Trends in pregnancy outcomes for women with gestational diabetes mellitus in Sweden 1998-2012 : a nationwide cohort study

Aim: To assess whether incidence of maternal and neonatal outcomes for women with or without gestational diabetes mellitus (GDM) have changed over time. Methods: Population-based cohort study in Sweden including all singleton pregnancies over the period 1998–2012. GDM was diagnosed following Diabetic Pregnancy Study Group 1991 criteria. Poisson regression or negative binomial regression was used to model yearly relative change in numbers of cases and incidence of the outcomes with 95% confidence intervals (CI), and yearly absolute change in birthweight z-score. Results: The study included 1 455 667 pregnancies. The number of pregnancies increased over time and the overall prevalence of GDM was 1%. For women with GDM there was a significantly decreasing trend in incidence per year for large for gestational age (LGA) (0.986, 95% CI 0.975 to 0.996), birthweight z-score (−0.012, 95% CI −0.017 to −0.007) and birth trauma (0.937, 95% CI 0.907 to 0.968). The trend for small for gestational age (SGA) among women with GDM increased by an odds ratio per year (1.016, 95% CI 1.002 to 1.029). No significant interaction tests for maternal characteristics were found. Trends in outcomes for women without diabetes were similar to those for women with GDM. Conclusions: This study shows that there were improvements in pregnancy outcomes for women with GDM between 1998 and 2012, although the incidence of SGA increased. Improvements followed similar trends in the background population. Inequalities in obstetric outcomes between women with GDM and those without have continued unchanged over 15 years, suggesting that new management strategies are required to reduce this gap.