All topics
lifeevergreen
Research-ready episode

This Diet Made People Eat 500 Extra Calories a Day

这套饮食让人每天多吃了500千卡

What this research must answer

A single-question constellation film. Atwater makes food energy measurable. The NIH ward then reveals that measurement cannot predict when people stop eating. Whole apples and almonds expose the missing role of structure. A swallowed balloon, gut signals and leptin show that appetite relies on indirect feedback. Food processing solves scarcity and safety while weakening the old coupling between energy and those signals. Eating rate resolves the 508-calorie mystery, a slow ultra-processed countercase prevents a simplistic villain story, and fat storage explains how a repeated meal-level mismatch becomes metabolic disease. Fasting and the large-meal absorption myth test the model before the film returns to Atwater's machine.

这是一部由单一问题牵引的科普纪录片。阿特沃特让食物能量变得可测量，NIH病房随后证明，测得出热量并不等于预测得了人何时停口。苹果和杏仁让食物结构显形，胃气囊、肠道信号和瘦素说明食欲依赖间接反馈。食品加工解决了匮乏、安全与保存，也让能量逐渐脱离过去伴随它而来的信号。进食速度解释508千卡之谜，慢速超加工食品的反例阻止我们把一切归咎于一个标签，脂肪储存则把一顿饭的偏差连接到代谢疾病。最后，16比8和大餐吸收迷思检验这套模型，影片再回到阿特沃特的机器。

Opening tension

In the winter of 1895 to 1896, four men took turns living inside Wilbur Atwater's sealed respiration calorimeter in Middletown, Connecticut. They were two laboratory assistants, a chemist and a physicist. Food and drink went in. Carbon dioxide and waste were measured coming out. The first report deferred the energy results, but Atwater's wider programme helped make human energy countable. Then, more than a century later, researchers at the US National Institutes of Health controlled every meal for twenty adults. Every plate and leftover was weighed. The participants could eat as much or as little as they wanted. On one menu, the same people chose to eat an average of 508 extra calories every day. They gained about 0.9 kilograms in two weeks. On the other, they lost about the same amount. The arithmetic worked perfectly after they ate. It could not explain why they kept eating. If the body obeys the law of energy, why can it fail to notice 500 calories until they are already inside?

Opening variants
  1. ·

    A. The controlled result

    Twenty adults lived in an NIH metabolic ward where every meal and leftover was weighed. On the ultra-processed menu, the same people chose to eat 508 extra calories a day and gained about 0.9 kilograms in two weeks. On the other menu, they lost about the same amount. What made the stop signal arrive so late?

    Fulfils the title immediately and opens a clean controlled mystery.

  2. ·

    B. The mouse that could not hear

    Douglas Coleman joined the blood supply of an obese diabetic mouse to another mouse. The partner stopped eating even with food beside it and could waste away. One mouse carried a powerful stop signal it could not hear. Decades later, the same human body could still eat 508 extra calories on one controlled diet.

    Opens on a memorable character and biological reversal before moving from leptin to the whole signal system.

  3. ·

    C. Four versions of an apple

    Four apple servings carried the same 125 calories and the same weight. Peeled apple segments made people fuller and reduced intake at lunch. Sauce and juice did less. Adding pectin to the juice at naturally occurring levels did not rebuild the effect of the apple. What information disappeared when the fruit lost its structure?

    Makes the invisible food matrix concrete and personally testable.

  4. ·

    D. The machine that could count a human

    In the winter of 1895 to 1896, four men took turns living inside Wilbur Atwater's sealed respiration calorimeter. They were two laboratory assistants, a chemist and a physicist. The first report deferred its energy results, but Atwater's wider programme helped make human energy countable. More than a century later, the NIH controlled every meal for twenty adults. On one diet, the same people chose to eat an average of 508 extra calories every day. The arithmetic worked after they ate. It could not explain why they kept eating.

    Turns the invention of calorie measurement into the exact mystery the film resolves and reaches the title result within the first minute.

How the episode moves
  1. 01

    If Atwater could measure human energy, why did the NIH participants choose 508 extra calories on one diet?

    Atwater's chamber established energy accounting, and modern labels estimate metabolizable energy from composition. Hidden oil and recipe variation explain many real-world errors, but the NIH team weighed food and leftovers, so hidden calories cannot explain the choice to eat more.

    USDA Bulletin 44 and USDA nutrition history; Hall et al., Cell Metabolism 2019

    Can foods with the same listed energy reach the body differently?

  2. 02

    Can foods with the same listed energy reach the body differently?

    Peeled apple segments produced more fullness than sauce or juice, and adding pectin to juice at naturally occurring levels did not restore the apple effect. Whole almonds also yielded less measured metabolizable energy than almond butter. Physical structure changes release and gastric processing.

    Flood-Obbagy and Rolls 2009; Mackie et al. 2020; Gebauer et al. 2016

    If structure changes delivery, what signal actually tells us to stop eating?

  3. 03

    What signal actually tells us to stop eating?

    Cannon and Washburn linked hunger pangs to stomach contractions, yet later balloon experiments showed that distension can increase fullness without reliably reducing intake. Soup, protein, viscosity and nutrient arrival show that meal termination emerges from several timed signals.

    Cannon and Washburn 1912; Geliebter 2006; Flood and Rolls 2007; gastrointestinal appetite reviews

    If the body has several brakes, why can long-term weight gain continue?

  4. 04

    If the body has several brakes, why can long-term weight gain continue?

    Coleman's parabiosis experiments and the discovery of leptin showed that fat tissue reports stored energy to the brain. Common obesity usually comes with high leptin, not an absent signal. Regulation responds more strongly to falling stores than to abundant stores.

    Coleman 2010; Zhang et al., Nature 1994; Blundell et al. 2010

    Why would an asymmetric system fail under one modern menu and not another?

  5. 05

    Why would an asymmetric system fail under one modern menu and not another?

    Cooking, milling, preservation and industrial formulation made energy safer, cheaper and easier to distribute. The same transformations can remove water, rupture structure and reduce chewing. The system optimizes safety, cost, shelf life and convenience, while satiety per calorie is rarely the primary target.

    Carmody et al. 2011; FAO food-processing and food-security review; Drewnowski and Specter 2004

    Which properties let energy outrun the body's feedback?

  6. 06

    Which properties let energy outrun the body's feedback?

    Energy density determines calories per gram, texture affects grams per minute, and reward can increase value for fat-carbohydrate combinations. The useful quantity is energy intake rate. Feedback from chewing, gastric stretch and intestinal nutrients takes time to accumulate.

    Forde et al. 2020; Teo et al. 2021; DiFeliceantonio et al. 2018

    Did the high-intake NIH menu actually deliver energy faster?

  7. 07

    Did the high-intake NIH menu actually deliver energy faster?

    Participants ate the ultra-processed menu at about 48 calories per minute versus 31. A later controlled comparison kept both diets ultra-processed and reduced daily intake by 369 calories when texture made eating slower. Processing marks a recurring bundle, not one toxin.

    Hall et al. 2019; Forde et al. 2020; Hamano et al. 2025

    How does a few hundred extra calories become metabolic disease?

  8. 08

    How does a few hundred extra calories become metabolic disease?

    Subcutaneous fat initially buffers surplus safely. Risk rises when susceptible bodies accumulate visceral and ectopic fat in liver and muscle, impairing insulin action and prompting pancreatic compensation. Reaven's Syndrome X connected the resulting cluster before obesity was part of its definition.

    Reaven 1988; ectopic-fat reviews and International Atherosclerosis Society position statement

    Can fasting or one huge meal escape the system?

  9. 09

    Can fasting or one huge meal escape the system?

    Time-restricted eating can remove eating opportunities, but randomized trials do not show a reliable calorie-independent advantage. The digestive tract can slow gastric emptying and has substantial reserve, so a large meal does not simply pass through unabsorbed. The practical answer is to improve the signals per calorie and reduce opportunities when that rule helps.

    Liu et al., NEJM 2022; Lowe et al., JAMA Internal Medicine 2020; NIDDK digestion overview; human intestinal reserve review

    Payoff. What was Atwater's machine measuring that the body itself could never see?

Story bank
  1. opening-pressure

    Four men inside Atwater's chamber

    In winter 1895 to 1896, two laboratory assistants, a chemist and a physicist took turns in four reported experiments while Atwater's team measured metabolic inputs and outputs.

  2. opening-pressure

    The 508-calorie ward

    Twenty adults ate both controlled diets in randomized order. On the ultra-processed diet they chose 508 more calories per day and gained about 0.9 kilograms in two weeks.

  3. mechanism

    Four forms of one apple

    Peeled segments, sauce, juice and juice with naturally occurring levels of pectin carried equal energy and weight, yet the segments produced more fullness and lower later intake.

  4. historical-turn

    Washburn swallows a balloon

    A stomach balloon and key presses created an early trace linking hunger pangs with stomach contractions, then later distension studies broke the simple empty-stomach theory.

  5. historical-turn

    Coleman's shared blood

    Parabiosis showed that a circulating signal could suppress eating powerfully and that an obese animal could produce a signal it could not hear.

  6. institution

    Processing solves scarcity

    Cooking, milling, preservation, refrigeration and formulation made food safer, stable and affordable while changing the physical signals accompanying energy.

  7. conceptual-reversal

    The slow ultra-processed diet

    Both menus remained ultra-processed, yet slowing texture-derived eating rate reduced intake by 369 calories per day, proving the category is not one mechanism.

  8. consequence

    Reaven's cluster without obesity

    Syndrome X linked insulin resistance, hyperinsulinaemia, dyslipidaemia and hypertension before obesity entered the cluster, forcing attention toward storage location and susceptibility.

  9. application

    The final causal map

    Oil awareness, intact structure, lower energy density, firmer texture, protein, structured fibre and fewer eating opportunities each repair a different part of the signal mismatch.

  10. return-payoff

    The machine and the body

    Atwater's chamber measured energy directly. The body could only infer it from the trail energy left behind.

How the system works
Input
Food energy arrives in a physical form with texture, volume, composition, structure and reward cues.
Transformation
Chewing, gastric processing and intestinal sensing generate delayed mechanical, neural and hormonal proxy signals while energy is consumed at a rate set by density and eating speed.
Output
Satiation ends the current meal, satiety delays the next one, and longer-term signals adjust appetite and expenditure.
Limit
The body never reads a direct calorie count. A high energy intake rate can deliver more energy before the proxies become strong enough to stop eating.
Mechanism cards

  1. Calorie estimation

    Input
    Measured or database-derived macronutrients and recipe information.
    Transformation
    General or food-specific energy factors convert composition into estimated metabolizable energy.
    Output
    Calories per serving for comparison and accounting.
    Limit
    The number does not encode physical matrix, eating speed, individual response or later compensation.
    Evidence
    USDA Atwater history and FDA label methodology

  2. Food matrix

    Input
    Intact cells, particle size, heat treatment and physical form.
    Transformation
    Chewing, grinding, cooking, gastric processing and enzymes alter access to nutrients.
    Output
    A food-specific rate and amount of nutrient release and gastric emptying.
    Limit
    The effect varies by food and never violates energy conservation.
    Evidence
    Flood-Obbagy and Rolls 2009; Mackie et al. 2020; Gebauer et al. 2016

  3. Satiation cascade

    Input
    Oral cues, chewing, gastric volume and nutrients reaching the intestine.
    Transformation
    Mechanical sensing, gastric emptying, vagal signalling and gut-peptide release integrate over time.
    Output
    Meal termination and post-meal satiety.
    Limit
    These are delayed proxy signals, not a direct calorie meter.
    Evidence
    Geliebter 2006; Flood and Rolls 2007; gastrointestinal appetite review

  4. Leptin regulation

    Input
    Changes in adipose stores.
    Transformation
    Fat cells secrete leptin and the brain integrates the signal with other inputs.
    Output
    Changes in appetite and energy expenditure, especially when stores fall.
    Limit
    Common obesity usually includes high leptin and reduced responsiveness, so adding leptin is not a universal off switch.
    Evidence
    Coleman 2010; Zhang et al. 1994; Blundell et al. 2010

  5. Energy intake rate

    Input
    Food energy density and texture-dependent grams consumed per minute.
    Transformation
    Calories arrive while oral, gastric and intestinal feedback is still accumulating.
    Output
    The amount of energy delivered before satiation ends the meal.
    Limit
    A major mediator and useful design target, not a complete explanation for every food or person.
    Evidence
    Forde et al. 2020; Teo et al. 2021; Hamano et al. 2025

  6. Food-system optimization

    Input
    Raw and perishable ingredients plus constraints on safety, price, labour and distribution.
    Transformation
    Cooking, preservation, refinement, formulation, packaging and logistics make food stable and convenient.
    Output
    Safe, consistent, portable and affordable energy.
    Limit
    Satiety per calorie is rarely the primary optimization target, and affordability remains a real constraint.
    Evidence
    Carmody et al. 2011; FAO food-security review; Drewnowski and Specter 2004

  7. Ectopic-fat pathway

    Input
    Repeated energy surplus and an individual's adipose-storage capacity and distribution.
    Transformation
    Subcutaneous storage expands, while susceptible bodies accumulate visceral fat and lipid in liver or muscle.
    Output
    Impaired insulin action, compensatory insulin production and clustered cardiometabolic risk.
    Limit
    Metabolic disease is heterogeneous and body weight alone cannot reveal storage location.
    Evidence
    Reaven 1988; Neeland et al. 2019
Where this changes a real decision

SignalA meal's visible portion, physical structure, energy density, protein and fibre context, texture and likely eating rate.

Decision ownerA diner, food-product designer or procurement owner choosing how energy will be packaged and consumed.

ThresholdAsk whether energy remains coupled to chewing, volume, structure and nutrient feedback, and whether the food can be eaten rapidly.

ActionPreserve intact structure where practical, make hidden fats visible, combine energy with protein and structured fibre, reduce excessive energy density or eating speed, and use timing rules only when they reduce opportunities.

ConsequenceThe body's proxy signals become more informative before a large surplus arrives, improving the odds of easier intake control without promising automatic weight loss.

Application chains

  1. Build a meal whose signals can keep up

    SignalVisible oil, physical structure, energy density, protein and fibre context, texture and eating speed.

    Decision ownerA shopper, cook or diner.

    ActionChoose intact forms where practical, make calorie-dense additions visible, add protein and structured fibre, and use soups or lower-density foods when they help slow intake.

    ConsequenceMore chewing, volume and nutrient feedback can arrive before a large surplus, making control easier without guaranteeing weight loss.

  2. Design satiety into processed food

    SignalControlled evidence on eating rate, energy density, texture and nutrient feedback.

    Decision ownerA food manufacturer or institutional procurement team.

    ActionChange texture, density, portion architecture, protein or fibre while preserving safety, price and shelf life.

    ConsequenceProducts may deliver energy more slowly and strengthen feedback per calorie, subject to product-level and long-duration testing.

  3. Use timing as an environment rule

    SignalA person repeatedly eats through late-night opportunities or decision fatigue.

    Decision ownerThe person choosing a sustainable eating routine.

    ActionUse a limited eating window when it reliably removes opportunities, without assuming the clock makes calories disappear.

    ConsequenceFewer decisions can reduce total intake for some people, while adherence and total energy remain decisive.

The limit this episode must keep

Ultra-processed is a useful population-level category, not a single causal molecule. In a controlled trial, two diets both dominated by ultra-processed food produced different intake when texture-derived eating rate changed. Whole nuts and nut butter also show that processing effects are food-specific. The conclusion is a mechanism map, not a purity rule.

Key sources