Diagnosis of Gestational Diabetes Mellitus: How Should We Measure Glucose?

Gestational diabetes mellitus (GDM) is associated with adverse health outcomes for mother and child, the incidence of which can be reduced by treatment (1). GDM diagnosis is based on an oral glucose tolerance test (OGTT) requiring accurate glucose measurements, as small errors can have a major impact on assessment of GDM prevalence (2). A major source of preanalytical error in measuring glucose is in vitro glycolysis, which can result in significant reduction of glucose concentration (5–7% per hour) (3). Guidelines recommend immediate centrifugation or the use of ice water slurry preanalytical storage containers to minimize glycolysis, but most laboratories use blood collection tubes containing antiglycolytic agents instead for practical reasons. Sodium fluoride (NaF), which is widely used, allows continued glycolysis for up to 4 h, lowering the measured glucose (4). Conversely, acidification of the sample by addition of citrate buffer provides immediate inhibition of glycolysis but is not widely used.

The current World Health Organization 2013 diagnostic criteria are based on the findings of the Hyperglycemia and Adverse Pregnancy Outcome (HAPO) study, where NaF tubes were used, with immediate placement of tubes in ice slurry storage and centrifugation shortly after each sample collection in many centers, or after collection of all samples if early centrifugation was not possible. The aim of this study was to compare glucose concentrations obtained at several OGTT time points in pregnant women, with use of four different types of collection tubes.

Here we conducted a substudy of 65 pregnant women referred for early (<20 weeks’ gestation) and late (24–28 weeks’ gestation) GDM screening using the 75-g 2-h OGTT at the Treatment of Booking Gestational Diabetes Mellitus (TOBOGM) (5) study site in Sweden, contributing to 294 samples (98 OGTTs on three time points). The substudy was approved by the Swedish Ethical Review Authority (Dnr 2020-05537 and 2021-00727). Blood samples were collected into tubes in a prespecified order: serum-separating tube (SST), lithium heparin, NaF-EDTA-citrate (granulate) (referred to as citrate tubes hereafter), and NaF. Tubes were immediately put on an ice block with spaces completely surrounding the tubes, centrifuged for 7 min at 2,400g at 4°C within 30 min (most within 15 min), and analyzed within 60 min (most within 30 min) from sample collection with use of a glucose hexokinase method (ADVIA XPT; Siemens Healthineers). There were missing data due to handling and laboratory issues; there were 282 valid samples for citrate, NaF, and lithium and 276 for SST. Statistical analysis was by a linear mixed model with tube type as a fixed factor and time (0, 60, 120 min) and early/late OGTT nested in patients as random factors.

With use of NaF as reference, only the citrate had a significantly different mean glucose concentration (higher by 0.08 mmol/L [1.44 mg/dL]; P < 0.001). SST (as used in TOBOGM) and NaF tubes had comparable mean glucose levels (0.02 mmol/L [0.36 mg/dL]; P = 0.13) (Table 1). The citrate tube bias, in millimoles per liter, was similar for all OGTT time point samples (Table 1). GDM prevalence (World Health Organization 2013 criteria) as assessed with citrate was 54.0% and with fluoride 44.9%.

The key findings are that with use of ice and centrifugation within 30 min and analysis within 60 min of sampling, the citrate tubes had a positive bias of almost 0.1 mmol/L (1.8 mg/dL), with no significant differences in comparisons between the SST, lithium, and NaF tubes. GDM prevalence increased by 9.1% with use of citrate in assessment for women at high risk. A potentially important difference between the protocol used here and that of HAPO is that some HAPO samples were kept in ice slurry until the end of the OGTT before centrifugation. Thus, use of citrate tubes, with less preanalytical loss, will result in findings of a higher prevalence of GDM in comparison with using NaF tubes following the HAPO protocol.

Unanswered here, and requiring more research, is the question of the effect of delayed centrifugation of NaF tubes that have been placed on ice, as used in some HAPO centers. As the intention is to use citrate tubes without ice, the effects of ice versus no ice with citrate tubes need to be determined. Our findings demonstrate that use of citrate tubes processed on ice with rapid centrifugation results in higher glucose values in comparison with use of NaF tubes processed in the same manner and will result in a higher assessed prevalence of GDM.

Limitations of this study are that included women were from a single site in Sweden and the protocol used was not identical to that of HAPO. We recommend an international, multicenter study comparing use of citrate tubes at room temperature and NaF tubes with processing identical to that used in HAPO. This should enable appropriate adjustment of OGTT criteria for use of citrate tubes in the diagnosis of GDM.