The Truth About Food

— Nothing's wrong with you. You're just at war with a system that's 2 million years older than you and doesn't play by the rules you read about.

— The arithmetic works. What doesn't work is the assumption that expenditure is a constant. You're subtracting from a fixed number, and the number is alive: it watches what you're doing and moves toward you. Imagine you cut a department's budget by 20% and expect spending to drop by exactly 20%. But the department is alive: it turns off the lights in the corridors, stops going to conferences, quietly cuts things that never show up in your report at all — and every month it comes to you asking for money a little more insistently.

— And you never made a single decision about it. The most detailed measurement of this comes from the contestants on the American show The Biggest Loser. 16 people went from an average of 148.9 kg to 90.6 over 30 weeks. Six years later — 131.6.

— Partly. But look at the second line. Their resting expenditure was 2607 kcal a day before the show, 1996 at the end, and 1903 six years later. Over those 6 years they gained 41 kg back — and their metabolism didn't come back up, it fell further.

— Less than when they were thin and already slowed down. Of that drop, the adaptive part — the decrease beyond what body composition explains — was 275 kcal a day at the end of the show and 499 six years later. And here I have to tell you what that number costs, or it'll lie to you. That's 14 people they managed to reach 6 years later, from the most extreme case we know of. It's not the norm for someone losing weight. For an ordinary person it's tens and low hundreds of kilocalories, not 500 and not 800.

— That 800 is one specific person. Danny Cahill, the winner of season eight: to hold his current 295 pounds he has to eat 800 kcal a day less than an ordinary man his size. That's one man's number from a news story, not the physiology of the species. The punishment for winning turned out to be lifelong — but it's his punishment, not yours.

— The opposite. The slowdown at the end of the show didn't predict who would regain at all: the correlation was −0.1, no significance. But there was a link with keeping the weight off, and it was positive, 0.59. The ones who slowed down the most were exactly the ones who, 6 years later, had kept off more of what they'd lost.

— It's the bill you get for winning. And it doesn't stop coming. There's an older and more careful measurement: if someone holds their weight 10% or more below where they started, their daily expenditure is on average 6 kcal per kilogram of lean mass lower. For a man with 55 kg of lean mass that's about 330 kcal a day he has to keep not eating, indefinitely.

— Indefinitely. Which is why the whole vocabulary of dieting — a course, a finish line, a result — is built wrong. There is no finish line, there's a different way of living.

— And you're right not to, because the other half of the truth is just as solid. In the early 1990s they took people who had spent years explaining to doctors that they weren't losing weight on a small diet, and put them under direct measurement. Their expenditure turned out to be normal. And they were eating on average 47% more than they recorded, and overstating their activity by about double.

— That's the interesting part — they weren't. They recorded honestly, intending to record the truth. Memory just works that way with food: it remembers a meal well and barely remembers what you ate standing up, off someone else's plate, on the way somewhere.

— I can't. Both are true at once, and that's exactly why the "willpower versus metabolism" argument has been running for 70 years. The adaptation is real and measured. The undercounting is real and measured. Nobody knows the proportion. Worse: Kevin Hall, the author of that very study on the show, wrote 6 years later that they may have been reading their own data from the wrong side — that it isn't about the weight loss at all, but about the monstrous volume of imposed exercise, and the body was damping down resting metabolism to keep total expenditure inside its usual corridor.

— The author himself proposed a different interpretation. That's not a measurement, it's a hypothesis. And another group, Martins and colleagues, argues on 71 participants that the adaptation half disappears if you measure it not during the deficit but after weight has stabilized. The argument is happening right now, and I don't know how it ends.

— It shouldn't. But there's a simpler thing you can use starting today, and it's about how a calorie isn't equal to a calorie even before any adaptation. Digesting protein costs 20–30% of its own calories, carbohydrates 5–10%, fat anywhere from zero to 3%.

— Are identical on the label. Of the chicken's 500, digestion itself takes 100–150. Of the butter's, 15 at most. The intake differs by roughly 100 kcal, and not one app on your phone shows you that.

— Because what it's defending isn't weight, it's fullness. And this is where it gets more interesting, because fullness is counted by completely different rules than the ones you think.