Does the mother’s blood glucose affect her baby? This is the underlying question which drives the interest in what is presently known as gestational diabetes mellitus. It is a different, although related question to that of does pregnancy cause diabetes? Whatever the question, the answer interests many health care professionals who have a concern about diabetes: specialist nurses, midwives, dieticians, and laboratory staff, as well as general practitioners, obstetricians, neonatologists, epidemiologists, and diabetologists.

A recent systematic review on recurrence of gestational diabetes tackles a number of problems in this field and comes up with some answers and some sensible suggestions. The presently agreed definition of gestational diabetes—glucose intolerance with onset or first recognition during pregnancy—which is said to affect between 4% and 12% of all pregnancies, and to be becoming more common, is nonetheless unsatisfactory. It does not distinguish between mothers who were already unknowingly diabetic (Type 2) before they became pregnant and those who became temporarily diabetic only during the pregnancy. It also demands demonstration of glucose intolerance by some form of glucose tolerance test, rather than a simpler demonstration of a higher than normal blood glucose level.

Most of the evidence about gestational hyperglycaemia dates back 50 years to studies in Boston, where an oral glucose tolerance test in the third trimester was used to define the diagnostic levels of blood glucose and validated by the eventual development of permanent Type 2 diabetes in the mother up to 25 years later. This, of course, is not the same as recurrence of transient hyperglycaemia in a subsequent pregnancy but does indicate that it is useful to know whether or not the mother remains diabetic after her pregnancy or becomes so at some later date.

A large international multicentre study of maternal glucose tolerance in the third trimester in normal (non-diabetic) pregnancy, HAPO (Hyperglycemia and Adverse Pregnancy Outcome), has recently reported that increasing levels of maternal blood glucose, including those previously thought to be normal in pregnancy, show a gradually increasing effect on a number of outcomes for the baby. These include birth weight, fatness, tendency toward neonatal hypoglycaemia, and the risk of delivery by caesarean section. This study, after worldwide discussion, will set the scene for future management protocols and guidelines for maternal blood glucose in pregnancy—whether it be called hyperglycaemia in pregnancy or gestational diabetes.

Methods and Key Results

A systematic review attempts to make sense of all the published research on a topic. The review by Kim et al has a narrow definition of aim, which makes the analysis simpler. The authors found only 13 articles from 1965 to 2006 which accurately investigated the recurrence of gestational diabetes in a second or subsequent pregnancy.

They found a reported recurrence rate of between 30% and 84%. Lower rates were found in white populations, and higher rates were found in Latina, African American, and Asian populations, although there was no systematic study of selected ethnic groups (international terminology and definition of ethnicity are by no means standard).

No other risk factors were consistently reported to predict actual recurrence of gestational diabetes, although the review identified the need for better definition of pre-existing diabetes, socio-economic status, and postpartum blood glucose screening and the importance of inter-pregnancy intervals. These factors, along with maternal age, obesity, and family history of diabetes have all been associated with the initial risk of gestational diabetes.

Perhaps the most surprising result is that recurrence is far from 100%. Does this mean that diabetes in pregnancy is different from other types of diabetes in at least some cases? This would support an environmental (nutritional- and exercise-based) cause for the initial diagnosis, rather than a genetic predisposition. And it further suggests that something useful by way of nutritional and activity management is possible.

Analysis of different studies from different countries, using different screening strategies and different diagnostic criteria, showed that these contributors to variance were not of clinical importance: the actual level of blood glucose appeared to have similar effects in different ethnic populations in different environments. However, a direct answer to the question under review was difficult because of the rarity of studies which included any measurement of blood glucose after the index pregnancy. This is seen as a public health problem rather than one for the obstetrician or neonatal paediatrician and is therefore only done under a few research protocols.

Clinical Implications

Recurrence of maternal hyperglycaemia in a subsequent pregnancy is very variable. It appears to be more common in those of non-white race/ethnicity. The classic risk factors of maternal age, parity, body mass index, and even oral glucose tolerance were inconsistent predictors of recurrence.

To improve the diagnosis of recurrent gestational diabetes, a more consistent and earlier antepartum screening process is needed. The present increase in maternal obesity at younger ages will produce earlier insulin resistance, and this will increase previous estimates of the risk of hyperglycaemia in pregnancy.

Postpartum screening of blood glucose after an index pregnancy remains suboptimal but is needed if effective management is to be offered to prevent recurrence in a subsequent pregnancy or in the longer-term risk of permanent Type 2 diabetes. The authors suggest that linkage of obstetric, neonatal, and diabetes databases would be an effective process to further these aims.

Finally, they draw attention to several prospective studies which provide evidence that appropriate intervention can both reduce recurrence of gestational diabetes and preserve β-cell function in women with a recent history of this condition. Both lifestyle and pharmacological interventions have been shown to be useful and will allow cost-effectiveness to be demonstrated. These interventions may be appropriately tailored by race/ethnicity. As the authors note, “Women may be responsive to improving their fetus’s health even more than their own.”

Further Reading

Readers interested in further information on gestational diabetes can view the Proceedings of the Fifth International Workshop-Conference on Gestational Diabetes, published in Diabetes Care. 2007;30(suppl 2):S105-S260.