FAT AND HYDRATION MONITORING BY ABDOMINAL IMPEDANCE ANALYSIS:
MODEL-BASED DATA INTERPRETATION

We investigated the sensitivity of the abdominal impedance Z measured with four electrodes to fat fractions and hydration in different abdominal compartments with a finite-element model. The compartments are subcutaneous fat, muscle, mesentery, spine. Assuming the same percent change of the fat fraction in all compartments the model predicts a sensitivity of 1.03 W per mm change of the subcutaneous fat layer thickness (SFL) compared to 1.1 W/mm measured.

• 60-70 % of the sensitivity stems from the SFL itself, 20-30 % from the mesentery and 10 % from fat in the muscle compartment.

The hierarchical model could be fitted satisfactorily to the experimental data yielding parameter values within the expected range.

• The observed regression slope of SFL and Z2 (1.1 W/mm) can be explained by a simultaneous change of the visceral and the subcutaneous fat fraction. i. e. the method is sensitive to the visceral fat.
• The hydration of muscle interferes strongly with the fat parameters. However, it can be clearly discerned whether a weight reduction is due to fat reduction or dehydration as the trend of the transimpedance is inverse in both cases.
• It is impossible to identify hydration and SFL independently with a 4-terminal arrangement. A better separation of different contributions can be achieved by adding at least one strategically placed electrode. The combined evaluation of Z2 and Z4 yields a significant improvement, Z2 yielding more information about central structures and Z4 characterizing surface-near compartments. The use of multifrequency data reduces the CV of the parameters when compared to the single frequency case.

Transversal abdominal cross section at the umbilicus