Eat Fat, Get Thin?

I recently received an e-mail from a high school science department head who is teaching a unit on nutritional science. He asked for my opinion of a YouTube video of a lecture advocating a high saturated fat diet. The speaker is Donald W. Miller, Jr., M.D., a cardiothoracic surgeon at my alma mater, the University of Washington. My correspondent commented, “I have a feeling that there is room for some skepticism.”   I agree: there’s a whole lot of room for skepticism.

An article based on that video lecture is available on Dr. Miller’s website.   It’s entitled “Enjoy Saturated Fats, They’re Good for You!”  If you want to judge for yourselves, I recommend the article over the video, as he is a poor public speaker.

Dr. Miller’s website contains a lot of disturbing material.  He appears to be a contrarian who disagrees with the consensus of scientific experts on a wide variety of topics, for instance:

  • Health Benefits of a Low-Carbohydrate, High-Saturated Fat Diet
  • Fighting Fluoride [fluoride is poison!]
  • Cardiac Surgeon Dr. Donald Miller Tells Dr. Dean Ornish to Take a Hike
  • Avoid Flu Shot, Take Vitamin D [flu is a Vitamin D deficiency disease?]
  • Questioning HIV/AIDS, Human-Caused Global Warming, and other Orthodoxies in the Biomedical Sciences
  • A User-Friendly Vaccination Schedule [no vaccinations before age 2, no live vaccines, etc.]

He refers to questionable sources of information like the Weston Price Foundation and the notorious AIDS denialist Peter Duesberg.

Based on the company it keeps, it would be tempting to write off his diet advice as crankery.  But I tried to ignore who was saying it and give a fair hearing to what he was saying. He begins by attacking the lipid hypothesis: the idea that fat in the diet causes elevated blood cholesterol which causes cardiovascular disease. This no longer needs attacking: mainstream recommendations have followed the evidence and have evolved from a low cholesterol diet to a low fat diet to a low saturated fat diet to avoidance of trans-fats and of excess calories from fat.  He echoes Gary Taubes’ low-carb arguments and his contention that government recommendations to avoid fat led to higher carbohydrate intake and thereby caused the obesity epidemic. That may be true, but it doesn’t justify his conclusion that the solution to the obesity epidemic is to replace carbohydrates in the diet with saturated fat.

He relies on epidemiologic data and evolutionary arguments to show that dietary fat is good: he says that breast-fed babies get high-fat milk, Eskimos and the Maasai eat a high-fat diet and don’t develop CVD (cardiovascular disease), the paleolithic diet was high in fat, and our genome has evolutionarily adapted to a high fat diet. In fact, the groups he cites do develop CVD, although they have a low incidence. He conveniently fails to consider any other possible explanations for why people like the Maasai might have less CVD: they are thin, get a lot of strenuous exercise, have different genes, and have a low life expectancy, often dying of infectious disease and trauma before they are old enough to have developed signs of heart disease.

He says protein is bad for you in excess: it causes nausea, diarrhea, etc.  He repeats  Gary Taubes’ arguments  to demonize carbohydrates.  He prefers fat, particularly saturated fat, and recommends it constitute 70% of our diet.

His epiphany came from reading an article in Nexus in 1999 that led him to question Dean Ornish’s low fat advice. He was further influenced by the book Eat Fat, Lose Fat by Mary Enig and Sally Fallon (co-founders of the Weston A. Price Foundation, which has been criticized by medical experts as purveying misleading information) and by the writings of notorious cholesterol denialist Uffe Ravnskov.

He cites data from the Framingham study showing that over age 50, there is no effect of either high or low cholesterol levels on mortality except for those whose cholesterol level is falling: for every 1% drop in cholesterol levels, there is an 11% increase in all-cause mortality. He doesn’t mention that the same study showed a robust correlation of cholesterol levels with mortality below age 50, and that the authors said the data for those over 50 is confounded by people whose cholesterol levels are not stable, but falling — perhaps due to diseases predisposing to death.  He cites an article by Castelli allegedly showing that increased saturated fat, cholesterol and calories in the diet are associated with decreased blood cholesterol and that people who ate more cholesterol, saturated fat and calories had lower weight and increased activity levels.  This was not an original study but a commentary accompanying an article about The Seventh Day Adventist Health Study showing that, in vegetarians who ate nuts and whole wheat bread, greater nut consumption was correlated with lower CVD mortality. In fact, Seventh Day Adventists have lower blood cholesterol levels, lower fat consumption, and lower mortality rates than the general population.

His is not a fair assessment of the literature, but an exercise in cherry-picking and confirmation bias. The consensus of experts who have reviewed all the published evidence is that high blood LDL cholesterol is a major risk factor for CVD and that lowering it reduces risk, with a greater effect on those with multiple risk factors.

He says, “A hundred years ago less than one in one hundred Americans were obese and coronary heart disease was unknown.”  It was hardly unknown, although it was admittedly less prevalent and may not have been diagnosed with that name. He is correct that obesity has become a serious problem and is a risk factor for diabetes, heart disease, and a number of other illnesses; but he goes too far when he claims

  • that it has been “proven” that CVD, diabetes, and cancer are nutrition diseases;
  • that a low fat/hi carb diet is a human-made disaster comparable to wars and economic meltdowns and to natural disasters like hurricanes, the Ice Age and asteroid impacts;
  • that soda pop trucks are Weapons of Mass Destruction.

He plays the conspiracy card.  He says that scientists who question state-sanctioned paradigms like low fat diets, mutations as a cause of cancer, and HIV as a cause of AIDS are denied grants and silenced. In fact, some of them are very loud for people who have been “silenced,” and they may have been denied grants simply because their research was not as worthy as competing grant proposals.

Taubes says as long as you avoid carbs you can eat all the protein and fat you want and not gain weight. Miller goes even further, claiming that a high fat diet will make you lose weight. I think any influence on weight loss is due to a decrease in calorie intake; admittedly, the satiety from fat might reduce total calorie consumption, but this has not been proven to have any practical effect on dieters.

Miller says

The ideal caloric ratio between carbohydrates, fats, and protein is carbohydrates, 10-15 percent; proteins, 15-25 percent; and fats, 60-70 percent of calories, with the majority of them being saturated fats. Among the different kinds of fats, saturated fats and monounsaturated fats are good; except for omega-3 and a small amount of omega-6 essential fatty acids, polyunsaturated fats are bad in the high quantities that they are eaten in a Western diet, particularly industrially processed vegetable oils; and trans fats are terrible. Saturated animal fat is best obtained from grass-fed beef and pastured chickens, along with nitrate-free, additive-free bacon and sausage; and seafood from wild, not farm-raised, fish.

Some of this sounds like ideology based on the natural fallacy. He starts each day with two tablespoons of coconut oil. He claims that eggs from pastured poultry are much richer in nutrients, that organic isn’t good enough, and that the “Sacred Cow” transforms green grass and sunlight into vitamins A, D, and K and that this process is enhanced when cows are contented, grass fed, and not confined. He cites no peer-reviewed studies as evidence for these claims, but cites his own articles and podcasts and unreliable books by Enig and Ravnskov, and he shows a video of Julia Child talking about how much she liked the French fries from McDonald’s when they were deep-fried in lard. In passing, he mentions that aspartame is neurotoxic and causes obesity; I think he’s wrong about that.  He also links Alzheimer’s to a deficiency of saturated fat and asthma to a lack of the dietary fat needed to make lung surfactant.

He says:

Saturated fats work best for losing weight. In a randomized, double-blind trial comparing the effects of coconut oil and polyunsaturated vegetable (soybean) oil in women with abdominal obesity, women who consumed coconut oil had a significant reduction in waist circumference (with no change in cholesterol levels). Women taking vegetable oil had no change in their waist size and had a statistically significant increase in LDL cholesterol and reduction in HDL cholesterol (Lipids 2009;44:596–601).

This is cherry-picking and misrepresentation. The study doesn’t support his claim: it was a small trial and in fact it showed no difference in weight, but only a change in waist circumference.

Miller manages to ignore the mountain of data that resulted in the current recommendations for the prevention and treatment of heart disease. It is probable that the dangers of saturated fat have been exaggerated, but no reputable cardiologist would claim that saturated fat is good for you or that fat should make up 70% of the diet. Cardiologists have become more liberal about allowing some cholesterol and fat in the diet, but they still consider fat intake as part of the complex equation of risk factors. Fat contains 9 calories per gram compared to 4 calories per gram for protein and carbohydrates, and the calories can add up fast to contribute to weight gain.

Serendipitously, I have on my desk an uncorrected proof of a book by two cardiologists that is scheduled for publication in January.  They mention that recent studies have questioned the saturated fat/CVD link but they suggest that those studies only compared one unhealthy diet to another. Reducing saturated fat in the diet is not good for the heart if you replace the fat with refined carbohydrates and more calories. The heart-healthy diet that is supported by the best evidence is the Mediterranean diet, although even there the evidence isn’t perfect, because no good controlled prospective studies have been done to show that switching to that diet actually reduces mortality. They recommend avoiding trans fats and limiting eggs, fatty meat, saturated fat, and salt, consistent with the guidelines of the American Heart Association and every other credible source of diet advice.

I see Dr. Miller as an example of an oft-repeated pattern. A conventional doctor becomes frustrated with his inability to help some of his patients, and he becomes seduced by an overly simplistic answer based on anecdote and speculation; then confirmation bias does the rest.  He convinces himself that his new treatment plan is working. He cherry-picks the literature, writes books and articles, proselytizes to the world, imagines conspiracies against the Truth he has found, and generally becomes far more enthusiastic than the evidence warrants.  He is sure he is right, and considers his experience proof enough; whereas a true scientist entertains other possible explanations, engages in meaningful debate with peers who disagree, and even tries to prove himself wrong. Listening to Miller gave me déjà vu all over again, and made me feel rather sad.


Posted in: Nutrition

Leave a Comment (52) ↓