Mitchell A D, Scholz A M, Conway J M
Growth Biology Laboratory, ARS, USDA, Beltsville, MD 20705-2350, USA.
J Anim Sci. 1998 Sep;76(9):2392-8. doi: 10.2527/1998.7692392x.
We evaluated the use of dual-energy x-ray absorptiometry (DXA) for measuring the gross body composition of small subjects in 97 pigs that ranged from 5 to 27 kg live body weight. Scans were performed using a Lunar DPXL densitometer in the pediatric mode (Version 3.8e). The DXA scans of the live pigs provided measurements of total fat, lean, and bone mineral content. After scanning, the pigs were killed, the entire body was ground, and samples were analyzed chemically (CHEM) for fat, protein, ash, and water content. We found that DXA significantly underestimated the percentage of fat in the body (DXA, 6.9+/-.33% vs CHEM, 10.9+/-.31%, P < .001). The correlation (r) between DXA and chemical measures of percentage fat was .86 and for grams of fat it was .96. Lean tissue mass measured by DXA was highly correlated with CHEM measurements of total grams of body water (r = .99), total grams of body protein (r = .94), and lean body mass (r = .99). The average DXA bone mineral content was within 2% of the amount estimated from total body ash and the correlation between the two values was .94. The relationships between DXA and CHEM measurements for percentages of body composition of pigs that weighed between 5 and 27 kg are described by the following regression equations: %fatCHEM = 5.22 + [.817.fatDXA], (r = .86, standard error of the estimate, SEE = 1.56); %proteinCHEM = -7.8 + [.256%leanDXA], (r = .35, SEE = 2.3); %waterCHEM = -5.2 + [.808-%leanDXA], (r = .59, SEE = 3.67). These results are consistent with previously reported results and suggest that even though direct use of DXA readings may not be sufficiently accurate, the high degree of correlation indicates that with proper calibration the DXA values can be used to predict body composition.
我们评估了双能X线吸收法（DXA）在测量97头体重在5至27千克之间的小猪总体身体成分中的应用。使用Lunar DPXL骨密度仪以儿科模式（版本3.8e）进行扫描。对活猪的DXA扫描提供了总脂肪、瘦肉和骨矿物质含量的测量值。扫描后，将猪处死，将整个身体磨碎，并对样本进行化学分析（CHEM）以测定脂肪、蛋白质、灰分和水分含量。我们发现，DXA显著低估了体内脂肪的百分比（DXA为6.9±0.33%，而CHEM为10.9±0.31%，P<0.001）。DXA与化学测量的脂肪百分比之间的相关性（r）为0.86，与脂肪克数的相关性为0.96。通过DXA测量的瘦组织质量与CHEM测量的体内总水分克数（r = 0.99）、体内总蛋白质克数（r = 0.94）和瘦体重（r = 0.99）高度相关。平均DXA骨矿物质含量在根据全身灰分估计量的2%以内，两者之间的相关性为0.94。体重在5至27千克之间的猪的身体成分百分比的DXA与CHEM测量值之间的关系由以下回归方程描述：%脂肪CHEM = 5.22 + [0.817×脂肪DXA]，（r = 0.86，估计标准误差，SEE = 1.56）；%蛋白质CHEM = -7.8 + [0.256×%瘦肉DXA]，（r = 0.35，SEE = 2.3）；%水分CHEM = -5.2 + [0.808×%瘦肉DXA]，（r = 0.59，SEE = 3.67）。这些结果与先前报道的结果一致，表明尽管直接使用DXA读数可能不够准确，但高度相关性表明通过适当校准，DXA值可用于预测身体成分。
