Chapter 34 Eating Disorders

⚠️ CONTENT NOTE, and please read it before the chapter.

⚠️ This chapter is about eating disorders. It discusses restriction, loss of control around food, distress about body and weight, and the medical consequences of all three.

⚠️ It contains NO calorie figures, NO weights, NO body-mass thresholds, NO durations of restriction, and NO description of methods. ⚠️ That is deliberate. Numbers in a chapter like this function as targets for some readers, and diagnostic thresholds get read as entry requirements.

⚠️ If you are unwell right now, §34.14 tells you where to go, and it is fine to read that section first and the rest later or not at all.

⚠️ If reading about this is difficult today, skip to §34.14 and come back another time. Nothing in this book requires you to read this chapter in order.

The Hook: The chapter Chapter 33 could not write

⚠️ Chapter 33 ended by handing something over.

⚠️ It described a cycle — restriction, preoccupation, disinhibition, self-blame, tighter restriction — and said the mechanism was ordinary physiology in ordinary people. ⚠️ It then said that for some people the cycle stops being a description of ordinary eating and becomes something else, and that this chapter is where that happens.

⚠️ Here is what "something else" means, and it is not a matter of degree.

⚠️ In Chapter 33's framework, the eating was maintained by a cue or by an interval. Change the cue, fix the earlier meal, and the behaviour changes.

⚠️ In an eating disorder, the behaviour is maintained by the disorder.

⚠️ Remove the cue and it finds another. Fix the earlier meal and the rules absorb it. Change the environment and the illness adapts, because what is driving the behaviour is no longer the environment — it is a self-sustaining system involving beliefs about food, weight, control and self-worth that has acquired its own momentum.

⚠️ Which is why the tools in Chapter 33 do not merely fail here. Several of them make things worse.

⚠️ Three things this chapter is going to insist on, stated at the front because they are the ones most often got wrong:

⚠️ 1. Eating disorders are not diets that went too far. ⚠️ Dieting is a risk factor, and a substantial one. It is not the same thing, and the great majority of people who diet do not develop an eating disorder.

⚠️ 2. They are not about vanity, and they are not a choice. ⚠️ They are psychiatric illnesses with genetic, neurobiological, psychological and social contributions, and among the highest mortality of any psychiatric category.

⚠️ 3. You cannot tell who has one by looking. ⚠️ This single misconception causes more delay to more people than anything else in the chapter, and §34.4 is about it.

34.1 What the disorders are, in words

⚠️ Described in words rather than criteria, deliberately.

⚠️ Formal diagnostic manuals specify thresholds — of frequency, of duration, of weight. Those thresholds exist to make clinical research comparable and to allocate treatment. ⚠️ Reproduced in a textbook they do something else: they get read as entry requirements, and people conclude they are not ill enough to deserve help.

⚠️ So: no thresholds here. If you want the criteria, they are in DSM-5-TR and ICD-11, and they are for clinicians.

⚠️ ANOREXIA NERVOSA. Persistent restriction of intake, an intense fear of weight gain or persistent behaviour that interferes with gaining weight, and a disturbance in how body weight or shape is experienced. ⚠️ Weight may be low; it does not have to be — see §34.4, because this is the single most consequential thing in the chapter.

⚠️ BULIMIA NERVOSA. Recurrent episodes of eating experienced as out of control, followed by behaviours intended to compensate, with self-evaluation heavily determined by shape and weight. ⚠️ Weight is typically in the range considered unremarkable, which is a large part of why it goes unrecognized.

⚠️ BINGE EATING DISORDER. Recurrent episodes of eating experienced as out of control, with marked distress, and without regular compensatory behaviour. ⚠️ It is the most common eating disorder, it was not formally recognized until 2013, and it is still widely mistaken for a lack of self-control rather than a psychiatric condition.

⚠️ ARFID — AVOIDANT/RESTRICTIVE FOOD INTAKE DISORDER. ⚠️ Restriction driven by sensory characteristics, by fear of aversive consequences such as choking or vomiting, or by an apparent lack of interest in eating — WITHOUT the body-image component. ⚠️ It occurs across the lifespan, is substantially over-represented in autistic people, and is frequently dismissed as fussy eating for years.

⚠️ OSFED — OTHER SPECIFIED FEEDING OR EATING DISORDER. ⚠️ The category for presentations that are clinically serious and do not match the pattern of the others.

⚠️ OSFED is not a mild category and the name misleads badly. It accounts for a large share of presentations, and outcomes are comparable to the named disorders. ⚠️ "Not fitting the criteria" is a statement about the criteria, not about how ill someone is.

⚠️ Two more that matter:

⚠️ RUMINATION DISORDER and PICA are recognized feeding and eating disorders and are outside this chapter's scope.

⚠️ And ORTHOREXIA — preoccupation with the healthfulness rather than the quantity of food — ⚠️ is not a formal diagnosis in DSM-5-TR or ICD-11. ⚠️ It describes something real and recognizable that clinicians see, and §34.9 takes it seriously without pretending it has a status it does not have.

34.2 ⚠️ Where Chapter 33's cycle stops being sufficient

⚠️ The transition, described mechanistically, because it is the chapter's central explanatory job.

⚠️ Four things maintain the system once it is running, and none of them is willpower:

⚠️ 1. Starvation itself. ⚠️ The Minnesota findings from Chapter 33 §33.6 are the point: restriction produces rigidity, preoccupation, low mood and social withdrawal. ⚠️ Those symptoms then look like personality, and they are physiology. Some of what appears to be the illness is the malnutrition, and it improves with nutrition — which is why nutritional restoration comes early in treatment even when the psychological work has barely begun.

⚠️ 2. Relief. ⚠️ The behaviours reduce anxiety in the short term. A behaviour that reliably relieves distress is powerfully reinforced regardless of its long-term cost, and this is why "but you know it's harmful" changes nothing.

⚠️ 3. Identity and control. ⚠️ The illness frequently becomes a source of structure, achievement or identity, particularly where other domains feel uncontrollable. ⚠️ This is why recovery can be experienced as loss, and why "you must want to get better" is a sentence that misunderstands the condition.

⚠️ 4. Social reinforcement. ⚠️ Restrictive behaviour is frequently praised. Compliments about weight loss are received by people whose weight loss is a symptom, and they land as encouragement.

⚠️ That fourth point is the one bystanders have most control over, and §34.12 returns to it.

34.3 ⚠️ Who gets them

⚠️ The stereotype is a young, thin, white, affluent, female teenager.

⚠️ The stereotype is wrong in every one of those five terms, and the wrongness has consequences that are measurable in delayed diagnosis.

⚠️ And two groups with clearly elevated risk that the stereotype excludes entirely:

⚠️ LGBTQ+ people, ⚠️ and gay and bisexual men in particular, show substantially elevated rates.

⚠️ AUTISTIC PEOPLE, ⚠️ who are over-represented in both anorexia nervosa and ARFID — and for whom standard treatment approaches, which lean heavily on flexibility and social eating, frequently need adapting rather than repeating louder.

🔬 Claim → Evidence → Verdict

CLAIM: You can identify who has an eating disorder from their appearance, sex, age or background.

⚠️ EVIDENCE: Prevalence data across demographic groups; consistent findings of diagnostic delay in men, older adults, minoritized groups and people in larger bodies; and the fact that most people with eating disorders are not underweight.

⚠️ VERDICT: ❌ NOT SUPPORTED. ⚠️ And this is not a merely-inaccurate belief. It is the mechanism by which people are told they are not ill.

34.4 ⚠️ The most consequential misconception in this chapter

⚠️ It deserves its own section.

⚠️ MOST PEOPLE WITH EATING DISORDERS ARE NOT UNDERWEIGHT.

⚠️ Binge eating disorder, bulimia nervosa, OSFED and atypical anorexia nervosa — in which every feature of anorexia is present except a low weight — together account for the great majority of presentations.

⚠️ What follows from that, and each of these is documented:

⚠️ People in larger bodies with restrictive eating disorders are routinely praised for their symptoms. ⚠️ The weight loss produced by an illness is received as an achievement, sometimes by clinicians.

⚠️ They wait longer for diagnosis, are referred less often, and are more likely to be advised to lose weight while unwell.

⚠️ And the medical consequences of restriction do not require a low weight. ⚠️ Cardiac, electrolyte, endocrine, bone and gastrointestinal complications follow from the BEHAVIOUR and from the RATE and DEGREE of change, not from where someone's weight sits on a chart.

⚠️ Which is why the clinical guidance in several countries now states explicitly that weight and BMI should not be used as the sole determinant of whether someone gets treatment.

⚠️ It is also why this chapter contains no weights. A reader who finds a number here would use it to decide whether they qualify, and that decision is exactly the one being made wrongly.

⚠️ Atypical anorexia nervosa deserves its name examined. ⚠️ It is not atypical, it is not milder, and its medical complications and psychological severity are comparable. The name is an artefact of a diagnostic system built around weight.

34.5 What this does to a body

⚠️ Described at the level of systems, without figures, because the point is the breadth rather than the thresholds.

⚠️ And the point that matters most clinically, from Chapter 29 §29.7:

⚠️ REFEEDING SYNDROME. Reintroducing nutrition to someone who has been substantially undernourished causes rapid shifts in phosphate, potassium, magnesium and fluid that can be fatal.

⚠️ Which is why nutritional restoration in someone who is significantly unwell is a medical procedure with monitoring, and not something to attempt on the advice of a book.

⚠️ This is one of the two places in this entire textbook where following general nutrition advice could kill someone.

⚠️ The gastrointestinal row deserves a second look, because it explains something families find baffling. ⚠️ Prolonged restriction slows gastric emptying. Eating then produces genuine pain, bloating and early fullness. ⚠️ The person is not inventing it and it is not resistance — and it resolves with sustained nutrition, which means the period when eating is hardest is the period when it is most necessary.

34.5b ⚠️ What else is usually going on

⚠️ Eating disorders rarely arrive alone, and treating them as though they do is one reason treatment fails.

⚠️ And one combination that deserves naming on its own, because it is dangerous and under-recognized:

⚠️ TYPE 1 DIABETES AND EATING DISORDERS.

⚠️ Living with type 1 diabetes requires exactly the things this chapter has described as risk factors: continuous attention to food, weighing and counting, numerical targets, and a clinical relationship in which weight is routinely discussed.

⚠️ Rates of disordered eating are elevated in this group, and the omission or reduction of insulin as a weight-control behaviour occurs and is medically very serious. ⚠️ It carries substantially elevated risk of acute complications, of long-term microvascular damage, and of death.

⚠️ It is frequently missed, because unexplained variability in glucose control and recurrent hospital admissions get attributed to non-adherence rather than investigated as a symptom.

⚠️ Why this section is here rather than in a footnote:

⚠️ Two of the eight rows describe conditions — autism and type 1 diabetes — where the STANDARD management of the condition involves behaviours that overlap with disordered eating. ⚠️ In both, the person is doing what they were correctly instructed to do, and the point at which it becomes an illness is genuinely hard to see from outside.

⚠️ The general principle: the eating disorder is rarely the only thing, and it is frequently not the first thing. ⚠️ Assessment that stops at the eating misses what is holding it in place.

34.6 Athletes, and where Devi's thread lands

⚠️ Devi Raghunathan has been in this book since Chapter 4 — her low energy availability and depleted iron stores were established in Chapter 15, and Chapter 23 discussed the performance consequences. ⚠️ This is where that thread arrives, and it arrives here rather than in the sports chapter for a reason.

⚠️ Her actual figures are in Chapter 15 and are not repeated here. ⚠️ Energy-availability thresholds in particular function as targets — "am I under the number?" is the wrong question and it is the one people ask — so this section describes the syndrome and points you to a clinician for the arithmetic.

⚠️ RED-S — RELATIVE ENERGY DEFICIENCY IN SPORT — describes the consequences of insufficient energy availability relative to training load.

⚠️ It affects bone, endocrine function, immunity, cardiovascular health, iron status, mood and performance itself. ⚠️ It occurs in all sexes. It occurs in athletes at every level. And it does not require a low body weight — the deficit is relative to expenditure, which is why an athlete can develop it while eating what would be an entirely adequate amount for a non-athlete.

⚠️ The relationship to eating disorders is genuinely two-directional, and both directions are common:

⚠️ This is why Chapter 23 §23.13 refused to give body-composition targets, and it is worth connecting the two: sports environments select for exactly the traits — discipline, tolerance of discomfort, willingness to control intake — that also characterize restrictive illness.

⚠️ Two things specific to athletes that make detection harder:

⚠️ The behaviours are praised. ⚠️ Weighing food, training through fatigue and controlling intake are described as dedication in a context where they would be flags anywhere else.

⚠️ Performance can improve before it collapses. ⚠️ Short-term performance gains from weight loss are real and temporary, and they function as confirmation.

⚠️ Amenorrhoea in athletes is not normal, is not a sign of training hard enough, and is a symptom. ⚠️ Saying so is one of the more useful things a coach can do, and the belief that it is normal remains widespread.

34.7 ⚠️ How nutrition advice becomes a vector — including this book

⚠️ The uncomfortable section, and it is here because leaving it out would be dishonest.

⚠️ Nutrition information is not neutral for a person who is unwell or vulnerable to becoming so.

⚠️ This book has done some of these things, and I would rather name them than not.

⚠️ It has given numbers — protein per kilogram, fibre in grams, energy availability. It has used the word "should." It has recommended tracking in several project checkpoints.

⚠️ What I have tried to do about it, so you can judge whether it worked:

⚠️ No calorie targets anywhere in the book. No goal weights, no goal body compositions. Chapter 23 §23.13 refused body-composition targets explicitly. Chapter 33 §33.9 and §33.10 argued against forbidden-food framing. This chapter contains no figures at all.

⚠️ And the thing that cannot be engineered away: a book that says "here is what the evidence supports about food" will be read by some people as a book that says "here is a better set of rules," and no amount of framing prevents that.

⚠️ If this book has become a rule set for you — if you find yourself using it to justify narrowing what you eat, and the narrowing feels compulsory rather than chosen — that is worth taking seriously, and §34.14 is the section.

⚠️ Which is also the honest answer to a question this book has raised repeatedly.

⚠️ Nutrition literacy is protective on average. It is not protective for everyone, and for a subset of people more information about food is not what is needed. ⚠️ Chapter 38 will make a claim about what this book is for; this section is the caveat that claim has to survive.

34.8 ⚠️ The myths, taken one at a time

🔬 CLAIM: Eating disorders are a phase that people grow out of.

⚠️ EVIDENCE: Course is variable; a substantial proportion recover fully, a substantial proportion improve partially, and a proportion have a long-term course. Duration before treatment predicts outcome.

⚠️ VERDICT: ❌ NOT SUPPORTED — and the belief causes delay, which is the single modifiable factor most associated with worse outcome.

🔬 CLAIM: Eating disorders are about vanity or attention-seeking.

⚠️ EVIDENCE: Heritability estimates are substantial. Neurobiological and metabolic contributions are increasingly well characterized. Mortality is among the highest of any psychiatric illness, driven by medical complications and suicide.

⚠️ VERDICT: ❌ NOT SUPPORTED.

🔬 CLAIM: They only affect young women.

⚠️ VERDICT: ❌ NOT SUPPORTED (§34.3).

🔬 CLAIM: Dieting causes eating disorders.

⚠️ EVIDENCE: Dieting is one of the more consistently identified risk factors, particularly in adolescence, and most people who diet do not develop an eating disorder.

⚠️ VERDICT: 🟡 UNCLEAR / IT DEPENDS. ⚠️ A real risk factor and not a cause. Both halves matter — "dieting is harmless" and "dieting causes eating disorders" are both wrong.

🔬 CLAIM: Families, and mothers in particular, cause eating disorders.

⚠️ EVIDENCE: The historical model has not held up. Family-based treatment — which explicitly enlists parents as a resource rather than treating them as the problem — has the best evidence base of any treatment for adolescents.

⚠️ VERDICT: ❌ NOT SUPPORTED — and the belief actively harmed families for decades.

🔬 CLAIM: Recovery means never thinking about food again.

⚠️ EVIDENCE: Reported recovery experiences and long-term follow-up describe a wide range of outcomes, including full recovery. "Never thinking about food" is not what most people describe, and nor is it what most people without eating disorders experience.

⚠️ VERDICT: 🟠 PROBABLY FALSE, and setting it as the standard makes real recovery look like failure.

🔬 CLAIM: Someone has to hit rock bottom, or want to recover, before treatment can work.

⚠️ EVIDENCE: Ambivalence about recovery is a characteristic feature rather than an obstacle to treatment. Early intervention is associated with better outcomes. Family-based treatment for adolescents does not require the young person to be motivated at the outset.

⚠️ VERDICT: ❌ NOT SUPPORTED — and it is one of the more dangerous things a person can be told, because it is used to justify waiting.

34.9 ⚠️ Orthorexia, and the clean-eating on-ramp

⚠️ Chapter 20 §20.10 described a values position borrowing a scientific costume. This is what happens when that costume fits too well.

⚠️ Orthorexia describes preoccupation with the healthfulness of food rather than the quantity — an escalating set of rules about what is permissible, distress when they are broken, narrowing of the diet, and impairment of social and occupational life.

⚠️ It is not a formal diagnosis. Saying so is not dismissing it.

⚠️ Clinicians describe seeing it; the presentations frequently meet criteria for OSFED or ARFID; and the research literature has not established it as a distinct entity with reliable measurement.

⚠️ The honest verdict: ⚗️ the construct is under-tested as a distinct disorder, and the pattern it names is real and is frequently clinically significant under another label.

⚠️ Why it is in this chapter and not in Chapter 20:

⚠️ It is the one presentation that dietary advice — including good dietary advice — most directly supplies. ⚠️ Every other eating disorder has to work around the culture's approval. This one is congruent with it.

⚠️ The distinguishing questions are not about the content of the rules — they are about what the rules are doing:

⚠️ Is the diet narrowing over time, and does it ever widen? ⚠️ What happens — emotionally — when a rule is broken? ⚠️ Have social occasions become difficult or avoided? ⚠️ How much of the day does this occupy? ⚠️ Is the eating still flexible, or is it compulsory?

⚠️ A person eating a well-evidenced diet can answer those questions comfortably. That is the difference, and it is not visible on a plate.

34.10 Food insecurity, and a risk factor that gets missed

⚠️ Chapter 32 §32.10 noted that food insecurity and obesity co-occur, and Chapter 33 §33.6 noted that food insecurity runs the restriction cycle without anyone calling it a diet.

⚠️ The connection to this chapter is direct and under-recognized:

⚠️ Food insecurity is associated with elevated rates of binge eating and of eating-disorder symptoms generally. ⚠️ Involuntary restriction produces the same physiology as voluntary restriction, and the cycle does not distinguish between them.

⚠️ What this means practically:

⚠️ A person in a household that runs short every month is being subjected to a restriction cycle they did not choose. ⚠️ Treating that as a psychological problem, without addressing the access problem, is Chapter 32's error arriving in a clinic.

34.11 ⚠️ The language, and why it is not fussiness

⚠️ The vocabulary around eating disorders does real work, and most of it is doing the wrong work.

⚠️ The last row is worth pausing on because it has become ambient.

⚠️ "Earning" food and "burning off" a meal are the exact cognitive structure of a compensatory behaviour, spoken casually by people who do not have an eating disorder, in front of people who do.

⚠️ And on the compliment problem, which is the most common accidental harm a bystander does:

⚠️ A person losing weight because they are unwell will be complimented on it, repeatedly, by people who mean well. ⚠️ Each compliment is reinforcement, and it arrives from someone whose opinion the person values.

⚠️ The practical rule: do not comment on anyone's body, ever, including favourably. ⚠️ It costs nothing, and you do not know who you are talking to.

34.12 What treatment actually is, and what helps

⚠️ Described honestly rather than optimistically, because false optimism is its own harm.

⚠️ Three honest statements about outcomes:

⚠️ Full recovery happens, and it happens frequently enough that it should be the expectation.

⚠️ Recovery is typically long, non-linear, and includes periods of deterioration that are part of the course rather than evidence of failure.

⚠️ And duration of untreated illness is one of the strongest modifiable predictors of outcome — ⚠️ which is the single most important practical fact in this section, because it means the belief in §34.8 that someone must want it first is not merely wrong but costly.

⚠️ What family and friends can do — and what they cannot:

⚠️ The most useful sentence available to a bystander is a version of: "I've noticed X. I'm not going to make you talk about it. I'd like to help you see someone, and I'll come with you."

⚠️ Note what it does not contain: a description of a body, a number, a demand, or a deadline.

34.13 ⚠️ Who this chapter is for — and where it stops

⚠️ That last row is the one that matters most, so it gets a plain answer.

⚠️ There is no threshold you have to reach. If your relationship with food is causing you distress or taking up your life, that is sufficient reason to talk to someone, and it does not require a diagnosis, a number, or anyone's permission.

⚠️ And where a nutrition textbook stops, stated plainly:

⚠️ This chapter can describe. It cannot assess, diagnose, treat or monitor. ⚠️ Every one of those requires a person who can see you, and several of them require someone who can take blood.

⚠️ The most useful thing a nutrition book can do here is tell you accurately what these illnesses are, dismantle the beliefs that delay help, and then get out of the way.

🧾 What delay costs

⚠️ Part VI costs everything. This one costs the thing that is actually expensive, and it is not food.

⚠️ There is no purchase in this table. That is the point of including it.

34.14 ⚠️ Where to get help

⚠️ Read this section even if you skipped the rest.

⚠️ If you are worried about yourself or someone else, you do not need to be certain, and you do not need to meet a threshold.

⚠️ Three things worth knowing before you go:

⚠️ You are allowed to ask for a second opinion if you are told you are not ill enough. ⚠️ That sentence is common, it is frequently wrong, and §34.4 is why.

⚠️ You can bring someone with you, and it helps.

⚠️ And you can go on behalf of someone else to ask what to do. ⚠️ You do not need their permission to seek advice about how to help them.

⚠️ Avoid: pro-eating-disorder communities, "what I eat in a day" content, calorie-tracking apps while unwell, and anything that supplies numbers. ⚠️ These are documented harms and they are extremely easy to find.

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

⚠️ What this field is bad at:

⚠️ Treatment research in eating disorders is harder than in almost any other area — recruitment is difficult, dropout is high, randomization to no-treatment is not ethical, and follow-up periods are rarely long enough for illnesses that run for years. ⚠️ The evidence base for adults with anorexia nervosa in particular is thinner than anyone would like.

⚠️ And research has historically over-studied young white women, which is part of why §34.3's corrections took as long as they did to arrive.

Spaced Review

From Chapter 33 §33.6: ⚠️ What did the Minnesota findings show, and how does §34.2 use them to explain why some of what looks like personality is physiology?

From Chapter 29 §29.7: ⚠️ What is refeeding syndrome, and why does it make nutritional restoration a medical procedure?

From Chapter 20 §20.10: ⚠️ What is a values position in a scientific costume, and how does §34.9 extend it?

From Chapter 32 §32.10: ⚠️ Why is food insecurity a risk factor here?

From Chapter 15 and 23: ⚠️ What is energy availability, and why did Chapter 23 §23.13 refuse to give body-composition targets?

From Chapter 25 §25.10b: ⚠️ Why does bone loss in adolescence matter for decades?

Project Checkpoint: A Different Kind of Checkpoint

⚠️ This checkpoint asks you to examine your own framework rather than to record anything about your eating. ⚠️ There is deliberately nothing to log, count or measure here.

Step 1 — ⚠️ Audit your vocabulary. For one week, notice §34.11's phrases in what you say and what you hear.

⚠️ How many times did you hear "cheat meal," "earning it," "being good," "guilt-free," or a comment on someone's body? __

⚠️ How many of those were yours? __

Step 2 — ⚠️ Retire two phrases. Pick the two you use most and stop.

Step 3 — ⚠️ Check your framework against §34.9's questions — not against your food.

⚠️ Is what I eat narrowing over time, and does it ever widen? ⚠️ What happens emotionally when I break one of my own rules? ⚠️ Have social occasions become harder? ⚠️ How much of my day does this occupy? ⚠️ Is my eating flexible, or is it compulsory?

⚠️ These are not diagnostic. They are the questions that distinguish a considered diet from a rule set, and if the answers concern you, §34.14 is the section.

Step 4 — ⚠️ Write the sentence. If you were worried about someone, what would you actually say?

⚠️ Draft it. Then check it contains no description of a body, no number, no demand and no deadline (§34.12).

Step 5 — ⚠️ Find your local service now, before you need it.

⚠️ Your country's national eating disorders organization: __

⚠️ Five minutes, once. The moment you need it is not the moment to start searching.

Chapter Summary

⚠️ Chapter 33's cycle is maintained by the environment. An eating disorder is maintained by itself — ⚠️ remove the cue and it finds another route, because what drives it is a self-sustaining system of beliefs about food, weight, control and self-worth.

⚠️ They are not diets that went too far, not vanity, and not a choice. ⚠️ Dieting is a real risk factor and not a cause. Both halves matter.

⚠️ §34.2: four things maintain the system — starvation itself (some of what looks like personality is physiology), short-term relief, identity and control, and social praise. ⚠️ The fourth is the one bystanders control.

⚠️ §34.3–34.4: the stereotype is wrong in all five of its terms, and MOST PEOPLE WITH EATING DISORDERS ARE NOT UNDERWEIGHT. ⚠️ People in larger bodies are praised for their symptoms, wait longer, and are sometimes advised to lose weight while unwell. ⚠️ Medical complications follow from the behaviour, not from a position on a chart — which is why this chapter contains no numbers at all.

⚠️ §34.5: consequences are systemic, and refeeding is a medical procedure (Ch 29 §29.7). ⚠️ Prolonged restriction makes eating physically painful, which means the period when eating is hardest is the period when it is most necessary.

⚠️ §34.6: RED-S runs in both directions, occurs in all sexes at every level, and does not require a low weight. ⚠️ Sports environments select for the traits restrictive illness runs on, the behaviours get called dedication, and performance can improve before it collapses.

⚠️ §34.7: nutrition advice — including this book — can be a vector. ⚠️ Named honestly rather than omitted.

⚠️ §34.8: ❌ a phase · ❌ vanity · ❌ only young women · ❌ families cause them · ❌ they must want it first · 🟠 recovery means never thinking about food · 🟡 dieting causes them.

⚠️ §34.9: orthorexia is ⚗️ under-tested as a distinct entity and names something real — and it is the one presentation the culture actively supplies. ⚠️ The distinguishing questions are about what the rules DO, not what they contain.

⚠️ §34.11: do not comment on anyone's body, ever, including favourably. ⚠️ And retire "earning" food and "burning it off" — that is the cognitive structure of a compensatory behaviour, spoken casually in front of people who have one.

⚠️ §34.12: full recovery happens and should be the expectation; recovery is long and non-linear; and duration of untreated illness is the strongest modifiable predictor of outcome.

⚠️ Which is why "they have to want it first" is not merely wrong. It is costly.

⚠️ §34.14: there is no threshold you have to reach.

What's Next

⚠️ Chapters 33 and 34 have been about the person doing the eating. Chapter 35 turns to a promise made about them.

⚠️ Personalized nutrition — genetic tests, microbiome tests, continuous glucose monitors, and the claim that the right diet for you is discoverable and different from everybody else's.

⚠️ It is the fastest-growing commercial sector in nutrition, it contains one genuinely impressive finding, and most of what is sold on the back of that finding does not do what it says.

⚠️ And it arrives immediately after this chapter for a reason worth noticing: an industry selling individually optimized rule sets, to people encouraged to monitor themselves continuously, is not a neutral development for everyone reading it.