Association and discriminatory performance of body roundness index for dysglycemia among adults in Ethiopia

Abstract

Body roundness index is a novel anthropometric measure of body fat based on height and waist circumference, but its association with dysglycemia remains understudied. We aimed to assess the association and discriminatory performance of between body roundness index and dysglycemia among adult patients at outpatient department of a tertiary hospital in Ethiopia. A hospital-based cross-sectional study was conducted. A binary logistic regression and receiver operating characteristic curve analysis were used to assess the association. A total of 427 adults aged ≥ 35 years were included in this study. The mean body roundness index of the study population was 4.22 ± 1.12. BRI was significantly associated with higher odds of dysglycemia. A graded increase in risk was observed across BRI quartiles, with participants in the highest quartile (Q4) showing markedly increased odds compared with Q1 (AOR: 16.6; 95% CI: 7.09–39.14). Similar associations were observed for prediabetes (Q4 vs. Q1: AOR: 11.39; 95% CI: 3.73–34.76) and type 2 diabetes (Q4 vs. Q1: AOR: 29.2; 95% CI: 9.84–86.59). A receiver operating characteristic analysis indicated that it had good discriminatory ability for dysglycemia (AUC 0.813, 95% CI: 0.772–0.854). When compared with other anthropometric indices, BRI demonstrated a comparable AUC to BMI (0.810; 95% CI: 0.769–0.850) and waist circumference (0.833; 95% CI: 0.796–0.871). The optimal body roundness index cutoff was 4.17 (sensitivity 73.6%, specificity 79.2%), with sex-specific cutoffs of 4.02 for males and 4.17 for females. Body roundness index showed a strong, graded association with dysglycemia.