Two Models of Why We Get Fat


This is part 1 of a two part series. Part 2 is here.

Why do we get fat? The prevailing model is energy balance, pictured above on the left. When people eat more than they burn off, they get fat. All calories are the same. Just do the math. In contrast, on the right is a picture of various kinds of food. In this model, people get fat because they eat the wrong kinds of food, like sugar. All calories are not the same because the body responds differently to different kinds of food. Now to be up front, I’m a big fan of science writer Gary Taubes and his book Why We Get Fat, where he argues high carb intake is to blame for obesity. But low carbohydrate diets are controversial, so instead of tackling this head on let’s step back and see why the energy balance model on the left is flawed.

The energy balance model suggests two reasons for why we get fat:

  1. Too little exercise. While exercise makes you stronger and healthier, it surprisingly doesn’t make you lose fat tissue. That’s a myth. More here.
  2. Too much food. The Hazda, a hunter gather society, don’t consume any more calories than people with desk jobs. This makes it unlikely that we are eating any more food than people into deep prehistory. And if the too much food theory were true we’d expect obesity to appear in the historical record as soon as the wealthy have plenty to eat. Yet the obesity epidemic only took off in the 1970’s, long after food was plentiful for even the middle class.

Let me quote Taubes on some numbers here:

Now, if you gain 40 pounds of fat over 20 years, that’s an average of two pounds of excess fat accumulation every year. Since a pound of fat is roughly equal to 3500 calories, this means you accumulate roughly 7000 calories worth of fat every year. Divide that 7000 by 365 and you get the number of calories of fat you stored each day and never burned – roughly 19 calories. Let’s round up to 20 calories, so we have a nice round number. So now the question: if all you have to do to become obese is store 20 extra calories each day on average in your fat tissue — 20 calories that you don’t mobilize and burn — what does overeating have to do with it? And why aren’t we all fat?

To put this in perspective, if you exercise regularly you burn an extra 300 calories each day. And then what happens? Your body automatically adjusts your food intake to match how much you burned off. It’s called “working up an appetite”. And if you stop exercising it also adjusts your food intake back down. These adjustments happen all the time and swamp the 20 calories a day needed for long term weight gain.

The energy balance model asks us to believe something as essential as food intake regulation breaks down in an ideal living environment with plenty of food and rest. If taken at face value we should expect animals, which share much of our regulatory framework, to have this same problem. So for example if zebras exercise too little or too much, and don’t constantly track their diets on their iPads, we’d expect they get fat or starve. Of course evolution doesn’t build such fragile solutions. The energy balance model ignores the central problem, which is to explain why an exquisitely tuned regulatory system for adjusting food intake breaks down in the first place.

So what about the other model? Well, in that model eating food your body is maladapted for leads to hormonal dysfunction, signally the body to create more fat cells than needed. In turn, this triggers the regulatory system to eat more. The correct analogy is how taking steroids builds excess muscle tissue, making you eat more to compensate. Think of junk food as fat steroids and you get the picture. We eat more because the body (incorrectly) decided to build too many fat cells, not the other way around. The energy balance model has it backwards.

Experimentally, with the energy balance model nutritionists should run studies where people eat less and exercise. And we should expect this to work. It generally fails. Constant hunger is a sign of an organism chronically out of balance. In contrast, with the kinds of food model we should run studies where people try eliminating various types of high risk foods, and then see which ones get the regulatory system working properly again. So maybe a study where we cut back on carbs, or other potentially bad types of food. As a class we might think of these as “subtractive diets”, since they subtract the bad food out of the diet, but then with that caveat you can eat as much as you want. As a bonus with this approach, burning willpower trying to exercise more and eat less isn’t required. You want your natural regulatory system to handle your food intake, as it should. The only willpower you need is to avoid certain classes of food.

Now of course Taubes points to insulin regulation as the failure point. Excessive carbs trigger insulin regulatory problems, in turn causing diabetes and getting fat. But there are other versions of the ”subtractive diet” model besides cutting carbs, such as this recent New York Times article on The Extraordinary Science of Addictive Junk Food. Another variation is not low carb diets per se, but low GI diets. The takeaway for me is the energy balance model has nothing scientifically useful to say, despite being current dogma. It allows well meaning but misguided nutritionists to cite trivial applications of second law of thermodynamics (energy in = energy out), and blames patients for lack of willpower and being fat. With that background, next week we’ll dig into the pros and cons of Taubes’ low carb thesis and his solid but occasionally over the top book.

This is part 1 of a two part series. Part 2 is here.

Categorized as Science

By Nathan Taylor

I blog at on tech trends and the near future. I'm on twitter as @ntaylor963.