Risk Factors That can Predict Antenatal Insulin Need in Gestational Diabetes.

BACKGROUND

This study was undertaken to assess the association between insulin need in gestational diabetes mellitus (GDM) and clinical features and laboratory parameters. Factors that can predict insulin need are also identified.

METHODS

Cases with GDM were included retrospectively from records. Cases which failed to achieve target blood glucose levels with medical nutrition therapy (MNT) and need insulin treatment were recorded. Risk factors which can predict antenatal insulin treatment (AIT) were identified as follows; the presence of diabetes in a first degree relative, body mass index prior to pregnancy, number of parity, history of GDM, macrosomic baby delivery (> 4,000 g), age, gestational week at time of diagnosis, body mass index during diagnosis, weight gain untill diagnosis, mean systolic and diastolic blood pressure, HbA1C level during diagnosis, and fasting plasma glucose on diagnostic oral glucose tolerance test. Presence of a statistical significance between those patient features and AIT was assessed. Independent predictors for AIT were evaluated.

RESULTS

A total of 300 cases were recruited from records, 190 cases (63.3%) were followed only with MNT until delivery and 110 cases (36.7%) were initiated AIT. The association between AIT and patient factors like presence of diabetes in the pedigree, week of gestation at which GDM was diagnosed, BMI during diagnosis, HbA1C levels, and fasting plasma glucose during diagnosis was found (P = 0.03; 0.008; 0.049; 0.001 and 0.001respectively). Multivariant analysis showed that fasting plasma glucose levels during diagnosis and HbA1C levels were independent risk factors for AIT. Fasting plasma glucose values that can predict AIT were identified > 89.5 mg/dL with 72.7% sensitivity and 62.6% spesifity (P < 0.001). Positive predictive value was 73% (P < 0.001). Also, HbA1C levels that can predict AIT was found to be > 5.485% with 65.3% sensitivity and 66.7% spesifitiy(P < 0.001) with a positive predictive value 68% (P < 0.001).

CONCLUSIONS

Independent predictors for AIT were found as fasting plasma glucose on OGTT and HbA1c levels during diagnosis in GDM. Cases with fasting plasma glucose ≥ 89.5 mg/dL or HbA1C ≥ 5.485% should be closely followed for AIT in specified centers.