What the best-controlled experiment ever run on this says, and what you can do with it tomorrow.
The advice is always the same: eat less, move more. It is arithmetically correct, and yet it fails for most people who apply it seriously. You hold out three weeks, you are hungry all the time, you give in, and you blame yourself for a lack of willpower.
That reading is probably wrong. Not because calories do not count, but because it treats appetite as a decision when it is a physiological response. And that response depends a great deal on what you eat, not only on how much.
In other words: the useful question is not “how do I eat less despite the hunger”, but “what makes me less hungry”.
In 2019, a team at the NIH ran a rare experiment: 20 adults housed in hospital for four weeks, fed exclusively by the researchers, with every plate weighed before and after.
Two diets, two weeks each, in a randomized order. One made of ultra-processed foods, the other of whole foods. The two were matched on calories offered, sugar, fat, salt, fiber and macronutrients. The only variable was the degree of processing.
Participants ate as much as they wanted. That is the whole point: they were not asked to hold back, what they took spontaneously was measured.
Result over two weeks, at equivalent nutritional composition
About 500 kcal of difference a day, with no instruction, no effort and nobody deciding it. The same people, two weeks apart, with foods equivalent on the label.
This is the important point: the difference does not come from a lack of willpower on one side and discipline on the other. These are the same individuals. What changed is what they were served.
Several mechanisms stack up, and none of them is psychological.
Calorie density. An ultra-processed food packs a lot of energy into little volume. The stomach fills up, but well after the calories have gone through. For the same felt fullness, you have swallowed more.
Eating speed. These products are designed to be soft, easy to chew, quick to swallow. Fullness signals take about fifteen minutes to reach the brain. Eating fast means eating before the signal arrives.
Protein. It is the most filling of the three macronutrients, and it also costs more to digest: part of its energy goes into its own absorption. Ultra-processed products are often low in it, despite flattering labels.
Fiber. It slows the stomach emptying and spreads out the arrival of glucose in the blood. A whole food carries it in its original matrix; a reconstituted product sometimes adds it back, but without that structure.
💡 This is also why fruit juice and whole fruit do not have the same effect at identical sugar. The fruit still has its matrix; the juice does not. You can compare the two in the table Fruit.
One lever is specific to carbs, independent of how processed the food is.
When a carb-heavy meal raises blood sugar quickly, the pancreas answers with a spike of insulin. Insulin is the storage hormone: while it is high, the body puts available energy away rather than drawing on its reserves. That is not a flaw, it is its job.
The problem is what follows. A strong insulin response often brings blood sugar back below where it started, one to three hours after the meal. The body then asks for sugar. That is the mid-morning craving, and it is nothing like a whim.
Two articles go into this mechanism: what happens after a meal, et glycemic index and glycemic load, which explains why quantity matters as much as speed.
And watch out for the label trap: most of the carbs in a grain product come from starch, which never appears on the “of which sugars” line. A muesli at “8.9g of sugar” can hold six times more in total carbs. That is the subject of hidden carbs.
None of these levers requires counting calories. They all work on hunger rather than against it.
Put a protein source in every meal.
This is the best value lever for fullness. Eggs, fish, meat, and on the vegetarian side far more options than people think.
Hard cheeses lead: 28g of protein per 100g of Comté, 35g for Parmesan, and not one gram of carbs. Then come skyr and cottage cheese (9 to 11g), fromage blanc, legumes, tofu and nuts.
One caveat on hard cheeses: at 400 kcal per 100g, they count as a serving, not as a free side. The tables Dairy & eggs, Meat, Fish and Seeds & nuts give the figures for comparison.
Read the “Carbohydrate” line, not the “of which sugars” line.
One move, turning the pack over. It is the one that exposes products presented as healthy that are not.
Replace before you remove.
Cutting a food out creates a gap; replacing it with a whole, more filling equivalent leaves none. Breakfast cereal for eggs, juice for the fruit, the afternoon cookie for a handful of nuts (Seeds & nuts table).
Fill the volume with vegetables.
Plenty of mass, plenty of fiber, few calories: it is the only way to increase what is on the plate while lowering its energy density. See the Vegetables table.
Measure once to calibrate.
Not to count every day, but to know what a typical day looks like. The macro calculator is exactly for that: you do it two or three times, adjust, and never come back to it.
A useful rule: aim for fullness, not restriction. If a change leaves you hungry, it will not last three weeks: it is not the right change, whatever it takes off in the short term.
| The received idea | What the measurement shows |
|---|---|
| A calorie is a calorie | True in thermodynamics, insufficient in practice: at identical calories offered, ultra-processed food gets 500 more eaten per day |
| It is a matter of willpower | The same people, two weeks later, spontaneously eat less when the food changes |
| You have to be hungry to lose weight | Hunger is the best predictor of a diet being abandoned |
| Only added sugar counts | Starch becomes glucose just as fast, and is not on the “of which sugars” line |
| You have to cut foods out | Replacing lasts better than removing |
The common thread: quantity is not what to tackle first, quality is. Quantity often follows by itself, because you are no longer hungry at the same moments.
⚕️ This article describes general mechanisms, it does not replace medical advice. Support from a doctor or dietitian is warranted in case of diabetes, ongoing treatment, pregnancy, or a history of disordered eating. Rapid, unexplained weight loss should always be checked by a professional.
ℹ️ Sources
• Randomized trial, ultra-processed versus whole food: Hall K.D. et al., Ultra-Processed Diets Cause Excess Calorie Intake and Weight Gain: An Inpatient Randomized Controlled Trial of Ad Libitum Food Intake, Cell Metabolism, 2019. Abstract on PubMed
• Full text: Cell Metabolism