Risk Factors for the Prediction of Hyperglycemia during Pregnancy – A Cohort Study from a Brazilian Public Primary Care Center

Abstract

BACKGROUND – In Brazil, the prevalence of maternal hyperglycemia is among the highest, costs are elevated and there is no evidence to recommend universal screening. OBJECTIVE – To evaluate the performance of risk factors (RF) for predicting hyperglycemia in pregnancy – Mild Gestational Hyperglycemia (MGH) or gestational Diabetes Mellitus (GDM) in public primary-care centers in Brazil. METHODS – A cohort study, including 514 women with a single pregnancy and no hyperglycemia. Study participants were evaluated at GA (gestational age) < or ≥ 20 weeks, and underwent a 75g-OGTT along with glycemic profile (GP) testing between 24 and 28 weeks. Clinical, anthropometric and laboratory data – fasting glucose (FG), glycated hemoglobin (HbA1c), basal insulin and lipid profile were obtained. The most common RF associations (OR and 95% CI and p <0.05) and different cut-off points were tested for the prediction of MGH-GDM. Predictive performance was assessed by Sensitivity/Specificity, negative predictive value NPV (negative predictive value) and false positive rates (FP; 1-Esp). RESULTS – At GA <20 weeks, age ≥25 years, WC (Waist circumference) ≥ 88 cm, BMI pre ≥25 kg/m 2 (pre gestational body mass index) and BMI gest ≥25 kg/m 2 (gestational body mass index ); at GA (gestational age) ≥20 weeks, age ≥25 years, BMI pre ≥25 kg/m 2 and TG (triglicerides) ≥150 mg/dL showed better performace for predicting MGH-GDM. Irrespective of gestational age, FG (Fasting glucose) ≥ 85 mg/dL, HbA1c ≥5.7% and HOMA-IR ≥2.71 were good predictors to rule out the risk of these complications. CONCLUSION – The results of this study should contribute to define the best diagnostic approach to MGH-GDM in our center and in others with similar characteristics.