RESIDENTIAL RADON EXPOSURE AND LUNG-CANCER AMONG NONSMOKING WOMEN

Background: Radon at sufficiently high concentrations is known to caus e lung cancer among underground miners and in experimental laboratory animals. Purpose: Our aim was to determine whether indoor levels of ra don are associated with a detectable increase in lung cancer. Nonsmoki ng women were selected because they offer the best opportunity to dete ct radon-related risk while minimizing the potentially confounding inf luences of cigarette smoking and occupation. Methods: A population-bas ed, case-control study of incident lung cancer was conducted in Missou ri. A total of 538 nonsmoking white women diagnosed with lung cancer b etween 1986 and 1992 and 1183 age-matched control subjects were identi fied from the Missouri Cancer Registry and from driver's license and M edicare listings, respectively. Information on lung cancer risk factor s was obtained by telephone interview. Year-long radon measurements we re sought in every dwelling occupied for the previous 5-30 years. Resu lts: Radon measurements covered 78% of the relevant residential period , and women reported being indoors for 84% of this time. The time-weig hted average radon concentrations were exactly the same for case subje cts and control subjects (1.82 pCi/L of air [pCi L(-1)]). Radon levels greater than 4 pCi L(-1) were experienced by 6.5% of the case subject s and 6.8% of the control subjects. For all data combined, there was l ittle evidence for a trend of lung cancer with increasing radon concen trations (two-tailed trend test, P = .99 continuous data analysis; P = .19 categorical data analysis). A positive dose-response trend was su ggested for the adenocarcinoma cell type and among directly interviewe d women (two-tailed trend test; P = .31 continuous data analysis; P = .04 categorical data analysis), but not for other histologies or among those who had surrogate interviews. Conclusions: The possibility of d etecting a risk from indoor radon in this study was maximized by (a) i ncluding a large number of nonsmoking women with high indoor occupancy , (b) conducting a large number of radon measurements near the time of the diagnosis of cancer, and (c) controlling for known causes of lung cancer. However, an association between lung cancer and the exposure to domestic levels of radon was not convincingly demonstrated. Implica tions: The magnitude of the lung cancer risk from radon levels commonl y found in U.S. dwellings appears low.