<< Previous

Pilgrim Study. In 1990, the Massachusetts Department of Public Health (MDPH) published a study (the “Southeastern Massachusetts Health Study”) of leukemia incidence for 22 towns in the southeastern area of the state. The purpose of the study was to determine if the incidence of leukemia could be associated with exposure to radiation from the Pilgrim nuclear power plant. The population studied consisted of people aged 13 years and older who were diagnosed between 1978 and 1986 with any type of leukemia, excluding one type known not to be associated with radiation.

The report’s findings included the following:

• Individuals with the highest potential for exposure to radiation emissions from Pilgrim (i.e., those who lived and/or worked the longest and closest to the plant) had almost four times the incidence of leukemia as those having the lowest potential for exposure (i.e., those who lived and/or worked the least amount of time and farthest from the plant).
• The study found an association between radiation released from the Pilgrim plant and leukemia incidence only among those cases diagnosed before 1984.
• No apparent relationship with the plant was observed for cases diagnosed between 1984 and 1986.

A review of the study, released in October 1992, found serious problems with the study’s methods and conclusions. The 16-month-long analysis was the work of an independent review panel, composed of six experts in epidemiology, appointed by MDPH and Boston Edison Co., which owned the Pilgrim plant.

Among the most serious flaws in methodology, according to the panel, were the following:

• It was very inconsistent with other radiation studies. There was a large disparity between the number of excess leukemia cases reported by the study (47) and the number expected using data from other radiation studies (0.52).
• The study failed to document, from vital records, any excess leukemia deaths during the study period, compared with leukemia mortality before the Pilgrim plant opened.
• The study failed to include towns on Cape Cod that were within the study area.
• In estimating how much radiation exposure that people living near the plant received, the study should have used alternative models of how radiation is dispersed.

The panel called for “a carefully designed new study” to address the concerns raised in its report.

Greenpeace Study. Greenpeace and Ernest Sternglass, an anti-nuclear activist, released “Nuclear Power, Human Health and the Environment: The Breast Cancer Warning in the Great Lakes Basin” in 1995. The study claims that women in 81 counties in the Great Lakes region, where there are 36 U.S. and Canadian nuclear power plants, have an increased risk of breast cancer mortality. It also claims the 1990 National Cancer Institute study and other studies that found no cancer-nuclear power connection failed to look at a sufficiently broad radius around the plants.

The study’s findings and methodology have drawn widespread criticism among scientists and in the news media:

• It provides no evidence that women in the 81 Great Lakes counties live closer to nuclear power plants, or that women were exposed to significantly higher levels of radiation, than women in nearby counties that Greenpeace did not choose to study.
• The study offers no detail on important characteristics of the women in those 81 counties, such as urbanization, ethnicity or socioeconomic profile, which would help evaluate whether “selection bias” is present. (The risk of dying from breast cancer is higher in urban areas and among certain ethnic groups.)
• Results can depend on the method that researchers use to compare data. Greenpeace chose to combine the data for all women in all 81 counties and compare the total with the U.S. average. The result was 3.2 excess cancer deaths per 100,000 women—an extremely small increase. But if Greenpeace had looked at data from each of the 81 Great Lakes counties individually, it would have found something different: In slightly more than half of the counties, the breast cancer death rates are somewhat lower than the U.S. average, and in slightly less than half of them, the death rates are somewhat higher than the U.S. average. With this method, there is no consistent increase in the breast cancer death rates in all of the 81 counties.
• The reason the NCI did not extend its study to a 100-mile radius around each plant—as Greenpeace claims was necessary—is that radioactive emissions from nuclear power plants are virtually nonexistent at that radius. The nearest plant neighbor gets less than one millirem of radiation exposure from the plant annually. This is less than the average person gets annually from watching television.