Al-Gindan Yasmin Y, Hankey Catherine R, Govan Lindsay, Gallagher Dympna, Heymsfield Steven B, Lean Michael E J
1Department of Human Nutrition,School of Medicine,College of Medical,Veterinary & Life Sciences,University of Glasgow,Royal Infirmary,Glasgow G4 0SF,UK.
2Department of Human Nutrition,Institute of Health and Well Being,College of Medical,Veterinary & Life Sciences,University of Glasgow,Glasgow G12 8QQ,UK.
Br J Nutr. 2015 Dec 14;114(11):1852-67. doi: 10.1017/S0007114515003670. Epub 2015 Oct 5.
The reference organ-level body composition measurement method is MRI. Practical estimations of total adipose tissue mass (TATM), total adipose tissue fat mass (TATFM) and total body fat are valuable for epidemiology, but validated prediction equations based on MRI are not currently available. We aimed to derive and validate new anthropometric equations to estimate MRI-measured TATM/TATFM/total body fat and compare them with existing prediction equations using older methods. The derivation sample included 416 participants (222 women), aged between 18 and 88 years with BMI between 15·9 and 40·8 (kg/m2). The validation sample included 204 participants (110 women), aged between 18 and 86 years with BMI between 15·7 and 36·4 (kg/m2). Both samples included mixed ethnic/racial groups. All the participants underwent whole-body MRI to quantify TATM (dependent variable) and anthropometry (independent variables). Prediction equations developed using stepwise multiple regression were further investigated for agreement and bias before validation in separate data sets. Simplest equations with optimal R (2) and Bland-Altman plots demonstrated good agreement without bias in the validation analyses: men: TATM (kg)=0·198 weight (kg)+0·478 waist (cm)-0·147 height (cm)-12·8 (validation: R 2 0·79, CV=20 %, standard error of the estimate (SEE)=3·8 kg) and women: TATM (kg)=0·789 weight (kg)+0·0786 age (years)-0·342 height (cm)+24·5 (validation: R (2) 0·84, CV=13 %, SEE=3·0 kg). Published anthropometric prediction equations, based on MRI and computed tomographic scans, correlated strongly with MRI-measured TATM: (R (2) 0·70-0·82). Estimated TATFM correlated well with published prediction equations for total body fat based on underwater weighing (R (2) 0·70-0·80), with mean bias of 2·5-4·9 kg, correctable with log-transformation in most equations. In conclusion, new equations, using simple anthropometric measurements, estimated MRI-measured TATM with correlations and agreements suitable for use in groups and populations across a wide range of fatness.
参考器官水平的身体成分测量方法是磁共振成像（MRI）。对总脂肪组织量（TATM）、总脂肪组织脂肪量（TATFM）和全身脂肪进行实际估计对流行病学研究很有价值，但目前尚无基于MRI的经过验证的预测方程。我们旨在推导并验证新的人体测量方程，以估计通过MRI测量的TATM/TATFM/全身脂肪，并将其与使用旧方法的现有预测方程进行比较。推导样本包括416名参与者（222名女性），年龄在18至88岁之间，体重指数（BMI）在15.9至40.8（kg/m²）之间。验证样本包括204名参与者（110名女性），年龄在18至86岁之间，BMI在15.7至36.4（kg/m²）之间。两个样本都包括不同种族/民族群体。所有参与者均接受全身MRI检查，以量化TATM（因变量）和人体测量数据（自变量）。在单独的数据集中进行验证之前，进一步研究了使用逐步多元回归开发的预测方程的一致性和偏差。具有最佳R²和Bland-Altman图的最简单方程在验证分析中显示出良好的一致性且无偏差：男性：TATM（kg）=0.198×体重（kg）+0.478×腰围（cm）-0.147×身高（cm）-12.8（验证：R²=0.79，变异系数（CV）=20%，估计标准误差（SEE）=3.8 kg）；女性：TATM（kg）=0.789×体重（kg）+0.0786×年龄（岁）-0.342×身高（cm）+24.5（验证：R²=0.84，CV=13%，SEE=3.0 kg）。基于MRI和计算机断层扫描的已发表人体测量预测方程与通过MRI测量的TATM高度相关：（R²=0.70 - 0.82）。估计的TATFM与基于水下称重的已发表全身脂肪预测方程相关性良好（R²=0.70 - 0.80），平均偏差为2.5 - 4.9 kg，大多数方程可通过对数转换进行校正。总之，使用简单人体测量指标的新方程能够以适合广泛肥胖程度的群体和人群使用的相关性和一致性来估计通过MRI测量的TATM。
