Assessment of risk factors and glycosylated haemoglobin in early pregnancy as predictors of diabetes in pregnancy

Abstract

Background: The aim of the study was to determine the performance of history of risk factors and universal HBA1c testing as screening tools for diabetes mellitus in the first trimester of pregnancy using OGTT as a gold standard. Methods: A prospective cross-sectional study conducted between 8 and 13±6 weeks in 305 consecutive pregnancies in the antenatal clinics of the University of Port Harcourt Teaching (UPTH) and Rivers State University Teaching Hospital (RSUTH) between January and August 2020. Each woman had oral glucose tolerance test (OGTT), and glycosylated haemoglobin (HBA1c) levels estimation. Multivariate logistic regression analysis was carried out with history of risk factors and HBA1c level as independent variables and OGTT as the dependent variable for the assessment of their predictive performances. Results: The prevalence of DM was 28.85%, 2.62% and 31.48% for GDM, pre-gestational and for both respectively. Family history of DM was associated with high specificity (91.4%) and negative predictive value (NPV) of 68.7% but low sensitivity (9.4%) and positive predictive value (PPV) (33.3%). The receiver operator characteristic curve for HBA1c revealed a significant area under the curve value: 0.653 (CI: 058-0.72), p<0.01. The optimal cut-off for HBA1c from Youden index was 5.25%. HBA1c levels had high specificity (88.5%) and NPV (75.2%) with low sensitivity (36.5%) and PPV (59.3%). Multivariate logistic regression analysis showed HbA1c as the only independent predictor of GDM (p=0.0001). Conclusions: The high prevalence of diabetes (31.48%), underscores the need for universal screening in early pregnancy. The high NPV and specificity of the risk factors for GDM and HBA1c levels better predict pregnancies that are not likely to develop GDM, but they are not suitable for diagnosis because of the low sensitivity and PPV.