Harvard Research Finds Cancer Mortality Tied to Nuclear Power Plant Proximity – ANS qualifies study as “deeply flawed”

A nationwide study led by researchers at the Harvard T.H. Chan School of Public Health has found a statistically significant association between proximity to operating nuclear power plants (NPPs) and higher cancer mortality rates in U.S. counties. Published in Nature Communications on February 23, 2026, the study analyzed data from 2000 to 2018, revealing that counties closer to NPPs experienced elevated cancer death rates even after adjusting for socioeconomic, environmental, and healthcare factors. The findings, while not establishing causation, underscore the need for further research into the health impacts of nuclear energy, particularly as it is increasingly promoted as a climate solution.

Methodology and Key Findings

The study employed a novel approach called continuous proximity, which measures the cumulative influence of multiple nearby NPPs rather than focusing on a single facility. Researchers used inverse-distance weighting to quantify how proximity to all operational NPPs—both in the U.S. and Canada—affected cancer mortality. County-level cancer mortality data was sourced from the Centers for Disease Control and Prevention (CDC), while NPP locations and operational histories were obtained from the U.S. Energy Information Administration (EIA).

The analysis adjusted for confounding variables including median household income, educational attainment, racial composition, smoking prevalence, body mass index (BMI), and access to healthcare. Even after these adjustments, the study estimated that approximately 115,000 cancer deaths (about 6,400 per year) nationwide were linked to proximity to NPPs. The association was strongest among individuals aged 65 and older, suggesting a potential dose-response relationship with age.

Contextualizing the Study

This research builds on earlier work by the same team, which found elevated cancer incidence in Massachusetts residents living near NPPs. The current study expands on that by applying a national-scale analysis, addressing previous limitations of localized studies that focused on single facilities. The continuous proximity method allows for a more nuanced understanding of how multiple NPPs might collectively influence health outcomes, rather than isolating the impact of any one plant.

Limitations and Cautionary Notes

The authors emphasized that the study does not prove a direct causal link between NPPs and cancer. Key limitations include the absence of direct radiation measurements and the assumption that all NPPs exert similar health impacts, which may not account for variations in plant design, safety protocols, or environmental conditions. Additionally, the study relies on observational data, meaning it cannot establish whether proximity to NPPs causes increased cancer mortality or if other unmeasured factors might explain the observed correlation.

Note: American Nuclear Society on a recent press release qualifies this study as deeply flawed and also states that “it proves nothing regarding cancer mortality and proximity to nuclear power plants”.
Full Press Release available below and here

Implications for Policy and Public Health

Senior author Petros Koutrakis, Akira Yamaguchi Professor of Environmental Health and Human Habitation, noted that the findings highlight the need for rigorous, long-term research into the health effects of nuclear energy. ‘Our study suggests that living near a NPP may carry a measurable cancer risk—one that lessens with distance,’ he said. ‘We recommend that more studies be done to address the issue of NPPs and health impacts, particularly as nuclear power is being promoted as part of the solution to climate change.’

The study’s results align with broader concerns about the environmental and health risks of nuclear energy. While nuclear power is often touted as a low-carbon energy source, its potential health impacts remain a subject of scientific debate. The findings add to a growing body of research suggesting that proximity to industrial facilities, including NPPs, may influence public health outcomes in ways that require further investigation.

American Nuclear Society responds to deeply flawed study on nuclear plant proximity

“Cannot establish causality”: Why the study published in Nature Communications proves nothing regarding cancer mortality and proximity to nuclear power plants
Thu, Feb 26, 2026, 9:39AM

WASHINGTON, D.C. — The American Nuclear Society (ANS), a nonprofit representing over 12,000 professionals in the fields of nuclear science and technology, issues the following response to “National Analysis of Cancer Mortality and Proximity to Nuclear Power Plants in the United States,” by Yazan Alwadi, Petros Koutrakis, et al., published February 23, 2026, in Nature Communications (doi: 10.1038/s41467-026-69285-4):

Experts with ANS, including health physicists and radiation protection specialists, have reviewed the study published in Nature Communications purporting to show associations between residential proximity to nuclear power plants and elevated cancer mortality rates across U.S. counties from 2000 to 2018. This study contains fundamental methodological shortcomings, acknowledged by the authors themselves, that prevent it from supporting any credible scientific conclusions.

The Bottom Line:

This flawed ecological study does not advance our understanding of radiological risk. The authors themselves state that their findings “cannot establish causality” and that their study “does not include dosimetry,” admissions that undermine the study’s central premise and that ANS urges journalists and policymakers to weigh carefully.

Core Scientific Issues:

1. The Authors’ Own Admissions

The paper explicitly acknowledges that its findings “cannot establish causality,” that it “does not include dosimetry,” and that its attributable deaths calculation “assumes a causal relationship between exposure and outcome,” an assumption the study’s own design cannot validate. These are not peripheral disclaimers; they go to the heart of whether this study can support the conclusions being drawn from it.

2. No Dose Assessment

The study includes no dosimetric measurements, environmental radiation monitoring data, dose modeling, or exposure pathway analysis. Even a basic calculation using publicly available plant emission data would demonstrate that radiation doses received by populations near U.S. nuclear plants are far too low to plausibly account for the cancer mortality burden the authors describe, making the absence of this foundational step disqualifying.

3. Proximity as an Invalid Proxy for Radiation Dose

The study’s sole exposure measure, an inverse-distance weighted proximity metric, is not a measure of radiation exposure and does not account for plant-specific emission profiles, wind patterns, topography, or water pathways. Using geographic proximity as a radiation dose proxy introduces severe exposure misclassification and cannot support the conclusions the study implies.

4. A Critical and Unaddressed Confounding Factor: Load-Serving Entity Territories

Nuclear power plants were deliberately sited near major population and industrial load centers, the high-demand service territories of load-serving entities where heavy manufacturing, fossil fuel combustion, and associated carcinogenic pollutants are most concentrated. No county-level statistical model can fully untangle those exposures from any signal attributed to a nuclear facility, representing a profound and unresolved confounding problem.

5. Ecological Study Design Cannot Support Causal Claims

This is an ecological study, widely recognized as the weakest epidemiological study design for establishing causation, and is appropriate only for generating hypotheses, not for estimating attributable deaths. Presenting county-level correlations as the basis for calculating over 115,000 “attributable cancer deaths” dramatically overstates what this study design can support.

6. Inadequate Confounding Control

While the authors include several county-level covariates, ecological analyses inherently cannot account for the full range of cancer risk factors present in the heavily industrialized, densely populated regions where nuclear plants are typically located. Smoking, occupational exposures, fossil fuel emissions, and socioeconomic factors, among others, remain inadequately controlled.

7. All-Cancer Outcomes Without Radiation Sensitivity Context

The study analyzes all malignant neoplasms combined rather than the radiation-sensitive cancers, such as leukemia and thyroid cancer, for which dose-response relationships are established. A credible study of this kind would demonstrate a biologically plausible dose-response relationship; this study does not.

The Broader Scientific Context:

Decades of rigorous research by the National Cancer Institute, the U.S. Nuclear Regulatory Commission, and international scientific bodies including the United Nations Scientific Committee on the Effects of Atomic Radiation have not found consistent evidence of elevated cancer risk near nuclear plants operating within regulatory limits. That is the science. This study does not change that nor add to it.

Conclusion:

This study does not meet the highest scientific standards, a fact the authors themselves explicitly acknowledge with the admission that “our findings cannot establish causality.” The absence of dose assessment, a fatally limited study design, unresolved industrial confounding, and the unwarranted attribution of over 115,000 cancer deaths to geographic correlation represent serious and compounding failures. ANS stands ready to assist reporters and stakeholders in contextualizing this research within the broader, well-established body of radiological health science.

About the American Nuclear Society (ANS):

Established in 1954, the American Nuclear Society (ANS) is an international professional organization of engineers, scientists, technologists, teachers, and healthcare providers devoted to the peaceful applications of nuclear science and technology. Its more than 12,000 members represent government, academia, research laboratories, medical facilities, and private industry. ANS’s mission is to advance, foster, and spur the development and application of nuclear science, engineering, and technology for the benefit of humanity.