Abstract
Background: Gestational diabetes mellitus (GDM) is among the most common metabolic complications of pregnancy, affecting between 5% and 14% of pregnancies worldwide. Early and accurate diagnosis is essential to reduce adverse maternal and fetal outcomes. Although the 2-hour oral glucose tolerance test (OGTT) remains the diagnostic gold standard, it is cumbersome, time-consuming, and poorly tolerated by many pregnant women.
Objective: To determine the diagnostic accuracy of glycosylated haemoglobin (HbA1c) for detecting GDM among pregnant females at 18 to 24 weeks of gestation, using the OGTT as the reference standard.
Methods: A comparative cross-sectional study was conducted at Jinnah Hospital, Lahore, enrolling 168 consecutive pregnant women aged 18–40 years at 18–24 weeks of gestation. All participants underwent concurrent HbA1c measurement and a 2-hour 75 g OGTT. Receiver operating characteristic (ROC) curve analysis was performed to identify the optimal diagnostic threshold, and standard diagnostic accuracy metrics were computed.
Results: The true prevalence of GDM in this cohort was 31.5%. HbA1c demonstrated outstanding overall discriminatory capacity, with an area under the ROC curve (AUC) of 0.980 (95% CI: 0.962–0.997; p < 0.001). An HbA1c threshold of 5.3% was identified as the optimal cut-off, yielding a sensitivity of 94.3%, specificity of 96.5%, positive predictive value (PPV) of 92.6%, and negative predictive value (NPV) of 97.4%.
Conclusion: HbA1c at a threshold of 5.3% demonstrates excellent diagnostic accuracy for GDM screening in this South Asian obstetric population, offering a reliable, patient-friendly alternative to support primary metabolic triage and reduce over-reliance on the OGTT