Evaluation of diabetes as a risk factor for pre-eclampsia

Methods The study was designed as a case-control study. It was based on the review of the medical charts of the subjects. The cases are women that are presented with diabetes during pregnancy. Diagnosis is based on the finding of fasting plasma glucose levels higher than 120 mg/dl or postprandial glucose levels higher than 180 mg/dl. We excluded subjects with chronic HTA, chronic renal disease and prior hepatic disease. The cases are patients that have been present in the pregnancy pathology clinic in our hospital from 2005 to 2012. The control group is made up of pregnant women with no diabetes. We excluded subjects with chronic HTA, chronic renal disease and prior hepatic disease. The controls have been present in the department of pregnancy pathology and obstetrics in the same period 2005-2012 (random selection). Statistical analysis: The following variables were taken into consideration: Maternal age, residence, education, parity, gestational age, blood glucose level, HbA1c measurement, personal history of diabetes, maternal birth weight, presence of polyhydramnios, previous history of stillbirth and incidence of preeclampsia. Data were analyzed using univariate analysis. The variables were divided and organized according to the objectives. The descriptive analysis of continuous variables consists of the number of observations, mean value, standard deviation, median value and minimal and maximal values. To explore the distribution of continuous variables we used standard graphic presentations like the histogram. To explore the distribution of the observations according to the categories, individual and cumulative frequencies were determined. Bivariate analysis was used to describe and test the association of different variables. Association of categorical variables was analyzed using the χ2 and Fisher’s exact test. The difference between the mean values of the continuous variables according to the categories was analyzed using the Student’s test. Confidence intervals were calculated based on the binominal distribution. To quantify the effect of a variable on another variable we used the odds homogeneity test. Score’s test for odds trend was used to test the association of two variables. To take into consideration the effect of confounding factors (parity, maternal age and maternal birth weight) on the association of diabetes with preeclampsia we used logistic multivariate analysis. Statistical analysis was performed using Intercooled STATA 9. 1.