When to Use Schofield, WHO, or PDRI Equations in Pediatric Energy Expenditure Estimation
Use the Schofield equation (weight and height) as the preferred method for calculating resting energy expenditure (REE) in pediatric patients, as it is least likely to underestimate REE compared to measured values. 1
Primary Recommendation: Schofield Equation
The ESPGHAN/ESPEN/ESPR/CSPEN guidelines explicitly recommend the Schofield equation for calculating REE in pediatric patients. 1 This recommendation is based on evidence showing that the Schofield equation using both weight and height was least likely to underestimate REE when compared to measured REE by indirect calorimetry. 1, 2, 3
Schofield Equation Formulas by Age:
- Ages 0-3 years: Boys: 59.5 × (weight in kg) + 30 kcal/day; Girls: 58.3 × (weight in kg) + 31 kcal/day 1, 4
- Ages 3-10 years: Boys: 22.7 × (weight in kg) + 504 kcal/day; Girls: 20.3 × (weight in kg) + 486 kcal/day 1, 4
- Ages 10-18 years: Boys: 17.7 × (weight in kg) + 658 kcal/day; Girls: 13.4 × (weight in kg) + 692 kcal/day 1, 4
WHO Equation: Historical Context Only
The WHO equations (derived from 1985 and 2004 FAO/WHO/UNU recommendations) were used in previous ESPGHAN guidelines from 2005, but current 2018 guidelines have shifted preference to the Schofield equation. 1 While WHO equations showed similar mean bias to Schofield equations at the population level, research demonstrates they are not accurate enough for individual patient use in hospitalized children. 5
PDRI: Not a Calculation Method
PDRI (Pediatric Dietary Reference Intake) is not an equation for calculating energy expenditure—it represents reference values for total energy requirements. The evidence provided does not support PDRI as a calculation method comparable to Schofield or WHO equations. The current guidelines base their energy recommendations on calculated REE using Schofield equations, then adding factors for physical activity and growth. 1
Critical Situations Requiring Indirect Calorimetry Instead
Do not rely on any predictive equation in the following scenarios—measure REE directly with indirect calorimetry: 1
- Children with suspected metabolic alterations 1, 4
- Severe malnutrition 1
- Moderate to severe failure to thrive in young infants (birth to 3 years) 3
- When initial weight management approaches have been unsuccessful 4
- Critically ill ventilated children, as all predictive equations (including Schofield) fail to predict within clinically accepted ranges in this population 6
Adjusting REE to Total Energy Expenditure
After calculating REE with Schofield equation, adjust for total energy needs: 1
- Physical activity level (PAL): Multiply REE by 1.2-1.3 for sedentary/hospitalized patients, 1.5 for light activity, 1.7 for moderate activity, or 2.0 for vigorous activity 4
- Growth requirements: Add 20 kcal/day for ages 6-12 months and prepubertal children, or 30 kcal/day during peak pubertal growth 4
- Disease factors: Adjust for conditions that increase (fever, inflammation, chronic disease) or decrease (hypothermia) REE 1
- Catch-up growth: Add additional calories based on growth deficit when recovering from malnutrition 1
Common Pitfalls to Avoid
All predictive equations have significant individual variability: Even the preferred Schofield equation accurately predicts REE in only approximately 40% of individual patients, despite good population-level accuracy. 2, 3 The standard deviation of bias ranges from 214-286 kcal/day across different equations. 5
Younger age and severe growth failure increase prediction error: Prediction equations are more likely to underestimate REE in infants under 3 years and children with moderate to severe failure to thrive. 3
Use actual body weight, not ideal weight: The Schofield equations were developed using actual weight, and body weight is the main predictor of energy expenditure. 4
Hospitalized children are not the same as healthy children: Research shows that WHO, Harris-Benedict, Schofield, and Oxford formulae should not be used to estimate REE in hospitalized children at the individual level due to large variability. 5