Estimating Abdominal Adipose Tissue with DXA and Anthropometry

Authors

  • Alison M. Hill,

    Corresponding author
    1. ATN Centre for Metabolic Fitness & Nutritional Physiology Research Centre, University of South Australia, Adelaide, Australia
    2. Discipline of Physiology, University of Adelaide, Adelaide, Australia.
    Search for more papers by this author
  • Joe LaForgia,

    1. ATN Centre for Metabolic Fitness & Nutritional Physiology Research Centre, University of South Australia, Adelaide, Australia
    Search for more papers by this author
  • Alison M. Coates,

    1. ATN Centre for Metabolic Fitness & Nutritional Physiology Research Centre, University of South Australia, Adelaide, Australia
    Search for more papers by this author
  • Jonathan D. Buckley,

    1. ATN Centre for Metabolic Fitness & Nutritional Physiology Research Centre, University of South Australia, Adelaide, Australia
    Search for more papers by this author
  • Peter R.C. Howe

    1. ATN Centre for Metabolic Fitness & Nutritional Physiology Research Centre, University of South Australia, Adelaide, Australia
    2. Discipline of Physiology, University of Adelaide, Adelaide, Australia.
    Search for more papers by this author

Nutritional Physiology Research Centre, University of South Australia, Adelaide, South Australia 5005, Australia. E-mail: alison.hill@adelaide.edu.au

Abstract

Objective: To identify an anatomically defined region of interest (ROI) from DXA assessment of body composition that when combined with anthropometry can be used to accurately predict intra-abdominal adipose tissue (IAAT) in overweight/obese individuals.

Research Methods and Procedures: Forty-one postmenopausal women (age, 49 to 66 years; BMI, 26 to 37 kg/m2) underwent anthropometric and body composition assessments. ROI were defined as quadrilateral boxes extending 5 or 10 cm above the iliac crest and laterally to the edges of the abdominal soft tissue. A single-slice computed tomography (CT) scan was measured at the L3 to L4 intervertebral space, and abdominal skinfolds were taken.

Results: Forward step-wise regression revealed the best predictor model of IAAT area measured by CT (r2 = 0.68, standard error of estimate = 17%) to be: IAAT area (centimeters squared) = 51.844 + DXA 10-cm ROI (grams) (0.031) + abdominal skinfold (millimeters) (1.342). Interobserver reliability for fat mass (r = 0.994; coefficient of variation, 2.60%) and lean mass (r = 0.986, coefficient of variation, 2.67%) in the DXA 10-cm ROI was excellent.

Discussion: This study has identified a DXA ROI that can be reliably measured using prominent anatomical landmarks, in this case, the iliac crest. Using this ROI, combined with an abdominal skinfold measurement, we have derived an equation to predict IAAT in overweight/obese postmenopausal women. This approach offers a simpler, safer, and more cost-effective method than CT for assessing the efficacy of lifestyle interventions aimed at reducing IAAT. However, this warrants further investigation and validation with an independent cohort.

Ancillary