Geographic Disparities in Hypertensive Disorders of Pregnancy: An Environment-Wide Association Study

Hypertensive disorders of pregnancy (HDP) are significant contributor to maternal and infant morbidity and mortality, and large geographic disparities exist. Although the importance of an individual’s surrounding environment as a potentially modifiable risk factor for HDP has been recognized and multiple environmental factors have been associated with HDP, only a few components in the environment have been assessed, often separately, with no consideration of the totality of environment. To address this, we conducted an environment-wide association study (EWAS) using the Florida Vital Statistics Birth Records including 2,119,324 women with conception dates between January 1, 2005 and December 31, 2014. EWAS is an agnostic and hypothesis-free approach. A wide range of environmental data from 12 sources were collected, harmonized, integrated, and spatiotemporally linked to the women based on their geocoded residential address and pregnancy period. A total of 5,510 factors characterizing women’s surrounding natural, built, and social environment during pregnancy (i.e. climate, air pollution, noise, greenness, walkability, food access, socioeconomics, social capital, housing, and safety) were included in the analysis. A random 50:50 split divided the data into training and testing sets. Environmental factors were individually examined after adjusting for age, race/ethnicity, education, marital status, WIC (Women, Infants, and Children Nutrition Program) recipient, smoking during pregnancy, pre-pregnancy BMI, parity, season, and year of conception. Mixed-effects models were used to account for the multilevel structure. A factor is deemed as “significant” if it has a false discovery rate-adjusted p-value (or q-value)<0.01 in the training set and a p-value<0.01 in the testing set. We identified 1,208 factors significantly associated with HDP, with strong associations observed between HDP and three categories of environmental factors: 1) neighborhood social capital (i.e. ZIP Code-level density of fitness and recreational sports centers, OR per standard deviation increase [OR sd ]: 0.34, 99% CI: 0.25, 0.46), 2) food access (i.e. low access tract at 10 miles, OR=1.27, 99% CI: 1.17, 1.47), and 3) bikeability (OR sd =0.86, 99% CI: 0.83, 0.89). EWAS is a promising approach to identify novel environmental factors associated with cardiovascular diseases. Environmental exposures appear to play a critical role, both in identifications of pregnant women at increased risk of HDP and in determinations of potential target for public health interventions.