Body composition analysis of pigs by dual-energy x-ray absorptiometry

Dual-energy x-ray absorptiometry (DXA) was evaluated as a method for measuring body composition of pigs. Forty-eight female pigs (10.2 to 60.5 kg) were killed and the whole bodies scanned on a DXA instrument. The DXA measurements provided readings of total tissue mass, percentage of fat, fat tissue mass, lean body mass, and bone mineral content. By chemical analysis, the whole body fat content of the pigs ranged from 9.3 to 24.3%, giving rise to DXA RST values (ratio of attenuation coefficients) ranging from 1.386 to 1.334. The average percentage body fat measured by DXA (18.2 +/- .9%) was not significantly different (P = .76) from the results by chemical analysis (17.8 +/- .6%); however, concordance correlation analysis revealed unacceptable accuracy in the DXA measurement due to a negative bias for smaller percentages and a positive bias for larger percentages. Total body fat measured by the two methods was also highly correlated (r = .989), and the average for DXA (7.31 +/- .62 kg) was not significantly different (P = .89) from the value for the chemical method (7.20 +/- .50 kg). The relationship between the DXA estimate of lean body mass and the amount of protein in the body by chemical analysis was described by the equation: kg protein=.227.kg DXA lean -1.28, with a correlation coefficient (r) of .968. The DXA measurements of total tissue mass and actual body weights were highly correlated (r = .999), with an acceptable concordance at the .05 level. Scans were also analyzed for regional composition of the front and back legs. It was generally difficult to determine anatomically how well soft tissue described by DXA regions corresponded to those dissected. The DXA and chemical results for regional analysis of back legs were in better agreement than those for the front legs. These results indicate that DXA may be used as a reliable method for measuring body composition of pigs, but needs more extensive calibration and may be more appropriate for total body rather than regional analysis.