Variations of blood cells in prediction of gestational diabetes mellitus

This study aims to evaluate the value of increasing inflammation in predicting gestational diabetes mellitus (GDM).

Subjects in this cross-sectional study included 302 GDM and 310 normal pregnancies before 20 weeks. Sociodemographic and pregnancy characteristics as well as blood parameters were acquired by maternal health booklet, medical records and laboratory information systems. Blood cell parameters were compared between the two groups by independent sample t-tests. Multivariate logistic regression, χ²-test, receiver operator characteristic curve (ROC), and Fisher's linear discriminant were performed to analyze the screening effects of variables in developing GDM.

Women with GDM had significantly higher neutrophil (NEU), lymphocyte (LYM), platelet (PLT) and erythrocyte (RBC) counts, and were positively correlated with GDM. NEU (odds ratios, OR, 1.22) and LYM (OR, 2.01) were independently associated with the development of GDM (P<0.001). The OR of the mean platelet volume (MPV) and mean cell volume (MCV) were 0.84 and 0.92, respectively (P<0.01 for both). The efficiency of Fisher's equations in correctly classifying cases of GDM from 4 to 20 weeks of gestation was 70.06%.

Maternal WBC, RBC, and PLT counts are important correlates of GDM. Increased volume of RBC and PLT might protect pregnant women from development of GDM.