Oral Presentation Australasian Diabetes in Pregnancy Society Annual Scientific Meeting 2016

The impact of ethnicity on predictive biomarkers for gestational diabetes mellitus (#21)

Arianne N Sweeting 1 2 , Glynis P Ross 1 2 , Jencia Wong 1 2 , Paul Williams 1 2 3 , Heidi Appelblom 4 , Heikki Kouru 4 , Mikko Sairanen 4 , Jonathan Hyett 1 2
  1. Royal Prince Alfred Hospital, Camperdown, NSW, Australia
  2. Sydney Medical School, University of Sydney, Sydney
  3. Department of Clinical Biochemistry, Royal Prince Alfred Hospital, Sydney, Australia
  4. Diagnostics, Perkin Elmer, Turku, Finland

Introduction: Biomarker-predictors may improve the early detection of gestational diabetes (GDM) when combined with maternal clinical risk factors1, but the degree to which their predictive utility is modified by ethnicity is unknown.

Aim: To assess the performance of several potential GDM biomarker-predictors in a large multi-ethnic Australian population and determine whether GDM can be predicted in the first trimester.

Methods: Routine biomarkers (PAPP-A, free β-HCG) were measured in serum prospectively at 11-13+6 weeks’ gestation in 224 women who developed GDM (ADIPS 1998 diagnostic criteria2) and 718 controls (n=942), undertaken at Royal Prince Alfred Hospital, Sydney. Novel biomarkers (adiponectin, leptin, PAI-2, lipocalin-2, triglycerides) were measured on retrieved samples. The relationship between biomarker-multiples-of-the-median (MoM) and GDM (1) overall and (2) stratified by ethnicity (Caucasian, East and South Asian) was examined with logistic regression. A multivariate GDM-prediction model was developed and evaluated using areas under the receiver-operating characteristic (AUROC) curve.

Results: Overall, PAPP-A- and adiponectin-MoM values were lower, and triglyceride-, leptin- and lipocalin-2-MoM values were higher, in women with GDM versus controls, respectively. Leptin- and lipocalin-2-MoM values were highest in Caucasians with GDM at 1.18, 1.25, respectively, compared to 1.04, 1.07 (East Asian-GDM) and 1.07, 1.04 (South Asian-GDM). PAPP-A-MoM was lowest in South and East Asians with GDM, at 0.69 and 0.81 respectively, compared to 0.88 (Caucasian-GDM). In contrast, differences in adiponectin- and triglyceride-MoM values between GDM and controls were consistent across ethnicity. The best performing GDM-prediction model overall combined clinical factors (age, BMI, ethnicity, mean arterial pressure), PAPP-A, triglycerides and lipocalin-2, achieving an AUROC of 0.93.

Conclusion: GDM can be accurately predicted in early pregnancy by combining novel and routine biomarkers to maternal clinical parameters. Biomarker performance varied by ethnicity, suggesting underlying differences in pathophysiology. The performance and validation of GDM-prediction models using biomarker-predictors will be influenced by the ethnic distribution of the local population.

  1. Savvidou et al. First-trimester prediction of gestational diabetes mellitus: examining the potential of combining maternal characteristics and laboratory measures. Diabetes. 2010;59: 3017-3022.
  2. Hoffman et al. Gestational diabetes mellitus – management guidelines. Med J Aust 1998;169:93–7.