The effect of established and gestational diabetes on pregnancy outcome

OBJECTIVE--To study the prevalence and type of glucose intolerance in pregnancy and the effect of different types on perinatal mortality and fetal size. DESIGN--A prospective case-control study with data collected by patient interview and examination of all available records during a 16-months period between 1984 and 1986. SETTING--A large maternity hospital in Kuwait where diabetes in pregnancy is common. SUBJECTS--The cases were a consecutive sample of 731 women, delivered during the study period, recorded in the labour ward register as being diabetic or having abnormal glucose tolerance, the control group was formed from the next woman in the register (provided she was not known to be diabetic). MAIN OUTCOME MEASURES--Type of diabetes followed the WHO classification, with subdivision depending on level of fasting plasma glucose. Type of perinatal death was examined in detail and birthweight centile calculated. RESULTS--Of the 731 cases, 22% were established diabetics, most were treated with oral hypoglycaemic drugs before pregnancy and insulin during pregnancy. Of those discovered during pregnancy, 43% were classified as gestational diabetes and the remainder as impaired glucose tolerance. Overall, 50% of cases were treated with insulin. Established diabetics had a perinatal mortality rate nearly four times greater than non diabetics (RR, 3.7, 95% CI 2.6 to 6.4) and for gestational diabetics RR was 2.0 95% CI 1.2 to 3.7). Unexplained deaths were particularly common, both in established diabetics (RR, 18.4, 95% CI 3.9 to 85.7) and in gestational diabetics (RR, 13.4, 95% (CI 2.9 to 61.6). Cases with impaired glucose tolerance had no stillbirths and had a lower perinatal loss than the controls, though this was not statistically significant. Heavier babies were seen in all case groups compared with controls, though the impaired glucose tolerance group had lower birthweights than the other two case groups. CONCLUSIONS--Type 2 diabetes was found to be common, most cases being diagnosed in pregnancy. Under the conditions found in Kuwait, diabetes, in the sense of a raised fasting glucose, is accompanied by a high rate of perinatal loss from unexplained stillbirth. This applies whether the condition was present before pregnancy or was discovered during pregnancy. Fetal macrosomia was also common in both situations. Impaired glucose tolerance, where fasting levels remain normal, does not appear to increase fetal loss, but may be associated with fetal macrosomia. As these women age they are likely to develop overt diabetes in the non-pregnant state, and subsequently to develop serious complications of this disease. Improving glycaemic control, both during preg