Comparison of segmental with whole-body impedance measurements in peritoneal dialysis patients

Segmental impedance measurements were obtained using nine electrode configurations in 21 male patients undergoing peritoneal dialysis PD before and after the fluid drainage. For each segment we analyzed the impedance Z and the impedance divided by the height H of the patient Z/H. Our objective was to compare different segmental measurements with whole-body measurements in peritoneal dialysis. The Wilcoxon test was used to analyze the change in impedance produced by a PD session. Pearson or Spearman correlation coefficients were used for continuous or discrete variables, respectively. Statistical significance was set at P < 0.05. Similar results were obtained for Z and Z/H. The correlation coefficients between the real R and imaginary X-c parts of segmental impedances after drainage were within the expected range for healthy population (0.46-0.70), but not before drainage for the abdomen (0.34) and the upper part of the leg (0.24). The correlation between the real part of whole-body and the real part of longitudinal segments in the limbs was high (r = 0.807-0.879). Furthermore, the imaginary part of whole-body showed a high correlation with the imaginary part of all longitudinal segments (r = 0.856-0.931). The high contribution of arm and leg impedances in the whole-body impedance produced high correlation between whole-body and segmental measurements in legs and arms. In agreement with other previous studies, a significant increase of the arm resistance was detected after fluid drainage. The drainage of fluids in PD patients produced significant changes in the measured real parts of impedance in all measured segments, but only the measurement in the abdomen showed a significant positive correlation (r = 0.533) with the extracted fluid volume. This low correlation indicates that the individual assessment of fluid volumes using segmental measurements will be highly inaccurate. (C) 2007 IPEM. Published by Elsevier Ltd. All rights reserved.