The sensitivity and specificity of screening tests are crucial to adequately detect, and thus efficiently prevent and treat, gestational diabetes mellitus (GDM). The fascinating report by van Leeuwen et al.1 was a meta-analysis on the 50-g glucose challenge test (GCT), spanning 26 papers. The oral glucose tolerance test (OGTT), following the ingestion of 75 or 100 g of glucose, was used as the reference/diagnostic test for the meta-analysis, which found similar results among studies including women with risk factors and studies with consecutive recruitment: a pooled sensitivity of 0.74 and 0.74, and a pooled specificity of 0.77 and 0.85, respectively. Interestingly, GCT displayed only modest accuracy for GDM screening, with a positive likelihood ratio (LR) of 3.2 and 4.9, and negative LR of 0.34 and 0.32, respectively. As expected, a higher threshold for a positive OGTT increased the sensitivity and decreased the specificity of the GCT.

Despite its relatively weak sensitivity and specificity values, the authors concluded that GCT is nevertheless acceptable for screening, but can’t replace the OGTT. This conclusion is in agreement with the recent guidelines of the International Association of Diabetes and Pregnancy Study Groups (IADPSG) and the American Diabetes Association (ADA). Both associations recommend OGTT as a one-step screening and diagnosis combination. This limits the scope of the meta-analysis reported by van Leeuwen et al., but the degree of adherence to IADPSG/ADA guidelines has yet to be established.

Other limitations are linked to the inherent drawbacks of GCT and OGTT, relating mainly to the length of time required to process blood samples in order to measure blood glucose (BG), as well as the actual time of day when women are tested. It is widely recognised that the nature of a sample (serum/plasma), its storage temperature, the time elapsed between sampling and centrifugation (in-tube glycolysis), and the use of a glycolysis inhibitor are all factors that can alter BG values. The impact of these factors was recently addressed in a study of 60 pregnant women. When tubes were processed systematically and centrifugation performed 5 minutes after blood collection, plasma BG values increased by 10 mg/l, uncovering fifteen cases of diabetes that were previously overlooked when the regular routine was used.2

Another key issue for screening accuracy is linked to the actual time of day when testing is performed. The literature supports a circadian variation of BG in pregnant women, with higher values for GCT and OGTT performed in the afternoon, observed in studies performed in relatively small cohorts.3,4 These all but forgotten results suggested that GCT cut-off values should be adjusted for morning versus afternoon testing. Unfortunately, no further recommendation has ever been issued regarding the best time of day for testing.

The bottom line is that GDM may all too often be overlooked for reasons easily circumvented, and relating mainly to sample handling and processing, for both GCT and OGTT alike. Accordingly, screening and diagnostic tests are lacking in sensitivity and specificity. Therefore, adherence to good laboratory practices should be strictly ensured and the issue of time of day for screening should be resolved.


  1. Top of page
  2. References
  • 1
    van Leeuwen M, Louwerse M, Opmeer B, Limpens J, Serlie M, Reitsma J, et al. Glucose challenge test for detecting gestational diabetes mellitus: a systematic review. BJOG 2012;119:393401.
  • 2
    Festa R, Carta M, Ceriello A, Testa R. Time is glucose, can’t miss gestational diabetes. Diabetes Technol Ther 2012;14:4446.
  • 3
    Galimberti D, Gernandez C, Joao M, Aparicio N, Alvarez T. Circadian variation of blood glucose response to an oral glucose load in pregnancy. Int J Gynaecol Obstet 1994;46:131.
  • 4
    Aparicio NJ, Joao MA, Cortelezzi M, Guz M, Sturgeon C, Galimberti DM, et al. Pregnant women with impaired tolerance to an oral glucose load in the afternoon: evidence suggesting that they behave metabolically as patients with gestational diabetes. Am J Obstet Gynecol 1998;178:105966.