BIOELECTRICAL-IMPEDANCE ANALYSIS FOR ESTIMATING BODY-COMPOSITION OF SOWS IN-VIVO

In 15 crossbred sows, bioelectrical impedance analysis (BIA) was performed to get regression formulas for total body water (GKW), fat (GKF), and protein (GKP). Animals were fasted 12 h, but had free access to water. BIA was performed by a four-terminal impedance plethysmograph, which supplies a current of 0.9 mA with 51.2 kHz. Adhesive silver chloride electrodes were used, animals were not sedated, standing on a dry floor. Live weight (LM, 145-225 kg), heart girth (BU, 112-133 cm), crown-rump-length (SSL, 127-160 cm), Resistance (Rs, 89-127 Ohms), Reactance (Xc, 18-38 Ohms) were measured. As reference method served the chemical analysis of crude water (Xw, 54.7-70.8%), fat (Xf, 4.6-23.3%), and protein (Xp, 16.4-20.1%) in the homogenized body to establish regression formulas. By addition of Rs and/or Xc to the regression formulas accuracy was better than without Rs and/or Xc. The best results for estimating the body composition were achieved with the following equations: GKW (kg) = 7.126 + 0.389 LM + 0.039 BU + 0.129 SSL(2)/Rs - 0.285 Xc (R(2) = 0.968 p < 0.001) max. error:-3.52 kg/+4.30 kg GKF (kg) = -27.510 + 0.296 LM + 0.197 BU - 0.183 SSL(2)/Rs + 0.613 Xc (R(2) = 0.820 p < 0.001) max. error:-5.78 kg/+8.56 kg GKP (kg) = -8.133 + 0.151 LM + 0.092 SSL - 0.011 BU2/Rs - 0.037 Xc (R(2) = 0.956 p < 0.001) max. error:-1.95 kg/+1.39 kg Equations for absolute contents (kg) were much better than for relative contents (% of total body mass). Most problems were found in estimating the body fat content from BIA regressions, but without Rs and/or Xc the deviation was even greater. This problem may possibly be diminished in progressive studies, i.e, in sows during lactation. Results stimulate for further research, espec. in type and position of electrodes, and for direct comparison with other equipment for estimating body composition in vivo.