Associations between pregnancy glucose measurements and risk of preterm birth: a retrospective cohort study of commercially insured women in the United States from 2003 to 2021

Abstract

To investigate associations between glucose measurements during pregnancy and risk of preterm birth (PTB). Retrospective cohort study of commercially insured women with singleton live births in the United States from 2003 to 2021 using longitudinal medical claims, socioeconomic data, and eight glucose results from different fasting and post-load tests performed between 24 and 28 weeks of gestation for gestational diabetes screening. Risk ratios of PTB (<37 weeks) were estimated via Poisson regression for z-standardized glucose measures. Non-linear relationships for continuous glucose measures were examined via generalized additive models. Elevations in all eight glucose measures were associated with increased risk (adjusted risk ratio point estimates: 1.05-1.19) of PTB for 196,377 women with non-fasting 50-gglucose challenge test (one glucose result), 31,522 women with complete 100-g, 3-hour fasting oral glucose tolerance test (OGTT) results (four glucose results), and 10,978 women with complete 75-g, 2-hour fasting OGTT results (three glucose results). Associations were consistent after adjusting for and stratifying by sociodemographic and clinical factors. Substantial non-linear relationships (U-, J-, and S-shaped) were observed between several glucose measurements and PTB. Elevations in various glucose measures were linearly and non-linearly associated with increased PTB risk, even before diagnostic thresholds for gestational diabetes.