Methods
We conducted a retrospective cohort study at a single outpatient health clinic between 2011 and 2018. The study group included healthy women at 24-28 weeks of gestation with one abnormally high value on their OGTT and a low glucose finding for the 3-hour hypoglycemia analysis (<60 mg/dl). The control group was comprised of women with high or normal 180-minute glucose values with one abnormal value in the OGTT. Excluded from the study were women with a multiple pregnancy, pre-gestational diabetes, or incomplete medical or obstetrical records. The Institutional Review Board approved the study and waived informed consent.
GDM was diagnosed by a two-step approach consisting of a 1-hour 50 g GCT at 24–28 weeks of pregnancy, followed by a 3-hour 100 g diagnostic OGTT if the GCT plasma glucose result was ≥140 mg/dL. Women with risk factors for developing GDM, such as a family history of diabetes in a first-degree relative, previous GDM, or previous macrosomia, underwent only diagnostic OGTT. GDM was diagnosed when at least one of the four OGTT glucose values was abnormally high by the Carpenter and Cousten criteria14.
Once diagnosed, all women were instructed to self-measure capillary glucose at least four times daily (fasting and one or two hours postprandial). The initial clinic visit included counseling by a nurse expert in diabetes, dietary consultation by a certified dietician, and consultation with a maternal-fetal medicine specialist. The biweekly visits were scheduled up to 36 weeks of gestation and weekly from that stage to delivery. Each visit included dietary consultation, blood pressure monitoring, urine dipstick test, review of daily glucose profiles, and decision-making regarding the need for adjusting therapy. Ultrasound scans for biophysical profiles and estimation of fetal growth were performed biweekly from 26 weeks of gestation, and non-stress testing was added at 34 weeks of gestation.
The desired level of glycemic control was defined as mean fasting blood glucose <90 mg/dl and mean one or two hours postprandial <140 mg/dL or <120 mg/dl, respectively. Women with fewer than 50 glucose measurements over the two weeks were not included in the final analysis. Women who had more than30% abnormally high values during a 2-week period of glucose monitoring were considered as having suboptimal glucose control.15 The detection of suboptimal glucose control led to an intervention aimed at modifying the diet and encouraging physical activity. Insulin treatment was initiated when fasting glucose levels were consistently >100 mg after two weeks of observing a strict diet or when suboptimal glucose control, as defined above, persisted after 2 weeks of dietary modifications and physical activity.
The need for induction of labor was determined by the level of glycemic control and estimated fetal weight. Induction of labor was offered at 38–39 weeks of gestation if the estimated fetal weight (EFW) was 3600-4000 gr or poor glycemic control was documented [defined as?]. Cesarean section was advised when the EFW was >4000 gr or for obstetrical indications. If the abovementioned criteria were not fulfilled induction of labor was advocated at 41 weeks of gestation.
Women with one abnormally high OGTT value were divided into two groups. Women in the study group had a low 180-minute value defined as a plasma glucose level of ≤60 mg/dl, and those in the control group had a 180-minute glucose value of >60 mg/dl. The following data were extracted from women’s computerized files and compared between the two groups: maternal demographics and clinical characteristics, periods of suboptimal glucose control, the need for insulin to achieve desired levels of glycemic control, glycemic profile, and obstetrical and neonatal outcome measures. The primary outcome was the rate of women with suboptimal maternal glycemic control. The secondary outcomes were the need for insulin treatment to achieve desired levels of glycemic control and the presence of obstetrical and neonatal complications, including mean birthweights, incidence of macrosomia (birthweight more than the 90th percentile or >4000 gr), and polyhydramnios (amniotic fluid index >25 cm or a maximal vertical pocket >8 cm).
Statistical analysis
Between-group comparisons of the study participants’ characteristics were performed by a 2-tailed Student’s t test for continuous variables, a Pearson χ2 test for categorical variables, and the Kruskal–Wallis test for continuous non-normal variables. A P value of less than or equal to .05 was considered significant. Multivariate logistic regression assessed the association between the various selected variables and suboptimal maternal glycemic control. Maternal age, parity, EFW, and delivery week were taken as continuous variables and the remaining variables were taken as categorical. All statistical analyses were performed with R version 3.5.2. (http://www.r-project.org)