of foods.

Glucose, as I’ve been arguing, is a nutrient: it has beneficial uses in the body. Nutrients generally deliver their greatest benefits when the body is deficient in them; few benefits when the body is replete; and often become toxic at high doses. Here is a figure from our book (p 4):

A “glycemic load” can be understood as a bolus of glucose delivered to the body. In a condition of glucose deficiency, a “glycemic load” is likely to be highly beneficial: it will be nourishing and repair the nutrient deficiency.

At higher levels of carb intake, a “glycemic load” is likely to be health-neutral – neither damaging nor beneficial. At very high carb intakes, a “glycemic load” may become dangerous.

So knowing a plant’s “glycemic index” or “glycemic load” cannot tell us whether it is good to eat some. That depends on the context of the rest of the diet. On a low carb diet, a safe starch is likely to be nourishing, regardless of its glycemic index.

Issues of Glycemic Control

In interpreting the safety of glucose, there is also the issue of whether postprandial increases in blood sugar can create transient toxicity effects. What is a dangerous level of blood glucose?

In diabetics, there seems to be no detectable health risk from glucose levels up to 140 mg/dl, but higher levels might have risks. Neurons seem to be the most sensitive cells to high glucose levels, and the severity of neuropathy in diabetes is correlated with how high blood glucose rises above 140 mg/dl in response to a glucose tolerance test. [8] In people not diagnosed with diabetes, there is also some evidence for risks above 140 mg/dl. [9]

For several reasons brief excursions above 140 mg/dl are probably not a problem for healthy people. However, for purposes of argument I’ll stipulate that a blood glucose level over 140 mg/dl probably does some mild harm.

Does eating a safe starch necessarily raise blood glucose above this level? No.

I offer as Exhibit A the experience of Haggus Lividus on Jimmy’s thread. Haggus measured blood glucose levels after consuming ~100 calories of rice and found that blood glucose levels peaked at 7.7 mmol/l = 139 mg/dl. Within an hour and fifteen minutes they were back at 5.8 mmol/l = 104 mg/dl. After sweet potatoes, blood glucose peaked at 6 mmol/l = 108 mg/dl.

These are safe levels of blood glucose – below 140 mg/dl at all times. Yet Haggus Lividus took these as levels to be unsafe!

Tom Naughton reports that a potato raises his blood glucose level to 175 mg/dl. This is, indeed, an unsafe blood glucose level.

But he eats a very low-carb diet, and very low-carb diets induce hormonal changes that lead to glucose conservation. One result of these changes is insulin resistance and impaired glucose tolerance.

Thus, an isolated glucose tolerance test is not necessarily a fair test of glycemic control in a very low-carb dieter. Were Tom to eat 400 calories per day from safe starches for a week, he might find his glycemic control was considerably improved. Or, he may find that he is somewhat diabetic and intolerant of carbs in all circumstances.

What is a normal blood glucose response to consumption of a starchy meal? Here is a view of blood glucose levels in normal people as measured by Professor JS Christiansen (from Ned Kock via CarbSane):

Although a majority maintain blood glucose levels below 140 mg/dl at all times, it is not unusual for blood glucose levels to enter the range 140 to 165 mg/dl for brief periods after meals. These measurements were all done in healthy young people.

Vegetables as Poor Glucose Sources

Some responders were understandably confused by a line Jimmy quoted out of context from our book: “don’t count vegetables as as a carb source – they are a fiber (and therefore a fat) source” (page 45).

The point is that vegetables are not usually helpful in repairing a glucose deficiency. A typical vegetable has about 80 carb calories per pound, half as glucose and half as fructose. The digestive tract typically consumes about 50 calories of glucose in digesting a pound of vegetable matter, due to intestinal and immune utilization. Some fructose may be converted to glycogen and then to glucose, but some may be converted to fat and much may be intercepted by gut bacteria. Fructose malabsorption is a widespread problem. So the net contribution of vegetables to the body’s glucose status is small and may be negative.

Since we recommend counting calories only for a few days until one learns how much one must eat to obtain our recommended 400 calories per day of glucose, there is no reason to include vegetables in calorie counting. Vegetables are recommended in our diet due to their micronutrient and fiber content, not their carbohydrate content.

Improved Weight Loss with Consumption of Safe Starches

Since many of Jimmy’s readers eat low-carb diets in the hope of losing weight. It may be of interest to them to know that some of our readers have experienced easier weight loss, reduced appetite, and diminished food cravings after adding “safe starches.” Our “Results” page has examples.

Part 3: Specific Replies

Readers may wish to open Jimmy’s post, Is There Any Such Thing as “Safe Starches” on a Low-Carb Diet?, in another window to follow along.

Colette Heimowitz does not seem familiar with our work, and to have misunderstood the basis for our recommendation of a modest amount of starch. We do not come from a “glucose mentality” and agree that fat and ketones are fine metabolic fuels. However, ketones do not eliminate glucose needs.

The fact that glucose can be formed via gluconeogenesis does not prevent the emergence of glucose deficiency conditions, because the degree of gluconeogenesis is hormonally controlled and may be insufficient to maintain all normal glucose functions.

Maintainance of blood sugar is not an indicator that there is no glucose deficiency.

Glycation is one thing, glycosylation and manufacture of GAGs and other glucose containing structural molecules of the human body is another. We agree that glycation is bad.

I’d like to thank Robb Wolf for his point of view, which is quite reasonable. I am not asserting that no one can do well on a very low-carb diet, only that as carb consumption approaches zero risks of health problems increase. That Robb himself experienced problems on sustained very low-carb is a helpful data point.

I’d like to thank Chris Masterjohn for his contribution. I think Chris has read enough of our work to know that we recommend ketogenic diets as a therapy for various conditions, including neurological disorders of all kinds, and generally hold that dietary adjustments are desirable in many health conditions. So we do not consider that a single macronutrient ratio applies to everyone, but we do believe that intolerance of a “normal” macronutrient ratio is diagnostic of a dysfunction of some kind.

Chris is quite right that it’s a “safe[r] bet” to meet the body’s physiological need for glucose in part by eating glucose. This reduces the risk of failing to provide adequate glucose for optimal cellular and extracellular function.

I’d like to thank Dr. Kurt Harris for his contribution. I discussed Dr. Harris’s post on my blog: https://perfecthealthdiet.com/?p=4802.

Dr Jonny Bowden makes an excellent point: a major advantage of starches over other carbohydrate sources is their lack of fructose. Glucose is, in general, a safer carb source than fructose.

Dr Robert Su directs us to an essay of his, which makes a lot of points that I agree with, but his evidence doesn’t imply the conclusion that all carbs should be excluded, nor does it address the main issues of our diet.

Tom Naughton and I share Irish ancestry, so if he has been extinguished due to lack of ancestral potatoes then so have I. Luckily for both of us, failure to consume safe starches, if that is what our ancestors did, is not so damaging to health as to necessarily result in early death and failure to leave descendants.

That his blood glucose rises to 175 mg/dl after consuming a potato indicates one of two things: his glucose regulation is irretrievably broken and he must never again eat a whole potato in isolation from other foods, or he is insulin resistant in order to conserve glucose and he should eat carbohydrates more often to improve his insulin sensitivity. Which is his optimal course is not something I can know.

Dr Richard Feinman may not have noticed but Shou-Ching and I were at the Ancestral Health Symposium and so were dozens of people following our diet; indeed, about two dozen people came up to us and told us that our diet had improved their health. The most frequently cited benefit was feeling better after adding safe starches to the diet, with relief of dry eyes the most common symptomatic improvement. So if symposium attendees were not dropping like flies, perhaps we deserve a bit of the credit.

Dr. Loren Cordain’s assertion that eating sugary plants like yams, sweet potatoes, and berries is preferable to eating starchy plants like rice and potatoes may be a defensible position, but we believe the evidence is strong that glucose is preferable to fructose as a carb source, and does not support the notion that rice or white potatoes are intrinsically dangerous foods.

Dana Carpender links to one of Mike Eades’s best posts, which we cite and quote in our book’s discussion of why wheat bran is unhealthy. However, it in no way rebuts our observations about the negative health effects of a deficiency of mucus arising from a glucose deficiency.

We agree with Dana that turnips, rutabaga, Jerusalem artichokes, and jicama are fine foods.

Anonymous Prominent Member makes a good point: adding carbs back into a very low-carb diet worked for me, but may not work for everyone. I agree with Anonymous Prominent Member’s point about the importance of practical experience. I think this is one of our greatest strengths. Thousands of copies of our book have been sold, and hundreds of people have reported results back to us. These reports have been overwhelmingly positive. On the blog, I answer questions from people with health problems, ask them to report back results, and many return weeks or months later to report cures or improvements. I invite Anonymous Prominent Member to review the case studies on our “Results” page.

Dr. Uffe Ravnskov can find the scientific studies in support of our views on our blog and in our book. Nowhere do we assert that it is impossible to survive on a zero-carb diet. Rather we assert that a zero-carb diet is suboptimal for health, and not robust to certain health problems, such as some infections.

I would like to thank Adele Hite for her generous statements that our “overall approach is very reasonable” and “may be useful to many people,” and for her engagement on issues of substance.

Adele links to Mike Eades’s excellent fiber post, which we cite approvingly in our book; see my comment to Dana Carpender. The issue Mike discussed, of an excess of mucus due to intestinal injury, is unrelated to the issue we discuss, of a mucus deficiency due to glucose deficiency.

About vitamin C, I think Jimmy may have given this issue quite a bit more prominence than it deserves. It happens that the incidence of kidney stones, glutathione deficiency, and vitamin C deficiency is increased on very low carb ketogenic diets for epilepsy, and other very low carb diets. I made a speculative post attempting to guess the causes of this. To answer Adele, part of the issue is likely a protein deficiency: the need to utilize protein for gluconeogenesis may induce a protein deficiency on an otherwise adequate dietary intake. Other factors are that vitamin C degrades through a pathway that generates oxalate in the kidneys, a risk factor for calcium oxalate stones. Vitamin C does indeed share insulin-dependent receptors with glucose, which implies that glucose competes with C but also that insulin promotes C entry into cells for recycling, so the overall effect of consuming carb-rich foods is unclear. On a low-carb diet adding a little dietary glucose is unlikely to be pro-inflammatory.

About cancer, this is an interesting scientific question. I’ve explained above why a glucose deficient diet can downregulate production of glycoproteins and other structural glucose-containing compounds. However, cancers often evolve an ability to take in glucose independently of insulin and other hormones that regulate glucose utilization in normal cells. As a result, one could argue that things would run the opposite way than Adele proposes: reducing dietary glucose, which generally does not reduce blood glucose levels, will not affect cancer metabolism, but will limit availability of glucose to normal cells for structural use.

I would like to thank Dr. Larry McCleary for addressing matters of scientific substance in a well-reasoned comment. He is quite right that cancers disable glycosylation by suppressing enzymes involved in it. Our reasoning, admittedly speculative, is that (a) the cancer cellular phenotype is a wayward phenotype characterized by reduced intercellular cooperation, cooperation that is largely mediated through glycoproteins, proteoglycans, and glycosylated proteins; (b) cells evolve the cancer phenotype in part by disabling the enzymes which glycosylate proteins; (c) therefore (the speculative inference) dietary steps which downregulate glycosylation may inadvertently serve to entrench or promote the cancer phenotype. This is speculative science, but speculation is the first step in scientific discovery.

Dr McCleary is quite right that depriving cancer cells of glucose is an attractive therapeutic strategy for cancer. However, except in the brain (where ketogenic dieting can significantly reduce glucose levels) this is a difficult strategy to implement. Blood glucose levels are maintained even through the late stages of starvation, and cancer cells can evolve insulin-independence and the ability to import glucose massively from blood. Paradoxically, eating some dietary carbs can even decrease average 24-hour blood glucose levels by increasing insulin sensitivity in normal cells.

I would like to thank Chris Kresser for an excellent comment sharing his clinical experience:

In cases where there is no significant metabolic damage, when I have these folks increase their carbohydrate intake (with starch like tubers and white rice, and fruit) to closer to 150g a day, they almost always feel better. Their hair loss stops, their body temperature increases and their mood and energy improves.

For people that are overweight and are insulin/leptin resistant, it’s a bit trickier. In some cases increasing carbohydrate intake moderately, to approximately 100g per day, actually re-starts the weight loss again. In other cases, any increase in carbohydrate intake – in any form – will cause weight gain and other unpleasant symptoms.

This corresponds precisely with our recommendations. Healthy people will do best on 100-150g per day; obesity is a heterogeneous disease and some will do best on a carb intake in that normal range, others (especially those who are more diabetic) will do best on very low-carb diets. Our “Results” page includes feedback from a number of people who lost weight on our diet better than on other low-carb diets.

I would like to thank Dr David Diamond for a thoughtful comment and for taking the time to read our blog. It is gratifying that he eats largely in accord with our recommendations and has had good results: “This has been my basic diet plan for the past 6 years and my blood lipids have responded in the right directions and I’ve lost about 25 lbs.”

Nowhere do we assert that slipping to 300-400 carb calories is dangerous; rather this is in our “safe range” of 200 to 600 carb calories per day and very close to our estimated optimum. However, I do think that for healthy people the potential harms from very low-carb are greater than the potential harms from excessive carb consumption, so it is perhaps safer to advise eating in the upper end of the range, since a large number of people will deviate from their target.

Dr. Diamond notes that “I haven’t actually seen adverse health outcomes for most people who eat 50-100 gm of carbs/day.” Interestingly, 50 g is sort of a magic number of carbs for many people: there are adverse health outcomes eating less than 50 g, but intake of 50 g or more tends to eliminate them. Our comment threads, and other sites such as PaleoHacks, are full of people who have reported this experience. So I would agree with Dr. Diamond’s statement, but argue that it supports our recommendation to eat at least 50 g of safe starches.

I’ve discussed the cancer issue elsewhere, but I appreciate Dr Diamond’s contribution.

Livin’ La Vida Low-Carb Reader’s carb intolerance is a difficult problem to deal with; I sympathize, and largely agree with what LLVLCR says. LLVLCR may wish to read my reply to Tom Naughton; I would say something similar in LLVLCR’s case. Low-carb is good, control of blood glucose is good, but it is not obvious that zero-carb is optimal.

I agree with Dr. Andreas Eenfeldt that those with diabetes and metabolic syndrome may do better with lower carb intake than is optimal for healthy people.

Dr. Eenfeldt may wish to visit our “Results” page to learn about the mucus deficiency issue on very low carb. It can generally be healed with the addition of 50 g starch to the diet; sometimes vitamin C supplementation is needed as well.

As noted elsewhere, blood glucose levels are not an indicator of the body’s glucose status, and will remain normal even when there is a serious glucose deficiency. Production of glycoproteins such as mucin is a much more sensitive indicator of whole-body glucose status.

Dr. Jeff Volek should be aware that if “there is no defined condition associated with not consuming carbs,” it may be because biomedical scientists have spent little to no time observing people who do not consume carbs. Dr Volek may consult our “Results” page for examples of people who have developed adverse health conditions from very low-carb dieting.

I’d like to thank Dr Jeffrey Gerber for sharing his very interesting clinical experience with cancer patients:

Patients who are ill such as cancer, post surgical, after the hospital are stressed and their basic metabolic rate is increased. In this situation I have found that there is an increased caloric demand. Patients require more calories from fat protein and carbs.

Cancer is a very complex disease and Dr. Gerber’s experience is a helpful reminder that knowledge of the Warburg effect, while helpful for understanding cancer, is not sufficient knowledge to design an anti-cancer diet.

Dr. Jack Kruse’s only substantive sentence is this: “I think avoiding anything that stimulates the IGF1 pathway is ‘smart’ based upon current knowledge and i think using a ketogenic diet is also prudent.”

The dominant dietary factors stimulating IGF-1 release are “protein and energy intake … and energy intake may be of greater importance.” Our diet is generally lower in protein than other low-carb diets, and as a nourishing diet with macronutrient intakes near the body’s utilization needs, it is highly effective at minimizing appetite and total energy intake, as perusal of our “Results” page will show.

Using a ketogenic diet is sometimes prudent. We recommend a ketogenic diet for many neurological disorders and brain cancers, and readers have used our version of the ketogenc diet to cure migraines and ameliorate genetic diseases such as Neurodegeneration with Brain Iron Accumulation (see our “Ketogenic Diet” category for more). We also recommend practices that introduce ketosis intermittently, such as daily intermittent fasting, to everyone as a good general health practice.

Since our diet minimizes IGF-1 and is frequently ketogenic, I would have expected Dr. Kruse to be more positive. Perhaps his reaction may have been just a reflex: more IGF1 reduction, more ketosis, more cowbell. Or perhaps he favors the ultimate in low-IGF1, high-ketosis diets: the Terri Schiavo diet.

Dr. Fred Pescatore should read our book. It is not true that 200 calories of starch will necessarily take a person out of ketosis. Consumption of medium chain triglycerides or coconut oil in conjunction with starches will trigger a mild ketosis, see Ketogenic Diets, I: Ways to Make a Diet Ketogenic, Feb 24, 2011.

Glycosylation of proteins occurs primarily intracellularly in the endoplasmic reticulum and Golgi bodies, not on cell membranes.

I thank Dr. Eric Westman for looking at our web site and trying to understand our diet. Hopefully this response will have made things clearer.

Peter Dobromylskyj of “Hyperlipid” has had a very busy year with a new daughter and new home, so I’m not in the least surprised that he hasn’t yet had time to read our book. I hope he will enjoy it when he does, as he is one of my favorite health writers.

The human glycome is much more than a lectin signaling system: it has a myriad of structural and functional roles, some of them discussed above.

Re “I can’t see glucose deficiency being a gut problem as this is the organ with the highest exposure to dietary glucose,” two factors which limit availability of dietary glucose to gut cells are (a) dietary glucose is absorbed in the small intestine but gut problems are most common in the colon where bacterial populations are highest, and (b) we are considering very low carb diets that provide little dietary glucose. A third factor to consider is that the gut, due to its mucin production, immune activity, and rapid turnover in cells and extracellular matrix, is a major consumer of glucose.

Cancer is an extremely complex and interesting disorder and I’ll be delighted to hear Peter’s ideas.

Peter makes an extremely important point: that minor dietary defects may take decades to reveal themselves. On the Standard American Diet, an unhealthy diet, it often takes 50 years for chronic diseases to appear. If there are problems with very low carb diets, we should not necessarily expect them to appear immediately.

Peter says “I don’t know,” but in truth we all don’t know: dietary science is complex and all of our positions are somewhat speculative. Thus humility is in order.

I thank Dr. William Davis for his assessment that our “diet seems a rational, workable program” and agree with him that diabetics will benefit from reducing starch consumption.

Valerie Berkowitz should be aware that we do recommend tomato consumption, but we do not consider it a “safe starch” because its calories are mainly in the form of sugars. Her other concerns are addressed above.

We agree with Diane Sanfilippo’s observations.

Dr William Yancy seems to be intelligent, reasonable, and unfamiliar with our diet. Perhaps the exposition above will help.

Dr Ann Childers links to an article in Discover magazine and avers that the humans of the Ice Age and the Inuit were “without cancer, diabetes, tooth decay, glutathione deficiency, vitamin C deficiency or gut dysbiosis.” These claims are unsupported. Nor is it the case that Ice Age humans ate zero-carb diets, nor any other humans who had access to starchy plants.

Dr Cate Shanahan makes two important points with which we wholeheartedly agree.

First is her observation that the protein quality of food, especially the presence of immuno-reactive proteins, is extremely important for health. Indeed, our “safe starches” are defined by their lack of these toxic or immunogenic proteins. We strongly agree with this point, and it is a centerpiece of our diet.

Second is the point that just because it is possible to manufacture glucose from protein does not mean that optimal amounts of glucose will actually be manufactured if none are eaten. It is well established that macrobiotic dieters, who eat low-fat diets, can develop lipid deficiencies, notwithstanding the fact that lipids can be manufactured from glucose. Something similar can happen on very low-carb diets, especially if dietary protein is insufficient.

Amy Kubal seems to be under the misimpression that I recommend 1 pound of safe starches daily for “everyone.” No, this is a recommendation for healthy people, I understand that some people with defects of metabolic regulation or neurological disorders will benefit from ketogenic diets or severe carb restriction.

I agree with her advice about the benefits of carbs following workouts. The reason we recommend not counting carb calories from vegetables was discussed above.

It is not obvious to me from her description that her recommended cancer diet differs much from ours.

Dr Robert Su refers us to a column of his. It makes a lot of points whose truth I acknowledge, but doesn’t address any of the arguments I’ve made, and certainly doesn’t support the conclusion that there is no benefit from dietary carbohydrate.

I applaud Mark Sisson’s comment. Primal and Perfect Health Diet are indeed extremely close, and Mark properly focuses on the important points, such as avoiding grains, fructose, and seed oils. Mark’s easygoing attitude toward unimportant differences is praiseworthy.

Dr Lauren Noel notes that other than a few minor cell types, “all tissues can run on ketones,” and supposes this refutes the need for dietary carbohydrate. However, although the brain can run on ketones, it turns out that ketones don’t diffuse well to the cortical areas of the brain, and the brain always requires some glucose even in extreme ketosis. Also, while ketones can replace glucose as a fuel, they cannot glycosylate proteins, or generate ROS in the manner needed by immune cells.

Dr Noel believes that eating white rice and sweet potatoes will aggravate Candida infections. Dietary carbs can feed Candida in the gut, but they also feed competing probiotic bacteria and promote intestinal barrier integrity and immune function, and thus their effect on the gut flora is complex. More importantly, ketosis promotes systemic invasion by Candida and glucose is needed for the immune defense to Candida, so a moderate carb intake is helpful to the defense against systemic Candida. As Candida is an effective intracellular pathogen that can flourish systemically, this is a very important consideration. No one with a Candida infection should eat a ketogenic diet. Dr Noel might wish to consult our “Results” page for a few cases in which fungal infections were exacerbated on very low-carb diets and cured on our diet.

Dr Daniel Chong is quite right that starches have been a part of the evolutionary human diet, since at least Australopithecus 3.5 million years ago. The history may go back even farther: recent anthropology speculates that the common human-chimp ancestor may have been bipedal and lived in open woodlands where starches but not sugary fruits were the predominant food.

Dr Greg Ellis is rather quick to assert that our work is “made up” and “constructed out of thin air” even though he acknowledges not having read our book, and is under the misimpression that we have “bought into the dangers of fat and cholesterol.” He asserts, “If you want to talk about toxins then glucose is at the top of the list” which is absurd; among sugars alone, fructose is more toxic than glucose. He asserts, “If glycosylation is truly important there is enough glucose available to perform this function without eating glucose or carbs” which is precisely the point at issue. He blames cancer on glycation of proteins, a highly dubious claim. He is unaware that fungi are eukaryotic organisms that have mitochondria.

Dr Ron Rosedale has written an extended commentary which deserves a considered response. Since he posted a series on Facebook to which I had already begun drafting a reply, I’ll finish that and post it on my blog next week. I thank Dr Rosedale for the time he’s given to this discussion.

Dr Joe Leonardi’s comments are intelligent, and it sounds as though his own dietary advice is excellent. I thank him for his contribution.

Dr BG makes an excellent point: that carbs do in practice improve the health of many paleo dieters, in part via improving adrenal function. Dr BG herself reports, “I feel ‘better’ on higher carbs for the adrenals.” Dr BG also notes, “On Paleohacks there are countless stories of people on VLC paleo who feel dizzy or lightheaded. H-E-L-L-O this is cardinal signs and symptoms of adrenal fatigue. Many of these folks are also doing HIIT and hard core CROSSFIT!” Of course, exercise utilizes glucose and will exacerbate any glucose deficiency.

These cases of improved health upon higher carb consumption should be a warning to those other writers who question whether it’s possible to have a glucose deficiency.

Zoe Harcombe appears to approach dietary science from premises similar to ours. She shares our nutrient-based view and general orientation.

On the issue of taste, we do recommend that starches be eaten as part of a meal in combination with sauces, vegetables, fats, and meats. So yes, rice will often be combined with curry. In our “Food Plate,” the body of the apple signifies foods that are best eaten as part of a meal – starches, vegetables, meats, soups, sauce – and the “pleasure foods” are good snacks or desserts.

Our bodies do need glucose, and it may be preferable to obtain it directly from diet than to have to manufacture it from protein.

An appropriate population of commensal bacteria tends to stabilize the gut and make it resistant to dysbiosis. Antibiotics, starvation of carbohydrates, and other factors that deplete gut bacteria may increase the risk of fungal or other infections.

We do recommend lower carb consumption for diabetics.

The amount of glucose in blood is not related to the amount of glucose the body consumes in a day. The “stock” of glucose in blood is continually replenished by a “flow” from the liver as tissues draw it down. It is the flow, integrated over 24 hours, which is the daily glucose consumption.

The Taubesian idea of intentionally creating a glucose deficiency to force the body to breakdown triglycerides for glycerol is a clever but flawed strategy for weight loss. Its chief defect is that triglycerides break down to about 11% glucose by calories, but the body’s glucose utilization is close to 30% of energy. As a result, this strategy cannot meet glucose needs without releasing free fatty acids beyond energy needs. If these are not successfully disposed of, then blood free fatty acid levels may become elevated, which leads to the phenomenon of “lipotoxicity” which can promote diabetes. Whether and to what degree glucose deficiency and lipotoxicity would occur in any attempt to execute such a strategy is an empirical matter, but no reader should assume that such a strategy is riskless.

There is room to disagree about the optimal level of glucose intake, and I hope Zoe will look into our arguments for a slightly higher carb consumption than she is used to.

Dr Stephen Phinney seems to be under the misimpression that my term “safe starches” refers to low glycemic index foods. No, it has nothing to do with the carbohydrate; “safe” means that after cooking the food lacks toxic, bioactive, or immunogenic proteins. It is about the plant proteins, not the carbs.

Dr Phinney avers that “there is no absolute human requirement for dietary carbohydrate.” I am not sure what “absolute” means, but I do believe that health will usually be improved if the diet includes some carbohydrate.

Dr. Phinney defends his statement by reference to blood sugar levels. As discussed above, blood glucose levels are not an adequate indicator of the body’s glucose status.

Re the issue of vitamin deficiencies, there are plenty of reports of nutrient deficiencies on clinical ketogenic diets, thus Dr Phinney’s need to include the adjective “well-formulated” before ketogenic diets. I agree with him on this point: it is possible to formulate ketogenic diets in such a way that they don’t generate nutrient deficiencies. However, it is perilously easy to misformulate them. Diets should be robust to error. When carbohydrate intake approaches zero, diets become less robust. Since few people know how to properly formulate a ketogenic diet, this has to be considered a risk to low carb diets.

On the issue of dysbiosis, I assume Dr Phinney will agree that some non-zero level of mucus production is optimal, and that a level of mucus production below that optimum impairs health.

Dr Richard Bernstein is the author of a book we frequently recommend to diabetics, so it’s unfortunate that he may have gotten the mistaken impression we recommend higher carbohydrate consumption for diabetics. Perhaps he’ll look more closely into our diet and reconsider his judgment.

References

[1] Nair KS et al. Leucine, glucose, and energy metabolism after 3 days of fasting in healthy human subjects.  Am J Clin Nutr. 1987 Oct;46(4):557-62. http://pmid.us/3661473.

[2] McCann JC, Ames BN. Adaptive dysfunction of selenoproteins from the perspective of the triage theory: why modest selenium deficiency may increase risk of diseases of aging. FASEB J. 2011 Jun;25(6):1793-814. http://pmid.us/21402715. McCann JC, Ames BN. Vitamin K, an example of triage theory: is micronutrient inadequacy linked to diseases of aging? Am J Clin Nutr. 2009 Oct;90(4):889-907. http://pmid.us/19692494.

[3] Stern R. Hyaluronan catabolism: a new metabolic pathway. Eur J Cell Biol. 2004 Aug;83(7):317-25.  http://pmid.us/15503855.

[4] Hassinen A et al. Functional organization of the Golgi N- and O-glycosylation pathways involves pH-dependent complex formation that is impaired in cancer cells. J Biol Chem. 2011 Sep 12. [Epub ahead of print] http://pmid.us/21911486.

[5] Satomaa T et al. Analysis of the human cancer glycome identifies a novel group of tumor-associated N-acetylglucosamine glycan antigens. Cancer Res. 2009 Jul 15;69(14):5811-9. http://pmid.us/19584298.

[6] An G et al. Increased susceptibility to colitis and colorectal tumors in mice lacking core 3-derived O-glycans. J Exp Med. 2007 Jun 11;204(6):1417-29.  http://pmid.us/17517967.

[7] Ohtsubo K et al. Pathway to diabetes through attenuation of pancreatic beta cell glycosylation and glucose transport. Nat Med. 2011 Aug 14;17(9):1067-75. http://pmid.us/21841783.

[8] Singleton JR et al. Increased prevalence of impaired glucose tolerance in patients with painful sensory neuropathy. Diabetes Care. 2001 Aug;24(8):1448-53. http://pmid.us/11473085. Hat tip Jenny Ruhl, http://www.phlaunt.com/diabetes/14045678.php.

[9] Ziegler D et al. Prevalence of polyneuropathy in pre-diabetes and diabetes is associated with abdominal obesity and macroangiopathy: the MONICA/KORA Augsburg Surveys S2 and S3. Diabetes Care. 2008 Mar;31(3):464-9. http://pmid.us/18039804.

An Anti-Cancer Diet

Our cancer series resumes today with some tentative advice for cancer patients. (Note: This post is designed for solid tumor cancers, not blood cancers. However, most of the advice would also be applicable to blood cancers.)

This series began with Toward an Anti-Cancer Diet (Sep 15, 2011). There we advocated trying to shift cells away from the cancer phenotype via 8 anti-cancer strategies.

Future posts will explore in detail how to implement those strategies via diet and lifestyle. Today, I’m just going to give a general overview of what I would do if I had cancer.

Eat the Perfect Health Diet

This may sound self-serving, but it’s my best advice. Our diet is designed to optimize health generally, and that’s exactly what you want to do against cancer.

I said in the introduction that cancer is a disease in which cells lose their “humanness” – their proclivity to collaborate with other human cells to create a human organism. Instead, they lose recently evolved features and “remember” an identity similar to that of our distant evolutionary ancestors from the early days of multicellular life. This regression is possible because we retain the genes of our primitive evolutionary ancestors, and silencing of only a few hundred genes may cause a human cell to resemble, genetically, bacteria or fungi.

Many gut bacteria can take on two modes of behavior – a commensal or harmless phenotype, or a virulent harmful phenotype – depending on whether their environment is benign. In beneficial environments, bacteria tend to be cooperative with their host; in harsh environments, bacteria begin to look out for their own interests “selfishly,” and begin to display virulence traits which harm their host but help them move to a better environment.

Something similar may happen with “proto-cancer” cells. In a healthy environment, they are pleased to cooperate with their host – to retain their “humanness.” But in a harsh environment, they are more likely to withdraw from their neighbors and go their own way. An abused cell is more likely to become a cancer cell.

This may sound like anthropomorphization, but the metaphor is probably sound. Bruce Ames has remarked upon the fact that almost every compound is a carcinogen in large enough doses. Why? Because any unbalanced environment is harsh, and any harsh environment makes the cell more likely to develop the cancer phenotype.

It’s not only by discouraging “cancer virulence” that a good diet helps. A healthy diet also optimizes immune function.

Immune function is highly variable. Under stress, we suppress immunity so that all the body’s resources are available to meet “fight or flight” needs. Contrariwise, peaceable happiness is stimulating to immune function. A nutrient-rich diet, savory meals, happiness, calm, restful time spent in conversation – all of these things tell the body it has no pressing concerns and that available resources can be devoted to immunity and healing.

After cancer diagnosis, from a similar medical condition, those who are under stress tend to succumb to cancer, while those who are happy, cheerful, and sociable tend to recover from it. It is believed that this difference is primarily due to improved immune function in those under less stress.

I believe that a healthy, tasty diet is also a stimulant for immune function. Make your food nourishing and enjoyable.

Specific Dietary Aspects

A few aspects of an anti-cancer diet deserve special mention. Let’s look at the PHD Food Plate:

Some aspects I would emphasize for cancer patients:

  • Safe starches. I recommend obtaining 400 to 600 glucose calories a day, mainly from safe starches. I believe it is important to avoid a glucose deficiency, since glycosylated proteins are the means of intercellular coordination, and defects in glycosylation are characteristic of the cancer phenotype. (See, eg, this paper.) You don’t want to aggravate this with a self-induced glucose deficiency.
  • Low omega-6 meats. Omega-6 fats can be very damaging to mitochondria and can promote metastasis. Our needs for them are minimal, and they are everywhere. It’s important to choose foods that minimize omega-6 levels. Among meats, prefer seafood, shellfish, and red meats; obtain eggs, milk, and organ meats from pastured and naturally raised animals. Eat tropical plant oils like coconut and palm.
  • Omega-3 and omega-6 balance. The diet should include some marine sources of omega-3 fats, like salmon or sardines.
  • Bone broth soups and gelatin (cooked collagen). Collagen is 30% of our body’s protein and forms much of the extracellular matrix scaffolding which is crucial to maintainance of tissue health. The extracellular matrix is broken down in cancer. An anti-cancer diet should be rich in cooked joint tissue, such as can be found in Ox Feet Broth soups. Vitamin C and sulfur, discussed below, are also required for collagen formation; be sure you’re not deficient in these.
  • Fermented vegetables, yogurt, and acids. A diverse portfolio of gut bacteria can be helpful to the fight against cancer by several mechanisms. Probiotic flora from fermented  foods help shield against the entry of cancer-promoting pathogens to the body through the gut; they generate by-products, like short-chain fats and vitamin K2, which have anti-cancer effects; and they can modulate immunity in a favorable direction. Acids such as vinegar and lemon juice can also favorably modify gut bacteria.
  • Vegetables, herbs, and spices.Fiber is probably beneficial against cancer. Butyrate, which is produced by gut bacteria from the digestion of many types of fiber including “resistant starch” from safe starches, has anti-cancer properties. Moreover, many vegetables and traditional herbs and spices have been shown to have anti-angiogenic effects. Foods with anti-angiogenic properties include:
    • Garlic.
    • Tomato.
    • Green tea.
    • Dark chocolate / cocoa.
    • Maitake mushroom.
    • Bok choy.
    • Kale.
    • Many berries.
    • Cherries.
    • Ginseng.
    • Turmeric.
    • Oregano.
    • Parsley.
    • Polyphenol-rich extra virgin olive oils.
  • Organ meats and egg yolks. It’s important to be well nourished, and organ meats like liver and egg yolks tend to be rich in micronutrients. They are much better than plant foods for compounds like phospholipids. In particular, choline (and its phospholipid form phosphatidylcholine) is important for methylation status and epigenetic functioning – an important element in cancer prevention.
  • Sea vegetables, sea salt, and seafoods. These are good sources of trace minerals such as iodine, which is a critical anti-cancer nutrient.

In general cancer patients should focus on the foods in the apple of the PHD Food Plate more than the “pleasure foods.” However, there’s nothing wrong with some berries, dark chocolate, pistachios, and whipped cream for dessert, and some red wine with dinner. Above all, it’s important to enjoy your food. Try to obtain from every meal a sense of pleasure and well being!

Supplements

Much more could be said on this topic than I’m going to say today. One could make a very long list of supplements that might help against cancer (also a long list of those that hurt). However, the crucial five from my point of view are in our recommended supplement list:

  • Vitamin D
  • Vitamin K2
  • Iodine
  • Selenium
  • Magnesium

The tricky one here is the iodine. Iodine dosage should be built up very slowly from a low level, so as not to disrupt thyroid function. (Hyperthyroidism can strongly promote cancer, and hypothyroidism can inhibit immune function and healing, so any thyroid dysfunction is a serious risk.) Start at 500 mcg or less, and increase the dose no faster than a doubling per month. If you get either hypothyroid or hyperthyroid symptoms from an increase in dose, back off a bit (eg instead of going directly from 500 mcg to 1 mg per day, go to 500 mcg and 1 mg on alternate days). Be patient, but try to build up to 12 mg/day over a 6 month period. Then stay there. Be sure to get 200 mcg/day selenium along with the iodine.

I also recommend a multivitamin, for general nourishment; and make sure there is no deficiency of vitamin C, zinc, copper, or chromium. Also, when it comes to antioxidants, more is not better. Avoid most antioxidant supplements other than glutathione, vitamin C, selenium, zinc, copper, and manganese.

For magnesium, I recommend taking a 200 mg oral supplement of magnesium citrate or a magnesium chelate. Epsom salt baths might not provide magnesium, but they can be a useful source of sulfur (in the form of sulfate) which assists collagen formation.

Vitamin C is an unusual case. It supports collagen formation, and for this purpose and to avoid a deficiency I strongly suggest taking 1 g per day. In higher doses, vitamin C may be helpful because it has anti-viral properties (see Fighting Viral Infections by Vitamin C at Bowel Tolerance, Sep 26, 2010), and most cancers are probably viral in origin. Linus Pauling, of course, advocated high doses of vitamin C – either taken orally to bowel tolerance, or intravenously. However, there are arguments on the other side. Vitamin C can protect cancer cells from immune attack, and also makes them resistant to chemotherapies. Clinical trials have not yet proven high-dose vitamin C therapy, but it may help against a subset of cancers caused by viruses sensitive to vitamin C therapy.

If sufficient amounts are not obtained from diet, then choline should be supplemented.

Intermittent Fasting, Intermittent Ketosis, Intermittent Protein Restriction

This is an extremely important cluster of strategies that are probably highly effective against cancer.

Their common trait is that all three promote autophagy, or “self-eating,” which is both a means for cells to cope with resource scarcity and a central part of the intracellular immune response.

When resources are abundant, cells allow aged organelles and junk proteins to accumulate. When resources are scarce, they turn on autophagy and digest unnecessary components, recycling the resources.

Autophagy is the dominant innate immune mechanism inside cells – the primary way cells kill bacteria and viruses.

Autophagy also recycles damaged mitochondria, which can be digested, enabling remaining healthy mitochondria to multiply. The result is a healthier mitochondrial population.

Since viruses and damaged mitochondria promote cancer, autophagy helps transform cells from the cancer phenotype back to the normal human phenotype.

Fasting, by inducing resource scarcity, promotes autophagy. Scarcity of amino acids, which can be achieved by a protein restricted diet, also promotes autophagy. And ketosis, which is part of the metabolic profile of starvation, also promotes autophagy.

Note in my section heading the shared word: “intermittent.” We don’t want to sustain fasts or protein scarcity too long; that could create malnourishment and cause more harm than good. Permanent ketosis may promote fungal infections. The most helpful course is probably to follow these strategies intermittently:

  • Engage in daily intermittent fasting: eat only within a 6 to 8 hour window each day. Within the fasting period, eat some coconut oil or MCT oil to promote ketosis.
  • Eat high protein for a few weeks while engaging in resistance exercise to build muscle; then low protein for a few weeks.

A Note on Ketogenic Diets

Since we wrote our book, we’ve become a bit less excited about the therapeutic potential of ketogenic diets.

Ketogenic diets have demonstrated effectiveness in brain cancers, and several considerations suggest that they would be helpful against all cancers:

  • Cancer cells are dependent on glucose metabolism, a phenomenon called the Warburg effect. In ketosis, blood glucose levels can be decreased – a fall from 90 to 65 mg/dl is achievable – and reduced glucose availability should retard cancer growth.
  • Mitochondria do well on ketones, and some studies had shown that provision of ketones can restore the ability of mitochondria to trigger apoptosis, or the programmed cell death of cancer cells.

It’s too early to judge, but a few scraps of data published recently have made ketogenic diets seem a bit less exciting then hoped.

First, the group of Michael Lisanti has published work suggesting that tumors can evade the metabolic restrictions of a ketogenic diet by manipulating neighboring normal cells. The idea (here is an overview) is that cancer cells release hydrogen peroxide, which causes a stress response in neighboring cells, stimulating them to release lactic acid, which the cancer cells can metabolize. This process can happen nearly as well on a ketogenic as on a normal diet, so the effectiveness of a ketogenic diet in starving the cancer cells is reduced.

The Lisanti group results are hardly conclusive – indeed so far as I know no other group has supported their claims – and there are plenty of skeptics. Jimmy Moore gathered responses from a panel of low-carb experts.

Second, clinical experience with ketogenic diets has not yet shown them to be highly effective. The sort of data we have is well represented by a recent report in Nutrition and Metabolism. Sixteen patients with advanced metastatic cancer were put on ketogenic diets. The results:

One patient did not tolerate the diet and dropped out within 3 days. Among those who tolerated the diet, two patients died early, one stopped after 2 weeks due to personal reasons, one felt unable to stick to the diet after 4 weeks, one stopped after 6 and two stopped after 7 and 8 weeks due to progress of the disease, one had to discontinue after 6 weeks to resume chemotherapy and five completed the 3 month intervention period.

The conclusion: a ketogenic diet “has no severe side effects and might improve aspects of quality of life and blood parameters in some patients.”

Clinical trials with control groups and more statistical power are needed to evaluate whether ketogenic diets have therapeutic effect. For now, I think the most prudent course is intermittent ketosis and intermittent ketogenic fasting, rather than a continuously ketogenic diet.

UPDATE: Mario makes a great point in the comments: fasting prior to chemotherapy reduces toxicity to normal cells but increases toxicity to cancer cells. It is quite likely that a ketogenic diet might have the same effect during chemotherapy. So the combination of intermittent ketogenic dieting with chemotherapy should be given consideration.

Circadian Rhythm Enhancement

Many diseases become more likely, or more severe, if circadian rhythms are disrupted. Enhancement of circadian rhythms may be therapeutic for these diseases.

I’ve blogged about circadian rhythm therapies for hypothyroidism (“Intermittent Fasting as a Therapy for Hypothyroidism,” Dec 1, 2010) and for sleep disorders, psychiatric disorders, neurodegenerative disorders, and obesity (“Seth Roberts and Circadian Therapy,” Mar 22, 2011).

Well, cancer is another disease for which circadian disruption may be damaging. The International Agency on Research on Cancer (IARC) has recently classified “shiftwork that involves circadian disruption” as “probably carcinogenic to humans.”

It’s plausible that circadian enhancement may be therapeutic for cancer. Tactics that enhance circadian rhythms include:

  • Exposure to mid-day sunlight.
  • Sleeping in total darkness during hours of darkness.
  • Confining eating to daylight hours.
  • Socializing – especially, looking at faces and talking – during daylight hours. Seth Roberts found that looking at images of human faces can substitute for actual socializing.
  • Exercising during daylight hours. Even low-level activity – like standing instead of sitting – helps.
  • In people who are melatonin deficient due to a brain immune response, supplementation of melatonin just before bedtime.

Curiously, circadian rhythm disruption seems to make chemotherapy more effective. Also, timing treatments to match circadian rhythms may double their effectiveness.

Exercise and Other Lifestyle Factors

A number of lifestyle factors are important for cancer recovery. David Servan-Schreiber’s Anti-Cancer has an excellent overview of the evidence.

A recent study in the Lancet found that every additional 15 min of daily exercise beyond 15 min a day reduced all-cancer mortality by 1%. Exercise appears to be therapeutic even for late stage cancers. A meta-review found that two and a half hours of exercise a week could lower a breast cancer patient’s risk of dying or cancer recurrence by 40 percent, and could reduce a prostate cancer patient’s risk of dying from the disease by about 30 percent.

However, exercise should not be exhausting. Rather, it should be restful and relaxing; or build muscle. Resistance exercise on the “Body by Science” model of one intense workout per week, with more time spent in restful recovery than in stress, is probably a good strategy. Long walks outdoors in nature, and relaxing exercises like yoga or tai chi, are also great approaches to cancer therapy.

Being sociable, happy, calm, and optimistic are all important factors for cancer recovery. Those who have companions they love, and a purpose for living that makes them happy, have the best prognosis. Be grateful for what you have, and make your body understand that life is worth living.

Dealing with Anorexia and Nausea

Anorexia and nausea can seriously impair the ability of cancer patients to eat a nourishing diet and maintain their strength.

I haven’t had time to research this aspect of the disease yet, but there do seem to be some dietary and lifestyle interventions that help.

For instance, exercise can correct anorexia.

Among dietary interventions, ginger has been reported to reduce chemotherapy-induced nausea, reducing incidence in one study from 93% to 55%. (Hat tip: Healthy Fellow.)

Ginger teas are a traditional Asian folk remedy. Slice some ginger root in water, boil it on the stove, add some rice syrup for sweetness, and drink up!

Under-Utilized Therapies

There are a few therapies which are rarely prescribed, but might be more helpful than chemotherapies in treating cancer:

  • Low-dose naltrexone.
  • Anti-viral drugs.
  • Anti-fungal therapies.

Low-dose naltrexone is taken at night before bed. It temporarily blocks opioid receptors, which leads the body to increase production of endorphins and enkephalins – immune compounds which interact with opioid receptors. The following day, the naltrexone is gone and the opioid receptors are working again, but the endorphins are still around. Taking LDN thus increases endorphin levels. Endorphins inhibit cancer proliferation, and may enhance anti-cancer immunity. Here is a recent paper on anti-proliferative effects of LDN against ovarian cancer: http://pmid.us/21685240. Here is a recent paper on LDN plus alpha lipoic acid as a therapy against pancreatic cancer: http://pmid.us/20042414. For a general overview, see http://lowdosenaltrexone.org/.

Viruses cause or contribute to most cancers, and thus anti-viral drugs have great potential. A few cancer-causing viruses are famous, such as the Human Papilloma Virus for which there is a vaccine; however, most of the viruses that cause cancer remain unknown, though we know they exist because genetic mutations that impair viral immunity greatly increase cancer incidence.

Mario Renato Iwakura recently sent me a link to a paper that nicely illustrates the potential of antiviral therapies against cancer. Cytomegalovirus, also known as human herpes virus 5, is a common virus that infects 40% of adults worldwide and 50% to 80% of Americans. However, it is found in almost 100% of human tumors. It seems to be difficult to get cancer if you haven’t been infected by cytomegalovirus.

From the paper abstract:

Medulloblastomas are the most common malignant brain tumors in children…. Human cytomegalovirus (HCMV) is prevalent in the human population and encodes proteins that provide immune evasion strategies and promote oncogenic transformation and oncomodulation…. Remarkably, all of the human medulloblastoma cell lines that we analyzed contained HCMV DNA and RNA and expressed HCMV proteins at various levels in vitro. When engrafted into immunocompromised mice, human medulloblastoma cells induced expression of HCMV proteins. HCMV and COX-2 expression correlated in primary tumors, cell lines, and medulloblastoma xenografts. The antiviral drug valganciclovir and the specific COX-2 inhibitor celecoxib prevented HCMV replication in vitro and inhibited PGE2 production and reduced medulloblastoma tumor cell growth both in vitro and in vivo.

Tumor growth declined by 72% when treated with Valcyte (valganciclovir) and an NSAID drug. A press release notes that these drugs have “relatively good adverse effect profiles” and that “antiviral drugs are selective and largely affect infected cells.”

Yet another antimicrobial approach that may be helpful against cancer is antifungal therapy. Most cancer patients develop systemic fungal infections, and fungal infections such as Candida promote metastasis and tumor growth, and may also suppress anti-cancer immunity. An effective antifungal therapy may significantly retard cancer progression.

Conclusion

Much more remains to be said, and it’s certain that we’ll refine these suggestions after more thoroughly studying the literature. But I think this basic approach to an anti-cancer diet can’t be too far wrong.

Our prayers and best wishes go out to all those who are battling cancer.

High LDL on Paleo Revisited: Low Carb & the Thyroid

One of the more mysterious conditions afflicting low-carb Paleo dieters has been high serum cholesterol. Two of our most popular posts were about this problem: Low Carb Paleo, and LDL is Soaring – Help! (Mar 2, 2011) enumerated some cases and asked readers to suggest answers; Answer Day: What Causes High LDL on Low-Carb Paleo? (Mar 4, 2011) suggested one possible remedy.

On the first post, one of the causes suggested by readers was hypothyroidism – an astute answer. Raj Ganpath wrote:

Weight loss (and VLC diet) resulting in hypothyroidism resulting in elevated cholesterol due to less pronounced LDL receptors?

Kratos said “Hypothyroidism from low carbs.” Mike Gruber said:

I’m the guy with the 585 TC. It went down (to 378 8 months or so ago, time to check again) when I started supplementing with iodine. My TSH has also been trending up the last few years, even before Paleo. So hypothyroidism is my primary suspect.

Those answers caused me to put the connection between hypothyroidism and LDL levels on my research “to do” list.

Chris Masterjohn’s Work on Thyroid Hormone and LDL Receptors

Chris Masterjohn has done a number of blog posts about the role of LDL receptors in cardiovascular disease. His talk at the Ancestral Health Symposium was on this topic, and a recent blog post, “The Central Role of Thyroid Hormone in Governing LDL Receptor Activity and the Risk of Heart Disease,” provides an overview.

His key observation is that thyroid hormone stimulates expression of the LDL receptor (1). T3 thyroid hormone binds to thyroid hormone receptors on the nuclear membrane, the pair (a “dimer”) is then imported into the nucleus where it acts as a transcription factor causing, among other effects, LDL receptors to be generated on the cell membrane.

So higher T3 = more LDL receptors = more LDL particles pulled into cells and stripped of their fatty cargo. So high T3 tends to reduce serum LDL cholesterol levels, but give cells more energy-providing fats. Low T3, conversely, would tend to raise serum cholesterol but deprive cells of energy.

Other Pieces of the Puzzle

Two other facts we’ve recently blogged about help us interpret this result:

We can now assemble a hypothesis linking low carb diets to high LDL. If one eats a glucose and/or protein restricted diet, T3 levels will fall to conserve glucose or protein. When T3 levels fall, LDL receptor expression is reduced. This prevents LDL from serving its fat transport function, but keeps the LDL particles in the blood where their immune function operates.

If LDL particles were being taken up from the blood via LDL receptors, they would have to be replaced – a resource-expensive operation – or immunity would suffer. Apparently evolution favors immunity, and gives up the lipid-transport functions of LDL in order to maintain immune functions during periods of food scarcity.

High LDL on Low Carb: Good health, bad diet?

Suppose LDL receptors are so thoroughly suppressed by low T3 that the lipid transport function of LDL is abolished. What happens to LDL particles in the blood?

Immunity becomes their only function. They hang around in the blood until they meet up with (bacterial) toxins. This contact causes the LDL lipoprotein to be oxidized, after which the particle attaches to macrophage scavenger receptors and is cleared by immune cells.

So, if T3 hormone levels are very low and there is an infection, LDL particles will get oxidized and cleared by immune cells, and LDL levels will stay low. But if there is no infection and no toxins to oxidize LDL, and the diet creates no oxidative stress (ie low levels of omega-6 fats and fructose), then LDL particles may stay in the blood for long periods of time.

If LDL particles continue to be generated, which happens in part when eating fatty food, then LDL levels might increase.

So we might take high LDL on Paleo as a possible sign of two things:

  • A chronic state of glucose deficiency, leading to very low T3 levels and suppressed clearance of LDL particles by lipid transport pathways.
  • Absence of infections or oxidative stress which would clear LDL particles by immune pathways.

The solution? Eat more carbs, and address any remaining cause of hypothyroidism, such as iodine or selenium deficiency. T3 levels should then rise and LDL levels return to normal.

Alternatively, there is evidence that some infections may induce euthyroid sick syndrome, a state of low T3 and high rT3, directly. And these infections may not oxidize LDL, thus they may not lead to loss of LDL particles by immune pathways. So such infections could be another cause of high LDL on Paleo.

Gregory Barton’s Experience

Gregory Barton is an Australian, 52 years old, living in Thailand, where he keeps goats, makes goat cheese and manages a large garden which can be seen on http://www.asiagoat.com/.

Gregory left a comment with an intriguing story, and I invited him to elaborate in a post. Here’s Gregory’s story. – Paul

Gregory’s Writing Begins Here

One of the claims of low carb dieting is that it will normalize the symptoms of metabolic syndrome. Blood pressure, blood sugar and blood lipids, it is claimed, will all come down on a low carb diet, in addition to weight. For most people this happens. But there is a significant minority of people on Paleo and other low carb diets whose blood lipids defy this claim. (See the list of low-carb celebrities with high LDL in this post.)

Why should this happen? Why should some people’s lipids fall on low carb while other people’s lipids rise? Suboptimal thyroid might be the proximate cause for lipids rising on a low carb or paleo diet. Broda Barnes and Lawrence Galton have this to say about thyroid disorders:

“Of all the problems that can affect physical or mental health, none is more common than thyroid gland disturbance. None is more readily and inexpensively corrected. And none is more often untreated, and even unsuspected.”  — Hypothyroidism: The Unsuspected Illness

I went very low carb in April in an effort to address metabolic issues, eating as little as 15grams carbohydrate per day. I had great results with blood pressure, sleeping, blood sugar and weight loss. But lipids bucked the trend.

I had expected triglycerides and cholesterol to drop when I cut the carbs, but they did the opposite: They surged. By July my total cholesterol was 350, LDL 280, and triglycerides bobbed around between 150 and 220.

I did some research and found several competing theories for this kind of surge:

  1. Saturated fat: The increase in saturated fat created a superabundance of cholesterol which the liver cannot handle. Also, Loren Cordain has claimed that saturated fat downregulates LDL receptors.
  2. Temporary hyperlipidemia: The surge in lipids is the temporary consequence of the body purging visceral fat. Jenny Ruhl has argued that within a period of months the situation should settle down and lipids should normalize.
  3. Hibernation: The metabolism has gone into “hibernation” with the result that the thyroid hormone T4 is being converted into rT3, an isomer of the T3 molecule, which prevents the clearance of LDL.
  4. Malnutrition: In March, Paul wrote that malnutrition in general and copper deficiency in particular “… is, I believe, the single most likely cause of elevated LDL on low-carb Paleo diets.”
  5. Genetics: Dr. Davis has argued that some combinations of ApoE alleles may make a  person “unable to deal with fats and dietary cholesterol.”

I could accept that saturated fat would raise my cholesterol to some degree. However, I doubted that an increase in saturated fat, or purging of visceral fat, would be responsible for a 75% increase in TC from 200 to 350.

There are two basic factors controlling cholesterol levels: creation and clearance. If the surge was not entirely attributable to saturated fat, perhaps the better explanation was that the cholesterol was not being cleared properly. I was drawn to the hibernation theory.

But what causes the body to go into hibernation? According to Chris Masterjohn, a low carb diet could be the cause. Although he does not mention rT3, he warns,

“One thing to look out for is that extended low-carbing can decrease thyroid function, which will cause a bad increase in LDL-C, and be bad in itself. So be careful not to go to extremes, or if you do, to monitor thyroid function carefully.”

If low carb is the cause, then higher carb should be the cure. Indeed, Val Taylor, the owner of the yahoo rT3 group, commented that “it is possible that the rT3 could just be from a low carb diet.” She says, “I keep carbs at no lower than 60g per day for this reason.”

Cortisol and Getting “Stuck” in Hibernation

So what about temporary hyperlipidemia? Bears hibernate for winter, creating rT3, but manage to awaken in spring. Why should humans on low carb diets not be able to awaken from their hibernation? There are many people who complain of high cholesterol years after starting low carb.

A hormonal factor associated with staying in hibernation is high cortisol. It has been claimed that excessively high or low cortisol, sustained over long periods, may cause one to get “stuck” in hibernation mode. One of the moderators from the yahoo rT3 group said:

High or low cortisol can cause rT3 problems, as can chronic illness. It would be nice if correcting these things was all that was necessary. But it seems that the body gets stuck in high rT3 mode.

James LaValle & Stacy Lundin in Cracking the Metabolic Code: 9 Keys to Optimal Health wrote:

When a person experiences prolonged stress, the adrenals manufacture a large amount of the stress hormone cortisol. Cortisol inhibits the conversion of T4 to T3 and favours the conversion of T4 to rT3. If stress is prolonged a condition called reverse T3 dominance occurs and lasts even after the stress passes and cortisol levels fall. (my emphasis)

What I Did

First, I got my thyroid hormone levels tested. A blood test revealed that I had T4 at the top of the range and T3 below range. Ideally I would have tested rT3, but in Thailand the test is not available. I consulted Val Taylor, the owner of the yahoo rT3 group, who said that low T3 can cause lipids to go as high as mine have and, “as you have plenty of T4 there is no other reason for low T3 other than rT3.”

I decided to make these changes:

  1. Increase net carbs to ~50 grams per day. Having achieved my goals with all other metabolic markers I increased carbs, taking care that one hour postprandial blood sugar did not exceed 130 mg/dl.
  2. Supplement with T3 thyroid hormone.
  3. In case the malnutrition explanation was a factor, I began supplementing copper and eating my wife’s delicious liver pate three times per week.

I decided to supplement T3 for the following reasons:

  • The surge in TC was acute and very high. It was above the optimal range in O Primitivo’s mortality data.
  • I increased carbs by 20-30g/day for about a month. TC stabilized, but did not drop.
  • The rT3 theory is elegant and I was eager to test my claim that the bulk of the cholesterol was due to a problem with clearance rather than ‘superabundance’.

What happened?

I started taking cynomel, a T3 supplement, four weeks ago. After one week triglycerides dropped from 150 to 90. After two weeks TC dropped from 350 to 300 and after another week, to 220. Last week numbers were stable.

Based on Paul’s recent series on blood lipids, especially the post Blood Lipids and Infectious Disease, Part I (Jun 21, 2011), I think TC of 220 mg/dl is optimal. As far as serum cholesterol levels are concerned, the problem has been fixed.

I believe that thyroid hormone levels were the dominant factor in my high LDL. Saturated fat intake has remained constant throughout.

My current goal is to address the root causes of the rT3 dominance and wean myself off the T3 supplement. I hope to achieve this in the next few months. My working hypothesis is that the cause of my high rT3 / low T3 was some combination of very low carb dieting, elevated cortisol (perhaps aggravated by stress over my blood lipids!), or malnutrition.

Another possibility is toxins: Dr Davis claims that such chemicals as perchlorate residues from vegetable fertilizers and polyfluorooctanoic acid, the residue of non-stick cookware, may act as inhibitors of the 5′-deiodinase enzyme that converts T4 to T3. Finally, Val Taylor claims that blood sugar over 140 mg/dl causes rT3 dominance. I couldn’t find any studies confirming this claim, and don’t believe it is relevant to my case. Val recommends low carb for diabetics to prevent cholesterol and rT3 issues but warns not to go under 60g carb per day.

Issues with T3 Supplementation

There are some factors to consider before embarking upon T3 supplementation:

  1. Preparation: In order to tolerate T3 supplement you have to be sure that your iron level and your adrenals are strong enough. This requires quite a bit of testing. I’ve read of people who cut corners with unpleasant results.
  2. Practicalities: T3 supplementation requires daily temperature monitoring in order to assess your progress. People who are on the move throughout the day would find this difficult.
  3. Danger: Once you get on the T3 boat you can’t get off abruptly. Your T4 level will drop below range and you will be dependent on T3 until you wean yourself off. If you stopped abruptly you could develop a nasty reaction and even become comatose.

My advice for anyone doing very low carb

As Chris Masterjohn said, in the quote above, if you are going to do very low carb, check your thyroid levels. I would add: Increase the carbs if you find your free T3 falling to the bottom of the range. It might be a good idea to test also for cortisol. A 24-hour saliva test will give you an idea whether your cortisol levels are likely to contribute to an rT3 issue. It might also be a good idea to avoid very low carb if you are suffering from stress – such as lipid anxiety!

Gregory Barton’s Conclusion

I also think my experience may help prove thyroid hormone replacement to be an alternative, and superior, therapy to statins for very high cholesterol. Statins, in the words of Chris Masterjohn,

“… do nothing to ramp up the level of cholesterol-made goodies to promote strength, proper digestion, virility and fertility.  It is the vocation of thyroid hormone, by contrast, to do both.”

Paul’s Conclusion

Thanks, Gregory, for a great story and well-researched ideas. The rapid restoration of normal cholesterol levels with T3 supplementation would seem to prove that low T3 caused the high LDL levels.

However, I would be very reluctant to recommend T3 supplementation as a treatment for high LDL on Paleo.  If the cause of low T3 is eating too few carbs, then supplementing T3 will greatly increase the rate of glucose utilization and aggravate the glucose deficiency.

The proper solution, I think, is simply to eat more carbs, to provide other thyroid-supporting nutrients like selenium and iodine, and allow the body to adjust its T3 levels naturally. The adjustment might be quite rapid.

In Gregory’s case, his increased carb consumption of ~50 g/day was still near our minimum, and he may have been well below the carb+protein minimum of 150 g/day (since few people naturally eat more than about 75 g protein). So I think he might have given additional carbs a try before proceeding to the T3.

Gregory had a few questions for me:

GB: What if one is glucose intolerant and can’t tolerate more than 60 grams per day without hyperglycemia or weight gain?

PJ: I think almost everyone, even diabetics, can find a way to tolerate 60 g/day dietary carbs without hyperglycemia or weight gain, and should.

GB: What if raising carbs doesn’t normalize blood lipids and one finds oneself ‘stuck in rT3 mode’?

PJ: I’m not yet convinced there is such a thing as “stuck in rT3 mode” apart from being “stuck in a diet that provides too few carbs” or “stuck in a chronic infection.” If one finds one’s self stuck while eating a balanced diet, I would look for infectious causes and address those.

Finally, if I may sound like Seth Roberts for a moment, I believe this story shows the value of a new form of science: personal experimentation, exploration of ideas on blogs, and the sharing of experiences online. It takes medical researchers years – often decades – to track down the causes of simple phenomena, such as high LDL on low carb. We’re on pace to figure out the essentials in a year.