Determinants of preeclampsia in women with type 1 diabetes.

AIMS

Despite improvement in diabetic care over the years, the incidence of hypertensive disorders of pregnancy is still very high. Therefore, the aim of our study was to determine risk factors for PE in women with T1DM.

METHODS

This study was a prospective, nested case-control study on a population of 165 women with T1DM. Women were divided into 3 subgroups: normotensive (N = 141), gestational hypertension (GH) (N = 8), and PE (N = 16). Clinical data were collected in the first trimester (< 12th week), in mid-pregnancy (20-24th weeks), and just prior to delivery (34-39th weeks). IR in the first trimester was quantified using the estimated glucose disposal rate formula (eGDR, milligrams/kilogram/minute). Simple logistic regression was used to search for factors associated with PE and GH. For multivariate comparisons, we used multiple logistic regression with stepwise selection.

RESULTS

All preeclampsia cases were diagnosed in primiparae. The presence of vasculopathy was the strongest determinant of PE (OR 10.8, 95% CI 3.27-35.97, P = 0.0001), followed by a history of chronic hypertension (6.05, 1.75-20.8, P = 0.004) and the duration of diabetes (1.11, 1.03-1.12, P = 0.009). However, chronic hypertension and duration of diabetes were no longer associated with PE after adjustment for the presence of vasculopathy. Higher gestational weight gain (GWG) was associated with PE, and this association remained significant after adjustment for first trimester body mass index (1.14, 1.02-1.28, P = 0.02). Both systolic and diastolic blood pressure assessed in the first trimester were significant determinants of PE; however, this association was no longer observed after adjustment for the presence of chronic hypertension. Glycated hemoglobin (HbA) levels from all 3 trimesters were significantly associated with PE (first trimester: 1.38, 1.01-1.87, P = 0.04; second trimester: 2.76, 1.43-5.31, P = 0.002; third trimester: 2.42, 1.30-4.51, P = 0.005). There was a negative association between eGDR and PE (0.66, 0.50-0.87, P = 0.003). Among lipids, triglycerides (TG) in all 3 trimesters were positively associated with PE, and this association was independent of HbA levels (first trimester: 5.32, 1.65-17.18, P = 0.005; second trimester: 2.52, 1.02-6.26, P = 0.05; third trimester: 2.28, 1.39-3.74, P = 0.001. We did not find any predictors of GH in the regression analysis among all analyzed factors.

CONCLUSIONS

Primiparity and diabetic vasculopathy seem to be the strongest risk factors for PE in women with type 1 diabetes. However, preexisting hypertension and higher GWG were also associated with PE in women with T1DM. Among laboratory results, higher HbA and TG levels in all 3 trimesters were associated with PE. The association between higher IR and PE in women with T1DM needs further study.