Guo S M, Roche A F, Houtkooper L
Department of Pediatrics, Wright State University, Yellow Springs, OH 45387-1695.
Am J Clin Nutr. 1989 Sep;50(3):435-43. doi: 10.1093/ajcn/50.3.435.
Fat-free mass (FFM) values calculated from densitometry by using a multicomponent model were significantly (p less than 0.05) larger than those from the Siri two-component model (465 males, 441 females), especially in children and females. The multicomponent model assumes the density of FFM varies by age and sex because of differences in its constituents. With FFM values from the multicomponent model as the dependent variable and impedance and anthropometry as the independent variables, prediction equations were derived by all possible subsets of regression (140 males, 110 females). These equations had RMSEs of 2.2-2.3 kg and CVs of 5.0-5.8%. Cross-validation results were excellent for young adults but less satisfactory for children for whom a different calf skinfold had been measured. The selected equations are applicable to healthy white individuals aged 7-25 y.
使用多组分模型通过密度测定法计算得出的去脂体重(FFM)值显著(p小于0.05)高于使用Siri双组分模型得出的值(465名男性,441名女性),在儿童和女性中尤为明显。多组分模型假定由于去脂体重的成分存在差异,其密度会随年龄和性别而变化。以多组分模型得出的去脂体重值作为因变量,阻抗和人体测量数据作为自变量,通过回归的所有可能子集推导预测方程(140名男性,110名女性)。这些方程的均方根误差为2.2 - 2.3千克,变异系数为5.0 - 5.8%。交叉验证结果对年轻人来说非常好,但对测量了不同小腿皮褶厚度的儿童来说不太令人满意。所选方程适用于7至25岁的健康白人个体。