Nutritional Assessment in Critically Ill Patients

Introduction to Nutritional Markers

The American College of Cardiology and other guideline societies recommend that neither prealbumin nor albumin should be used as primary markers of nutritional status in critically ill adults with suspected malnutrition, impaired liver function, and impaired kidney function, because both are negative acute-phase reactants that reflect inflammation rather than true nutritional state, and both are confounded by renal and hepatic dysfunction 1, 2

Prealbumin has a half-life of 2-3 days, making it theoretically more responsive to acute nutritional changes compared to albumin's 20-day half-life, in patients with normal liver and kidney function 1
The normal threshold for prealbumin is ≥30 mg/dL, and for albumin is ≥4.0 g/dL in stable chronic kidney disease patients, according to the American Journal of Kidney Diseases 1, 2

Both proteins are negative acute-phase reactants that decline during inflammation, infection, or critical illness regardless of actual nutritional intake, as stated by the American College of Cardiology 1, 2
Inflammatory cytokines directly downregulate hepatic synthesis of both proteins, even when protein and caloric intake are adequate, in patients with impaired liver function 3
An elevated C-reactive protein negates the relationship between these proteins and actual protein intake, in critically ill patients 2
Prealbumin levels are artificially elevated in renal failure due to impaired kidney degradation, making interpretation unreliable in patients with impaired kidney function 1

The Nutritional Risk Screening 2002 (NRS-2002) should be used, which incorporates BMI, weight loss, food intake, and disease severity, as recommended by the American Society for Parenteral and Enteral Nutrition 2
Subjective Global Assessment (SGA) has been shown to be more reliable than albumin in detecting protein-energy wasting, in patients with malnutrition 3
Body composition assessment should be preferred to biochemical markers when diagnosing and monitoring malnutrition, according to the American College of Nutrition 3

Food intake should be monitored daily using semi-quantitative methods in malnourished patients, as recommended by the American Dietetic Association 2
Nutritional intervention should be triggered when intake is ≤50% of energy requirements over 3 days, in critically ill patients 2
C-reactive protein should be measured alongside any visceral protein measurement to interpret inflammatory status, as stated by the American College of Cardiology 2

Low serum albumin is strongly associated with both mortality and cardiac disease in chronic kidney disease patients, according to the American Journal of Kidney Diseases 3
In hemodialysis patients, death risk increases by 6% for every 0.1 g/dL decrease in serum albumin, as reported by the American Society of Nephrology 3

Albumin or prealbumin should not be interpreted in isolation without considering inflammatory status, as recommended by the American College of Cardiology 2
Hypoalbuminemia should not be assumed to be solely due to malnutrition when inflammation may be the primary driver, in critically ill patients 3
Prealbumin levels should not be used to assess nutritional status in renal failure patients due to artificial elevation, as stated by the American Journal of Kidney Diseases 1

Prealbumin has a shorter biological half‑life (approximately 2–3 days) compared with albumin (≈ 20 days), which has been proposed as a reason it might respond more quickly to acute nutritional changes【4】【5】.
However, systematic reviews of the available studies conclude that there is insufficient evidence to demonstrate that prealbumin is a more sensitive or accurate index of nutritional status than albumin in critically ill adults【4】【5】【6】 (evidence level: low/insufficient).

In patients with impaired kidney function, prealbumin concentrations are often artificially elevated because reduced renal catabolism slows its clearance; therefore, prealbumin values are unreliable for nutritional assessment in this population【4】【5】 (evidence level: moderate).

Studies have shown that changes in prealbumin do not consistently parallel alterations in other objective nutritional measures, indicating limited utility as a standalone marker【4】【6】 (evidence level: moderate).