Presentation June 2004 in Brazil by M. Cristina G. B. e Silva.
BACKGROUND: The phase angle (PA) is a measure obtained directly through Electric Bioimpedance (EB), like the arctangent of the ratio between reactance and resistence. Studies suggest that it can represent an indirect measuring of the body cellular mass (BCM), thus playing an important role as a prognostic indicator in diverse clinical situations. As such, devices such as Bodystat®, in which the impedance value is supplied rather than the resistance (R) and reactance (Xc) values, have limited use in such situations, as the attainment of the PA becomes impossible. In more up-to-date multi-frequency models, it is possible to estimate resistance, reactance and PA through a program that uses the impedances at frequencies of 5, 50, 100 and 200 kHz.
OBJECTIVE: Verify the PA estimation precision and accuracy through the program made available by Bodystat® (BPA) with the PA obtained directly through the resistance and reactance values in an EB device from RJL Systems® (RJLPA), which is more commonly used in scientific work.
METHODS: Individuals simultaneoulsy underwent EB exams on two devices (BIA QUANTUM and Bodystat QuadScan 4000) in accordance with adequate standardization (NIHT, 1996). R and Xc values were obtained at 50 kHz (BIA Quantum) and impedance values were obtained at frequencies of 5, 50, 100 and 200 kHz (Bodystat Quadscan® 4000). For the first device, the PA was obtained through the equation AF = arctangent (Xc/R)* 57.296. For the second device, the Bodystat® Phase Angle Software Program - version 1.0 - was used with the impedances at the 4 frequencies cited above. Exams were carried out in duplicate with the use of the RJL and Bodystat electrodes, simultaneously. Statistical tests were carried out to assess the precision, accuracy and concordance coefficient (CC) between the two measurements. Bland-Altman graphs were used to assess measurement agreement on a single individual.
RESULTS: The sample included 70 EB exams carried out on individuals (39 healthy volunteers and 23 patients with neoplasia in chemiotherapy), with an average age of 42.7 ± 17.5 years, 18 of which were male and 44 were female. With the use of the RJL electrodes it was observed that, on average, the BPA overestimates the RJLPA by 0.074 (variation from - 0.376 to + 0.228). The precision and accuracy of both measurements are excellent: 0.987 and 0.997, respectively, with a CC of 0.984. Using the Bodystat electrodes it was observed that, on average, the BPA overestimates the RJLPA by 0.025o (variation from - 0.341 to + 0.291o). Precision and accuracy between the two measuements remained excellent: 0.987 and 1.0, respectively, with a CC of 0.987.
CONCLUSION: The use of multi-frequency bioimpedance devices with the assistance of estimation programs enables the PA to be obtained even without the direct input of resistance and reactance measures. Thus, the device can be used in clinical situations where the estimation of this variable (PA) is required. The device can also make possible the utilization of distinct formulas supplied by the fabricant for estimating body composition.