Many undernourished patients need nutritional support, which aims to increase lean body mass. Oral feeding can be difficult for some patients with anorexia or with eating or absorption problems. Behavioral measures that sometimes enhance oral intake include the following:

• Encouraging patients to eat
• Heating or seasoning foods
• Providing favorite or strongly flavored foods
• Encouraging patients to eat small portions
• Scheduling around meals
• Assisting patients with feeding

If behavioral measures are ineffective, nutritional support—oral, enteral tube, or parenteral nutrition—is indicated, except sometimes for dying or severely demented patients (see Nutritional Support: Nutritional Support for Dying or Severely Demented Patients).

Predicting Nutritional Requirements

Nutritional requirements are predicted so that interventions can be planned. Requirements can be estimated by formulas or measured by indirect calorimetry. Indirect calorimetry requires use of a metabolic cart (a closed rebreathing system that determines energy expenditure based on total CO₂ production), which requires special expertise and is not always available. Thus, total energy expenditure (TEE) and protein requirements usually are estimated.

Energy expenditure: TEE varies based on the patient's weight, activity level, and degree of metabolic stress (metabolic demands); TEE ranges from 25 kcal/kg/day for people who are sedentary and not under stress to about 40 kcal/kg/day for people who are critically ill. TEE equals the sum of

• Resting metabolic rate (RMR, or resting energy expenditure rate), which is normally about 70% of TEE
• Energy dissipated by metabolism of food (10% of TEE)
• Energy expended during physical activity (20% of TEE)

Clinical Calculator

Undernutrition can decrease RMR up to 20%. Conditions that increase metabolic stress (eg, critical illness, infection, inflammation, trauma, surgery) can increase RMR but rarely by > 50%.

The Mifflin–St. Jeor equation estimates RMR more precisely and with fewer errors than the commonly used Harris-Benedict equation, usually providing results that are within 20% of those measured by indirect calorimetry. The Mifflin–St. Jeor equation estimates RMR as follows:

TEE can be estimated by adding about 10% (for sedentary people) to about 40% (for people who are critically ill) to RMR.

Protein requirements: For healthy people, protein requirements are estimated at 0.8 g/kg/day. However, for patients with metabolic stress or kidney failure and for elderly patients, requirements may be higher (see Table 1: Nutritional Support: Estimated Adult Daily Protein Requirement).

Table 1
PrintOpen table
Estimated Adult Daily Protein Requirement Condition Requirement (g/kg of ideal body wt/day) Normal 0.8 Age > 70 yr 1.0 Kidney failure without dialysis (GFR < 25 mL/min/1.73 m2) 0.6–0.75 Kidney failure with dialysis 1.2 Metabolic stress (eg, critical illness, trauma, burns, surgery) 1.5

Assessing Response to Nutritional Support

There is no gold standard to assess response. Clinicians commonly use indicators of lean body mass such as the following:

• Body mass index (BMI)
• Body composition analysis
• Body fat distribution (see Undernutrition: Physical examination and Obesity and the Metabolic Syndrome: Body composition analysis)

Clinical Calculator

Nitrogen balance, response to skin antigens, muscle strength measurement, and indirect calorimetry can also be used.

Nitrogen balance, which reflects the balance between protein needs and supplies, is the difference between amount of nitrogen ingested and amount lost. A positive balance (ie, more ingested than lost) implies adequate intake. Precise measurement is impractical, but estimates help assess response to nutritional support. Nitrogen intake is estimated from protein intake: nitrogen (g) equals protein (g)/6.25. Estimated nitrogen losses consist of urinary nitrogen losses (estimated by measuring urea nitrogen content of an accurately obtained 24-h urine collection) plus stool losses (estimated at 1 g/day if stool is produced; negligible if stool is not produced) plus insensible and other unmeasured losses (estimated at 3 g).

Response to skin antigens, a measure of delayed hypersensitivity, often increases to normal as undernourished patients respond to nutritional support. However, other factors can affect response to skin antigens.

Muscle strength indirectly reflects increases in lean body mass. It can be measured quantitatively, by hand-grip dynamometry, or electrophysiologically (typically by stimulating the ulnar nerve with an electrode).

Levels of acute-phase reactant serum proteins (particularly short-lived proteins such as prealbumin [transthyretin], retinol-binding protein, and transferrin) sometimes correlate with improved nutritional status, but these levels correlate better with inflammatory conditions.

Key Points

• Behavioral measures may avert the need for nutritional support.
• Predict the patient's energy requirements based on weight, sex, activity level, and degree of metabolic stress (eg, due to critical illness, trauma, burns, or recent surgery).
• Normal protein requirement is 0.8 mg/kg/day, but this amount is adjusted if age is > 70 or if the patient has kidney failure or metabolic stress.
• Assess the response to nutritional support by indicators of lean body mass and/or other indicators (eg, nitrogen balance, response to skin antigens, muscle strength measurement, indirect calorimetry).

Last full review/revision April 2013 by David R. Thomas, MD

Content last modified April 2013