ANS National
 Meetings/
 Sessions

November 1994

John Cameron

(University of Wisconsin)

4. What Does the Nuclear Shipyard Worker Study Tell Us?

                "The Nuclear Shipyard Workers Study¹ (NSWS) was designed to determine whether there was an excess risk of leukemia or other cancers associated with exposure to low levels of gamma radiation. The study was conducted at a cost of more than $10 million through a contract from the U.S. Department of Energy (DOE) to Matanoski at Johns Hopkins University, School of Hygiene and Public Health. The study was initiated in 1978. The study was completed in 1987. The results have never been published in the scientific literature. In 1991 the DOE made the 437-page study available to the public with a 2-page press release. In the debate on the health effects of low-level radiation, this study, which is perhaps the largest and best of its kind, with better dosimetry than most studies of large exposed populations, is largely ignored. It is fair to assume that if the result of this study had indicated a statistically significant 24% INCREASE in cancer mortality among the radiation-exposed shipyard workers, instead of a 24% DECREASE, there would have been extreme media and political and scientific attention (or perhaps even for a 2 to 4% statistically insignificant increase).

                "An NSWS technical advisory panel (TAP) chaired by Upton reviewed results and advised on the research. The author of this paper was a member of the TAP and of the NSWS Radiation Dosimetry Advisory Committee.

                "The study groups were selected from 3 database of almost 700,000 shipyard workers, including about 108,000 nuclear workers. The three study groups consisted of 28,542 nuclear workers with 3 working lifetime dose equivalent (DE) equal to or greater than 5 mSv (0.5 rem), referred to here as NW>5; 10,462 nuclear workers with (1 working lifetime DE <5 mSv, referred to here as NW <5; and 33,352 nonnuclear workers, referred to as NNW. Five mSv is the amount of radiation a typical person receives from natural background radiation in ~20 months.

                "All three study groups were balanced in the initial sample to provide comparability on basic demographic characteristics to make between-group comparisons appropriate. The study included exposures received from the beginning of nuclear ship overhauls in the 1960s until the end of 1981.

                "Deaths in each of the groups were classified as due to all causes, leukemia, lymphatic and hematopoietic cancers (LHC), mesothelioma, and lung cancer. Mesothelioma was the only cancer that showed a significantly higher incidence for all groups. Mesothelioma deaths are considered a marker For asbestos exposure. There were only 36 mesothelioma deaths during the entire study of nearly a million person-years (in 70,730 workers).

                "Asbestos is also associated with lung cancer. The slightly higher incidence of lung cancer for all three groups compared to the general population may have been due to asbestos exposure.

                "Both nuclear worker groups had a lower death rate from leukemia and lymphatic and hematopoietic cancers than the nonnuclear group. All three groups had lower LHC death rates than the general population. Table I summarizes the data.

                "The most significant and surprising finding of the NSWS research was that the nuclear workers with the greatest radiation exposure, a cumulative lifetime occupational dose equivalent of 5 mSv or more, had a standardized mortality rate (SMR) of deaths from all causes of only 0.76 that for their age and sex in the general population, while the nonnuclear workers had an SMR of 1.0. The standard deviation of the SMR was ~0.015; i.e., the mortality rate for the nuclear workers was ~16 standard deviations below that of the nonnuclear worker group!

                "The occupational exposure to the nuclear shipyard workers is comparable to the cumulated effective dose equivalent they received From natural radiation. Their total radiation, occupational plus natural, is comparable to natural radiation exposures in some parts of the world.

                "This study is probably the best scientific evidence, of many scientific data sources, to show that low levels of ionizing radiation exposure are without health hazard. The results clearly contradict the conclusions of BEIR that even small amounts of radiation have risk (in BEIR V and earlier reports), which have been largely based on the data from the Japanese atomic bomb survivors, who largely received their radiation exposures in very brief, high dose rate conditions and who are also now demonstrating that effective radiation health effects thresholds exist in the range of 20 to 200 rem."

1. "Health Effects of Low-Level Radiation in Shipyard Workers," U.S. Department of Energy Final Report (June 1991).