TO THE EDITOR: A 32-year-old white woman 24 weeks into her uncomplicated first pregnancy had a fasting blood glucose level of 77 mg per dL (4.3 mmol per L) that increased to 142 mg per dL (7.9 mmol per L) one hour after a 50-g glucose loading test, which was interpreted as a positive screening result for gestational diabetes mellitus. She declined further testing and was told by her physician that she was putting herself and her baby at unnecessary risk. She wanted to know the evidence basis for this statement.

Two randomized controlled trials, both with approximately 500 pregnant women in each arm, compared intensive treatment for gestational diabetes with usual care.^1,2 Interventions included dietary counseling and insulin therapy for those with persistently high blood glucose levels. Women in the intervention groups gained less weight (difference of approximately 2 kg [4.4 lb]) and had less preeclampsia, fewer newborns with birth weights greater than 4 kg (8 lb, 13 oz), and a lower risk of shoulder dystocia compared with those in the control groups. Based on these trials, the U.S. Preventive Services Task Force recently recommended routine screening for gestational diabetes after 24 weeks' gestation.³ However, the results of these trials may not be generalizable to this patient because one study included women with fasting blood glucose levels up to 140 mg per dL (7.8 mmol per L),¹ and the other recruited mostly nonwhite participants.²

A multinational study of more than 20,000 women found that the fasting glucose level is as accurate as one- or two-hour levels after a 75-g glucose load in predicting maternal and infant morbidity, and that the risks increase linearly with glucose levels.⁴ Based on this study, the patient's fasting glucose level put her at a slightly increased risk of birth injuries (1.3% vs. 1.1%) and of preeclampsia (4.8% vs. 4.0%). Others have proposed a fasting blood glucose threshold of 85 mg per dL (4.7 mmol per L) to identify women who do not have gestational diabetes.⁵

In light of this information, the patient decided against further testing or treatment, and at 41 weeks' gestation delivered a healthy girl weighing 3.3 kg (7 lb, 4 oz). Her physician's approach was inconsistent with the evidence on the benefits of screening for gestational diabetes, and likely caused her unnecessary anxiety. We recommend a personalized approach to such screening, taking into consideration patient preferences and being careful not to exaggerate potential benefits or minimize potential harms.