Predictive Potential of the Triglycerides-Glucose Index for Type 2 Diabetes in Benin: A Descriptive Cross-Sectional Study

Abstract

This study investigates the Triglycerides-Glucose (TyG) index's predictive potential for type 2 diabetes (T2D) in the Beninese population. Among 850 participants, including 327 with T2D, 25 with prediabetes, and 491 without diabetes, various risk factors were assessed. Anthropometric data, fasting blood glucose, cholesterol levels, and insulin were measured, and indices (HOMA1, HOMA2, TyG) were calculated. Statistical analyses utilized SigmaPlot 14, employing parametric and non-parametric tests based on data distribution. Data were presented as mean ± SD for quantitative, and proportions for qualitative variables. Normality was checked via the Shapiro-Wilk test. Parametric (ANOVA, t-test) and non-parametric (Kruskal-Wallis, Mann-Whitney) tests were employed based on data distribution. Pearson's chi-square test compared proportions, and associations were evaluated with odds ratios (OR). ROC curves assessed diagnostic performance (p<0.05). Results revealed a strong association between insulin resistance (IR) and T2D, with significant correlations to body mass index, waist circumference, blood glucose, cholesterol levels, and triglycerides. The TyG index correlated significantly with HOMA1-IR and HOMA2- IR. Elevated TyG index levels were observed in subjects with prediabetes, T2D, visceral obesity, age ≥ 45 years, family history of diabetes, and hypertension. The TyG index demonstrated good diagnostic performance, with an area under the curve of 0.82, outperforming other ratios. The TyG index emerged as a reliable marker for IR, exhibiting strong correlations with established HOMA models. It proved efficient, particularly in subjects with fasting blood glucose levels below 1 g/L, suggesting its potential as a predictive tool for T2D in the Beninese population.