Restriction and bingeing

What thirty-six men in Minnesota demonstrated

In 1944 a group of healthy volunteers — screened for stability, chosen partly for it — agreed to eat roughly half of what they needed for six months, so that researchers could learn how to refeed a starving Europe. What the study is remembered for is not the refeeding protocol.

The men became obsessed with food. They read cookbooks for pleasure. They collected recipes, hoarded utensils, drew out meals for hours, talked about eating and little else. They became irritable, withdrawn and unable to concentrate. Several developed episodes of uncontrolled eating that appalled them, and some of that behaviour persisted for months after normal food was restored.

These were not people with a weak relationship to food. They had no relationship to food worth remarking on until the restriction created one. That is the finding: the behaviour was manufactured by the deprivation, in ordinary men, on purpose.

Why the body treats a diet as an emergency

It has no way to tell the difference between a shortage you chose and one you did not. The signals it reads are energy in, energy stored, and how long the gap has run — and none of those carry your intention.

So it does what it has always done in a shortage. Attention narrows onto food, because noticing food is how a hungry animal survives. Fullness signalling weakens. The reward attached to eating goes up, so the same meal is more compelling than it was a week ago. This is not weakness being revealed; it is a system working exactly as built, on someone who has decided the system is the enemy.

And it escalates rather than settling. The longer and harder the restriction, the stronger the pull — which is why the pattern so often ends in an episode that feels wildly out of proportion to the rule that started it.

The part that makes it a cycle

What turns an episode into a loop is what happens next. The episode is read as proof of a character flaw, and the response to a character flaw is a stricter rule. The stricter rule produces a stronger pull. The stronger pull produces a larger episode, which is read as further proof.

Each turn of that makes the next one more likely, and the person in it experiences the whole thing as evidence about themselves rather than as a predictable response to the thing they keep doing.

It is worth saying plainly, because it is the part people rarely hear: the fact that this is predictable does not make it a small problem. Predictable and serious are not opposites.

When this stops being a pattern

There is a line, and it is not drawn by how much someone eats in an episode. It is drawn by what the eating is doing to the rest of a life.

Some markers that it has been crossed: episodes accompanied by a real sense of loss of control rather than simple over-eating; anything done afterwards to compensate — vomiting, laxatives, punitive exercise, fasting the next day; food or body shape occupying so much attention that work, study or relationships are suffering; and secrecy, which is one of the most reliable signals of all.

None of that is diagnosis, and this page cannot do that. It is the point at which the right next step stops being a different eating strategy and starts being a person — a GP, a psychologist, a helpline. Eating disorders have the highest mortality of any psychiatric illness and they respond well to treatment, and both halves of that sentence are reasons to make the call early rather than late.

What this means for tracking anything, including with us

We make an app that counts things, so we have an obvious interest here and should declare it. Measuring what you eat is genuinely useful to some people and genuinely harmful to others, and which one you are is not determined by how disciplined you are.

If a number on a screen sets the tone of your day, if you have started eating around the app rather than using it, or if seeing a total makes you want to compensate — that is not a sign to track more carefully. Close it. That advice costs us a user and it is the correct advice.