Controversy grows over redefinition of gestational diabetes

One of the freshest fights over the definition of disease is happening with gestational diabetes. A recent redefinition more than doubles the patient pool and labels up to one in five pregnant women as unwell.1 The new criteria arose at a conference in California in 2008, were published in 2010, and are now being endorsed around the world, including by the Australasian Diabetes in Pregnancy Society. Yet the expanded definition is the subject of sustained criticism and will be further scrutinised at a United States National Institutes of Health (NIH) consensus development conference, by an adjudicating panel free of members with financial or reputational conflicts of interest, to be held in Maryland in October this year.

Proponents argue that new evidence linking raised maternal blood glucose levels with risks to the fetus and newborn — especially being large for gestational age — warrants a lower threshold for diagnosis of gestational diabetes and routine screening.1 In response, critics claim the evidence of benefit from intervention for newly diagnosed women and their offspring is far from robust, and the magnitude of any benefit is modest at best. They fear that mass over-medicalisation may emerge, producing little benefit, large cost and significant harm.2,3

One of those who believe the new definition has “gone too far” is Canadian Edmond Ryan, a University of Alberta endocrinologist who was on the planning committee for the NIH conference, where he will also be presenting. Ryan has expressed concerns about the poor reproducibility of the oral glucose tolerance test proposed for routine screening, which could produce a high rate of false-positive results.2 He has also questioned how predictive maternal blood glucose level is of adverse outcomes,2 arguing the mother’s weight is at least as important.4 He worries that many women may be labelled and monitored unnecessarily, with unneeded interventions, as well as potentially negative consequences for their insurance status. “There [are] no randomised controlled data showing the benefits of treating people diagnosed with these new criteria”, he told me. While there is evidence from trials showing some benefit from diet or lifestyle modification for women with mild gestational diabetes, those trials did not use the expanded diagnostic criteria of the new approach.1

Australian Jeremy Oats, Honorary Professorial Fellow at the University of Melbourne, is Secretary-General of the International Association of Diabetes and Pregnancy Study Groups, which oversaw the redefinition. He told me that the group spent a lot of time “looking at all available quality evidence” and that there is “significant preventable morbidity” the new approach could minimise.

Whatever the disagreement over the new criteria, there is no doubt they will mean more diagnoses, with one study estimating they would apply to one in eight pregnant women in the Wollongong area.5 That study’s lead author was the director of the Illawarra Diabetes Service, Robert Moses, who is also an associate editor of Diabetes Care and well aware of the controversy. While he acknowledges its “potential for some harm from the over-medicalisation of pregnancy” because of increased obstetric and paediatric interventions, he told me that more widespread dietary advice was a positive and that “the chance of international consensus is too valuable to miss”. Other authors have stated candidly that although the new definition means “thousands of women will be falsely identified as at-risk”, it also provides a “unique opportunity” to advocate behaviour modification.6

Nearly 20 years ago, an article in the BMJ asked whether gestational diabetes was a “non-entity”, and whether screening pregnant women might do more harm than good.7 University of Auckland diabetes specialist Tim Cundy, who will also present at the NIH conference, argues those questions are still valid today. He believes many of the problems ascribed to mild gestational diabetes actually relate to maternal obesity, which enhanced public health strategies should target. “If we flood the clinics with women with relatively trivial elevations in blood sugar who don’t really need to be there, as there is little convincing evidence that they benefit from treatment, then we’re diluting the resources for women with type 1 and type 2 diabetes who really do need intensive specialist treatment”, he told me.

Whatever the decision of the NIH adjudicating panel, this debate raises wider questions about defining disease. Should being “at risk” really be a condition in its own right? Should panels defining diseases more broadly consider social and environmental determinants of health? And should we all be routinely informed about the arbitrary, uncertain and controversial nature of labels like gestational diabetes, before they are offered to us?