Objective: Accuracy of hand-held calorimeters and predictive equations in the assessment of energy requirements have not been studied for the Saudi population. The aims of the present study were to assess the accuracy of resting energy expenditure (REE) measurement in a sample of overweight and obese Saudi males, using the BodyGem device (BG) with Whole Room Calorimetry (WRC) as a reference; and to evaluate the accuracy of predictive equations.
Methods: 38 subjects (mean±SD, age 26.8±3.7 years, BMI, 31.0±4.8) were recruited during the period from 5 February 2007 to 28 March 2008. REE was measured with WRC and BG and also calculated using 7 prediction equations. Mean differences, bias, percent of bias, accurate estimation, underestimation and overestimation were calculated.
Results: Repeated measures with the BG were not significantly different (accurate prediction: 81.6%; bias %; 1.1±6.3, p > 0.24) with limits of agreement ranging from + 242 to – 200 kcal. REE measured by BG was significantly less than WRC values (accurate prediction: 47.4%; bias %; 11.0±14.6, p < 0.05) with unacceptably wide limits of agreement. Harris-Benedict, Schofield and WHO equations were the most accurate, estimating REE within 10% of measured REE, but none seem appropriate to predict the REE of individuals.
Conclusion: There was poor agreement between the REE measured by WRC compared to BG or predictive equations. The BG assessed REE accurately in 47.4% of the subjects on an individual level.