Ömer Osman Eroğlu, Cansın Eroğlu, Sait Erbey, Murat Polat, Bilge Erbey, Mehmet Alican Sapmaz, Çağanay Soysal
Background/Objectives: Universal screening for gestational diabetes mellitus (GDM) is performed at 24-28 gestational weeks, but recommendations for earlier risk stratification differ across guidelines. Overweight pregnant women (body mass index [BMI] 25.0-29.9 kg/m2) without established major GDM risk factors constitute a clinical gray zone for which current guidelines provide no explicit directive. We evaluated the first-trimester diagnostic performance of two composite anthropometric indices-the Body Roundness Index (BRI) and A Body Shape Index (ABSI)-for subsequent IADPSG-defined GDM within this specific clinical niche. Methods: This prospective single-center cohort study enrolled 198 consecutive overweight pregnant women without major GDM risk factors at 11 + 0 to 13 + 6 gestational weeks; 186 participants completed the 75 g OGTT at 24-28 weeks and were analyzed. Anthropometric measurements were obtained by trained nurses blinded to outcome. Glucose was assayed by hexokinase enzymatic reference method (intra-assay CV ≤ 1.5%). Discrimination, calibration, reclassification (NRI, IDI), decision curve analysis, and three independent internal validation procedures were performed. Results: GDM prevalence was 18.28% (34/186). BRI achieved an AUC of 0.905 (95% CI 0.835-0.974) and ABSI an AUC of 0.889 (95% CI 0.815-0.963), both significantly greater than BMI (AUC 0.705; DeLong p < 0.001). The optimal BRI cutoff of 4.45-interpreted as a cohort-specific analytic anchor rather than a transferable clinical threshold-yielded 85.3% sensitivity and 90.8% specificity. Adding BRI to BMI improved discrimination by 45.6% (IDI). Bootstrap optimism remained below 0.01 across all models. Conclusions: Following internal validation through bootstrap optimism correction and cross-validation, first-trimester BRI and ABSI provided strong supplementary risk stratification for GDM in overweight pregnant women without major risk factors. External validation in independent populations is required before clinical implementation.