ESTIMATING THE ATTRIBUTABLE RISK OF RESIDENTIAL RADON AMONG NONSMOKING WOMEN - INITIAL RESULTS AND METHODOLOGIC CHALLENGES

A population-based, case-control study of incident lung cancer among w omen in Missouri (United States) who were lifetime nonsmokers and long -term exsmokers was conducted between 1986 and 1992. The study include d 538 lung cancer cases and 1183 population-based, age matched control s. Information on lung cancer risk factors was obtained by interviewin g cases, next-of-kin of cases, and controls. Year-long radon measureme nts were also sought in every dwelling occupied for the previous 5-30 y. The Missouri study was among the first studies specifically-designe d to evaluate the effects of residential radon on lung cancer risk. Th e mean radon lever found in homes was 1.6 pCi/L. This level of radon e xposure is somewhat higher than that observed in the United States as a whole (mean 1.25 pCi/L). A small nonsignificant risk was found for s tudy subjects exposed to a median domestic radon concentration of 4 pC i/L (25 y time-weight average). Since only a small fraction of the pop ulation is exposed at this level, it is estimated that the population attributable risk (PAR) for domestic radon is between 1 and 4% in Miss ouri. The growing body of residential studies has not clearly shown an elevation in lung cancer risk due to low-level, long-term radon expos ure. It may be difficult to demonstrate a statistically significant in crease in risk due in part to the inherent methodologic difficulties a ssociated with assessing the potential carcinogenic effect of low-leve l radon exposure, errors in reconstructing past radon exposures, and p opulation mobility which tends to homogenize exposures. A complementar y method of radon exposure measurement, now being used in a second epi demiologic evaluation of lung cancer in Missouri, which uses heirloom grass to assess actual historic cumulative dose, is discussed, and may help strengthen future studies. Copyright (C) 1996 Elsevier Science L td.