Frailty in Ageing Research Group (FRIA), Vrije Universiteit Brussel (VUB), Laarbeeklaan 103, 1090, Brussels, Belgium.
Radiology Department, University Hospital Brussels, Laarbeeklaan 101, 1090, Brussels, Belgium.
Clin Nutr. 2017 Jun;36(3):869-875. doi: 10.1016/j.clnu.2016.04.026. Epub 2016 Apr 28.
BACKGROUND & AIMS: No generalizable formulas exist that are derived from bioelectrical impedance analysis (BIA) for predicting appendicular lean mass (ALM) and fat mass (AFM) in sarcopenic older adults. Since precision of regional body composition (BC) data in multicentre trials is essential, this study aimed to: 1) develop and cross-validate soft tissue BIA equations with GE Lunar and Hologic DXA systems as their reference 2) to compare our new ALM equation to two previously published models and 3) to assess the agreement between BIA- and DXA-derived soft tissue ratios as indicators of limb tissue quality.
Two-hundred and ninety-one participants with functional limitations (SPPB-score 4-9; sarcopenia class I or II, measured by BIA) were recruited from 18 study centres in six European countries. BIA equations, using DXA-derived ALM and AFM as the dependent variable, and age, gender, weight, impedance index and reactance as independent variables, were developed using a stepwise multiple linear regression approach.
Cross-validation gave rise to 4 equations using the whole sample: ALM (kg) = 1.821 + (0.168height/resistance) + (0.132weight) + (0.017reactance) - (1.931sex) [R = 0.86 and SEE = 1.37 kg] AFM (kg) = -6.553 - (0.093* height/resistance) + (0.272weight) + (4.295sex) [R = 0.70 and SEE = 1.53 kg] ALM (kg) = 4.957 + (0.196* height/resistance) + (0.060weight) - (2.554sex) [R = 0.90 and SEE = 1.28 kg] AFM (kg) = -4.716 - (0.142* height/resistance) + (0.316weight) + (4.453sex) - (0.040*reactance) [R = 0.73 and SEE = 1.54 kg] Both previously published models significantly overestimated ALM in our sample with biases of -0.36 kg to -1.05 kg. For the ratio of ALM to AFM, a strong correlation (r = 0.82, P < 0.0001) was found between the mean estimate from BIA and the DXA models without significant difference (estimated bias of 0.02 and 95% LOA -0.62, 0.65).
We propose new BIA equations allowing the estimation of appendicular lean and fat mass. Our equations allow to accurately estimate the appendicular lean/fat ratio which might provide information regarding limb tissue quality, in clinical settings. Furthermore, these BIA equations can be applied to characterize sarcopenia with Hologic and Lunar reference values for BC. Previously published BIA-based models tend to overestimate ALM in sarcopenic older adults. Users of both GE Lunar and Hologic may now benefit from these equations in field research.
目前尚无基于生物电阻抗分析（BIA）的可推广公式，用于预测老年肌少症患者的四肢瘦体重（ALM）和脂肪量（AFM）。由于多中心试验中区域体成分（BC）数据的精度至关重要，本研究旨在：1）开发并交叉验证使用 GE Lunar 和 Hologic DXA 系统作为参考的软组织 BIA 方程；2）比较我们新的 ALM 方程与两个先前发表的模型；3）评估 BIA 和 DXA 衍生软组织比作为肢体组织质量指标的一致性。
从六个欧洲国家的 18 个研究中心招募了 291 名有功能障碍的参与者（SPPB 评分 4-9；通过 BIA 测量为 I 或 II 期肌少症）。使用 DXA 衍生的 ALM 和 AFM 作为因变量，年龄、性别、体重、阻抗指数和电抗作为自变量，使用逐步多元线性回归方法开发 BIA 方程。
对整个样本进行交叉验证后得到 4 个方程：ALM（kg）= 1.821 +（0.168身高/电阻）+（0.132体重）+（0.017电抗）-（1.931性别）[R=0.86，SEE=1.37 kg]；AFM（kg）= -6.553 -（0.093身高/电阻）+（0.272体重）+（4.295性别）[R=0.70，SEE=1.53 kg]；ALM（kg）= 4.957 +（0.196身高/电阻）+（0.060体重）-（2.554性别）[R=0.90，SEE=1.28 kg]；AFM（kg）= -4.716 -（0.142身高/电阻）+（0.316体重）+（4.453性别）-（0.040电抗）[R=0.73，SEE=1.54 kg]。两个先前发表的模型在我们的样本中明显高估了 ALM，偏差为-0.36 至-1.05 kg。对于 ALM 与 AFM 的比值，BIA 和 DXA 模型的平均估计值之间存在很强的相关性（r=0.82，P<0.0001），且没有显著差异（估计偏差为 0.02，95%LOA -0.62，0.65）。
我们提出了新的 BIA 方程，可用于估计四肢瘦体重和脂肪量。我们的方程可以准确估计四肢瘦/脂肪比，这可能为肢体组织质量提供信息，适用于临床环境。此外，这些 BIA 方程可用于使用 Hologic 和 Lunar 的 BC 参考值来特征化肌少症。先前发表的基于 BIA 的模型往往会高估老年肌少症患者的 ALM。使用 GE Lunar 和 Hologic 的用户现在可以从这些方程中受益于 Field Research。
