Measurement of Visceral Adipose Tissue in Persons With Spinal Cord Injury by Magnetic Resonance Imaging and Dual X-Ray Absorptiometry: Generation and Application of a Predictive Equation.

Purpose: Dual energy X-ray absorptiometry (DXA) and magnetic resonance imaging (MRI) permits quantification of visceral adipose tissue (VAT). However, DXA has not been validated against MRI in persons with chronic spinal cord injury (SCI). A predictive equation was generated from the measurement of VAT by MRI, a "gold" standard to quantitate VAT, compared to that of DXA, a method with several practical advantages.

Method: DXA and MRI scans were performed in 27 participants with SCI. MRI multiaxial images were captured for VAT analysis. DXA-VAT was quantified at the android region (DXA-VAT) using enCore software. Android regions of DXA and MRI were matched using android height. Volumes of multiaxial MRI-VAT and subcutaneous adipose tissue (SAT) were quantified for the android region (MRI-VAT, MRI-SAT) and total trunk (MRI-VAT). Linear regression analysis was used to establish the proposed predication equations. The prediction equations were then applied to an independent sample that consisted of 98 participants with SCI. Bland-Altman analysis was used to determine the limits of agreement.

Results: DXA-VAT predicted 92% of the variance in MRI-VAT (SEE = 252.5, p < 0.0005) and 85% of the variance in MRI-VAT (SEE = 1526.9, p < 0.0005). DXA-SAT predicted 81.5% of the variance in MRI-SAT (SEE = 458.2, p < 0.0005). Bland-Altman analysis revealed a high level of agreement between MRI-VAT and DXA-VAT (mean bias = 58.45 cm). A predicted mean DXA-VAT of 995.2 cm was estimated as the population-specific cut-off point for high levels of VAT.

Conclusion: DXA-VAT may accurately predict MRI-VAT in persons with SCI. The ability of DXA to detect VAT changes in longitudinal studies in persons with SCI should be performed.