Clinical Nutrition

Abstract

Clinical nutrition sits between two facts that the field keeps trying to collapse. First: undernutrition, sarcopenia, and disease-related malnutrition are associated with infection, delayed recovery, and death in observational cohorts and in hospital audits. Second: when large randomised trials have tested whether delivering more calories, more protein, a specialised formula, or an earlier parenteral supplement changes those hard outcomes, the typical result in intensive care has been a null — and occasionally a harm. The Swiss EFFORT trial of individualised support in medical inpatients at nutritional risk is not the same experiment as EFFORT Protein in ventilated high-risk ICU adults; they share a word and almost nothing else (Schuetz et al., 2019; Heyland et al., 2023).

Assessment is not one act. Screening tools (MUST, NRS-2002, MNA) estimate risk. Subjective global assessment and the GLIM construct attempt diagnosis. NUTRIC kept severity physiology and dropped the classical nutrition history. Albumin and prealbumin track inflammation more faithfully than intake. Energy equations disagree on the same body. Protein targets written as 1.2–2.0 g/kg in critical-care guidelines failed their best randomised test of a higher versus usual prescription. Refeeding syndrome is a real phosphate-and-thiamine physiology; it is not a licence to withhold food indefinitely.

Nutrition support is route engineering before it is outcome science. Oral supplements, enteral tubes, and parenteral admixtures have indications, contraindications, formulations, and complications that are better documented than the mortality claims attached to them. CALORIES and NUTRIREA-2 did not show that early enteral feeding reduces death when parenteral feeding is a feasible alternative; NUTRIREA-2 found more digestive ischaemia with early isocaloric enteral feeding in shocked ventilated adults (Harvey et al., 2014; Reignier et al., 2018). Disease chapters must say when an ICU finding does not travel — to surgery, oncology, malabsorption, liver, kidney, cardiometabolic disease, neurology, wounds, or frailty. This article compares ASPEN, ESPEN, NICE, and KDOQI where they disagree, and it treats consensus as consensus.

Clinical nutrition did not begin as a lifestyle subject. It began as the hospital problem of the patient who could not, or would not, eat enough to meet a catabolic illness — and of the staff who could see wasting and still lacked a vocabulary that separated starvation from inflammation, intake from absorption, and a risk score from a diagnosis. The early instruments were crude and honest: weight, intake, oedema, a hand on the temple and the shoulder. Later instruments became numerical and less honest when a laboratory value or a formula bag was allowed to stand in for that examination.

The gold coin of this Part is simple and still failed in practice. A low albumin is not a malnutrition diagnosis. Neither is a calorie deficit computed from an equation, a MUST score, or a low plasma glutamine. Each of those objects has a use. None of them is the others. The rest of the article is an attempt to keep them from collapsing.

02 What “malnutrition” has been asked to mean

The word has been asked to do too much work. It has meant simple starvation; disease-related wasting with inflammation; obesity with depleted lean mass; a screening score above a cut-point; a GLIM phenotype; a billing code; and, in careless speech, any abnormal nutrition-related laboratory value. Those are not one object. A methods paper that treats them as one object will generate a prevalence, an association with death, and an apparent mandate to “feed more” that no trial has earned.

An international consensus committee proposed an etiology-based diagnosis that distinguished starvation-related, chronic disease-related, and acute disease- or injury-related malnutrition and moved the argument off isolated laboratories toward a clinical cluster (Jensen et al., 2010). Later Academy / ASPEN and ESPEN documents developed that frame for adult practice. The Global Leadership Initiative on Malnutrition (GLIM) proposed a two-step structure: a validated screen, then diagnosis requiring at least one phenotypic criterion (weight loss, low BMI, or reduced muscle mass) and one etiologic criterion (reduced intake or assimilation, or inflammation). That is a consensus diagnostic frame. It is not a treatment trial. It does not say that meeting GLIM and then delivering a commercial formula changes mortality.

Cachexia in cancer or advanced organ failure can satisfy a malnutrition phenotype while the dominant biology is inflammatory wasting that food does not reverse. Sarcopenia can exist in a person with a normal or high BMI. Protein-energy wasting in kidney disease is a related but separately named construct (Ikizler et al., 2020). The article keeps those names. Collapsing them into “the patient is malnourished, therefore support” is the error Part Five exists to catch.

03 Screening is not diagnosis

Hospital screening tools were built to find people who warrant a closer look, not to name a disease. The Malnutrition Universal Screening Tool (MUST) combines BMI, unplanned weight loss, and an acute-disease effect on intake. NRS-2002 adds a severity-of-disease axis to impaired nutritional status and was designed for the hospital, including the ICU (Kondrup et al., 2003). The Mini Nutritional Assessment was built for older adults (Guigoz, 2006). Subjective global assessment (SGA) is a structured clinical judgement from history and physical findings, not a laboratory panel (Detsky et al., 1987). Screening tools themselves have been evidence-graded; many are better at finding risk than at predicting who will benefit from a later intervention (Skipper et al., 2012).

NUTRIC is a different object. It was constructed from age, APACHE II, SOFA, comorbidities, days from hospital to ICU, and, in the original form, interleukin-6, and it was offered as a way to identify patients most likely to benefit from aggressive nutrition. A later validation paper kept that logic (Rahman et al., 2016). The score is a severity instrument wearing a nutrition badge. It does not contain weight loss or dietary history. Later megatrials that actually delivered more calories or more protein to high-risk ventilated adults did not confirm a general rule that “more feeding helps the high-risk ICU patient” (Chapman et al., 2018; Heyland et al., 2023). Using NUTRIC as if it were MUST is a category error.

The Swiss EFFORT trial is the cleanest randomised test of a screen-to-support pathway outside the ICU. Schuetz and colleagues randomised medical inpatients at nutritional risk to individualised nutritional support versus standard hospital food. The protocolised arm improved a composite of adverse clinical outcomes and reduced mortality in that ward population (Schuetz et al., 2019). That result does not travel automatically into shock, mechanical ventilation, or dialysis. It also does not license a commercial formula as the active ingredient; the contrast was a structured assessment-and-support process versus usual hospital meals.

Matrix A — Screening versus diagnosis. Tools are not interchangeable. Citations name the defining paper or consensus, not a licence to treat.

04 Anthropometry and body composition

Weight is the oldest vital sign in this discipline and still the most abused. A falling weight in the presence of oedema, ascites, or a large resuscitation can hide wasting. A stable weight in a person who has lost muscle and gained water is not reassurance. BMI is a population sorting tool. It is a poor individual diagnosis of lean-mass depletion, and it is a worse one in older adults and in many Asian populations where the same BMI corresponds to a different fat–lean split. Mid-upper-arm circumference and calf circumference are field approximations of muscle; they are not DXA.

Body-composition methods — DXA, CT at a lumbar vertebra, bioimpedance, ultrasound of the quadriceps — measure different physical objects with different error structures. CT muscle area at L3 became popular in oncology and ICU research because the scan already existed. It is a research phenotype, not a bedside screen, and it is not a feeding indication by itself. Bioimpedance in the oedematous critically ill patient is a conductivity problem wearing a nutrition label. The article treats composition data as strongly supported for describing wasting and speculative as a sole trigger for a support regimen.

05 Biochemical assessment and its traps

The characteristic laboratory error in this field is to read a plasma concentration as a grocery list. Albumin and prealbumin fall with the acute-phase response. They are prognostic. They are not, in isolation, measures of recent intake, and the etiology-based consensus already moved adult malnutrition diagnosis off those proteins (Jensen et al., 2010). A rising prealbumin after a week of feeding may track resolving inflammation as much as it tracks nitrogen.

Phosphate, potassium, and magnesium fall when carbohydrate is reintroduced after starvation; that is refeeding physiology, taken up in Part Two, not proof that the patient was “phosphate deficient” in the dietary sense before the meal. Iron studies, B12, folate, 25-hydroxyvitamin D, zinc, and selenium each have indication-specific uses and each can be deranged by inflammation, renal clearance, binding proteins, or assay interference. Low plasma glutamine in the ICU is a marker of a catabolic circulation. Replacing it in ventilated adults with multiorgan failure increased death in REDOXS (Heyland et al., 2013). That trial is the field’s most expensive lesson in the difference between a biochemical abnormality and a nutritional deficiency.

Nitrogen balance is a research assay with incomplete collections and an assumption that the unmeasured losses are small. A positive balance is not hypertrophy and is not survival. Urinary urea nitrogen in the ICU is a stress readout as much as a protein-need calculator.

06 Dietary assessment

In the hospital the useful dietary question is often short: what has been eaten in the last five days, and can the gut receive food now? Food records, 24-hour recalls, and food-frequency questionnaires were built for epidemiology. They misclassify individuals. In critical care they are usually impossible. Calorie counts on a ward tray are better than memory and still miss what was taken from a visitor’s bag.

Estimated energy and protein intake should be compared with an estimated requirement that names its method (Part Two). A gap on a clipboard is not an outcome. The Swiss EFFORT investigators treated the gap as a problem to be closed by a protocolised process in medical inpatients; they did not treat a food-frequency questionnaire as a diagnosis (Schuetz et al., 2019). Outpatient dietary assessment is a different literature — and a sibling article problem for macronutrient patterns, not a substitute for the hospital examination.

Energy need is a rate, not a personality. Resting energy expenditure (REE) can be measured by indirect calorimetry when the circuit is closed enough and the FiO₂ is not too high. In health, predictive equations are a convenience. In critical illness they are a guess with a confidence interval wide enough to contain both “trophic” and “full” feeding. Mifflin and St Jeor derived a REE equation in healthy adults that outperformed Harris–Benedict in that sample (Mifflin et al., 1990). Penn State and Ireton-Jones modifications attempt to carry the guess into the ventilated patient. None of these equations is a mortality target.

The gold coin of this Part: two honest energy equations can disagree on the same patient by several hundred kilocalories. Guidelines that write “25 kcal/kg” or “70% of measured EE” are choosing a convention. ASPEN/SCCM 2016 and the 2022 update, and ESPEN’s ICU guideline, do not use identical conventions, and they do not use identical readings of the same calorie megatrials (McClave et al., 2016; Compher et al., 2022; Singer et al., 2019).

The randomised calorie literature in ventilated adults is now large enough to constrain slogans. EDEN compared about 400 with about 1,300 kcal/day for six days in acute lung injury and found no difference in ventilator-free days or 60-day death; the fuller arm had more gastrointestinal intolerance (Rice et al., 2012). PermiT held protein similar and reduced non-protein calories to 40–60% versus 70–100% of calculated need for up to fourteen days; 90-day mortality did not differ (Arabi et al., 2015). TARGET used a blinded energy-dense versus routine enteral formula at the same volume in 3,957 ventilated adults, delivered about 600 kcal/day more in the dense arm, and did not improve 90-day survival (Chapman et al., 2018). NUTRIREA-3 compared early low versus standard calorie and protein intakes in shocked ventilated adults and did not find a day-90 mortality cost to the lower early prescription (Reignier et al., 2023).

Those trials do not say that energy is irrelevant in a starving medical inpatient, a postoperative patient who can eat, or a person living with obesity who is not critically ill. They say that the guideline habit of treating “full calculated calories by day 3 in the ICU” as a survival intervention failed its best tests. Permissive underfeeding, in the sense PermiT actually tested, is a non-protein calorie contrast with protein held similar. It is not a protein-restriction trial and not a clinic diet.

08 Protein requirements and the ICU overshoot

Healthy-adult protein requirement methods — nitrogen balance and indicator amino-acid oxidation — are the subject of the sibling amino-acid article. They estimate a rate that prevents negative balance in health. They are not hypertrophy prescriptions and they are not ICU laws. Older-adult consensus statements have argued for higher protein than the historical 0.8 g/kg healthy-adult RDA, often in the neighbourhood of 1.0–1.2 g/kg or more when sarcopenia or illness is present (Deutz et al., 2014). That is expert synthesis plus physiology. It is not EFFORT Protein.

Critical-care guidelines wrote higher still. ASPEN/SCCM 2016 placed many ICU adults in a 1.2–2.0 g/kg band, with still higher figures named for burns, trauma, and continuous renal replacement (McClave et al., 2016). ESPEN’s ICU guideline favoured a lower central figure, often cited as about 1.3 g/kg, after progression from the acute phase (Singer et al., 2019). Observational ICU cohorts in which patients who received more protein did better cannot separate indication, survival bias, and feeding tolerance from a causal protein effect (Weijs; Allingstrup — observational, not randomised).

EFFORT Protein is the randomised test those bands needed. Heyland and colleagues assigned 1,301 mechanically ventilated adults at high nutritional risk to a higher versus usual protein prescription. Time to discharge alive from hospital did not improve. Secondary signals included possible harm in acute kidney injury (Heyland et al., 2023). Keeping the 1.2–2.0 g/kg band as a clinical law after that trial is a decision to prefer the guideline committee over the outcome experiment. NUTRIREA-3, which lowered both calories and protein early in shock, likewise failed to show that the standard early prescription was necessary for day-90 survival (Reignier et al., 2023).

None of this retires protein as a substrate for muscle in rehabilitation, in older outpatients, or in the ward patient who can eat. It retires the habit of treating a high ICU protein target as a mortality intervention. The sibling high-protein-diet and amino-acid articles own the outpatient and supplement arguments; this title owns the hospital overshoot.

09 Micronutrient assessment

Micronutrient assessment in clinical nutrition is indication-led, not panel-led. Thiamine before or with carbohydrate after prolonged starvation is a refeeding precaution with a long clinical history, not a megatrial. Water-soluble vitamins are lost in dialysis and in high-output gastrointestinal fluids. Fat-soluble vitamins track cholestasis, malabsorption, and inflammation-altered binding proteins. Trace elements in parenteral nutrition are a compounding and contamination problem as much as a requirement problem (Part Three).

Pharmaconutrition — high-dose single nutrients given as if they were drugs — has a worse record than replacement of a documented deficiency. REDOXS tested glutamine and antioxidants in ventilated multiorgan failure and found that glutamine increased in-hospital and six-month death; antioxidants were null (Heyland et al., 2013). SIGNET and related glutamine/selenium ICU programmes did not rescue the biochemical-deficiency story. Immune-modulating high-protein enteral programmes have not produced a clean outcome win that would justify treating “immunonutrition” as a standard of ICU care. Industry-adjacent formula trials that report nitrogen balance, length of stay, or infection composites without a pre-specified mortality or durable-function primary are underpowered for the claim on the bag.

Vitamin D, zinc, and selenium each have disease-specific literatures in which a low plasma value is common and a replacement trial is either null, mixed, or confined to a surrogate. This article does not convert those literatures into a standing supplement list. A documented deficiency in a named pathway (B12 malabsorption, night blindness in cholestasis, scurvy) is a different object from a low 25-hydroxyvitamin D in acute inflammation.

10 Refeeding syndrome

Refeeding syndrome is the shift of phosphate, potassium, and magnesium into cells, with thiamine consumption, when carbohydrate oxidation restarts after a period of starvation or near-starvation. Cardiac arrhythmia, fluid overload, and Wernicke physiology are the feared clinical expressions. The physiology is established. The incidence depends on the definition, the population, and whether anyone measured a phosphate (Mehanna et al., 2008).

NICE nutrition-support guidance made refeeding a reason to start feed cautiously in people with very low BMI, little recent intake, or abnormal electrolytes, and to give thiamine. That is a guideline precaution, not a randomised factorial of start-rate versus outcome. The opposite error is also live: treating any phosphate dip as a reason to keep a hungry patient nil by mouth for days. Hypophosphataemia after feeding is a reason to replace electrolytes and to think about rate; it is not a diagnosis of “the gut has failed.”

Who is at risk is more specific than “everyone who has been unwell.” Prolonged starvation, alcohol-use disorder, anorexia nervosa, and prolonged unfed postoperative states are the classical settings. A three-day ICU fast in a previously nourished adult is not the same risk as a month of near-fasting. Definitions that count any phosphate fall after feed starts will find a high incidence and a weak link to arrhythmia. Definitions that require clinical events will find a lower incidence and a clearer harm. The article reports both and does not invent a start-rate for any person.

The first route is the mouth. That sentence is more radical than it sounds in a field that industrialised tubes and bags. Oral nutrition includes hospital food, snacks, texture-modified diets, and oral nutritional supplements. The Swiss EFFORT contrast in medical inpatients at nutritional risk was a structured process that used oral support as the default, not a randomised comparison of two commercial cartons (Schuetz et al., 2019). Industry trials of a named sip feed against “standard care” that already includes poor mealtime assistance are tests of a system as much as of a product.

Contraindications to oral intake are mechanical or neurological: unsafe swallow, obstruction, severe vomiting, or an explicit nil-by-mouth decision for a procedure. They are not a low albumin. Dysphagia assessment belongs to speech-and-language and endoscopic literatures; this article records that texture modification can reduce intake even as it reduces aspiration risk — a trade the IDDSI framework tried to standardise, not a licence to puree every older adult.

The gold coin of this Part sits in the middle of the document on purpose. The gut works, until it does not — and parenteral nutrition is not a moral failure. Culture in the last two decades treated the central line as a confession. CALORIES and NUTRIREA-2 are the randomised refusal of that sermon when both routes are feasible.

12 Enteral nutrition

Enteral nutrition is delivery of a formulated liquid into the gut through a tube — nasogastric, nasojejunal, gastrostomy, or jejunostomy. It uses the portal circulation and the gut barrier. It does not require that the patient be conscious. It does require that the gut be able to receive fluid without vomiting, massive residual, or ischaemia.

ASPEN and ESPEN both prefer the gut when it works. That preference is physiology and infection-control culture. It is not, after CALORIES and NUTRIREA-2, a mortality result. Harvey and colleagues randomised 2,400 unplanned ICU admissions to early parenteral versus early enteral support and found no 30-day mortality difference (33.1% versus 34.2%) (Harvey et al., 2014). Reignier and colleagues randomised 2,410 shocked ventilated adults to early isocaloric enteral versus parenteral nutrition and found no day-28 mortality difference, with more digestive complications — including bowel ischaemia — in the enteral arm (Reignier et al., 2018). “EN first” survives as engineering when the gut is usable. It does not survive as “EN reduces death.”

Gastric residual volume as a feeding-pause rule is a habit with a weak outcome base. Intolerance is real; a single residual number is not a laparotomy. Post-pyloric tubes trade placement difficulty for less gastric pooling; they have not been shown, at megatrial scale, to change death. Energy-dense commercial formulas were the tool in TARGET: same volume, more calories, no 90-day survival gain (Chapman et al., 2018). That is the cleanest available test of the claim that a manufactured energy-dense product, used to hit a calorie target, improves a hard endpoint in ventilated adults. It failed.

13 Parenteral nutrition

Parenteral nutrition is an intravenous admixture of amino acids, glucose, lipid, electrolytes, vitamins, and trace elements. Short-term peripheral PN is osmolarity-limited. Central PN is a line-infection and thrombosis problem. Long-term home PN is a life-saving therapy in irreversible intestinal failure and a chronic liver-and-bone disease problem.

EPaNIC is the trial that broke the habit of early supplemental PN as a completeness ritual. Casaer and colleagues randomised 4,640 ICU adults to supplemental PN within 48 hours versus not before day 8. Late initiation led to earlier live ICU discharge and fewer ICU infections. Death at ICU, hospital, and 90 days did not differ. The case-mix included a large cardiac-surgery fraction and tight glycemic control (Casaer et al., 2011). “Late PN saves lives” over-reads the primary story. “EPaNIC is every underfed ICU patient” ignores the case-mix. Doig and colleagues later tested early PN in critically ill adults with short-term relative contraindications to early EN and did not reproduce EPaNIC as a simple opposite (Doig et al., 2013). The two trials enrolled different gaps. They are not a pair of refutations; they are two different questions.

Heidegger and colleagues tested supplemental PN to close an energy gap after day 3 in ICU patients not meeting targets enterally; the reported benefit was nosocomial infection, not mortality. Infection endpoints in unblinded nutrition trials are fragile. The article records the trial as a gap-closure experiment, not as a licence to start PN on arrival.

14 Indications, contraindications, formulations

A usable indication is a gut that cannot meet need for long enough that starvation becomes the dominant problem — obstruction, high-output fistula, short bowel, severe mucositis, or prolonged ileus — or a swallow that is unsafe and a tube that is not yet placed. A usable contraindication to EN is gut ischaemia, uncontrolled obstruction, or a surgical instruction that the anastomosis must not be loaded. A usable contraindication to PN is a working gut plus reliable oral or enteral access, or a line situation in which infection risk dominates a short anticipated gap. These are physiological statements. They are not doses.

Formulations are industrial objects. Standard polymeric enteral formulas differ in energy density, protein density, fibre, and osmolality. Disease-specific labels — “renal,” “hepatic,” “diabetic,” “immune-modulating” — encode a composition hypothesis. The hypothesis is only as good as the outcome trial. TARGET’s energy-dense product delivered more calories and did not improve survival (Chapman et al., 2018). Immune-modulating cocktails (arginine, nucleotides, fish oil) have a mixed surgical-infection literature and a poor ICU mortality literature; REDOXS is a harm trial for high-dose glutamine in shocked ventilated multiorgan failure (Heyland et al., 2013). SIGNET tested glutamine and selenium added to PN in critically ill adults and did not establish those additives as outcome therapy (Andrews et al., 2011).

Parenteral lipid emulsions have migrated from soy-heavy to mixed-oil and fish-oil-containing products. The biochemical rationale is fatty-acid composition and phytosterol load. The outcome literature is smaller than the marketing. Liver injury in long-term PN is multifactorial: overfeeding, sepsis, lack of enteral stimulation, and lipid dose. Switching the oil is not a substitute for reducing a calorie surplus.

15 Complications and monitoring

Enteral complications: misplaced tubes in the lung, sinusitis, aspiration, diarrhoea, constipation, peristomal leak, buried bumper, and — in shock — bowel ischaemia (Reignier et al., 2018). Parenteral complications: central-line bloodstream infection, thrombosis, hyperglycaemia, hypertriglyceridaemia, electrolyte swings, manganese accumulation, and intestinal-failure-associated liver disease. Refeeding belongs here as well as in Part Two: the first days of any route can drop phosphate.

Monitoring that earns its keep: tube or line position, glycaemia, electrolytes including phosphate, fluid balance, liver enzymes in long-term PN, and whether the patient is actually receiving what was prescribed. Monitoring that does not earn its keep as a sole dashboard: albumin, a single gastric residual, and a calorie-adequacy percentage treated as a quality trophy after TARGET. Functional measures — sitting, walking, grip, return home — are closer to what patients notice and farther from what formula trials usually power.

Hyperglycaemia during nutrition support is both a marker of stress and a consequence of the glucose infusion. Tight glycemic control was part of the EPaNIC environment; it is not a free variable one can ignore when importing that trial into a unit that runs looser glucose. Overfeeding remains a more plausible harm than underfeeding in the first ICU days of a previously nourished adult. That sentence is an ICU sentence. It is not a community-nutrition slogan.

The thematic turn is here. The interesting question is no longer whether an unfed patient suffers. Observational association already answered that in the weak sense. The interesting question is which intervention, in whom, against which endpoint — and why so many disease-specific recommendations cannot answer it. Critical care is the one domain that actually possesses large outcome trials. That is not a licence to treat the ICU as the law of the hospital.

EPaNIC, EDEN, PermiT, TARGET, CALORIES, NUTRIREA-2, NUTRIREA-3, REDOXS, and EFFORT Protein are the spine (Casaer et al., 2011; Rice et al., 2012; Arabi et al., 2015; Chapman et al., 2018; Harvey et al., 2014; Reignier et al., 2018; Reignier et al., 2023; Heyland et al., 2013; Heyland et al., 2023). Read together they say: early full calorie delivery is not a survival intervention in ventilated adults; early EN is not a survival intervention when PN is available; high-dose glutamine in shocked multiorgan failure can kill; a higher protein prescription in high-risk ventilated adults did not speed discharge alive. Surviving Sepsis and society ICU nutrition guidelines still write targets. Those targets are committee readings of this spine, not replacements for it.

17 Surgery and enhanced recovery

Elective surgery is not septic shock. Enhanced-recovery programmes shortened length of stay by changing fluids, opioids, mobilisation, and the habit of prolonged preoperative fasting (Fearon et al., 2005). Carbohydrate-containing preoperative drinks and early postoperative eating are components of that bundle. Isolating “immunonutrition for every GI anastomosis” from the bundle is how a composition hypothesis becomes a standing order. ESPEN surgical guidelines have been more willing than some North American documents to discuss arginine-containing formulas in selected oncologic GI surgery; the evidence is infection and length of stay, often industry-adjacent, and it is not TARGET. ESPEN surgical nutrition texts are guideline documents: a mix of randomised fragments and consensus.

Emergency laparotomy, anastomotic leak, and open abdomen return the problem to ICU physiology. Importing ERAS snack rules into an ischaemic bowel is the illicit move in reverse.

18 Oncology

Cancer malnutrition is a mixture of anorexia, obstruction, treatment mucositis, and cachexia. ESPEN cancer nutrition guidelines and the expert-group statement on cancer-related malnutrition describe screening, treatable obstruction, and the limits of feeding in progressive cachexia (Arends et al., 2017; Arends et al., 2021). Parenteral nutrition at the end of life is a goals-of-care question, not a calorie-adequacy trophy. Neutropenic “low-microbial” diets have a thinner evidence base than their popularity. The ICU calorie nulls do not say that a patient receiving induction therapy who cannot swallow should be left to a tray. They say that a ventilated patient with ARDS is not a leukaemia ward.

Home EN and home PN in head-and-neck and GI oncology are access therapies. Disease-specific “cancer formulas” with extra n-3 fatty acids have surrogate literatures (eicosanoids, weight) that have not been converted into survival trials of adequate size. The article grades those products as plausible composition, unverified hard outcome.

19 Gastrointestinal disease and malabsorption

Here the gut is the disease. Inflammatory bowel disease, coeliac disease, pancreatic exocrine insufficiency, bile-salt deficiency, short bowel, and radiation enteropathy reduce assimilation. ESPEN IBD guidance treats exclusive enteral nutrition as an induction option in Crohn’s disease, especially in children, and treats PN as support when the gut is unusable, not as primary anti-inflammatory therapy (Forbes et al., 2017). Albumin in IBD is inflammation and losses; feeding it with amino acids does not close the fistula.

Short bowel is the clearest indication for long-term PN and for later glucagon-like peptide-2 analogue therapy — a pharmacologic literature this article notes without adopting. Fistula output, stoma height, and remaining bowel length predict dependence better than a MUST score. Nightingale and colleagues’ short-bowel guidance is physiology first. Importing ICU “trophic feeding” into a 50-centimetre jejunum is a category error. So is treating every high stoma as a reason to withhold oral food when oral food is the adaptation stimulus.

Malabsorption is not low intake. A three-day food record can look adequate while faecal elastase or a 75-gram fat collection says otherwise. Pancreatic enzyme replacement and bile-acid binders are drugs with indications; they are not “nutrition support” in the EN/PN sense and they are not dosed here.

20 Liver and kidney disease

Cirrhosis wastes muscle while the extracellular water expands. BMI lies. Ammonia-era protein restriction for every cirrhotic is a retired harm; ESPEN liver guidance moved toward avoiding unnecessary protein restriction in compensated disease and toward recognising sarcopenia as a transplant-waitlist problem. Late-night oral carbohydrate in cirrhosis is a small-trial literature on muscle and fuel, not a survival megatrial. Ascites is not a feeding indication. A low sodium diet for ascites is a fluid-management tactic, not a malnutrition therapy.

Kidney disease reverses the protein story at the dialysis threshold. KDOQI 2020 is a GRADE-labelled guideline that is still, in long stretches, expert opinion (Ikizler et al., 2020). Protein restriction in metabolically stable non-dialysis CKD is a different evidence structure from higher protein targets once dialysis or a catabolic AKI begins. Those are not one sentence. The EFFORT Protein secondary signal in AKI now sits against older “higher protein in AKI” consensus and must not be silently dropped or exported to the outpatient dialysis unit (Heyland et al., 2023). Phosphorus and potassium in CKD are diet, binders, residual function, and assay — not automatic proof of dietary excess or deficiency. Protein-energy wasting is the kidney-specific wasting name; it is not GLIM by another label, and it is not a licence to ignore volume.

21 Cardiovascular disease, diabetes, and obesity

These are pattern-and-drug diseases more than tube-and-bag diseases. AHA dietary guidance is a food-pattern document (Lichtenstein et al., 2021). Diabetes nutrition-therapy statements are the same kind of object. They are not EN protocols. Hospitalised heart-failure patients are salt-and-volume problems with a sarcopenia underlay; cardiac cachexia does not reverse because a carton is opened. Diabetes nutrition support is glycaemic monitoring of whatever route is chosen, not a search for a magical “diabetic formula.” Disease-specific diabetic EN products change carbohydrate quality; they have not earned a mortality claim.

Obesity in the ICU produces the opposite error from starvation: underestimation of lean mass, overestimation of calorie need from actual body weight, and a temptation to “use the admission as a diet.” Hypocaloric high-protein ICU prescriptions for obesity appear in ASPEN documents as consensus (McClave et al., 2016; Compher et al., 2022). They were not the TARGET contrast. Bariatric surgery and GLP-1 receptor-agonist eras have created new malnutrition-after-weight-loss problems — thiamine, iron, B12, protein — that belong to those literatures and to the sibling metabolic articles. This title records the category: obesity does not immunise against nutrient deficiency, and weight loss is not a nutrition-support endpoint in the ICU.

22 Neurological disease, wounds, frailty, and sarcopenia

Stroke, motor-neuron disease, and Parkinsonism make swallow the rate-limiting step. ESPEN neurology guidance is a tube-timing and goals-of-care document as much as a calorie document. Early gastrostomy in acute stroke has a randomised history (FOOD) that this article treats as a timing-and-complication literature, not as proof that a formula brand changes disability. Prolonged nasogastric feeding is a comfort and dignity problem.

Pressure injuries are associated with wasting and immobility. Nutrition is one of several supports; it is not a topical dressing. Trials powered on wound area are usually too small for death and often too small for healing. Formula marketing that treats a wound as a protein-and-zinc deficiency awaiting a carton is the biochemical error in another costume.

Frailty and sarcopenia are geriatric objects. EWGSOP2 defined sarcopenia as low muscle strength plus low muscle quantity or quality, with poor performance as severity (Cruz-Jentoft et al., 2019). ESPEN geriatric nutrition and hydration guidance and the PROT-AGE protein statements argue for protecting intake and protein in older adults (Deutz et al., 2014). Those are consensus-plus-physiology documents. They are not EFFORT Protein. A restriction diet that shrinks an older adult’s dinner in the name of a longevity hypothesis is the illicit export of a metabolic idea into a muscle problem — the red-team case owned jointly with the Nutrition-for-Aging sibling. Exercise is not optional scenery in sarcopenia trials; protein without the training stimulus is a weaker object.

Society documents are not a chorus. ASPEN/SCCM 2016 and the 2022 adult critically ill update, ESPEN’s ICU guideline, NICE nutrition-support guidance, and KDOQI 2020 read overlapping literatures and still disagree on timing of supplemental PN, on protein ceilings, and on how much expert opinion a GRADE table may carry (McClave et al., 2016; Compher et al., 2022; Singer et al., 2019; Ikizler et al., 2020). NICE CG32 / the later nutrition-support quality standard is a UK health-system document: screen, assess, and do not invent a feeding ritual. It is not an ICU protein-dose trial.

Matrix B — Guideline disagreement is two committee readings of the same nulls, not a wording quarrel. Amounts in the table are published guideline or trial parameters, not instructions.

KDOQI versus ICU protein documents is a different disagreement: the population reverses. Disease-society food-pattern statements (AHA; diabetes nutrition therapy; adult obesity guidance) are not EN manuals and should not be footnoted as if they were (Lichtenstein et al., 2021; Wharton et al., 2020). ESPEN organ guidelines (surgery, cancer, IBD, liver, neurology, geriatrics) are the right documents for those wards and are still mixed evidence-plus-consensus (Arends et al., 2017; Arends et al., 2021; Forbes et al., 2017).

24 RCT architecture: early versus late, protein, permissive underfeeding

Heterogeneity is not a polite caveat. EPaNIC’s cardiac-surgery fraction, EDEN’s ARDS six-day window, PermiT’s matched protein, TARGET’s blinded volume, NUTRIREA-2’s shock, and EFFORT Protein’s high-risk ventilated sample are different machines. A meta-analysis that pools “more nutrition versus less” without those axes will report a summary that no clinician can use.

Mortality versus surrogate endpoints is the other axis. Length of stay, infection, calorie adequacy, nitrogen balance, and gastric residuals are not death and are not durable function. An intermediate “win” does not retire a mortality null. Swiss EFFORT’s ward composite and mortality signal is a process result in medical inpatients (Schuetz et al., 2019). EFFORT Protein’s null on time-to-discharge-alive is a dose result in ventilated high-risk adults (Heyland et al., 2023). Confusing the two because they share a word is a methods failure.

Permissive underfeeding, in the only sense PermiT licensed, is reduced non-protein calories with protein held similar (Arabi et al., 2015). It is not a protein-restriction diet and not a longevity fast. Early versus late feeding is a clock on a specific route. EPaNIC timed supplemental PN. It did not time the first sip on a ward.

25 Immunonutrition, specialised formulas, pharmaconutrients

Immunonutrition promised that arginine, nucleotides, and fish oil would change infection. In elective GI cancer surgery the literature is mixed and often industry-proximate; endpoints are usually infection or stay. In the ICU the literature includes a large harm trial for glutamine (Heyland et al., 2013) and a null-to-negative record for routine cocktails (Andrews et al., 2011). Specialised renal, hepatic, and diabetic formulas are packaging of electrolyte and carbohydrate hypotheses. They are rarely powered for mortality. TARGET is the reminder that even a well-made energy-dense product, used to deliver the calories the guideline asked for, can fail the endpoint that would justify the bag (Chapman et al., 2018).

Micronutrient supplementation as pharmaconutrition is the REDOXS lesson. Replacement of a documented deficiency in a named pathway remains ordinary medicine. High-dose single-nutrient therapy in inflamed, shocked adults is a drug trial and should be read as one.

26 Adversarial review

The six questions were put to specialist adversarial readings in critical care, renal medicine, gastroenterology, and nutrition methodology. The answers converge. They are not a tie.

Are guideline recommendations evidence-based or expert consensus? Both, and the stamps must stay separate. ASPEN and ESPEN disagree on early supplemental PN and on protein ceilings after reading the same megatrials. Reprinting a numeric target as if those trials had shown the target to be superior is consensus in an evidence costume. KDOQI is GRADE-labelled and still, in long stretches, expert opinion (Ikizler et al., 2020).

Are ICU findings generalized beyond ICU populations? Constantly, and in both directions. TARGET, PermiT, EDEN, and NUTRIREA-3 enrolled ventilated adults for days to weeks. Exporting “calories did not change death” to a cancer clinic or a longevity diet is illicit. Importing CKD restriction lore into shocked AKI is the same move reversed. Swiss EFFORT is the ward counterexample that must not be relabelled as an ICU protein trial (Schuetz et al., 2019; Heyland et al., 2023).

Are biochemical abnormalities mistaken for nutritional deficiency? Yes. Albumin is inflammation. Glutamine in the ICU is a catabolic marker; replacing it increased death (Heyland et al., 2013). Phosphate and potassium in CKD are composites of diet, drugs, and residual function. The etiology-based consensus and later GLIM frame already moved diagnosis off isolated laboratories (Jensen et al., 2010). A later chapter that re-elevates albumin as a feeding indication contradicts Part One.

Are industry-manufactured formulas supported by clinical outcomes? They are supported by composition. TARGET used a commercial energy-dense product, delivered more calories, and did not improve 90-day survival (Chapman et al., 2018). Renal and GI specialty products are typically powered for laboratories or tolerance. Meeting a specification is not a mortality or function claim.

Are disease-specific interventions sufficiently powered? Calorie and route megatrials are powered for death. Protein, immunonutrition, wound formulas, and “high-risk subgroup” literatures usually are not. EFFORT Protein was large enough to test high versus usual protein in ventilated high-risk adults and did not improve time-to-discharge-alive (Heyland et al., 2023). AKI and severity strata are secondary.

Are mortality and functional outcomes distinguished from biomarkers? Better in the megatrials than in the commentary. Commentary recenters nitrogen balance, residuals, and calorie-adequacy percentages. Keep the trial’s own endpoint in the sentence. Length of stay is not function. nPCR is not survival.

27 Standing constraint

This document describes published research. It is not medical advice. It is not an individualized treatment protocol. No diet, formula, dose, route, schedule, or prescription is recommended for any person. Amounts and durations appear only as parameters of named studies or as labelled guideline statements, with the population attached. Clinical authority remains with qualified professionals and with the person in front of them.

The numbered list is generated from verified NCBI records and from primary-agency guideline pages checked against the instrument. Identifiers that could not be resolved were not printed. Author–year citations in the text correspond to this list.

1. Andrews PJ, Avenell A, Noble DW, Campbell MK, Croal BL, Simpson WG, et al.. Randomised trial of glutamine, selenium, or both, to supplement parenteral nutrition for critically ill patients. BMJ. 2011;342:d1542.
   PMID 21415104 · doi:10.1136/bmj.d1542
2. Arabi YM, Aldawood AS, Haddad SH, Al-Dorzi HM, Tamim HM, Jones G, et al.. Permissive Underfeeding or Standard Enteral Feeding in Critically Ill Adults. N Engl J Med. 2015;372(25):2398-408.
   PMID 25992505 · doi:10.1056/NEJMoa1502826
3. Arends J, Baracos V, Bertz H, Bozzetti F, Calder PC, Deutz NEP, et al.. ESPEN expert group recommendations for action against cancer-related malnutrition. Clin Nutr. 2017;36(5):1187-1196.
   PMID 28689670 · doi:10.1016/j.clnu.2017.06.017
4. Casaer MP, Mesotten D, Hermans G, Wouters PJ, Schetz M, Meyfroidt G, et al.. Early versus late parenteral nutrition in critically ill adults. N Engl J Med. 2011;365(6):506-17.
   PMID 21714640 · doi:10.1056/NEJMoa1102662
5. Compher C, Bingham AL, McCall M, Patel J, Rice TW, Braunschweig C, et al.. Guidelines for the provision of nutrition support therapy in the adult critically ill patient: The American Society for Parenteral and Enteral Nutrition. JPEN J Parenter Enteral Nutr. 2022;46(1):12-41.
   PMID 34784064 · doi:10.1002/jpen.2267
6. Cruz-Jentoft AJ, Bahat G, Bauer J, Boirie Y, Bruyère O, Cederholm T, et al.. Sarcopenia: revised European consensus on definition and diagnosis. Age Ageing. 2019;48(1):16-31.
   PMID 30312372 · PMC6322506 · doi:10.1093/ageing/afy169
7. Detsky AS, McLaughlin JR, Baker JP, Johnston N, Whittaker S, Mendelson RA, et al.. What is subjective global assessment of nutritional status?. JPEN J Parenter Enteral Nutr. 1987;11(1):8-13.
   PMID 3820522 · doi:10.1177/014860718701100108
8. Deutz NE, Bauer JM, Barazzoni R, Biolo G, Boirie Y, Bosy-Westphal A, et al.. Protein intake and exercise for optimal muscle function with aging: recommendations from the ESPEN Expert Group. Clin Nutr. 2014;33(6):929-36.
   PMID 24814383 · PMC4208946 · doi:10.1016/j.clnu.2014.04.007
9. Doig GS, Simpson F, Early PN Trial Investigators Group. Early parenteral nutrition in critically ill patients with short-term relative contraindications to early enteral nutrition: a full economic analysis of a multicenter randomized controlled trial based on US costs. Clinicoecon Outcomes Res. 2013;5:369-79.
   PMID 23901287 · PMC3723321 · doi:10.2147/CEOR.S48821
10. Fearon KC, Ljungqvist O, Von Meyenfeldt M, Revhaug A, Dejong CH, Lassen K, et al.. Enhanced recovery after surgery: a consensus review of clinical care for patients undergoing colonic resection. Clin Nutr. 2005;24(3):466-77.
    PMID 15896435 · doi:10.1016/j.clnu.2005.02.002
11. Forbes A, Escher J, Hébuterne X, Kłęk S, Krznaric Z, Schneider S, et al.. ESPEN guideline: Clinical nutrition in inflammatory bowel disease. Clin Nutr. 2017;36(2):321-347.
    PMID 28131521 · doi:10.1016/j.clnu.2016.12.027
12. Harvey SE, Parrott F, Harrison DA, Bear DE, Segaran E, Beale R, et al.. Trial of the route of early nutritional support in critically ill adults. N Engl J Med. 2014;371(18):1673-84.
    PMID 25271389 · doi:10.1056/NEJMoa1409860
13. Heyland D, Muscedere J, Wischmeyer PE, Cook D, Jones G, Albert M, et al.. A randomized trial of glutamine and antioxidants in critically ill patients. N Engl J Med. 2013;368(16):1489-97.
    PMID 23594003 · doi:10.1056/NEJMoa1212722
14. Heyland DK, Patel J, Compher C, Rice TW, Bear DE, Lee ZY, et al.. The effect of higher protein dosing in critically ill patients with high nutritional risk (EFFORT Protein): an international, multicentre, pragmatic, registry-based randomised trial. Lancet. 2023;401(10376):568-576.
    PMID 36708732 · doi:10.1016/S0140-6736(22)02469-2
15. Ikizler TA, Burrowes JD, Byham-Gray LD, Campbell KL, Carrero JJ, Chan W, et al.. KDOQI Clinical Practice Guideline for Nutrition in CKD: 2020 Update. Am J Kidney Dis. 2020;76(3 Suppl 1):S1-S107.
    PMID 32829751 · doi:10.1053/j.ajkd.2020.05.006
16. Jensen GL, Mirtallo J, Compher C, Dhaliwal R, Forbes A, Grijalba RF, et al.. Adult starvation and disease-related malnutrition: a proposal for etiology-based diagnosis in the clinical practice setting from the International Consensus Guideline Committee. JPEN J Parenter Enteral Nutr. 2010;34(2):156-9.
    PMID 20375423 · doi:10.1177/0148607110361910
17. Kondrup J, Rasmussen HH, Hamberg O, Stanga Z, Ad Hoc ESPEN Working Group. Nutritional risk screening (NRS 2002): a new method based on an analysis of controlled clinical trials. Clin Nutr. 2003;22(3):321-36.
    PMID 12765673 · doi:10.1016/s0261-5614(02)00214-5
18. Lichtenstein AH, Appel LJ, Vadiveloo M, Hu FB, Kris-Etherton PM, Rebholz CM, et al.. 2021 Dietary Guidance to Improve Cardiovascular Health: A Scientific Statement From the American Heart Association. Circulation. 2021;144(23):e472-e487.
    PMID 34724806 · doi:10.1161/CIR.0000000000001031
19. McClave SA, Taylor BE, Martindale RG, Warren MM, Johnson DR, Braunschweig C, et al.. Guidelines for the Provision and Assessment of Nutrition Support Therapy in the Adult Critically Ill Patient: Society of Critical Care Medicine (SCCM) and American Society for Parenteral and Enteral Nutrition (A.S.P.E.N.). JPEN J Parenter Enteral Nutr. 2016;40(2):159-211.
    PMID 26773077 · doi:10.1177/0148607115621863
20. Mehanna HM, Moledina J, Travis J. Refeeding syndrome: what it is, and how to prevent and treat it. BMJ. 2008;336(7659):1495-8.
    PMID 18583681 · PMC2440847 · doi:10.1136/bmj.a301
21. Mifflin MD, St Jeor ST, Hill LA, Scott BJ, Daugherty SA, Koh YO. A new predictive equation for resting energy expenditure in healthy individuals. Am J Clin Nutr. 1990;51(2):241-7.
    PMID 2305711 · doi:10.1093/ajcn/51.2.241
22. Muscaritoli M, Arends J, Bachmann P, Baracos V, Barthelemy N, Bertz H, et al.. ESPEN practical guideline: Clinical Nutrition in cancer. Clin Nutr. 2021;40(5):2898-2913.
    PMID 33946039 · doi:10.1016/j.clnu.2021.02.005
23. National Heart, Lung, and Blood Institute Acute Respiratory Distress Syndrome (ARDS) Clinical Trials Network, Rice TW, Wheeler AP, Thompson BT, Steingrub J, Hite RD, et al.. Initial trophic vs full enteral feeding in patients with acute lung injury: the EDEN randomized trial. JAMA. 2012;307(8):795-803.
    PMID 22307571 · PMC3743415 · doi:10.1001/jama.2012.137
24. National Institute for Health and Care Excellence. Nutrition support for adults: oral nutrition support, enteral tube feeding and parenteral nutrition (CG32). NICE. 2006.
    NICE CG32
25. Nightingale J, Woodward JM, Small Bowel and Nutrition Committee of the British Society of Gastroenterology. Guidelines for management of patients with a short bowel. Gut. 2006;55 Suppl 4(Suppl 4):iv1-12.
    PMID 16837533 · PMC2806687 · doi:10.1136/gut.2006.091108
26. Rahman A, Hasan RM, Agarwala R, Martin C, Day AG, Heyland DK. Identifying critically-ill patients who will benefit most from nutritional therapy: Further validation of the "modified NUTRIC" nutritional risk assessment tool. Clin Nutr. 2016;35(1):158-162.
    PMID 25698099 · doi:10.1016/j.clnu.2015.01.015
27. Reignier J, Boisramé-Helms J, Brisard L, Lascarrou JB, Ait Hssain A, Anguel N, et al.. Enteral versus parenteral early nutrition in ventilated adults with shock: a randomised, controlled, multicentre, open-label, parallel-group study (NUTRIREA-2). Lancet. 2018;391(10116):133-143.
    PMID 29128300 · doi:10.1016/S0140-6736(17)32146-3
28. Reignier J, Plantefeve G, Mira JP, Argaud L, Asfar P, Aissaoui N, et al.. Low versus standard calorie and protein feeding in ventilated adults with shock: a randomised, controlled, multicentre, open-label, parallel-group trial (NUTRIREA-3). Lancet Respir Med. 2023;11(7):602-612.
    PMID 36958363 · doi:10.1016/S2213-2600(23)00092-9
29. Schuetz P, Fehr R, Baechli V, Geiser M, Deiss M, Gomes F, et al.. Individualised nutritional support in medical inpatients at nutritional risk: a randomised clinical trial. Lancet. 2019;393(10188):2312-2321.
    PMID 31030981 · doi:10.1016/S0140-6736(18)32776-4
30. Singer P, Blaser AR, Berger MM, Alhazzani W, Calder PC, Casaer MP, et al.. ESPEN guideline on clinical nutrition in the intensive care unit. Clin Nutr. 2019;38(1):48-79.
    PMID 30348463 · doi:10.1016/j.clnu.2018.08.037
31. Skipper A, Ferguson M, Thompson K, Castellanos VH, Porcari J. Nutrition screening tools: an analysis of the evidence. JPEN J Parenter Enteral Nutr. 2012;36(3):292-8.
    PMID 22045723 · doi:10.1177/0148607111414023
32. TARGET Investigators, for the ANZICS Clinical Trials Group, Chapman M, Peake SL, Bellomo R, Davies A, Deane A, et al.. Energy-Dense versus Routine Enteral Nutrition in the Critically Ill. N Engl J Med. 2018;379(19):1823-1834.
    PMID 30346225 · doi:10.1056/NEJMoa1811687
33. Vellas B, Villars H, Abellan G, Soto ME, Rolland Y, Guigoz Y, et al.. Overview of the MNA--Its history and challenges. J Nutr Health Aging. 2006;10(6):456-63; discussion 463-5.
    PMID 17183418
34. Wharton S, Lau DCW, Vallis M, Sharma AM, Biertho L, Campbell-Scherer D, et al.. Obesity in adults: a clinical practice guideline. CMAJ. 2020;192(31):E875-E891.
    PMID 32753461 · PMC7828878 · doi:10.1503/cmaj.191707

29 Evidence handling

Study type is labelled in the reporting sentence. Animal and in-vitro results are never phrased as human outcomes. Screening scores, diagnostic consensus frames, predictive energy equations, nitrogen balance, guideline targets, and hard clinical endpoints are treated as different objects. When a newer null trial and an older positive observational association conflict, both are kept and the design reasons are named. ICU findings are not exported to ward or outpatient populations without a named warrant. Isolated laboratory values are not treated as deficiencies. Industry formula composition is not treated as an outcome trial.

Project 06 was queried read-only. Identity verification passed; Project 05 remains NOT_IMPORTED. The 06 integration layer is peptide- and biochemistry-heavy and thin on hospital nutrition primaries (no title-level EPaNIC, MUST, NRS-2002, ASPEN ICU, NICE, or KDOQI holdings). Those 06 hits were used as orientation, not as the evidence base. 07_Peptide_News is a journalism/discovery layer and is not cited as scientific evidence; Aspen Pharmacare title collisions were discarded. Full-text retrieval preferred PMC open access. Paywalled papers were used at metadata and abstract level when a verified identifier existed and the claim did not require a hidden table.

Specialist adversarial memos in critical care, renal medicine, gastroenterology, and nutrition methodology were written before the draft was locked and are filed under editorial/. Their hardest objections are answered in Part Five. They are not published apparatus.

30 Limitations

This is a scientific article, not a systematic review with a registered protocol. Coverage is deep on ICU megatrials, screening-versus-diagnosis, and guideline disagreement, and thinner on paediatric nutrition, intestinal-failure surgery, and inborn errors of metabolism. Several official requirement and screening papers that are historically central could not be assigned a verified PMID in this build and were therefore described as society documents or omitted from the numbered list rather than guessed. The literature harvest was live on 20 August 2026; later corrections and retractions after that date are not in the file.

Figures are original schematics. They are not care pathways, not measured mass-balances, and not dosing charts. No third-party published figure has been reproduced.

31 Glossary

AND/ASPEN. Academy of Nutrition and Dietetics / American Society for Parenteral and Enteral Nutrition adult-malnutrition consensus frame.

EN / PN. Enteral nutrition / parenteral nutrition.

GLIM. Global Leadership Initiative on Malnutrition; phenotype plus etiology after a screen; a consensus diagnostic frame, not a treatment trial.

MUST / NRS-2002 / MNA / NUTRIC / SGA. Screening or assessment instruments. NUTRIC is a severity construct; SGA is structured clinical diagnosis.

Permissive underfeeding (PermiT sense). Reduced non-protein calories with protein held similar in ventilated adults; not a clinic protein-restriction diet.

Swiss EFFORT vs EFFORT Protein. Ward individualised-support trial versus ICU higher-versus-usual protein trial. The name is shared; the experiment is not.