Objective: To compare standardized prediction equations to a hand-held indirect calorimeter in estimating resting energy and total energy requirements in overweight women.
Design: Resting energy expenditure (REE) was measured by hand-held indirect calorimeter and calculated by prediction equations Harris-Benedict, Mifflin-St Jeor, World Health Organization/Food and Agriculture Organization/United Nations University (WHO), and Dietary Reference Intakes (DRI). Physical activity level, assessed by questionnaire, was used to estimate total energy expenditure (TEE).
Subjects: Subjects (n=39) were female nonsmokers older than 25 years of age with body mass index more than 25.
Statistical analyses: Repeated measures analysis of variance, Bland-Altman plot, and fitted regression line of difference. A difference within +/-10% of two methods indicated agreement.
Results: Significant proportional bias was present between hand-held indirect calorimeter and prediction equations for REE and TEE (P<0.01); prediction equations overestimated at lower values and underestimated at higher values. Mean differences (+/-standard error) for REE and TEE between hand-held indirect calorimeter and Harris-Benedict were -5.98+/-46.7 kcal/day (P=0.90) and 21.40+/-75.7 kcal/day (P=0.78); between hand-held indirect calorimeter and Mifflin-St Jeor were 69.93+/-46.7 kcal/day (P=0.14) and 116.44+/-75.9 kcal/day (P=0.13); between hand-held indirect calorimeter and WHO were -22.03+/-48.4 kcal/day (P=0.65) and -15.8+/-77.9 kcal/day (P=0.84); and between hand-held indirect calorimeter and DRI were 39.65+/-47.4 kcal/day (P=0.41) and 56.36+/-85.5 kcal/day (P=0.51). Less than 50% of predictive equation values were within +/-10% of hand-held indirect calorimeter values, indicating poor agreement.
Conclusions: A significant discrepancy between predicted and measured energy expenditure was observed. Further evaluation of hand-held indirect calorimeter research screening is needed.