Association of Historical Redlining with Gestational Diabetes Mellitus: The Mediating Role of Body Mass Index and Area Deprivation Index
Objective
We investigated the association between historic redlining and risk of gestational diabetes (GDM), and if this relationship is mediated by maternal obesity and area-level deprivation.
Research design and methods
This retrospective study included 86,834 singleton pregnancies from Kaiser Permanente Southern California’s (KPSC) health records (2008–2018). Redlining was assessed using digitized Home Owners' Loan Corporation (HOLC) maps, with patient’s residential addresses geocoded and assigned HOLC grades (A,B,C,D) based on their geographic location within HOLC-graded zones. For GDM cases, exposure was assigned based on address at diagnosis date; for non-cases, it was assigned based on address during the 24th–28th gestational week. Health records were combined with area deprivation index (ADI) from 2011–2015 census data. Mixed-effect logistic regression models assessed associations between redlining and GDM, with mediation by BMI and ADI evaluated using inverse odds ratio weighting. Models were adjusted for maternal age, education, race and ethnicity, neighborhood level income, and smoking status.
Results
Among the 10,134 (11.67%) GDM cases, we found increased risk of GDM in B ("Still desirable," adjusted odds ratio [aOR] 1.20, 95% confidence interval [CI] 0.99-1.44), C ("Definitely declining," aOR 1.22, 95% CI 1.02-1.47), and D ("Hazardous, i.e., redlined," aOR 1.30, 95% CI 1.08-1.57) graded neighborhoods compared to the "Best" graded zone. Pre-pregnancy BMI and ADI mediated 44.2%, and 64.5% of the increased GDM risk among mothers in redlined areas.
Conclusions
