Predictive value of glycosylated hemoglobin levels for large for gestational age infants in women with pregestational diabetes according to body mass index

Objective: This study aimed to evaluate whether body mass index (BMI) modifies the predictive ability of glycosylated hemoglobin (HbA1c) levels for large for gestational age (LGA) infants in women with pregestational diabetes. Methods: This multicenter retrospective cohort study included 402 pregnant women with pregestational diabetes from 17 university hospitals in South Korea. Pregestational diabetes was defined as early diagnosis of gestational diabetes mellitus, type 1 diabetes, or type 2 diabetes. Participants were categorized on the basis of their BMI (<25 vs. ≥25 kg/m²). The primary outcome was LGA birth. Logistic regression analysis was performed to assess the association between trimester-specific HbA1c levels and the risk of LGA births in each BMI group. Receiver operating characteristic (ROC) curve analysis was used to evaluate predictive performance. Results: Among 402 women with pregestational diabetes, 183 (45.5%) had pre-pregnancy BMI <25 kg/m² and 219 (54.5%) had BMI ≥25 kg/m². In the BMI <25 kg/m² group, higher HbA1c levels in all trimesters were significantly associated with an increased risk of LGA births, with the strongest association being observed in the third trimester (odds ratio [OR], 2.284; 95% confidence interval [CI], 1.458-3.579). In women with BMI ≥25 kg/m², only the third-trimester HbA1c level was significantly associated with the risk of an LGA birth (OR, 1.455; 95% CI, 1.069-1.981). ROC analysis indicated acceptable to moderate discrimination for LGA prediction in women with BMI <25 kg/m² (area under the curve [AUC], 0.70-0.82), but poor to modest discrimination in those with BMI ≥25 kg/m² (all AUCs ≤0.65). Conclusion: The predictive utility of HbA1c levels for LGA varied on the basis of the pre-pregnancy BMI and was greater in women with BMI <25 kg/m² than in those with BMI ≥25 kg/m².