目的：生物阻抗分析(Bioelectrical impedance analysis, BIA)是目前常用來估測身體組成的測量方法。一般研究指出八極板式的生物阻抗分析可經由測量四肢的瘦肉質量(Lean Body Mass, LBM)進而估測出全身的骨骼肌肉量(Skeletal Muscle Mass, SMM)，本文研究目的在於應用雙能X光吸收儀(Dual-energy X-ray absorptiometry, DXA)來評估單頻八極板生物阻抗分析儀於瘦肉質量估測的精確度，並建立適合台灣中老年人的瘦肉質量的估測方程式。方法：以台灣地區行動正常的中老年人為受測者，應用站立式單頻八極板生物阻抗分析儀BC-418(Tanita Co., Tokyo, Japan; 簡稱BIA_8)測量其左、右上下肢合併計算的四肢瘦肉質量，並以DXA測量相同項目的結果當標準來比較分析。結果：共71名中老年人(男性29名、女性42名，平均年齡60.76±6.26歲)接受測試。其BIA_8於上肢、下肢與四肢估測的結果分別為4.15±1.21kg、16.00±4.51kg與20.15±5.79kg；DXA的測量結果為3.77±1.27kg、11.69±2.89kg與15.46±4.09kg，比較BIA_8與DXA的測量結果其對應的相關係數分別為0.960、0.961與0.973(p＜0.001)。用Bland-Altman分析兩裝置的一致性差異(Limits of agreement, LOA)分別為-0.342至1.094kg、0.320至8.309kg與0.596至8.768kg。結論：本研究顯示BIA_8與DXA於上、下肢與四肢的瘦肉質量的測量結果有高正相關性，但BIA_8估測結果顯著高估DXA測量結果；未來進一步擴大收案和修正估測結果，將可更適合應用於於台灣地區中老人的上、下肢瘦肉質量的估測，及有效測量其瘦肉質量或者骨骼肌肉量。
Objectives. Eight-electrode bioelectrical impedance analysis (BIA) can estimate the skeletal muscle mass (SMM) in the total body by measuring the appendicular lean body mass (LBM). The purpose of the present study was to assess the precision of the estimation for the lean body mass by a single-frequency eight-electrode bioelectrical impedance analyzer via using duel energy x-ray absorptiometry, and to establish the estimation equation suitable for the assessment of lean body mass of middle-aged and elderly people in Taiwan. Methods. The study group consisted middle-aged and elderly people with normal mobility in Taiwan. The lean body mass in the left, right, upper and lower limbs and the four limbs was measured by a standing-posture eight-electrode bioelectrical impedance analyzer- BC418 (Tanita Corp, Tokyo, Japan, termed as BIA_8), and the results were compared with those obtained from duel energy x-ray absorptiometry (DXA). Results. 71 people (29 males and 42 females, aged 60.76±6.26 years) were recruited. The estimation results in the upper, lower and four limbs by BIA_8 were 4.15±1.21kg, 16.00±4.51kg and 20.15±5.79kg, respectively, and those by DXA were 3.77±1.27kg, 11.69±2.89kg and 15.46±4.09kg, respectively. The correlation coefficients between the measurements by BIA_8 and the corresponding results by DXA were 0.960, 0.961 and 0.973 (p＜0.001). The limits of agreement (LOA) between the two devices by Bland-Altman analysis were -0.342kg to 1.094kg, 0.320kg to 8.309kg, and 0.596 to 8.768kg, respectively. Conclusion. The measurements of lean body mass in upper, lower and four limbs by BIA_8 were highly positively correlated with those by DXA. However, the assessments by BIA_8 were overestimated and significantly higher than the measurements by DXA. It is suggested that the estimation results should be adjusted to be more suitable for the assessment of the lean body mass in upper and lower limbs of middle-age and elderly people in Taiwan in future. So it could be used to measure their lean body mass or SMM effectively.