Chapter 29 — ★ Clinical Nutrition: When Food Becomes Treatment, and When Feeding Can Kill

The Hook: Nine days

Ivor Petrakis is 80 now.

Chapter 25 left him doing well — 78 kg, protein at 85 g a day, a lunch club twice a week, and able to stand up from a chair without using his arms.

In March he slipped on a wet pavement and fractured his hip.

⚠️ What follows is his oral intake over the first nine days of that admission, reconstructed from the notes.

⚠️ Nine days. Roughly 3,150 kcal total. About 350 a day, against a requirement in the region of 1,900.

⚠️ He lost 5 kg. He could no longer stand from a chair unaided. He was discharged to a rehabilitation bed instead of home.

⚠️ And here is the thing that makes this a chapter rather than a scandal: not one of those decisions was wrong.

Fasting for theatre is correct. Bumping a list happens. Nausea after anaesthesia is expected. X-ray had to be done. The possible procedure had to be planned for.

⚠️ Nobody decided to starve Ivor Petrakis. Nine individually reasonable decisions did it, and no single person was in a position to see the total.

That is what clinical malnutrition looks like, and it is why this chapter exists.

Part V has been about nutrition as prevention. ⚠️ This chapter is about nutrition as treatment — and about the fact that in this setting it can be got wrong in both directions.

⚠️ The through-line: this is where Chapters 24, 25 and 26's inversions all arrive at once, and where "eat less" becomes the most dangerous sentence in the book.

And there is one place — §29.8 — where feeding someone too enthusiastically will kill them.

🏃 Fast Track: §29.1–29.4 (the problem and the cascade), §29.8 (⚠️ refeeding), §29.12 (⚠️ the inversions). Forty minutes.

🔬 Deep Dive: §29.9 (⚠️ critical illness, where the trials overturned the intuitions), §29.14 (⚠️ end of life).

29.1 The scale of it

⚠️ Hospital malnutrition is common, consequential, and largely invisible.

Prevalence estimates on admission vary by setting and tool but commonly land somewhere between a fifth and a half of patients — highest in older adults, oncology, gastrointestinal disease and critical care.

What it is associated with:

⚠️ And it is frequently undetected. Weight is often not measured on admission. Screening tools exist, are mandated in many systems, and completion rates are poor.

⚠️ The uncomfortable framing: a hospital is a place people go to get better, and a substantial proportion of them leave nutritionally worse than they arrived.

⚠️ On definitions: the GLIM criteria — a 2018 global consensus — set out a two-step approach: screen first, then diagnose using at least one phenotypic criterion (unintentional weight loss, low BMI, reduced muscle mass) plus at least one aetiologic criterion (reduced intake or assimilation, or disease burden and inflammation). ⚠️ The muscle mass criterion matters — Chapter 25's sarcopenia, arriving in a hospital bed.

29.2 Screening: the tools, and why it fails

Validated, quick, free.

⚠️ Each takes minutes. So why is it not done?

1. ⚠️ Weight isn't measured. Scales are absent, broken, or the patient can't stand. ⚠️ You cannot screen without a weight, and this is the commonest failure point.

2. It's a form, and forms compete. Screening is one of many admission tasks and it is rarely the one anyone is asked about.

3. ⚠️ Nobody owns the result. A completed MUST score with no triggered action is worse than useless — it documents that someone noticed.

4. ⚠️ And nutrition has no monitor. Fluid balance is charted, oxygen saturation alarms, blood pressure triggers a review. ⚠️ Nobody's intake sets off anything, and a food chart is only as good as the person who fills it in.

⚠️ Ivor was screened on admission. He scored medium risk. Nothing was triggered by it.

🪞 Learning Check-In: predict the trial results

Stop here. §29.9 is coming and I want your predictions first.

⚠️ Here is the clinical situation. A patient is admitted to intensive care — septic, ventilated, requiring vasopressors. They cannot eat. They will be there for a week or more, and they are catabolic: breaking down muscle at a rate you cannot match by feeding.

⚠️ The intuition is obvious: feed them, promptly, to full estimated requirements, to limit the losses.

That intuition has been tested repeatedly in large randomized trials. Write your predictions before reading on.

1. ⚠️ Does starting full-dose nutrition EARLY improve survival compared with starting lower and building up?

2. Does delivering MORE calories in the first week improve outcomes?

3. ⚠️ Does delivering MORE protein improve outcomes?

4. If any of your answers is "no" — ⚠️ propose a mechanism. Why would feeding a starving, catabolic patient less be better?

5. And the honest one: ⚠️ how confident are you? This is a case where the physiological reasoning is clean, the intuition is strong, and the trials are large.

Keep your answers. §29.9 asks you to compare.

29.3 The nil-by-mouth cascade

⚠️ Ivor's table, generalized — because these are the mechanisms, and every one is fixable.

⚠️ What actually works, and none of it is complicated:

⚠️ Protected mealtimes — non-urgent activity stops · red tray or equivalent systems flagging who needs help · food charts that someone reads · ⚠️ re-feeding immediately when a procedure is cancelled · opening the packaging · getting the dentures · treating the nausea · ⚠️ and someone whose job it is to look at the total.

⚠️ The single highest-yield intervention in this section is that when a procedure is cancelled, the patient eats. It requires no equipment, no budget and no training, and it is the one most often missed.

29.3b ⚠️ The other half of the problem: lying still

Because Ivor did not only stop eating. He stopped moving, and the two compound.

⚠️ Bed rest causes rapid loss of muscle mass and function, and the rate is faster than most people expect — studies of healthy volunteers in enforced bed rest show measurable loss of leg lean mass and strength within DAYS, and older adults lose faster than younger ones.

⚠️ Add acute illness and the rate accelerates further, because inflammation drives catabolism.

💡 ⚠️ This is why nutrition and mobilization are the same intervention, and why treating them separately fails.

Protein without loading does much less (Chapter 25 §25.13). ⚠️ And loading without protein does less than it could. A patient who is fed adequately and left in bed still loses muscle; a patient who is mobilized and underfed has nothing to build with.

⚠️ Ivor got neither for the first six days.

⚠️ Which makes the practical instruction unusually simple and unusually neglected: get people out of bed, and feed them. Early mobilization after hip fracture is standard practice precisely because of this, and it is frequently delayed for reasons — pain, dizziness, staffing — that are individually reasonable and cumulatively identical to §29.3's cascade.

29.4 The routes, in order

⚠️ THE GOVERNING PRINCIPLE: if the gut works, use it. And the least invasive adequate route wins.

On oral nutritional supplements: 🟢 to ✅ — in malnourished patients, particularly older adults, ONS have shown reduced complications and, in some analyses, reduced mortality. ⚠️ They work best as an ADDITION to food, taken between meals rather than instead of them — and "sip feeds left on the locker, untouched, for three days" is the standard failure mode.

⚠️ And Chapter 25 §25.14's point holds here: home-style food fortification — milk powder, cheese, cream, oil, nut butters — is cheaper, better accepted, and frequently as effective.

29.5 Enteral nutrition

A tube into the stomach or small bowel: nasogastric or nasojejunal for shorter term, gastrostomy (PEG) or jejunostomy for longer.

⚠️ The safety point that matters most:

⚠️ NASOGASTRIC TUBE MISPLACEMENT INTO THE LUNG IS A RECOGNIZED NEVER EVENT AND HAS KILLED PEOPLE.

Position must be confirmed before use — aspirate pH testing (a pH of 5.5 or below is generally accepted) and/or radiographic confirmation, interpreted by someone competent to do so. ⚠️ The whoosh test and bubbling in water are not valid checks and were abandoned for good reason.

Other complications: aspiration · tube blockage and displacement · site infection · refeeding (§29.8) · and diarrhoea, which is ⚠️ frequently attributed to the feed and is more often caused by medications — sorbitol-containing liquid preparations, magnesium, antibiotics — or by C. difficile (Chapter 27).

⚠️ How it is actually delivered, because the practicalities determine whether it works:

⚠️ And enteral feeding is not confined to hospital. Home enteral nutrition supports a substantial population — people with neurological conditions, head and neck cancer, and children with complex needs — living ordinary lives with a tube. It requires training, community dietetic support and equipment supply, and it is one of the clearer successes of clinical nutrition.

⚠️ The commonest practical failures are mundane: blocked tubes from inadequate flushing, feeds held for hours around drug administration and never made up, and — the recurring theme — nobody totalling what was actually delivered against what was prescribed.

⚠️ PEG feeding in advanced dementia

A case where the evidence contradicts the intuition strongly enough that most professional bodies now advise against it.

⚠️ In advanced dementia, PEG feeding has NOT been shown to prolong survival, prevent aspiration pneumonia, improve pressure ulcer healing, improve function, or improve comfort.

⚠️ Aspiration continues, because the aspiration is largely of saliva and refluxed gastric contents, not of food. And the tube carries its own burdens: procedural risk, restraint to prevent pulling it out, and loss of the pleasure and human contact of being fed.

⚠️ The recommended alternative is careful hand feeding — assisted, unhurried, with modified textures where indicated, accepting that intake will be limited.

⚠️ This is a difficult conversation with families, and §29.14 is about why.

29.6 Parenteral nutrition

Nutrition delivered intravenously, bypassing the gut entirely.

⚠️ Indications are narrow: a gut that does not work or cannot be reached — short bowel syndrome, bowel obstruction, high-output fistula, severe prolonged ileus, and some perioperative and oncology situations.

⚠️ It is not a nutritional upgrade. It is a rescue route with substantial risk:

Catheter-related bloodstream infection — ⚠️ the major one, and a leading cause of avoidable harm · venous thrombosis · ⚠️ intestinal failure-associated liver disease with prolonged use · metabolic derangement · refeeding syndrome (§29.8) · and cost, plus the need for a specialist nutrition support team.

⚠️ Parenteral nutrition was historically over-used, on the reasoning that if some nutrition is good, guaranteed delivery must be better. The trials in §29.9 are a substantial part of why that changed.

29.7 ⚠️ Refeeding syndrome

The section where feeding kills.

The mechanism

In prolonged starvation, the body shifts to catabolism. ⚠️ Intracellular stores of phosphate, potassium and magnesium are depleted — but SERUM levels can remain normal, because the body maintains them at the expense of the cells. Thiamine stores deplete within weeks.

⚠️ Then you feed carbohydrate.

⚠️ The consequences: cardiac arrhythmia, cardiac failure, respiratory failure, rhabdomyolysis, seizures, Wernicke's encephalopathy, and death.

⚠️ Hypophosphataemia is the hallmark, and it can develop within 24–72 hours of starting feeding.

⚠️ The historical origin is worth knowing: it was described in liberated prisoners of war and famine survivors who died after being fed. People survived starvation and were killed by food.

⚠️ Who is at risk

Criteria of the kind used in national guidance:

HIGH RISK if ONE or more of: BMI under about 16 · unintentional weight loss over ~15% in the past 3–6 months · little or no intake for more than ~10 days · ⚠️ low pre-feeding potassium, phosphate or magnesium

OR TWO or more of: BMI under about 18.5 · weight loss over ~10% in 3–6 months · little or no intake for more than ~5 days · ⚠️ a history of alcohol misuse, or of drugs including insulin, chemotherapy, antacids or diuretics

⚠️ Populations to think of: anorexia nervosa (Chapter 34) · alcohol use disorder · prolonged fasting or food insecurity · post-bariatric surgery · advanced cancer · older adults after a long period of poor intake · ⚠️ and anyone whose oral intake has been minimal in hospital for a week, which brings us back to Ivor.

⚠️ Management

⚠️ THIAMINE BEFORE FEEDING. Not with it. Before it. Plus a B-vitamin preparation, continued during the early days.

⚠️ START LOW, GO SLOW. In the highest-risk patients, guidance commonly begins around 5–10 kcal/kg/day, increasing gradually over several days to a week.

⚠️ Check phosphate, potassium and magnesium before feeding and daily thereafter. Replace aggressively. Do NOT wait for normal levels before starting — start low and replace alongside.

Monitor fluid balance and cardiac status; consider cardiac monitoring in severe cases.

⚠️ The instinct — "this person is starving, feed them properly" — is exactly the instinct that causes the harm. This is the single clearest example in the book of a correct-sounding intervention being dangerous, and it is why "eat more" needs a rate attached.

29.8 ⚠️ Critical illness: where the intuitions were wrong

Now compare your Learning Check-In predictions.

The intuition: a septic, ventilated, catabolic patient is losing muscle rapidly, so feed them fully and promptly to limit the damage.

⚠️ Large randomized trials have repeatedly failed to support it.

⚠️ Proposed explanations, none settled:

1. ⚠️ Endogenous energy production. A critically ill patient is generating substantial energy from their own tissue. Feeding on top of that produces overfeeding, even when the delivered amount looks like a deficit.

2. ⚠️ Suppression of autophagy. Feeding — particularly protein and amino acids — suppresses autophagy (Chapter 21 §21.6), which may be needed for cellular clearance and repair during acute illness. Which is a striking place for that mechanism to reappear, and a much better-supported use of it than the fasting literature made.

3. Hyperglycaemia and metabolic stress.

⚠️ The current consensus, roughly: start enteral nutrition early but at a LOW rate, advance gradually across the first week, do not chase full estimated targets during the acute phase, and avoid overfeeding.

⚠️ And note what this is NOT: it is not an argument for starving critically ill patients, and it does not extend beyond the acute phase.

⚠️ The part that gets forgotten: the recovery phase. Patients leave intensive care profoundly catabolic, with ICU-acquired weakness, and requirements that are high and sustained for months. ⚠️ That is where nutrition matters most and where it is delivered least — because by then they have left the unit that was counting.

💡 ⚠️ Compare your predictions.

Most people — including most clinicians before the trials — predict that early full feeding helps. If you did, you were reasoning correctly from clean physiology, and the trials said otherwise.

⚠️ Which is Chapter 19 §19.5 and Chapter 13's antioxidants, arriving in intensive care: a mechanism that is real, an inference that is reasonable, and an outcome that had to be measured.

29.9 Perioperative nutrition

Where the evidence has changed practice, and where it hasn't yet.

⚠️ Preoperative fasting. "Nil by mouth from midnight" persists widely and is not supported. Modern anaesthetic guidance permits clear fluids up to about two hours and a light meal up to about six hours before. ⚠️ Prolonged fasting produces discomfort, dehydration, insulin resistance and misery, and confers no additional safety.

⚠️ Preoperative carbohydrate loading — a clear carbohydrate drink up to two hours before surgery — is a component of ERAS (Enhanced Recovery After Surgery) protocols. 🟢 It reduces postoperative insulin resistance and improves patient-reported wellbeing; effects on hard outcomes are more modest.

⚠️ Early postoperative feeding. Waiting for bowel sounds or flatus before feeding is largely obsolete. 🟢 Early oral intake after most abdominal surgery is safe and associated with faster recovery.

⚠️ Prehabilitation — nutrition plus exercise plus psychological preparation in the weeks before major surgery. 🟢 Growing evidence, particularly in colorectal and major abdominal surgery, for improved functional recovery and reduced complications. ⚠️ The window between diagnosis and surgery is usually treated as dead time and is a genuine opportunity.

Immunonutrition — formulations containing arginine, omega-3 and nucleotides. 🟡 ⚠️ Mixed evidence, and a good example of a plausible mechanism outrunning its outcome data.

And ✅ the strongest one: ⚠️ preoperative nutritional optimization in malnourished patients reduces postoperative complications. Which means the malnourished patient should be identified in clinic, not on the ward.

29.10 Drug–nutrient interactions

Chapter 16 §16.6 covered supplements. ⚠️ Here is the clinical set.

⚠️ The warfarin row is the one most often got wrong in both directions. Patients are told to avoid green vegetables; the correct advice is to eat a consistent amount. Avoidance costs them folate, fibre and vitamin K stability.

29.11 ⚠️ The inversions, all at once

This is what I promised at the end of Chapter 28, and it is the chapter's structural point.

⚠️ The general rule that emerges, and it is worth carrying out of Part V:

Almost all population dietary advice is calibrated for a person whose problem is EXCESS. A substantial number of people in a hospital have the opposite problem, and the advice inverts completely.

⚠️ Applying Chapter 22's recommendations to Ivor Petrakis would have killed him more slowly than nine days of nil by mouth, and just as certainly.

29.12 ⚠️ The restriction pattern, and the general rule

Chapter 26 §26.11b named it in cancer and kidney disease. Chapter 27 found it in IBS. Chapter 28 found it in food intolerance testing. ⚠️ Here is the fifth instance and the general rule.

In hospital, restriction arrives as a diet order. ⚠️ "Diabetic diet." "Low-salt diet." "Renal diet." "Soft diet." Each is reasonable in principle and each reduces intake.

⚠️ In an acutely unwell inpatient with poor oral intake, a restrictive therapeutic diet is frequently the wrong trade — and a "diabetic diet" that halves someone's intake to improve a glucose reading is a bad exchange, which is why most modern guidance de-emphasizes restrictive inpatient diets in favour of adjusting medication.

⚠️ THE GENERAL RULE, across all five instances:

A person who is unwell, motivated, and looking for something to control is unusually likely to be handed a restriction — by a clinician, a test, a protocol, or themselves.

⚠️ The question is almost never "what should be removed?" It is "are they getting enough, and of what?" — and any restriction should be specific, time-limited, and justified against what it costs.

29.12b Who is actually responsible

⚠️ The structural question underneath everything above, and it has an unsatisfying answer.

In most hospitals, nutrition is everybody's job and nobody's specialty.

⚠️ The gap is ownership of the TOTAL. Every professional above is doing their job correctly. Nobody's job is to notice that a patient has eaten 350 kcal a day for nine days — which is exactly the failure §29.1 described, expressed as an organizational chart.

⚠️ What demonstrably helps: nutrition support teams · screening with an automatic referral trigger rather than a score in a box · protected mealtimes · ⚠️ and putting nutrition on the ward round, which costs nothing and changes what gets noticed.

⚠️ And a word about hospital food itself, which is the standard target and is only part of the problem. It is frequently criticized and frequently the wrong thing to criticize — ⚠️ because a meal that is nutritionally excellent, delivered on time, and removed untouched because the patient was at X-ray, has failed for reasons that have nothing to do with the menu. Improving the food without fixing §29.3's cascade improves very little.

⚠️ What a visitor can do, which is more than it sounds

Because most readers of this book will be relatives before they are clinicians, and the visitor is frequently the only person present for a whole meal.

⚠️ None of this requires permission, expertise or confrontation, and collectively it is a larger intervention than anything else available to a non-clinician.

The visitor is the only person in the building whose job is one patient.

29.13 ⚠️ When feeding should stop

The hardest section, and it belongs in a nutrition book because families ask nutrition questions.

⚠️ In the last days and weeks of life, artificial nutrition generally does not prolong life or improve comfort, and it can cause harm — fluid overload, increased respiratory secretions, oedema, nausea, and aspiration.

⚠️ The distinction that matters, and it needs saying plainly:

A dying person is not dying because they have stopped eating. They have stopped eating because they are dying.

⚠️ Reversing the causality — which is intuitive and almost universal among families — leads to requests for feeding that cannot help and may hurt.

What is appropriate: ⚠️ comfort feeding — offering small amounts of whatever is wanted, for pleasure, without targets. Mouth care, which matters enormously for comfort and is frequently the actual answer to "they're thirsty." And sips of anything they like.

⚠️ And the family's distress deserves to be taken seriously rather than managed around. Feeding someone is what you do when you love them. Being told to stop feels like being told to stop caring, and a clinician who does not address that directly has not finished the conversation.

⚠️ What usually helps: naming that food is love, giving them something to do — mouth care, moistening lips, offering tastes — and being explicit that stopping artificial feeding is not withdrawal of care.

29.14 Who this chapter is about

⚠️ How firmly I hold these

🧾 What it costs

⚠️ This chapter's economics run backwards from the rest of the book, and the inversion is instructive.

⚠️ The pattern that has held for thirteen chapters — the best-evidenced option is the cheapest — holds here too, and more starkly.

The interventions with the strongest evidence in this chapter are: weigh people, feed them when a procedure is cancelled, open the packaging, get them out of bed, and give thiamine before feeding. ⚠️ All of them are free or nearly free.

And the expensive end — parenteral nutrition, prolonged admission, rehabilitation placement — is substantially downstream of the free end not being done.

What we don't know

⚠️ The optimal protein dose and timing in critical illness — genuinely unresolved and currently contested. Whether prehabilitation improves hard outcomes or mainly functional recovery. How to deliver adequate nutrition in the post-ICU recovery phase, which is where it probably matters most. ⚠️ And whether any of the screening tools change outcomes rather than merely identifying risk — which is an uncomfortable question that has been asked and not cleanly answered.

29.15 What to actually do

⚠️ If you are a patient, or a relative:

1. ⚠️ Ask what they have actually eaten today. Not "are you eating?" — what, and how much. 2. ⚠️ Open the packaging. Cut the food up. Get the dentures. Sit them up. 3. ⚠️ If a procedure is cancelled, ask for food immediately. 4. Ask whether they have been weighed and screened, and what was done about it. 5. ⚠️ Bring food they like, if allowed.

⚠️ If you work in healthcare:

6. ⚠️ Weigh people. It is the commonest failure point. 7. ⚠️ Assess refeeding risk BEFORE feeding anyone who has eaten little for days. Thiamine first. 8. ⚠️ Question restrictive inpatient diets in someone who isn't eating. 9. ⚠️ Re-feed after cancelled procedures. 10. And look at the total — ⚠️ because nine reasonable decisions produced Ivor's nine days, and nobody was looking at the sum.

What happened to Ivor

He spent five weeks in a rehabilitation bed. ⚠️ He went home.

What made that possible was Chapter 25's work: he arrived at the fracture with adequate protein intake, a resistance training habit, corrected B12 and vitamin D, and a body that had six kilos of margin. ⚠️ The version of Ivor from Chapter 25's opening — 72 kg, protein at 0.5 g/kg, unable to rise from a chair — would probably not have gone home.

⚠️ Which is the argument for everything in Chapter 25, stated in the only way that ever really persuades anyone: what happens when something goes wrong.

Nutritional status is not a number on a chart. It is the margin you have when you fall.

⚠️ And what he said at his follow-up, which is the sentence this chapter is built around:

"Nobody was unkind. Not one person. I just didn't eat for a week and it took five weeks to get back."

Spaced Review

1. (Chapter 21) Autophagy reappears in §29.8. ⚠️ How is its use here different from Chapter 21's?

⚠️ In Chapter 21 it was a mechanism with no established human dose-response, used to justify a protocol (⚗️). Here it is one proposed explanation for a finding that has already been demonstrated in large randomized trials — the mechanism is being invoked to explain an OUTCOME, not to predict one. ⚠️ Same biology, opposite epistemic position.

2. (Chapter 25) Why is Ivor's 5 kg loss over nine days worse than it sounds?

⚠️ Because in an 80-year-old a large fraction of rapid loss is lean tissue (Chapter 25 §25.13), and anabolic resistance means it is difficult to regain. The functional consequence was immediate — he could no longer stand from a chair unaided — and that determined his discharge destination.

3. (Chapter 26) §29.12 says a "diabetic diet" in a poorly-eating inpatient is often the wrong trade. Why?

⚠️ Because the restriction reduces intake in someone who is already not eating enough, to improve a glucose reading that could be managed by adjusting medication instead. Chapter 26's tiers apply: glycaemia is highly responsive and rapidly correctable by other means; lean mass loss is not.

Project Checkpoint: The Clinical Picture

Component twenty-nine. ⚠️ This one is mostly about someone else.

Step 1 — ⚠️ If you or someone you know is in hospital, or heading there, ask these five:

What have they eaten today — actually? _ Have they been weighed? ⚠️ Have they been fasted for a procedure that didn't happen? _ ⚠️ Can they open the packaging and reach the tray? Do they have their dentures? ____

Step 2 — ⚠️ The refeeding screen, which takes thirty seconds and is the highest-stakes item in this book:

• [ ] BMI under ~16, or under ~18.5 with other risk
• [ ] Weight loss over ~10–15% in 3–6 months
• [ ] ⚠️ Little or no intake for more than 5–10 days
• [ ] History of alcohol misuse
• [ ] ⚠️ Low potassium, phosphate or magnesium before feeding

⚠️ Any of these and feeding needs to start LOW, with THIAMINE FIRST, and with daily electrolytes.

Step 3 — Your own medication–nutrient list (§29.10, Chapter 16):

⚠️ Warfarin: consistency, not avoidance. Levothyroxine: separate from calcium, iron and coffee.

Step 4 — ⚠️ And the conversation worth having before it's needed:

If someone you love reaches the point where they cannot eat — do you know what they would want?

⚠️ §29.13's distinction is much easier to hold in advance than at the bedside.

Next checkpoint (Chapter 30): reading a label properly — Part VI turns practical.

29.16 ⚠️ Part V verdict re-audit

Promised at Chapter 22 and Chapter 24, and this is the point.

⚠️ The affirmative share has nearly tripled. The question the audit exists to answer is whether that is honest.

⚠️ I think it is, and here is the argument:

1. ⚠️ The questions changed, not the standard. Part IV adjudicated contested public arguments — sugar, seed oils, organic, fasting. Those are contested precisely BECAUSE the evidence is ambiguous; a chapter about a settled question doesn't generate an argument. Part V asked applied clinical questions, and applied clinical questions have better evidence.

2. ⚠️ Part V's ✅s are mostly not nutrition-science claims at all. Adrenaline for anaphylaxis. Coeliac testing requiring gluten exposure. Resistance training against sarcopenia. NG tube position checks. Folic acid preconception. These come from clinical medicine, where randomized trials and mechanistic certainty are far more available than in dietary epidemiology.

3. ⚠️ And the 🟡 count kept rising too — from 20 to 60. If the standard had loosened, uncertainty would have shrunk. It didn't; both ends grew, because there were simply more questions.

⚠️ The honest caution I'd attach: a reader who arrives at Chapter 29 and concludes "nutrition science is actually quite settled" has drawn the wrong lesson from a chapter about intensive care. Part VI and Part VII return to questions where the answer is "it depends," and ⚠️ Part IV's 🟡 column — thirteen entries on the questions people argue about most — remains the honest picture of the field's core.

Chapter Summary

⚠️ Nine individually reasonable decisions produced 350 kcal a day for nine days, 5 kg of loss, the ability to stand lost, and a rehabilitation bed instead of home. Nobody decided to starve Ivor Petrakis.

⚠️ §29.3b's pairing is the practical core: nutrition and mobilization are the SAME intervention. Protein without loading does much less; loading without protein has nothing to build with. Ivor got neither for six days.

⚠️ §29.11's inversion table is Part V's structural conclusion: almost all population dietary advice is calibrated for a person whose problem is EXCESS, and a substantial number of people in a hospital have the opposite problem. Energy density up. Fortify everything. Sugar is irrelevant. Oral supplements ARE ultra-processed and they save lives. ⚠️ "Eat less" is the most dangerous sentence in the book.

⚠️ §29.12 closes the restriction pattern's fifth instance with the general rule: a person who is unwell, motivated and looking for something to control is unusually likely to be handed a restriction — by a clinician, a test, a protocol, or themselves. The question is almost never "what should be removed?" It is "are they getting enough, and of what?"

⚠️ And §29.16's re-audit answers the question this book has been tracking since Chapter 16. The affirmative share went 15.4% → 25.3% → ~42%, and the rise is honest: the QUESTIONS changed, not the standard. Part IV adjudicated contested public arguments, which are contested precisely because the evidence is ambiguous. Part V asked applied clinical questions, and Part V's ✅s are largely clinical-medicine claims — adrenaline, coeliac testing, tube position, resistance training — where randomized evidence is far more available than in dietary epidemiology. ⚠️ The 🟡 column grew too, from 20 to 60, which is what you would NOT see if the standard had loosened.

The one thing to remember: ⚠️ When a procedure is cancelled, the patient should eat. It requires no equipment, no budget and no training, and it is the intervention most often missed.

What's Next

Part V closes here. Chapter 30 opens Part VI — Practical Nutrition — and the register changes again.

⚠️ No more physiology. No more trials. Part VI is about the things you actually do: reading a label properly, planning meals, feeding a household on a budget, the psychology of why you eat what you eat, eating disorders, personalized nutrition, and the food system you're buying from.

Chapter 30 starts with the label — which is a regulated document, designed by committee, shaped by lobbying, and much more informative than it looks once you know which parts are meaningful and which are decoration.