Body roundness index shows a short-term, sex-specific advantage over body mass index in predicting incident diabetes: a longitudinal prospective cohort study.

PURPOSE: Diabetes Mellitus (DM) is accompanied by insulin resistance and excessive adiposity. The body roundness index (BRI), an emerging anthropometric measure, comprehensively reflects visceral fat derived from height and waist circumference. The purpose of this study was to evaluate the association between BRI and DM risk, and to investigate whether BRI serves as a potential predictive indicator for DM incidence. METHODS: A total of 3,159 participants (1,286 men and 1,873 women) were sourced from the Korean genome and epidemiology study (KoGES) from 2001 to 2022. We used Kaplan-Meier analysis to estimate the cumulative incidence of DM across BRI quartiles. Multivariable Cox proportional hazards regression models were performed to examine the association between BRI levels and incident DM. The discriminative power of BRI and BMI was compared using the time-dependent receiver operating characteristic (ROC) curve analysis. RESULTS: During a median 17.5-year follow-up period, 766 DM events occurred. The cumulative incidence of DM was significantly increased with higher BRI quartiles in the Kaplan-Meier curves (log-rank p-value < 0.001). Multivariable Cox proportional hazards regression analysis uncovered a dose-response association between BRI and risk of incident DM in both sexes, with higher hazard ratios (HRs) in women (Q4:Q1, adjusted HR 2.52, 95% CI 1.79-3.54). Additionally, time-dependent ROC curve proved that BRI was superior to BMI in short-term (4-5 years) discriminatory performance, particularly in women (5-year AUC: BRI 0.708 vs. BMI 0.662, p = 0.042). CONCLUSIONS: In this cohort study, baseline BRI was associated with an increased risk of DM incidence, and BRI showed superior discriminatory performance relative to BMI specifically for short-term (4-5 year) prediction, particularly among women.