Predicting Gestational Diabetes from the First Trimester - an exploration of alternative screening processes
Gestational diabetes is recognised to be an increasingly common and treatable cause of maternal and infant morbidity. Screening and formal diagnosis of gestational diabetes have been at the centre of much of the scientific debate. This thesis aims to explore the ability of a panel of biomarkers in the first trimester of pregnancy to predict gestational diabetes and examines their link with the clinical complications of GDM. 248 women with known risk factors for GDM were recruited in the first trimester and had serum samples taken for c-reactive protein, sex hormone binding globulin, adiponectin anhydroglucitol. They underwent oral glucose tolerance testing at 28 weeks gestation and had prospectively collected perinatal outcome data. Cord blood c-peptide levels were drawn as an indicator of fetal hyperglycaemia. Our study found that adiponectin and SHBG demonstrated a correlation to the risk of onset of GDM in the univariate analysis. However, after adjustment for BMI, family history and ethnicity in the multivariate analysis SHBG loses significance. Following adjustment for BMI, family history and ethnicity 1st trimester 1,5 AG becomes a significant predictor of GDM. Mean 1,5 AG levels are significantly lower in the first trimester in women that will go on to develop GDM. However, none of these biomarkers exhibited a specific threshold value at which onset of GDM could be predicted absolutely with acceptable sensitivity and specificity. Our analysis of the relationship between sex hormone binding globulin values in the first trimester and neonatal hyperinsulinaemia (cord blood c-peptide> 90th centile/1.7µL) shows that in women with a high SHBG (>350nmol/L) early in pregnancy have an odds ratio of 2.7 to have a non-hyperinsulinaemic neonate. Our study also explored the patient experience of the OGTT. We found that despite the OGTT having adverse physical effects and being inconvenient, most women find it an acceptable test and would agree to be screened again in the future. BMI >30kg/m2 and non-Caucasian ethnicity remain as independent predictors of GDM status in the adjusted analysis of the history-identified risk factors. Adiponectin and 1,5 AG remain independent predictors of GDM status after adjusting for all other variables.