Endocrine Abstracts

Aim: We aimed determinants of insulin usage in pregnant women with impaired carbohydrate metabolism (ICM).

Methods: We collected 164 pregnant women in any trimester who were referred to endocrinology unit due to suspicion of ICM by obstetricians. Diagnostic laboratory criteria were as follows: fasting glucose ≥126 mg/dl in first trimester; 1 hour 50 g challenge test glucose ≥180 mg/dl; any value exceeding threshold during OGTT (fasting glucose ≥92 mg/dl, 1 hour glucose ≥180 mg/dl, 2 hour glucose ≥153 mg/dl). Those with previous diagnosis of DM were also included. Insulin therapy was initiated in case of measurements exceeding target values (at least 2 fasting glucose ≥95 mg/dl and either 1 hour or 2 hour postprandial glucose ≥140 mg/dl ≥120 mg/dl, respectively) despite medical nutrition therapy.

Results: Mean and s.d. values were: age 31.4 ± 4.7 years, gestational age (GA) at initial visit to endocrinology unit 25.9 ± 5.2 weeks, GA at initiation of insulin 26.7 ± 7.8 weeks, number of pregnancies 2.3 ± 1.4, TSH 1.76 ± 1.21 mIU/l, and hemoglobin A1c 5.82 ± 1.06%. Insulin users (n = 90) had higher fasting glucose (P = 0.001), 2 hour OGTT-glucose level (P = 0.004), and hemoglobin A1c (P = 0.001) at diagnosis in comparison to non-insulin users (n = 74). Insulin users who gave birth before 36 weeks had higher hemoglobin A1c than those after 36 weeks (6.32 ± 1.68% vs 5.62 ± 1.68%, P = 0.014). GA at diagnosis was positively correlated with GA at insulin initiation (r = 0.634, P = 0.001). GA at diagnosis was significantly lower in subjects who gave birth before 36 weeks in comparison to those after 36 weeks (22.0 ± 7.8 weeks vs 26.2 ± 5.6, P = 0.01). There was a positive correlation in-between (r = 0.466, P = 0.002). GA at insulin initiation was positively correlated delivery date (before vs after 36 weeks) (r = 0.510, P = 0.006). Insulin dosage was negatively correlated with delivery date (final total dosage: r = −0.453, P = 0.016). GA at insulin initiation was negatively correlated with initial basal insulin dose (r = −0.283, P = 0.007). Insulin dosage increased significantly from baseline to final visit (basal: 2.53 ± 2.38 vs 3.73 ± 3.97 U, bolus: 2.27 ± 3.45 vs 5.20 ± 6.12 U, total 4.80 ± 3.88 vs 11.98 ± 12.09 U). A positive correlation was detected between initial and final total insulin dosage (r = 0.643; P = 0.001).

Conclusion: Pregnant women need basal insulin more frequently than bolus insulin. Insulin userswith higher hemoglobin A1c level and those who need higher insulin dosage are at higher risk for delivery before 36 weeks. Early insulin initiation avoids delivery before 36 weeks. Basal insulin demand is higher in women who were initiated insulin at an earlier GA.