Abstract
OBJECTIVE: The present study examined the ability of the body roundness index (BRI) to predict the incidence of diabetes and prediabetes among adults in the USA. METHOD: The study enrolled 11,980 adults aged ≥ 20 years from the National Health and Nutrition Examination Survey (NHANES). Logistic regression served as the primary method for analyzing the relevant link between BRI and the incidence of diabetes and prediabetes, including univariate analysis, multivariate regression analysis, smooth curve fitting analysis, and subgroup analysis. What's more, receiver operating characteristic (ROC) analysis was applied to confirm the predictive values of BRI for diabetes and prediabetes. RESULTS: Each unit higher than BRI was associated with a 17% increased risk of diabetes and prediabetes after covariate adjustments (OR: 1.17, 95% CI: 1.07-1.27). Those with BRI in the high scores (Q4) possessed an increased likelihood of having diabetes and prediabetes than individuals in the reference group (OR: 1.83, 95% CI: 1.29-2.58). A smooth curve fitting analysis revealed a non-linear trend. The results across all population subgroups were uniform to those of the total population. The ROC curve indicated that the BRI was the best predictor of diabetes and prediabetes among other anthropometric indices. CONCLUSIONS: Diabetes and prediabetes occurrence rates and BRI have a positive and non-linear relationship in American adults. The BRI indices could function as predictive markers for diabetes and prediabetes.