Table 8

Impact of detection and management of maternal diabetes mellitus on stillbirth and perinatal mortality


Location and Type of Study


Stillbirths/Perinatal Outcomes

Reviews and meta-analyses

Russell et al. 2007 [67]


Review. 3 RCTs included, 1 reported perinatal outcomes.

Assessed the impact of treatment of gestational diabetes on perinatal outcomes.

Serious perinatal complication (shoulder dystocia, nerve injury, fracture, or perinatal death): 67% reduction (1 RCT).

Macrosomia: 53% reduction (1 RCT).

2 of 3 studies lacked power to detect a difference in perinatal outcomes.

Tuffnell et al. 2003 [73].


Review (Cochrane). 3 RCTs and quasi-RCTs included (N = 223 pregnant women with impaired glucose tolerance)

Assessed the effect of treatments for impaired glucose tolerance on perinatal outcome.

PMR: Insufficient data to assess.

Reduction in BW > 90% percentile: RR = 0.55, 95% CI: 0.19–1.61) [NS]

Boulvain et al. 2001 [84]


Review (Cochrane). 1 RCT included (N = 200 term diabetic women).

Assessed the effects of a policy of elective delivery (intervention) vs. expectant management (controls) on maternal and perinatal mortality and morbidity.

PMR: RR not estimable

[0/100 in both groups].

Mukhopadhyay et al. [81]

United Kingdom.

Review (non-Cochrane). 5 RCTs included (N = 182 diabetic pregnant women).

Compared the impact of continuous subcutaneous insulin infusion (CSII) (intervention) with multiple daily insulin injections (MDI)/intensive conventional therapy (ICT) (controls).

SBR: OR = 2.50 (95% CI: 0.53 – 11.77) [NS]; P = 0.25

[6/94 (6.4%) vs. 1/88 (1.1%) in intervention and control groups, respectively].

Intervention studies

Hod et al. 2008 [82]

Multicentre, multinational. 63 study sites in 18 countries.

Open-label RCT. N = 322 pregnant type I diabetic women (N = 157 intervention group, N = 165 controls).

Assessed the impact of mealtime insulin Aspart (IAsp) (intervention) with human insulin (HI) (controls), both in combination with basal neutral protamine Hagedorn (NPH) insulin.

SBR: one in each group.

PMR: 14 vs. 22/1000 births in intervention and control groups, respectively.

Bancroft et al. 2000 [174]

UK (West Yorkshire). Hospital-based study (district general hospital and large teaching hospital).

RCT. Pregnant women (N = 68) with impaired glucose tolerance.

Compared the effects in a group that monitored plasma glucose up to 4× daily (intervention) vs. an unmonitored group (controls). Median plasma glucose measurements in intervention group = 118 (range: 0–500); 19% of women in the monitored group treated with insulin.

PMR: 0/36 vs. 0/32 in intervention vs. control groups, respectively [NS]

NMR: 0/36 vs. 0/32 in intervention vs. control groups, respectively [NS]

Karmon et al. 2009 [175]


Retrospective cohort study.

Measured pregnancy outcomes in women with diet-controlled gestational diabetes subject to a routine policy of labour induction at 40 weeks.

Rates of dystocia, congenital malformation, and macrosomia higher in offspring of diet-controlled diabetic patients than non-diabetic patients.

Perinatal mortality: no difference when adjusted for confounders.

SBR: Higher in non-diabetic women after 40 weeks (likely due to policy of induction of diabetic women at 40 weeks).

Langer et al. 1994 [74]

USA (Texas). Population-based.

Prospective study. Pregnant women (N = 2461; N = 1145 intervention, N = 1316 diabetic controls) with gestational diabetes and a non-diabetic control group (N = 4922).

Compared the impact of intensified management (intervention) vs. conventional management (diabetic controls) vs. non-diabetic controls on adverse pregnancy outcomes.

SBR: 1/1000 (N = 1145) vs. 4/1000 (N = 1316) vs. 4/1000 (N = 4922) in the intervention, diabetic controls and non-diabetic controls, respectively.

Macrosomia, Caesarean section, metabolic complications, shoulder dystocia, NICU days, and respiratory complications lower among intervention group than diabetic controls; comparable to non-diabetic controls.

Observational studies

Aucott et al. 1994 [83]


Prospective study. Pregnant patients (N = 78) with pre-gestational diabetes vs. matched non-diabetic pregnant controls (N = 78).

Compared the impact on stillbirth of treatment with insulin (exposed) by either infusion pump (N = 20) or split-dose therapy (N = 58) vs. matched controls (unexposed).

SB: 1/78 vs. 0/78 in exposed vs. unexposed groups, respectively.

Fadel et al. 1982 [176]


Observational study. N = 84 women with gestational diabetes, N = 23 women with pre-gestational diabetes.

Compared the impact of a protocol of intensive diabetes management including frequent prenatal visits, strict metabolic control, antepartum monitoring including estriols and contraction stress tests, liberal hospitalization policy, fetal lung maturity assessment, and intrapartum fetal monitoring.

SB: 1 antepartum SB among women with gestational diabetes (due to true knot in cord); 0 SB in women with pre-gestational diabetes.

Caesarean section: 15.4% among gestational diabetics, 56.5% among pre-gestational diabetics.

Macrosomia: 20% among gestational diabetics, 13% among pre-gestational diabetics.

Banerjee et al. 2004 [77]

India. Antenatal clinic.

Prospective study. N = 240 women with gestational (GDM) and pre-gestational (PGDM) diabetes mellitus.

All women were placed on exercise, diet, and/or insulin therapy. And were divided into three groups based on blood glucose levels and HbA1C: Tight Glycaemic Control (TC), Acceptable glycaemic control (AC) and uncontrolled glycaemic group (UC).

PMR: 4.16% vs. 18.2% vs. 22.2% in the TC, AC and UC subgroups of GDM.

PMR: 0% vs. 20% vs. 40% in TC, AC and UC subgroups of PGDM.

Gonzalez-Quintero et al. 2007 [76]

USA. Centralised perinatal database.

Retrospective study on prospective data. N = 3,218 women participating in the outpatient GDM management programme.

Compared the impact on stillbirth of women whose blood glucose levels were controlled vs. those uncontrolled.

SBR (n,%): 2 (0.1) vs. 4 (0.3) in women with controlled vs. uncontrolled GDM.

Huddle et al 1993 [79]

South Africa (Soweto).

Retrospective study. Patient records (N = 733 women; N = 348 with gestational diabetes).

To assess the impact on perinatal mortality of the use of a specialised service for diabetic pregnant women (intervention) vs. non-use of the service (controls).

PMR: 3.7% vs. 15.6% in intervention vs. controls, respectively.

McElvy et al. 2000 [78]


Retrospective before-after study. Pregnant Type-1 diabetic women enrolled peri-conceptionally (N = 306; N = 111 before, N = 103 during, N = 92 after programme).

To evaluate the impact on perinatal mortality of a focused preconceptional and early pregnancy programme for Type 1 diabetes including strict glucose control and antepartum fetal surveillance from 32 weeks gestation until delivery.

PMR: 3% vs. 2% vs. 0% before, during, and after the programme, respectively.

Landon et al. 1992 [80]

USA (Ohio). 2 teaching hospitals.

Prospective cohort study. Pregnant women (N = 114) with insulin-dependent diabetes. Non-stress testing was begun weekly at 28–30 wks and 2× weekly at 32 wks (N = 1676 NSTs performed (14.7+/-3.2 tests per patient)).

8% of tests (N = 134) non-reactive, necessitating a biophysical profile.

To determine whether maternal vascular disease and/or glycaemic control are associated with fetal condition in diabetic pregnancies by comparing the effect of reactive (exposed) vs. non-reactive (unexposed) NST on perinatal outcomes.

Fetal death (miscarriage + SB): 1/114.

N = 10 deliveries among patients with abnormal test results.

No significant differences in ambulatory glucose profile data in exposed vs. unexposed groups.

No significant differences in glycaemic parameters in women delivered for suspected fetal compromise vs. nonintervention group.

8/20 (40%) women with nephropathy or hypertension required delivery for fetal well-being, vs. 2/94 women (2%) without nephropathy or hypertension but with abnormal test results (P < 0.001).

Nachum et al. 2001 [177]


Prospective controlled study. Pre-gestationally and gestationally diabetic women (N = 681 women; N = 801 pregnancies) recruited 1986–1989.

Compared the impact of diabetic pregnancies managed by hospitalisation vs. those managed by ambulatory care.

NMR: 1/394 vs. 0/407 in hospitalised vs. ambulatory care groups, respectively.

Haws et al. BMC Pregnancy and Childbirth 2009 9(Suppl 1):S5   doi:10.1186/1471-2393-9-S1-S5

Open Data