Interview Summary

Access the study: https://doi.org/10.1093/ajcn/nqab287

I’ll begin by asking a question fundamental to this work. Why care so much about carbohydrates?

Great question, Kelly. Carbohydrates amount to at least half the calories in a typical diet today, which is interesting from a historical and evolutionary perspective. Because of the three major nutrients we eat, protein, fat, and carbohydrate, carbohydrate is the only one for which humans have virtually no requirement. Think of Northern populations, especially in the Ice Ages but also up to recently, such as the Inuit, that had access to only animal products and could eat plant products like berries maybe one or two or three months a year at most. So for nine months a year, they were eating only fat and protein. And yet, those populations were healthy. The women were fertile; they could breastfeed. And children grew normally. So recognizing that there’s no absolute requirement for carbohydrates, the question then becomes: How much carbohydrate and what kind would be optimal for health and allow for the greatest flexibility, diversity and enjoyment in our diets?

So David, if the body doesn’t have an innate need for these, presumably there’s no biological driver to go out and seek these, why in the heck are people eating so much of this?

Well, carbohydrates are delicious. And the food industry certainly knows that and has taken advantage of that. In fact, when you step back and ask: What are the foods that we tend to binge on? They may have a combination of key flavors and nutrients. Oftentimes, we hear sugar, salt and fat. But I’ll argue that there are virtually no binge foods that are just fat. Do people actually binge on butter? I mean, butter is very tasty. You might enjoy an initial bite. But very few people, perhaps with the exception of a major eating disorder, would sit down and eat a quarter pound, a stick of butter. But there are all sorts of high-carbohydrate binge foods. Sugary beverages are 100% sugar. Bread, baked potato chips, popcorn, especially the low-fat versions, these are easy to binge. And from one perspective, the key difference is the hormone insulin. Fat does not raise insulin. And so fat is digested slowly, and doesn’t get directly stored in large amounts into body tissue. It has to be metabolized more slowly. Whereas carbohydrates, especially the processed ones, when eaten in large amounts, raise insulin to high levels. That insulin directs those incoming calories into storage. And a few hours later, blood sugar crashes and we get hungry again and are ready to have another blood sugar surge by indulging the next time in those foods.

So what question specifically was your study designed to address?

We conducted a large feeding study that had two parts. The parent study had 164 young and middle-aged adults, who were at least a little bit overweight, ranging from overweight to having obesity. And the first thing we did was bring their weight down by providing them all of their foods, delivered foods to their home, in a calorie-restricted way. You know, you cut back calories, and of course you’re going to lose weight for a while. It doesn’t address why people get hungry, and why they regain weight. But in the short term, we cut their calories, and they lost 10% to 12% of their weight. Then we stabilized them at their new, lower body weight, and then randomly assigned them to one of three groups: low, moderate or high-carbohydrate diets. And we kept them on these three different diets for another five months. And during this time, we were again delivering all of the meals to the participants. This was over 100,000 prepared meals throughout this time, so it was a really major effort. And during this low, moderate, and high-carbohydrate diet period, we adjusted calories to keep their weight the same. We wanted to keep them at that weight-loss anchor, 10% to 12% below where they started. The first study looked at what happened to their metabolism and their energy expenditure. And we found that when people were on the low-carb diet at the same weight as the other groups, they were burning about 200 calories a day more. So the study raised an interesting possibility, that the kind of calories you eat can affect the number of calories you burn. That from a biological perspective, all calories are not alike to the body.

David, this is fascinating work. I’d like to ask a strategy question. So this was an extremely intensive study of 164 people. And you mentioned the people were provided all their meals, very careful measurement and things like that. So the same amount of money, you could have studied many more people but just done a less intensive study with less supervision and fewer measurements of outcome. So why do the study in such an intensive way?

Right, there’s always going to be a trade-off in design considerations. And you’ve identified a classic trade-off. You can study fewer people more intensively, or more people less intensively. Most weight loss trials have chosen the second route. They take a lot of people, and they try to study them for a long period of time, or at least some of them do: a year or ideally two years or longer. The problem is that without an intensive intervention, so what are we talking about? These studies would oftentimes have participants meet with a nutritionist once a month. They would get written educational materials, and maybe other kinds of behavioral support. But that’s about it. And without greater levels of support and intervention, people characteristically can’t stick to these diets over the long term. Maybe they make changes for two, three or four months. But by six months or a year, they’re largely back to eating what they were originally. And the different diet groups don’t look much different. So if the groups didn’t eat in much of a different way throughout most of the study, why would we expect to see any differences in outcomes, such as weight or energy expenditure, or cardiovascular disease risk factors? So these studies don’t test a dietary hypothesis very well. It leads to the mistaken conclusion that all diets are alike. Really, what the conclusion of these studies has to mean is that we need more intensive intervention in our modern toxic environment, if you will, to promote long-term change. And it’s only when we get that long-term change can we actually figure out which diet is better and for whom.

So you’ve explained how the study was done and why you did it. What did you find?

So the first leg of the study, which was published in “BMJ” late in 2018, so just before the pandemic, showed that the kinds of calories you’re eating can affect the number of calories you burn. And, that by cutting back on the total and processed carbohydrates, you can increase your metabolic rate. And that could be a big help in the long-term management of a weight problem. You know, you want your body on your side rather than fighting you when you’re trying to maintain weight loss. And a faster metabolism would be a tremendous help if this is a reproducible finding and applies to the general population. We recently published in the September “American Journal of Clinical Nutrition,” a second part of the study. And that asks: How do these different diets, low, moderate and high carbohydrate, affect cardiovascular disease risk factors? It’s one thing to lose weight. Maybe a low carbohydrate diet helps you lose weight. But if your cardiovascular disease risk factors go up, that might not be such a good trade-off. So that’s the aim of the second study. Because low-carbohydrate diets are often very high in saturated fat. So we wanted to find out what were the effects of this low-carbohydrate, high-saturated-fat on a range of risk factors.

So tell us specifically some of the cardiovascular risk factors that changed. And if you would, place the changes that you found in your participants in a context. Like are these big-deal changes? Are they small changes? Or put it in context, if you would?

The big problem with saturated fat is that it clearly raises LDL cholesterol, low-density lipoprotein cholesterol, which is a classic cardiovascular disease risk factor. It’s the main one that’s targeted by many of the drugs, such as statins. Yeah, I think there’s no question that on a conventional high-carbohydrate diet, a lot of saturated fat is harmful. So the combination of bread and butter is not a good one. But the question we wanted to ask was: What happens when you get rid of a lot of that bread? Does the saturated fat still comprise a major risk factor? And so our low-carbohydrate diet was exceptionally high in saturated fat, as is characteristic of how these low-carb diets are usually consumed. It had 21% saturated fat, which compares to the 7% saturated fat on the high-carb, low-fat diet that’s oftentimes recommended to people at risk for heart disease.

So what did we find? Well, the first thing we found was that LDL cholesterol was not adversely affected at all. There was no difference in LDL cholesterol between those getting 21% versus 7% saturated fat. Suggesting that when you substitute saturated fat for processed carbohydrates, from the standpoint of this key risk factor, it’s pretty much a wash. However, the low-carbohydrate, high-saturated-fat diet benefited a range of other risk factors that go along with what we call the metabolic syndrome, the insulin resistance syndrome. Specifically, we saw strongly significant, from a statistical perspective, improvements in triglycerides, that’s the total amount of fat in the bloodstream; HDL cholesterol, that’s the good cholesterol that you want to be higher; and other lipids that indicate overall levels of insulin resistance. Suggesting that insulin resistance was improving. And we know that low-carbohydrate diets show promise for diabetes in other studies, in part because they do tend to improve insulin resistance and lower blood sugar. And so our study suggests that if you are pursuing a low-carbohydrate diet, and we can talk about the different degrees of restriction of carbohydrate, and at the same time you’re reducing the processed carbohydrates, then the saturated fat might not really be such a problem.

So then if you take all this information in this, as I said, exquisitely designed intensive study and distill it into what dietary recommendations would be, what do you think is a reasonable proportion of fat, carbohydrate and protein in the diet? And what sort of things should people think about as they want to lose weight and keep the weight off?

One key qualification I need to mention is even though this was an intensive study with a relatively large number of people for a feeding study of this magnitude, we still don’t know how generalizable these findings are to people at different ages, different body weights, different levels of susceptibility. So no one study can inform a change of clinical practice like this, especially in the world of nutrition where there’s so many complicated and interacting factors. I will also venture to say that there’s no one diet that’s going to be right for everybody. We know that some people can do perfectly well on a high-carbohydrate, low-fat diet. I mean, think of the classic Asian agrarian societies where rates of obesity and diabetes are very low. But those societies tend to be highly physically active and the people insulin-sensitive. America is characterized by high levels of overweight and obesity, sedentary lifestyle. And these create insulin resistance as a highly prevalent problem. For societies such as ours, we think that high-carbohydrate diets that are raising insulin levels on the background of insulin resistance is a recipe for metabolic problems. And so for Americans, especially those struggling with weight, pre-diabetes, and even more so diabetes, a reasonable first step is to cut back on the processed carbohydrates. And I think that’s an intervention that increasingly few experts would argue with. We’re talking about concentrated sugars and refined grains. Where we start to get into the controversy is whether carbohydrates should be further reduced down to say 20% as in our study, which still leaves room for some unprocessed grains, beans, and a couple of servings of whole fruit a day, or even lower to what’s called the ketogenic diet that’s less than 10%. And that’s where you really have to give up most carbohydrates and focus just on the proteins and fats. I think for people with diabetes, such a strict approach looks appealing in preliminary research studies. But again, this is going to need more research. And I would caution anybody with diabetes or anybody who’s thinking about a ketogenic diet to discuss these kinds of dietary changes with their healthcare provider.

I realize your study wasn’t meant to address this issue that I’m about to raise, but I’d appreciate hearing your instincts. One key, of course, to any recommended nutrition plan or diet, if you’d like to call it that, is whether people will stick to it. What do your instincts tell you, or data if you have it, on how readily people can adhere to this sort of an approach over the long term compared to other kinds of approaches?

Great question. And I’ll approach that by saying: We all understand that if diet is a problem that’s contributing to obesity, diabetes, heart disease, other chronic health problems, then we have to change our diet in one way or another regardless of what the mechanisms are. So I’ll return the question to you by saying: Which do you think is going to be easier for most people over the long term: cutting back calories, getting hungry and trying to ignore that very intense drive to eat, or getting rid of certain kinds of foods that may be triggering hunger and making it so much harder to stick to a lower calorie intake?

As a doctor, as a pediatrician, and as a researcher, and also myself, I try to do N of 1 experiments on myself with any kind of a nutrition approach I might use with patients or with research participants. I’ve found that it’s so much easier to just give up the processed carbohydrates and enjoy a range of other very satisfying, delicious, higher fat foods. And oftentimes, in my experience personally and I hear as reported by patients that the cravings for these highly processed carbohydrates go down. And lastly, I’ll just say, it’s not that these processed carbohydrates are inherently so irresistibly delicious. I mean, white bread, these common binge foods, white bread, unbuttered popcorn, baked potato chips, even though these are almost 100% carbohydrate yet they’re commonly binged on not because they’re so incredibly tasty. But I would argue because they’re producing changes in our body that are driving overeating. So it’s not that they’re so tasty and we’re getting so much enjoyment. We’re eating these foods because we’re driven to metabolically. And once you come off that blood sugar rollercoaster, it becomes much easier to say no.

When you mentioned before that with one approach, you’re kind of fighting your body; and another approach, your body is becoming your ally in this process, I thought of going to the beach and, you know, you can go out and try to swim against the waves coming in, or you can ride the waves toward the beach. And one, of course, is a lot easier than the other. And it sounds that’s kind of what you’re talking about, isn’t it?

When you line up biology and behavior, and clearly behavior, psychology, and our food environment are all factors that are going to have to be addressed. We don’t want to make it much harder for people. So we do need to think in systems dynamics: the food supply, the environment. But on a strictly individual level, when you line up biology with your behavior, the effort required to accomplish your goals becomes less. You know, this is characteristic of so many areas of medicine and research. This is why we aim to identify the cause of a problem when you treat a cause. So let’s use the example of fever. Fever you could say is a problem of heat balance: too much heat in the body, not enough heat out. And so from that perspective, you could treat any fever by getting into an ice bath. Couldn’t you, right? The ice would pull the excess heat out of your body. But is that an effective treatment for fever? No, of course not. Because your body’s going to fight back violently with severe shivering, blood vessel constriction. And you’re going to feel miserable and you’re going to get out of that ice bath quickly. In the case of obesity, the timeframe is much longer, but similar kinds of responses occur. The body fights back against calorie restriction because calorie restriction, according to this way of thinking, is an effect. It’s not the cause. If the cause is the body’s been triggered to store too much fat, then we have to address that problem by lowering insulin levels and producing a more stable blood sugar pattern after eating. If that happens, then the effort that you put into cutting back calories goes a lot further.