Resting energy expenditure in children and adolescents with cerebral palsy: accuracy of available prediction formulas and development of population-specific methods

INTRODUCTION: Energy requirements are difficult to estimate in children with cerebral palsy (CP). Resting energy expenditure (REE), necessary to implement personalized nutritional interventions, is most commonly estimated using prediction formulae since indirect calorimetry, the reference method, is not available in all nutrition units. The aims of the present study were: (1) to evaluate the accuracy of the most commonly used REE prediction formulae developed for healthy children, in children with CP; (2) to assess the accuracy of the REE population-specific formula for CP children proposed in our preliminary report; (3) to develop new population-specific methods. METHODS: REE was measured by indirect calorimetry in 100 children and adolescents with spastic quadriplegic cerebral palsy (SQCP) and estimated on the basis of predictive formulas selected by the clinicians [World Health Organization (WHO), Harris-Benedict, Schofield weight, Schofield weight & height, Oxford, Mifflin formulae and a population-specific formula for CP children developed in our preliminary report]. RESULTS: 100 children with SQCP (35 girls, 35%) classified as level V according to gross motor function classification system (GMFCS-V); 64% with oral nutrition, 29% total enteral nutrition (nasogastric tube feeding, percutaneous endoscopic gastrostomy, percutaneous endoscopic transgastric jejunostomy) and 7% mixed nutrition. The median (IQR) REE was 41.96 (17.5) kcal/kg/day. Statistical analysis highlighted a proportional bias between the indirect calorimetry and all considered predictive formulae for REE determination. By studying the relationship between the bias and the mean values of REE, specific conversion equations were obtained. With a pre-specified model having as predictors the variable weight and the variable Triceps Skinfold (TSF) and, as response the variable REE measured by indirect calorimetry, a predictive nomogram was developed to estimate the REE in this population of children. CONCLUSIONS: We suggest using predictive formulae for healthy children with caution, and where possible carrying out indirect calorimetry to assess REE in children with CP. However, we propose a new tool which could be developed to become an additional help for assessment of REE in the clinical practice. Future objectives will be to obtain a larger sample size, in a multicenter perspective study, to build a specific predictive model for the REE of the studied population.