Quantification of Scan Analysis Errors in GE Lunar DXA Visceral Adiposity in Adults

J Clin Densitom. 2021 Apr-Jun;24(2):287-293. doi: 10.1016/j.jocd.2020.06.004. Epub 2020 Jun 25.

Abstract

Utilization of dual-energy X-ray absorptiometry is increasing in clinical settings and the fitness industry as a viable tool to assess total and regional body composition, including visceral adiposity. Previous research using small samples (<50) has described several pitfalls in patient positioning, scan acquisition, and/or analysis that alter regional body composition values. Our aim was to quantify the largest probable error in measures of total, android, gynoid, and visceral fat caused by incorrect placement of the head cut-line, in a large sample of adults. Total body images (N = 436) from 196 women and 67 men (20-85 years) scanned on a GE Lunar Prodigy densitometer were analyzed using enCORE software in 2 ways: (1) placing the head cut-line just beneath the bony protuberance of the chin according to manufacturer recommendation (correct method); (2) placing the head cut-line at the lowest point below the chin and just above the soft tissue at the shoulders (incorrect method). All other cut-lines were fixed. Mean differences in adiposity were examined using Lin's concordance correlation coefficient; equality of means and variances were evaluated using Bradley-Blackwood F-tests. The limits of agreement were displayed as Bland-Altman plots and calculated as the mean difference ±1.96 times the standard deviation of the difference. Correlation coefficients for paired comparisons of adiposity for correct vs incorrect cut-line placement ranged from 0.983-0.999 for all variables (all p < 0.001). Significant mean differences were 172 ± 130, 201 ± 168, 65 ± 122, and -143 ± 336 g for android, gynoid, visceral, and total fat mass, respectively (all p < 0.0001). These differences exceeded our site's least significant change in 66%, 37%, 29%, and 4% of participant scans for android, gynoid, visceral, and total fat mass, respectively. Our findings underscore the importance of careful review of the manufacturer's auto analysis and consistency in conducting serial scans to ensure accurate and precise measures of regional body fat.

Keywords: DXA scan analysis; Regional body composition; visceral adiposity.

MeSH terms

  • Absorptiometry, Photon
  • Adipose Tissue
  • Adiposity*
  • Adult
  • Body Composition*
  • Female
  • Humans
  • Intra-Abdominal Fat / diagnostic imaging
  • Male