Chapter 33 ★ 🚪 The Psychology of Eating

🚪 THRESHOLD CONCEPT. ⚠️ This chapter contains one of the twelve ideas in this book that change what everything else means. Once you have it, a great deal of Parts IV, V and VI will look different — and several things you have been told about your own behaviour will stop making sense in the way they used to.

The Hook: The chapter that had to come last

⚠️ Chapters 30, 31 and 32 all assumed the same thing, and none of them examined it.

⚠️ They assumed that once the label is decoded, the plan is written, the list is made, the money is accounted for and the food is in the house — a person eats what they intended to eat.

⚠️ Theo Vasquez has read all three.

⚠️ His fridge, on a Tuesday in March, contained: the components from Sunday, eggs, milk, a bag of carrots, two tins of beans, frozen vegetables, and the bad-night meal on the shelf where it lives. ⚠️ His plan for Tuesday said "rice bowl, 12 minutes."

⚠️ At 21:40 he ate most of a bag of tortilla chips standing up, and then made the rice bowl anyway.

⚠️ He was not hungry at 21:40. He knew he was not hungry at 21:40. He could have told you, at 21:39, exactly what was about to happen and why it was a bad idea.

⚠️ Everything in Chapters 30 to 32 worked, and it did not help.

⚠️ Here is the standard explanation, and it is wrong: he lacked willpower.

⚠️ Here is the reason it is wrong. If willpower were the mechanism, then the same person, on the same day, with the same reserves of resolve, would behave the same way in front of any food. ⚠️ Put a bag of tortilla chips in front of Theo at 21:40 and it disappears. Put a bag of raw carrots in front of Theo at 21:40 and it does not. Same person, same hour, same willpower.

⚠️ Which means willpower is not what varied. Something about the FOOD varied, and it interacted with something in him that was not hunger.

🚪 THE THRESHOLD.

⚠️ Hunger is not one system. It is at least two.

⚠️ One is HOMEOSTATIC — the body's energy-regulation system, which tracks stores, signals need, and is the thing you feel when you have not eaten since breakfast.

⚠️ The other is HEDONIC — the reward system, which responds to palatability, cue, novelty, context and expectation, and which will happily initiate eating in a person who is not remotely short of energy.

⚠️ AND MOST DIETS FIGHT THE WRONG ONE.

⚠️ They apply homeostatic tools — eat less, restrict, count, delay — to a hedonic problem. And the hedonic system does not respond to those tools. ⚠️ Worse: restriction makes it stronger.

⚠️ Both halves of that matter, and the second half is the one people miss.

⚠️ This is not "hunger is real and cravings are fake." Both systems are real, both are physiological, both have measurable substrates, and neither is a moral category. ⚠️ The hedonic system is not a character defect that a better person wouldn't have. It is a normally functioning reward system doing exactly what it evolved to do, in an environment it did not evolve in (Chapter 22).

33.1 The homeostatic system, briefly

⚠️ Chapter 24 did most of this and it is not going to be re-litigated. What matters here is the shape.

⚠️ The body tracks its own energy stores and defends them, using signals that operate on different timescales:

⚠️ Two distinctions worth having precisely, because they are used loosely everywhere:

⚠️ SATIATION ends a meal. SATIETY delays the next one. ⚠️ A large glass of water produces satiation and no satiety. A bowl of lentils produces both. Most "filling food" advice conflates them and then wonders why the effect doesn't last.

⚠️ And the finding from Chapter 24 §24.5 that this chapter builds on directly:

⚠️ After weight loss, the largest and most durable change is in APPETITE, not metabolic rate.

⚠️ Ghrelin rises. Satiety hormones fall. The changes persist for a year or more. The body's response to losing weight is to make you hungrier, and to keep doing so.

⚠️ That is the homeostatic system working correctly. It is not a malfunction and it is not psychological.

33.2 ⚠️ The hedonic system, and why it is not "wanting food"

⚠️ Here is the part that is usually explained badly.

⚠️ The reward system is not a system for enjoying things. It is a system for LEARNING WHAT TO APPROACH. ⚠️ And it separates into two components that feel like one thing and are not:

⚠️ These come apart, and the fact that they come apart is the single most useful thing in this chapter.

⚠️ You can want something intensely and not much enjoy it. This is not a paradox and it is not self-deception — it is two systems with different substrates producing different outputs.

⚠️ Theo's tortilla chips at 21:40 were high in WANTING and moderate in LIKING. He has said as much: "I didn't even enjoy them that much."

⚠️ That sentence is usually treated as evidence of irrationality. It is evidence of accurate introspection about two different systems.

⚠️ What drives WANTING, and none of these is hunger:

⚠️ CUES. The sight of the packet. The time on the clock. The sofa. The end of a work task. ⚠️ Cues acquire motivational force through learning, and they keep it.

⚠️ PALATABILITY. ⚠️ Combinations of fat, sugar, salt, texture and aroma that do not occur together in whole foods and do occur together in engineered ones (Chapter 22 again, and this is the mechanism behind that chapter's finding).

⚠️ VARIETY. ⚠️ Sensory-specific satiety — you tire of the food you are eating, and not of a different food. This is why the meal you could not finish makes room for dessert, and it is why a buffet increases intake reliably.

⚠️ CONTEXT AND EXPECTATION. ⚠️ Portion size, plate size, packaging, price, the presence of others, and simply what you were told the food was.

⚠️ NOVELTY AND UNPREDICTABILITY. ⚠️ Reward systems respond more strongly to variable outcomes than to reliable ones.

⚠️ Notice what is absent from that list: energy need.

⚠️ The hedonic system does not consult the homeostatic system before firing. It can, and routinely does, initiate eating in someone who is in energy surplus.

33.3 🚪 Why the wrong-tool problem is the whole chapter

⚠️ Now the threshold, in operational form.

⚠️ Read the second and fourth rows together. They are the two halves of the same error, in opposite directions, and almost all popular diet advice commits one of them.

⚠️ The second row is the common one: applying homeostatic tools to hedonic eating.

⚠️ "Just eat less" is an instruction to the homeostatic system. The homeostatic system was not the one that opened the packet. ⚠️ Telling it to be quieter does nothing to a reward system responding to a cue, and it degrades your energy status, which — see §33.6 — makes the reward system MORE responsive.

⚠️ You lower your defences against the system that wasn't the problem, and strengthen the one that was.

⚠️ That is not a metaphor. It is why restriction reliably produces preoccupation with food.

⚠️ The fourth row is rarer and does real damage: treating genuine hunger as a hedonic impulse to be resisted. ⚠️ "You're not hungry, you're bored" said to someone who has eaten 1,200 kcal since Sunday. This is a teaching error with a direct route into Chapter 34.

⚠️ And the diagnostic question this produces, which is the single most practically useful thing in the chapter:

⚠️ NOT "am I really hungry?" — which is unanswerable in the moment and invites self-interrogation that helps nobody.

⚠️ INSTEAD: "what STARTED this?"

⚠️ A cue — a time, a place, a packet, a feeling, the end of something — or an interval since food?

⚠️ Because the answer determines which tool works, and using the wrong one is worse than doing nothing.

33.4 The food environment is a psychological intervention

⚠️ Chapter 22 established that what is available changes what is eaten, in a controlled setting, in the same people. This is why.

⚠️ Availability, proximity and effort are cue variables, and cue variables act on the hedonic system directly, without requiring any decision at all.

⚠️ This reframes something Part VI has been recommending for three chapters without explaining.

⚠️ Chapter 31's bad-night plan is not a nutrition intervention. It is a CUE intervention.

⚠️ It works because at the moment of collapse — the moment the hedonic system is most responsive and the homeostatic system is genuinely depleted — the decision has already been made, and made by someone who was not depleted.

⚠️ And Chapter 31's shopping list is the same move, one step upstream:

⚠️ The list is a decision made by Sunday-you, in a kitchen, unhungry, about what Wednesday-you will have available at 21:40. It is not a planning tool. It is a mechanism for putting your intentions somewhere your reward system cannot reach them.

⚠️ Which is why "keep it out of the house" outperforms "have more self-control" in essentially every comparison anyone has run. You are not being asked to be stronger. You are being asked to make the decision once, in advance, in a state where it is easy.

33.5 🪞 Learning Check-In

⚠️ Before you read the next section. Answer, then keep reading, then come back.

⚠️ 1. You have eaten a full dinner. You are pleasantly full. Someone brings out a dessert you like. ⚠️ You eat it. Which system fired, and what triggered it?

⚠️ 2. A person restricts their eating for six weeks and loses weight. They report thinking about food constantly. ⚠️ Is that a psychological failing, a physiological consequence, or both?

⚠️ 3. Someone tells you they crave chocolate because their body needs magnesium. ⚠️ Using this chapter's framework, what is wrong with that explanation — and what would have to be true for it to be right?

⚠️ 4. You are asked to design a single intervention for someone who eats well all day and then eats a large amount in the evening while watching television. ⚠️ What do you change, and what do you deliberately NOT change?

⚠️ Write your answers down. §33.9 and §33.12 will grade them, and question 3 is the one most people get wrong for an interesting reason.

33.6 ⚠️ Restriction, preoccupation, and the cycle that runs through the rest of the book

⚠️ This section sets up Chapter 34 and does not resolve it. That division is deliberate.

⚠️ The finding is old, robust, and consistently under-taught: restriction increases the salience of food.

⚠️ The best-known demonstration is the Minnesota Starvation Experiment (Ancel Keys, 1944–45) — 36 conscientious objectors, semi-starvation for six months. ⚠️ It is a Tier 1 named study and its findings are not in dispute:

⚠️ Participants became preoccupied with food to the point of obsession — collecting recipes, reading cookbooks, talking about food constantly. ⚠️ They reported irritability, apathy, social withdrawal and depression. ⚠️ Some developed ritualized eating behaviours. And during refeeding, many ate far beyond apparent need and reported that fullness did not register.

⚠️ These were psychologically screened, healthy young men with no history of disordered eating.

⚠️ The behaviours were produced BY the restriction. They were not pre-existing traits revealed by it.

⚠️ That distinction is the whole point, and it is the same structural-not-personal move that Chapter 31 §31.2 and Chapter 32 §32.11 made — arriving now for the third time, in the place it matters most.

⚠️ The cycle, stated plainly:

⚠️ Two things about that cycle deserve emphasis.

⚠️ First, step 5 is load-bearing. ⚠️ The self-blame is not an epiphenomenon — it is what drives the escalation at step 6. A person who interpreted step 4 as "predictable physiological consequence" rather than "moral failure" would not tighten the restriction, and the cycle would not amplify.

⚠️ Second, restriction here is not only dieting.

⚠️ Food insecurity produces the same cycle. Chapter 32 §32.10's finding that food insecurity and obesity co-occur has this as one of its proposed mechanisms: cyclical restriction and over-consumption tracking the pay cycle.

⚠️ A household that runs short in week three every month is running a restriction–disinhibition cycle whether or not anyone has called it a diet.

⚠️ Chapter 34 takes this into clinical territory — where the cycle becomes self-sustaining, where it meets body image, perfectionism, control and trauma, and where it stops being a description of ordinary eating. ⚠️ This chapter's job is to establish that the mechanism is normal physiology before that chapter shows what happens when it is not.

33.6b ⚠️ Children, pressure, and the two directions restriction backfires

⚠️ Chapter 25 §25.6 made this point once. It belongs here properly, because it is the cleanest demonstration in the book that the mechanism in §33.6 is not about dieting.

⚠️ Two interventions, opposite intentions, and both increase the behaviour they were meant to suppress.

⚠️ The mechanism is the same one as §33.6, arriving before anyone has ever heard the word diet.

⚠️ Restriction of a specific food raises its salience — the effect is not learned from diet culture, it is what the reward system does when access to a valued stimulus becomes unreliable. ⚠️ And pressure converts eating from something you do into something done to you, which is a reliable way to reduce liking for anything.

⚠️ The child who eats an entire packet at a friend's house is not badly behaved. They are behaving exactly as the restriction predicts.

⚠️ What the evidence supports instead — and it is uncomfortable for most people's instincts:

⚠️ Repeated NEUTRAL exposure. ⚠️ Offering a food, without comment, many times. The literature on this suggests it often takes far more exposures than a parent will manage before concluding a child "doesn't like" something — ⚠️ and that comment-free is the operative word, because praise works about as badly as pressure.

⚠️ The division of responsibility, which is the most useful single idea in child feeding:

⚠️ The adult decides WHAT is offered, WHEN, and WHERE. The child decides WHETHER and HOW MUCH.

⚠️ It is uncomfortable, it removes the adult from the part of the transaction they most want to control, and it is what the evidence supports.

⚠️ Alma Ortiz eats nine foods (Chapter 11). ⚠️ The intervention that made it worse was the one everybody tries first: negotiating at the table. The intervention that helped was the one that looks like doing nothing — putting a tenth food on the table, weekly, saying nothing about it, and being prepared for it to come back untouched fourteen times.

⚠️ Marisol's summary: "The hardest part was not mentioning it."

⚠️ And the reason this section sits in a chapter about adult eating: many adults are running the outputs of exactly these processes. ⚠️ The forbidden-food dynamic in §33.9 and the clean-your-plate override of satiation are frequently thirty years old and were installed with the best of intentions.

33.7 Emotional eating, described accurately

⚠️ This term is used to mean four different things, at least three of which are normal.

⚠️ What the evidence actually supports:

⚠️ Acute stress reduces intake in some people and increases it in others, fairly reliably by individual — ⚠️ the split is roughly comparable in size in both directions, which is why "stress makes you eat" and "stress makes you lose your appetite" are both things people say with confidence.

⚠️ Chronic stress, poor sleep and low mood are associated with higher intake of highly palatable food specifically — ⚠️ 🟢, and the mechanism is plausibly a shift in reward sensitivity rather than in hunger.

⚠️ And the practical point that matters more than the mechanism:

⚠️ If eating is doing a job — regulating a feeling, marking the end of a day, providing the only reliably pleasant twenty minutes in it — then removing the eating without replacing the job fails.

⚠️ It fails reliably, and the failure is then attributed to the person rather than to the intervention.

⚠️ The useful question is therefore not "how do I stop emotional eating" but ⚠️ "what is this eating doing, and what else could do it?" — and sometimes the honest answer is that nothing else is available right now, in which case the eating is the best available option and should be left alone.

33.8 Habit, and why it is good news

⚠️ A large fraction of eating is not decided at all. It is cued, automatic and stable.

⚠️ Habit has three components: a CUE, a BEHAVIOUR, and a context that binds them. ⚠️ Once established, the behaviour runs with minimal deliberation, which is why it survives intentions, knowledge and resolutions unchanged.

⚠️ Why this is good news rather than bad:

⚠️ A behaviour that runs automatically requires no willpower to maintain. The same property that makes an unwanted habit hard to stop makes a wanted habit nearly free to keep.

⚠️ The 21-day figure has no serious evidential basis. The studies that exist suggest a median nearer two to three months with enormous individual variation, and some behaviours never become automatic at all.

33.9 ⚠️ Cravings, and the magnesium question

⚠️ Learning Check-In question 3, answered.

⚠️ The claim: you crave chocolate because your body needs magnesium.

⚠️ What is wrong with it — and it is worth being precise, because the claim has a shape that recurs:

⚠️ 1. If the body could detect and target specific micronutrients through craving, the craving would be for the richest available source. ⚠️ Nuts, seeds, beans and leafy greens contain more magnesium than milk chocolate. Nobody craves spinach.

⚠️ 2. Cravings are strongly culturally patterned. ⚠️ Chocolate craving is common in some countries and rare in others with comparable magnesium status. A physiological detection mechanism should not have a nationality.

⚠️ 3. The craving tracks the FOOD, not the nutrient. ⚠️ People who crave chocolate are not satisfied by magnesium supplements, and are satisfied by chocolate.

⚠️ 4. And the general form of the error: it is Chapter 2's healthy-user template inverted — ⚠️ a correlation between a food and a compound, converted into a causal claim about a mechanism nobody tested.

⚠️ Verdict: ❌ Not supported. ⚠️ Cravings are learned, cue-driven and specific to foods, not to nutrients.

⚠️ And what WOULD have to be true for it to be right, because the question was posed that way deliberately:

⚠️ You would need a detection mechanism for the nutrient, a route from that detection to a food-specific motivational state, and evidence that supplying the nutrient by another route abolishes the craving. ⚠️ The one clear case where something like this happens is SODIUM — sodium appetite is a genuine, specific, physiologically driven appetite with a known mechanism.

⚠️ Which is the honest version: specific nutrient appetites exist, sodium is the clear one, thirst is arguably another — and there is no good evidence that the mechanism extends to magnesium, chocolate, or the general run of things people crave.

⚠️ What cravings actually respond to:

⚠️ Cue exposure · restriction of the specific food (which increases craving for it — a well-replicated finding and a direct argument against forbidden-food lists) · stress · low mood · sleep debt · and hormonal cycling in some people, reliably enough to be worth planning around rather than arguing about.

33.9b ⚠️ Eating with other people, which is most eating

⚠️ A large environmental effect, well replicated, and almost entirely absent from advice — which is odd, because for most of human history and most people now, eating is a social act.

⚠️ Three separate effects, and they push in different directions.

⚠️ 1. SOCIAL FACILITATION. ⚠️ Meals eaten with other people are larger and longer than meals eaten alone, and the effect scales with the number of people. ⚠️ The mechanism is mostly duration — you stay at the table longer, and time at the table predicts intake.

⚠️ 2. MODELLING. ⚠️ People adjust intake toward what their companions eat, substantially, and without noticing. ⚠️ This is one of the more reliably replicated findings in the area, it works in both directions, and it operates on what people choose as well as how much.

⚠️ 3. IMPRESSION MANAGEMENT. ⚠️ People eat less in front of people they want to impress, and the effect is not symmetric across groups — it falls hardest on those for whom eating is most scrutinized, which is a Chapter 34 problem in embryo.

⚠️ Effects 1 and 2 can oppose each other, which is why "eating with others makes you eat more" and "eating with others makes you eat better" are both defensible and both incomplete.

⚠️ The honest version: the company changes the meal, and WHICH company determines the direction.

⚠️ What follows from this practically:

⚠️ And the point that matters most for anyone trying to change how they eat:

⚠️ You are not the unit of intervention. The household is.

⚠️ Chapter 31 §31.13b treated this as a logistics problem — who plans, who shops, who cooks. ⚠️ It is also a psychological one: an individual changing their eating inside a household that hasn't is fighting a modelling effect at every meal, and losing to it is not a willpower failure.

⚠️ Which is the section's argument in one line, and it is this chapter's threshold again in a different costume: the situation you eat in is doing more work than the decisions you make in it.

33.10 ⚠️ Mindful and intuitive eating, assessed honestly

⚠️ Two popular approaches, and this book's usual standard applies.

⚠️ INTUITIVE EATING — rejecting external diet rules and eating in response to internal hunger and fullness cues.

⚠️ MINDFUL EATING — attending to the sensory experience of eating, slowly, without distraction.

⚠️ The last row is where the honest criticism sits, and it follows directly from this chapter's threshold.

⚠️ Intuitive eating is a homeostatic strategy. It asks you to listen to the homeostatic system. ⚠️ That is excellent advice for someone whose problem is that they have been overriding it with rules — which is a great many people, and for them the approach is genuinely good.

⚠️ It says much less about the hedonic system, which is not a signal to be listened to and does not become accurate when attended to. ⚠️ "Listen to your body" is not useful guidance about a bag of tortilla chips at 21:40, because the body is not what is talking.

⚠️ The fair summary: intuitive eating is a good corrective for rule-driven, restrictive, externally-controlled eating, and a weak tool against cue-driven eating in a food environment engineered to produce it.

⚠️ It is treating the right problem for some people and the wrong one for others, which is this chapter's threshold applied to a movement.

⚠️ And one thing it gets unambiguously right, which the evidence supports and this chapter endorses: forbidden-food lists increase craving for the forbidden food. ⚠️ Permission genuinely reduces salience. That is not a slogan — it is §33.9's restriction finding, and it is one of the better- replicated results in this area.

33.11 Sleep, alcohol, and the two easiest levers nobody mentions

⚠️ Short section, high yield.

⚠️ SLEEP. Short sleep is associated with higher intake, and the experimental work supports a causal component.

⚠️ Mechanisms: ghrelin up, leptin down, more hours awake in which to eat, and — probably most importantly — increased reward responsiveness to palatable food specifically. ⚠️ Tired people do not eat more broccoli.

⚠️ Verdict: 🟢 Probably true, and it is the intervention with the best ratio of effect to effort in this entire chapter — which is why it is striking how rarely eating advice mentions it.

⚠️ ALCOHOL. Three separate effects, usually collapsed into one:

⚠️ It supplies energy (about 7 kcal/g) that does not register as food (Chapter 12 — Theo's 6%). ⚠️ It acutely stimulates appetite. ⚠️ And it impairs the exact executive processes that any intention depends on.

⚠️ Which makes it a hedonic-system amplifier and a restraint-system suppressant simultaneously. ⚠️ The 21:40 problem and the 22:40 problem are usually the same problem with a drink in the middle.

⚠️ Neither of these is a moral recommendation. They are the two variables that most reliably change the difficulty of everything else in this chapter, and they are almost never on the list.

33.12 ⚠️ This is not a character flaw

⚠️ The third use of a move this book has now made in three consecutive chapters, and the most important one.

⚠️ Chapter 31 §31.2: the failure modes of a meal plan are properties of the PLAN, not the person. ⚠️ Chapter 32 §32.11: what looks like a competence problem is usually a resource problem. ⚠️ And here: what is called a willpower problem is usually a mismatch between the tool and the system.

⚠️ The reasons this matters practically rather than just kindly:

⚠️ 1. The self-blame is load-bearing. ⚠️ §33.6's step 5 is what drives the escalation at step 6. A person who reads a lapse as information tightens nothing. A person who reads it as failure restricts harder, and the cycle amplifies.

⚠️ 2. The diagnosis determines the tool. ⚠️ "I need more willpower" leads to trying harder at the thing that was already the wrong intervention. "That was a cue at a predictable time" leads to changing the cue, which works.

⚠️ 3. The environment is genuinely engineered. ⚠️ Chapter 22's Group 4 foods are formulated, tested and optimized for exactly the properties §33.2 lists. Losing to a professionally optimized stimulus is not a personal deficiency.

⚠️ 4. And willpower is the wrong unit of analysis regardless. ⚠️ Whatever the state of the ego-depletion literature — and it is contested, with the original effects failing to replicate at their claimed sizes — the practical point does not depend on it. ⚠️ People who appear to have more self-control mostly report exercising it LESS often, because their environments generate fewer conflicts.

⚠️ The people who look disciplined are usually not winning more fights. They are having fewer.

⚠️ What this does not mean, said plainly so it is not misread: none of this removes agency. ⚠️ You can change cues, environments, schedules, sleep and availability, and those changes work. ⚠️ The argument is not that behaviour is uncontrollable. It is that the control operates upstream of the moment, and the moment is where everyone has been told to apply it.

33.13 What actually works, ranked

⚠️ Assembled from the above, ordered by evidence and effect size, and honest about which are which.

⚠️ Nine of the thirteen work by changing the situation rather than the person.

⚠️ That distribution is the chapter's argument in a table.

33.13b ⚠️ Who this chapter is for — and who it isn't

⚠️ Part VI has done this in every chapter and this one needs it most, because the framework above is genuinely useful for some readers and genuinely insufficient for others.

⚠️ If the right-hand column is describing you, this chapter's tools are not wrong — they are addressed to a different problem, and applying them harder is the specific failure mode §33.6 predicts.

⚠️ Chapter 34 is written for the right-hand column. It is not a more intense version of this chapter. It is a different one, and it says where to get help.

⚠️ And one row that belongs to nobody's column: if you are currently food insecure, the cycle in §33.6 is being driven by something no psychological intervention addresses. ⚠️ Chapter 32 §32.8 is the relevant section, and it is not a lesser answer.

🧾 What this chapter is worth

⚠️ Part VI costs everything, so this one should too — and the honest accounting here is unusual, because most of the value is in what you STOP buying.

⚠️ The systemic-problems-as-purchases pattern, one last time: an engineered food environment produces a predictable failure, and an industry sells you a personal remedy for it.

⚠️ The remedy that works is free and is not a purchase. It is deciding once, in advance, what is in the house.

33.14 What we don't know, and how firmly I hold this

⚠️ And what this field is bad at, stated in the way Chapter 29 §29.11 established:

⚠️ Much of the behavioural nutrition literature consists of small laboratory studies, in undergraduates, measuring a single meal, in a room, with no follow-up. ⚠️ A substantial number of its best-known findings have replicated poorly — the portion-size and plate-size literature especially, and one prominent food-psychology lab had a large body of work retracted.

⚠️ I have tried to lean on the findings that survived and to mark the ones that are shakier. ⚠️ Where a claim in this chapter is 🟡, that is usually why.

Spaced Review

From Chapter 22: ⚠️ What did the ward trial find, and which system in this chapter explains it?

From Chapter 24 §24.5: ⚠️ After weight loss, what changes more — appetite or metabolic rate? How long does it persist?

From Chapter 21: ⚠️ What is ghrelin entrainment, and why does it make meal timing a lever?

From Chapter 31 §31.11: ⚠️ What is the bad-night plan, and what does §33.4 say it actually is?

From Chapter 32 §32.10: ⚠️ Why do food insecurity and obesity co-occur, and which mechanism in §33.6 is one of the proposed answers?

From Chapter 12: ⚠️ What fraction of Theo's intake was alcohol, and what three things does §33.11 say alcohol does?

From Chapter 2: ⚠️ What is the healthy-user template, and how does §33.9 show the magnesium claim is a version of it?

Project Checkpoint: Find Your Cue

⚠️ Progressive project, Phase 5.

Step 1 — ⚠️ For one week, record every eating occasion you did not plan. No judgement, no calories. Four columns.

Step 2 — ⚠️ Apply §33.3's diagnostic to each one.

⚠️ NOT "was I really hungry." Ask: WHAT STARTED THIS?

⚠️ A cue (time, place, packet, feeling, end of something) — or an interval since food?

Step 3 — ⚠️ Count.

Cue-initiated: _ Interval-initiated: _

⚠️ Most people are surprised by this ratio in one direction or the other, and both are informative.

Step 4 — ⚠️ Take the most frequent CUE and change the situation, not the response.

The cue: __ ⚠️ What I will change about the SITUATION: __ ⚠️ What I am explicitly NOT going to do: try harder at that moment.

Step 5 — ⚠️ Take the most frequent INTERVAL and fix the earlier meal.

⚠️ If evening eating follows an inadequate lunch, the intervention is lunch. It is not the evening.

Step 6 — ⚠️ And the honesty check.

⚠️ "Is any of my eating doing a JOB — regulating something, marking something, providing the only pleasant twenty minutes in the day?" __

⚠️ If yes: what else could do that job? And if the honest answer is "nothing available right now," leave it alone. That is a legitimate answer (§33.7).

Chapter Summary

🚪 THE THRESHOLD: hunger is two systems — homeostatic and hedonic — and most diets fight the wrong one.

⚠️ The homeostatic system tracks energy stores and defends them — ghrelin, leptin, satiation and satiety signals — and after weight loss it responds by making you durably hungrier (Ch 24 §24.5).

⚠️ The hedonic system responds to cue, palatability, variety, context and novelty, and does not consult energy need before firing. ⚠️ It separates into WANTING and LIKING, which come apart — which is why "I didn't even enjoy them that much" is accurate introspection rather than irrationality.

⚠️ §33.3: the error runs in both directions. ⚠️ Applying "eat less" to cue-driven eating fails and makes it worse; treating genuine hunger as an impulse to resist produces preoccupation. ⚠️ So the useful question is not "am I really hungry" but "WHAT STARTED THIS?" — a cue, or an interval.

⚠️ §33.4: availability and cues act directly, without a decision. ⚠️ Which reveals what Part VI has been recommending without explaining: the bad-night plan and the shopping list are CUE interventions, made by an unhungry person in advance, in a state where the decision is easy.

⚠️ §33.6: restriction increases food salience — the Minnesota Starvation Experiment produced obsessive food preoccupation in screened, healthy men with no prior disordered eating. ⚠️ The behaviours were produced by the restriction, not revealed by it. ⚠️ And the cycle's step 5 — self-blame — is load-bearing, because it is what drives the escalation. ⚠️ Food insecurity runs the same cycle without anyone calling it a diet (Ch 32 §32.10).

⚠️ §33.9: nutrient-specific cravings are ❌ not supported beyond sodium. Cravings are learned, cue-driven and food-specific — and restricting a specific food increases craving for it.

⚠️ §33.10: intuitive eating is a good corrective for rule-driven eating and a weak tool against cue-driven eating. It is treating the right problem for some people and the wrong one for others.

⚠️ §33.11: sleep is the best effort-to-effect intervention in the chapter, and alcohol amplifies the hedonic system while suppressing restraint — the same problem with a drink in the middle.

⚠️ §33.13: nine of the thirteen things that work operate by changing the situation rather than the person.

⚠️ And §33.12: the people who look disciplined are usually not winning more fights. They are having fewer.

What's Next

⚠️ §33.6 described a cycle and stopped deliberately.

⚠️ Chapter 34 is where restriction, preoccupation and disinhibition stop being a description of ordinary eating and become something else — where the cycle becomes self-sustaining, where it meets body image, control, perfectionism and trauma, and where the framework in this chapter stops being sufficient.

⚠️ It is written carefully. It contains no numbers that could function as targets, it signposts where to get help, and it does not tell anybody to just eat normally.

⚠️ It is also the chapter most likely to be personally relevant to a reader, which is why it is in a book about nutrition science at all.