Predictive Effects of Different Insulin Resistance Indexes on Diabetes Risk in Hypertensive Population / 中国医学科学院学报

ABSTRACT

Objective To explore the roles of different insulin resistance indexes[triglyceride-glucose (TyG),triglyceride (TG)/high-density lipoprotein cholesterol (HDL-C),and metabolic score for insulin resistance (METS-IR)]and combinations of two indexes in predicting diabetes risk in hypertensive population. Methods The survey of hypertension was conducted for the residents in Wuyuan county,Jiangxi province from March to August in 2018.The basic information of hypertensive residents was collected by interview.Blood was drawn on an empty stomach in the morning and physical measurements were carried out.Logistic regression model was employed to analyze the relationship between different insulin resistance indexes and diabetes,and the area under the receiver operating characteristic curve was used for evaluating the predictive effects of each index on diabetes risk. Results A total of 14 222 hypertensive patients with an average age of (63.8±9.4) years old were included in this study,including 2616 diabetic patients.The diabetic hypertensive population had higher TyG (t=50.323,P<0.001),TG/HDL-C (Z=17.325,P<0.001),and METS-IR (t=28.839,P<0.001) than the non-diabetic hypertensive population.Multivariate analysis showed that each insulin resistance index was positively correlated with diabetes risk.The area under curve of each insulin index was in a descending order of TyG (0.770)> METS-IR (0.673)> TG/HDL-C (0.620).The difference in the area under curve between two indexes was statistically significant[TyG vs.TG/HDL-C (Z=42.325,P<0.001);TyG vs.METS-IR(Z=17.517,P<0.001);METS-IR vs.TG/HDL-C (Z=10.502,P<0.001)]. Conclusions Elevated insulin resistance indexes can increase the risk of diabetes.TyG and the combination of indexes outperform TG/HDL-C and METS-IR in the prediction of diabetes.