Last month, the American Journal of Clinical Nutrition (AJCN) published a perspective piece that stirred up tensions in the worlds of epidemiology, physiology, and nutrition. Doctors and researchers argued on Twitter; commenters defended and derided related opinion pieces, and lay audiences read headlines such as, “Overeating isn’t the primary cause of obesity” and “Study finds primary cause of current obesity epidemic.”
The paper’s lead author, David Ludwig, MD, PhD, says the journal saw more than 140,000 views on the paper in its first few weeks of publication, which he said, “for an academic nutrition journal, is exceptional.”
But there’s more to this perspective piece than its reception. While supporters applauded what they say is a long-overdue reset of the narrative on weight and metabolism, the model and its central premises prompted fierce criticism.
The controversy reveals just how much more there is to learn about how the body responds to food, even as all sides ultimately agree on basic nutritional and metabolic interventions.
The Carbohydrate-Insulin Model
In the piece, Ludwig, an endocrinologist and researcher at Boston Children’s Hospital and professor of pediatrics and nutrition at Harvard, lays out a “carbohydrate-insulin model” of obesity. In it, insulin plays the lead role in a chain of events that starts with high-glycemic carbohydrates and ends with conservation of energy, increased hunger, and ultimately, more fat storage.
This increased hunger and diminished energy expenditure begets a vicious cycle that can be averted with a low-carb, low glycemic diet. “Given the choice between bread and butter, for years we focused on getting rid of the butter,” said Ludwig. “But maybe between the two, the bread is the bigger issue.”
Essentially, Ludwig says, we have the causal chain backwards. It’s not that people are eating too much, calorie for calorie, and thus gaining weight. Rather, the high-glycemic carbs we eat set off a hormonal pathway that causes our bodies to store too much energy, so we are hungrier and eat more.
The current leading mechanistic model — the energy balance theory — Ludwig said, is “not addressing the cause, it’s addressing a downstream effect, and so we’d be like somebody with fever, trying to treat the problem by getting into an ice bath.”
The Argument Against ‘Calories in, Calories Out’
Ludwig and co-authors say that the obesity treatment community is still operating within a framework that goes hand-in-hand with the energy balance model. An energy imbalance comes from the law of physics that state energy cannot be created or destroyed — so if you take in more than you expend, it has to go somewhere. Ludwig and others say the theory is still missing the “why.” Essentially, energy balance explains what happens when people gain weight, but why is it happening?
“The whole world thinks obesity is about energy balance,” said Robert Lustig, MD, a professor of pediatrics in the division of endocrinology at the University of California San Francisco, who outlined a similar earlier model of hyperinsulinemia leading to weight gain in 2006. “Therefore, it’s calories in, calories out. Therefore, it’s about two behaviors, gluttony and sloth. Therefore, if you’re fat, it’s your fault. Therefore, diet and exercise. Therefore, any calorie can be part of a balanced diet.”
And indeed, recent studies and reports suggest that stigmatization of people with obesity is alive and well, negatively impacting almost every aspect of life, especially healthcare. As a statement from Nature Medicine put it, “altering body weight is not as simple as eating less and moving more.”
That statement also emphasized that hormonal and metabolic adaptations out of someone’s control mean that diet and exercise are often only minimally effective. People can be considered “obese” while being free of metabolic issues just as others can, and do, have metabolic problems at a “normal” weight.
Ludwig’s perspective piece seeks to upend elements of this narrative by reframing a mechanistic model that he holds partially responsible.
But some say the models and the social stigma aren’t one and the same.
Criticisms: Argument Framing and Energy Balance
Critics say that the authors use the energy balance model in a misleading way for their argument’s purposes, conflating the principle with what amounts to ineffective dietary advice.
“The common recommendation, ‘eat less, be more physically active,’ that we tell people doesn’t work very well. Results are not that successful,” said Samuel Klein, MD, director of the Center for Human Nutrition at the Washington University School of Medicine in St. Louis. “But the ‘why’ doesn’t mean that energy balance is wrong. We’re unable to get people to actually reduce their energy intake adequately in order to do that.”
Klein and others point to the Endocrine Society’s 2017 scientific statement on obesity pathogenesis for a more full explanation of the theory.
The authors address this in their “criticisms” section, writing that none of “the recent dismissals of the CIM [carbohydrate-insulin model], provide a mechanistically oriented, testable model addressing dietary drivers of obesity, beyond the recurring focus on widely available, inexpensive, energy-dense, highly palatable, processed foods.”
Ludwig said what influences palatability isn’t universal or well defined yet, either, and that this explanation of the rise of obesity isn’t adequate.
Many say their perspective pushes the conversation forward. “I think one of the main arguments in the paper is, obesity is a huge problem, and we obviously have not been good at fixing it because of how prevalent it is,” said Jorge Moreno, MD, an internal medicine physician at Yale who specializes in obesity. He said while low-carb for weight loss isn’t new, linking low-glycemic diets directly to weight loss is compelling. “We can’t use the old model as our driver for change.”
The article refers to an earlier article in Science critical of the carbohydrate-insulin model by Kevin Hall, PhD, a section chief at the National Institute of Diabetes and Digestive and Kidney Diseases at the NIH, and John Speakman, PhD, of the Energetics Research Group at the University of Aberdeen. This article details the energy balance model and takes aim at some of the key points of the CIM.
Hall will also be an author of a response to Ludwig’s paper set to appear in the AJCN.
Ludwig and another author are associate editors of the journal, which is widely regarded as one of the most authoritative in the field of nutrition. The journal also turned around the perspective piece — though not a study or a review — relatively quickly. While Journal Guide estimates an average of 26 weeks from submission to online publication for the AJCN, this paper was published online within 12 weeks of its submission — though Ludwig says that’s typical for a perspective piece.
The Landscape Today
Others who have joined the conversation say that the research has been trying to quantify what actually drives energy intake and expenditure that leads to an energy imbalance, and that they’re not just pushing the outdated agenda of a bygone low-fat era.
“Obesity pathophysiology is very complicated, and it doesn’t fit in a sort of energy-in-energy-out perfect little equation because the complexities of both sides of that equation are so vast,” said Karl Nadolsky, DO, an endocrinologist specializing in diabetes, metabolism, and obesity. “They [Ludwig, et al.] use the real complexities to try to push this model that really ultimately supports their dietary bias for some reason,” but one diet or another doesn’t have to be mutually exclusive, he said.
Nadolsky said he and others recognize there’s more to obesity than just “calories in, calories out,” and he tailors recommendations for patients accordingly — but that an energy imbalance is certainly involved in weight gain. He said one adds fat when energy intake outmatches energy expenditure over time, which amounts to an imbalance. But how, exactly, a patient addresses that imbalance — and with which foods — is another problem.
The Carb Conundrum
That’s where the carbohydrates come into play. The authors conclude that based on their model, a low-carb diet would “decrease the insulin-to-glucagon ratio, enhance lipolysis and fat oxidation, and result in lower spontaneous food intake.”
They also disclose in their acknowledgements that five of the authors receive royalties from books on low-carb diets, and consulted for or started companies that promote low-carb diets. Among them, Ludwig has written a book that promotes a low-carb diet, although he said he has never accepted money from the food industry.
Regarding the conflicts of interest, Ludwig said, “There’s nothing unusual about this,” and pointed to drug studies where authors are closely tied to pharmaceutical companies, and food industry interests that fueled much of the “low-fat” wave of the 80s and 90s.
“Writing books, including for the public, is what academics have been encouraged to do for centuries, and selectively impugning a controversial idea … because people are writing books, I don’t think advances the dialogue,” he said.
The authors propose that a high-glycemic, high-carbohydrate diet leads to a spike in post-meal insulin, a higher insulin-to-glucagon ratio, which, along with a host of other factors (genetic, hormonal, microbiotic, fatty acid type consumed), leads to greater insulin sensitivity in body fat and increased fat storage.
From there, the energy sources that the body uses (glucose, ketones, fatty acids) are depleted, which causes the body to want to conserve energy (thus, expending less of it) and eat more (or increase intake), bringing on obesity and metabolic problems. The authors cite studies that support low-carb diets for long-term weight loss.
Research Diet Wars
Here, one encounters the classic pitfalls of nutrition research. To find support for the carb-insulin model, large, long-term, randomized controlled trials with humans eating either a high-carb diet or a low-carb diet would show that high-carb diets led them to eat more, expend less energy, and ultimately gain body fat. But long-term trials like these in a controlled setting (think: humans confined to a clinic and fed every meal) are hard to conduct, because they are ethically tricky, expensive, and resource-intensive.
Feasible alternatives to this gold-standard setup have serious drawbacks: long-term studies that ask people to document what they ate can be unreliable (humans are notoriously bad at documenting their own food intake accurately), short-term studies that closely control food intake can fail to grasp the long-term effects of a diet, and mouse studies don’t capture the complexity of human physiology. Adherence to diets, as in real life, is difficult, and consistently poor in nutrition studies.
“You can support anything that you want by looking at the literature, because it’s all over the place,” said Klein. “I think the bulk of literature clearly shows that this carbohydrate insulin model is not correct, and that it doesn’t exist.”
Supporters, on the other hand, say the literature supports their theory, and call for future research to test it.
Meta-analyses and reviews have come out on both ends of the carb question, concluding that low-carb doesn’t confer a significant advantage, but also that low-carb diets lead to slight weight loss over time.
Any difference in weight loss, Klein said, is probably due to eating fewer calories. When you cut carbs, you cut out many fatty, sugary items, too. There is limited evidence that very low-carb, high-fat (“good fats”) diets can help people feel more full.
The Mechanism Itself
Central to the debate is a disagreement about the role of insulin. In the paper, insulin is a key driver (although not the only one) of food intake, leading us to store more fat, which in turn lowers our energy expenditure, and makes us hungrier. In the energy balance theory, insulin plays a role, but not the lead.
Critics also say that there’s not enough evidence that low-carb diets cause the body to expend more energy, or that increased insulin is a direct cause of increased food intake.
“Stimulating fuel uptake by adipose tissue — there’s no evidence that goes back to stimulate your brain to eat more food,” said Klein. What’s more, he said the initial insulin spike doesn’t just happen with high-glycemic foods — it happens with small amounts of any carbohydrate.
Lustig, who pioneered the earlier paper on a model similar to the CIM in 2006, said it’s clear: insulin is what influences energy expenditure — but it’s driven by leptin.
He said he thought a discussion of leptin’s role, fructose and its conversion of fat in the liver, and how insulin release might affect three fat depots (subcutaneous, visceral, and liver) were missing. “Those are the things that would have made this paper much better. But the concept is right.”
Agreement and an Individualized Approach
Ludwig says he welcomes a debate, and hopes to push the conversation forward. “We explicitly leave open the possibility of common ground, but I think that it’s likely that the truth will be in the middle,” he said, “and we call on opponents to be civil with each other. That is to me, what science is about … to consider new ideas, even if those ideas are provocative.”
Still, it can be easy to miss what specialists in obesity, nutrition, and endocrinology agree on. For one, cutting back on sugar and highly processed foods, and recommending low-glycemic foods to patients.
“Replacing the high-saturated-fat foods and refined processed foods that have both types of fat with nuts and seeds and olive oil, improves, not only the weight, but also the cardiometabolic data,” Nadolsky said. “So these things all go together, and they don’t have to be mutually exclusive.”
He said helping individuals find the right approach for them is most important of all — and it might be low-carb, or it might be something else.
Although there’s disagreement in the framing of “easily available, highly palatable” foods as the main cause of obesity, most agree that easily available, highly processed carb-rich foods contribute in a big way. There are also more effective, non-diet interventions for obesity than ever before, like bariatric surgery, said Klein, and semaglutide (Wegovy), a weight management medication.
There are also many other factors at play when it comes to what people eat and how the body responds to it: genetics, social lives, hormonal differences, the gut microbiome, even a mother’s diet. Obesity affects Black and Hispanic adults disproportionately. There’s widespread agreement that there should be more research on these influences.
The truth is, there’s still much to be learned about the complexities of how food interacts with human metabolism.
“The regulation of food intake is really still very much an emerging science that we don’t know a lot about,” said Klein.
The academic discourse, some say, can sometimes miss the forest for the trees. “Why it creates this sort of political religious divide just blows my mind,” said Nadolsky. “We’re all on the same team, let’s just get it right and give the best guidance.”
Last Updated October 12, 2021