Content area
Full Text
International Journal of Obesity (2004) 28, 14791484
& 2004 Nature Publishing Group All rights reserved 0307-0565/04 $30.00www.nature.com/ijoPAPERValidation of the BodyGemt hand-held calorimeterEL Melanson1*, LB Coelho1, ZV Tran2, HA Haugen3, JT Kearney3 and JO Hill11Center for Human Nutrition, University of Colorado Health Sciences Center, Denver, CO, USA; 2Department of Preventive
Medicine and Biometrics, University of Colorado Health Sciences Center, Denver, CO, USA; and3HealtheTech, Inc., Golden,CO, USAOBJECTIVE: To assess the validity and reliability of a hand-held indirect calorimeter.DESIGN: Resting metabolic rate (RMR) was measured on two separate mornings.
SUBJECTS: A heterogeneous sample of 41 healthy adults.MEASUREMENTS: RMR using both a metabolic cart (Sensormedics 2900, SM-2900) and a hand-held indirect calorimeter
(BodyGemt, BG).RESULTS: There were no trial-to-trial differences in RMR measured by the BG (67567163 vs 66977163 kJ/day) or the SM-2900
(64007163 vs 63967167 kJ/day). RMR measured by the BG was significantly higher than that measured by the SM-2900
during both trials. In a sample of 10 subjects, the energy cost of holding the BG in position was determined to be (0.1770.04 kJ/
min, or 255784 kJ/day). After applying this adjustment, the differences between systems were no longer significant during trial
1 (mean difference 101767 kJ/day) or trial 2 (46775 kJ/day). In overweight and obese individuals, RMR measured by the
BodyGemt was more accurate than that estimated by the HarrisBenedict equations.CONCLUSION: The BodyGemt provides valid and reliable measurements of RMR. The BodyGemt produces significantly higher
values than the Sensor Medics 2900 indirect calorimeter, with the increase largely due to an increased energy demand required
to hold the BG in position.International Journal of Obesity (2004) 28, 14791484. doi:10.1038/sj.ijo.0802643Published online 24 August 2004Keywords: resting metabolic rate; metabolic cart; fat-free mass; HarrisBenedict equations; reliabilityIntroductionThe HarrisBenedict equations1 are widely used in clinicalsettings and nutritional assessment. Although the originalequations can be applied over a wide range of age and bodytypes to obtain a reasonable estimate of resting metabolicrate (RMR), only between 50 and 70% of the variability inRMR is explained by the equations, and the equationssystematically overestimate measured RMR by at least 5%.2Although numerous equations have been subsequentlydeveloped,37 they have generally not improved upon thislevel of random error.2 The use of metabolic carts is thestandard in research settings by which RMR is measured.However, because of the cost and complexity, metabolic cartsare not a...