Magnesium and Insulin Sensitivity

From a paper based on US NHANES nutrition and health survey data (1):

During 1999–2000, the diet of a large proportion of the U.S. population did not contain adequate magnesium... Furthermore, racial or ethnic differences in magnesium persist and may contribute to some health disparities.... Because magnesium intake is low among many people in the United States and inadequate magnesium status is associated with increased risk of acute and chronic conditions, an urgent need exists to perform a current survey to assess the physiologic status of magnesium in the U.S. population.

Magnesium is an essential mineral that's slowly disappearing from the modern diet, as industrial agriculture and industrial food processing increasingly dominate our food choices. One of the many things it's necessary for in mammals is proper insulin sensitivity and glucose control. A loss of glucose control due to insulin resistance can eventually lead to diabetes and all its complications.

Magnesium status is associated with insulin sensitivity (2, 3), and a low magnesium intake predicts the development of type II diabetes in most studies (4, 5) but not all (6). Magnesium supplements largely prevent diabetes in a rat model* (7). Interestingly, excess blood glucose and insulin themselves seem to reduce magnesium status, possibly creating a vicious cycle.

In a 1993 trial, a low-magnesium diet reduced insulin sensitivity in healthy volunteers by 25% in just four weeks (8). It also increased urinary thromboxane concentration, a potential concern for cardiovascular health**.

At least three trials have shown that magnesium supplementation increases insulin sensitivity in insulin-resistant diabetics and non-diabetics (9, 10, 11). In some cases, the results were remarkable. In type II diabetics, 16 weeks of magnesium supplementation improved fasting glucose, calculated insulin sensitivity and HbA1c*** (12). HbA1c dropped by 22 percent.

In insulin resistant volunteers with low blood magnesium, magnesium supplementation for four months reduced estimated insulin resistance by 43 percent and decreased fasting insulin by 32 percent (13). This suggests to me that magnesium deficiency was probably one of the main reasons they were insulin resistant in the first place. But the study had another very interesting finding: magnesium improved the subjects' blood lipid profile remarkably. Total cholesterol decreased, LDL decreased, HDL increased and triglycerides decreased by a whopping 39 percent. The same thing had been reported in the medical literature decades earlier when doctors used magnesium injections to treat heart disease, and also in animals treated with magnesium. Magnesium supplementation also suppresses atherosclerosis (thickening and hardening of the arteries) in animal models, a fact that I may discuss in more detail at some point (14, 15).

In the previous study, participants were given 2.5 g magnesium chloride (MgCl2) per day. That's a bit more than the USDA recommended daily allowance (MgCl2 is mostly chloride by weight), in addition to what they were already getting from their diet. Most of a person's magnesium is in their bones, so correcting a deficiency by eating a nutritious diet may take a while.

Speaking of nutritious diets, how does one get magnesium? Good sources include halibut, leafy greens, chocolate and nuts. Bone broths are also an excellent source of highly absorbable magnesium. Whole grains and beans are also fairly good sources, while refined grains lack most of the magnesium in the whole grain. Organic foods, particularly artisanally produced foods from a farmer's market, are richer in magnesium because they grow on better soil and often use older varieties that are more nutritious.

The problem with seeds such as grains, beans and nuts is that they also contain phytic acid which prevents the absorption of magnesium and other minerals (16). Healthy non-industrial societies that relied on grains took great care in their preparation: they soaked them, often fermented them, and also frequently removed a portion of the bran before cooking (17). These steps all served to reduce the level of phytic acid and other anti-nutrients. I've posted a method for effectively reducing the amount of phytic acid in brown rice (18). Beans should ideally be soaked for 24 hours before cooking, preferably in warm water.

Industrial agriculture has systematically depleted our soil of many minerals, due to high-yield crop varieties and the fact that synthetic fertilizers only replace a few minerals. The mineral content of foods in the US, including magnesium, has dropped sharply in the last 50 years. The reason we need to use fertilizers in the first place is that we've broken the natural nutrient cycle in which minerals always return to the soil in the same place they were removed. In 21st century America, minerals are removed from the soil, pass through our toilets, and end up in the landfill or in waste water. This will continue until we find an acceptable way to return human feces and urine to agricultural soil, as many cultures do to this day****.

I believe that an adequate magnesium intake is critical for proper insulin sensitivity and overall health.


* Zucker rats that lack leptin signaling

** Thromboxane A2 is an omega-6 derived eicosanoid that potently constricts blood vessels and promotes blood clotting. It's interesting that magnesium has such a strong effect on it. It indicates that fatty acid balance is not the only major influence on eicosanoid production.

*** Glycated hemoglobin. A measure of the average blood glucose level over the past few weeks.

**** Anyone interested in further reading on this should look up The Humanure Handbook

Read More

21.00 | 0 comments

Dissolve Away those Pesky Bones with Corn Oil

I just read an interesting paper from Gabriel Fernandes's group at the University of Texas. It's titled "High fat diet-induced animal model of age-associated obesity and osteoporosis". I was expecting this to be the usual "we fed mice industrial lard for 60% of calories and they got sick" paper, but I was pleasantly surprised. From the introduction:

CO [corn oil] is known to promote bone loss, obesity, impaired glucose tolerance, insulin resistance and thus represents a useful model for studying the early stages in the development of obesity, hyperglycemia, Type 2 diabetes [23] and osteoporosis. We have used omega-6 fatty acids enriched diet as a fat source which is commonly observed in today's Western diets basically responsible for the pathogenesis of many diseases [24].

Just 10% of the diet as corn oil (roughly 20% of calories), with no added omega-3, on top of an otherwise poor laboratory diet, caused:

• Obesity
• Osteoporosis
• The replacement of bone marrow with fat cells
  
• Diabetes
• Insulin resistance
  
• Generalized inflammation
• Elevated liver weight (possibly indicating fatty liver)
  

Hmm, some of these sound familiar... We can add them to the findings that omega-6 also promotes various types of cancer in rodents (1).

20% fat is less than the amount it typically takes to make a rodent this sick. This leads me to conclude that corn oil is particularly good at causing mouse versions of some of the most common facets of the "diseases of civilization". It's exceptionally high in omega-6 (linoleic acid) with virtually no omega-3.

Make sure to eat your heart-healthy corn oil! It's made in the USA, dirt cheap and it even lowers cholesterol!

Read More

17.43 | 0 comments

Saturated Fat and Insulin Sensitivity

Insulin sensitivity is a measure of the tissue response to insulin. Typically, it refers to insulin's ability to cause tissues to absorb glucose from the blood. A loss of insulin sensitivity, also called insulin resistance, is a core part of the metabolic disorder that affects many people in industrial nations.

I don't know how many times I've seen the claim in journal articles and on the internet that saturated fat reduces insulin sensitivity. The idea is that saturated fat reduces the body's ability to handle glucose effectively, placing people on the road to diabetes, obesity and heart disease. Given the "selective citation disorder" that plagues the diet-health literature, perhaps this particular claim deserves a closer look.

The Evidence

I found a review article from 2008 that addressed this question (1). I like this review because it only includes high-quality trials that used reliable methods of determining insulin sensitivity*.

On to the meat of it. There were 5 studies in which non-diabetic people were fed diets rich in saturated fat, and compared with a group eating a diet rich in monounsaturated (like olive oil) or polyunsaturated (like corn oil) fat. They ranged in duration from one week to 3 months. Four of the five studies found that fat quality did not affect insulin sensitivity, including one of the 3-month studies.

The fifth study, which is the one that's nearly always cited in the diet-health literature, requires some discussion. This was the KANWU study (2). Over the course of three months, investigators fed 163 volunteers a diet rich in either saturated fat or monounsaturated fat.

The SAFA diet included butter and a table margarine containing a relatively high proportion of SAFAs. The MUFA diet included a spread and a margarine containing high proportions of oleic acid derived from high-oleic sunflower oil and negligible amounts of trans fatty acids and n-3 fatty acids and olive oil.

Yummy. After three months of these diets, there was no significant difference in insulin sensitivity between the saturated fat group and the monounsaturated fat group. Yes, you read that right. Even the study that's selectively cited as evidence that saturated fat causes insulin resistance found no significant difference between the diets. You might not get this by reading the misleading abstract. I'll be generous and acknowledge that the (small) difference was almost statistically significant (p = 0.053).

What the authors decided to focus on instead is the fact that insulin sensitivity declined slightly but significantly on the saturated fat diet compared with the pre-diet baseline. That's why this study is cited as evidence that saturated fat impairs insulin sensitivity. But anyone who has a basic science background will see where this reasoning is flawed (warning: nerd attack. skip the rest of the paragraph if you're not interested). You need a control group for comparison, to take into account normal fluctuations caused by such things as the season, eating mostly cafeteria food, and having a doctor hooking you up to machines. That control group was the group eating monounsaturated fat. The comparison between diet groups was the 'primary outcome', in statistics lingo. That's the comparison that matters, and it wasn't significant. To interpret the study otherwise is to ignore the basic conventions of statistics, which the authors were happy to do. There's a name for it: 'moving the goalpost'. The reviewers shouldn't have let this kind of shenanigans slide.

So we have five studies through 2008, none of which support the idea that saturated fat reduces insulin sensitivity in non-diabetics. Since the review paper was published, I know of one subsequent study that asked the same question (3). Susan J. van Dijk and colleagues fed volunteers with abdominal overweight (beer gut) a diet rich in either saturated fat or monounsaturated fat. I e-mailed the senior author and she said the saturated fat diet was "mostly butter". The specific fats used in the diets weren't mentioned anywhere in the paper, which is a major omission**. In any case, after 8 weeks, insulin sensitivity was virtually identical between the two groups. This study appeared well controlled and used the gold standard method for assessing insulin sensitivity, called the euglycemic-hyperinsulinemic clamp technique***.

The evidence from controlled trials is rather consistent that saturated fat has no appreciable effect on insulin sensitivity.

Why Are We so Focused on Saturated Fat?

Answer: because it's the nutrient everyone loves to hate. As an exercise in completeness, I'm going to mention three dietary factors that actually reduce insulin sensitivity, and get a lot less air time than saturated fat.

#1: Caffeine. That's right, controlled trials show that your favorite murky beverage reduces insulin sensitivity (4, 5). Is it actually relevant to real life? I doubt it. The doses used were large and the studies short-term.

#2: Magnesium deficiency. A low-magnesium diet reduced insulin sensitivity by 25% over the course of three weeks (6). I think this is probably relevant to long-term insulin sensitivity and overall health, although it would be good to have longer-term data. Magnesium deficiency is widespread in industrial nations, due to our over-reliance on refined foods such as sugar, white flour and oils.

#3: Sugar. Fructose reduces insulin sensitivity in humans, along with many other harmful effects (7).

As long as we continue to focus our energy on indicting saturated fat, it will continue distracting us from the real causes of disease.


* For the nerds: euglycemic-hyperinsulinemic clamp (the gold standard), insulin suppression test, or intravenous glucose tolerance test with Minimal Model. They didn't include studies that reported HOMA as their only measure, because it's not very accurate.

** There's this idea that pervades the diet-health literature that all saturated fats are roughly equivalent, all monounsaturated fats are equivalent, etc., therefore it doesn't matter what the source was. This is beyond absurd and reflects our cultural obsession with saturated fat. It really irks me that the reviewers didn't demand this information.

*** They did find that markers of inflammation in fat tissue were higher after the saturated fat diet.

Read More

21.00 | 0 comments

What's the Ideal Fasting Insulin Level?

Insulin is an important hormone. Its canonical function is to signal cells to absorb glucose from the bloodstream, but it has many other effects. Chronically elevated insulin is a marker of metabolic dysfunction, and typically accompanies high fat mass, poor glucose tolerance (prediabetes) and blood lipid abnormalities. Measuring insulin first thing in the morning, before eating a meal, reflects fasting insulin. High fasting insulin prevents the escape of fat from fat tissue and causes a number of other metabolic disturbances.

Elevated fasting insulin is a hallmark of the metabolic syndrome, the quintessential modern metabolic disorder that affects 24% of Americans (NHANES III). Dr. Lamarche and colleagues found that having an insulin level of 13 uIU/mL in Canada correlated with an 8-fold higher heart attack risk than a level of 9.3 uIU/mL (1; thanks to NephroPal for the reference). So right away, we can put our upper limit at 9.3 uIU/mL. The average insulin level in the U.S., according to the NHANES III survey, is 8.8 uIU/mL for men and 8.4 for women (2). Given the degree of metabolic dysfunction in this country, I think it's safe to say that the ideal level of fasting insulin is probably below 8.4 uIU/mL as well.

Let's dig deeper. What we really need is a healthy, non-industrial "negative control" group. Fortunately, Dr. Staffan Lindeberg and his team made detailed measurements of fasting insulin while they were visiting the isolated Melanesian island of Kitava (3). He compared his measurements to age-matched Swedish volunteers. In male and female Swedes, the average fasting insulin ranges from 4-11 uIU/mL, and increases with age. From age 60-74, the average insulin level is 7.3 uIU/mL.

In contrast, the range on Kitava is 3-6 uIU/mL, which does not increase with age. In the 60-74 age group, in both men and women, the average fasting insulin on Kitava is 3.5 uIU/mL. That's less than half the average level in Sweden and the U.S. Keep in mind that the Kitavans are lean and have an undetectable rate of heart attack and stroke.

Another example from the literature are the Shuar hunter-gatherers of the Amazon rainforest. Women in this group have an average fasting insulin concentration of 5.1 uIU/mL (4; no data was given for men).

I found a couple of studies from the early 1970s as well, indicating that African pygmies and San bushmen have rather high fasting insulin. Glucose tolerance was excellent in the pygmies and poor in the bushmen (5, 6, free full text). This may reflect differences in carbohydrate intake. San bushmen consume very little carbohydrate during certain seasons, and thus would likely have glucose intolerance during that period. There are three facts that make me doubt the insulin measurements in these older studies:

1. It's hard to be sure that they didn't eat anything prior to the blood draw.
2. From what I understand, insulin assays were variable and not standardized back then.
   
3. In the San study, their fasting insulin was 1/3 lower than the Caucasian control group (10 vs. 15 uIU/mL). I doubt these active Caucasian researchers really had an average fasting insulin level of 15 uIU/mL. Both sets of measurements are probably too high.
   

Now you know the conflicting evidence, so you're free to be skeptical if you'd like.

We also have data from a controlled trial in healthy urban people eating a "paleolithic"-type diet. On a paleolithic diet designed to maintain body weight (calorie intake had to be increased substantially to prevent fat loss during the diet), fasting insulin dropped from an average of 7.2 to 2.9 uIU/mL in just 10 days. The variation in insulin level between individuals decreased 9-fold, and by the end, all participants were close to the average value of 2.9 uIU/mL. This shows that high fasting insulin is correctable in people who haven't yet been permanently damaged by the industrial diet and lifestyle. The study included men and women of European, African and Asian descent (7).

One final data point. My own fasting insulin, earlier this year, was 2.3 uIU/mL. I believe it reflects a good diet, regular exercise, sufficient sleep, a relatively healthy diet growing up, and the fact that I managed to come across the right information relatively young. It does not reflect: carbohydrate restriction, fat restriction, or saturated fat restriction. Neither does the low fasting insulin of healthy non-industrial cultures.

So what's the ideal fasting insulin level? My current feeling is that we can consider anything between 2 and 6 uIU/mL within our evolutionary template, although the lower half of that range may be preferable.

Read More

14.00 | 0 comments

Diabetics on a Low-carbohydrate Diet, Part II

I just found another very interesting study performed in Japan by Dr. Hajime Haimoto and colleagues (free full text). They took severe diabetics with an HbA1c of 10.9% and put them on a low-carbohydrate diet:

The main principle of the CRD [carbohydrate-restricted diet] was to eliminate carbohydrate-rich food twice a day at breakfast and dinner, or eliminate it three times a day at breakfast, lunch and dinner... There were no other restrictions. Patients on the CRD were permitted to eat as much protein and fat as they wanted, including saturated fat.

What happened to their blood lipids after eating all that fat for 6 months, and increasing their saturated fat intake to that of the average American? LDL decreased and HDL increased, both statistically significant. Oops. But that's water under the bridge. What we really care about here is glucose control. The patients' HbA1c (glycated hemoglobin; a measure of average blood glucose over the past several weeks) declined from 10.9 to 7.4%.

Here's a graph showing the improvement in HbA1c. Each line represents one individual:

Every single patient improved, except the "dropout" who stopped following the diet advice after 3 months (the one line that shoots back up at 6 months). And now, an inspirational anecdote from the paper:

One female patient had an increased physical activity level during the study period in spite of our instructions. However, her increase in physical activity was no more than one hour of walking per day, four days a week. She had implemented an 11% carbohydrate diet without any antidiabetic drug, and her HbA1c level decreased from 14.4% at baseline to 6.1% after 3 months and had been maintained at 5.5% after 6 months.

That patient began with the highest HbA1c and ended with the lowest. Complete glucose control using only diet and exercise. It may not work for everyone, but it's effective in some cases. The study's conclusion:

...the 30%-carbohydrate diet over 6 months led to a remarkable reduction in HbA1c levels, even among outpatients with severe type 2 diabetes, without any insulin therapy, hospital care or increase in sulfonylureas. The effectiveness of the diet may be comparable to that of insulin therapy.

Diabetics on a Low-carbohydrate Diet
The Tokelau Island Migrant Study: Diabetes

Read More

21.40 | 0 comments

Palmitic Acid and Insulin Resistance: a New Paradigm

We've been having an interesting discussion in the comments about a recently published paper by Dr. Stephen C. Benoit and colleagues (free full text). They showed that a butter-rich diet causes weight gain and insulin resistance in rats, compared to a low-fat diet or a diet based on olive oil. They published a thorough description of the diets' compositions, which is very much appreciated!

They went on to show that infusing palmitic acid (a 16-carbon saturated fat) directly into the brain of rats also caused insulin resistance relative to oleic acid (an 18-carbon monounsaturated fat, like in olive oil). Here's a representation of palmitic acid. The COOH end is the acid end, and the squiggly line is the fatty end. Thus it's called a "fatty acid", various forms of which are the fat currency of the body.

One of the most interesting things about this study is the butter group that the investigators fed the same number of calories as the low-fat group (this is called pair-feeding). This group did not become overweight, and did not experience elevated fasting insulin and blood glucose relative to the low-fat group*. This shows clearly that the adverse effects of the butter diet were primarily due to the fact that rodents overeat when fed a high-fat diet.

Unfortunately, the paper doesn't provide longitudinal food intake data so we have no idea how many calories the rats in each group ate, beyond knowing that the low-fat group and the pair-fed butter group ate the same amount. We have no assurance that rats in the butter group and olive oil group ate the same number of calories over time. Rats eat less of foods they find bitter. This probably accounts, at least in part, for the beneficial effects of things like blueberry extracts on rodent models of disease. Olive oil may taste bitter to a rat, particularly when it's 20% of the diet by weight. Butter is tasty to calves, humans and rats alike.

Now we arrive at the speculative part of the post. I've been pondering a tough question for months. Palmitic acid has aroused universal ire for its supposed effects on lipid metabolism and insulin sensitivity**. But that leaves us with a puzzling paradox: palmitic acid is precisely the fatty acid that the liver produces when we eat carbohydrate. Our bodies contain the enzymes necessary to desaturate palmitic acid, making it monounsaturated. Why don't we use them? Why does the liver choose to secrete palmitic acid into the bloodstream unmodified? A fundamental metabolic process like this does not evolve by accident.

Here's the hypothesis. I believe that palmitic acid in the bloodstream does promote insulin resistance in rodents and probably humans as well. But there's a twist: it's probably not pathological at all; it's simply serving as a reversible signal to conserve blood glucose. This is similar to the hormone glucagon, which increases glucose production by the liver in response to falling blood glucose. Let's imagine an average person's eating habits throughout the day. Breakfast is at 8:00 am, lunch is at noon, and dinner is at 7:00 pm. The meals are about 45% carbohydrate, 40% fat and 15% protein. Let's imagine the fat consumed is animal fat, which contains some palmitic acid (25-30% of fatty acids).

The carbohydrate will be absorbed, partially turned into palmitic acid in the liver, and exported as VLDL particles. The amount of palmitic acid produced depends on the intake of starch and fructose, and will be relatively small except in the case of high carbohydrate or fructose consumption. Dietary fat will be absorbed in the intestine and sent out directly as chylomicrons (another lipoprotein particle). This is delayed relative to glucose absorption, such that the palmitic acid from both sources will enter the bloodstream at a similar time (peaks roughly 4 hours post-meal). Here is a hypothetical graph of blood glucose and blood palmitic acid at different points throughout this person's day (based on data such as these):
Notice a pattern? The concentrations of blood glucose and palmitic acid in the blood are approximately opposite one another. The brain responds to palmitic acid by temporarily decreasing the insulin sensitivity of other tissues, because it uses palmitic acid as a signal to begin conserving blood glucose while insulin is still elevated. Glucagon increases glucose secretion by the liver, and palmitic acid makes sure the glucose isn't removed from the bloodstream too quickly. I believe we're looking at a well-coordinated system designed by evolution to ensure that the glucose content of the blood remains stable after a meal.

There are two other scenarios in which this type of system would be advantageous. Let's imagine Nanook the Inuit has just killed a caribou in September. He eats some of the meat and organs with a generous slab of backfat. Large male caribou in the fall can carry a deposit of subcutaneous fat on their back that weighs up to 50 pounds. This fat is about 50% saturated, and roughly 25% palmitic acid. Here's a quote from the book My Life With the Eskimo, published by the anthropologist Vilhjalmur Stefansson in 1913:

The largest slab of back fat which I have seen taken from a Caribou on the Arctic coast was from a bull killed near Langton Bay early in September, the fat weighing 39 pounds. A large bull killed by Mr. Stefansson on Dease River in October had back fat 72 mm. in thickness (2 7/8 inches). Comparing the thickness of this with the Langton Bay specimen, the back fat of the Dease River bull must have weighed at least 50 pounds.

As the food is digested, Nanook's insulin rises to allow amino acids from the protein to be absorbed into his tissues from his bloodstream. But wait, insulin also tells tissues to absorb glucose, and the meal contained virtually no carbohydrate. Nanook is in danger of hypoglycemia. Fortunately, his brain detects the palmitic acid from the meal and signals his tissues to become resistant to the glucose-transporting effect of insulin. At the same time, glucagon signals the liver to release glucose into the bloodstream. His blood glucose remains stable.

The next week, the herd of caribou has moved on and there's no prey in Nanook's territory. He has to live on his own body fat for two days while he hunts. Fortunately, human body fat is about 20% palmitic acid. As fat is released into his bloodstream, the brain detects the palmitic acid and reduces peripheral insulin sensitivity. This helps Nanook's body conserve glucose and use his own body fat as fuel instead.

Over a wide range of fat, carbohydrate and calorie intakes, this system works to maintain stable blood glucose. These three scenarios all illustrate why palmitic acid would be helpful by causing temporary insulin resistance in situations where blood glucose needs to be conserved.

Back to the paper. The authors also showed that force-feeding rats large amounts of palmitic acid and calories (much more than would be present in animal fat) causes changes associated with insulin resistance in the brain. What I believe they have done is overstimulate this natural pathway for regulating insulin sensitivity by feeding unnatural amounts of palmitic acid.

Rats fed the butter diet at the same number of calories as the low-fat group did not exhibit metabolic dysfunction, showing that a reasonable amount of palmitic acid is compatible with metabolic health in this species. I believe this is even more true in humans, given our evolutionary history with animal fat and carbohydrate, both of which contribute palmitic acid to the circulation. Our deep-seated fear of saturated fat may have caused us to mistake a natural aspect of mammalian metabolism for a pathological process.


* The pair-fed butter group did show a lowered sensitivity to insulin, but given its normal weight, normal fasting insulin, and normal blood sugar, it really cannot be said to exhibit metabolic dysfunction in my opinion. Human "metabolic syndrome" involves overweight and elevated fasting insulin, which these rats did not have. Furthermore, the investigators did not show that the insulin sensitivity of the pair-fed butter group was different than a pair-fed olive oil group (they didn't make that comparison), so the finding doesn't implicate saturated fat specifically. Insulin sensitivity is determined in part by carbohydrate intake. This is normal. The more carbohydrate the body has to dispose of, the better it gets at handling it. On a high-fat diet, you don't need much insulin sensitivity to keep blood glucose in the normal range, because you aren't ingesting much glucose. On the other hand, in high-fat (low carbohydrate) diet trials on insulin-resistant people, insulin sensitivity often improves, however this is not the case in healthy insulin-sensitive people.

** The idea of palmitic acid's effect on insulin sensitivity is based largely on animal models and cell culture data. A long-term (rather than temporary and reversible) effect of palmitic acid on insulin sensitivity has never been convincingly demonstrated in humans, to my knowledge. After reviewing the literature, I've also concluded that a long-term, biologically significant effect of saturated fats in general on insulin sensitivity has not been convincingly demonstrated. I'll save that for another post.

Read More

10.00 | 0 comments

Diabetics on a Low-carbohydrate Diet

Diabetes is a disorder of glucose intolerance. What happens when a diabetic eats a low-carbohydrate diet? Here's a graph of blood glucose over a 24 hour period, in type II diabetics on their usual diet (blue and grey triangles), and after 5 weeks on a 55% carbohydrate (yellow circles) or 20% carbohydrate (blue circles) diet:


The study in question describes these volunteers as having "mild, untreated diabetes." If 270 mg/dL of blood glucose is mild diabetes, I'd hate to see severe diabetes! In any case, the low-carbohydrate, high-fat diet brought blood glucose down to an acceptable level without requiring medication.

It's interesting to note in the graph above that fasting blood glucose (18-24 hours) also fell dramatically. This probably reflects improved insulin sensitivity in the liver. The liver pumps glucose into the bloodstream when it's necessary, and insulin suppresses this. When the liver is insulin resistant, it doesn't respond to the normal signal that there's already sufficient glucose, so it releases more and increases fasting blood glucose. When other tissues are insulin resistant, they don't take up the extra glucose, also contributing to the problem.

Glycated hemoglobin (HbA1c), a measure of average blood glucose concentration over the preceding few weeks, also reflected a profound improvement in blood glucose levels in the low-carbohydrate group:

At 5 weeks, the low-carbohydrate group was still improving and headed toward normal HbA1c, while the high-carbohydrate group remained at a dangerously high level. Total cholesterol, LDL and HDL remained unchanged in both groups, while triglycerides fell dramatically in the low-carbohydrate group.

When glucose is poison, it's better to eat fat.

Graph #1 was reproduced from Volek et al. (2005), which re-plotted data from Gannon et al. (2004). Graph #2 was drawn directly from Gannon et al.

Read More

21.18 | 0 comments

Paleolithic Diet Clinical Trials Part IV

Dr. Staffan Lindeberg has published a new study using the "paleolithic diet" to treat type II diabetics (free full text). Type II diabetes, formerly known as late-onset diabetes until it began appearing in children, is typically thought to develop as a result of insulin resistance (a lowered tissue response to the glucose-clearing function of insulin). This is often followed by a decrease in insulin secretion due to degeneration of the insulin-secreting pancreatic beta cells.

After Dr. Lindeberg's wild success treating patients with type II diabetes or glucose intolerance, in which he normalized the glucose tolerance of all 14 of his volunteers in 12 weeks, he set out to replicate the experiment. This time, he began with 13 men and women who had been diagnosed with type II diabetes for an average of 9 years.

Patients were put on two different diets for 3 months each. The first was a "conventional diabetes diet". I read a previous draft of the paper in which I believe they stated it was based on American Diabetes Association guidelines, but I can't find that statement in the final draft. In any case, here are the guidelines from the methods section:

The information on the Diabetes diet stated that it should aim at evenly distributed meals with increased intake of vegetables, root vegetables, dietary fiber, whole-grain bread and other whole-grain cereal products, fruits and berries, and decreased intake of total fat with more unsaturated fat. The majority of dietary energy should come from carbohydrates from foods naturally rich in carbohydrate and dietary fiber. The concepts of glycemic index and varied meals through meal planning by the Plate Model were explained [18]. Salt intake was recommended to be kept below 6 g per day.

The investigators gave the paleolithic group the following advice:

The information on the Paleolithic diet stated that it should be based on lean meat, fish, fruit, leafy and cruciferous vegetables, root vegetables, eggs and nuts, while excluding dairy products, cereal grains, beans, refined fats, sugar, candy, soft drinks, beer and extra addition of salt. The following items were recommended in limited amounts for the Paleolithic diet: eggs (≤2 per day), nuts (preferentially walnuts), dried fruit, potatoes (≤1 medium-sized per day), rapeseed or olive oil (≤1 tablespoon per day), wine (≤1 glass per day). The intake of other foods was not restricted and no advice was given with regard to proportions of food categories (e.g. animal versus plant foods). The evolutionary rationale for a Paleolithic diet and potential benefits were explained.

Neither diet was restricted in calories. After comparing the effects of the two diets for 3 months, the investigators concluded that the paleolithic diet:

• Reduced HbA1c more than the diabetes diet (a measure of average blood glucose)
• Reduced weight, BMI and waist circumference more than the diabetes diet
• Lowered blood pressure more than the diabetes diet
  
• Reduced triglycerides more than the diabetes diet
• Increased HDL more than the diabetes diet

However, the paleolithic diet was not a cure-all. At the end of the trial, 8 out of 13 patents still had diabetic blood glucose after an oral glucose tolerance test (OGTT). This is compared to 9 out of 13 for the diabetes diet. Still, 5 out of 13 with "normal" OGTT after the paleolithic diet isn't bad. The paleolithic diet also significantly reduced insulin resistance and increased glucose tolerance, although it didn't do so more than the diabetes diet.

As has been reported in other studies, paleolithic dieters ate fewer total calories than the comparison group. This is part of the reason why I believe that something in the modern diet causes hyperphagia, or excessive eating. According to the paleolithic diet studies, this food or combination of foods is neolithic, and probably resides in grains, refined sugar and/or dairy. I have my money on wheat and sugar, with a probable long-term contribution from industrial vegetable oils as well.

Were the improvements on the paleolithic diet simply due to calorie restriction? Maybe, but keep in mind that neither group was told to restrict its caloric intake. The reduction in caloric intake occurred naturally, despite the participants presumably eating to fullness. I suspect that the paleolithic diet reset the dieters' body fat set-point, after which fat began pouring out of their fat tissue. They were supplementing their diets with body fat-- 13 pounds (6 kg) of it over 3 months.

The other notable difference between the two diets, besides food types, was carbohydrate intake. The diabetes diet group ate 56% more carbohydrate than the paleo diet group, with 42% of their calories coming from it. The paleolithic group ate 32% carbohydrate. Could this have been the reason for the better outcome of the paleolithic group? I'd be surprised if it wasn't a factor. Advising a diabetic to eat a high-carbohydrate diet is like asking someone who's allergic to bee stings to fetch you some honey from your bee hive. Diabetes is a disorder of glucose intolerance. Starch is a glucose polymer.

Although to be fair, participants on the diabetes diet did improve in a number of ways. There's something to be said for eating whole foods.

This trial was actually a bit of a disappointment for me. I was hoping for a slam dunk, similar to Lindeberg's previous study that "cured" all 14 patients of glucose intolerance in 3 months. In the current study, the paleolithic diet left 8 out of 13 patients diabetic after 3 months. What was the difference? For one thing, the patients in this study had well-established diabetes with an average duration of 9 years. As Jenny Ruhl explains in her book Blood Sugar 101, type II diabetes often progresses to beta cell loss, after which the pancreas can no longer secrete an adequate amount of insulin.

This may be the critical finding of Dr. Lindeberg's two studies: type II diabetes can be prevented when it's caught at an early stage, such as pre-diabetes, whereas prolonged diabetes may cause damage that cannot be completely reversed though diet. I think this is consistent with the experience of many diabetics who have seen an improvement but not a cure from changes in diet. Please add any relevant experiences to the comments.

Collectively, the evidence from clinical trials on the "paleolithic diet" indicate that it's a very effective treatment for modern metabolic dysfunction, including excess body fat, insulin resistance and glucose intolerance. Another way of saying this is that the modern industrial diet causes metabolic dysfunction.

Paleolithic Diet Clinical Trials
Paleolithic Diet Clinical Trials Part II
One Last Thought
Paleolithic Diet Clinical Trials Part III

Read More

15.00 | 0 comments

The Diet-Heart Hypothesis: Oxidized LDL, Part II

In the last post, I presented the evidence that oxidized LDL (oxLDL) is a dominant factor in the arterial disease known as atherosclerosis, although probably not the only factor. In this post, I'll describe some of the major contributors to oxLDL.

Polyunsaturated Fats Increase LDL Oxidation

The serum concentration of oxLDL is strongly influcenced by diet. One dietary determinant of oxLDL is dietary polyunsaturated fat (PUFA). PUFA are inherently susceptible to oxidative damage, compared to monounsaturated and saturated fats. The predominant PUFA in the modern diet is linoleic acid, found excessively in industrial seed oils like corn oil, sunflower oil, safflower oil, cottonseed oil and soy oil. LDL is naturally rich in linoleic acid, even in cultures such as the Kitavans who have a very low dietary intake of it. However, LDL content of linoleic acid does correlate with dietary intake, and the Kitavans have a comparatively small amount of linoleic acid in their LDL, relative to industrial cultures.

There have been a number of media reports in the last few years proclaiming that monounsaturated fat reduces LDL oxidation compared to saturated and polyunsaturated fat. This is rather implausible on the surface, so let's take a closer look. There are two ways to measure oxLDL:

1. Measure it directly from the blood
   
2. Take normal LDL from the blood, expose it to copper in a test tube, and see how fast it oxidizes

The first reflects actual oxLDL in the blood, whereas the second reflects "susceptibility to oxidation" and has a dubious relationship with actual oxidized LDL in the bloodstream. This results in statements like the following (ref):

LDL resistance to copper-induced oxidation, expressed as lag time, was highest during the MUFA-rich diet (55.1±7.3 minutes) and lowest during the PUFA(n-3)– (45.3±7 minutes) and SFA- (45.3±6.4 minutes) rich diets.

This was published in a paper by P. Mata and colleagues in 1996. They fed 42 volunteers one of four different diets for 5 weeks each: one rich in saturated fat, one rich in monounsaturated fat, one rich in linoleic acid PUFA, and one rich in linoleic acid plus omega-3 PUFA. They emphasized the finding quoted above, as did the media. But there's an embarrassing piece of data buried in the paper that the authors, and the media, ignored (thanks to Chris Masterjohn for pointing this out). Here's what they saw when they looked directly at LDL oxidation in their volunteers:

Oops! LDL oxidation in the two PUFA groups was increased by more than 31%. The difference between the leftmost two groups and the rightmost two was statistically significant. As one would expect, oxidized LDL is proportional to the amount of PUFA in LDL, which is proportional to dietary PUFA. This somehow got left out of the abstract and media reports. The same investigators published a similar report a year later.

In another diet trial, participants were placed on one of two diets for 5 weeks: a low-fat, high PUFA diet low in vegetables; or a low-fat, high PUFA diet high in vegetables. The authors were forthright about their findings, so I'll let them summarize:

The median plasma OxLDL-EO6 increased by 27% (P less than 0.01) in response to the low-fat, low-vegetable diet and 19% (P less than 0.01) in response to the low-fat, high-vegetable diet. Also, the Lp(a) concentration was increased by 7% (P less than 0.01) and 9% (P=0.01), respectively.

This is the diet mainstream cardiologists have been prescribing to heart attack patients for 40 years. The trials I mentioned above are the only three I'm aware of in which fat quality was manipulated and oxLDL was directly measured (the first two were based on subsets of the same data). They all suggest that replacing saturated fat with PUFA increases oxLDL.

I suspect that the effect has less to do with the decrease in saturated fat and more to do with the increase in PUFA, although there's no way to know for sure. In the Lyon Diet-Heart trial, which I believe was the most successful diet trial of all time, linoleic acid was reduced to 3.6% of calories, but saturated fat was also reduced. Another reason is that there are numerous low-fat, low PUFA, high-carbohydrate cultures that have low levels of atherosclerosis and heart attacks. The Kitavans, for example, don't seem to have heart attacks or strokes (although no autopsies have been done so we don't know how much atherosclerosis they have).

They get 69% of their calories from high-glycemic starchy tubers, and their 21% fat comes mostly from coconut so it's highly saturated. Their blood lipids are low in omega-6 linoleic acid and very saturated. But there's a little surprise in the data: their lipids are full of palmitic acid (saturated), despite the fact that their diet contains very little of it. The reason is that their livers are turning all that carbohydrate into saturated fat, which is what happens when you eat more carbohydrate than you can burn immediately or store as glycogen. The moral of the story is that you don't need to eat saturated fat to have saturated LDL: a high-carbohydrate diet can accomplish the same thing, especially if it has a high glycemic index.

Fat-Soluble Antioxidants Decrease LDL Oxidation

LDL carries fat-soluble antioxidants, predominantly vitamin E and coenzyme Q10 (CoQ10). One form of vitamin E, alpha-tocopherol, slows atherosclerosis in most animal models but has shown equivocal results in human trials. There is even the suggestion that it may increase LDL oxidation under some circumstances. I don't recommend supplementing with vitamin E. However, the first line of antioxidant defense in LDL is provided by CoQ10. CoQ10 unequivocally reduces LDL oxidation in human subjects, and potently reduces atherosclerosis in animal models.

CoQ10 has a special relationship with cardiovascular health. Levels are reduced in individuals with cardiovascular disease and high oxLDL. Whether this is cause or effect, it's difficult to say. However, supplementing with CoQ10 has been repeatedly shown to be effective for high blood pressure and congestive heart failure. There has been one controlled trial of CoQ10 (120 mg/day) supplementation for the prevention of heart attacks, which reduced cardiac events including deaths by 45%, compared to a group receiving B vitamins. The CoQ10 group showed a large reduction in plasma lipid oxidation. This is a promising result and the experiment should be repeated.

CoQ10 is not an essential nutrient, although food does contribute a small portion of our total CoQ10 use. The large majority of CoQ10 is synthesized by the body itself, and this is dependent on a number of essential nutrients, including vitamin B2, B3, B5, B6, B12, vitamin C and folic acid. Thus, the body's synthesis of CoQ10 is dependent on overall nutritional status. Sub-clinical deficiency of any of these vitamins can hypothetically contribute to reduced CoQ10 production and thus oxLDL. This is potentially a big problem since modern Americans get more than half their calories from nutrient-poor refined foods. Liver is the single best source of many of these vitamins, and also holds the title of Most Nutritious Food on the Planet. It's also rich in CoQ10.

CoQ10 synthesis declines with age and is reduced in people with disorders involving oxidative stress, like cardiovascular disease. It's also greatly reduced by the cholesterol-lowering drugs statins. I'm not generally in favor of supplements, but CoQ10 seems to have a lot of promise and nothing but positive side effects that I'm aware of. CoQ10 deficiency may be a common theme in a number of modern disorders.

Excess Blood Sugar and Fructose Increase LDL Oxidation

Both type I and type II diabetes are associated with higher levels of oxLDL, therefore, prolonged high blood glucose may contribute to LDL oxidation due to glycosylation of the LDL protein ApoB. Fructose consumption increases oxLDL relative to glucose. Fructose is a very powerful glycosylating agent (binds non-specifically to other molecules, causing damage). Although it isn't present at high levels in the general circulation, it does interact with blood lipids in the hepatic portal vein as it moves from the digestive tract to the liver to be turned into fat (palmitic acid). Peter at Hyperlipid has written extensively about the role of glycosylation in LDL oxidation.

The Diet-Heart Hypothesis: The Verdict

The diet-heart hypothesis, the idea that dietary saturated fat and cholesterol raise blood cholesterol and thus increase heart attack risk, is a half-century embarrassment to the international scientific community. It requires willful ignorance of the fact that saturated fat does not increase total cholesterol or LDL in humans, in the long term. It requires a simplistic view of blood lipids that ignores the potentially harmful effects of replacing animal fats with carbohydrate or industrial seed oils. Worst of all, it requires selective citation of the literature on diet modification trials.

I have to conclude that if dietary saturated fat and cholesterol play any role whatsoever in cardiovascular disease, it's a minor one that's trumped by other factors. Industrial seed oils and sugar are likely to play an important role in cardiovascular disease.

Read More

10.00 | 0 comments