Harvard Study Links Higher Cancer Death Rates To Living Near Nuclear Plants

Millions of Americans live within a short drive of a nuclear power plant and never think twice about it. A major Harvard study suggests proximity may deserve closer scrutiny. Researchers found that counties closer to operational nuclear facilities across the United States had consistently higher cancer death rates at the county level than counties farther away, a pattern that held firm across nearly two decades of national data and persisted even after accounting for poverty, race, smoking, healthcare access, and other factors that typically drive cancer mortality.

The findings land at a particularly charged moment. Nuclear energy has surged back into political favor as a low-carbon answer to climate change, with bipartisan support in Congress and a growing number of proposed reactor projects. What gets less attention is the geographic reality: more than 90 commercial reactors operate across 28 states, and tens of millions of Americans live in their orbit. Depending on where a person happens to live, that proximity may be associated with differences in cancer mortality.

The study, published in Nature Communications by researchers at the Harvard T.H. Chan School of Public Health, cannot prove that nuclear plants cause cancer. But its scope, covering more than 3,000 counties and nearly 19 years of mortality records, makes it one of the most thorough U.S. assessments of this question to date.

A Map of Nuclear Risk Across the Country

Most earlier studies on nuclear plants and cancer drew a circle around a single facility and compared health outcomes inside and outside that boundary. The Harvard team went bigger. Rather than focusing on one plant in one region, they mapped cumulative nuclear exposure across the entire country, accounting for every operational reactor within roughly 125 miles of each county. Counties sitting near multiple plants, or very close to a single large one, were treated as more heavily exposed. Counties in more remote regions were treated as less so.

That approach revealed a striking geographic divide. The Midwest, Northeast, and parts of the Southeast, home to the densest concentrations of U.S. reactors, carried the heaviest estimated exposure. Large stretches of the West and Great Plains, where nuclear plants are scarce, showed far lower levels. It’s worth noting that 200 kilometers, roughly 125 miles, was the outer boundary of the model, and counties weren’t simply sorted as inside or outside that radius. Instead, proximity was weighted continuously, meaning a county right next to a plant counted far more heavily than one near the edge of that boundary. Cancer death records from the Centers for Disease Control and Prevention were then layered onto that map, covering adults 35 and older from 2000 through 2018. After adjusting for income, smoking, race, obesity, hospital access, and other variables, the proximity-cancer mortality relationship remained.

The Pattern Strengthens With Age

Across nearly every age group and both sexes, counties with higher nuclear proximity scored higher cancer death rates. The gap was smallest among adults in their late 30s and early 40s and widened steadily with age. The strongest link in the data appeared in females aged 55 to 64 and males aged 65 to 74. In terms of raw numbers, though, the greatest estimated burden of attributable deaths for females fell in the 65 to 74 age group, where more people are alive and dying of cancer to begin with. It’s a distinction worth keeping in mind: the age group with the sharpest statistical signal and the age group with the most deaths are not always the same.

Over the full study period, approximately 115,586 cancer deaths were estimated to be linked to nuclear plant proximity under the study’s modeling assumptions. The authors are explicit that this calculation assumes a causal relationship between proximity and cancer, which the study itself cannot confirm. That uncertainty is real and should temper how the number is read. Still, the scale is worth contextualizing. A separate study published in the journal Science estimated that coal plant emissions were linked to roughly 20,900 deaths per year from all causes. The nuclear figure, spread across 19 years and counting only cancer deaths, represents a meaningful fraction of that annual coal toll, though the two studies measured different outcomes and cannot be directly compared.

The age skew in the data fits what scientists already understand about radiation biology. Radiation-linked cancers tend to have long latency periods, meaning they can take a decade or more to develop after exposure. Long-term studies of atomic bomb survivors in Japan found that elevated cancer risks initially concentrated in leukemia, then expanded over decades to include cancers of the stomach, colon, lung, breast, and thyroid. That slow-building pattern may help explain why proximity effects in the Harvard data appeared most clearly in older adults.

Conflicting Research, New Scale

Researchers have debated the nuclear-cancer question for decades. Large national studies in the United States, United Kingdom, and Canada have generally found no clear link. But research from France, Spain, South Korea, and Germany has reported elevated rates near certain facilities. A German study found that children under five living within five kilometers of a nuclear plant had more than twice the expected rate of leukemia.

The Harvard team argues that many earlier studies were structurally limited, mostly because they focused on a single plant in a narrow area, which makes it harder to detect patterns that only emerge at scale. By building a continuous national proximity measure across all operational plants, the Harvard analysis was designed to capture what smaller studies couldn’t.

The Geography of Risk

The practical question the study raises is as much political as it is scientific. Some Americans live in states with no nuclear plants. Others live in the dense reactor corridors of the Midwest and Northeast, where multiple facilities may lie within a hundred miles in different directions. That disparity reflects decades of siting decisions made largely without systematic health surveillance at the national scale.

Researchers called for follow-up work to identify which specific cancer types are most affected by proximity, since all malignancies were grouped together in this analysis. They also called for investigation into the precise pathways, whether through air, water, or soil, by which low-level radioactive emissions might reach nearby populations over time.

“While current evidence remains insufficient to draw definitive causal conclusions regarding cancer risks among populations living near nuclear plants,” the researchers wrote, “our findings highlight an important area for future investigation.” As the United States weighs a larger role for nuclear power in its energy future, the health geography of who lives near these facilities, and whether that proximity may be associated with differences in cancer mortality, is a question worth taking seriously.

---

Disclaimer: This article is based on an observational, county-level study and does not establish that nuclear power plants cause cancer. It does not constitute medical advice. Readers with health concerns should consult a qualified medical professional.

---