Letters

Debate over screening for gestational diabetes

Screening should take place only in context of good quality controlled trials

Scientific uncertainty is mirrored in clinical practice in Italy

Evidence from randomised controlled trial is needed

Screening should take place only in context of good quality controlled trials

EDITORJarrett and Soares et al present arguments respectively against and in favour of screening for gestational diabetes.¹ Jarrett has been arguing objectively against gestational diabetes as a useful concept for many years.² In contrast, many obstetricians and diabetologists have aggressively sought and treated this condition despite a lack of real evidence. Rather, they have worked on the assumption that there is a continuum from normal glucose tolerance through gestational diabetes to frank diabetes and that gestational diabetes must be a less severe form of diabetes.

Screening for gestational diabetes presents a massive organisational and financial cost as well as a stressful burden to the women concerned. As Walkinshaw has shown, there is no evidence that attempts to control carbohydrate metabolism in women labelled as having gestational diabetes usefully alter perinatal outcome; rather, they lead to unnecessary interference.³ Had either Jarrett or Soares et al referred to Walkinshaw's systematic review then one of the three arguments of Soares et al for screening could have immediately been dismissed as lacking in supportive evidence. Their other two arguments are no more certain: that the presence of gestational diabetes predicts an increased risk of maturity onset diabetes in later life, and that management of carbohydrate metabolism in the mother may prevent the child from developing long term medical problems as a result of an adverse intrauterine environment.

Sufficient evidence probably exists to support the case that a degree of glucose intolerance predicts an increased risk of frank diabetes in later life. What is less clear is whether, as a result of subjects at increased risk being identified, any intervention can alter disease progression. Surely if there were such evidence, confining screening for such glucose intolerance solely to women who happen to be pregnant would be massively unfairwhat about men with an increased risk of maturity onset diabetes, and childless women?

The hypothesis regarding long term metabolic effects of an adverse (carbohydrate) intrauterine environment requires confirmation by prospective studies. In discussing this hypothesis Soares et al ignore the difference between frank diabetes and gestational diabetes once more.

Until the relevance of gestational diabetes is known, screening for it should take place only in the context of good quality controlled trials. In the meantime we should continue to ensure good preconceptional and antenatal control of frank diabetes and aim to detect the few women who have frank diabetes not diagnosed before pregnancy. The benefits of treating such women have been shown by many authors.