OBJECTIVE: There are considerable differences in published prediction algorithms for resting energy expenditure (REE) based on fat-free mass (FFM). The aim of the study was to investigate the influence of the methodology of body composition analysis on the prediction of REE from FFM. DESIGN: In a cross-sectional design measurements of REE and body composition were performed. SUBJECTS: The study population consisted of 50 men (age 37.1+/-15.1 years, body mass index (BMI) 25.9+/-4.1 kg/m2) and 54 women (age 35.3+/-15.4 years, BMI 25.5+/-4.4 kg/m2). INTERVENTIONS: REE was measured by indirect calorimetry and predicted by either FFM or body weight. Measurement of FFM was performed by methods based on a 2-compartment (2C)-model: skinfold (SF)-measurement, bioelectrical impedance analysis (BIA), Dual X-ray absorptiometry (DXA), air displacement plethysmography (ADP) and deuterium oxide dilution (D2O). A 4-compartment (4C)-model was used as a reference. RESULTS: When compared with the 4C-model, REE prediction from FFM obtained from the 2C methods were not significantly different. Intercepts of the regression equations of REE prediction by FFM differed from 1231 (FFM(ADP)) to 1645 kJ/24 h (FFM(SF)) and the slopes ranged between 100.3 kJ (FFM(SF)) and 108.1 kJ/FFM (kg) (FFM(ADP)). In a normal range of FFM, REE predicted from FFM by different methods showed only small differences. The variance in REE explained by FFM varied from 69% (FFM(BIA)) to 75% (FFM(DXA)) and was only 46% for body weight. CONCLUSION: Differences in slopes and intercepts of the regression lines between REE and FFM depended on the methods used for body composition analysis. However, the differences in prediction of REE are small and do not explain the large differences in the results obtained from published FFM-based REE prediction equations and therefore imply a population- and/or investigator specificity of algorithms for REE prediction.
OBJECTIVE: There are considerable differences in published prediction algorithms for resting energy expenditure (REE) based on fat-free mass (FFM). The aim of the study was to investigate the influence of the methodology of body composition analysis on the prediction of REE from FFM. DESIGN: In a cross-sectional design measurements of REE and body composition were performed. SUBJECTS: The study population consisted of 50 men (age 37.1+/-15.1 years, body mass index (BMI) 25.9+/-4.1 kg/m2) and 54 women (age 35.3+/-15.4 years, BMI 25.5+/-4.4 kg/m2). INTERVENTIONS: REE was measured by indirect calorimetry and predicted by either FFM or body weight. Measurement of FFM was performed by methods based on a 2-compartment (2C)-model: skinfold (SF)-measurement, bioelectrical impedance analysis (BIA), Dual X-ray absorptiometry (DXA), air displacement plethysmography (ADP) and deuterium oxide dilution (D2O). A 4-compartment (4C)-model was used as a reference. RESULTS: When compared with the 4C-model, REE prediction from FFM obtained from the 2C methods were not significantly different. Intercepts of the regression equations of REE prediction by FFM differed from 1231 (FFM(ADP)) to 1645 kJ/24 h (FFM(SF)) and the slopes ranged between 100.3 kJ (FFM(SF)) and 108.1 kJ/FFM (kg) (FFM(ADP)). In a normal range of FFM, REE predicted from FFM by different methods showed only small differences. The variance in REE explained by FFM varied from 69% (FFM(BIA)) to 75% (FFM(DXA)) and was only 46% for body weight. CONCLUSION: Differences in slopes and intercepts of the regression lines between REE and FFM depended on the methods used for body composition analysis. However, the differences in prediction of REE are small and do not explain the large differences in the results obtained from published FFM-based REE prediction equations and therefore imply a population- and/or investigator specificity of algorithms for REE prediction.
Authors: M Marra; I Cioffi; R Sammarco; C Montagnese; M Naccarato; V Amato; F Contaldo; F Pasanisi Journal: Int J Obes (Lond) Date: 2017-02-28 Impact factor: 5.095