Regional Intra‐Subject Variability in Abdominal Adiposity Limits Usefulness of Computed Tomography

Abstract

Computed tomography (CT) and magnetic resonance imaging, the most accurate methods of abdominal fat measurement, have been applied using a number of protocols, ranging from single-slice area determination to multiple-slice volume calculation. The aim of this study was to assess the validity of single-slice CT for abdominal fat area measurement by estimating the intra-subject variability in abdominal fat areas and comparing the ranking of subjects across four contiguous abdominal levels. Nineteen premenopausal women (age, 35.3 +/- 1.4 years; mean +/- SE) were studied. CT was used to measure intra-abdominal fat (IAF) area, percentage of total intra-abdominal area (%IAF), subcutaneous abdominal fat (SAF) area, and IAF/SAF at four adjacent cross-sectional lumbar levels (L2-L4). Intra-subject variability (percentage) was defined as SD/mean x 100. Total body fat was measured by DXA, which was further analyzed for central abdominal fat. Mean body mass index was 24.9 +/- 1.0 kg/m(2). The average (range) intra-subject variability was 28% (8% to 61%) for IAF, 46% (19% to 124%) for %IAF, 26% (14% to 38%) for SAF area, and 19% (7% to 71%) for IAF/SAF. The pattern of this variability was not uniform between subjects, because their ranking by IAF area was markedly different at each CT level. We demonstrated significant intra-subject variability in CT-measured adipose tissue areas across four predetermined sites. This resulted in a difference in the ordering of subjects by IAF at each of the four imaging sites, suggesting that the usefulness of single-slice CT in the assessment of abdominal adiposity in premenopausal women may be limited, particularly when performed for the purpose of making comparisons between subjects based on abdominal fat area.