Presented as an Abstract at the FELANPE Nutrition Congress in Montevideo, Uruguay on 7th November, 2005
Translated from the original Portuguese presentation.
INTRODUCTION: The clinical utility of electrical bio-impedance has become increasingly widespread in spite of its limitations in some clinical situations. Different bio-impedance devices are commercialized although it is still uncertain whether the resistance values and reactance obtained from each one can be compared. This information is extremely useful for multi-centre studies, when attempting to produce standard equations in each population. Confirmation of comparable results would require the use of one sole instrument for obtaining data.
A previous study demonstrated that the percentages of fat and muscle obtained from different instruments are not comparable. However, this difference may occur due to the use of different equations that are specific to each device, or different resistance and reactance values obtained by the devices.
OBJECTIVE: To compare the resistance, reactance and phase angle results obtained by 5 single-phase and multi-frequency BIA devices in healthy volunteers.
METHODS: Voluntary employees from 3 hospitals of Pelotas, RS, were submitted to the examination according to the recommended standards. 50 kHz frequency devices were used. BIA Quantum (RJL Systems: 1); BIM 4 (Impedimed: and the Quadscan multi-frequency (Bodystat® and Xitron 4000 B (Xitron: 5). Resistance, reactance and the 50 kHz frequency phase angle were measured in each device. In the Quadscan device, this data was obtained from the impedance of the 5, 50, 100 and 200 kHz frequencies and transformed by the appropriate Bodystat® Phase Angle software. The data analyzed in version 8.0 of the STATA programme, using the concordance correlation coefficient (CCC) and concordance limits according to Bland Altman (LDC).
RESULTS: The sample was taken from 51 healthy volunteers, 65% of which were females. Their average age ranged between 31 ± 8. With regard to Resistance, all devices showed satisfactory CCCs (varying between 0.974 and 0.999). The average difference between the devices varied between - 13,1 a 6,8 ohms. With regard to Reactance, only devices 1 and 4 had a CCC = 0.969. All the other comparisons were unsatisfactory. And with regard to the phase angle, only devices 1 and 4 had a CCC = 0.988. All other comparisons were unsuitable.
CONCLUSION: The use of different BIA devices may result in incomparable results of the estimated body composition not due to the use of different equations but to the different Resistance and Reactance values. Care must be taken when using different devices in multi-centre studies, depending upon the variables used. In this study, only the resistance, reactance and phase angle results of Quantum and Bodystat Quadscan BIA devices may be considered comparable.