Interview Summary

View The carbohydrate-insulin model: a physiological perspective on the obesity pandemic paper

David Ludwig MD, PhD is Professor of Nutrition at the Harvard School of Public Health, and Professor of Pediatrics in the Harvard Medical School. He has published innumerable books and papers on nutrition, contributors to obesity and diabetes, and what might be done with both practice and policy to improve things. He has a real remarkable breadth and scope of his work. David, Time Magazine once named you a warrior in work on obesity. This is exactly how I see you as well. You’re really challenging the traditional ways of thinking, and as I said, you’ve broken new ground. So I’m proud to say that you and I have been friends for a number of years, and I’m also proud to say that we’ve written a number of things together. So thanks so much for being with us today. It’s a real honor to have you.

Thanks, Kelly. Great to be with you. And I’m sitting here in my office looking at a plaque I have on the wall of an op-ed we wrote for the Washington Post almost two decades ago, so it’s been a real honor and productive pleasure to know you.

The pleasure has been mine. So let’s talk about the paper. So in this paper, you and your co-authors challenged the widely-embraced energy balance model. So can you say what the energy balance model is?

Well, the notion of energy balance is really just a restatement of physics, the first law of physics that says, that speaks to energy conservation, and it’s commonly interpreted that in order to gain weight, you have to have a positive energy balance, that is you have to consume more calories than you burn off, and that to lose weight, you have to reverse that. You have to have a negative energy balance. You have to consume fewer calories than you burn off. But we argue first off that this doesn’t tell us anything about causality, cause and effect, what’s actually driving obesity. We use the example of a fever. Of course, a fever can only happen if the body generates more heat than it dissipates, more heat into the body than heat out of the body. But that’s obvious that’s, it’s, you know, we don’t need to be emphasizing that in textbooks. We don’t need to be teaching patients that notion. The question is what’s cause and what’s effect? And the conventional way of thinking is that the positive energy balance is driving weight gain, is causing obesity. So we’re surrounded by all these convenient, inexpensive, energy-dense, hyper-palatable, highly tasty foods. We lose control. We overeat them. We don’t burn off those excess calories with our modern lifestyle, and so those excess calories get forced into fat cells, and we gain weight. So ultimately this view considers all calories are alike to the body, and that we have to eat fewer calories, and ideally burn more of them off by exercise to address the problem. So that’s the conventional way of thinking.

So you have a different, and very science-based explanation for all of this that I’ll get to in a minute, but before we do that, why did the field come to adopt this energy balance model?

Well, it does seem to make sense, and certainly over the short term, we know that this way of viewing things applies. If you force feed an animal, or if we just intentionally overeat ourselves, we can gain weight, and conversely, if we put ourselves on a low calorie diet, we can lose weight for a while, but characteristically, we know the body isn’t a, you know, an inert energy storage depot. The body fights back in a dynamic way against changes in body weight and in energy balance, and this is something that almost every dieter has experienced, right? If it were just a matter of eating less and moving more, 150 calories less a day, that’s a serving of juice, 150 calories out more a day, that’s walking moderately for half hour, then virtually every weight problem should be solved within, you know, months to at most, a few years, but that’s not the case. Very few people can adhere to, can stay with low calorie diets for very clear reasons. The first thing that happens is we get hungry, and hunger isn’t a fleeting feeling. It’s a primary biological signal that the body wants more calories. And even if we could, those few of us who are highly-disciplined, and can resist hunger, the body fights back in other ways, most notably by slowing down metabolism, which means that to keep the weight coming off, even as we’re getting hungrier. We have to keep eating less and less, because the body’s getting more efficient. So the conventional way of thinking about things, all calories are alike, calorie in calorie out, just eat less and move more. Doesn’t seem to address the difficulty that people are facing, and recognize that despite a lot of attention to calorie balance, the obesity epidemic is getting worse and worse every year. I mean, the data just from the last year suggests that the weight gain during the pandemic was even faster than it was just prior.

Well, let’s talk for a minute about what’s at stake here. So vast numbers of people in the United States, both adults and children are overweight. This is increasingly becoming true of essentially every country in the world. The amount of weight that people have been gaining seems to be going up over time, and people find it very difficult, perhaps for the reasons you mentioned, to lose weight and keep it off, so it’s a pretty dire situation then, and given the health consequences of excess weight, and the psychosocial implications of things, there’s really a lot at stake here, isn’t there?

Certainly so. We know that in childhood, obesity can affect virtually every organ system in the body, and set the stage for a lifetime increased risk of diabetes, heart disease, even many cancers. Among adults, the majority, and in fact 70% of adults in the United States have at least overweight, if not obesity, and this is becoming a huge driver of the chronic health burden on the healthcare system, and which so many patients themselves experience, in terms of diabetes, risk for heart disease, fatty liver, orthopedic problems, sleep apnea. So we have a problem that has gotten so much attention, and yet keeps getting worse with every effort that we can bring to bear. My coauthors and I have this new paper in American Journal of Clinical Nutrition, arguing it’s time for new thinking. And the carbohydrate insulin model that we are proposing is perfectly consistent with the laws of physics around energy balance, but it suggests that we’ve been coming at the problem in exactly the opposite way than would be most effective.

So let’s talk about that. So if you have a different explanation than the traditional energy balance model, what is it exactly?

So the usual way of thinking, as we considered earlier is that overeating causes weight gain, and that certainly happens in the short term, but that model has a hard time explaining why people are gaining weight year after year, and their bodies are wanting to hold onto those calories. So we argue that a metabolic perspective would better explain this continuing creep upward in the so-called body weight set point. So the carbohydrate insulin model suggests that we’ve had it backwards, that overeating is not the primary cause of weight gain, that the body’s process of gaining weight, and storing too much fat is driving overeating. So overeating and a positive calorie balance certainly has to exist. That’s a law of physics, but it’s a downstream effect. It’s not at the source of the problem. And so this may sound a little surprising. How could the body gaining weight cause us to overeat? Well, let’s take the example of an adolescent during the growth spurt. We know a teenager might consume hundreds, or a thousand calories more than he or she might have a few years earlier, and that adolescent is growing really quickly, but which comes first? Is the overeating that that child is doing causing the growth, or is the rapid growth and the deposition of many calories into new body tissue causing that adolescent to get hungry and to eat more? Neither explanation violates any law of physics, but they have radically different implications to how we understand growth, and what we might do about growth disorders. In the case of the adolescent, it’s clearly the other way. It’s the growth that’s driving the overeating, and how do we know that? Well, Kelly, neither you or I, no matter how much we’re going to eat or overeat are going to grow any taller. So something in the body is regulating hunger, based on the needs of growth, and we argue that the same thing is happening in the case of obesity, that the aspects of our diet, importantly, including the processed carbohydrates that flooded our diet during the low fat years, that these are triggering fat cells in the body to hoard too many calories, to hold onto too many calories, so there are fewer calories available for the muscle, the liver, and the brain, and our body recognizes that. We get hungry, and we eat more as a consequence.

You mentioned the highly processed foods, especially carbohydrates that bombarded the American scene during the low fat craze. Explain more about that.

These processed carbohydrates, that at one point, just 20 to 30 years ago, people thought, and you can find many examples of this written in the literature. In fact, the first food guide pyramid is a clear illustration of the fact that all fats were considered unhealthy, because they have so many calories per bite, more than twice the calories per gram than carbohydrates. Whereas the bottom of the food guide pyramid, you know, we were supposed to eat six to 11 servings of grains, many of which were highly processed. Sugar was considered benign, and a good way to, and this is what they said, dilute out fat calories. The problem is that these processed carbohydrates, white bread, white rice, potato products, virtually all of the prepared breakfast cereals, and of course, concentrated sugars, and sugary beverages. So when you eat these foods in substantial amount, and it’s worse if the meal is also low in fat and protein, because they tend to slow down digestion. So if you just eat a lot of these processed carbohydrates, the body digests it into glucose literally in minutes. So blood sugar shoots upwards 10, 20, 30 minutes later, and that causes a lot of the hormone insulin to be produced. I sometimes refer to insulin as the Miracle-Gro for your fat cells, just not the sort of miracle you want happening in your body. We know that when a person with diabetes gets started on insulin, they’ll typically gain weight, and if insulin is given in excess dose, they’ll gain a lot of weight. So insulin is the hormone that promotes fat storage, and we argue that basically just endocrinology 101, all these processed carbohydrates, by stimulating more insulin than we would normally make on a less processed, lower carbohydrate diet, are driving too many of the incoming calories from a meal into storage and fat cells, instead of into muscle where they can burn. And so when you store, all it takes us to store one gram of fat too much a day to explain basically the whole of obesity, if one looks from childhood to adulthood.

So David, provide some context for this, if you would. So what fraction of the American diet is comprised of these kinds of foods, and what would that number be if people followed the recommended dietary guidelines you suggested?

Well, back in the 1950s, it’s not as if Americans were extremely healthy. We had much higher rates of heart disease, although much of that related to smoking, and we of course, had many fewer medications, and surgical procedures to help prevent or treat heart disease. But at that time, obesity rates were much, much lower, you know, about only one third of the rates they are today. And at that time in the 1950s, Americans ate about 40% of their calories as fat, and about 40% as carbohydrate, and maybe 15 to 20% as protein. Because of concerns around saturated fat and heart disease, which then got generalized to all fats being bad, well, we got the low fat diet of the 1980s, nineties, and the beginning of the century. Fat came down as a proportion of our diet. Carbs went up, but also the processing of those carbs. We got foods like the fat-free SnackWells cookies, a whole range of these fat-reduced products that simply took out fat, dumped in sugar and starch. These are after all processed foods, so they’re not going to be putting in fruits and vegetables. And these products were considered healthy. We ate them as we were told to eat them, and at that time, obesity rates really exploded. And we’re arguing that this is not just an association, that this change to our diet has played an important role in driving obesity, and that by bringing both the total amount of carbohydrates down, not necessarily a very low carb or ketogenic diet, but bringing them back down, maybe to what might oftentimes be characterized as a Mediterranean diet, focusing on getting rid of the processed carbs, eating more of the delicious and nutritious high fat foods, like nuts and nut butters, olive oil, avocado, even real dark chocolate. All of these high fat high, calorie foods look a whole lot healthier than the processed carbohydrates do in the best cohort studies.

You know, it’s a somewhat hopeful message, isn’t it? Because you’re not just telling people you have to eat less of everything, but there are actually some things that are quite delicious where you can eat more, and maybe that hope will lead more people to try this sort of approach.

That is exactly the issue with the conventional approach. If all calories are alike, and overeating is the primary problem, then we really just have to control our appetites. We have to discipline ourselves. Yes, clearly the conventional thinking recognizes that environment has a lot to do with it, and psychology of behavior, but ultimately, one way or another, you have to cut back on calories, because overeating is driving the problem. But if the driver is at the fat cells, if the foods that we’re eating are triggering our fat cells to store too many calories, and that’s what’s causing the hunger and the overeating, then just eating less doesn’t solve the problem, and it actually could make it worse by slowing down your metabolism. So this model argues that a focus on what you eat, not how much is more effective. You focus on controlling the quality of the foods, importantly, the processed carbs, but there are other aspects that can help hormonal and metabolic response. That’s what the person focuses on, and we let the body, based on our hunger levels, and satiety levels, determine how much we need to satisfy metabolic requirement.

So you’ve got what we call in the field a testable hypothesis, that people will do better if they follow the approach that you’ve mentioned, compared to the traditional approach. And you put that to a test in a study that we’re going to be talking about in a second podcast. But before we get to that, what sort of pushback, if you had, as your paper has been published, are corporate interests involved in this picture at all?

Yeah, let me just say that we recognize that these ideas are not fully proven. There are animal studies, we’ve done one of them that provides what we could call a proof of concept, that when you give rodents, and this has been reproduced by many different groups. This is a very rigorous finding. When you give rodents high glycemic index, versus low-glycemic index starch, so that’s fast-digesting, versus slow-digesting starch. You keep everything else the same, the ones that get the fast-digesting starch, that’s like, all of those processed carbs we’re eating that raise insulin a lot, well, they in fact show this whole sequence of events. Their insulin levels initially go up, they start getting fatter, and their energy expenditure goes down. They start moving less, and if you restrict their calories to that of the control animal, they’re still fatter, because more calories wound up getting stored than burnt in muscle. So they wind up getting more fat tissue, and less lean tissue, even at the same total body weight when you prevent their weight from going up. So we argued that there’s no way to explain that finding based on the conventional, calorie in, calorie out way of thinking. We need to examine whether this applies in humans, and to whom, you know? It may be that one model explains certain situations, or certain people better than the other, but it is a testable hypothesis. Unfortunately, this debate has become polarized, and we, in our article, specifically invite opponents to work with us on generating common ground. There’s plenty of basis for common ground already, and in our article, which is freely available online at American Journal of Clinical Nutrition. We put out a diagrammatic model in which each step leads to another step, and each of these steps is testable. So we can figure out what we got right, what needs improvement, you know, and where common ground is. After all, this is what science is supposed to be about, to come up with new ways of thinking for intractable problems.

You know, you reminded me when you talked about the animal studies of work that occurred many decades ago on something that people in the field were referring to as the cafeteria diet. And I remember the slide that I used for years in my own talks that was given to me by Ted Van Itallie, one of the pioneers in the obesity field, that showed a rat sitting on top of basically a junk food diet, where they take animals, and in the cage, they would put Cheetos and Hershey bars, and marshmallows, and things like that. And the animals would eat a lot of those things, and gain an enormous amount of weight. But people were really attributing the weight gain to the fact that these were highly palatable foods. The animals would eat a lot of it just because it tasted really good, and that would bring a lot of calories, and that was the reason for the weight gain. And what you’re saying is just, “Wait a minute, what happens to be that food that goes in there is a really important part of the picture,” And that’s been proven by controlling the calories in the experiment that you set.

Well, I think that’s a really great point that you raised that it’s easy to think in the cafeteria diet model, that the animals are getting fat because of the tastiness of the food, but these studies can’t distinguish tastiness, and whatever that means, and we could come back to that point, because tastiness is elusive. It’s a very squishy term to define, for reasons we can consider, but it’s impossible in these studies to distinguish tastiness from the nutrient content of the foods, and they tend to be full of sugar and processed carbs. In fact, the few studies that have aimed to disentangle this provide clear support for the carbohydrate insulin model that tastiness by itself, when you control nutrients, does not result in obesity, but the nutrients, even in a bland or untasty diet does result in weight gain in animals.

Fascinating science. So, David, what do you think are some of the main policy implications of all this?

Well, there has been push back. Some of that relates to just the difficulty of paradigm change, amidst scientific uncertainty. You know, we need ultimately to be all working together on all sides of this. But in addition, there’s resistance from the food industry that loves the notion that all calories are alike. All calories are alike, and there are no bad foods, and that you can drink a sugary beverage, have any kind of junk food, as long as you eat less of other things, or burn off those calories with physical activity. Whereas if this way of thinking, involving the carbohydrate insulin model, this opposite cause and effect conception is correct, then those foods have adverse effects on our metabolism above and beyond their calorie content. And that from that perspective, you really, can’t just outrun a bad diet, that we really need to be thinking about how our food is influencing our hormones and metabolism, otherwise we’re going to set ourselves up for failure, and that’s not a message that many, although not all in the food industry like to hear, because it requires corporate responsibility for helping to create the nutritional nightmare that confronts so many of us, and especially children throughout so much of their days.

You reminded me about an interesting parallel with tobacco here, where the tobacco companies, you know, long after it was known that cigarettes were killing people, just said that it’s not the tobacco that’s killing the people, it’s the fact that they’re just consuming too much of it, and the food companies have made very much that same argument. And then the tobacco researchers said, “No, tobacco is bad in any amount, and even a little of it can be harmful.” And that’s not totally true of the processed foods you’re talking about. I’m assuming people can have them in small amounts, but the parallel really kind of exists there, doesn’t it? That these things are risky, and dangerous really, after you go beyond whatever that small amount is, and then you’re going to have trouble, no matter what you’re doing elsewhere in your diet?

The metaphor with tobacco is useful to a point, although it can also elicit some strong responses, because obviously, tobacco products aren’t needed for survival, food clearly is. But I do think that there are some parallels that if these highly processed carbohydrates are undermining our metabolism, and also triggering, in part because of the metabolic changes. Fat cells communicate with the brain in many ways, including by releasing or withholding nutrients. If these foods are also triggering pathways in the brain that make managing calorie balance increasingly difficult, then we do really begin to need to think about food way beyond calorie issues, and that all calories aren’t alike, and that the food industry may indeed have to manage the food supply in a way that makes weight control easier rather than harder.

The paper we were discussing today was published in September, 2021 in the American Journal of Clinical Nutrition and is available to the public.