Antepartum glucose tolerance test results as predictors of type 2 diabetes mellitus in women with a history of gestational diabetes mellitus: A systematic review

Background: Women with a history of gestational diabetes mellitus (GDM) are at high risk for type 2 diabetes mellitus (T2DM).Objective: We reviewed prospective studies of antepartum glucose tolerance test results as risk factors for development of T2DM among women with a history of GDM.Methods: We searched 4 electronic databases and hand-searched 13 journals for literature published through January 2007. The search strategy consisted of medical subject headings and text words for GDM, T2DM, and other relevant terms. Articles were excluded for the following reasons: (1) not written in English; (2) no human data; (3) no original data; (4) <90% of sample was diagnosed with GDM without a separate analysis for women with GDM; (5) case report or series; (6) diagnosis of GDM not based on 3-hour 100-g oral glucose tolerance test (OGTT) or 2-hour 75-g OGTT; (7) T2DM not evaluated as outcome; (8) no relative measure of association or incidence reported; or (9) design did not address antepartum OGTT as a predictor of T2DM. Two investigators independently reviewed citations, performed serial data abstraction on full articles, and assessed the quality of each article. Data were abstracted for study participants and characteristics, T2DM diagnosis, length of follow-up, regression model covariates, and measures of association and variability.Results: Of 11,400 unique citations, we identified 11 articles that evaluated antepartum glucose testing and risk of T2DM in women with a history of GDM. Five studies found that the fasting blood glucose (FBG) on the antepartum diagnostic OGTT was a significant predictor of T2DM (odds ratio [OR] range: 11.1–21.0; relative risk [RR] range: 1.37–1.5; relative hazard [RH] = 2.47). Risk of incident T2DM was predicted by the antepartum 2-hour OGTT plasma glucose in 3 studies (OR range: 1.02–1.03; RR = 1.3) and by the antepartum OGTT glucose AUC in 3 other studies (OR range: 3.64–15; RH = 2.13). Overall, study quality was limited by high losses to follow-up (>20% in 6 studies) and short duration. Few studies adjusted for adiposity, an established diabetes risk factor.Conclusion: FBG, OGTT 2-hour blood glucose, and OGTT glucose AUC appeared to be strong and consistent predictors of subsequent T2DM among women who met diagnostic criteria for GDM using the OGTT.