AS700 - The WHI Radon Risk Investigation (WHI Risk It?) Study

Investigator Names and Contact Information

Eric Whitsel (eric_whitsel@med.unc.edu)

Eric Whitsel

Introduction/Intent

Although stroke is a major contributor to disability and mortality worldwide, established stroke risk factors have not fully explained the observed burden of or spatial variation in its incidence. Indeed, nearly a third of strokes are cryptogenic, thereby compelling a search for environmental factors that may play an important, but to date, inadequately characterized role in stroke pathogenesis and epidemiology. One such factor—radon—is a naturally occurring, odorless gas generated via decay of radioactive elements in rocks and soils. The gas enters homes through basement and foundation cracks, making unrecognized exposures both ubiquitous and the second-leading, modifiable cause of lung cancer in the US, after smoking. Although radon’s role in lung carcinogenesis is well-known, relatively little is known about its role in the pathogenesis and epidemiology of cerebrovascular disease. Yet extant studies of the radon-stroke association suggest that radon exposure is associated both with pathologic precursors of stroke and incident stroke, itself. Their collective limitations nonetheless leave a substantial gap in the state of the science that could otherwise highlight putative means of reducing radon exposures and their common cerebrovascular sequelae in the US. To fill this gap, we will therefore estimate: [1] pre- and post-mitigation home radon exposures across the US (1986-2018), [2] the association between time-varying home radon exposure and incident stroke, [3] the association between time- varying home radon exposure and clonal hematopoiesis of indeterminate potential (CHIP), and [4] mediation of the radon-stroke association by CHIP. We will do so among a demographically and environmentally diverse group of 194,015 participants in three NHLBI-supported cohorts that have reviewed and approved this proposal. Quantitatively estimating radon exposures and their uncertainties at their accurately geocoded addresses will involve a national, spatiotemporal prediction and two-stage cross-validation based on time-varying, short- and long-term mean, pre- and post-mitigation home radon concentrations; home radon testing statistics / policies; and housing characteristics; plus, time-invariant aerial gammaradioactivity, geology, and soil features. Estimating radon-stroke and radon-CHIP associations also will involve cohort-specific, multivariable-adjusted, Bayesian, hierarchical, time-to-event and logistic regression accounting for exposure uncertainty and competing mortality, followed by inverse varianceweighted, random-effects meta-analysis. Finally, estimating mediation of the radon-stroke association by CHIP will involve quantifying direct, indirect, and total effects of radon on stroke. The previously unparalleled focus of such estimation on a modifiable risk factor, a plausible mechanism of disease, and a common, clinically and societally burdensome health outcome heretofore ignored by EPA and WHO uniform radon testing and mitigation policies will thereby position the study to help characterize the environmental epidemiology and somatic genomic pathogenesis of stroke and other health outcomes in ways that may lead to the identification of plausible targets for new public health screening and prevention efforts.