TL;DR
Get oils, diffusers and self-care delivered free — and shop member deals
- Fast, free delivery on millions of items
- Access to Prime Big Deal Days deals on October 6–7
- Prime Video, Amazon Music and more included
A commentary in The Ecomodernist says a method used in Harvard-led studies of nuclear power plants and cancer also produces apparent cancer associations for unrelated landmarks, including Costco. The comparison challenges the method’s interpretation, but does not by itself establish that the underlying studies are invalid or that proximity has no health effect.
The Ecomodernist says a statistical method used in recent studies linking proximity to nuclear power plants with cancer outcomes can also generate apparent associations for unrelated landmarks, including Costco and private colleges. The critique challenges how the studies’ results should be interpreted, but the reported associations do not establish that those landmarks—or nuclear plants—cause cancer.
The commentary says researchers led by Yazan Alwadi at Harvard T.H. Chan School of Public Health published a Massachusetts study of cancer incidence in December 2025, followed by a nationwide study of cancer mortality in March, according to the chronology in the source. It also cites later papers examining selected cancer types and total mortality. The authors of those studies described associations with proximity, not proof of cause and effect, the commentary says.
The studies assign a proximity score to ZIP codes or counties within set distances of nuclear plants, with closer areas weighted more heavily and scores accumulating where areas are near multiple plants. The commentary says the researchers then use statistical models and other variables to estimate relative risks and calculate deaths attributed to proximity. Its authors report applying a similar approach to other landmarks and obtaining large estimated associations, including a claim that more than 2.2 million cancer deaths could be attributed to Costco over 2000–2018. That is the commentators’ model-derived result, not a confirmed count of deaths caused by Costco.
The critique’s central point is that geographic proximity is not a direct measure of radiation exposure. Its authors also say an attempt to reproduce the published results using code and answers provided by the study authors did not match, and that they reconstructed much of the analysis themselves. The source does not provide enough detail to independently assess that replication or the alternative landmark calculations.
Why the Landmark Comparison Matters
The comparison illustrates a risk in reading population-level associations as evidence of a specific hazard. If a method produces striking estimates for unrelated places, readers and researchers need to examine whether the result reflects the exposure of interest, other differences between communities, model choices, or chance. The commentary argues that the landmark results weaken confidence in the interpretation of the nuclear-plant studies.
That argument is not, on its own, a definitive rebuttal of the Harvard-led research. A deliberately chosen comparison may expose limitations in a model, but it does not show exactly which assumptions are responsible or settle the question of whether people living near plants face any additional risk. For the public, the distinction matters: an association is not a causal finding, and a model’s estimate is not a direct tally of deaths verified as caused by a site.
As an affiliate, we earn on qualifying purchases.
How the Proximity Studies Work
As described by The Ecomodernist, the Massachusetts analysis included areas within 120 kilometres of plants, while national studies used radii of up to 200 kilometres. The scoring approach gives areas nearer a site a greater weight. It also combines contributions from multiple plants, so a location’s score may reflect several sites rather than distance from only the nearest one.
The studies use those scores in regressions with additional variables to estimate associations with cancer incidence, cancer mortality, or total mortality. The commentary says the results were used to calculate relative risks and attributable fractions for the period 2000 to 2018. It argues that measured exposure data would be more informative than distance alone and points to monitored occupational radiation exposure as a comparison. Its discussion of worker doses is not a direct measurement of exposure among nearby residents, however, and does not by itself answer what residents received.
“Proximity is not exposure.”
— The Ecomodernist commentary
What the Critique Does Not Establish
The source is an argumentative commentary, not an independent review of the studies. It does not provide the full analysis code, all model specifications, or enough detail to verify its reported landmark estimates. The basis for the Costco death estimate and other comparisons therefore remains unconfirmed here, including how sensitive they are to choices of landmarks, covariates, and geographic boundaries.
It is also unclear from the material supplied whether the researchers have responded to the replication claims or to the landmark comparisons. The studies’ reported associations do not demonstrate that nuclear plants caused cancer, while the critique’s examples do not establish that nearby residents received no exposure or faced no risk. Those questions require evidence about exposure and careful evaluation of the competing analyses.
Further Checks on the Models
The next step is independent scrutiny of both the original studies and the commentary’s replication, including publication of reproducible code, clear documentation of model choices, and comparisons with direct exposure measurements where available. Researchers’ responses to the reported replication gap could clarify whether it resulted from coding, data, or differences in analytical decisions.
Until that work is available, the reported numbers should be treated as model-based associations, not counts of deaths proven to have been caused by a nuclear plant or another landmark. The material supplied does not identify a scheduled review, correction, or response from the study authors.
Key Questions
Did the Harvard-led studies prove that nuclear plants cause cancer?
No. As described in the source, the studies reported associations with proximity and did not establish cause and effect.
Does the Costco estimate mean Costco caused millions of cancer deaths?
No. The figure is a model-derived estimate reported by the commentary, not a verified count of deaths caused by Costco. The source does not provide enough information to independently check the calculation.
Why does the critique compare nuclear plants with landmarks?
The commentators say they applied a similar proximity-based method to other locations to test whether it produces apparent associations even where a causal connection is not established. They use the results to question the method’s interpretation.
What remains unknown about cancer risk near nuclear plants?
The material does not settle whether nearby residents receive measurable radiation exposure or whether any such exposure changes cancer risk. Distance-based statistical associations alone cannot answer those questions.
Source: hn
Halloween Picks
halloween
As an affiliate, we earn on qualifying purchases.
