This study aimed to investigate the relationship between gestational diabetes mellitus (GDM) screening methods and GDM incidences. In 2018, a national questionnaire was administered at 231 institutions (56.6%) of all 408 perinatal medical centers in Japan. Of 100,485 women, 2,982 (3.0%) were diagnosed with GDM during their first pregnancy period (FPP) and 7,289 (7.3%) were diagnosed with GDM during their middle pregnancy period (MPP). The proportion of women diagnosed with GDM during FPP and MPP using 95 mg/dL as the cutoff value (CV) for random plasma glucose (PG) at FPP (4.3% and 9.2%) was significantly higher than that of women diagnosed with GDM using 100 mg/dL as the CV for random PG (2.7% and 6.9%, p < 0.0001, respectively). Compared with women screened for GDM using "random PG and random PG," women who were screened for GDM using "random PG and 50-g glucose challenge test (GCT)" had a significantly higher incidence of GDM (6.6% versus 8.9%, p < 0.0001). Using random PG and 50-g GCT, the incidence of GDM among women diagnosed at MPP using a CV of 95 mg/dL at FPP was significantly higher than that of women diagnosed using a CV of 100 mg/dL (16.5% versus 7.8%: p < 0.0001). While, using "random PG and random PG," the incidences of GDM among women were similar between institutions using a CV of 100 mg/dL and those using a CV of 95 mg/dL at FPP (6.7% versus 6.9%: p = 0.3581). This study showed random PG as a first-step screening method in MPP may overlook women with GDM.
Keywords: Cutoff value; Gestational diabetes mellitus; Glucose challenge test; Random plasma glucose level; Two-step method.