IEEE TRANSACTIONS ON BIOMEDICAL ENGINEERING, VOL. 52, NO. 6, JUNE 2005 975 Fat and Hydration Monitoring by Abdominal Bioimpedance Analysis: Data Interpretation by Hierarchical Electrical Modeling Hermann Scharfetter*, Patricia Brunner, Michael Mayer, Bernhard Brandstätter, and Helmut Hinghofer-Szalkay Abstract—In a previous publication, it was demonstrated that the abdominal subcutaneous fat layer thickness (SFL) is strongly correlated with the abdominal electrical impedance when mea- sured with a transversal tetrapolar electrode arrangement. This article addresses the following questions: 1) To which extent do different abdominal compartments contribute to the impedance? 2) How does the hydration state of tissues affect the data? 3) Can hydration and fat content be assessed independently? For simulating the measured data a hierarchical electrical model was built. The abdomen was subdivided into three compartments (subcutaneous fat, muscle, mesentery). The true anatomical struc- ture of the compartment boundaries was modeled using finite-ele- ment modeling (FEM). Each compartment is described by an elec- trical tissue model parameterized in physiological terms. Assuming the same percent change of the fat fraction in the mesentery and the SFL the model predicts a change of 1,24 /mm change of the SFL compared to 1,1 /mm measured. 42% of the change stem from the SFL, 56% from the mesentery and 2% from changes of fat within the muscle compartment. A 1% increase of the extracellular water in the muscle is not discernible from a 1% decrease of the SFL. The measured data reflect not only the SFL but also the visceral fat. The tetrapolar electrode arrangement allows the measurement of the abdominal fat content only if the hydration remains constant. Index Terms—Bioimpedance, fat monitoring, electrical tissue modeling, finite elements, subcutaneous fat, visceral fat. I. INTRODUCTION T HERE exists a huge number of publications dealing with the estimation of body composition from measurements of the electrical impedance of body segments (bioimpedance analysis (BIA); see, e.g., [1]). The assessment of the human body composition in terms of body fat (BF) and lean [fat free mass (FFM)] fractions in varyious compartments is of particular interest in different dis- ciplines such as sports sciences, clinical nutrition, assessment of the nutritional status of the elderly population, management of obesity and associated risk factors (cardiovascular diseases, Manuscript received April 6, 2004; revised November 14, 2004. This work was supported in part by the Austrian Science Fund under Project P16413. As- terisk indicates corresponding author. *H. Scharfetter is with the Institute of Medical Engineering, Graz University of Technology, 8010 Graz, Austria (e-mail: scharfetter@bmt.tu-graz.ac.at). P. Brunner and M. Mayer are with the are with the Institute of Medical Engi- neering, Graz University of Technology, 8010 Graz, Austria. B. Brandstätter is with the Institute of Electrical Measurement and Measure- ment Signal Processing, Graz University of Technology, 8010 Graz, Austria. H. Hinghofer-Szalkay is with the Institute for Systems Physiology, Graz Uni- versity of Technology, 8010 Graz, Austria. Digital Object Identifier 10.1109/TBME.2005.846733 diabetes). However, the reliability of all impedance methods presented so far is very limited. The relative deviation of the BF measured with BIA and reference methods has been inves- tigated in various validation studies [1]–[5]. When compared to underwater weighting the BF as determined with BIA shows a standard error of the difference between 2% and 5% of the total body mass [1]–[5]. As the BF ranges between 5%–40% in men and 10%–45% in women, the percent errors are comparatively large, especially in lean subjects where they can go up to several tens of percent. The errors are at least partly due to the low direct sensitivity of BIA to the quantity of adipose tissue. Instead, the electrical cur- rent passes essentially through the fat-free, highly conducting regions. It was found [6] that the fat contributes significantly to the overall conductance only in considerably obese people so that the muscle represents the main conductor. The error prop- agation from the estimated FFM to the derived BF makes the method highly susceptible to interferences, e.g., with the water content of the FFM. The redistribution of body fluids, e.g., due to orthostatic processes, can cause considerable errors in the estimation of BF by BIA. Moreover, the data obtained from so-called whole-body measurements (measured from wrist to ankle) contain only little information about the trunk, as the latter contributes only with 5%–10% to the total impedance [5]. Essentially the same is true for the leg-to-leg or arm-to-arm con- figuration, which has also been suggested. Therefore, a more di- rect method with high sensitivity to the adipose tissue in central body compartments is highly desirable. Actually most fat mass estimators are based on the idea that the fat mass can be obtained by subtracting FFM, which is neg- atively correlated with impedance, from body mass. To reduce the error due to poor correlation between impedance and fat mass usual estimators consider also other anthropometric data such as body mass, height, sex and age. The best reproducibility is mostly obtained from such estimators in which the relative weight of the impedance is comparatively low. This is somewhat unsatisfactory because in that way impedance measurements do not really contribute as much diagnostic information as they the- oretically contain. Intuitive quantities of interest are an index for the subcuta- neous fat mass and the fraction of adipose tissue in the muscular and in the mesenteric compartments. However, it is not trivial to define these parameters precisely. A more general analysis might include the frequency dependence of the conductivities of the different tissues as they contain information about some microstructural properties, e.g., about the hydration state. 0018-9294/$20.00 © 2005 IEEE