Abstract
Background: Globally, the prevalence of metabolic syndrome (MetS) is on the rise, especially in Arab countries, which emphasizes the need for reliable ethnic-specific biochemical screening parameters. Methods: Two hundred twenty-one Lebanese adults were enrolled in this cross-sectional study. Biochemical parameters including Homeostasis Model Assessment (HOMA), Triglyceride and Glucose index (TyG), ratio of Triglycerides to High-Density Lipoprotein Cholesterol (TG/HDL-C), Atherogenic Index of Plasma (AIP), and Visceral Adiposity Index (VAI) were assessed for their prediction of MetS. Analysis of covariance, logistic regression, expected-versus-observed case ratio were used to determine model calibration, concordance statistic, area under the receiver operating characteristic curve (AUC) and 95% confidence intervals (CIs), sensitivity, specificity, and negative and positive predictive values (PPV, NPV). Results: The prevalence of MetS was 44.3%. All biochemical parameters were significantly associated with MetS, with a strong model discrimination (c-statistic between 0.77 and 0.94). In both sex categories, TyG best predicted MetS (females: cut-off value, 8.34; males: cut-off value, 8.43) and showed good estimation among females, but overestimation among males. HOMA had the lowest discriminatory power in both sex categories. Conclusions: This study suggests that TyG best predicts MetS, while HOMA has the lowest predictive power. Future larger studies need to focus on harmonizing ethnic specific cut-offs and further validating our results.