Female workers at nuclear-weapons plants - including Savannah River Site - have a lower risk of early death than the female population at large, a study has found.
But the study's architect warned Thursday that a common bias in such studies might have skewed the results. The findings might be influenced by the "healthy worker" effect, said Gregg S. Wilkinson of the State University of New York at Buffalo.
That effect stems from the fact that to hold a job, workers must maintain good health - meaning they would be less likely to die than the general population, Dr. Wilkinson said.
Dr. Wilkinson studied the health records and death certificates of 67,976 white females who worked at 12 nuclear-weapons plants, including SRS, from their start-up until 1980.
The number of minority females was too small to be significant statistically, Dr. Wilkinson said.
Researchers performed the study using the "standard mortality ratio," he said. For a specific cause of death, the ratio compares the number of actual deaths to the number of deaths that normally would be expected.
At a ratio of 100, the actual deaths equal the number of expected deaths, Dr. Wilkinson said. But in nearly every case, the weapons-plant workers had a ratio lower than 100, he said, indicating fewer deaths than expected.
For example, SRS had a ratio of 79 for all causes of death, and a ratio of 73 for deaths caused by cancer. The site had a ratio of 66 for all heart diseases.
For all sites, the ratio was 76 for all causes of death and 76 for all cancer deaths.
The study did find a higher incidence of deaths from mental disorders among the workers studied, with a ratio of 141. Researchers expected 14.8 deaths but found 21, Dr. Wilkinson said.
Among 21,440 women monitored for radiation exposure at 10 plants, the researchers found numbers that suggest a higher risk of cancer.
For example, the numbers indicate a 5 percent increase in the risk of breast cancer for each rem of radiation a worker receives. Rem, short for roentgen equivalent man, is a unit designed to measure radiation exposure in humans.
The study also indicated a 13 percent increase in the risk of leukemia for each rem.
But those figures had wide "confidence intervals," meaning the actual risk probably ranges from equal to above that of the general population, Dr. Wilkinson said.
The study also did not account for other factors that influence workers' health, such as their economic or educational status, whether they smoke, or the area in which they live, Dr. Wilkinson said.
Further studies, already under way, should provide more detailed analyses of health and mortality risks at specific sites, Dr. Wilkinson said.
SRS statistics indicated a risk of leukemia six times higher than the population at large, but the finding was based on one death out of 794 women. Dr. Wilkinson called the statistic "unstable."
Reach Brandon Haddock at (706) 823-3409.