Introduction: Visceral adipose tissue (VAT) has been shown to contribute to obesity-associated metabolic derangements. The measurement of VAT typically involves imaging techniques such as magnetic resonance imaging (MRI), but these are expensive and time-consuming. Some dual-energy X-ray absorptiometry (DXA) scanners provide an option for measuring VAT, but this has not been validated during dynamic weight loss. Thus, the aim of this study was to compare the agreement between DXA and MRI in tracking changes in VAT during weight loss.
Methods: This study involved 101 postmenopausal women with obesity (mean age 57.9±4.2 years and mean BMI 34.5±2.5 kg/m2) who took part in the TEMPO Diet Trial (Australian New Zealand Clinical Trials Registry Reference Number 12612000651886). Half of the women (n=50) were randomised to 12 months of moderate energy restriction (25–35% energy restriction) with a food-based diet (“MODERATE” intervention), and the other half (n=51) were randomised to 4 months of severe energy restriction (65–75%) with a total meal replacement diet followed by moderate energy restriction until 12 months (“SEVERE” intervention). Abdominal VAT was measured with DXA (Discovery W model, Hologic Inc, Bedford, MA, USA) and MRI (Discovery MR750 3.0T from GE Healthcare, Milwaukee, WI, USA) at 0, 4, 6 and 12 months after diet commencement. Bland Altman plots were used to compare the two techniques.
Results: The change in VAT from baseline (0 months) at 4, 6 and 12 months, as determined using MRI, was 16.8±11.8, 24.5±16.1 and 26.4±23.4% of baseline values for women in the MODERATE intervention, and 48.8±18.4, 58.5±18.3 and 57.1±18.6% for those in the SEVERE intervention. DXA underestimated the change in VAT at all time points, albeit this underestimation was not significant at 4 months (1.6% and 7.5% for the MODERATE and SEVERE interventions, respectively), and became significant with time with greater weight loss (6.4% and 13.7% at 6 months for the MODERATE and SEVERE interventions, respectively; 10.5% and 18.8% at 12 months). These underestimations represent 9.5-15.4% of the MRI-determined change in VAT at 4 months, 23.4-26.1% at 6 months, and 33.0-39.8% at 12 months.
Conclusion: DXA appears to have limited capability to evaluate changes in average VAT with greater weight loss which may be the consequence of substantial changes in body composition, at least in post-menopausal women with obesity.