Gestational diabetes in Manitoba during a twenty-year period.

PURPOSE This retrospective cohort study was designed to examine the prevalence and risk factors of gestational diabetes mellitus (GDM) in Manitoba. METHODS A total of 324,605 deliveries by 165,969 women were reported to Manitoba Health in the years 1985-2004. Data on maternal ages, delivery dates, GDM, self-declared First Nation (FN) status, rural or urban residence and previous GDM were collected for the study. Data were analyzed using multivariate logistic regression models. RESULTS The prevalence of GDM during the 20-year period was 2.9%, which was 2.3% in 1985-1989 and 3.7% in 1999-2004 (P < 0.01). The trend of increase in the prevalence of GDM continued after major modifications on the screening and diagnostic criteria for GDM in 1998. The prevalence of GDM in FN women was 3-times greater than that in non-FN women. Higher prevalence of GDM was detected in FN pregnant women living in rural areas compared to those in urban areas (P < 0.01), which was opposite for non-FN pregnant women living in rural and urban areas. The prevalence of GDM in pregnant women > or =35 yr was 2.3-fold higher than that in those < 35 yr (P < 0.01). The recurrent rate of GDM was 44.4%. Adjusted odds ratios of GDM for FN status, advanced age, a history of GDM and rural living were 2.2, 2.4, 25.1 and 0.8, respectively. CONCLUSIONS The prevalence of GDM is increased in Manitoba. FN status, advanced age and a history of GDM, but not rural living, are independent predictors for GDM.