Influence of gestational diabetes on diabetes risk and glycemic control in a retrospective population-based cohort

Background. Racial/ethnic specific estimates of the influence of gestational diabetes mellitus on type 2 diabetes remain underexplored in large population-based cohorts. We estimated racial/ethnic differences in the influence of gestational diabetes on diabetes risk and glycemic control in a multiethnic population-based cohort of postpartum women.

Methods. Hospital discharge and vital registry data for New York City (NYC) births between 2009 and 2011 were linked with NYC A1C Registry data between 2009-2017. Those with baseline diabetes (n=2,810) were excluded for a final cohort of 336,276 births. Gestational diabetes on time to diabetes onset (two A1c tests of ≥ 6.5% from 12 weeks postpartum onward) or glucose control (first test of A1c<7.0% following diagnosis) was assessed using Cox regression with a time varying exposure. Models were adjusted for sociodemographic and clinical factors and stratified by race/ethnicity.     

Results. The cumulative incidence for diabetes was 11.8% and 0.6% among those with and without gestational diabetes. Gestational diabetes status on diabetes risk was aHR 11.5 (95% CI:10.8, 12.3), overall, with slight differences by race/ethnicity. Gestational diabetes was associated with lower likelihood of glycemic control (aHR: 0.85, 95% CI: 0.79 0.92), with the largest negative influence among Black (aHR: 0.77, 95% CI: 0.68, 0.88) and Hispanic (aHR: 0.84, 95% CI: 0.74, 0.95) individuals. Adjustment for screening bias and loss to follow-up modestly attenuated-racial/ethnic differences in diabetes risk but had little influence on glycemic control.

Conclusions. Understanding racial/ethnic differences in the influence of gestational diabetes on diabetes progression is critical to disrupt lifecourse cardiometabolic disparities.