| Summary: | Aims: To evaluate the performance of adiponectin in predicting insulin need during pregnancy. Methods: Case-control analysis, nested from an ongoing prospective cohort, including 254 singleton pregnancies of women with gestational diabetes mellitus (GDM) at a tertiary teaching hospital. Patients were divided into 2 groups (DIET group and INSULIN group) and were compared regarding clinical and laboratory variables. After univariate analysis, variables that presented statistical significance or clinical relevance were included in the logistic regression. Two prediction models (INITIAL model: clinical and laboratory data available after the diagnosis of GDM; and INITIAL plus adiponectin model) were developed to evaluate their accuracy in predicting insulin need. A p value < 0.05 was considered significant. Results: 202 patients (79.5%) were in the DIET group and 52 (20.5%) were in the INSULIN group. Prior GDM, presence of hypertension, first trimester fasting blood glucose value, fasting and first hour values at the oral glucose tolerance test and number of abnormal results at this test were associated with insulin requirement during pregnancy. In the prediction model, adiponectin was not a predictor of insulin need. Conclusions: Considering that the addition of adiponectin does not improve the performance of the prediction model, it might not be used as a predictor of insulin need for glycemic control during pregnancy.
|
|---|