Of Course Overeating Makes Us Fat... Except When It Doesn't?

What a pair of experiments--one a thought experiment, one a century old--tell us about the illogic of overeating.

“There just never seemed to be enough food to satiate [him]. After school, he would eat a footlong sub before his mother’s home-cooked dinners, even after having a hefty lunch of homemade chicken, rice and vegetables and his favorite snacks, granola bars and Buncha Crunch.”

Question of the day: Is the young man described in this paragraph from a 2015 New York Times article getting fatter?

This kind of eating behavior is typically considered a cause of obesity, so is this young man on the way? Is he already obese?

These questions, of course, are impossible to answer, not with the information given.

Conventional wisdom says we get fat because we eat more calories than we expend—we eat too much—but the quote only tells us about intake, not expenditure.¹ We know he’s eating a lot, but we don’t know if he’s overeating.

To know if this young man is getting fatter, we have to know far more about him. He might spend hours a day in physical activity. Training for marathons? Maybe that’s why he’s always hungry. Maybe he’s lifting weights and building muscle. Or maybe he’s just growing. An adolescent with a growth spurt? That might explain the hunger. Some combination of all of the above?²

So what’s the point? For the past 80 years, the conventional wisdom on the cause of obesity is that we get fat because we eat too much—we overeat or over consume or even over nourish. Another way to make this point is that we get fat because we’re in positive energy balance, using the jargon that comes with the common thinking on obesity that it is a disorder of energy balance.

This is why authoritative articles on the cause of obesity often came with simplistic diagrams like this one, from a 2001 article by two of the leading researchers in the field (and cited over 3000 times):

Spiegelman, BS, and Flier, JS. 2001. Obesity and the Regulation of Energy Balance. Cell. Feb 23;104:531-43.

Clearly to know whether or not we’re in energy balance, whether or not we’re eating too much, requires knowing both the intake and expenditure side of the equation.

But that’s still not enough information to solve the question I asked. If we now define overeating as eating so much that it has harmful consequences, it requires knowing whether those consequences have actually panned out. For instance, someone can be in positive energy balance, taking in more energy than they expend, if they’re building muscle, but we wouldn’t say they’re overeating unless they were putting on excessive fat with the muscle.

Here’s where energy balance thinking gets circular and I would say fatally so: to know if we’re overeating, we must know if we’re getting too fat. But if our explanation for why we get fat is that we overate, then overeating becomes nothing more than a synonym for fat accumulation. The conventional wisdom is tautological: It claims overeating causes obesity when it's really just describing the same phenomenon—getting excessively fat—in two different words.

In my books, particularly in the epilogue of Good Calories, Bad Calories, I criticized nutrition and obesity research as functioning “like a religion” rather than a science, based in part on this almost religious belief that overeating causes obesity. Yet by the 1930s, European authorities—particularly in Germany and Austria—had largely rejected this energy balance framework, replacing it with the concept of obesity as a neuroendocrine/constitutional disorder of fat storage and metabolism.

What happened in obesity research in this era was a common progression in science. The physicians of this era who were thinking about the cause of obesity had inherited a way of conceptualizing it, as an energy balance problem, that was dependent on the limited tools available to study obesity in the late 19th Century. As the understanding of medical science in general evolved, and of endocrinology in particular, the more critical of these physician scientists came to realize that this energy balance framework was actually an impediment to understanding the nature of the disorder.

They had to reject it to make progress, and they did; just as chemists in the late 18th century had to overturn the notion of phlogiston to make progress, and epidemiologists studying infectious diseases had to rid their discipline of the miasma thinking in the late 19th Century, and physicists had to reject the idea of aether.

The catch, as I’ve written, is that this German/Austrian school of medical science effectively vanished with the Second World War³ and the rejection of the energy balance framework vanished with it. In the post-war years, young American physicians, new to the science of obesity, rebuilt it on the basis of energy balance and the community of obesity researchers never looked back. They certainly didn’t bother reading the pre-World War 2 German language literature to find out what they might have missed.

As a result, even people who live with obesity, who have struggled and failed to maintain a healthy weight for as long as they can remember, who know they don’t eat any more than their relatively lean friends and relatives do, would come to believe that they’re burdened by excessive amounts of fat because they somehow eat too much. What could be more obvious?

One reason why this conceptual problem has never been resolved is because the authorities who have grown up believing in the conventional wisdom of energy balance have always had trouble taking seriously any idea that assumed otherwise. This is group think at work, but it’s compounded by how difficult it is to do experiments that can differentiate between these two hypotheses/paradigms. Absent some kind of definitive experimental evidence, the obesity authorities could always insist that what they had believed all along somehow had precedence.

In 1920, for instance, Carl von Noorden, the German internist and diabetes specialist who was more or less the father of this energy balance thinking, wrote of the competing hypothesis that obesity is a fat storage disorder: “Undoubtedly, [the] theory offers many new insights. Still, it is not yet firmly established enough to shake the previously accepted doctrine that emphasizes the behavior of energy balance… For now, it's advisable to adhere to the old doctrine.”

And physicians and obesity researchers continue to adhere to that old doctrine today. By now they’ve long forgotten where it even came from. It’s dogma and to question it is to establish that you’re not a serious enough thinker to be taken seriously. (I can live with that.)

Are thought experiments enough to dismiss an absurd hypothesis?

The first half of my 2011 book, Why We Get Fat, discusses many of the conceptual problems with this energy balance thinking, but today I wanted to use a thought experiment to demonstrate the kind of absurdity that can arise from this thinking.

Worth noting is that thought experiments have often been at the heart of scientific progress—Einstein’s use of thought experiments, for instance, led him to derive the laws of relativity. Thought experiments allow us to conceptualize the implications of a hypothesis to see if it makes sense and, if it doesn’t, whether it can or should be salvaged by modifications. They’re another way of testing hypotheses in a world in which hypothesis and test is the essence of progress.

Thought experiments have the advantage over the physical kind in that they’re free and anyone can do them; they require access to no laboratory or physical equipment. Anyone can do them. They’re not bounded by reality or practicality (and they don’t require permission from an internal review board). All that they require is the thinking, although that implies the kind of common sense that may not be, as Voltaire once suggested, all that common.

So now let’s get to that energy balance thinking. It’s derived from the first law of thermodynamics, which is the law of energy conservation. In short, energy is conserved and so the change of energy in a system (ΔE) is equal to the energy that goes in minus the energy that goes out:

ΔE = E_in - E_out

This is pretty simple. It’s like saying that the change in the amount of money in your bank account (Δ$) is equal to the money that you deposit and the interest earned ($_in) minus the money you withdraw ($_out). Energy conservation is a law of physics and probably the easiest of all of the laws of physics to understand.

The problem is that that the law of energy conservation doesn’t actually say anything about the cause of obesity, as I’ll explain. Beginning in the late 1990s and working under the assumption that obesity is a penalty of gluttony and sloth, the physicians thinking back then about the cause of the disorder—I’ll now call them obesitologists, for lack of a better word⁴—decided that this thermodynamic law somehow scientifically confirmed their thinking about gluttony and sloth.

They did so by changing the equal sign in the energy conservation equation to an arrow of causality. Instead of thinking to themselves that E_in - E_out equals ΔE, which means only that the two sides of the equation are equivalent, they think that E_in - E_out causes ΔE.

From one perspective, this can seem legitimate. After all, if we manipulate one side of the equation—increase E_in, say, which means food consumption goes up, while keeping E_out the same—then we force the other side of the equation to change. The two are, after all, equivalent, forever locked in step by the equal sign. So if we make food consumption (E_in) go up, then we make ΔE go up. Voilà, as the French like to say, we have caused overeating, which has seemingly caused the system to accumulate energy. Now ΔE has gone up—that’s the de facto definition of positive energy balance—and since the obesitologists thought of E as the energy stored in fat tissue, that means we’re getting fatter.

While all of this is true, it regrettably says nothing about the cause of obesity because ΔE itself, after all, can go up for any number of other reasons, a number limited only by our imagination. If it does—say we inject a drug that causes fat cells to multiply—then E_in and E_out will have to change as well, because the two sides of the equation are equal. One side goes up, the other side goes up. So maybe getting fatter—ΔE increases because of our drug in this scenario—causes this overeating thing. It makes make people hungrier (E_in goes up) or more sedentary (E_out goes down). Maybe it even slows metabolism (E_out goes down), or some combination of all of these possibilities.

Now here’s the challenge of the thought experiment: let’s find an examples when ΔE increases in humans for a reason that is clearly not overeating. In that case, what happens to appetite?

Back to our insatiable adolescent… now a star in the NBA.

This is where the opening quote comes in. As it turns out the young man that paragraph was referring to is a basketball player. He was becoming famous then, hence a New York Times profile, and certainly is now: Karl-Anthony Towns, aka KAT, formerly of the Minnesota Timberwolves and currently (although perhaps not for long) the New York Knicks.

When KAT was a college freshman at Kentucky in 2015 and destined for stardom, he was already 7’ tall and 250 pounds (with size 20 sneakers). Here’s what he looked like:

And it was KAT’s insatiable appetite that the Times profile was discussing in the paragraph I quoted: the hefty lunches, foot-long subs, mother’s (assuredly hefty) home-cooked dinners, all followed by his “favorite snacks, granola bars and Buncha Crunch.”

A technical term for this kind of voracious hunger is hyperphagia. If you were familiar with the obesity literature and only read that opening quote; if you didn’t know you were reading about a kid who was growing to 7’ and probably worked out 3 to 4 hours every day, you’d certainly suspect that this was hyperphagia.

Now comes the thought experiment: imagine KAT had a fraternal twin. Let’s call him Anthony-Karl Towns, or AKT for short. Unlike KAT, AKT only grew to be 6’ tall, but his appetite was equally insatiable during his teens. ”There just never seemed to be enough food to satiate” AKT either. AKT also grew to be 250 pounds.

But because AKT was a foot shorter than his famous brother, he was obese. AKT had a BMI of 34, well into the obese range. KAT had a BMI of 25.

Here’s what AKT might have looked like had he also played college basketball (with thanks to ChatGPT’s image generator for the revision):

And here’s the point: We know in Karl-Anthony’s case that his appetite was insatiable because he was growing. His growth (ΔE increasing so dramatically) was the cause (the reason why E_in was so large). The fact that KAT probably worked out 3+ hours every day probably didn’t hurt.

In this case, transforming the conservation of energy law into a causal hypothesis would look like this:

ΔE → E_in - E_out

But notice this is not purporting to be a causal hypothesis for KAT’s growth, not a reason why he grows. We know that already: KAT was growing because his body was surging with testosterone, growth hormone and insulin-like growth factor. His growth was the result of these endocrine phenomena at work. In a word, hormones.

The transformation of the energy law in this case gives gives us a causal hypothesis for his insatiable appetite. KAT was eating a lot and never satisfied because he was growing. (And, again, his workouts and whatever they added to E_out would have increased his appetite further).

But now the situation with Anthony-Karl, his brother, is the exact opposite, or at least it is if we buy into the energy balance thinking. Because AKT was both insatiable and burdened by excessive fat (unlike KAT), the obesity authorities blame AKT’s progression to obesity, his positive ΔE, on his appetite. Now they assume he gets fat because he’s insatiable and eats so much. Now the energy conservation law is assumed to be a hypothesis of obesity and the arrow flips to the left.

ΔE ← E_in - E_out

So if we know that a particular individual is obese—AKT, in this case—then the law of energy conservation tells us why. He overate. If we know that a particular is lean—i.e., KAT—it tells us only about his appetite and expenditure.

Even more bizarre or inane (or pick your dismissive adjective), we can quantify ΔE with reasonable accurately for both KAT and AKT, and its essentially identical. That’s easy.

Over the course of their lives, ΔE is their body weight (minus water weight, times whatever unit of conversion we want to use to convert body mass into energy). It’s 250 pounds times our conversion factor for each. That’s their positive energy balance since the day they were conceived, and it’s (essentially) identical. If we establish how much of their body weight is fat mass and how much lean, we can calculate precisely how much energy they’ve accumulated over the course of their lives. AKT’s will be greater because AKT has more fat, with its dense calories, than KAT does. [With thanks to David Ludwig for the post-publication correction.]

But the obesity authorities will tell us this positive energy balance in KAT’s case was a cause and in AKT’s it was an effect. KAT’s ΔE was driven upward by growth and sex hormones; AKT’s ΔE, despite being identical, was driven upward by overeating.

And here’s the key point: one of these is a fact of medical science and the other is an assumption. We know that KAT grew to 7’ and 250 pounds and that his appetite was insatiable because of hormonal phenomena that drove the growth. No authorities would disagree with that. All that theorizing about AKT and his overeating-caused-obesity is the assumption, based only on preconceptions about the penalties of gluttony and sloth, and a logically unjustifiable manipulation of the law of energy conservation.

Wouldn’t the simplest possible hypothesis be that the same causal pathway that’s at work in KAT be at work in his (imaginary) twin brother also? Both are endocrine phenomena, one of vertical growth, the other of, well, horizontal?

Isn’t that the most likely explanation?

English bulldogs and a historical precedent for the same thinking

As I mentioned in my last post, I’m working on a book on the history of obesity research. The research has me going through the literature, year by year, article by article, since it begins as a serious endeavor in the late 19th Century. I’m up to the 1930s, when the thinking on obesity should have shifted, when von Noorden’s advice to “adhere to the old doctrine” had been mostly rejected.

Why mostly? While the European obesitologists came to agree that most people with obesity suffered from some kind of neural-endocrine, fat-storage disorder, it was still possible that exceptions existed. Clearly some people with obesity ate enormous amounts of food and drank enormous amounts of alcohol. Shakespeare’s Falstaff was the iconic example.

In those cases, wasn’t it reasonable to assume that these people were obese because of their insatiable appetites? Wasn’t it possible, if not likely, that these people had overeating obesity, as some obesitologists were calling it, even if they represented only a small minority of all people with obesity. Those making this claim, though, tended to be those who had spent their careers thinking about energy metabolism—about E_in and E_out. They were not paying attention to the evolving science of endocrinology and certainly not the evolving research on fat storage itself.

In the 1920s, both of these were brand new sciences. Hormones like insulin, glucagon and growth hormone had only been discovered early in the decade, and even a rudimentary understanding of what these hormones did required that researchers be able to purify them in the quantities necessary to use in experiments. That development alone ate up years of work.

As the science of endocrinology evolved through the 1920s, though, it provided more and more insights into the relationship between hormones and fat storage and between fundamental behaviors—eating, drinking, and physical activity, specifically—and underlying physiological states.

By 1929, the endocrinology research had offered up an observation that did the job then of what my KAT-AKT thought experiment does now. Three physiologists from Harvard Medical School had experimented with the effect of growth hormone extracts on growth itself. They took two English bulldog puppies from the same litter, studied them for three weeks—they “were found to grow and behave as normal puppies should”—and then gave the smaller of the two dogs daily injections of the pituitary extracts that we know today to be growth hormone.

Within weeks, the dog getting the extracts was clearly growing faster. Here’s a photo of the two dogs at three months. The dog on the right was getting the growth hormone.

By the fifth month, as the Harvard researchers wrote, rapid growth was continuing in the injected dog and, crucially, “The appetite became noticeably greater.”

After 18 months, the dog receiving the growth hormone injections weighed twice as much as the control dog. When the dogs were later autopsied, it was clear that the dog getting the injections had all the symptoms of the condition known as acromegaly (less technically, gigantism, when it occurs in children). Physicians had long assumed that acromegaly was produced by tumors in the pituitary gland, increasing the secretion of whatever hormone that gland (and particularly the anterior lobe) secreted. The 1929 research was seen as confirming that mechanism experimentally, even as only one animal was needed (thankfully) to demonstrate it.

Julius Bauer, a University of Vienna endocrinologist who was already well known for his work linking endocrinology and genetics (also a new science) to constitutional diseases, then used this Harvard observation to make the necessary point about obesity. Writing with a young American endocrinologist, Solomon Silver of Mt. Sinai Hospital in New York, and publishing in an English language journal, Bauer pointed out an obvious implication: an abnormally increased appetite alone “does not necessarily result in obesity.”

In their studies in the experimental production of acromegaly in dogs by the injection of [growth hormone] extract, they noted rapid growth and enormous increase in appetite…. One would not say that these dogs grew because they ate more. They ate more as an inescapable result of a newly instituted tendency to growth which is a specific action of the anterior pituitary lobe. The dogs did not grow because they ate more, but they ate more because they were growing. The increase in appetite is a result, not a cause.

If the insatiable appetite could be a result of growth, as the Harvard research confirmed, then what was necessary to cause the excessive growth of fat tissue? As Bauer and Solomon proposed, that would be a predisposition that manifested itself through the endocrine system and the nervous system.

If we locate the genesis of obesity in a constitutional destiny of the tissues of certain people to store fat then we can understand a disturbed balance between energy intake and expenditure as a necessary sequel of this almost irresistible tendency. We can also understand those cases of obesity, common in the experience of everyone, where in spite of rigorous diet and exercise, we are unable to reduce the fat from the areas where it is most obvious.

And that could be true for all those whose bodies are storing excessive fat, whether their appetites rival Falstaff’s (as KAT’s assuredly did, minus, I’m assuming, the alcohol), or whether they spend their lives fighting the tendency to store fat and resisting the urges to eat more than the bare minimum.

1. Expenditure, as I’m using the term, includes excretions.

2. If he had uncontrolled type 1 diabetes that could also explain the insatiable hunger, but in that case he might be emaciated. In this case, insulin deficiency would be the proximate cause of both the insatiable hunger and the emaciation.

3. It didn’t help that all the major proponents of this hypothesis, with one notable exception, were Jewish and lost their positions in Germany in 1933 or in Austria by 1938.

4. I’m introducing this word mid-way through a Substack post because I’m at a loss how to conveniently describe the folks studying obesity in the pre-War era. Some were physicians who had many obese patients—occasionally because they were at famous European spas as the physician-in-residence—some had scientific backgrounds and did actual experiments; some just read the literature and then tried to summarize it in journal articles or text book chapters. Obesitologists seems to sum them all up in a way that’s short and useable.