Endocrine Abstracts

Gestational Diabetes Mellitus (GDM) affects approximately 1 in 7 pregnancies in Ireland. Whilst the mainstay of GDM treatment is lifestyle modification, a significant proportion of women require pharmacological therapy to achieve target blood glucose levels. This work examines if maternal characteristics at the time of GDM diagnosis can predict which women will require treatment intensification. The EMERGE randomised controlled trial (NCT02980276) compared the efficacy of early metformin (vs placebo) in addition to lifestyle modification for the treatment of GDM. For the present work, anonymised data were extracted from the EMERGE database. The primary outcome of interest was insulin initiation. Candidate predictor variables known to be associated with insulin resistance were also extracted, as measured at the time of GDM diagnosis. Multivariate logistic regression analyses were performed in each treatment subgroup to identify which variables were independent predictors of insulin initiation. In the placebo group (n = 262), the independent predictors of requiring pharmacological treatment were: gestational age (P = 0.001), HbA1c (P = 0.029), and fasting glucose on the oral glucose tolerance test (OGTT) (P = 0.000). In the group who received metformin (n = 259), the independent predictors of requiring additional insulin treatment were: gestational age (P = 0.002), HbA1c (P = 0.003), fasting glucose (P = 0.016), 1-hour glucose (P = 0.028), and 2-hour glucose on OGTT (P = 0.012). This analysis shows that commonly measured parameters can be used at the time of GDM diagnosis to estimate the probability of subsequently requiring pharmacological treatment of hyperglycaemia, with moderate accuracy. Once externally validated, these models may allow a more personalised approach to GDM care in the Irish population.