Objectives To explore the relationship between the levels of maternal oxidative stress and glycaemia during pregnancy and to compare the predictive values of 8-epimer of prostaglandin F2alpha (8-isoPGF2α) and mean arterial pressure (MAP) in midpregnancy for the development of hypertensive complications in later pregnancy.
Design Prospective observational study as an ancillary study to the Hyperglycaemia and Adverse Pregnancy Outcome (HAPO) study.
Setting Obstetric clinics and wards of a university teaching hospital in Hong Kong.
Population Selected women with singleton pregnancies attending the antenatal clinic.
Methods Pregnant women who met HAPO inclusion criteria were recruited for the study. Glucose tolerance was assessed by a 75-g 2-hour oral glucose tolerance test (OGTT) at 24–32 weeks of gestation. Fasting plasma samples for 8-isoPGF2α estimation and urine samples for 8-isoPGF2α and 2,3-dinor 8-isoPGF2α assays were collected and blood pressures measured during the OGTT visit. Random plasma and urine samples were also obtained at 34–37 weeks. Glucose results were unblinded to the attending obstetrician if limits preset under the HAPO protocol were met.
Main outcome measures Maternal plasma 8-isoPGF2α and urinary 8-isoPGF2α and 2,3-dinor 8-isoPGF2α both at the time of OGTT (24–32 weeks) and at 34–37 weeks of gestation. Incidence of pre-eclampsia and gestational hypertension.
Results Of the 408 women who attended for OGTT at 24–32 weeks, two met the glucose criteria for unblinding and 25 had missing 8-isoPGF2α values and thus were excluded from analysis. Of the 381 women, 338 (88.7%) attended for random plasma samples at 34–37 weeks. Significant correlations were observed between maternal fasting plasma isoprostane and both fasting (r= 0.20; P < 0.001) and 2-hour (r= 0.39; P < 0.001) plasma glucose levels at the time of OGTT. Gestational hypertension/pre-eclampsia occurred in 17 (4.2%) women, and at the time of OGTT, they had significantly higher fasting plasma 8-isoPGF2α (P < 0.001), urine 8-isoPGF2α (P < 0.005) and urine 2,3-dinor 8-isoPGF2α to creatinine ratios (P < 0.001), as well as higher MAP (P < 0.001) than women who remained normotensive. At 34–37 weeks, only random plasma 8-isoPGF2α was significantly higher (P < 0.001) among the women with gestational hypertension/pre-eclampsia.
Conclusions Plasma markers of oxidative stress were positively correlated with plasma glucose at the time of OGTT (24–32 weeks). Women who subsequently developed gestational hypertension/pre-eclampsia had significantly higher plasma and urine markers of oxidative stress at the time of OGTT but only higher plasma markers at 34–37 weeks. Plasma 8-isoPGF2α appears to be a very good predictor of subsequent gestational hypertension/pre-eclampsia when measured at the time of OGTT, but its ability to discriminate deteriorates as pregnancy advances.