Diagnosis of gestational diabetes mellitus: it is time for international consensus.

Gestational diabetes mellitus (GDM) has been defined as any degree of glucose intolerance with an onset or first recognition during pregnancy (1–3 ). Fetal complications of GDM include macrosomia (large baby, which leads to birth injuries), shoulder dystocia, and neonatal hypoglycemia, and adverse outcomes for the mother are an increased risk of cesarean delivery, preeclampsia, and hypertension during pregnancy, as well as a significantly higher risk of subsequent type 2 diabetes. Treating women with GDM reduces at least some of the adverse outcomes (4 ). Estimates of the prevalence of GDM range from 1% to 28%, reflecting the different diagnostic criteria and populations studied. The increasing prevalence of GDM in the US is congruent with the marked increase in obesity and type 2 diabetes. Pregnant women have been evaluated for diabetes for more than 50 years. Notwithstanding the 5 international workshops devoted to GDM that have been held between 1979 and 2005, there is lack of agreement concerning the optimal method to identify “any degree of glucose intolerance.” The criteria for both screening and diagnosing GDM vary considerably among countries and often between diabetes and obstetric organizations in a single country (5 ). Screening recommendations range from none (do not screen) to selective (screen only those at high risk) to universal. Approaches for screening tests include fasting glucose, random glucose, or, more commonly, a glucose challenge in which the patient ingests 50 g glucose (regardless of the time of the last meal). If the 1-h postload plasma glucose concentration exceeds the threshold, a full diagnostic oral glucose tolerance test (OGTT) is performed; however, 2 thresholds [140 mg/dL (7.8 mmol/L) or 130 mg/dL (7.2 mmol/L)] are generally used. Approximately 15% of women have concentrations that exceed the higher cutoff, which identifies approximately 80% of women with GDM, whereas the use of the lower cutoff increases the sensitivity to 90% but includes approximately 25% of pregnant women. Diagnosis is made with an OGTT, but, again, the criteria vary. The O’Sullivan and Mahan criteria (Table 1) form the basis for the majority of diagnostic approaches. Proposed in 1964, these criteria were derived from 752 unselected, asymptomatic pregnant women who underwent a 3-h OGTT (6 ). GDM was defined as 2 or more values 2 SDs above the mean. Although these arbitrary criteria were established to predict the subsequent (postpartum) development of diabetes (not necessarily to identify pregnancies with an unfavorable outcome), they became widely adopted to diagnose GDM. The cutoffs have been modified to compensate for changes in glucose measurement. In the O’Sullivan and Mahan study, glucose was measured in whole blood with the Somogyi–Nelson method. The National Diabetes Data Group (NDDG) (7 ) advised in 1979 that plasma be the preferred sample for glucose analysis. The NDDG raised the cutoffs for GDM (Table 1), because glucose concentrations in plasma are approximately 11% higher than in whole blood (owing to the higher concentration of water in plasma). Another concern is that the Somogyi–Nelson method, which quantifies glucose according to its ability to reduce copper, is not specific for glucose. Therefore, Carpenter and Coustan (8 ) lowered the cutoffs (Table 1) to reflect the more specific enzymatic assays that replaced the colorimetric assays clinical laboratories had previously used. To convert the O’Sullivan and Mahan values, Carpenter and Coustan used the formula: Plasma glucose concentration (whole-blood glucose concentration 5 mg/dL) 1.14. In contrast to the 2-step procedures outlined above, the WHO recommends that GDM be diagnosed via a 1-step procedure that uses the same OGTT performed to diagnose diabetes in nonpregnant individuals, i.e., 75 g of glucose, with only the fasting and 2-h samples analyzed (Table 1). Thus, although the OGTT is universally used for diagnosis, there is no accord. The main issues in dispute are a glucose load of 100 g vs. 75 g, a duration of 2 h vs. 3 h, the cutoffs, and whether 1 or 2 high values are necessary. A notable flaw in the GDM diagnostic criteria is that they have been based on the risk of future hyper1 Department of Laboratory Medicine, National Institutes of Health, Bethesda, MD. * Address correspondence to this author at: Department of Laboratory Medicine, Bldg. 10, Rm. 2C306, 10 Center Dr., Bethesda, MD 20892. Fax 301-402-1885; e-mail [email protected]. Received August 9, 2013; accepted August 29, 2013. Previously published online at DOI: 10.1373/clinchem.2013.206920 2 Nonstandard abbreviations: GDM, gestational diabetes mellitus; OGTT, oral glucose tolerance test; NDDG, National Diabetes Data Group; HAPO, Hyperglycemia and Adverse Pregnancy Outcome; IADPSG, International Association of Diabetes and Pregnancy Study Groups. Clinical Chemistry 60:1 141–143 (2014) Opinions