The relationship of smoking with gestational diabetes: a large population-based study and a matched comparison

Abstract

Research questionIs smoking associated with an increased risk for gestational diabetes (GDM)? DesignA retrospective population-based study using data from the United States Healthcare Cost and Utilization Project–Nationwide Inpatient Sample. A dataset of all deliveries that took place between 2004 and 2014 (inclusive) was created. The population was divided into two groups according to smoking status and matched 1:1 for age, obesity, race, insurance type, thyroid disease, drug use and chronic hypertension. Delivery outcomes were compared between the two matched groups and the control group of non-smokers as a whole. A multivariate logistic analysis was performed adjusting for any statistically significant confounding baseline effects not matched for. ResultsThe study identified 9,096,788 births between 2004 and 2014, of which 443,590 (4.9%) were maternal smokers. When compared with the entire group of non-smokers as controls, there was an increased risk of GDM, adjusted odd ratio (aOR) 1.10 (95% confidence interval [CI] 1.07–1.14, P < 0.001). This relationship was of a mild nature and there were many confounding variables that differed between the groups. As such a matched control group was developed to compare the risk of GDM in the smokers and non-smokers. An increased risk for GDM among smokers remained, with an aOR of 1.07 (95% CI 1.05–1.10, P < 0.001). ConclusionsThe current data suggest that smoking is a small but significant risk factor for GDM with an aOR of 1.07.