Clinical Nutrition: Malnutrition, Critical Care & Perioperative Support

Citation model. Claims grounded in AllNutrition's trust-scored library carry an inline bracketed reference [n] linking to the References section, which lists each source's evidence level and AllNutrition trust score (0–1). Where an AllNutrition query returned an overall evidence_strength/consensus_level, those labels are surfaced in the Evidence Review. Only sources actually returned by the tool are cited; no trust scores are invented. The landmark trials EPaNIC, PermiT, EDEN, and NUTRIREA are named for historical context — they are inseparable from how ICU nutrition guidelines evolved — but claims attributed to them here are grounded in the guideline and meta-analytic sources that synthesize their findings, not in primary-trial citations the tool did not return.

---

1. Introduction

Malnutrition is not a peripheral finding in hospitalized patients — it is a disease-modifying comorbidity that clinicians routinely under-recognize. A patient admitted for pneumonia, decompensated heart failure, or a bowel resection carries a second, silent diagnosis whenever intake has been inadequate relative to requirement: disease-related malnutrition, which independently worsens infection rates, length of stay, and mortality [1]. Critical illness converts this baseline vulnerability into an acute metabolic emergency: the stress response to sepsis, trauma, or major surgery drives a hypercatabolic state that dismantles lean tissue for fuel, a process nutrition support can blunt but not fully reverse [2][3].

For a generation, ICU nutrition practice followed an intuitive syllogism: sick patients are catabolic, catabolism destroys protein and calories, therefore feed early and feed fully. That syllogism turned out to be wrong in ways that reshaped the field. Landmark trials in the 2010s — EPaNIC, PermiT, EDEN, and the NUTRIREA series — showed that early, full-dose nutrition (particularly parenteral) in the acute phase of critical illness could increase infections and prolong organ support without improving survival, while permissive underfeeding and hypocaloric strategies performed as well or better, a synthesis reflected in current ESPEN ICU guidance [4]. This module traces that paradigm shift: how malnutrition is diagnosed (GLIM) and screened (MUST, NRS-2002); why the EFFORT trial reframed individualized nutrition support in medical inpatients as mortality-relevant [5]; how enteral and parenteral routes are chosen; how calorie and protein dosing is timed across the phases of illness; how refeeding syndrome is prevented; and how perioperative nutrition, ERAS, and immunonutrition fit into surgical care of malnourished, sarcopenic, and frail patients.

2. Learning Objectives

By the end of this module, the learner will be able to:

1. Apply the GLIM criteria to diagnose malnutrition and describe the diagnostic performance of MUST and NRS-2002 as screening tools [3][6][7][8].
2. Describe the metabolic stress response to critical illness — its phases, hormonal drivers, and mechanisms of protein catabolism — and distinguish it from simple starvation [2][9][10].
3. Summarize the design and findings of the EFFORT trial and explain why it changed the standard of care for nutritional risk screening and support in medical inpatients [5][11].
4. Articulate the evidence-based indications for enteral versus parenteral nutrition, including the "if the gut works, use it" principle and its exceptions [4][12].
5. Explain the paradigm shift in ICU nutrition timing and dose — from early aggressive feeding toward permissive underfeeding — and its evidentiary basis [4][13][14].
6. Describe the current controversy over protein dosing in critical illness, including the EFFORT-Protein findings [15][16][17].
7. Explain the pathophysiology of refeeding syndrome and apply prevention strategies (thiamine, phosphate, potassium, caloric restriction) [18][8].
8. Apply perioperative nutrition and ERAS principles — carbohydrate loading, minimized fasting, and immunonutrition — and identify their evidence base and limits [19][20][21].
9. Recognize how sarcopenia and frailty modify surgical risk and describe prehabilitation strategies [22][23].

3. Scientific Foundations

3.1 The metabolic stress response and catabolism of critical illness

Critical illness triggers a neuroendocrine and inflammatory cascade fundamentally different from simple starvation. In starvation, the body adapts to conserve lean mass — insulin falls, glucagon and cortisol rise, and after roughly 24–36 hours of fasting the body shifts toward fat oxidation, sparing protein [18]. In critical illness this adaptive economy is overridden by cytokine-driven hypermetabolism: catecholamines, cortisol, and glucagon surge while insulin resistance develops, and skeletal muscle is broken down for gluconeogenic substrate regardless of energy delivery [2][9]. The illness trajectory moves through an early catabolic phase, a hypermetabolic phase that can last weeks, and — in severe or prolonged cases — a state of ongoing inflammation and treatment-resistant muscle wasting sometimes termed persistent inflammation-immunosuppression-catabolism, in which nutrition alone cannot reverse lean-mass loss [9]. Multicenter energy-expenditure studies show metabolic rate rising, peaking around day 10, then declining, with hypo-, normo-, and hypermetabolic phenotypes coexisting within a single ICU population — undermining one-size-fits-all caloric prescriptions [10]. In severe burns, beta-blockade can blunt the hypermetabolic response and reduce energy expenditure by roughly 20%, illustrating how strongly catecholamine drive shapes this state [9].

Muscle proteolysis persists even when nutrition is delivered, because inflammation and immobilization independently induce anabolic resistance — a blunted capacity of muscle to convert amino acids into new protein [2][16]. Amino acids also act as signaling molecules ("metabokines") beyond their role as building blocks, regulating the balance between synthesis and breakdown [2]. This is why increasing calories or protein early in illness does not straightforwardly reverse catabolism: timing, dose, and route must respect the phase-specific biology of the stress response rather than an intuitive "more is better" logic [16][17].

3.2 Disease-related malnutrition: definitions, GLIM, and prevalence

The Global Leadership Initiative on Malnutrition (GLIM) consensus, now on its five-year update, requires at least one phenotypic criterion (non-volitional weight loss, low BMI, or reduced muscle mass by validated measurement) and at least one etiologic criterion (reduced food intake/assimilation, or disease burden/inflammation); severity is graded solely by the phenotypic criteria [3]. GLIM explicitly advises against using hypoalbuminemia as a primary marker of malnutrition, since albumin reflects inflammation more than nutrient status [3]. Disease-related malnutrition (DRM) prevalence in hospitalized adults ranges widely by setting and criteria — roughly 19–37% in general internal medicine wards, 63–65% when GLIM/SGA criteria are applied broadly, and up to 85% in some resource-limited settings — and is consistently associated with longer length of stay, more than double the risk of infectious complications, and higher post-discharge mortality [1]. Inadequate food intake is the single strongest predictor of DRM (odds ratio ~5.1) [1].

3.3 Refeeding physiology

Refeeding syndrome is the mirror image of the adaptive starvation state: when carbohydrate is reintroduced after depletion, an abrupt insulin surge drives glucose, phosphate, potassium, and magnesium into cells via glycolysis and Na⁺/K⁺-ATPase activation, causing a precipitous fall in their serum concentrations despite whole-body depletion having been present all along [18]. Hypophosphatemia — a fall below 0.65 mmol/L, or a drop exceeding 0.16 mmol/L from baseline — is the hallmark warning sign, and thiamine, an essential cofactor for carbohydrate metabolism, is often already depleted in malnourished patients and further consumed by the refeeding glucose load [18]. Parenteral dextrose provokes a more pronounced insulin response than oral or enteral routes and is linked to more severe electrolyte deficiencies, one more reason enteral feeding is preferred when the gut is usable [18][8].

4. Clinical Relevance

Every hospitalized patient — medical, surgical, or critically ill — sits somewhere on a spectrum from adequately nourished to severely malnourished, and that position changes the risk calculus of nearly every other intervention: infection risk, wound healing, ventilator liberation, and discharge destination. The physician who fails to screen for malnutrition, or who defaults to reflexive "aggressive early feeding" in the ICU without regard to the phase of illness, is now practicing against the evidence. Conversely, the physician who understands GLIM, EFFORT, and the ICU nutrition-timing trials can identify at-risk patients within 24–72 hours [1], deliver route- and dose-appropriate nutrition that measurably reduces mortality [5], anticipate and prevent refeeding syndrome, and optimize a frail or sarcopenic surgical patient before an operation rather than reacting to its complications afterward [22][23].

5. Evidence Review

Established (high confidence):

• GLIM is a validated, consensus-based framework for diagnosing malnutrition using phenotypic and etiologic criteria; hypoalbuminemia should not be used as a primary marker. evidence_strength: moderate, consensus_level: moderate [3].
• Individualized nutritional support guided by risk screening (EFFORT trial, NRS-2002 ≥3) reduces 30-day mortality (adjusted OR 0.65) and adverse composite outcomes in at-risk medical inpatients. evidence_strength: strong, consensus_level: moderate [5][11].
• Enteral nutrition is preferred over parenteral when the gut is functional, with fewer infectious complications; parenteral nutrition is reserved for gut failure or when enteral routes cannot meet ≥50–60% of needs. evidence_strength: strong, consensus_level: mixed [4][12].
• Refeeding syndrome is preventable with anticipatory thiamine, cautious caloric advancement, and frequent phosphate/potassium/magnesium monitoring. evidence_strength: moderate, consensus_level: moderate [18][24].
• Preoperative carbohydrate loading and shortened fasting do not increase aspiration risk and reduce insulin resistance and discomfort; length-of-stay benefit is most consistent in major abdominal surgery. evidence_strength: strong, consensus_level: moderate [19][7].

Probable:

• Hypocaloric/permissive underfeeding in the first ICU week, followed by progressive advancement, is preferred over early full-target feeding — consistent with guideline synthesis of EPaNIC, PermiT, EDEN, and NUTRIREA-era evidence [4][13].
• Early oral/enteral feeding after most GI surgery is safe and speeds recovery without increasing anastomotic leak (esophageal/cervical anastomoses remain a cautious exception) [6][7].
• Combining early nutrition with early mobilization better preserves muscle mass and function than either alone in the ICU [22][2].

Emerging:

• Adapting screening tools (removing disease-severity items) may better predict response to nutrition intervention, distinct from tools optimized purely for mortality prognosis [11].
• Metabolic-phenotyping (persistent hypo-, normo-, hypermetabolism) to individualize energy dosing beyond fixed weight-based equations [10].
• Vasopressor-dose thresholds (e.g., norepinephrine ≥0.3–0.5 µg/kg/min) as decision points for enteral nutrition in shock. evidence_strength: strong, consensus_level: mixed [25][4].

Controversial:

• Optimal protein dose: observational data link higher intake (~1.5–2.0 g/kg/day) to better survival, but the EFFORT-Protein RCT found high-dose protein (≥2.2 g/kg/day) did not improve discharge-alive time or 60-day mortality, and worsened outcomes in AKI patients off renal replacement therapy. evidence_strength: strong, consensus_level: moderate [15][16][17].
• Immunonutrition reduces infections and length of stay in major GI cancer surgery meta-analyses, but evidence quality is rated low by some sources and optimal timing/duration is unsettled. evidence_strength: strong, consensus_level: moderate [20][6].
• Early enteral nutrition in septic shock/high vasopressor states: mixed mortality signal across cohorts, reflecting confounding by indication and unresolved RCT-level uncertainty [25].

Unsupported / overstated:

• That early, full-target caloric and protein delivery in acute critical illness is uniformly beneficial — RCTs and guideline syntheses show it can increase infections and organ-support duration, and worsen outcomes in some subgroups [4][17].
• Routine reliance on serum albumin as a nutritional marker; it reflects inflammation, not intake [3].

6. Practical Clinical Applications

Screening and diagnosis. Screen all hospitalized patients within 24–72 hours of admission using a validated tool — NRS-2002 for acute/surgical/critical care populations (AUC ~0.80–0.85 against GLIM in several cohorts) or MUST in general wards [7][12]. Weight- and BMI-based tools including MUST are recognized to under-detect risk in fluid-overloaded states, since edema can mask weight loss [12]. Confirm the diagnosis with GLIM (≥1 phenotypic + ≥1 etiologic criterion); grade severity from the phenotypic criteria alone [3].

Route selection. Default to oral intake → oral nutritional supplements → enteral nutrition before parenteral nutrition whenever the gut is functional [4][6]. Reserve parenteral nutrition for gut failure, bowel ischemia, high-output fistula without distal access, or when enteral nutrition cannot deliver ≥50–60% of needs after the first week, monitoring for catheter-related and metabolic complications once started [4][6]. Delay or minimize enteral nutrition during uncontrolled shock, active GI bleeding, bowel ischemia, or very high vasopressor requirements; resume low-dose (trophic) feeding once perfusion targets are met, typically within 24–48 hours [4][25].

Timing and dose in the ICU. Avoid early full-target feeding in the first ICU week; favor hypocaloric or trophic dosing initially, advancing toward 70–100% of measured (or estimated) energy expenditure by day 4–7, consistent with the guideline synthesis of EPaNIC/PermiT/EDEN/NUTRIREA-era evidence [4][14]. Use indirect calorimetry where available; otherwise use adjusted body weight in obese patients, since standard predictive equations agree poorly with measured expenditure [10]. Target protein delivery of ~1.2–1.5 g/kg/day, progressively achieved, with caution against routinely pushing to ≥2.2 g/kg/day, particularly with acute kidney injury off renal replacement therapy [15][16][17]. In obesity, favor an iso-caloric, high-protein strategy using adjusted body weight, since sarcopenic obesity is easily missed [26].

Refeeding syndrome prevention. Identify high-risk patients (significant weight loss, prolonged low/no intake, alcohol use disorder, chronic malnutrition) before starting nutrition. Give thiamine before or with the first feed; restrict initial calories (commonly 20–25 kcal/kg/day or lower, sometimes 40–50% of goal); advance over 3–5 days while monitoring phosphate, potassium, and magnesium at least daily during the highest-risk window [18][24][8].

Perioperative and surgical nutrition. Apply ERAS carbohydrate loading (~800 mL of a 12–12.5% maltodextrin drink the evening before surgery, ~400 mL two hours before induction); permit clear fluids to 2 hours and solids to 6 hours preoperatively for most elective patients [19][7]. Resume oral/enteral intake within 24 hours (often 6–12 hours) after most GI surgery; chewing gum, coffee, and early mobilization support return of bowel function [6]. Consider immunonutrition (arginine, omega-3s, nucleotides) for 5–7 days preoperatively before major GI cancer surgery, recognizing the evidence is protocol-dependent (Grade B) [20][6]. Screen for sarcopenia and frailty before elective major surgery (SARC-F/SARC-CalF, CT muscle assessment, grip strength) and offer multimodal prehabilitation (protein/energy optimization, vitamin D repletion, exercise) — even one-week programs have reduced 30-day complication rates in some cohorts [22][23].

When not to push nutrition support: uncontrolled shock, active/uncontrolled GI hemorrhage, bowel ischemia or obstruction, and, for aggressive dosing specifically, uncorrected severe electrolyte derangement suggestive of refeeding risk.

7. Clinical Pearls

• "If the gut works, use it" is not a slogan to apply reflexively — but when the gut is working, enteral nutrition should almost always win over parenteral.
• Early in critical illness, restraint is a therapeutic choice, not a failure to act; the evidence base overturned "more is better" for both calories and, in some subgroups, protein.
• A normal serum phosphate on hospital day 1 in a malnourished patient tells you nothing about refeeding risk — the danger appears after feeding begins.
• High protein prescriptions in a patient with acute kidney injury not yet on dialysis deserve a second look; EFFORT-Protein suggests this may be a subgroup where less is more.
• Malnutrition screening is only useful if it changes practice within 24–72 hours — a NRS-2002 score filed away unread saves no one.
• Preoperative carbohydrate loading is one of the few interventions in this module where doing something 2 hours before surgery — rather than nothing for 12 — is both safer and more effective.

8. Common Misconceptions

• "Sicker ICU patients need more calories, sooner." Trial evidence (EPaNIC, PermiT, EDEN, NUTRIREA-era studies, synthesized in current ESPEN guidance) shows early full-target feeding — especially parenteral — can increase infection and organ-support duration without a mortality benefit [4].
• "More protein is always better in critical illness." EFFORT-Protein and related meta-analyses found no consistent benefit from very high protein doses, and possible harm in AKI subgroups [15][16][17].
• "Albumin is a nutrition marker." GLIM explicitly discourages this; albumin tracks inflammation, not intake [3].
• "NPO after midnight is the safe default before surgery." Prolonged fasting increases metabolic stress and dehydration without reducing aspiration risk compared to carbohydrate loading and 2-hour clear-fluid rules [19][7].
• "Refeeding syndrome only happens with obvious starvation (anorexia nervosa)." It also occurs with unrecognized chronic poor intake, alcohol use disorder, oncology patients, and prolonged NPO medical/surgical inpatients [18][24].
• "Gastric residual volume monitoring prevents aspiration." Contemporary guidance notes routine GRV checks do not reliably reduce aspiration risk and may promote underfeeding [8].

9. Summary

Malnutrition is a common, underdiagnosed, and independently harmful comorbidity across medical, surgical, and critically ill inpatients, now diagnosable in a standardized way via GLIM and detectable early via NRS-2002 or MUST screening. The EFFORT trial demonstrated that acting on that risk — individualized, dietitian-guided nutritional support started within 48 hours — measurably reduces mortality and adverse outcomes in medical inpatients. Critical illness superimposes a distinct hypercatabolic stress response on this baseline risk, and the ICU nutrition literature of the last fifteen years overturned the intuitive "feed early, feed fully" paradigm: EPaNIC, PermiT, EDEN, and the NUTRIREA trials, together with EFFORT-Protein, showed that permissive underfeeding, delayed/hypocaloric initiation of parenteral nutrition, and moderate rather than maximal protein dosing generally produce equal or better outcomes than aggressive early feeding, particularly in the first ICU week and in patients with acute kidney injury. Enteral nutrition remains preferred whenever the gut is usable, parenteral nutrition is reserved for gut failure, and refeeding syndrome — a predictable, preventable complication of reintroducing nutrition after depletion — requires anticipatory thiamine, cautious caloric advancement, and vigilant electrolyte monitoring. In the perioperative setting, ERAS principles (carbohydrate loading, minimized fasting, early postoperative feeding, and selective immunonutrition) improve recovery, and sarcopenia/frailty screening with multimodal prehabilitation is increasingly central to surgical risk reduction. Across all these settings, the unifying clinical skill is matching the route, timing, and dose of nutrition to the specific phase of illness and the specific patient — not applying a single reflexive strategy regardless of context.

10. References

Ordered by evidence strength / relevance. Evidence level and AllNutrition trust score (0–1) as returned by the tool.

1. Prevalence, Risk Factors, and Clinical Management of Disease-Related Malnutrition in Hospitalized Patients. Nutrients (2024). Observational — trust 0.727.
2. Amino Acids as Metabokines in Hypercatabolic States: Rethinking Nutritional Protein-Based Strategies Beyond Caloric Support. Nutrients (2026). Review — trust 0.8.
3. The GLIM consensus approach to diagnosis of malnutrition: A 5-year update. Clinical Nutrition (2025). Guideline — trust 0.922.
4. ESPEN guideline on clinical nutrition in the intensive care unit. Clinical Nutrition (2018). Guideline — trust 0.828.
5. Individualised nutritional support in medical inpatients at nutritional risk: a randomised clinical trial (EFFORT). The Lancet (2019). RCT — trust 0.791.
6. ESPEN guideline on clinical nutrition in surgery – Update 2025. Clinical Nutrition (2025). Guideline — trust 0.907.
7. ESPEN guideline: Clinical nutrition in surgery. Clinical Nutrition (2017). Guideline — trust 0.82.
8. Malnutrition and Cachexia in Inpatients With Acute Cardiac Conditions: A Scientific Statement From the American Heart Association. Circulation (2026). Guideline — trust 0.917.
9. What does "PICS" mean in major burns? Persistent critical illness or post-intensive care syndrome. Burns (2026). Review — trust 0.73.
10. Time course of energy expenditure in persistent critical illness: a prospective multicentre study. ClinicalTrials.gov (2026). Observational — trust 0.85.
11. Adaptation of nutritional risk screening tools may better predict response to nutritional treatment: secondary analysis of EFFORT. American Journal of Clinical Nutrition (2024). RCT — trust 0.825.
12. Prospective validation of five malnutrition screening and assessment instruments among medical inpatients. Clinical Nutrition (2022). RCT — trust 0.785.
13. Nutritional support after hospital discharge improves long-term mortality in malnourished adult medical patients: Systematic review and meta-analysis. Clinical Nutrition (2022). Systematic review — trust 0.81.
14. Body weight definitions for estimating energy expenditure in intensive care, a prospective monocentric observational study. Clinical Nutrition ESPEN (2026). Observational — trust 0.887.
15. High-protein supplementation in critically ill patients: a systematic review, meta-analysis and umbrella review of existing evidence. Frontiers in Nutrition (2026). Systematic review — trust 0.812.
16. The impact of protein intake on kidney adverse events in critically ill patients: a systematic review and meta-analysis. International Urology and Nephrology (2025). Systematic review — trust 0.738.
17. Low-protein diet for chronic kidney disease: Evidence, controversies, and practical guidelines (includes EFFORT-Protein trial synthesis). Journal of Internal Medicine (2025). Review — trust 0.74.
18. Nutritional Support Strategies for Refeeding Syndrome in ICU Patients: A Review of Current Evidence. Journal of Multidisciplinary Healthcare (2026). Review — trust 0.725.
19. Preoperative carbohydrate loading: evolution, trends, and future directions. Frontiers in Nutrition (2026). Systematic review — trust 0.842.
20. Immunonutrition Decreases Postoperative Complications in Gastrointestinal Cancer - A Systematic Review and Meta-analysis of Randomized Controlled Trials. Advances in Nutrition (2026). Systematic review — trust 0.825.
21. Nutrition and Exercise Interventions During Hospitalization in Frail or Sarcopenic Patients: A Scoping Review of Intervention Configurations and Evidence Gaps. Nutrients (2026). Review — trust 0.925.
22. Association between malnutrition and frailty and postoperative mortality in older surgical patients. Clinical Nutrition (2026). Observational — trust 0.7.
23. Editorial: The role of nutrition in enhancing surgical recovery and outcomes. Frontiers in Nutrition (2026). Review — trust 0.867.
24. Protein nutritional support in critically ill patients: pathophysiological basis, clinical evidence, and areas of uncertainty – a narrative review. Frontiers in Medicine (2026). Review — trust 0.715.
25. Association between vasopressor dose and feeding intolerance in critically ill patients receiving enteral nutrition: A retrospective cohort study from the MIMIC-IV database. Clinical Nutrition ESPEN (2026). Observational — trust 0.637.
26. Nutrition Therapy in Critically Ill Patients with Obesity: An Observational Study. Nutrients (2025). Observational — trust 0.712.

Supporting sources also surfaced: ESPEN guideline on nutritional support for polymorbid medical inpatients (Clinical Nutrition 2023, guideline, trust 0.897); ESPEN practical guideline: nutritional support for polymorbid medical inpatients (Clinical Nutrition 2024, guideline, trust 0.89); NRS2002 outperforms GNRI and PG-SGA SF (Nutrition 2026, observational, trust 0.785); Optimal timing of enteral nutrition initiation — network meta-analysis (Frontiers in Nutrition 2026, systematic review, trust 0.818); Early supplemental parenteral nutrition shortens ventilation and ICU stay in patients ≥60 years (Clinical Nutrition ESPEN 2026, RCT, trust 0.835); Consumption of a 12.5% carbohydrate after 8h fasting does not delay gastric emptying (Frontiers in Nutrition 2026, RCT, trust 0.853); Intestinal Rehabilitation Plan for Patients with Colorectal Cancer (J Multidiscip Healthc 2026, systematic review, trust 0.842); Early versus delayed enteral nutrition in septic shock: target trial emulation (Frontiers in Nutrition 2026, observational, trust 0.65).