Proximity t o Nuclear Power Plants and Cancer Risk: A Multi-Scale Epidemiological Investigation in the United States"

Abstract

Purpose This dissertation provides a comprehensive, first-level investigation of the associations between residential proximity to nuclear power plants and cancer outcomes in the United States. While nuclear energy is promoted as a low-carbon alternative to fossil fuels, longstanding concerns remain regarding potential health risks from chronic low-dose radiation exposures. Prior studies have been hampered by limited geographic scope, crude exposure definitions, and inconsistent analytic frameworks. This work addresses these gaps through three interlinked studies that systematically evaluate cancer mortality and incidence across multiple spatial scales, outcomes, and designs. Methods We developed a consistent spatial exposure metric based on inverse-distance proximity to operational nuclear power plants, updated annually to account for plant operations. Using national cancer mortality data (2000–2018) and Massachusetts cancer registry data (2000–2018), Dissertation Advisor: Petros Koutrakis Yazan Alwadi we applied generalized estimating equation Poisson regression models for longitudinal analyses and cross-sectional Poisson regression for site-specific incidence. Models incorporated demographic, socioeconomic, environmental, and healthcare access covariates, and accounted for spatial clustering. Attributable fractions (AFs) and counts were estimated to characterize the cancer burden. Results Proximity to nuclear power plants was consistently associated with increased cancer risk, with relative risks strongest among older adults. National analyses estimated over 115,000 cancer deaths attributable to nuclear plant proximity during 2000–2018, including 54,349 female deaths (AF: 1.5%, 95% CI: 0.7–2.3%) and 61,237 male deaths (AF: 1.6%, 95% CI: 0.7–2.5%). Cancer site-specific analyses indicated 33,915 attributable female deaths and 25,586 male deaths from breast, colon, and lung cancers, with lung cancer accounting for the largest share. In Massachusetts, proximity was associated with 10,815 attributable cancer cases among females (AF: 4.1%, 95% CI: 2.4–5.7%) and 9,803 among males (AF: 3.5%, 95% CI: 1.8–5.2%), with elevated risks across multiple cancers, including lung, prostate, breast, colorectal, bladder, melanoma, leukemia, thyroid, uterine, kidney, and pancreatic cancers. Risk declined with increasing distance, with the highest relative risks observed within 30 km of facilities. Conclusion This dissertation represents the most comprehensive evaluation to date of cancer outcomes in relation to nuclear power plant proximity in the United States. By systematically resolving key technical challenges that limited prior studies, it establishes a foundation for more refined investigations. The consistent proximity-based associations observed here underscore the importance of future research incorporating dosimetry, exposure-specific pathways, individual-level effect estimates, and cohort designs. As nuclear energy re-emerges in climate and energy policy debates, these findings highlight the need to integrate rigorous epidemiologic evidence into broader assessments of its long-term public health implications.