Abstract
Dear Editor:
We read with interest the article written by Polsky et al. 1 highlighting the point that pregnancy in women with type 1 diabetes mellitus (T1DM) represents a time when they are motivated to adopt new technology. We share our experience regarding the technology successfully applied in routine clinical practice.
We retrospectively analyzed the patterns of insulin use during pregnancy among 14 pregnant women with T1DM whose glucose was tightly controlled with continuous subcutaneous insulin infusion (CSII). Inclusion criteria in this study included a hemoglobin A1C ≤7% and body mass index <25 kg/m 2 at the time of conception, the diagnosis of diabetes mellitus at least 2 years before conception, and treatment using an insulin pump for at least 3 months before conception. Patients with severe hypoglycemia requiring the help of another person and/or who had ketoacidosis within 2 years before conception were not included. All patients were followed monthly by an experienced diabetologist and an obstetrician in face-to-face visits. The women were studied between 2011 and 2018 at the University Hospital Centre Zagreb; the diabetologist visits were conducted at the Department of Endocrinology and Diabetes and the Department of Obstetrics and Gynaecology (State Referral Centre for Diabetes in Pregnancy).
The records were selected from 387 pregnancies in women with T1DM, of whom 19 (5%) were treated with insulin pumps (Medtronic MiniMed™ Paradigm™ Veo™) from the time of conception to delivery. However, five patients treated with insulin pump did not meet the inclusion criteria and were not analyzed (they were either followed by another diabetologist and/or had poor glycemic control). A small percentage of patients treated with insulin pump corresponds roughly to the small percentage of all adult T1DM patients treated with insulin pump in Croatia.
Usage of the insulin pump was started in adulthood in six patients before they became pregnant. The indications for CSII in women are quite strict due to low accessibility of insulin pumps in Croatia. These include frequent and/or severe hypoglycemia, appropriate lifestyle habits, inability to reach glycemic goals despite good compliance, early onset of chronic complications, and preconception with labile diabetes. The other eight patients began using an insulin pump during their childhood. They were originally followed by a pediatric diabetologist but, in adulthood, transitioned to adult diabetologist. The patients did not adjust the basal pump rate themselves, although they used a temporary basal to change basal rate for a specific period of time and bolus calculator with preset insulin dosing algorithms. Two patients wore glucose sensors by employing the low glucose suspend option.
At the time of conception, the median age of the patients was 28.5 years (25–34 years), the median body mass index was 23.1 kg/m2 (19–25 kg/m2), and the median weight gain during pregnancy was 13 kg (4–16 kg). The average time from pump initiation to conception was 19 months (3–78 months). All patients successfully delivered by cesarean section (standard practice for women with T1DM in the facility) without significant peripartum complications. All newborns (eight boys, six girls) were closely monitored after delivery; only one newborn had transitory neonatal hypoglycemia and was admitted to the neonatal intensive care unit. None had jaundice.
The median week of delivery was 38 (37–39 weeks), the median APGAR score was 10 (7–10), and the median birth weight was 3190 g (2670–4440 g). When corrected for gender and gestational age, four (28%) newborns were large for gestational age, whereas the others were appropriate for gestational age. At the time of conception, the median A1C was 6.5% (4.8%–7.0%). In the last trimester, the A1C was significantly lower, with a median of 5.25% (4.6%–7.1%), (p = 0.020).
The median total daily insulin dose (TDD) significantly increased from the time of conception to week 32 of pregnancy (from 34.7 IU [13.4–50.6 IU] to 43.85 IU [17.5–87.0], p = 0.006). However, the TDD per kilogram of body weight (TDD/kg) did not significantly change (1.88 IU/kg [1.27–4.53 IU/kg]) at conception versus 1.72 IU/kg [1.08–4.0 IU/kg] at 32 weeks. The basal/prandial insulin ratio at conception was 52/48 and it changed significantly to 38/62 in the third trimester (p = 0.001).
Insulin is the standard of care for the treatment of T1DM. Constant insulin adjustment is necessary to keep up with changing insulin requirements throughout pregnancy. 2 Rapid, flexible, and precise dose of basal insulin and the use of bolus calculator settings for prandial insulin make insulin pump an optimal solution for pregnant women. In our patients, the TDD increased by 21% throughout pregnancy. Surprisingly, the TDD per kilogram remained the same.
Keeping in mind specific insulin demands in second and third trimesters because of postprandial hyperglycemia, we made consecutive changes in the insulin pump settings with a bolus calculator. The adjusted settings resulted in the reduction of the carbohydrate-to-insulin ratio followed by an increase in the prandial insulin dose but a decrease in the basal dose.
Our findings point to the importance of recalibration of insulin doses beyond mere weight-based dosing in pregnant patients with T1DM. This approach, which we found to yield a stable TDD per kilogram of exogenous insulin during pregnancy, enables optimal glycemic control without hyperinsulinization. In short, optimal glycemic control can be accomplished by modification of the basal/prandial ratio adjusting the settings in a bolus calculator. Tight glycemic control for pregnant women with T1DM requires both properly used state-of-the-art technology and personal experience.