Usefulness of four surrogate indexes of insulin resistance in middle-aged population in Hefei, China

Abstract

Objective Previous study have shown that lipid accumulation product (LAP), visceral adiposity index (VAI), triglyceride/high-density lipoprotein cholesterol ratio (TG/HDL-C) and triglycerides/glucose index (TyG index) could be simple clinical indicators of insulin resistance (IR) based on anthropometric and/or biochemical parameters. However, the rational and preferred surrogate marker of IR in different population has yet to be validated. The aim of this study was evaluating the practicability of the LAP, VAI, TG/HDL-C, and TyG in predicting IR in middle-aged Chinese population. Methods A cross-sectional study was conducted in 569 Chinese participants (mean age was 48.5; man 67.7%), and each participant completed a questionnaire survey, anthropometric measurement, and biochemical testing. One-way ANOVAs, Chi-squared test, Pearson’s correlation, and multiple logistic regression were used to evaluate the association between VAI, LAP, TG/HDL-C, and TyG with IR. To correctly discriminate individuals with insulin resistance, a receiver operating characteristic (ROC) analysis was conducted for each evaluated variable and the overall diagnostic accuracy was quantified using the area under the ROC curve (AUC). The AUC of evaluated variables were compared using a nonparametric approach. The optimal cut-off points were determined by the Youden’s index, and the corresponding sensitivity and specificity were provided. Results Significant positive correlation was identified between HOMA-IR with TG/HDL-C (r = 0.306), VAI (r = 0.217), LAP (r = 0.381), and TyG (r = 0.371), respectively (all p < .001). After adjustment for potential confounders of IR, compared with the lowest tertiles, odds ratio (95% CI) having IR in the highest tertiles of TG/HDL-C, VAI, LAP and TyG were 6.07 (2.89–12.71), 10.89 (4.37–27.13), 4.68 (2.00–10.92), and 12.20 (5.04–29.56). The area under ROC curves to predict HOMA-diagnosed IR was 0.773 for TG/HDL-C, 0.767 for VAI, 0.806 for LAP, and 0.800 for TyG, respectively. Among those, LAP showed the greatest value of AUC [0.806 (0.763–0.850)] and highest specificity (0.804). Conclusion Compared with other indicators, the LAP and TyG are simple, relatively accurate, clinically available surrogate markers of insulin resistance in middle-aged population in Hefei, China. Among 4 evaluated parameters, the LAP have the highest specificity and the TyG have the highest sensitivity. Key Messages LAP and TyG could be used as simple and alternative methods to identify the individuals at risk for insulin resistance. LAP and TyG have relatively high predictive ability in diagnosis of IR compared with VAI and TG/HDL-C. No significant difference is observed between LAP and TyG in the ability of predicting insulin resistance.