How a ‘fatally, tragically flawed’ paradigm has derailed the science of obesity

I’ve been a science reporter for 40 years. I’ve wanted to assume that the experts I interview can be trusted to understand their subjects. Put simply, to get it right. But watching researchers in the field of obesity almost blindly follow a failed paradigm has led me to cross a line that few journalists ever do, to publicly embrace and promote a minority opinion that many in the obesity field think is quackery.

For nearly a century, obesity research has been predicated on the belief that the cause of the disorder “is an energy imbalance between calories consumed and calories expended,” to quote the World Health Organization. By this ubiquitous thinking, obesity is an energy balance disorder: People get fat because they take in more calories than they expend. They stay lean when they don’t.

This is the central dogma of obesity science. Virtually all obesity research is interpreted in the context of this balance principle; all related public health discussions, not just on obesity but on all the common chronic diseases that associate with it, as well as the very nature of a healthy diet, rely fundamentally on its implications.

The concept that obesity is essentially an energy balance disorder, a problem of calories in and calories out, is “the number one concept that starts everything,” as Harvard researcher Bruce Spiegelman said when I interviewed him in 2019. “If you’re a serious scientist,” he added, “it’s just kind of obvious.”

I might have embraced this thinking as well if the prevalence of obesity had not risen relentlessly for the past half century; if obesity — along with type 2 diabetes, its partner in pathology — had not become the dominant non-Covid health crisis of our time. But I can’t.

The twin epidemics of obesity and diabetes have become a public health crisis, one the director of the World Health Organization has called a slow-motion disaster and the World Bank has called a ticking time bomb. Those assessments were made prior to the appearance of Covid-19, for which obesity and diabetes are second only to advanced age in elevating the likelihood of bad outcomes. The failure to make meaningful progress either treating or preventing obesity cannot be ignored.

This is why I am now a co-author, along with 16 influential academic researchers, of a lengthy review the American Journal of Clinical Nutrition is publishing on Sept. 13. The principal author is David Ludwig, a professor of pediatrics at Harvard Medical School and of nutrition at the Harvard T.H. Chan School of Public Health. We argue that the reason so little progress has been made against obesity and type 2 diabetes is because the field has been laboring, quite literally, in the sense intended by philosopher of science Thomas Kuhn, under the wrong paradigm.

This energy-in-energy-out conception of weight regulation, we argue, is fatally, tragically flawed: Obesity is not an energy balance disorder, but a hormonal or constitutional disorder, a dysregulation of fat storage and metabolism, a disorder of fuel-partitioning. Because these hormonal responses are dominated by the insulin signaling system, which in turn responds primarily (although not entirely) to the carbohydrate content of the diet, this thinking is now known as the carbohydrate-insulin model.

Its implications are simple and profound: People don’t get fat because they eat too much, consuming more calories than they expend, but because the carbohydrates in their diets — both the quantity of carbohydrates and their quality — establish a hormonal milieu that fosters the accumulation of excess fat.

Despite the unorthodoxy of this thinking, it is an easy case to make by studying the history of obesity research, as I’ve had to do as a journalist investigating the institutional failure to rein in the obesity/diabetes epidemic.

A century ago, the general thinking on obesity still allowed for two equally commonsensical ways to conceive of the pathology of the disorder. “The medical profession in general believes that there are two kinds of obese persons,” is how Louis Newburgh of the University of Michigan described this schism in 1930 in the first of two papers in which he claimed to have settled the issue, “those who have become fat because they overeat or under-exercise; and those composing a second group whose adiposity is not closely related to diet, but is caused by an endocrine or constitutional disorder.”

This endocrine (hormonal) or constitutional concept may have been best phrased not by a scientist or physician but by playwright George Bernard Shaw in his 1910 work “Misalliance.” “It’s constitutional,” says the character John Tarleton. “No matter how little you eat, you put on flesh if you’re made that way.”

By this thinking, promoted in the decades between the World Wars by prominent German and Austrian physician-researchers, some people are born predisposed to accumulate excess fat just as some are predestined to grow tall. However hungry or physically inactive they might be is an effect of the manifestation of this predisposition, not a cause.

That the environment plays a critical role is a given. The undeniable evidence is the enormous increase in the prevalence of obesity worldwide. In the U.S., 12% of Americans lived with obesity 60 years ago; more than 40% do today. Something has changed in the environment — in diets or lifestyles — to trigger such a dramatic rise in the prevalence of obesity. But is it nature or nurture that the environment triggers, behavior or physiology, minds or bodies?

In April 1929, at a time when the scientific discipline of obesity research constituted a handful of physicians in the U.S. and Europe musing on causality, Newburgh claimed to have established experimentally and definitively the primacy of overeating as its manifest cause. The articles he published promoting his research — comparisons of the rate of weight loss during calorie-restriction of maybe half a dozen obese subjects to one lean control — successfully shifted thinking in the field ever after. This makes Newburgh as close as obesity science gets to a transformational figure, the most influential researcher in the history of the science, although his contribution — wildly overblown, as it clearly was — has long been forgotten. When the Los Angeles Times reported on Newburgh’s work in 1932, the headline put it in appropriately blunt perspective: “Just Gluttony Causes Obesity. Michigan Professor Strips Defense of Portly.”

One clear effect of Newburgh’s work is that it defined the fundamental question that obesity researchers believed they had to answer: Why the gluttony, rather than why the excessive fat accumulation?

Beginning with the first animal models of obesity in the late 1930s (more on that shortly), researchers, with the very rarest of exceptions, have conceived of their work as elucidating the psychological, genetic, and physiological determinants of eating behavior. They’ve assumed, without justification, that what they observed translated directly and mechanistically to the excessive accumulation of fat in fat tissue. Ask a simple question, as I have as a journalist, like “Why is it that some of us fatten easily and some of us don’t — just as some breeds of livestock fatten easily and others don’t?” and obesity researchers can’t answer it because, curiously enough, that’s not what they study.

Had physicians in the 1930s thought to conceive of obesity as a fat accumulation disorder — an assumption-free definition of the problem — rather than as an energy-balance problem, they might have studied the physiological mechanisms that regulate fat storage. That science was worked out between the 1930s and 1960s, but by physiologists working with animal models who didn’t think of themselves as studying human obesity. Driven by a series of technological advances in laboratory assays, these researchers had identified in detail how hormones, enzymes, and the nervous system cooperate to move fat around the body, burn it for fuel, and store it as necessary. But this research evolved independent of the scientific thinking on obesity itself, as though the two phenomena — the storage of fat in adipose tissue and obesity itself, the storage of excess fat — had nothing in common.

In 1957, Columbia University’s Hilde Bruch, then the nation’s leading authority on childhood obesity, commented on precisely this problem. “Looking at obesity without preconceived ideas,” she wrote, “one would assume that the main trend of research should be directed toward an examination of abnormalities of fat metabolism, since by definition excessive accumulation of fat is the underlying abnormality. It so happens that this is the area in which the least work has been done.” In 1973, after this science had mostly been elucidated, Bruch was still remarking on its absence: “Amazing how little of this increasing awareness,” she wrote, “is reflected in the clinical literature on obesity.”

The same statement holds true today. Medical textbooks discuss the physiological mechanisms central to fat storage and metabolism — fat synthesis (lipogenesis), mobilization of fat from fat cells (lipolysis), fat storage, and burning fat for energy (oxidation) — implying that subtle disruptions in these processes could easily cause individuals to accumulate excess fat, but such explanations appear only in metabolism and endocrinology chapters. Discussions of obesity itself start and end with energy balance and, with no exceptions that I have found, omit the science of fat metabolism almost entirely. Textbooks dedicated to obesity will describe the causes of the disorder as “multifactorial” and “complex,” but the seemingly countless factors deemed relevant are all assumed to influence fat accumulation only indirectly through how much we eat and exercise.

Considering the slow-motion-disaster quality of the obesity/diabetes epidemic, it’s reasonable to ask why obesity researchers themselves so rarely if ever question their causal assumptions. (Why leave it up to a journalist to ask these questions instead?) The answer here is simple: Not only does it seem intuitively obvious that people who struggle with obesity eat too much — particularly so if you yourself are lean, as Newburgh was — but the energy balance thinking itself is perceived as an inescapable consequence of physical law. This is why articles and textbook chapters and institutional websites on obesity often begin with a discussion of the first law of thermodynamics and its perceived implications. The fact that energy is conserved, as a 2019 review in the journal Nature Metabolism explained, dictates that changes in the fat stored in the human body “must necessarily be a result [my italics] of changes in energy input (food intake) and/or energy expenditure.” To question this belief is to be accused — as I have been — of magical thinking, of believing that the laws of physics don’t hold for humans.

Despite the ubiquity of this thinking, though, it is clearly wrong, although “not even wrong” (as the physicist Wolfgang Pauli once described an ill-conceived idea) might be more appropriate. This is among the almost incomprehensible aspects of obesity science. The idea that obesity is caused by positive energy balance is not an implication of thermodynamics or any other law of physics.

Positive energy balance — more calories consumed than expended — is a description of what happens when people gain weight. As the energy stored in their bodies increases, so does their body mass. The increase in body mass is the positive energy balance. This reality says nothing about why it happens. After all, children are in positive energy balance when they’re growing — as Gustav von Bergmann, the leading German authority on internal medicine in the pre-World War II decades argued then — but that’s not why they’re growing. Women during pregnancy are in positive energy balance, as is the child in their womb, but that’s no explanation for why their body mass is increasing.

Consider using the identical logic to describe, say, why people get wealthy. Economists would (I hope) be embarrassed by a money-balance theory of wealth: People get rich because they take in more money than they spend. Clearly wealthy people did. We know that because they’re wealthy. The increase in wealth is the positive money balance. But this says nothing about how or why they accumulate such wealth. In obesity research, this tautological logic — saying the same thing in two different ways but offering no explanation for either — was allowed to become the central dogmatic truth.

Another of the remarkable aspects of this history is that the fundamentally tautological nature of energy-balance thinking has been so infrequently discussed. I could find only a handful of researchers who ever did so publicly. One was the Harvard nutritionist Jean Mayer, who wrote in 1954, “To attribute obesity to ‘overeating’ is as meaningful as to account for alcoholism by ascribing it to ‘overdrinking.'”

The only researcher I have found who has elaborated in this century on the not-even-wrongness of energy balance thinking is Thorkild Sørensen of the University of Copenhagen. Sørensen has published more than 400 articles on obesity, writing and directing research projects with many prominent researchers. He addressed the problem first in an article in 2008 and then later in a book chapter. But Sørensen says he has made little progress getting even his friends in the field to understand. “The first thing that happens,” he told me, “is that they think I am crazy … or stupid. They think I am denying the physics of thermodynamic law. I have to tell them that this is not what I’m doing. I tell them they’re making an assumption about the interpretation of thermodynamic law that they have no basis to do …. I sense sometimes that I’m kind of convincing them, then seconds later they’re back again to the typical thinking about positive energy balance is driven by input minus output.”

That the energy balance thinking offers no explanation for why people become obese is by no means its only serious failing. Follow almost any implication of energy balance to its logical conclusion, as physician-scientists thinking about obesity once did, and profoundly disturbing problems arise.

Among the most troubling is that it inescapably transforms a physiological disorder — the accumulation of excess body fat — into a behavioral disorder, a character flaw. This makes fat-shaming a seemingly unavoidable consequence.

Here it helps to quantify exactly what energy imbalance implies. To maintain a healthy weight, by this thinking, requires that people match their intake to their expenditure perfectly. Overshooting on average by just 10 calories a day — the calories in a single potato chip — translates into gaining a pound of fat yearly, 10 pounds of excess fat per decade. In just 30 years, that tiny imbalance will transform anyone from lean to obese. Those 10 calories a day of positive energy balance are less than half of 1% of the calories a typical American consumes daily. That’s the overeating that has to be explained.

That raises other seemingly inescapable questions: One is how does anyone stay lean when it requires this perfect energy balance to do so? Another is why wouldn’t anyone in the process of becoming obese consciously adjust so they don’t — consume one or even a dozen fewer potato chips than otherwise and stop the fat from accumulating? This is why Jules Hirsch of Rockefeller University, among the most celebrated of late-20th century obesity researchers, told me in a 2002 interview that after 40 years of research he still didn’t know why people get fat to begin with. It’s why Columbia University’s Rudy Leibel, equally as celebrated, told me that should I ever think I understand obesity “it will probably indicate you’ve lost your mind.”

It’s easy to imagine why already obese individuals might not be able to exert the concerted effort necessary to become lean. But why didn’t they prevent themselves from becoming obese initially? If obesity is caused by a positive energy balance, avoiding or preventing it should be effortless.

Because the energy-balance logic demands an answer, Newburgh offered up the implication in his articles and, by doing so, catalyzed the transformation of the scientific perception of obesity from a chronic, disabling physiological disorder into a character or psychological defect. Children get fat, Newburgh wrote, not because of a constitutional predisposition but because they’d “been deliberately trained to overeat by their parents”; and most adults get fat because they suffer from “various human weaknesses such as over-indulgence and ignorance.” If they didn’t, wouldn’t they have eaten less and not become obese?

By the 1950s, this logic had been institutionalized. Suggestions that obesity was a hormonal disorder were dismissed as “lame excuses” (the Mayo Clinic’s leading obesity expert in a 1957 book chapter) for the obese to avoid doing what comes naturally to lean people: eat in moderation. Authorities in the obesity field were now becoming psychologists and psychiatrists. People overate and got fat, these experts declared, because of “unresolved emotional conflicts” or to relieve “the nervous tensions of life.” It’s still common to see obesity blamed today on “comfort eating,” as one 2017 textbook describes it, “to reduce negative emotions, such as anger, loneliness, boredom and depression.”

To avoid the direct implications of the logic that people with obesity merely lack the willpower of those who are lean, some authorities instead blame genetics for making people overeat, determining, as one Oxford researcher told the New York Times in 2019, “who’s going to reach for the doughnuts?” Others blame the food industry for making the doughnuts (or other ultraprocessed foods, the terminology de jour) available to be overeaten, if not so addictive as to be impossible to refuse.

Since not everyone is obese or overweight, some people clearly do balance their intake to their expenditure even in an environment where food is everywhere. Shouldn’t a viable hypothesis of obesity be able to explain why some people remain lean and others don’t without implicitly or even explicitly blaming character?

This problem is solved by simply defining obesity as what it clearly is: a disorder of excessive fat accumulation. This perspective was argued in the 1920s and 1930s first by the German internist Gustav von Bergmann and then by University of Vienna endocrinologist Julius Bauer, the “noted Vienna authority on internal diseases,” as the New York Times then described him.

This conception focused on what Bauer called the “exaggerated tendency of some tissues to store fat.” Men and women fatten differently, he pointed out; and both in very specific areas. Sex hormones clearly influence fat accumulation independent of energy balance. Whatever mechanisms are at work locally, Bauer argued, should be the prime suspects systemically in causing obesity. “Like a malignant tumor or like the fetus, the uterus or the breasts of a pregnant woman,” Bauer wrote in 1929, the fat tissue in a person predisposed to obesity “seizes on foodstuffs, even in the case of undernutrition. It maintains its stock, and may increase it independent of the requirements of the organism. A sort of anarchy exists; the adipose tissue lives for itself and does not fit into the precisely regulated management of the whole organism.”

By the late 1930s, Bauer and von Bergmann’s fat-storage hypothesis was winning over influential researchers in the U.S. and Europe, but it didn’t survive the war years. The German and Austrian research establishments evaporated with World War II, ending the era of German/Austrian dominance in medical science; the lingua franca of medicine shifted from German to English, and the hypothesis lost its constituency.

This shift coincided with the development of the first animal models of obesity, allowing researchers, for the first time, to study obesity experimentally. Now obesity researchers in America, readers and citers of Newburgh and not of Bauer or von Bergmann, settled the debate between the two competing paradigms of obesity. But they did so incorrectly, interpreting their observations only in the energy balance context, seemingly unaware that another context or hypothesis or paradigm even existed.

The very first animal model of obesity set the precedent. Using a device created for brain surgery, the technique required placing a precisely targeted lesion in the middle of an animal’s hypothalamus near its underside (the ventromedial region). The hypothalamus sits directly above the pituitary gland at the base of the brain and is hardwired to organs throughout the body via the nervous system, including fat tissue. It’s also connected indirectly through the hormonal secretions of the endocrine system. Because animals with these lesions in the hypothalamus often ate voraciously and grew obese, John Brobeck, then a physiologist at Yale, proposed in 1946 that the hypothalamus must be controlling eating behavior. This is how it’s been perceived ever since.

But animals with lesions in the ventromedial region of the hypothalamus would also grow obese, or at least get significantly fatter, when they didn’t eat any more than a lean animal. Prevent the animal from eating excessively — meaning control for the overeating, in the language of experimental science — and these animals get fat anyway. Borrowing from Shaw’s Tarleton, they put on flesh anyway, even at levels of food intake that would not cause weight gain in normal animals. This observation would go unexplained or be ignored entirely.

Once Brobeck assumed that overeating (he called it hyperphagia, a term that is still in use) was the reason why these animals with ventromedial lesions in the hypothalamus got fat, obesity researchers in the post-World War II years perceived their scientific obligation as elucidating how the hypothalamus knows enough to moderate eating and maybe energy expenditure as well, and how that awareness breaks down in obesity. The hypotheses that dominated thinking from the 1950s onward have been attempts to answer this question, proposing, for instance, that the signal to the hypothalamus was blood sugar (Jean Mayer) or circulating fatty acids (Gordon C. Kennedy of the University of Cambridge in the U.K.).

While researchers have since created many animal models of obesity — genetically, surgically, or manipulated by diet — one observation is remarkably consistent. Although researchers have rarely thought to control for energy intake in their experiments, when they did, testing whether their animals get fatter than lean controls even when eating as little or less food, they almost invariably report that they do. “These mice will make fat out of their food under the most unlikely circumstances, even when half-starved [my italics],” as Jean Mayer wrote about obese mice he studied in the 1950s. This fundamental observation directly challenges the notion that obesity is caused by poorly regulated eating behavior. It supports the hypothesis that poorly regulated eating behavior is a consequence of an animal’s constitutional drive to accumulate excess fat. These observations, too, have been ignored.

The most conspicuous example of the disconnect between the energy-balance model and the actual evidence from animal models comes from the most famous mutant mice — named the ob/ob mice — ever discovered. The first appeared in a litter at the Jackson Laboratory in Maine in 1949. It was ob/ob mutants that led Jeffrey Friedman and his Rockefeller University colleagues to discover the hormone leptin and identify defects in the leptin gene as responsible for the animal’s obesity.

Leptin’s discovery in 1994 was hailed as the holy grail of obesity research: the “putative signal” secreted by fat tissue to inform the hypothalamus that fat is accumulating, and so eating and expenditure must be adjusted to compensate. Tens of thousands of articles have since been published relevant to leptin’s role in obesity.

Despite this body of literature, much of it from Friedman’s own lab, documenting leptin’s direct effect on fat metabolism and storage (Bauer and von Bergmann’s fat-accumulation hypothesis), the research community has assumed, as Friedman first did, that leptin is a satiety hormone (Newburgh’s energy-balance thinking). According to the Endocrine Society website, leptin “helps inhibit hunger and regulate energy balance so the body does not trigger hunger responses when it does not need energy.” Hence, when leptin is absent or the hypothalamus is resistant to leptin’s signaling, the result is positive energy balance — obesity.

But even the ob/ob mutants make fat out of their food when, quite literally, half-starved. Friedman’s research was based on Douglas Coleman’s research at the Jackson Laboratory and Coleman, who shared a Lasker Foundation Award in 2010 with Friedman, dedicated two papers to this observation: “Even when maintained on 50 percent of normal food intake,” Coleman reported in 1985, “mutants still become obese.”

If an animal is amassing fat even when deprived of any opportunity to eat a decent meal, even when semi-starved, surely leptin is doing something profound to fat storage other than regulating appetite or even appetite and expenditure. When I asked Friedman about this just before Covid-19 hit in the winter of 2020, he said he had no explanation and hadn’t thought about it in a while. “Clearly some metabolic alterations push nutrients into adipose tissue, in some ways to the detriment of other tissues,” he said. But if that can happen independent of how much an animal eats, why wouldn’t we suspect the same thing happens in humans?

My reading of the history of obesity science is that none of this would have happened had physicians thinking about what causes obesity paid any meaningful attention, as Bruch suggested in 1957, to the evolving research on fat metabolism itself.

A series of technological innovations that began in the mid-1930s allowed researchers to understand first how metabolically active fat was —”mobilization and deposition of fat go on continuously, without regard to the nutritional state of the animal [again, my italics],” as described in a seminal 1948 review — and then how hormones and the central nervous system very carefully regulated these processes.

By the mid-1960s, researchers studying fat storage and metabolism had established unambiguously that the hormone insulin dominated the regulation of fat storage. While insulin works conspicuously to control blood sugar — defects in insulin production and responsiveness are primary causes of diabetes — it does so partly by stimulating the uptake of fat into fat cells, inhibiting its release and inhibiting its use as energy in non-adipose tissue. No tissue in the body is as sensitive to insulin’s action as fat tissue — “exquisitely sensitive,” as these researchers often described it.

With the invention of a technique to measure circulating levels of insulin — the radioimmunoassay, developed by Solomon Berson and Rosalyn Yalow in 1960 (for which Yalow, after Berson’s death, won a Nobel Prize in 1977) — researchers learned that animals with hypothalamic lesions and ob/ob mice are both hyperinsulinemic, as are people with obesity and with type 2 diabetes, offering a parsimonious explanation for the mechanism of their excessive fat accumulation.

Some of the most influential researchers studying obesity and diabetes — including Berson and Yalow themselves — proposed primary roles for insulin in fat accumulation and obesity. Influential researchers in a discipline known as physiological psychology (with roots going back to Ivan Pavlov) — Jacques Le Magnen of the Collège de France in Paris, specifically, and Mark Friedman of the Monell Center in Philadelphia (a coauthor on the American Journal of Clinical Nutrition review) — did the same. But all these ideas have failed to take hold, as obesity researchers continued to insist that energy balance, or lack thereof, was the mechanistic explanation and an indisputable truth.

This is the danger with the kind of dogmatic status that the energy balance notion achieved so early and prematurely in obesity research. Scientists and philosophers of science have commented on this problem for centuries. “Men who have excessive faith in their theories or ideas are not only ill prepared for making discoveries; they also make very poor observations,” Claude Bernard wrote in “The Introduction to the Study of Experimental Medicine” in 1865. “Of necessity, they observe with a preconceived idea, and when they devise an experiment, they can see, in its results, only a confirmation of their theory. In this way they distort observation and often neglect very important facts because they do not further their aim.”

In the case of obesity, the hormonal/constitutional hypothesis also encountered dogmatic resistance in response to its single most direct practical implication: Diets that can successfully resolve obesity are not those that induce us to eat less, per energy-balance thinking, but those that reduce circulating levels of insulin, accomplished most effectively by replacing dietary carbohydrates — sugars, starchy vegetables and grains, and the like — with fat.

Physicians and diet book authors have been promoting carbohydrate-restricted, high-fat diets — ketogenic diets, now commonly known as keto — for going on 200 years, most famously Robert Atkins, a New York cardiologist. By arguing, as Atkins and others did, that fat could be lost without limiting calories by fixing the hormonal dysregulation of fat storage — restricting what one eats, not how much — these books were treated as de facto quackery. By advocating that we eat fat-rich foods, they were considered dangerous.

Over the course of my research, I came to find the arguments of these diet books compelling, so much so that I’ve written four books myself on the subject and co-founded a not-for-profit organization in 2012, the Nutrition Science Initiative, to help fund research that might shed meaningful light on the causes of common obesity. (The results were encouraging but ambiguous.)

While researchers continue to study fat metabolism and its role in obesity, they typically assume that whatever they’re observing plays a subordinate role to that of energy imbalance. This, again, is a danger of dogmatic thinking. In Stockholm, for instance, Karolinska Institute researchers have reported that fat is stored longer in the fat cells of people who are obese than it is in those who are lean. Mobilization of fat from fat cells in these people is inhibited, precisely as would be predicted by Bauer and von Bergmann’s thinking that the fat tissue in those predisposed to obesity is dysregulated. In 2019, the Karolinska researchers reported that the older we get, the more fat “turnover” in fat cells slows, a phenomenon that they estimate could account for a 200-pound person adding 30 pounds of excess fat in a decade without eating any more than he or she ever did. Still, they could not or would not shake the energy-balance context, suggesting that because overeating was the “main driver for obesity,” it might be the “failure to reduce calorie intake in aging” to compensate for this deaccelerating fat turnover that leads to weight gain.

The researchers most willing to question the energy balance logic are those who still practice as physicians and regularly treat patients with obesity. These physicians, an ever-growing but still small minority, find that when they induce their patients to restrict carbohydrates but not calories, their patients can achieve and maintain a healthy weight with relative ease and get healthier in the process. When this approach has been used for people with type 2 diabetes, as the San Francisco-based start-up Virta Health has been doing, the results have been unprecedented.

For researchers and public health authorities to make real progress against obesity and the obesity‑diabetes epidemics, they will have to shed their energy-balance thinking, their obsessive focus on how much people eat and exercise. Instead, borrowing again from Hilde Bruch in 1957, they’ll have to focus on fat metabolism and storage itself, “since by definition excessive accumulation of fat is the underlying abnormality.” If they think of obesity as it simply and clearly is, a disorder of excess fat accumulation, they might actually figure it out.

Gary Taubes is a science and health journalist, author, and co-founder of the Nutrition Science Initiative.