DXA: can it be used as a criterion reference for body fat measurements in children?

Obesity (Silver Spring). 2008 Feb;16(2):457-62. doi: 10.1038/oby.2007.81.


Objective: Dual-energy X-ray absorptiometry (DXA) is often cited as a criterion method for body composition measurements. We have previously shown that a new DXA software version (Hologic Discovery V12.1) will affect whole-body bone mineral results for subjects weighing <40 kg. We wished to reanalyze pediatric whole-body scans in order to assess the impact of the new software on pediatric soft-tissue body composition estimates.

Methods and procedures: We reanalyzed 1,384 pediatric scans (for ages 1.7-17.2 years) using Hologic software V12.1, previously analyzed using V11.2. Regression analysis and ANCOVA were used to compare body fat (total body fat (TBF), percentage fat (%BF)), and non-bone lean body mass (LBM) for the two versions, adjusting for gender, age and weight.

Results: Software V12.1 yielded values that were higher for TBF, lower for LBM, and unchanged for DXA-derived weight in subjects weighing <40 kg. Body composition values for younger, smaller subjects were most affected, and girls were more affected than boys. Using the new software, 14% of the girls and 10% of the boys were reclassified from the "normal" %BF range to "at risk of obesity," while 7 and 5%, respectively, were reclassified as obese.

Discussion: Hologic's newest DXA software has a significant effect on soft-tissue results for children weighing <40 kg. The effect is greater for girls than boys. Comparison of TBF estimates with previous studies that use older DXA instruments and software should be done with caution. DXA has not yet achieved sufficient reliability to be considered a "gold standard" for body composition assessment in pediatric studies.

Publication types

  • Evaluation Study
  • Research Support, U.S. Gov't, Non-P.H.S.

MeSH terms

  • Absorptiometry, Photon / instrumentation
  • Absorptiometry, Photon / methods*
  • Adipose Tissue*
  • Adolescent
  • Age Factors
  • Body Composition
  • Body Weight
  • Child
  • Child, Preschool
  • Female
  • Humans
  • Infant
  • Male
  • Obesity / pathology
  • Regression Analysis
  • Reproducibility of Results
  • Risk Factors
  • Sensitivity and Specificity
  • Sex Factors
  • Software*