Go to main content
Type 1
Type 2

Gary Taubes on Calories vs. Carbs, The Case Against Sugar, and More

By Adam Brown, Brian Levine, and Kelly Close

Best-selling author on why he believes “calories in-out” is not an explanation for obesity, why we need better science, and nutrition recommendations for people with diabetes

Over the past 15 years, Gary Taubes has become one of the most vocal and insightful authorities in the nutrition field. Initially trained as a journalist covering physics, Gary has brought a meticulous scientific and historical perspective to all his books: Good Calories, Bad Calories (2008); Why We Get Fat: And What To Do About It (2011); and the newest, The Case Against Sugar (2016). Each book is masterfully researched and chock-full of “Wow! Seriously?” moments – Gary is a student of history and science, and it’s remarkable what he digs up. He shows how certain people, events, misinterpretations, and measurement techniques can color a field for years.

Although he is happy to admit that the science behind his arguments is not yet completely rock solid, three of Gary’s biggest messages are:

  • We need far, far better nutrition science and long-term studies answering important questions. “The nature of a healthy diet should not be a matter of opinion but of fact,” as he notes in our interview below.

  • The common explanation of weight gain, “eating more calories than are burned,” is meaningless – it’s similar to saying “a room gets more crowded when more people enter than leave.” We want to know why the room gets more crowded. Gary argues for a “hormonal hypothesis” of obesity, based on the idea that excessive intake of highly-refined carbohydrates (e.g., sugar, white flour) leads to insulin secretion, causing accumulation of fat. Despite decades of research behind this idea, Gary argues it has been largely ignored and forgotten.

  • Instead of focusing on dietary fat, Gary suggests we should be focusing on the quantity and types of carbohydrates in our diets (especially sugar).

In this interview, we sat down with Gary to discuss his outstanding and insightful new book, The Case Against Sugar; the state of nutrition science; recommendations for people with diabetes; and far more! Click to jump to a particular section.

I. The Case Against Sugar – Is it a direct cause of type 2 diabetes and obesity? Is it addictive? What policies would make the biggest difference for our nation’s health? How much sugar is too much?

II. On Calories In-Out, Nutrition Studies, and Guidelines – Why do you believe that "calories-in, calories-out" is meaningless, and how is it different from a hormonal hypothesis of obesity? What would Gary change in nutrition guidelines? What studies would Gary hope to see funded? What about critics who say “nutrition science is too hard?”

III. Questions and Next Steps – What nutrition question has Gary been asking lately? What does he think about Virta Health? What does Gary wish people would ask him?


diaTribe: Your book points the finger at sugar as a direct cause of type 2 diabetes and obesity. If you had to rate the evidence on a scale of 1 (we have no clue if that is true) to 10 (it’s 100% likely), where would you put it?

GARY: A 7 or an 8. I think the evidence is more compelling than it ever was for the hypothesis that dietary fat causes heart disease, or the kind of argument that the American Heart Association is now pushing that we should replace the saturated fat we consume (butter, dairy, and animal products) with polyunsaturated fats from vegetable oils.  

One way to think about this question is to ask whether [we should conduct] a randomized controlled trial to test the hypothesis.

This trial might randomize a few tens of thousands of people to consume a relatively sugar-free diet and compare them with an equal number consuming sugars as usual. It would have to run long enough (e.g., 10 years) to see if there is any meaningful difference in mortality and other health outcomes. It would be a difficult trial to do, but with some innovative thinking and enough money, it could probably be done.

diaTribe: Is sugar addictive? What does the research tell us?

GARY: My favorite line about this is from Charles Mann, an exceedingly talented journalist/historian who wrote in his last book (1493) that scientists debate today whether sugar “is actually an addictive substance, or if people just act like it is.” We can state pretty conclusively that rodents find sugar as addictive as any other drug of abuse, including heroin, but the research on humans is harder to do and less well developed. But, as Mann said, we certainly act like it is. And if you have children, as I do, I’m not sure I care what conclusions the researchers reach on this. Clearly, sugar has a hold on my children, or at least one of them, that no other food does. (Not even burritos, although, I’ll admit, it’s close.)

diaTribe: If you were appointed to lead nutrition policy in the US, what changes do you think would make the biggest difference on the health of our nation? 

GARY: This is the kind of question that requires far more thought than I’m about to give it. But with that caveat, here goes:

  1. Convince all Americans to cut back on their sugar consumption dramatically, particularly sugary beverages.

  2. Take the focus of our attention off saturated fat and put it on the quality and quantity of carbohydrates we consume.

  3. Improve the quality of nutrition science (and scientists) such that all these issues of controversy can be resolved with rigorous experiments. The nature of a healthy diet should not be a matter of opinion but of fact. And for those who insist we know what a healthy diet is and it doesn’t need to be tested, I counter that when 100 million or so Americans are obese and 30 million have diabetes, we clearly don’t understand something about healthy eating.

  4. Get the obesity research community, physicians, and public health authorities to understand that obesity has to be a hormonal-regulatory disorder, with insulin playing a primary role. To say that obesity is caused by consuming more calories than we expend is almost incomprehensibly naïve.

  5. (repeated for emphasis) Convince all Americans to cut back on their sugar consumption dramatically, particularly sugary beverages.

diaTribe: If someone came up to you and said, “I just want to know for my kids and myself: how much sugar is too much,” what would you say? (Or put differently, what is your personal threshold?)

GARY: I honestly don’t know how to answer this question. The last chapter of The Case Against Sugar is a meditation on what it means to say don’t eat too much sugar when it’s quite possible that sugar is the fundamental dietary trigger of diabetes — i.e., without sugar in our diets, diabetes would be as rare a disease as lung cancer would be without cigarettes.

We don’t say smoking too many cigarettes causes lung cancer; we say smoking does. And we say it because, well, it does. The evidence is convincing. And we don’t counsel smokers to cut back so that they’re not over the “too much” limit; we counsel them to quit.

If sugar causes diabetes, does the concept of “too much” mean anything? And, of course, even if it does, it’s going to be different for everyone and we won’t actually know if we’ve consumed too much until we’re diagnosed. And what if it is addictive or at least so habit forming as to be effectively addictive?

So I’m ducking the question. For myself, I basically don’t eat or drink sugar. I lost my sweet tooth, so I don’t miss it. And I don’t want to drink or eat it because then my sweet tooth will come back, and likely with a passion. I try to keep the sugar consumption of my kids pretty low, but they eat and drink it (particularly when I’m not around). Keep in mind, though, that we have no family history of obesity and/or diabetes, and so whatever damage sugary sweets and beverages might do to my kids now can probably be undone later, and maybe these “moderate” amounts of sugar won’t do any harm. If we had a family history, I’d probably be a little less accommodating with what my kids eat.


diaTribe: All of your books note that the calories-in, calories-out model of obesity is descriptive, not explanatory – it’s like saying a room gets more crowded when more people enter than leave. But what we want to know is the WHY – why does the body gain weight? You suggest that a “hormonal hypothesis” of obesity – driven by the different impact of carbs, fat, and protein on fat storage – plays the central role. Why has this proven so hard to get accepted? Will we ever know the answer?

GARY: This “calories-in, calories-out” notion is meaningless. To say something like “excess calories cause accumulation of excess body fat” is a circular statement. It’s logically identical to saying that “excess money causes accumulation of excess wealth.” If you read that in an article about economics, you would think you were reading a satire. This model of obesity causality tells us nothing. It should be a crisis for the field.

The obesity research community, in their belief that this “energy balance” hypothesis had to be true, neglected to incorporate the breakthroughs in endocrinology that occurred in the 1960s and 1970s after Rosalyn Yalow and Solomon Berson developed the radioimmunoassay. For the first time, this allowed researchers to measure hormone levels in the bloodstream accurately. By 1965, it was clear that insulin was the hormone primarily regulating fat accumulation in fat cells (adipocytes), and this pointed the finger at insulin as a primary player in obesity – which is, after all, a disorder of excess fat accumulation. The implication, though, was that the carbohydrate content of our diets was perhaps uniquely fattening. The dogmatic believers in the energy balance notion had decided such a thing could simply not be true.

As such, by 1980 or so the research on the hormones and enzymes that regulate fat accumulation in fat cells was rendered by the authorities to be irrelevant to a disorder of excess fat accumulation. Researchers today still neglect to discuss the basic regulation of fat metabolism when they discuss causes of obesity, even though it is, in effect, textbook endocrinology.

I don’t think the scientific community will ever make meaningful progress on understanding obesity, its treatment, or prevention, without going back to this very fundamental 1960’s-era science and incorporating it back into their thinking. Virtually all animal models of obesity support this endocrine-regulatory (carb-insulin) hypothesis of obesity, but the researchers don’t pay attention to this literature. The same kinds of studies have to be done in humans, too.

The catch is that human trials come with all sorts of issues that animal trials don’t. And to do this work rigorously is expensive and time-consuming, perhaps an order of magnitude more so than animal experiments. As such, the NIH is loath to fund it and most researchers – trapped as they are in this energy balance thinking – don’t actually see the necessity. They will simultaneously insist, as an Endocrine Society scientific statement recently did, that the calories-in, calories-out hypothesis must be right, while also acknowledging that they don’t actually know why people gain weight or how to reverse it. I find the situation perverse almost beyond understanding, despite spending much of the last 17 years trying. If the studies are to be done, it likely will be those of us living on the outside of the establishment that make them happen.

diaTribe: Based on what you’ve learned, if you were to re-write the US Dietary Guidelines and nutrition guidelines from professional societies (e.g., American Diabetes Association, American Heart Association), what are the biggest changes you would suggest?

GARY: This one is tough. I’d have them say that for those who are overweight, obese, living with diabetes, or likely to become so in the future, the healthiest diet is one that avoids added sugars and replaces most to all of the remaining carbohydrates with healthy sources of dietary fat. And then we’d have to debate what the word “healthy” implies. And we’d have to point out that “overweight, obese, living with diabetes, or likely to become so in the future” constitutes the great bulk of the American population.

diaTribe: If you had an unlimited budget, what are the three nutrition studies you would love to see done?

GARY: They fall into two categories. One is the kind of study where we’d look at the effect of major elements of the diet on hard end points – not just risk factors. You must run the studies long enough to see whether more folks get heart attacks or cancer or die prematurely in one group rather than another. For instance, saturated fats are supposed to be bad for us — at least so still says the American Heart Association — and polyunsaturated fats from vegetable oils (corn, canola and soy oil) are supposed to be good. Before we tell the entire American public to consume vegetable oils, which are relatively new to human diets and certainly new compared to saturated fats from animal sources, maybe we should make sure they’re really not harmful. Let’s spend the money to find out. This would be another randomized trial with tens of thousands of people followed for a decade or so. It could be done with the will and the resources.

By the same token, there are those like me who believe that very low-carb, high-fat diets are good for us and these tend to be rich in animal products as a source of fat and protein. And then there are those who think we should eat mostly plant diets or even vegan or vegetarian diets for our health. While I would say that the likely dietary trigger of obesity, diabetes, heart disease and cancer are the carbohydrates we consume — specifically sugars and refined grains — these folks would argue that it’s the meats in the diet and the dairy products. (Of course, folks who follow vegan/vegetarian dietary instructions tend to give up sugars and cheap flour products as well.) This can all be tested.

I think before we tell Americans to consume vegetarian or vegan diets for their health, when these, too, are relatively new concepts for most human populations, we should test to see if they really are beneficial, or if the benefit comes from avoiding the same processed carbohydrates that are avoided in low-carb, high fat and paleo diets.

The second type of experiment that needs to be done is those that can disassociate the effect of the carbs, fat, and protein we eat on fat accumulation vs. the calories in our diet. The conventional wisdom on why we get fat is that we consume more calories than we expend. So these studies would directly test this belief that obesity is an energy balance disorder (calories in/out) and its implication that the amount of calories is the critical variable in determining fat gain, rather than just a convenient measure of the amount or dose of the various carbs, fat, and protein consumed. As with many controversies in science, whether or not we know the answer will depend on who you ask and what they believe today.

diaTribe: What do you tell people who say, “Nutrition science is too hard to do well?” or “These studies are too difficult?”

GARY: I recently had a discussion with a major proponent of plant-based diets, who argued that these randomized trials are just too difficult and expensive to do. He said we have to give the best advice we can based on the incomplete evidence and that I was dreaming if I thought these studies could ever be done. Here’s what I told him:

“How about this for something I am suggesting: science is exceedingly difficult to do right, and the questions we have left to answer happen to be those that are ludicrously expensive to test experimentally and exceedingly difficult to do right. But that’s not an excuse for not spending the money and putting in the work if it’s important, and it’s certainly not an excuse to say that it’s not necessary. 

“The physics community spent roughly $10 billion on a single experimental device – the Large Hadron Collider at CERN – to discover (i) whether the Higgs Boson exists (which it virtually had to according to the Standard Model) and (ii) whether there was something new they could learn beyond the standard model that might be a clue to understanding the universe in a way they don’t already. They did this because we have decided that it’s worth it.

“The nuclear physics community has so far spent roughly $50 billion developing nuclear fusion because we think nuclear fusion is important to the future of the human race. It’s likely another $50 billion will have to be spent before we have a working fusion reactor that actually produces more energy than it consumes, assuming such a thing is possible, which the $50 billion spent so far doesn’t tell us. We do this because we have decided that these are vitally important to who and what we are as a species and, in the latter case, to our survival as a species.

“Pharmaceutical companies regularly spend $1 billion to develop drugs to treat diseases that are related to obesity and diabetes, and that will often be used by a small percentage of the population afflicted.

“The US Fitzgerald, the destroyer that recently collided with a cargo vessel off the Sea of Japan, cost $1.5 billion and is one of 62 similarly expensive ships in its class. Our nation spends this money because we think it’s important to the defense of the US population and, well, world peace.

“Obesity and diabetes alone supposedly cost the US healthcare system roughly one billion dollars PER DAY.

“Isn’t it worth spending this kind of money to find out whether we’re fooling ourselves about the nature of a healthy diet, particularly as these obesity and diabetes epidemics are blowing up worldwide?”

I’d bet in this day and age we could do five to ten randomized controlled trials for one day’s worth of direct medical costs from obesity and diabetes and, with them, answer reliably virtually all of the critically important nutrition questions.

Coming from my perspective – trained to think by physicists and having spent much of my professional life studying episodes in which scientists fooled themselves and so simply got the wrong answers – I do not understand why we wouldn’t do such tests, and why we wouldn’t want to argue that they’re necessary. They’re inexpensive compared to the money we spend on other vital questions and issues.


diaTribe: What nutrition question have you been asking lately?

GARY: With the rise of the vegan/vegetarian movement and propaganda films like What the Health, I want to know whether eating a vegan or vegetarian diet is really healthy and whether the benefits outweigh the risks. And not just whether it’s healthy for adults, but whether it’s healthy for kids, because some of the adults who get scared out of eating animal products by What the Health are going to feed their kids vegan/vegetarian diets as well.

My bias is these diets are going to cause health issues for a substantial proportion of the folks who try to live on them, but that’s just my bias. There are precious few clinical trials addressing this question, and as more and more young adults go vegan or vegetarian, we’re going to have more young parents putting their kids on these diets, and we have no idea what kind of health issues that might cause. Maybe it’s fine and the proponents of this way of eating are absolutely right, but it’s a radical departure from how humans have eaten through virtually all of the last 2 million years. I think it’s urgently important to know the risk in relationship to the purported benefits.

diaTribe: Are you optimistic about companies like Virta Health that try to take nutrition into the realm of industry, align incentives, and prove the value of eating fewer carbs?

GARY: I think Virta Health is as likely as any organization to change the world. If their business model is viable, and that remains to be seen, and they continue to demonstrate that they can get people with type 2 diabetes off their meds and losing significant weight in the process, and if they continue to show that they can save insurers, corporations, and payer-providers money, then their approach is going to spread widely.

By targeting folks with type 2 diabetes and showing that they can reverse this disease, get people off meds, and make them healthy, then it negates or end-runs arguments that there’s something inherently unhealthy about eating low-carb, high-fat diets. It also counters the argument that the healthiest diet is a vegetarian or vegan diet, which would imply the problem is animal products and not the carbs we’re consuming.

Thanks so much for joining us, Gary! Those interested in learning more should definitely pick up The Case Against Sugar – it’s a fantastic and illuminating read.