Determination Of Resting Energy Expenditure And Accuracy Of Predictive Equations For Overweight And Obese People

Objective:Indirect calorimetry(IC)is used to measure the resting energy expenditure(REE)of overweight and obese people.At the same time,using various prediction equations to estimate and compare the values,choosing the REE prediction equation suitable for clinical application of Chinese overweight and obese people.Methods:221 subjects were recruited for light physical labor and no underlying disease,94 males(normal weight: 11;overweight: 29;obesity: 54),127 females(normal weight:29;overweight:49;obesity:49).The body composition was determined by bioelectrical impedance analysis(BIA)method,and the level of REE was measured by indirect calorimetry(IC)and was estimated using these 29 REE prediction equations.The REE measured by indirect calorimetry is the gold standard,and the accuracy of the REE prediction equations is analyzed.When predicted REE ranged between 90% to 110% of REE measured by indirect calorimetry,it was considered as accurate prediction.Predicted REE lower that 90% of measured REE was considered as an underprediction and predicted REE higher that 110% of mREE as an overprediction.The concordance between the REE measured value and the predicted value of the equations is tested by correlation and Bland-Altman method to evaluate the estimation effect of the existing REE prediction equations.Results:The median age of the subjects was 32 years and the interquartile range was 14.In the same BMI group,male height,weight,fat free mass,body cell volume,abdominal circumference,upper arm circumference,upper arm muscle circumference and measured REE values were significantly higher than females(P<0.05),while female body fat mass and body fat percentage were significantly higher than that of males(P<0.05).The measured REE in the normal weight group was most strongly correlated with the upper arm muscle circumference(r=0.730);The measured REE in the overweight group was most strongly correlated with body cell volume(r=0.727);The measured REE in the obese group was most strongly correlated with fat free mass(r=0.794);There was a negative correlation between the measured REE and body fat mass in normal weight and overweight groups(r=-0.320;r=-0.253).There was no significant correlation between the measured REE and body fat mass in obese group(P=0.077).The measured REE of the overweight and obese groups was higher than that of the normal weight,and the difference was statistically significant(P<0.001).The obese group was higher than the overweight group(P<0.001).The difference between the measured values and predicted values of the Lazzer(BC)equation,the Bernestin et al equation,and the Bernestin(BC)equation was between-4% and 8%.And the difference was not statistically significant(P>0.05).The remaining REE prediction equation estimates were significantly higher than the measured values(P<0.001).In all people,the highest accuracy of Bernestin(BC)equation was 71%(r=0.836,P<0.001),the predicted value of the Lazzer(BC)equation was negatively correlated with the measured value(r=-0.169,P=0.012).The Livingston equation was the most strongly correlated with the measured REE in the male normal weight group(r=0.755,P<0.05),and the predicted accuracy was 18.2%;There was no significant correlation between the predicted values of all prediction equations and the measured values in male overweight group(P>0.05).The predicted values of all prediction equations were significantly correlated with the measured values in the male obese group(r=0.595-0.718,P<0.001),and the highest accuracy of Bernestin(BC)was 70.4%(r=0.706,P<0.001).There was no significant correlation between all the predicted values and the measured values in the female normal weight group(P>0.05).The highest accuracy of the Bernestin(BC)equation was 79.6%(r=0.439,P=0.002)in the female overweight group.The predicted values of all REE predictive equations were significantly correlated with the measured values(r=0.580-0.662,P<0.001)in the female obese group,and the highest accuracy of the Bernestin equation was 69.4%(r=0.613,P<0.001).The Bland-Altman method is used to test the consistency between the measured values and the predicted values of these 29 REE prediction equations.The residual of the measured values and the predicted values of all predictive equations is(-419.2~154.2)kcal/d.The 95% limits of agreement of the residual between all predicted values and measured values is beyond the clinically acceptable range.Conclusion: The Bernestin(BC)prediction equation is closest to indirect calorimetry in male overweight/obese and female overweight people,and the Bernestin prediction equation is closest to indirect calorimetry in female obese people.The Bernestin(BC)prediction equation is closest to indirect calorimetry in all normal weight,overweight,and obese people.Most predictive equations overestimate REE in normal,overweight,and obese subjects,and the Lazzer(BC)equation underestimated REE in all subjects.Therefore,it is not recommended to use the prediction equation to estimate the individual REE,but when the indirect calorimetry method cannot be used,the Bernestin(BC)and Bernestin equations are more accurate.The main factors affecting REE in overweight/obese people are fat free mass and body cell volume.The effect of body fat mass on REE is not clear.REE in overweight/obese people is higher than normal weight,and obese people is higher than overweight people.