Deurenberg P, Deurenberg-Yap M, Schouten F J M
Department of Nutrition and Epidemiology, Wageningen University, The Netherlands.
Eur J Clin Nutr. 2002 Mar;56(3):214-20. doi: 10.1038/sj.ejcn.1601303.
To test the impact of body build factors on the validity of impedance-based body composition predictions across (ethnic) population groups and to study the suitability of segmental impedance measurements.
Cross-sectional observational study.
Ministry of Health and School of Physical Education, Nanyang Technological University, Singapore.
A total of 291 female and male Chinese, Malays and Indian Singaporeans, aged 18-69, body mass index (BMI) 16.0-40.2 kg/ m2.
Anthropometric parameters were measured in addition to impedance (100 kHz) of the total body, arms and legs. Impedance indexes were calculated as height2/impedance. Arm length (span) and leg length (sitting height), wrist and knee width were measured from which body build indices were calculated. Total body water (TBW) was measured using deuterium oxide dilution. Extra cellular water (ECW) was measured using bromide dilution. Body fat percentage was determined using a chemical four-compartment model.
The bias of TBW predicted from total body impedance index (bias: measured minus predicted TBW) was different among the three ethnic groups, TBW being significantly underestimated in Indians compared to Chinese and Malays. This bias was found to be dependent on body water distribution (ECW/TBW) and parameters of body build, mainly relative (to height) arm length. After correcting for differences in body water distribution and body build parameters the differences in bias across the ethnic groups disappeared. The impedance index using total body impedance was better correlated with TBW than the impedance index of arm or leg impedance, even after corrections for body build parameters.
The study shows that ethnic-specific bias of impedance-based prediction formulas for body composition is due mainly to differences in body build among the ethnic groups. This means that the use of 'general' prediction equations across different (ethnic) population groups without prior testing of their validity should be avoided. Total body impedance has higher predictive value than segmental impedance.
测试身体形态因素对不同（种族）人群基于阻抗的身体成分预测有效性的影响，并研究分段阻抗测量的适用性。
横断面观察性研究。
新加坡南洋理工大学卫生与体育学院。
共291名18 - 69岁的华裔、马来裔和印度裔新加坡男女，体重指数（BMI）为16.0 - 40.2kg/m²。
除测量全身、手臂和腿部的阻抗（100kHz）外，还测量人体测量参数。阻抗指数计算为身高²/阻抗。测量臂长（伸展度）和腿长（坐高）、手腕和膝盖宽度，并据此计算身体形态指数。使用氧化氘稀释法测量总体水（TBW）。使用溴化物稀释法测量细胞外液（ECW）。使用化学四室模型测定体脂百分比。
根据全身阻抗指数预测的TBW偏差（偏差：测量值减去预测的TBW）在三个种族群体中有所不同，与华裔和马来裔相比，印度裔的TBW被显著低估。发现这种偏差取决于身体水分分布（ECW/TBW）和身体形态参数，主要是相对（身高）臂长。在校正身体水分分布和身体形态参数的差异后，各民族间偏差的差异消失。即使在校正身体形态参数后，使用全身阻抗的阻抗指数与TBW的相关性也比手臂或腿部阻抗的阻抗指数更好。
研究表明，基于阻抗的身体成分预测公式的种族特异性偏差主要是由于不同种族群体之间身体形态的差异。这意味着应避免在未经事先有效性测试的情况下，对不同（种族）人群使用“通用”预测方程。全身阻抗比分段阻抗具有更高的预测价值。
