Accuracy of resting energy expenditure calculations in unselected overweight and obese patients.

AIMS: Measurements of resting energy expenditure (REE) were compared with the data of 14 equations to determine their accuracy. METHODS: REE measurements by indirect calorimetry in 1,032 unselected overweight and obese men (n = 306) and women (n = 726) were compared with calculations by 14 different formulas. RESULTS: The mean (± SD) values calculated with the Owen, Robertson and Reid and WHO-I equations were not significantly different from our measurement of 1,682 ± 441.9 kcal/24 h. The values obtained with the Livingston, Mifflin, Müller and Bernstein equations were significantly different but still within a range of ±100 kcal/24 h. For females, the best comparison was observed with the Müller equation which, however, differed substantially in males. For men, the Cunningham equation was best, but it gave the worst comparison in women. A good individual match was only obtained with the equation of Robertson and Reid in 34% of the men and with the Owen equation in 38% of the women. All other formulas were less accurate. Drug treatment for 55% of the subjects had no effect on the mismatch between calculated and measured data. CONCLUSION: Calculations of REE with most equations seem to be valid in a group analysis but they are not helpful for the estimation of an obese patient's individual energy expenditure.