Continuous Subcutaneous Insulin Infusion During Pregnancy in Women with Complicated Type 1 Diabetes Is Associated with Better Glycemic Control but Not with Improvement in Pregnancy Outcomes

Abstract

Background:

The aim of this study was to evaluate maternal and fetal pregnancy outcomes of women with type 1 diabetes managed on continuous subcutaneous insulin infusion (CSII) compared with multiple daily insulin injections (MDI).

Subjects and Methods:

Pregnancy outcomes were assessed retrospectively in women with type 1 diabetes who were patients of the Diabetes Clinic of North Karelia Hospital (Joensuu, Finland) between 2000 and 2012. The medical records of 72 women experiencing 135 pregnancies and data of their infants were retrospectively reviewed.

Results:

In total, 48 pregnancies were treated with CSII and 87 with MDI. Women on CSII treatment were older and had more diabetes complications compared with women on MDI. No significant differences in glycated hemoglobin (HbA1c) levels were observed between the CSII and MDI groups before or during pregnancy. Maternal or fetal outcomes did not differ between the treatment groups. However, among women with complicated diabetes, HbA1c levels were significantly lower in the CSII group until the second trimester (prepregnancy, 7.22% vs. 8.14%, respectively [P = 0.034]; first trimester, 6.85% vs. 7.87% [P < 0.001]; second trimester, 6.41% vs. 7.03% [P = 0.029]) without an increased rate of maternal hypoglycemia.

Conclusions:

Pregnancy outcomes were similar regardless of insulin treatment modality. Although using an insulin pump did not result in improvement of pregnancy outcomes, it allowed for better glycemic control in pregnancies of women with complicated diabetes. Therefore, it is worth considering in high-risk T1DM pregnancies, especially if good glycemic control is not achieved otherwise.

Introduction

Pregnancies of women with type 1 diabetes mellitus (T1DM) are associated with an increased risk of maternal and fetal adverse outcomes. An important cause is hyperglycemia, especially at the time of conception and in the first trimester of pregnancy.¹ It has been demonstrated that even a slightly raised glycated hemoglobin (HbA1c) level during early pregnancy in women with T1DM is associated with an increased risk for fetal malformations.² In addition, women with T1DM have increased risk of miscarriage, stillbirth, macrosomia, and neonatal hypoglycemia. Common maternal complications of pregnancies of women with T1DM are preeclampsia, preterm labor, and cesarean section.^2

–7

Glycemic control in pregnancies of women with diabetes is still inadequate despite the development of insulin analogs and insulin delivery methods.⁴ Pregnancy itself makes it difficult to maintain stable glycemic control, necessitating frequent adjustment of insulin doses. During the first trimester, the risk of hypoglycemia increases because of improved insulin sensitivity. In addition, the effort to achieve better glycemic control can predispose one to hypoglycemia. As the pregnancy progresses, insulin resistance and hyperglycemia increase, leading to fetal hyperinsulinemia and macrosomia.⁸ Good glycemic control throughout the pregnancy significantly reduces the risk of adverse outcomes of pregnancies in women with diabetes.^2,9,10

In nonpregnant adults with T1DM, continuous subcutaneous insulin infusion (CSII) has been demonstrated to decrease HbA1c levels more than multiple daily insulin injections (MDI),¹¹ but the benefit of CSII during pregnancy is controversial. In pregnant women with T1DM, several small randomized^12
–14 and observational studies^15

–19 have found no significant difference in glycemic control between the two delivery methods. However, some other observational studies have suggested that CSII can improve HbA1c better or earlier than MDI during pregnancy in T1DM women.^20

–23 In terms of pregnancy outcomes, several previous studies have shown that CSII and MDI are comparable.^12

–18 In contrast, the retrospective comparison of Talaviya et al.²⁴ found that CSII treatment resulted in improved pregnancy outcomes, as well as better glycemic control, in T1DM women.

The evidence is inconclusive whether one insulin delivery method (CSII or MDI) is safer and superior to another in improving glycemic control and pregnancy outcomes. In addition, previous studies have included mainly women with noncomplicated diabetes. Our study aimed to evaluate the efficacy and safety of CSII or MDI in the pregnant T1DM population, including a high proportion of women with complicated diabetes, in routine daily practice.

Subjects and Methods

Subjects

This retrospective study was conducted at the Hospital District of North Karelia in Finland. In the North Karelia district, the care of all pregnant T1DM patients is centralized to our secondary care hospital. We evaluated the data of women who had T1DM prior to pregnancy and were treated in the Diabetes Clinic of North Karelia Central Hospital (Joensuu) during their pregnancy over the period of 2000–2012. We included all pregnancies for each of the women (miscarriages, terminations of pregnancy for medical reasons, stillbirths, and live births), except for terminations without medical reasons. T1DM diagnosis and diabetes complications were assessed according to the clinical diagnosis documented in the medical records (International Statistical Classification of Diseases and Related Health Problems codes E10.1–E10.9). We excluded 10 pregnancies because the MDI treatment was switched to CSII treatment during the pregnancy. Finally, altogether, 135 pregnancies of 72 women and the data of their infants born between 2000 and 2012 were analyzed. The health status of the infants was followed for at least 12 months after the birth.

The ethics committee of the North Savo Hospital District (which includes the North Karelia Hospital District) approved this research on February 12, 2013.

The choice of insulin therapy was based on whether a woman achieved optimal glycemic control (HbA1c level of <7.0–7.5% [53–58 mmol/mol], preferably <7.0% before or at the beginning of the pregnancy, which is the recommendation of the Finnish Diabetes Association). If the women on MDI did not achieve the target, switching to CSII was suggested. Patients already receiving CSII prior to pregnancy continued the treatment. Both groups received similar clinical support before and during their pregnancy. Pregnancy was considered as planned if a woman with diabetes received prepregnancy care before conception to optimize health issues. A diabetologist and a diabetes nurse followed the diabetes patients during pregnancy in the diabetes clinic, and in the obstetric clinic, a midwife (nurse) and an obstetrician followed the patients. Generally, the patients visited the hospital outpatient clinics every 4 weeks.

Data collection

Maternal medical records were reviewed to obtain demographic data, information concerning diabetes complications, medication, glycemic control and delivery data. Body mass index was calculated with the documented weight during pregnancy visits, but height was self-reported. Severe hypoglycemia of the mother was defined as hypoglycemia with unconsciousness or requiring assistance from another person to administer oral carbohydrate or intravenous glucose infusion. Mild hypoglycemia included self-monitored plasma glucose measurements below 3.0 mmol/L. HbA1c level was determined using the turbidimetric inhibition immunoanalysis method, with the normal values ranging from 4.0% to 6.0%. The method was the same during the entire study period of 2000–2012.

Outcome data were abstracted from the medical records. Preeclampsia was defined using the diagnosis made by an obstetrician—generally defined as blood pressure ≥140/90 mm Hg with proteinuria (≥0.5 g). Preterm delivery included delivery before 37 completed weeks. Data on newborns included gestational age at delivery, birth weight, Apgar score (0, 5, and 10 min after delivery). A newborn was considered large for gestational age if birth weight was above the 90^th percentile according to growth charts for the Finnish population. We reviewed data concerning permanent congenital anomalies. We also reviewed neonatal complications, such as development of respiratory distress syndrome, hyperbilirubinemia, perinatal acidosis (asphyxia), and hypoglycemia. Neonatal hypoglycemia was defined as a glucose level <1.5 mmol/L or requiring intravenous glucose infusion.

Statistical analysis

Statistical analyses were carried out using the IBM (Armonk, NY) SPSS statistics programs (Statistical Package for the Social Sciences). Some mothers had several pregnancies. Thus we processed them as a cluster of the same mother containing unequal numbers of pregnancies and applied the complex samples analysis method with probability-proportional-to-size sampling. A complex sample χ² test was used to analyze the statistical significance of the differences of frequency estimates of weighted categorized variables. A complex sample general linear model estimated means test was used in the assessment of the differences of weighted continuous variables between the groups. A complex sample logistic regression analysis was applied to analyze the association of explanatory variables for different outcomes. Data in tables are presented as mean ± SD values.

Results

In total, 135 pregnancies in 72 women with T1DM were included. In all, 48 pregnancies were treated with CSII and 87 with MDI. Six women were treated with MDI or CSII during different pregnancies. There were 109 (80.7%) live births and 26 (19.3%) miscarriages. Of these, 25 were first trimester miscarriages, and one abortion was performed because of fetal malformation in gestational Week 22. There were no stillbirths. Five infants (4.6%) had congenital anomalies. These included one caudal regression leading to an abortion, one moderate sacral dysgenesia, and three cardiac malformations (abnormality of the vena cava, aortic dilatation, and atrial septal defect with pulmonary and tricuspidal valve regurgitations).

Because of the long study period, several mothers had more than one pregnancy. Altogether, two women had six pregnancies, two women had five pregnancies, two women had four pregnancies, 11 women had three pregnancies, 17 women had two pregnancies, and 38 women experienced one pregnancy over the period of 2000–2012. Parity did not differ between the CSII and MDI treatment groups.

Baseline characteristics are presented in Table 1. Women on the CSII treatment had more diabetes complications (79.2% vs. 39.1%; P = 0.019), including hypertension. There were significantly more smokers in the MDI-treated group compared with the CSII-treated group (14.9% vs. 0%; P = 0.001). Altogether, 88 (65.2%) pregnancies were planned: 81.2% in the CSII group and 56.3% in the MDI group (P = 0.052). The CSII-treated group used more rapid insulin analogs than the MDI-treated group (91.7% vs. 75.9%; P = 0.033).

Table 1.

Characteristics of Type 1 Diabetes Pregnancies by Treatment Group

	All			Complicated diabetes ^a
	CSII (n = 48) (35.5%)	MDI (n = 87) (64.5%)	P value ^b	CSII (n = 38) (52.8%)	MDI (n = 34) (47.2%)	P value ^b
Age (years)	31.3 ± 4.2	26.4 ± 5.0	<0.001	31.7 ± 4.5	27.9 ± 5.3	0.003
BMI (kg/m²) at beginning	26.9 ± 3.2	25.7 ± 4.1	NS	26.8 ± 3.4	26.7 ± 4.4	NS
Weight (kg)
At beginning	73.2 ± 11.5	70.7 ± 11.5	NS	72.4 ± 11.9	71.9 ± 11.8	NS
At delivery	86.5 ± 12.9	82.2 ± 12.8	NS	86.5 ± 13.7	83.4 ± 14.6	NS
Change	12.9 ± 5.6	11.9 ± 4.5	NS	13.1 ± 6.1	12.0 ± 4.8	NS
Diabetes complications	38 (79.2)	34 (39.1)	0.019
Retinopathy	37 (77.1)	23 (26.4)	0.003	37 (97.4)	23 (67.6)	0.006
Microalbuminuria	9 (18.8)	10 (11.5)	NS	9 (23.7)	10 (29.4)	NS
Retinopathy and microalbuminuria	9 (18.8)	7 (8.0)	NS	9 (23.7)	7 (20.6)	NS
Hypertension	14 (29.2)	11 (12.6)	0.013	14 (36.8)	11 (32.8)	NS
Duration of diabetes (years)	20.2 ± 7.5	14.0 ± 7.4	0.062	21.8 ± 7.0	18.9 ± 7.5	NS
BP (mm Hg)
Systolic	124.8 ± 12.6	126.0 ± 14.1	NS	124.5 ± 13.9	131.3 ± 13.9	NS
Diastolic	73.3 ± 7.4	75.9 ± 11.2	NS	72.6 ± 7.7	79.2 ± 12.9	0.021
Smoking	0 (0)	13 (14.9)	0.001	0 (0)	4 (11.8)	0.001
Planned pregnancy	39 (81.2)	49 (56.3)	0.052	31 (81.6)	18 (52.9)	NS
Nulliparity	10 (20.8)	29 (33.3)	NS	7 (18.0)	8 (13.0)	NS
HbA1c (%)
Prepregnancy	7.20 ± 0.86	7.88 ± 1.60	NS	7.22 ± 0.82	8.14 ± 1.36	0.034
First trimester	6.88 ± 0.68	7.32 ± 1.29	NS	6.85 ± 0.59	7.87 ± 1.13	<0.001
Second trimester	6.47 ± 0.52	6.64 ± 1.10	NS	6.41 ± 0.53	7.03 ± 0.53	0.029
Third trimester	6.55 ± 0.46	6.67 ± 0.93	NS	6.52 ± 0.48	6.85 ± 0.92	NS
Duration of pregnancy (weeks)	36.8 ± 1.6	36.9 ± 2.2	NS	36.6 ± 1.7	36.2 ± 3.1	NS
Total daily insulin dose
At beginning (U)	39.0 ± 9.3	51.5 ± 20.3	0.067	38.4 ± 7.9	54.5 ± 17.4	0.002
At delivery (U)	82.0 ± 29.1	88.0 ± 36.8	NS	80.2 ± 26.4	87.7 ± 27.0	NS
Increase (%)	112 ± 62	73 ± 54	0.062	117 ± 56	62 ± 36	0.059
Rapid insulin analog	44 (91.7)	66 (75.9)	0.033	35 (92.1)	29 (82.9)	NS

Data are mean ± SD values or n (%) as indicated.

Complications include retinopathy, microalbuminuria, and hypertension.

Because there were several pregnancies of the same mothers, P values were calculated by the complex sample general lineal model for weighted continuous variables and by the complex sample χ² test for weighted categorized variables.

BMI, body mass index; BP, blood pressure; CSII, continuous subcutaneous insulin infusion; HbA1c, glycated hemoglobin; MDI, multiple daily insulin injections; NS, not significant (P > 0.100).

In the subgroup of women with complicated diabetes, CSII-treated women had better HbA1c levels before pregnancy and in the first and the second trimesters compared with MDI-treated women (P = 0.034, <0.001, and 0.029, respectively). CSII-treated women had a lower insulin dose at the beginning and a greater increase of insulin dose during pregnancy (%) than did the MDI-treated group (P = 0.002 and P = 0.059, respectively).

The overall mean HbA1c level before pregnancy was 7.63 ± 1.41%, and among women with complicated diabetes it was 7.63 ± 1.18%. CSII-treated women achieved a target of HbA1c level of <7.0% during the first trimester, whereas MDI-treated women did not achieve this until the second trimester. There was only one incidence of ketoacidosis that occurred in MDI-treated women. In total, 32 pregnancies (29.3%) were complicated by mild but weekly hypoglycemia, and 12 pregnancies (11.0%) had severe hypoglycemia episodes. The prevalence of mild and severe hypoglycemia was similar between the CSII and MDI groups.

Maternal or fetal outcomes did not differ between the CSII and MDI groups overall or among the subgroup of women with complicated diabetes (Table 2). The associations of maternal and fetal pregnancy outcomes with treatment groups were further analyzed using logistic regression analyses (Table 3). There were no significant differences according to treatment modality.

Table 2.

Maternal and Fetal Outcomes by Treatment Groups

	All			Complicated diabetes ^a
	CSII	MDI	P value ^b	CSII	MDI	P value ^b
Maternal
Number	48	87		38	34
Miscarriage	11 (22.9)	15 (17.2)	NS	8 (21.1)	8 (23.5)	NS
Cesarean section	26 (70.3)	41 (56.9)	NS	20 (66.7)	19 (73.1)	NS
Preeclampsia	7 (18.9)	18 (25.0)	NS	7 (23.3)	12 (46.2)	NS
Preterm delivery	11 (29.7)	21 (29.2)	NS	11 (36.7)	12 (46.2)	NS
Fetal
Number	37	72		30	26
Congenital anomalies	1 (2.7)	4 (5.6)	NS	1 (3.3)	2 (7.7)	NS
LGA	19 (51.4)	28 (38.9)	NS	15 (50.0)	9 (34.6)	NS
Hypoglycemia	32 (86.5)	54 (75.0)	NS	25 (83.3)	21 (80.7)	NS
Asphyxia	3 (8.1)	5 (6.9)	NS	2 (6.7)	3 (11.5)	NS
Respiratory distress	7 (18.9)	19 (26.4)	NS	7 (23.3)	8 (30.8)	NS
Hyperbilirubinemia	11 (29.7)	21 (29.2)	NS	11 (36.7)	8 (30.8)	NS
Apgar score
0 min	8.22 ± 1.66	8.10 ± 1.78	NS	8.03 ± 1.73	8.23 ± 1.48	NS
5 min	8.68 ± 0.78	8.53 ± 1.05	NS	8.60 ± 0.86	8.54 ± 1.10	NS
10 min	8.89 ± 0.40	8.69 ± 0.76	NS	8.86 ± 0.45	8.74 ± 0.75	NS
Birth weight (kg)	3.84 ± 0.67	3.75 ± 0.74	NS	3.80 ± 0.77	3.58 ± 0.64	NS
Hospital stay (days)	6.9 ± 7.3	6.9 ± 7.4	NS	7.9 ± 7.8	8.3 ± 7.8	NS

Data for maternal factors represent data for all pregnancies. Data for fetal factors represent data for all live births. Data are mean ± SD values or n (%) as indicated.

Complications include retinopathy, microalbuminuria, and hypertension.

Because there were several pregnancies of the same mothers, P values were calculated by complex sample general lineal model for weighted continuous variables and by complex sample χ² test for weighted categorized variables.

CSII, continuous subcutaneous insulin infusion; LGA, large for gestational age; MDI, multiple daily insulin injections; NS, not significant (P > 0.100).

Table 3.

Association of Maternal and Fetal Outcomes of Pregnancies with Continuous Subcutaneous Insulin Infusion Treatment by Logistic Regression Analyses in the Subgroup of Complicated Diabetes (Complex Sample Method)

	Model 1 (adjusted for age)	P value	Model 2 (adjusted for age and HbA1c; 2nd trimester)	P value
Maternal
Miscarriage	1.28 (0.17–6.07)	NS	^a
Cesarean section	0.63 (0.06–6.34	NS	0.63 (0.06–6.34)	NS
Preeclampsia	0.20 (0.04–1.03)	0.055	0.19 (0.03–1.18)	0.074
Preterm delivery	1.84 (0.48–7.12)	NS	1.83 (0.48–7.04)	NS
Fetal
Congenital anomalies	0.31 (0.03–3.77)	NS	1.11 (0.02–54.88)	NS
LGA	0.86 (0.15–5.06)	NS	1.06 (0.14–7.92)	NS
Hypoglycemia	0.76 (0.14–4.28)	NS	0.88 (0.14–5.41)	NS
Respiratory distress	0.83 (0.21–3.19)	NS	1.07 (0.29–3.88)	NS
Asphyxia	0.33 (0.05–2.19)	NS	0.42 (0.07–2.47)	NS
Hyperbilirubinemia	1.13 (0.17–7.27)	NS	0.89 (0.13–6.29)	NS
Hospital stay ≥4 days	0.68 (0.18– 2.62)	NS	0.48 (0.08–2.92)	NS

Complications of diabetes include retinopathy, microalbuminuria, and hypertension. Data are odds ratio (95% confidence interval).

Not applicable because miscarriages happened before the second trimester.

HbA1c, glycated hemoglobin; LGA, large for gestational age; NS, not significant (P > 0.100).

In the subgroup of women with complicated diabetes, the use of CSII tended to decrease the risk of preeclampsia, but the difference was not significant (age-adjusted odds ratio = 0.20; P = 0.055) (Table 3). After further adjustment for hypertension, HbA1c level during the second trimester, and change of insulin dose, the association of CSII use with the risk of preeclampsia was not significant (P = 0.102).

Discussion

We retrospectively assessed pregnancy outcomes of women with T1DM treated with CSII or MDI. We did not find significant differences in HbA1c levels between the CSII and MDI groups before or during the pregnancy. Pregnancy outcomes were similar between the CSII- and MDI-treated groups. However, among women with complicated diabetes, the CSII-treated women had significantly better glycemic control than those who were treated with MDI. Previous observational and randomized studies^{12

–18,20,24
–26} were inconclusive regarding whether or not using CSII or MDI treatment during pregnancy of T1DM patients provided better glycemic control or pregnancy outcomes. The randomized studies were small and were conducted between 1985 and 1993, when insulin analogs were not available.^12
–14,26 Thus, these results are not directly applicable to today's practice. Large randomized controlled trials on this subject are lacking.

Glycemic control

We found that in the pregnancies of women with complicated diabetes, CSII treatment resulted in better glycemic control before pregnancy and in the first and the second trimesters than MDI treatment. It is important that CSII treatment did not increase rates of maternal ketoacidosis, severe hypoglycemia, or neonatal hypoglycemia. In contrast, two previous studies reported that CSII may be associated with higher rates of maternal diabetic ketoacidosis^14,26 or more severe hypoglycemia.¹⁴ Recent retrospective studies have demonstrated that CSII-treated women had lower HbA1c levels than MDI-treated women,^23,24 or at least women already using CSII before pregnancy.²¹ In other retrospective studies, women using CSII achieved good glycemic control earlier during pregnancy and required less insulin than did MDI-treated women.^20,22 Taken together, CSII treatment can be considered as effective and safe as MDI in achieving good glycemic control during pregnancies in women with diabetes.

Pregnancies of women with T1DM are associated with an increased risk of adverse pregnancy outcomes, mainly due to hyperglycemia. This relationship of glucose level to pregnancy outcomes is a continuum. Therefore, it is essential to obtain glycemic levels as close to normal as possible at conception and during pregnancy.¹ The American Diabetes Association and American Congress of Obstetricians and Gynecologists recommend maintaining an HbA1c level of <7.0% (53 mmol/mol) before the pregnancy and <6.0% (42.0 mmol/mol) during the pregnancy.^27,28 The recommendation of the Finnish Diabetes Association at conception and during pregnancy is an HbA1c level of <7.0–7.5% (<53–58 mmol/mol), preferably <7.0%, which is less strict but still emphasizes avoiding maternal hypoglycemia.²⁹

In the present study, the mean HbA1c level before pregnancy was 7.63%. CSII-treated women achieved a target HbA1c level of <7.0% during the first trimester, whereas MDI-treated women did not achieve this until the second trimester. Some studies have reported better HbA1c levels of pregnant T1DM patients compared with our results. In two Spanish studies, prepregnancy HbA1c levels were 6.2–6.8%^30,31 and in an Israeli study, the levels were 6.9–7.1%²⁶; however, a different HbA1c analysis method was used. In contrast, an Indian study found that the mean HbA1c level prior to pregnancy was above 8.0%,²⁴ using an unknown HbA1c analysis method. The HbA1c analysis methods may partly explain these differences, but the HbA1c levels may also reflect different attitudes of patients to preparing for pregnancy or unequal standards of pregnancy care in various countries.

Pregnancy outcomes

The results of our study are in agreement with the previous findings, that despite better metabolic control, CSII use did not improve pregnancy outcomes of women with T1DM compared with MDI use.^20,22,23 In a recent retrospective Canadian study of 387 pregnancies of T1DM patients, CSII users had lower HbA1c levels but no improvement in pregnancy outcomes compared with MDI users.²³ This study included the largest diabetes population with a high number of CSII users (113 pregnancies) from seven different centers. Their diabetes population had fewer complications of diabetes (hypertension, 3.6%; retinopathy, 11.4%) compared with our population (hypertension, 18.5%; retinopathy, 44.4%). Thus, our population represents a more unfavorable distribution of T1DM in pregnant women.

The evidence regarding whether CSII treatment has any advantage or disadvantage over MDI treatment in terms of pregnancy outcomes is still sporadic and inconclusive. One small Indian study indicated that in pregnancies of T1DM women, CSII treatment resulted in better glycemic control and improved fetal outcomes compared with MDI treatment.²⁴ In that study, the rates of abortion, preterm labor, cesarean section, and neonatal hypoglycemia were less, and the Apgar score was higher, in CSII-treated women compared with MDI-treated women. In contrast, two studies have demonstrated that CSII use could even have an unfavorable effect on pregnancy outcomes. An Irish study demonstrated that despite better glycemic control, CSII-treated women had more cesarean sections compared with MDI-treated women.²¹ Furthermore, a large Spanish retrospective study found that CSII use resulted in a higher risk of large for gestational age and perinatal mortality, and CSII with lispro insulin had a higher risk of miscarriage.³¹

Among women with complicated diabetes, the use of CSII tended to decrease the risk of preeclampsia (age-adjusted odds ratio = 0.20; P = 0.055). This association disappeared, when hypertension, which is associated with a risk of developing preeclampsia, was taken into account. In the subgroup of women with complicated diabetes, CSII users had lower diastolic blood pressure in early pregnancy. Diabetes itself increases the risk of a hypertensive pregnancy two- to fourfold compared with women without diabetes. This risk is related to the presence of diabetic nephropathy and poor glycemic control.³² In the present study, the rate of microalbuminuria did not differ between treatment groups, but CSII users had better glycemic control.

In our study, the overall rate of neonatal hypoglycemia (defined as a glucose level of <1.5 mmol/L or requiring intravenous glucose infusion) was surprisingly high, at 78.9%. These numbers are clearly higher compared with other reported frequencies of 15–40%,^20,22,23,26 although a comparison is difficult because of differences in definitions. Neonatal hypoglycemia may be explained by maternal glucose levels and insulin dosage used during delivery.³³ The review of various protocols for glucose control during the delivery of women with diabetes demonstrated that a target glucose level of 4.0–7.0 mmol/L was safe and resulted in a low rate of neonatal hypoglycemia.³³ In our hospital, the treatment protocol of glycemic control with CSII during delivery is the same for all T1DM women, and this protocol was unchanged during the study period. We used the Finnish recommendations for the glucose target during delivery, which is 4.0–7.0 mmol/L.²⁹ Because the frequency of neonatal hypoglycemia was equally high in the CSII and MDI groups, the insulin protocol used at delivery, rather than the treatment modality during the pregnancy, might be the factor predisposing one to neonatal hypoglycemia.

Strengths and limitations

The present study included a higher proportion of women with complicated T1DM than previous studies. We also compared the use of two insulin delivery methods in a subpopulation of women with complicated diabetes. Because the data were from one secondary care center, we managed to get very comprehensive data of baseline measures and outcomes, including data regarding miscarriages. The main limitation is that this was a retrospective observational study. The treatment groups were not randomized, but women were selected to either CSII or MDI by clinical decision, aiming to achieve optimal glycemic control. Moreover, the CSII group differed in many aspects from the MDI group. Women treated with CSII were older, and they had more diabetes complications than women with MDI. Nevertheless, these differences did not result in worse glycemic control or pregnancy outcomes of CSII-treated women compared with MDI-treated women. In fact, the efficacy and safety of both treatment regimens were similar. The post hoc power analysis was carried out for some outcomes with obviously clear, but statistically nonsignificant, differences between the CSII and MDI group, showing that the power values were relatively low (e.g., 28.9% for large for gestational age and 33.0% for cesarean sections). Thus, because of the small sample size, for some outcome factors the study was underpowered in demonstrating differences between treatment groups.

Conclusions

In conclusion, this study confirms that pregnancies of women with T1DM can be safely treated either with CSII or MDI with similar maternal and fetal outcomes. CSII allowed for better glycemic control in the pregnancies of women with complicated diabetes. Therefore, an insulin pump is worth considering when planning pregnancies in women with high-risk T1DM, especially if good glycemic control is not achieved otherwise.

Footnotes

Acknowledgments

The authors gratefully acknowledge the colleagues from the Diabetes and Obstetric Clinics for their valuable contribution to the treatment of patients. The authors acknowledge the Research Committee of the Kuopio University Hospital Catchment Area for the State Research Funding (project QCARE, Joensuu, Finland).

Author Disclosure Statement

No competing financial interests exist.

P.K. planned the study and the study design, collected the data from medical records, and carried out the statistical analyses. T.L., T.W., and M.J. also contributed to the study design. P.K. wrote and edited various versions of the manuscript. T.W., M.J., and T.L. participated in the interpretation of data, reviewed and commented the manuscript. T.L. provided expert advice in statistics and contributed to discussions. All authors read and approved the final version of the manuscript.

References

Negrato

, Mattar

, Gomes

: Adverse pregnancy outcomes in women with diabetes. Diabetol Metab Syndr, 2012; 4:41–46.

Suhonen

, Hiilesmaa

, Teramo

: Glycaemic control during early pregnancy and fetal malformations in women with type I diabetes mellitus. Diabetologia, 2000; 43:79–82.

Boulot

, Chabbert-Buffet

, d'Ercole

, et al.: French multicentric survey of outcome of pregnancy in women with pregestational diabetes. Diabetes Care, 2003; 26:2990–2993.

Evers

, de Valk

, Visser

: Risk of complications of pregnancy in women with type 1 diabetes: nationwide prospective study in the Netherlands. BMJ, 2004; 328:915–921.

Eidem

, Vangen

, Hanssen

, et al.: Perinatal and infant mortality in term and preterm births among women with type 1 diabetes. Diabetologia, 2011; 54:2771–2778.

Dunne

, Avalos

, Durkan

, et al.: ATLANTIC DIP: pregnancy outcomes for women with type 1 and type 2 diabetes. Ir Med J, 2012; 105(5 Suppl):6–9.

Klemetti

, Nuutila

, Tikkanen

, et al.: Trends in maternal BMI, glycaemic control and perinatal outcome among type 1 diabetic pregnant women in 1989–2008. Diabetologia, 2012; 55:2327–2334.

Lenhard

, Kinsley

: Insulin therapy for the treatment of type 1 diabetes during pregnancy. J Matern Fetal Neonatal Med, 2014; 12:1270–1275.

Hanson

, Persson

, Thunell

: Relationship between haemoglobin A1c in early type 1 (insulin-dependent) diabetic pregnancy and the occurrence of spontaneous abortion and fetal malformation in Sweden. Diabetologia, 1990; 33:100–104.

10.

Holmes

, Young

, Patterson

, et al.: Optimal glycemic control, pre-eclampsia, and gestational hypertension in women with type 1 diabetes in the diabetes and pre-eclampsia intervention trial. Diabetes Care, 2011; 34:1683–1688.

11.

Golden

, Sapir

: Methods for insulin delivery and glucose monitoring in diabetes: summary of a comparative effectiveness review. J Manag Care Pharm, 2012; 18(6 Suppl):S1–S17.

12.

Botta

, Sinagra

, Angelico

M-C

, et al.: Comparison of intensified traditional insulin therapy and micropump therapy in pregnant women with type 1 diabetes mellitus. Minerva Med, 1986; 77:657–661.

13.

Burkart

, Hanker

, Schneider

HPG

: Complications and fetal outcome in diabetic pregnancy: Intensified conventional versus insulin pump therapy. Gynecol Obstet Invest, 1988; 26:104–112.

14.

Nosari

, Maglio

, Lepore

, et al.: Is continuous subcutaneous insulin infusion more effective than conventional insulin therapy in the treatment of pregnant diabetic women?. Diabetes Nutr Metab, 1993; 6:33–37.

15.

Lapolla

, Dalfrà

, Masin

, et al.: Analysis of outcome of pregnancy in type 1 diabetics treated with insulin pump or conventional insulin therapy. Acta Diabetol, 2003; 40:143–149.

16.

Cypryk

, Kosiński

, Kamińska

, et al.: Diabetes control and pregnancy outcomes in women with type 1 diabetes treated during pregnancy with continuous subcutaneous insulin infusion or multiple daily insulin injections. Pol Arch Med Wewn, 2008; 118:339–344.

17.

Cyganek

, Hebda-Szydlo

, Katra

, et al.: Glycemic control and selected pregnancy outcomes in type 1 diabetes women on continuous subcutaneous insulin infusion and multiple daily injections: the significance of pregnancy planning. Diabetes Technol Ther, 2010; 12:41–47.

18.

Volpe

, Pancani

, Aragona

, et al.: Continuous subcutaneous insulin infusion and multiple dose insulin injections in Type 1 diabetic pregnant women: a case-control study. Gynecol Endocrinol, 2010; 26:193–196.

19.

Mello

, Biagioni

, Ottanelli

, et al.: Continuous subcutaneous insulin infusion (CSII) versus multiple daily injections (MDI) of rapid-acting insulin analogues and detemir in type 1 diabetic (T1D) pregnant women. J Matern Fetal Neonatal Med, 2015; 28:276–280.

20.

Bruttomesso

, Bonomo

, Costa

, et al.: Type 1 diabetes control and pregnancy outcomes in women treated with continuous subcutaneous insulin infusion (CSII) or with insulin glargine and multiple daily injections of rapid-acting insulin analogues (glargine-MDI). Diabetes Metab, 2011; 37: 426–431.

21.

Neff

, Forde

, Gavin

, et al.: Pre-pregnancy care and pregnancy outcomes in type 1 diabetes mellitus: a comparison of continuous subcutaneous insulin infusion and multiple daily injection therapy. Ir J Med Sci, 2014; 183:397–403.

22.

Wender-Ozegowska

, Zawiejska

, Ozegowska

, et al.: Multiple daily injections of insulin versus continuous subcutaneous insulin infusion for pregnant women with type 1 diabetes. Aust N Z J Obstet Gynaecol, 2013; 53:130–135.

23.

Kallas-Koeman

, Kong

, Klinke

, et al.: Insulin pump use in pregnancy is associated with lower HbA1c without increasing the rate of severe hypoglycaemia or diabetic ketoacidosis in women with type 1 diabetes. Diabetologia, 2014; 57:681–689.

24.

Talaviya

, Saboo

, Joshi

, et al.: Pregnancy outcome and glycemic control in women with type 1 diabetes: A retrospective comparison between CSII and MDI treatment. Diabetes Metab Syndr, 2013; 7:68–71.

25.

Carta

, Meriggi

, Trossarelli

, et al.: Continuous subcutaneous insulin infusion versus intensive conventional insulin therapy in type I and type II diabetic pregnancy. Diabete Metab, 1986; 12:121–129.

26.

Chen

, Ben-Haroush

, Weismann-Brenner

, et al.: Level of glycemic control and pregnancy outcome in type 1 diabetes: a comparison between multiple daily insulin injections and continuous subcutaneous insulin infusions. Am J Obstet Gynecol, 2007; 197:404.e1–5.

27.

Kitzmiller

, Block

, Brown

, et al.: Managing preexisting diabetes for pregnancy: summary of evidence and consensus recommendations for care. Diabetes Care, 2008; 31:1060–1079.

28.

ACOG Committee on Practice Bulletins. ACOG Practice Bulletin. Clinical management guidelines for Obstetrician-Gynecologists. Number 60. Pregestational diabetes mellitus. Obstet Gynecol, 2005; 105:675–685.

29.

Vääräsmäki

, KaajaR , Kröger

, et al.: Timinonen S: Treatment of Diabetic During Pregnancy [in Finnish]. Finnish Diabetes Association. 2012. www.diabetes.fi/files/2169/Diabeetikon_hoito_raskauden_aikana.pdf (accessed June 11, 2012 ).

30.

Gimenez

, Conget

, Nicolau

, et al.: Outcome of pregnancy in women with type 1 diabetes intensively treated with continuous subcutaneous insulin infusion or conventional therapy. A case-control study. Acta Diabetol, 2007; 44:34–37.

31.

Chico

, Saigi

, García-Patterson

, et al.: Glycemic control and perinatal outcomes of pregnancies complicated by type 1 diabetes: influence of continuous subcutaneous insulin infusion and lispro insulin. Diabetes Technol Ther, 2010; 12:937–945.

32.

Gordin

, Forsblom

, Groop

, et al.: Risk factors of hypertensive pregnancies in women with diabetes and the influence on their future life. Ann Med, 2014; 46:498–502.

33.

Ryan

, Al-Agha

: Glucose control during labor and delivery. Curr Diab Rep, 2014; 14:450–454.