Low predictive value of traditional risk factors in identifying women at risk for gestational diabetes

Abstract

There is no worldwide agreement on the best way to screen for gestational diabetes mellitus (GDM), and different diagnostic methods have been developed in order to identify women at risk. The aim of this study was to evaluate the prevalence and predictive value of the traditional risk indicators for GDM in a large group of Caucasian women. We evaluated the frequency distribution of age, body mass index (BMI), prior macrosomia, prior GDM, and family history of diabetes of 1,414 pregnant women with GDM and 1,011 healthy pregnant women. The distribution of risk factors in both groups was different and significantly higher in GDM women. The cut-off value for age was 28 years, and 23 kg/m2 for BMI. The accumulation of two or more risk factors was frequent in GDM, but not in healthy women. By multiple logistic regression, there were significant interactions between independent variables of interest and GDM (OR: 3.19; p<0.001; sensitivity: 57.9%, specificity: 69.8%). The strongest predictors were prior GDM (OR: 4.35;95% CI: 2.42-7.82) and a family history of diabetes (OR: 3.03; 95% CI: 2.47-3.72); less predictive were age (OR: 1.69;95% CI: 1.44-1.99), BMI (OR: 1.50; 95% CI: 1.28-1.77), and prior macrosomia (OR: 1.64; 95% CI: 1.19-2.26). Selective screening based on traditional risk factors for GDM had relatively low sensitivity, and identified <60% of Caucasian women at risk. The cut-off value for BMI as a risk indicator (23 kg/m2) was lower than that proposed by guidelines about screening for GDM.