Central adiposity markers, plasma lipid profile and cardiometabolic risk prediction in overweight-obese individuals

BACKGROUND: Waist circumference (WC) is the currently recommended marker of central fat for cardiometabolic risk screening. Alternative surrogate markers have been recently proposed to better reflect the metabolic impact of central fat accumulation per se, based on WC normalization by height (Weight-to-Height Ratio - WtoH; Body Roundness Index - BRI) or body mass index (BMI) without (A Body Shape Index - ABSI) or with inclusion of plasma triglyceride and HDL-cholesterol concentrations (Visceral Adiposity Index - VAI). METHODS: We investigated associations between WtoH, BRI, ABSI or VAI and insulin resistance (HOMA-index) or metabolic syndrome (MetS) in a general population cohort from the North-East Italy Mo.Ma. study (n = 1965, age = 49 ± 13 years, BMI = 26.7 ± 5.2 kg/m2). Baseline values were also evaluated as predictors of future insulin resistance and MetS in overweight-obese individuals undergoing 5-year follow-up (Ow-Ob) (n = 263; age = 54 ± 9, BMI = 30,7 ± 4,1). RESULTS: Compared to WC or BMI, basal WtoH and BRI were similarly associated with baseline HOMA and MetS prevalence after multiple adjustments (P < 0.001) and all markers similarly predicted 5-year HOMA and MetS (P < 0.001). Under basal conditions, superimposable results were observed for VAI whereas ABSI was less accurate or unable to identify baseline HOMA and MetS (p < 0.05 vs WtoH-BRI-VAI-WC-BMI). VAI had highest 5-year risk predictive value in Ow-Ob [ROC Area Under the Curve (AUC) VAI > WtoH-BRI-WC-BMI; p < 0.05] while no predictive value was in contrast observed for ABSI (ROC AUC ABSI < WtoH-BRI-WC-BMI; p < 0.05). Using alternate formulae with plasma lipid inclusion in ABSI and removal from VAI calculations completely reversed their 5-year predictive value and AUC. CONCLUSIONS: The current findings do not support replacement of WC with height-normalized anthropometric central fat surrogate markers to predict cardiometabolic risk in the general and overweight-obese population. BMI-normalization impairs risk assessment unless plasma lipid concentrations are available and included in calculations.