[Association between amniotic fluid insulin in the second trimester, maternal glucose tolerance and fetal malformations].

INTRODUCTION: Maternal hyperglycemia during gestational diabetes leads to fetal hyperinsulinemia, which is associated with increased perinatal morbidity and mortality. Amniotic fluid insulin levels are therefore considered by some researchers to be ideal parameters to use in diagnosing gestational diabetes and making decisions about correct therapy. There are various recommendations about determining gestational diabetes early in pregnancy (< 24 weeks) by measuring amniotic fluid insulin. This study tests this association -- taking additional risk factors into account -- in a group of pregnant women who had genetic indications requesting for amniocentesis (AC).

MATERIALS AND METHODS: All pregnant women who came to our clinic for genetically-determined amniocentesis from April 10, 1995 - Jan. 31, 2000 and who were between 12 and 24 weeks were included in our study. After a sample of amniotic fluid was taken, the laboratory performed a competitive radio-immuno-assay to determine the insulin concentration. O'Sullivan's cut-off values were used in diagnosing gestational diabetes. Since not all pregnant woman in our clinic were screened for gestational diabetes, we gathered our data retrospectively by checking all birth records; these were available in our clinic's data archive.

RESULTS: A total of 483 pregnant women were included in our study. 22 (4.6 %) of them were classified as gestational diabetics. The average value for amniotic fluid insulin was 1.21 mU/L +/- 0.89. The insulin values for the entire study population exhibited a weekly increase of 0.1 mU/L from the 12th through the 24th week. The insulin concentrations for the 22 gestational diabetics were not significantly higher than those of the non-diabetics (1.05 mU/L vs. 1.0 mU/L; p = 0.34). In the 90 (th) percentile and above of the amniotic fluid insulin levels (2.2 mU/L) for the entire study population, the rate of gestational diabetics was at 11.8 % three times that of the non-diabetics, at 3.7 % (p = 0.021). Among the risk factors for gestational diabetes, an increased body mass index (BMI) value correlated significantly with increased insulin concentration (p < 0.001). The patients at and above the 90th percentile also had significantly higher BMI values (p = 0.002). In the multivariate analysis, the following influences were determined to be significant: maternal body mass index (p < 0.001) and the gestational age (p < 0,001), not the mere diagnosis of "gestational diabetes". A significant association was not found between elevated insulin values in amniotic fluid and the child's birth weight, APGAR values, pH-levels and blood glucose values. However, a significant association was found regarding fetal malformations and chromosome abnormalities.

CONCLUSION: Even very low concentrations of insulin can be identified in amniotic fluid early in the pregnancy. The values increase during the course of the pregnancy. There is a positive correlation between maternal weight (BMI) and insulin levels in the amniotic fluid. Pregnant women with gestational diabetes have higher insulin levels in their amniotic fluid. The multivariant analysis shows, however, that this association can be traced to the maternal BMI and the time point during the pregnancy when the AC was performed. Malformations, especially those with a neural tube defect, are an additional cause for elevated insulin values in amniotic fluid.