LAUSR

Abstract BACKGROUND: Intra-abdominal fat (IF) and visceral adipose tissue (VAT) can be assessed using magnetic resonance (MRI) or dual-energy X-ray absorptiometry (DXA). Surrogates to estimate IF include “Visceral-adiposity index” (VAI), “Lipid-accumulation product” (LAP) and “Relative Fat Mass Index” (RFMI). METS-IR, a non-insulin-based insulin resistance index is strongly related to the metabolic syndrome traits OBJECTIVE: To develop a novel index combining METS-IR and anthropometric measures to estimate IF using DXA as gold standard. METHODS: We developed METS-VAT using METS-IR and the waist-height index (WHtI) by non-linear regression analyses. An inception (n=363) and a validation (n=184) cohorts were used. Lean, overweight and obese subjects were assessed with laboratory measurements, anthropometry and DXA. We also assessed IF by MRI in a subset of the validation cohort which included individuals with and without T2D. Finally, we evaluated the capacity of METS-VAT to detect increased IF defined as Visceral Fat Area (VFA) >100 cm2 by MRI. The same analysis were applied for subjects with VAT above 80th percentile and compared them with VAI, LAP, RFMI, METS-IR, BMI and WHtI. RESULTS: We defined METS-VAT as: METS-VAT=6.913+0.008[Ln(METS-IR)^3 ]+3.592[Ln(WHtI)^3 ]+0.350(Sex).Estimated Intra-abdominal fat (g)=e^6.913+0.008[Ln(METS-IR)^3 ]+3.592[Ln(WHtI)^3 ]+0.350(Sex). In the inception cohort, METS-VAT had the highest correlation with IF (ρ=0.716, 95%CI 0.608-0.808) compared to VAI (ρ=0.193, 95%CI 0.069-0.305), LAP (ρ=0.523, 95%CI 0.438-0.619), RFMI (ρ=0.429, 95%CI 0.302-0.539), BMI (ρ=0.576, 95%CI 0.502-0.657), WHtI (ρ=0.622, 95%CI 0.561-0.691) and METS-IR (ρ=0.299, 95%CI 0.177-0.540) adjusted for age and T2D. Furthermore, in the validation cohort, METS-VAT had a significant correlation with IF assessed with DXA (ρ=0.652, 95%CI 0.542-0.737) compared to VAI (ρ=0.333, 95%CI 0.218-0.445), LAP (ρ=0.503, 95%CI 0.389-0.606), RFMI (ρ=0.183, 95%CI 0.042-0.308) and METS-IR (ρ=0.650, 95%CI 0.576-0.735), WHtI (ρ=0.426, 95%CI 0.279-0.568) and BMI (ρ=0.625, 95%CI 0.536-0.713) adjusted for sex. Exploring the diagnostic capacity for increased IF and DXA-VAT in the validation cohort using ROC curves, we found that METS-VAT had the highest AUC to identify increased DXA-VAT (AUC=0.860, 95%CI 0.799-0.921) compared to BMI (AUC=0.800, 95%CI 0.724-0.876), LAP (AUC=0.792, 95%CI 0.718-0.866), WHtI (AUC=0.764, 95%CI 0.689-0.839), VAI (AUC=0.670, 95%CI 0.576-0.765) and RFMI (AUC=0.457, 95%CI 0.365-0.585). METS-VAT also had the highest AUC for increased IF by MRI-S (AUC=0.807, 95%CI 0.728-0.887) compared to BMI, LAP, WHtI, VAI and RFMI. CONCLUSION: METS-VAT is a novel index strongly correlated with VAT assessed by DXA and IF by MRI and has better performance compared with other visceral adipose tissue surrogates.