A recent letter in Risk Analysis presents an hypothesis concerning radon (progeny) as a possible risk factor for breast cancer." In addition to plateout on synthetic clothes-including brassieres-which the authors discuss, radon is fat soluble and can deliver an internal dose to breast tissue. (" The authors do not discuss the many risk factors for breast cancer including family history, high socioeconomic status, age at pregnancy, and obesity.") These factors can change over time, thus influencing the reported increase in breast cancer. In addition, improved case finding can also be a factor in the reported increase.
The authors also present a cancer map which shows some positive qualitative relationship between breast cancer mortality and EPA radon zones. Mortality is not an adequate indicator of risk for this disease; incidence data are preferred, but are not available for the entire United States.
In order to determine the association between radon levels and breast cancer incidence, we examined data from Iowa, a state with the highest mean radon screening level and the highest percentage of homes in the U.S. with radon screening measurements exceeding the EPA guideline of 148 Bq/ml (4 pCi/L). This amounts to 71% of homes exceeding this guideline as measured in the lowest livable area (unfinished or finished basement or first floor) using a random survey. The methodology of this analysis was similar to that described elsewhere.(4)
Population data for White female residents of each county in the state of Iowa were obtained from the U.S. Census Bureau for the 1980 and 1990 periods. Intercensal estimates were also provided by the census bureau for the years 1973-1979, 1981-1989, and 1991-1993.
Breast cancer cases were identified through the State Health Registry of Iowa, a member of the National Cancer Institute's Surveillance, Epidemiology, and End Results (SEER) Program since 1973. Eligibility was restricted to newly identified in situ and invasive breast cancers diagnosed between 1973 and 1993 among White female residents of the state of Iowa.
Residential radon data were obtained from randomly selected homes whose occupants participated in either a joint U.S. Environmental Protection Agency/ Iowa Department of Public Health radon screening survey conducted in the first half of 1989. All radon measurements were obtained from a single detector placed in the lowest livable area of the home during the winter. Open-faced charcoal canisters were deployed for 2 days at each site. Radon measurements were assigned to a county using the residential address. After counties with fewer than ten radon readings were excluded from the analysis, 77 counties remained out of the original 99.
Age-adjusted breast cancer incidence rates for 1973-1993 were calculated directly for each of the three groupings of counties. The standard population was the 1970 U.S. population divided into 5-year age groups. The 95% confidence interval was then calculated in order to determine statistical significance. The rates were stratified into low (<8 pCi/L), medium (8-10 pCi/L), and high (>10 pCi/L) radon levels. Arithmetic mean values of radon were utilized.
Results are presented in Table 1. A total of 32,171 cases are distributed across 1732 radon readings. Despite an 83% increase in average radon level, there is no increase in invasive breast cancer risk in the highest radon counties of Iowa. None of the results were statistically significant. There was no dose response trend (i.e., no increased risk as radon levels increased). Separate analysis for in situ breast cancer revealed identical conclusions.
Table 1. Age-Adjusted Incidence Rates for Invasive Breast Cancers in 77 Iowa Counties (White females, 1973-1993, stratified by radon levels)
Radon level
Number of counties
Number of radon readings
Average radon reading(a)
Number of breast cancer(b) cases
Breast cancer incidence rates(c)
<8
24
600
6.9
11,357
93.9
8-10
23
545
8.9
9.748
92.2
>10
30
587
12.6
11,066
93.9
(a) pCi/L.
(b) Invasive breast cancer.
(c) Per 100,000 per year. Age adjusted to the 1970 U.S. Standard Population.
It is recognized that descriptive epidemiology is hypothesis-generating and that an analytic design would give more refined data on radon exposure and would include data on other risks factors for breast cancer. An annual average for radon is a preferred measurement compared to a short-term reading (e.g., 2 days). The same comments apply to the letter mentioned above. Regardless, these Iowa data do not support the hypothesis that radon increases the risk of breast cancer.
REFERENCES
- P. M. Kildea, Jr. and T. C. Lee, "Radon Plateout on Synthetic Fibers as a Possible Risk Factor in Breast Cancer" (letter), Risk Analysis 16(l), 1-2 (1996).
- N. H. Harley and E. S. Robbins, "Rn-222 Alpha Dose to Organs Other Than Lung," Radiation Protection Dosimetry 45(1-4 Suppl.), 619-622 (1992).
- J. L. Kelsey, "Breast Cancer Epidemiology: Summary and Future Directions," Epidemiologic- Reviews 15(l), 256-263 (1993).
- J. S. Neuberger et al., "Residential Radon Exposure and Lung Cancer: Evidence of an Urban Factor in Iowa," Health Physics 66(3),263-269 (1994).
John S. Neuberger University of Kansas School of Medicine R William Field University of Iowa College of Medicine Material presented on this home page constitutes opinion of the author.
Copyright © 1997 Steven J. Milloy. All rights reserved. Site developed and hosted by WestLake Solutions, Inc.