Noppa H, Andersson M, Bengtsson C, Bruce A, Isaksson B
Am J Clin Nutr. 1979 Jul;32(7):1388-95. doi: 10.1093/ajcn/32.7.1388.
A variety of anthropometric measurements was made in a randomized population sample of middle-aged women in five age strata in whom body composition was estimated from total body potassium and total body water determined by whole body counting and isotope dilution technique, respectively. No significant differences with age were found for total body potassium or total body water. A significant age difference was found for body fat mass. Simple linear correlations between anthropometric variables and body fat mass estimates were found to be 0.90, 0.86, 0.77 for body weight, buttock circumference, and sum of triceps and subscapular skinfold thicknesses, respectively. Multiple regression analysis showed that these three variables accounted for 80 to 91% of the variation in body fat in the different age strata studied. Multiple regression equations for prediction of body fat from anthropometric variables are given. In about two-thirds of the subjects, the difference between predicted body fat mass and estimated body fat mass was found to be less than +/- 2.5 kg.
对处于五个年龄层的中年女性随机抽样群体进行了多种人体测量，分别通过全身计数和同位素稀释技术测定总体钾和总体水，据此估算身体成分。未发现总体钾或总体水随年龄有显著差异。但发现体脂量存在显著的年龄差异。人体测量变量与体脂量估算值之间的简单线性相关性分别为：体重0.90、臀围0.86、肱三头肌和肩胛下皮褶厚度之和0.77。多元回归分析表明，这三个变量在不同研究年龄层中占体脂变化的80%至91%。给出了根据人体测量变量预测体脂的多元回归方程。在约三分之二的受试者中，预测体脂量与估算体脂量之间的差异小于±2.5千克。
