Effects of variations in regional composition on soft tissue measurements by dual-energy X-ray absorptiometry

OBJECTIVE: To assess the effect of variations in regional composition on the accuracy of dual-energy X-ray absorptiometry (DXA) in humans by placing packets of lard, water and ground beef, to simulate fat, very lean tissue and high density lean material, on the thighs and abdomen. SUBJECTS: 28 healthy males and females (mean age: 25 y; mean % fat: 24.6%). METHODS: Up to six consecutive total body DXA scans (baseline plus five scans with added packets) at 16 cm/s were performed on each subject (Lunar DPX-L, software version 1.3y). Eight conditions were examined: one lard packet (n = 18), two stacked lard packets, one ground beef packet and one water packet (all n = 10) alternately placed on the thighs and abdomen. Masses (kg) and percentage fats (% fat) of the lard, water and ground beef packets were 1.45 and 90%, 2.80 and 4% and 1.80 and 24%, respectively. RESULTS: Differences between actual (baseline plus packet) and measured composition were assessed using paired t-tests (p < 0.05). Under all conditions except with two abdominal lard packets, actual and measured total mass were similar (p > 0.05). Percentage fat of one lard packet was significantly underestimated when placed on the abdomen (90% actual vs 52% measured). The % fat of two lard packets was underestimated when placed on the thighs (90% vs 77%) and abdomen (90% vs 47%) while the % fat of the water packet was overestimated in both locations. CONCLUSION: The Lunar DXA underestimates the % fat of lard placed over the abdomen moreso than when placed over the thighs and overestimates % fat of water in both locations independent of thickness. Ground beef is accurately detected.