I’ve been around this diet and health realm for just a small fraction of time that Dr. John McDougall has been around. I remember sitting down with him with my dad while interviewing him for The China Study over a decade ago and appreciating one of his incisive observations from his years of practice. “People love to hear good things about their bad habits,” he said. How true this is.
This manifests itself in the preternatural ability for the same “Eat your favorite foods, get healthy” message to keep springing back in the popular nutrition world over and over. It is dusted off and resurrected every few years in some superficially different form and the darn thing just doesn’t go away. And while there is the evolution of message from book to book, the core is the same: Eat only your favorite, richest foods and you can lose weight.
Of course, today I’m talking specifically about the low-carb message. It reached mass awareness with Dr. Atkins, then has returned in the overlapping messaging of The South Beach Diet, Protein Power, The Paleo Diet, a couple of anti-wheat diets and now the eat fat diets. In conjunction with other books written in a more ‘scientific narrative’ style from journalists Gary Taubes and Nina Teicholz, the low-carb message has been enjoying something of a renaissance over the past few years.
This is the ebb and flow of popular nutrition information. Like the tides of the oceans, but much less calming, I am convinced that it is unavoidable. With my articles and books, I contribute to it as much as anyone else, but of course I’m on the side of plant-based diets. So let me explain how I interpret the latest round of low-carb messaging and why I remain disagreeable with its basic tenets. Let me specifically, too, comment on its latest incarnation: fat will save us from the evils of carbs, specifically sugars and grains. Of course, as is probably true with all persistent messages, there are some important truths and lessons to come out of the low-carb messages, and these are important to understand, acknowledge, and incorporate into our understanding of optimal nutrition.
It’s possible to think of the low-fat, near-vegetarian diet of the past half-century as an uncontrolled experiment on the entire American population, significantly altering our traditional diet with unintended results.
It’s like the myth that the Inuit don’t have heart disease; if you repeat a falsehood often enough it can very nearly seem truthful. The story has an intuitive appeal. We all remember frequently hearing messages to eat less fat over the past several decades. We have been inundated with fat free food products, including fat free cooking, candies, and drinks. But it doesn’t mean we followed the recommendations in a meaningful way.
In fact, we most certainly have not been eating a low-fat diet, and certainly have not been eating a near-vegetarian diet at any point in the past 50 years in America. That’s a ridiculous claim. To even make the claim, low-carb authors refer to data that show that Americans have been eating more calories over the past several decades, and that a disproportionate amount of those additional calories are coming from carbohydrates (sugar along with refined grains). This means that fat’s contribution to our total calorie intake, as a percentage of total calories, has slightly gone down.
But we haven’t been cutting the amount of fat we eat. This is from a paper finding that fat intake, as a percentage of total calorie intake, has decreased:
… although the percentage of energy from fat has decreased, the total amount of fat consumed has not decreased in the setting of an overall increase in energy intake, primarily from carbohydrates. Even normal-weight men and women consume at least 33% of calories from fat, which could be considered a high-fat diet as absolute fat intake has not decreased but the proportion is smaller because of the overall increase in energy intake (reference cited).
We’ve continued eating lots and lots of animal foods in addition to added fats. Added fat in the food supply has increased significantly since 1970, and the availability of cheese has dramatically increased over the past 40 years, according to the USDA (Figure 1).
Similarly, total meat consumption has increased over the past 40 years, as it did, dramatically, throughout the 1900s. While red meat has leveled off or decreased in recent times, we’ve more than made up for that with increased poultry consumption. This may be one of the most significant outcomes of the low-fat recommendations: Americans increasingly adding poultry to their diet.
Let’s say you have a friend who eats scrambled eggs, toast and butter, and sausage for breakfast, a lunch of cold cut sandwiches with cheese and mayonnaise, and dinner of roast ham and baked beans. It’s a good ol’ standard American diet. Then one day your friend starts drinking several bottles of soda throughout the day, in addition to making minor tweaks to their meals (grilled chicken with a smaller portion of ham, extra turkey sausage instead of normal sausage, margarine instead of butter, extra cheese). Would you say they’re truly experimenting with a low-fat diet?
It’s true that refined carbohydrate consumption (sugar, for example), has increased disproportionately over the past 40 years, but please don’t characterize that as a low-fat dietary trial. That’s essentially what the setup of the low-carb formula claims and it’s not truthful.
Many questioned Keys’ scientific conclusions, but he was vigorous in criticizing anyone who challenged him. He was a dominant, persuasive, and charismatic man who convinced the world of his hypothesis.
The next conversation in the high-fat, low-carb book is to explain how we got it so terribly wrong with the low-fat message, with a brief review of the evidence. Ancel Keys is a popular figure to attack, because his work on heart disease was prominent and well-known. Many popular authors level a bunch of personal attacks at Keys, suggesting essentially that the low-fat history was based on a highly manipulated presentation of the data, questionable honesty, and a bully that forced the world to his point of view.
Of course, the fat recommendation was based on far more than diet and heart disease. Populations around the world had been observed to a have lower rates of heart disease while consuming a lower fat, particularly animal fat, diet, but also lower rates of diabetes and cancer. Referencing both observational studies and animal experiments, the prominent 1982 Diet, Nutrition, and Cancer expert panel report said, “The committee concluded that of all the dietary components it studied, the combined epidemiological and experimental evidence is most suggestive for a causal relationship between fat intake and the occurrence of cancer.”
An important part of any low-carb message, then, is explaining away the data showing lower rates of common western diseases in populations eating more plant-based diets, lower in added fats and sugars. The way they achieve this is to say that observational studies show correlations, but cannot prove causation. This statement is absolutely correct, but don’t take the next step in believing that observational research should be thrown out.
If you look around the world and see that most healthy, trim traditional populations with low-rates of common chronic western diseases ate a relatively unprocessed, low-fat plant-based diet, then it seems to me that you have a very steep hill to climb to go on to claim that the polar opposite diet, a high-fat, high-meat diet, is actually the real key to longevity and health. Observations can generate hypotheses for further research and can also serve as important context to come back to for evaluating results from other types of studies.
Another strategy is for the low-carb proponents to point out the inaccuracies of observational research. This is also valid. Dietary surveys are quite inaccurate. These inaccuracies, though, are often likely to lead to a null result when a relationship between two variables may actually exist. When you find relationships between diet and disease, and start to see consistent findings that plants or plant-components are linked to healthier outcomes, it becomes a more convincing finding, especially when it has been shown that the relationship is biologically plausible.
Yet another strategy is claiming that research is too small, or too “controversial” (even if it’s just controversial among those who don’t like its findings). Reverse heart disease? Early stage prostate cancer? Turn bad genes off and good genes on? It must be too small a group, or too controversial. Unfortunately, all of these superficial arguments are employed by the naysayers.
Some science does matter, of course, according to low-carb advocates, and it comes in three flavors: Research showing that there is no relationship between fat and disease, research in favor of very low-carbohydrate diets, and science that seems related but is actually tangential to the claim at hand. Let’s call them the three sisters of the low-carb message. At the end you can harvest a menu plan full of meat and fat.
1) The “no relationship” studies: There’s lots of studies showing no relationship between diet and disease. Tons of them. Including some exceptionally well-done studies. Take, for example, the Women’s Health Initiative, which found that a modestly lower fat diet did not significantly reduce the risk of breast cancer or heart disease (Notice I said lower, not low, as they only got down to about a 25% fat diet briefly before slipping on the diet and ending the study closer to a 30% fat diet). This is a valuable finding. We now know that maintaining a western style dietary pattern and choosing low fat meats and low-fat dairy for a few years, using poultry instead of red meat, for example, is not likely to make a big dent on cancer, or on heart disease. They found that a transient, modest reduction in fat intake alone, while not really altering the overall dietary pattern (the diet remained very high animal food and very low in fiber and antioxidants), doesn’t do much. There was marginal, transient weight loss among subjects and no significant benefit on chronic disease.
Great! This is valuable, and I think this was an amazing study for its scope and its execution. But let’s be careful how we discuss these results. It doesn’t prove that fat is healthy. It doesn’t prove that there’s no benefit from all types of low-fat diets. It doesn’t really illuminate at all the promise of the original dietary patterns seen in observational studies for preventing breast cancer or heart disease. Those low-processed food, plant-based dietary patterns weren’t tested. What was tested was reducing fat intake very modestly, in relative isolation, without altering other components of the diet. Interesting? Yes. Useful? Sure. Proof there’s no harm from high-fat, animal based diets? No way.
Lots of studies show no relationship between diet and disease, and there’s lots of reasons for this to be the case, even if there is truly a strong relationship between diet and disease. You have to be careful about any argument that uses a null result as strong evidence after throwing away mountains of research in the previous breath.
2) Pro-low carb studies: There are certainly several short-term trials now showing that very low carbohydrate, high fat diets may result in lower weight, insulin and blood sugar and higher HDL (the “healthy” cholesterol, though this moniker is flawed). There are also some well-done studies  now showing that lots of added sugar is bad in numerous ways (added sugar is not the same as whole, fresh fruit). It’s more than just empty calories. It will rot your teeth, worsen your cholesterol panel and increase fat inside and on the outside of your body. It will make your diet more nutrient-deficient, contribute to excess calorie intake, and fuel an addictive food relationship that encourages processed food consumption.
In addition, there are studies showing the Mediterranean diet is healthy. Mediterranean diets are more plant-based with fish and added vegetable oils (canola or olive, mostly) and reduced animal foods in general. They have shown benefit for dementia and heart disease compared to the standard western diet. In fact, if you take someone eating a high-fat, standard western diet and tell them to consume a Mediterranean diet and even give them a liter of olive oil to consume weekly, the addition of oil has been shown to significantly reduce stroke risk compared to continuing the high-fat, western diet. This provides fuel for the argument that added fats are healthy.
The study mentioned, while useful and interesting, is extra-confusing for the public because the press communicating its findings calls the high-fat control diet a low-fat diet. Researchers suggested people in the control arm limit their fat, but none of the study subjects actually lowered their fat consumption in any significant way. They consumed a very high fat diet throughout and actually decreased, marginally, their intake of fiber throughout the study. Nonetheless, the intervention utilizing a Mediterranean diet with olive oil demonstrated reduction in strokes. And there was a trend to fewer heart-related issues as well. Taken together, there was a 30% reduction in heart attacks, strokes, and deaths from either of these. Regarding heart attacks specifically, the number of heart attacks was 3.1 per 1000 people per year in the olive oil group compared to 3.9 per 1000 people per year in those eating the control high-fat diet. Does this mean that olive oil is a keystone of the optimal diet?
A low-carb book usually doesn’t get into these questions. It suffices to simply cite the study and say that added oils reduce cardiovascular disease and leave it at that.
3) Tangential Information: And then a variable portion of science in the low-carb book formula is tangential biomedical facts. For example, we know that the fat is crucial in a wide variety of functions throughout the body. It is important in making surfactant, which allows lungs to function. It is important in brain function. In fact, the brain is 60% fat.
This is interesting and true, but does it have much bearing on nutritional intake? It makes for a catchy story. “Eat more fat to build a better brain,” we might hear. But it doesn’t work that way. We can’t eat lungs and expect to breathe better, for example. Nutrition and physiology is a bit more complicated.
Now you’re beginning to see how a fad diet book is created. You can claim that we’ve trialed a low-fat (almost vegetarian!) diet while getting fatter and more diabetic than ever, that we got there by worshipping observational research, which is useless to begin with. Accordingly, it turns out that personally flawed, compromised researchers, namely Ancel Keys, pulled the wool over the world’s eyes in the service of their own ego and agenda. Looking at the more recent research that counts, we can now put together a story with plenty of citations that shows no benefit of the prevailing recommendation to consume low-fat diets, the dangers of added sugars, and then claim short-term benefits on our biomarkers (weight, triglycerides, glucose, HDL cholesterol) with very low carb diets. We can even explain the benefits using tangential facts (feed the brain, which is fat, etc…).
The layperson can get mightily confused, very quickly. It seems so scientific!
It’s important to learn from the studies showing benefits of low-carbohydrate diets. Three of the main lessons include:
Unfortunately, in addition to discounting broad swaths of science, the low-carb claims just outright ignores other research.
Very low-fat, plant-based diets with minimal processed foods have been observed to correlate with very low heart disease rates. In animal studies and in human studies they lower cholesterol. They cause weight loss. Most importantly, they can reverse advanced heart disease , something that has never been shown with a high-fat, high meat diet.
Animal studies show that very low-carbohydrate, high-fat diets may improve biomarkers while at the same time decimating the arteries, growing atherosclerotic plaques at a far more prodigious rate than a lower-fat, higher fiber diet. We now know that plant foods rich in fiber increase the body’s production of endothelial progenitor cells, the repair cells that bolster the vital lining of your blood vessels. And we know that humans in some of these very low-carbohydrate dietary trials show no improvement, or even worsening, of their arterial function (flow mediated dilatation) after a period of time on the diet, even though other biomarkers may improve . Observational studies find that those people eating lower carbohydrate diets have poorer functioning small arteries and higher risk of death, particularly death from heart disease  .
We could go through a similar exploration of evidence ignored linking low-fat, unprocessed plant-based diets and reduced burden of diabetes and obesity.
Let me mention a few final facts about the current low-carb push for more fat. There’s a lot of discussion about fat and brain health. People ask me about risks for dementia and a low-fat diet, having heard that added fats are essential for brain health. Cardiovascular disease (blood vessel disease) plays a large role in dementia, and I just mentioned a small sample of research, which suggests the diet that is best for your blood vessels: a whole-food, plant-based diet naturally low in fats. This will help your brain. We also know that whole-food, plant-based diets naturally low in fat can prevent and reverse diabetes and metabolic syndrome, which are also linked to dementia.
East Asian/Pacific populations that have been noted historically to have lower rates of Alzheimer’s dementia traditionally have eaten a more plant-based diet lower in fat. Their diet and lifestyle is changing, with significantly increasing rates of Alzheimer’s dementia in the Pacific region. This comes after several decades of a dietary transition, characterized in China by increasing animal foods and added oils and decreasing whole grains and beans (and only more recently added sugars). In populations followed over time, those who eat more of a plant-rich Mediterranean diet, marked by subjects consuming more fruits, vegetables, whole grains, fish, and less red meat and dairy, more unsaturated fats vs saturated fats and moderate alcohol consumption have dramatically lower rates of cognitive impairment and dementia. This is a dramatically different diet from the low-carb diets popularly espoused. This is actually a relatively high-carbohydrate diet.
I have no concerns about risk of dementia on a no-added oil, whole-food, plant-based diet supplemented with B12. In fact, I have concerns about brain health in people consuming a diet with lots of added fat and animal foods.
Lots of evidence is left out or discarded in the low-carb formula. Even worse, evidence can be inaccurately presented, or often is presented without references, so the veracity of the claims simply can’t be checked. In a more recent popular book, Dr. Ornish’s study results were incorrectly characterized. The author writes that the low-fat group gained weight, when they actually lost weight. An important metabolic study was inaccurately described. The author writes that the low-carb subjects burned more calories, which was the opposite of the actual finding; the low-fat group burned more calories. I trust these are honest mistakes, but the consistent omission or discounting of contextual research is a more problematic, obviously deliberate, strategy of science denialism in most low-carbohydrate plans.
The most important part of the low-carbohydrate formula is to tell people they can eat their favorite “bad foods” abundantly and get trim and healthy. This is where the food guides and recipes come in.
In a recent popular book, readers are asked to eliminate all grains, sweeteners/sugars, beans (with a couple exceptions), starchy plants (with the exception of optional 2-4 cups, maximum, per week, of limited starchy foods) and all fruit (except for small amounts of select fruit). Other things are also off the menu, including refined oils, dairy, gluten, and alcohol.
Even with a salad every meal, restricting the starchy plants this much means that most calories provided by this diet will be from either added fats or meats. This is a low carbohydrate diet. It sounds great to suggest that readers eat lots of non-starchy vegetables, but even if people eat pounds of these non-starchy veggies, it will still only amount to a few hundred calories a day (and of course most readers won’t come close to consuming pounds). For most, the vast majority of calories consumed on this diet will be provided by fat and protein; in other words, it’s another low-carbohydrate, high-fat, high protein diet.
The recipes in this particular book are consistent with this. Other than snacks (2), sauces, and smoothies, which are based on fat, about 75% of all the recipes are centered on meat. The rest of the non-meat dishes are focused on fat (coconut oil, avocado, olive oil, nuts, etc…).
In addition, no fewer than 11 supplements are recommended at baseline, and four additional supplements are optional. You can conveniently buy the first 21 day supply for $213 on the author’s website.
So there you have it. If you can write clearly and communicate well, you too can now write a low-carbohydrate book. You better do it quickly, though. This incarnation of the low-carb diet will only be here for a short while. It’ll come back in the future as a slightly different message, perhaps with a focus on protein again?
Or perhaps, more optimistically, the argument for plant-based diets will continue to gain momentum, based on sound science, and we’ll truly have a future that moves us beyond the low-carb diet cycle. I surely hope so!
Copyright 2024 Center for Nutrition Studies. All rights reserved.