The ultimate complication of type 1 diabetes in combination with pregnancy is maternal death, which may result from complications of the pregnancy itself, diabetes and associated diseases, or causes related to neither pregnancy nor diabetes. From the clinical point of view, the greatest reward would be to identify possible preventable causes of maternal death beforehand, especially because these patients are under close surveillance during the entire pregnancy and postpartum period.

The reported incidence of maternal mortality of pregnant type 1 diabetic women has been ∼0.5% (1,2), which is 5–20 times higher than that of the general obstetric population. However, these estimates date back prior to 1980, and because of the developments in both the obstetric management and the treatment of diabetes during the past 20 years, we have estimated the risk of death and analyzed the causes of maternal mortality in a large single referral center for all pregnant type 1 diabetic women from Southern Finland.

Between 1975 and 1997, 972 type 1 diabetic women delivered, or intended to deliver, in the Department of Obstetrics and Gynecology at the University Central Hospital of Helsinki. This is the referral center for all pregnant diabetic women in Southern Finland (population 1.5 million). If the diabetes classification was not evident based on the diagnoses established at pediatric or adult endocrinology units, it was confirmed by the undetectable plasma C-peptide levels. All patients were followed-up for an interval of 1–6 weeks during pregnancy and within 1–2 months after delivery. Maternal deaths (during pregnancy or within 42 days after delivery) were recorded, and the causes of death were determined by a forensic medical autopsy or by clinical findings.

Of the 972 women, 5 (0.51%, 95% CI 0.17–1.20) died during pregnancy or in the postpartum period (Table 1). Four these women had a duration of diabetes >20 years; two of the deaths were caused by hypoglycemia and one by ketoacidosis. Patient 1 suffered a brain stem infarction the night after cesarean section. The diagnosis was made by neurological senior consultants on the basis of clinical findings. She never regained consciousness. Patient 2 died from the complications of a massive unintentional spinal anesthesia for an elective cesarean section. She could not be intubated because of severe rheumatoid arthritis, and a tracheostomy was performed too late. Patient 3 was found unconscious at her home at 14 weeks of gestation. At that time, she was hypoglycemic and had cardiac ventricular fibrillation. She died from anoxic brain damage at 24 weeks of gestation. Patient 4 had labile diabetes with wide fluctuations in blood glucose levels and frequent episodes of hypoglycemia. She was found dead at her home at 10 weeks of gestation, and, after combining the clinical data and the findings of a forensic autopsy, the cause of her death was assigned to severe hypoglycemia. Patient 5 had dilatation and curettage for a missed abortion at 9 weeks of gestation. Thirteen days later, she became disoriented and suddenly lost consciousness, and when an emergency team arrived, she was already dead. A forensic chemical investigation revealed ketoacidosis and intoxication by trimipramine, ethylmorphine, and temazepam.

The relative death rate of type 1 diabetic Finnish women increases with the duration of diabetes and is highest at 30–34 years of age (3). In Finnish women with a duration of type 1 diabetes between 20 and 25 years, the relative death rate was 8.9 times higher than in nondiabetic Finnish women of the same age (3). On the other hand, the maternal mortality in Finland during the 1980s was 4.7 deaths per 100,000 births. Based on these data and this study, the mortality of type 1 diabetic mothers was 109 times greater than the general population and 3.4 times greater than nonpregnant type 1 diabetic women when calculated in person-years (each diabetic pregnancy was considered as 1 person-year).

None of the deaths were definitely associated with unsuspected diabetic complications that have a high maternal mortality rate, such as ischemic heart disease, although vascular disease might have been a contributing factor in the death of one of our patients (Patient 1). As reported in previous studies (1), anesthetic complications are an important cause of maternal death in diabetes. Other directly obstetrical deaths were not observed in our study.

The tight metabolic control of diabetes during pregnancy that is mandatory for the normal development of the fetus may expose the mother to life-threatening cases of hypoglycemia. Two deaths in our study could be assigned to “dead-in-bed syndrome” (4). Both incidents took place during the first half of pregnancy when, in particular, nocturnal hypoglycemic events are known to be prevalent. Whether such a pregnancy predisposes diabetic mothers to dead-in-bed syndrome or triggers subsequent mechanisms is equivocal; regardless, these deaths may amount to 24% of all deaths in young diabetic patients (4). However, in ours and other studies (1), these deaths might have been at least theoretically preventable, and we feel that first-trimester care of preganant diabetic women must focus on hypoglycemia.

1
Gabbe S-G, Mestman J-H, Hibbard L-T: Maternal mortality in diabetes mellitus: an 18-year survey.
Obstet Gynecol
48
:
549
–551,
1976
2
Cousins L: Pregnancy complications among diabetic women: review 1965–
1985
. Obstet Gynecol Surv
42
:
140
–149,
1987
3
Lounamaa R: Mortality in Finnish Patients With Insulin-Dependent Diabetes Mellitus. Helsinki, The Social Insurance Institution, 1993
4
Sovik O, Thordarson H: Dead-in-bed syndrome in young diabetic patients.
Diabetes Care
22(Suppl. 2)
:
B40
–B42,
1999

Address correspondence and reprint requests to Pekka Leinonen, MD, Department of Obstetrics and Gynecology, P.O.B. 140, Helsinki, FIN-00029, Finland. E-mail: pekka.leinonen@hus.fi.