Neonatal Outcomes Among Neonates of Women With and Without Type 1 Diabetes in Sweden From 2010 to 2022

ABSTRACT

Aim

This register‐based study aimed to investigate differences in adverse neonatal outcomes between neonates born to mothers with and without type 1 diabetes in Sweden.

Methods

A nationwide cohort study was conducted using the Swedish Medical Birth Register, including all singleton births from 22 gestational weeks between 2010 and 2022. The neonatal outcomes among neonates of women with and without type 1 diabetes were studied with descriptive statistics and logistic regressions. Adjusted odds ratios (aORs) with 95% confidence intervals are presented.

Results

The study consisted of 1 402 394 neonates, and 0.5% had a mother with type 1 diabetes. Neonates of mothers with type 1 diabetes had increased odds of stillbirth (aOR 2.54; 95% CI 1.98–3.28), congenital malformations (aOR 1.66; 95% CI 1.50–1.84), preterm birth (aOR 3.80; 95% CI 3.41–4.23), large for gestational age (aOR 22.0; 95% CI 20.9–23.1), hypoglycaemia (aOR 45.85; 95% CI 43.60–48.22), and neonatal death (aOR 4.1; 95% CI 3.00–5.83).

Conclusion

Neonates of mothers with type 1 diabetes continue to experience poorer neonatal outcomes, highlighting the need for more research on the causes and how care can be improved for their mothers during pregnancy.

Neonatal outcomes among neonates of women with and without type 1 diabetes in Sweden from 2010 to 2022 by Goldberg A. et al.

Abbreviations

aOR

adjusted odds ratio

BMI

body mass index

CI

confidence interval

GW

gestational weeks

HbA1c

haemoglobin A1c

ICD

international classification of diseases

LGA

large for gestational age

MBR

swedish medical birth register

SGA

small for gestational age

Summary.

• This nationwide cohort study of 1 402 394 singleton pregnancies compared the neonatal outcomes between neonates of women with and without type 1 diabetes.

• Neonates of mothers with type 1 diabetes had increased odds of stillbirth, congenital malformations, preterm birth, hypoglycaemia, and neonatal death.

• More research is needed on the cause and how care can be improved, specifically how new medical technology can improve neonatal outcomes when the mother has type 1 diabetes.

1. Introduction

Sweden has one of the highest rates of type 1 diabetes in the world, with an incidence of 39.6 per 100 000 children during the years 2009–2013, and only surpassed by Finland with an incidence of 60.9 during the same years [1]. Type 1 diabetes is associated with adverse neonatal outcomes [2]. Desoye and Carter have summarised the underlying mechanism of adverse neonatal outcomes, attributing it to fetal hyperinsulinism induced by maternal hyperglycemia [3]. Hyperglycemia and hyperinsulinism accelerate metabolism and create oxidative stress in the fetus's cells. Furthermore, increased metabolism enhances oxygen consumption in utero [3]. Hyperglycemia and vasculopathy, common in women with type 1 diabetes, cause placental insufficiency by creating a smaller fetoplacental blood volume and inadequate placental blood flow and gas exchange [4]. The increased demand for oxygen and decreased supply of gas exchange lead to an undersupply of oxygen in the fetus [3]. An undersupply of oxygen and escalated metabolism in the fetus are the leading causes of increased adverse perinatal outcomes such as stillbirth, congenital malformations, preterm birth, and being born large for gestational age (LGA) [2, 3].

In a Swedish national cohort study from 1999, type 1 diabetes in pregnancy was associated with considerably increased rates of adverse obstetric and perinatal outcomes, for example, increased odds of stillbirth, congenital malformations, preterm birth, and LGA [2]. Since then, new diabetes medical technology and treatments have been introduced, contributing to increased time in the range of the goal p‐glucose. The new medical technology also shows promising results in improving the p‐glucose levels during a type 1 diabetes pregnancy [5]. Therefore, the incidence of neonatal complications among neonates of mothers with type 1 diabetes should have decreased.

This register‐based study aimed to investigate differences in adverse neonatal outcomes between neonates born to mothers with and without type 1 diabetes in Sweden.

2. Methods

We conducted a nationwide cohort study of pregnancies among women with and without type 1 diabetes. Individual information was obtained from the Swedish Medical Birth Register (MBR). MBR contains information on antenatal, obstetric, and neonatal care prospectively collected from standardised forms in more than 98% of pregnancies at a gestational age of 22 weeks or later in Sweden since 1973. The data in MBR were extracted from standardised regional electronic health records available at all antenatal clinics and hospitals and are considered of high quality [6]. Data were extracted from antenatal checkbox entries, where midwives marked the presence of the investigated variable in the antenatal records, and from diagnostic codes recorded in the medical records. Demographic data and diseases of the women, together with perinatal outcomes, were entered and classified in the health records at the time of hospital discharge.

2.1. Study Population

We included all singleton births at 22 or more completed gestational weeks (GW) in Sweden between 2010 and 2022 (N = 1 402 394). For almost 98% of all pregnant women, a fetal ultrasound in the first or second trimester determines GW. In Sweden, all pregnant women are offered a first‐ or early second‐trimester ultrasound scan free of charge to accurately date the pregnancy and scan for congenital anomalies [7].

2.2. Exposure

According to the International Classification of Diseases (ICD)‐10 diagnosis, type 1 diabetes was defined as O24.0B‐F, X, and E10.

2.3. Covariates

Maternal age was determined from the neonate's date of birth and each woman's personal registration number and categorised into the following groups: below 20 years of age, 20–24, 25–29, 30–34, and older than 34 years. BMI was calculated at the first antenatal visit and categorised as underweight (BMI < 18.5 kg/m²), normal weight (BMI 18.5–24.9 kg/m²), overweight (BMI 25.0–29.9 kg/m²), or obesity (BMI > 30.0 kg/m²). Parity was divided into primiparous or multiparous. Smoking use was defined as smoking or nonsmoking. Using snuff was defined as snuff user or not using snuff. The maternal country of birth was categorised as either Nordic countries or elsewhere. Prematurity was categorised into < 34 GWs and < 37 GWs. A low Apgar score was defined as < 7 at 5 min. Neonatal mortality was defined as death on Days 0–27.

ICD‐10 was used to classify complications before, during, and after pregnancy for both the mother and the neonate. Each diagnosis identified the following ICD‐10: chronic kidney disease N 18.9, pregnancy‐induced hypertension O14.0–2, congenital malformation Q00‐99. Neonates were identified as LGA or small for gestational age (SGA) according to the definition weights > 2 SD above or < 2 SD below the mean for average fetal growth, Swedish reference data, and identified with ICD‐10 codes P08.1 and P05.0–1 [8]. Shoulder dystocia, hyperbilirubinemia, and hypoglycemia in neonates were identified by O66.0, P58, P59, and P70.0–4 respectively. Neonatal plexus injury, neonatal seizures, and neonatal sepsis were identified by P14.0, P90, and P36 respectively. Chronic hypertension and stillbirth were identified from the checkboxes.

2.4. Statistical Analysis

We compared the characteristics of women with type 1 diabetes and without type 1 diabetes using numbers and proportions, which were tested with the chi‐square test and presented as p‐values. The risks of adverse neonatal outcomes were compared between newborns of mothers with and without type 1 diabetes. Logistic regression was used to calculate crude and adjusted odds ratios (aORs) with a 95% confidence interval (CI) in relation to newborns of mothers without type 1 diabetes. The adjusted model included birth year, BMI, smoking, use of snuff, maternal height, maternal age, parity, and country of birth, all of which were chosen because of their potential to influence both the exposure and the outcome. Maternal age, height, BMI, smoking, snuff use, and parity were included as covariates in the regression models, despite their potential role as mediators in the causal pathway between type 1 diabetes and neonatal outcomes. Adjustment for these variables aimed to estimate the direct association between maternal type 1 diabetes and neonatal outcomes, while accounting for relevant maternal characteristics that are routinely considered in clinical risk assessments. In a sub‐analysis, binary logistic regression was used to assess whether there were statistically significant differences in the risk of congenital malformations across BMI categories, using normal BMI as the reference category. Two models were used, one for women with type 1 diabetes and another for women without it, and presented as a p‐value. All statistical analyses were performed using IBM SPSS Statistics version 28.0 (IBM Corp., Armonk, NY, USA).

2.5. Ethics

The Swedish Ethical Review Authority approved the study on 2020‐03‐16 (2020‐00113).

3. Results

During the study period 2010–2022, we identified 7398/1402394 (0.5%) women with type 1 diabetes. Compared with the comparison group, women with type 1 diabetes were taller, had a higher BMI, and were primipara more often. Smoking and using snuff in early pregnancy and being born in a Nordic country were more common for mothers with type 1 diabetes. Chronic kidney disease and hypertension before pregnancy occurred more often in women with type 1 diabetes than in the comparison group (Table 1).

TABLE 1.

Maternal characteristics for women with and without type 1 diabetes (N = 1 402 394).

	No type 1 diabetes	Type 1 diabetes	p
	n = 1 394 996	n = 7398
	n (%)	n (%)
Age (years)
11–19	14 224 (1.0)	68 (0.9)	< 0.001
20–24	154 157 (11.1)	845 (11.4)
25–29	425 814 (30.5)	2401 (32.5)
30–34	490 107 (35.1)	2557 (34.6)
> 34	310 694 (22.3)	1527 (20.6)
Maternal height (cm)
130–162	386 366 (27.7)	1755 (23.7)	< 0.001
163–171	682 167 (48.9)	3686 (49.8)
≥ 172—	277 009 (19.9)	1715 (23.2)
Data missing	58 889 (4.2)	234 (3.3)
Maternal BMI
< 18.5	29 289 (2.1)	44 (0.6)	< 0.001
18.5–24.9	737 824 (52.9)	3019 (40.4)
25.0–29.9	345 430 (24.8)	2453 (33.2)
≥ 30	189 527 (13.6)	1451 (19.6)
Data missing	92 926 (6.7)	431 (5.8)
Primiparas	596 421 (42.8)	3419 (46.2)	< 0.001
Daily smoking	63 807 (4.6)	356 (4.8)	0.022
Data missing	69 862 (5.0)	418 (5.7)
Snuff use	17 359 (1.2)	137 (1.9)	< 0.001
Data missing	40 677 (2.9)	172 (2.3)
Nordic maternal country of birth a	1 026 104 (73.6)	6705 (90.6)	< 0.001
Chronic kidney disease	5953 (0.4)	79 (1.1)	< 0.001
Chronic hypertension	7179 (0.5)	195 (2.6)	< 0.001

Note: Abbreviations: BMI, body mass index (kg/m²); Cm, centimetre.

^a p < 0.05 is concidered statistical significant.

The odds of stillbirth and congenital malformations was elevated among neonates of type 1 diabetes mothers (aOR 2.54; 95% CI 1.98–3.28, and aOR 1.66; 95% CI 1.50–1.84), respectively (Table 2, Figures 1, 2, 3). In the sub‐analysis, the prevalence of congenital malformations was significantly higher among underweight and obese women with type 1 diabetes compared to their normal‐weight counterparts. Similarly, an increased prevalence was observed among newborns of obese women without type 1 diabetes compared to those born to normal‐weight women without type 1 diabetes (Figure 3). Furthermore, neonates of mothers with type 1 diabetes were more often born preterm before GWs 34 and before GWs 37 (aOR 3.80; 95% CI 3.41–4.23, and aOR 6.54; 95% CI 6.19–6.91), respectively (Table 2).

TABLE 2.

Neonatal complications among infants of mothers with (n = 7398) and without (n = 1 394 996) type 1 diabetes.

	No type 1 diabetes	Type 1 diabetes	Crude OR	Adjusted OR a
	n (%)	n (%)	OR (95% CI)	(95% CI)
Stillbirth	4769 (0.3)	62 (0.8)	2.46 (1.92–3.17)	2.54 (1.98–3.28)
GW < 34	18 107 (1.3)	359 (4.9)	3.88 (3.49–4.32)	3.80 (3.41–4.23)
GW < 37	63 714 (4.6)	1750 (23.7)	6.47 (6.13–6.84)	6.54 (6.19–6.91)
Shoulder dystocia	3535 (0.3)	149 (2.0)	8.10 (6.86–9.55)	8.58 (7.26–10.14)
Plexus injury	1649 (0.1)	65 (0.9)	7.49 (5.84–9.61)	7.35 (5.72–9.44)
Apgar < 7 at 5 min	22 129 (1.6)	380 (5.1)	3.38 (3.04–3.75)	3.30 (2.68–3.66)
Neonatal seizure	2015 (0.1)	36 (0.5)	3.38 (2.43–4.70)	3.21 (2.31–4.48)
LGA	45 907 (3.3)	3079 (41.6)	21.0 (20.0–22.0)	22.0 (20.9–23.1)
SGA	32 879 (2.4)	76 (1.0)	0.43 (0.34–0.54)	0.48 (0.38–0.60)
Congenital malformations	45 257 (3.2)	395 (5.3)	1.68 (1.52–1.86)	1.66 (1.50–1.84)
Hypoglycaemia	20 534 (1.5)	2947 (39.8)	44.3 (42.2–46.5)	45.9 (43.6–48.2)
Jaundice	25 412 (1.8)	469 (6.3)	3.65 (3.32–4.01)	3.54 (3.22–3.89)
Neonatal death Day 0–27	1748 (0.1)	36 (0.5)	3.92 (2.81–5.45)	4.18 (3.00–5.83)
Neonatal Sepsis	9292 (0.7)	138 (1.9)	2.84 (2.40–3.36)	2.64 (2.22–3.13)

Abbreviations: GW, gestational week; LGA, large for gestational age; SGA, small for gestational age.

^a Adjusted for: maternal age, maternal height, body mass index, parity, smoking, using snuff, maternal country of birth (Nordic country or all other countries), and birth year.

FIGURE 1.

Stillbirth among newborns of women with and without type 1 diabetes.

FIGURE 2.

Congenital malformations among newborns of women with and without type 1 diabetes.

FIGURE 3.

Congenital malformations stratified on maternal body mass index (BMI) categories among newborns of women with and without type 1 diabetes. Two separate models were used, one for women with and one for women without type 1 diabetes, with BMI 18.5–24.9 serving as the reference category in each model.

Moreover, 41.6% of the neonates of type 1 diabetes mothers were born LGA compared with 3.3% in the comparison group (aOR 22.0; 95% CI 20.9–23.1), and the odds of hypoglycaemia was increased among neonates of mothers with type 1 diabetes (aOR 45.85; 95% CI 43.60–48.22). Furthermore, neonates of mothers with type 1 diabetes had a fourfold higher neonatal mortality (aOR 4.1; 95% CI 3.00–5.83) compared with those in the general population (Table 2).

4. Discussion

The main findings of this population‐based study were the increased perinatal risks among neonates of women with type 1 diabetes compared to neonates of women without type 1 diabetes. Specifically, we found an increased odds of stillbirth, congenital malformations, preterm birth, LGA, neonatal hypoglycaemia, and neonatal death. Still, we saw improved perinatal outcomes for neonates of mothers with type 1 diabetes compared with a Swedish register study, including data on 5089 women with type 1 diabetes and 1 260 207 women without type 1 diabetes during the years 1991 to 2003 [2]. Specifically, we found a decrease in stillbirth from an incidence of 1.5% to 0.8%. Conversely, we found an increased incidence of preterm birth and increased prevalence of LGA compared to the same Swedish register study [2].

The incidence of stillbirth was higher among women with type 1 diabetes compared to women without type 1 diabetes, with a minor trend line towards lower rates during the study period of 13 years. Additionally, a lower proportion of stillbirth rates in women with type 1 diabetes during 2010–2022 compared to 1991–2003 may reflect improvements in diabetes management, including glycemic control [2, 9]. While national and regional guidelines recommend maintaining HbA1c levels < 48 mmol/mol before pregnancy, this target remains higher than normoglycemia [10, 11, 12, 13]. Notably, Swedish regions differ in their recommended HbA1c levels before and during pregnancy, ranging from 38 to 48 mmol/mol in some local guidelines to < 57 mmol/mol in others [13]. Previous studies have demonstrated an increased risk of stillbirth when HbA1c exceeds 48 mmol/mol [9, 14], However, neither of these studies examined the risk of stillbirth in women with HbA1c levels between 42 and 48 mmol/mol [9, 14]. An HbA1c level below 42 mmol/mol is considered normal. Since glucose levels are the primary distinguishing factor between women with and without type 1 diabetes, comparing women with type 1 diabetes who achieve HbA1c levels < 42 mmol/mol to those with levels > 42 mmol/mol, as well as to women without type 1 diabetes, may offer a more appropriate basis for assessing potential causal effects on stillbirth risk in this population.

The elevated prevalence of congenital malformations among neonates of mothers with type 1 diabetes observed in this study is consistent with prior research [2, 15, 16]. Eriksen et al. reported an increased risk of congenital malformations even with a HbA1c < 53 mmol/mol before or during the first trimester [15], while Ludvigson et al. reported a twofold increased risk of cardiac defect in neonates of mothers with type 1 diabetes and a HbA1c even below 48 mmol/mol [16]. These findings suggest that current glycemic targets may still be suboptimal for fetal outcomes. However, other diabetes‐related factors, beyond glycemia, may also contribute to the increased malformation risk in type 1 diabetes pregnancies.

Elevated BMI might play a part in the increased risk of congenital malformations in neonates of women with type 1 diabetes. There was an increased prevalence of congenital malformation for every elevated level of BMI in women with type 1 diabetes, although this was only significant with obesity. Our findings align with and strengthen the conclusions of Persson et al. regarding the synergy between elevated BMI and type 1 diabetes in the occurrence of congenital malformations [17]. Additionally, elevated BMI makes it more challenging to detect congenital malformations with an ultrasound [18]. Simultaneously, for women with type 1 diabetes, the highest risk of congenital malformation was seen in the underweight BMI category. Women with type 1 diabetes who are underweight are often undertreated with insulin and therefore also frequently have hyperglycemia [19]. Therefore, we suspect that the reason for our observed association between low BMI and congenital malformations is primarily due to hyperglycemia.

The increased risk of preterm birth for neonates of mothers with type 1 diabetes compared to neonates of mothers without type 1 diabetes might also partly be explained by hyperglycaemia. Ludvigsson et al. described an increased risk of both iatrogenic and spontaneous preterm birth for neonates of mothers with type 1 diabetes compared to neonates of mothers without type 1 diabetes, and the risk increased for every level of elevated HbA1c [20]. Women with type 1 diabetes have other added risk factors for preterm birth compared to women without type 1 diabetes; for example, they tend to develop pre‐eclampsia more often [2]. In our study, the risk of preterm birth for neonates of mothers with type 1 diabetes has escalated compared to two Swedish register studies from 2009 and 2019 [2, 20]. LGA is suggested to elevate the risk of preterm birth, both iatrogenic and spontaneous [21]. Consequently, the acceleration of LGA in neonates of mothers with type 1 diabetes, seen in this study, could be one more of the explanations for increased preterm birth in this group.

Hence, we found an increased prevalence of LGA and a decreased prevalence of SGA when the mother had type 1 diabetes, compared to mothers without type 1 diabetes and compared with previous research. Both hyperglycaemia and elevated BMI explain LGA in neonates of mothers with type 1 diabetes [2, 8]. An LGA neonate is linked to increased risks for both the mother and the neonate, like shoulder dystocia and plexus injury; findings also increased in this study [22]. Shoulder dystocia and plexus injury in neonates of mothers with type 1 diabetes are explained not only by the increased amount of LGA but also because they are built disproportionately and with broader shoulders [23, 24]. Prepregnancy consulting focusing on a normal BMI is recommended in some countries. Still, no recommendations exist on how women with type 1 diabetes best achieve a normal BMI. We suggest that future research should focus on how prepregnancy consulting can help women with type 1 diabetes strive for a normal BMI before pregnancy and explore if it affects the prevalence of LGA.

Although this study concurs with earlier research on the increased prevalence of hypoglycaemia in neonates of mothers with type 1 diabetes, our findings should be carefully interpreted. When the mother has type 1 diabetes during pregnancy, her neonate is tested for hypoglycaemia following birth. Among neonates of mothers without type 1 diabetes, only those with risk factors or clinical signs of hypoglycaemia are tested. We therefore suspect that the prevalence of hypoglycaemia (1.5%) in the comparison group of our study is underestimated.

Neonates of mothers with type 1 diabetes still have worse health outcomes compared to the general Swedish birthing population. In 1989, the St Vincent Declaration stated that pregnancies complicated with type 1 diabetes should have the same health outcomes as pregnancies without type 1 diabetes [25]. The Swedish Health and Medical Act complies with the St Vincent Declaration by declaring that care should be received equally [26]. This study demonstrates that the St Vincent Declaration and the Swedish Health and Medical Act have not yet been fulfilled when it comes to the life and health of neonates of mothers with type 1 diabetes.

4.1. Strengths and Limitations

The strength of this study was the ability to study the whole birthing population for a long period. Furthermore, our comprehensive register study prospectively collected data without selection bias and was able to study rare outcomes, like stillbirth. Limitations of this study were the inability to adjust for all the potential diabetic complications, such as preexisting micro‐ and macroangiopathy, and the lack of control for the women's p‐glucose before and during pregnancy. Another limitation is that only pregnancies with GW ≥ 22 were included, and that can lead to selection bias; hence, abortions occur before GW 22 in Sweden. Furthermore, we could not control for planned or missed abortions due to congenital malformations. Women with type 1 diabetes have been described to undergo planned abortion [27] and missed abortion [28] more often than women without type 1 diabetes, leading to a risk of underestimating the prevalence of congenital malformations in neonates of mothers with type 1 diabetes.

5. Conclusion

This study concluded that neonates born to mothers with type 1 diabetes faced higher risks compared to neonates born to mothers without type 1 diabetes, highlighting the need for a better understanding of the causation and how care can be improved for their mothers. Further research should explore how advancements in medical technology can improve the outcomes for neonates of mothers with type 1 diabetes.

Author Contributions

Alexandra Goldberg: conceptualization, investigation, writing – original draft, methodology, visualization, writing – review and editing, formal analysis, data curation. Eva Wiberg‐Itzel: writing – review and editing, supervision, conceptualization. Joanna Tingström: conceptualization, writing – original draft, supervision. Cecilia Ekéus: conceptualization, investigation, funding acquisition, methodology, visualization, writing – review and editing, formal analysis, project administration, supervision, resources.

Conflicts of Interest

The authors declare no conflicts of interest.

Goldberg A., Wiberg‐Itzel E., Tingström J., and Ekéus C., “Neonatal Outcomes Among Neonates of Women With and Without Type 1 Diabetes in Sweden From 2010 to 2022,” Acta Paediatrica 114, no. 12 (2025): 3191–3198, 10.1111/apa.70230.