Müller B, Merk S, Bürgi U, Diem P
Abteilung für Endokrinologie und Diabetologie, Universität Bern, Inselspital, Bern.
Praxis (Bern 1994). 2001 Nov 8;90(45):1955-63.
The aim of this study was to evaluate the currently available predictive equations for basal metabolic rate (BMR) in subjects with obesity class II and III, and to assess the contribution by the components of a two-compartment model of body composition, namely the lean body mass (LBM) and the fat mass (FM) to the prediction. A second objective was to examine the reliability of the Harris Benedict equation in obese subjects, especially with a weight > or = 120 kg.
In 43 patients (age range 18 to 61 years, 5 men, 38 women) with obesity class II and III (body mass index, BMI, mean +/- SD 45.6 kg/m2 +/- 5.4 kg/m2, range 37.1-58.6 kg/m2) basal metabolic rate BMR was determined using indirect calorimetry (Deltatrac MBM, Datex, Instrumentarium Corp., Helsinki, Finnland) and the components of body composition were determined using the bioelectrical-impedance-analysis (BIA) method (BIA, Akern-Gerät, RJL Systems, Detroit). Calculated BMR was compared with measured BMR.
The best fitting equations for predicting BMR in these 43 severe und morbidly obese subjects were the Harris-Benedict (ratio calculated BMR to measured, BMR mean +/- SD%; correlation coefficient r = 101 +/- 12.9; 0.69), the Jensen (101.5 +/- 12.3; 0.74), the Nelson (99.3 +/- 11.4; 0.76) and the Cunningham equation (98.9 +/- 11.7; 0.74). The predictive value of the original Harris-Benedict equation was slightly different from modified Harris-Benedict equation, which was recalculated by Roza et al. (101.1 +/- 12.9; 0.69 vs. 99.7 +/- 12.8; 0.69). In the group of the 22 subjects with a body weight > or = 120 kg ratio of estimated values for BMR using original Harris-Benedict equation to measured BMR was 102.2 +/- 15.4% (mean +/- SD%, r = 0.61), respectively 93.2 +/- 14.5% (r = 0.50) when weight was set at 120 kg due to current recommendations. The ratio calculated BMR/measured BMR according to the Nelson equation in this subgroup was 101.0 (12.1/0.74).
In patients with obesity class II and III the equation of Harris-Benedict predicted the average BMR with acceptable precision for clinical use and was better fitting than most of the currently available predictive equations for basal metabolic rate (BMR). However, the recalculated version (by Roza et al.) was more accurate and should therefore be used instead of the original equation: BMR (men) = 88.362 + 4.799 x (length) + 13.397 x (weight) - 5.677 x (age); BMR (women) = 447.593 + 3.098 x (length) + 9.247 x (weight) - 4.330 x (age). The Nelson equation, including not only LBM but FM as additional predictor, was the best predicting equation ([108 LBM + 16.9 FM]0.239). Harris-Benedict equation had sufficient precision also in extreme obese subjects with a body weight > or = 120 kg, so there is no need for adaptation.
本研究旨在评估目前可用于预测II级和III级肥胖受试者基础代谢率（BMR）的方程，并评估身体成分双室模型的组成部分，即瘦体重（LBM）和脂肪量（FM）对预测的贡献。第二个目的是检验哈里斯-本尼迪克特方程在肥胖受试者中的可靠性，尤其是体重≥120 kg的受试者。
对43例II级和III级肥胖患者（年龄范围18至61岁，5名男性，38名女性）（体重指数，BMI，均值±标准差45.6 kg/m²±5.4 kg/m²，范围37.1 - 58.6 kg/m²），使用间接测热法（DeltaTrac MBM，Datex，Instrumentarium Corp.，赫尔辛基，芬兰）测定基础代谢率BMR，并使用生物电阻抗分析（BIA）方法（BIA，Akern-Gerät，RJL Systems，底特律）测定身体成分的组成部分。将计算得到的BMR与测量得到的BMR进行比较。
在这43例严重及病态肥胖受试者中，预测BMR的最佳拟合方程为哈里斯-本尼迪克特方程（计算得到的BMR与测量得到的BMR之比，均值±标准差%；相关系数r = 101±12.9；0.69）、詹森方程（101.5±12.3；0.74）、尼尔森方程（99.3±11.4；0.76）和坎宁安方程（98.9±11.7；0.74）。原始哈里斯-本尼迪克特方程的预测值与罗扎等人重新计算的修正哈里斯-本尼迪克特方程略有不同（101.1±12.9；0.69对99.7±12.8；0.69）。在体重≥120 kg的22名受试者组中，使用原始哈里斯-本尼迪克特方程估算的BMR值与测量得到的BMR之比分别为102.2±15.4%（均值±标准差%，r = 0.61），按照当前建议将体重设为120 kg时该比值为93.2±14.5%（r = 0.50）。在该亚组中，根据尼尔森方程计算得到的BMR/测量得到的BMR之比为101.0（12.1/0.74）。
对于II级和III级肥胖患者，哈里斯-本尼迪克特方程预测平均BMR的精度在临床上可接受，并且比目前大多数可用于基础代谢率（BMR）预测的方程拟合效果更好。然而，重新计算的版本（由罗扎等人计算）更准确，因此应使用该版本而非原始方程：BMR（男性）= 88.362 + 4.799×（身高）+ 13.397×（体重） - 5.677×（年龄）；BMR（女性）= 447.593 + 3.098×（身高）+ 9.247×（体重） - 4.330×（年龄）。尼尔森方程不仅将LBM作为预测因子，还将FM作为额外的预测因子，是最佳的预测方程（[108 LBM + 16.9 FM]0.239）。哈里斯-本尼迪克特方程在体重≥120 kg的极度肥胖受试者中也具有足够的精度，因此无需调整。
