Comparison of different insulin resistance surrogates to predict hyperuricemia among U.S. non-diabetic adults

Although it has been well-acknowledged that insulin resistance (IR) plays a critical role in the development of hyperuricemia (HU), specific relationship between IR and HU in non-diabetic patients remains rarely studied, and there is still no large-scale research regarding this issue. This study aims to explore the association between triglyceride glucose (TyG), TyG with body mass index (TyG-BMI), the ratio of triglycerides divided by high-density lipoprotein cholesterol (TG/HDL-C), metabolic score for insulin resistance (METS-IR), and the risk of HU in non-diabetic patients in The United States of America.

Data from the National Health and Nutrition Examination Survey (NHANES) enrolling a representative population aged ≥18-year-old were included to calculate these four indexes. Logistic regression analysis was applied to describe their associations and calculate odds ratios (OR) while the Receiver Operating Characteristic curve was utilized to assess the prediction ability of these four indexes.

A total of 7,743 people (3,806 males and 3,937 females, mean age: 45.17 ± 17.10 years old) were included in this study, among whom 32.18% suffered from HU. After adjustment for sex, age, ethnicity, education background, smoking status, drinking status, systolic blood pressure (SBP), diastolic blood pressure (DBP), metabolic equivalent values (METs), total cholesterol, low-density lipoprotein cholesterol, and estimated glomerular filtration rate, it showed that all four indexes were closely related to HU. Compared with the lowest quartile, OR of the highest quartile of these four indicators for HU were as following respectively: TyG: 5.61 (95% CI: 4.29-7.32); TyG-BMI: 7.15 (95% CI: 5.56-9.20); TG/HDL-C: 4.42 (95% CI: 3.49-5.60); METS-IR: 7.84 (95% CI: 6.07-10.13). TyG, TyG-BMI, TG/HDL-C and METS-IR had moderate discrimination ability for HU, with an AUC value of 0.66 (95% CI: 0.65-0.68), 0.67 (95% CI: 0.65-0.68), 0.68 (95% CI: 0.67-0.69) and 0.68 (95% CI: 0.66-0.69) respectively. Each index showed better prediction ability for HU risk in females than in males.

It was found that the risk of HU was positively associated with the elevation of TyG, TyG-BMI, TG/HDL-C and METS-IR in a large-scale population of U.S., and TyG-BMI and METS-IR have a better ability to identify HU in both genders.