Assessment of Differential Uranium Exposure and its Association with Hypertension and Elevated Blood Pressure in American Indian Communities in the Strong Heart Family Study

Abigail Onderwyzer Gold

Name: Abigail Onderwyzer Gold
School: Vagelos College of Physicians and Surgeons, Class of 2024
Mentor: Ana Navas-Acien, MD, PhD, MPH

View Presentation Poster

Abstract

Urinary uranium (UU) is present in higher concentrations in American Indian (AI) participants in the Strong Heart Study compared to the general US population, possibly due to groundwater contamination. Cardiovascular disease is the leading cause of death among AIs, of which hypertension is a risk factor. This study aimed to evaluate the association between uranium exposure and incident hypertension and elevated blood pressure levels among AIs in the Northern and Southern plains, and Southwest. We included 1,456 participants from the Strong Heart Family Study from three study centers with complete baseline visit data (1998-1999 or 2001–2003) and follow-up data (2001-2003 and/or 2006–2009), and with no diabetes or hypertension at baseline. We used generalized estimating equations with independence correlation structure conditional on family membership to estimate the association of uranium exposure with change in systolic blood pressure (SBP) and diastolic blood pressure using linear regressions, and with hypertension incidence using Poisson regression with robust variance. At follow-up, 17.4% of participants developed hypertension. Comparing the highest to lowest UU quartiles, the relative risk of incident hypertension was 1.44 (95% CI: 1.04, 1.99) and the mean difference (95% CI) in SBP at follow-up was -2.02 (-3.86, -0.18) mmHg. This second association was attenuated when other metals were added into the model (-1.49; -3.34, 0.35 mmHg). These findings suggest moderate increased risk for hypertension at chronic levels of uranium exposure typical of the Southwest and Great Plains, which is inconsistent with decreased SBP levels at follow-up. Disentangling this inconsistency requires additional research.