Study of indoor radon and lung cancer in France
National Radiation Environment (NRE - VII), 20-24 May, 2002, Rhodes, Grèce
H. Baysson, F. Jourdain, D. Laurier, F. Ducloy, M. Tirmarche
This paper presents the preliminary results of the French case-control study of indoor radon and lung cancer.The study was carried out in hospitals of 5 regions in France: Auvergne, Ardennes, Brittany, Languedoc and Limousin. Subjects with primary lung cancer and aged <75 years on their date of diagnosis were interviewed within 6 months of diagnosis. Subjects were eligible if they had lived in the study area for at least 25 out of the last 35 years. For each case, 2 controls were identified from the pool of hospital patients of the same region, and were individually matched to cases by sex and age (± 5 years). Each hospital control was a person who satisfied the study residence requirements and who was not hospitalised for a disease strongly related to tobacco. For each subject, a face-to-face interview conducted during hospitalisation allowed to collect information about residential history, past tobacco consumption, potential exposure to lung carcinogens during occupational life, and socio-economic status. For each dwelling occupied by study subjects during the last 35 years, two Kodalpha LR 115 radon detectors were installed during 6 months, one in the living room area and another in the bedroom. Cumulated radon exposure over the period of 5 to 35 years before the interview was calculated from these measurements, taking into account the period of occupation in each dwelling.
A total of 688 cases and 1428 controls were included in the study. Ninety percent of the cases and controls were men. The percentage of ever smokers was of 93% among cases and of 73% among controls. On average, 2.2 dwellings have been occupied by each subject during the last 35 years. Of a total number of 4523 dwellings, radon concentration has been measured in 2714 dwellings (60%).
The arithmetic mean of radon levels in these dwellings was 130 Bq/m3 among cases, and 127 Bq/m3 among controls. The lowest measured concentration was 6 Bq/m3 while the highest was 4991 Bq/m3. The number of years covered by measurements was 20, among cases and controls, corresponding to 66% of the total period of 5 to 35 years. An analysis of the distribution of missing radon measurements showed that they were associated mainly with the study region, the delay since occupation, and the number of dwellings occupied. Other associations between the number of years with missing radon measurement, tobacco consumption and the radon concentration level were observed.
In order to limit the probability of bias, the calculation of the cumulative exposure to radon has been limited to a period of 20 years (5 to 25 years before the interview). The percentage of missing data was then reduced to 10%, among cases and controls. Remaining missing data have been replaced by the mean radon concentration among controls from the same region and with the same dwelling characteristics. After adjustment on sex, age, region, smoking and past occupational exposure, the estimated relative risk per 100 Bq/m3 was 1.05, at the borderline of statistical significance (95% confidence interval = [0.99 – 1.12]).
One of the main limits of indoor radon case-control studies is the retrospective reconstruction of past exposure to radon. A detailed analysis of the missing data is therefore warranted to evaluate the impact on the results and the risk of bias. The relative risk estimated in this study is compatible with those derived from previous indoor radon case-control studies and also from studies of underground miners. Nevertheless, the estimated risk linked to indoor radon exposure is small, especially in comparison to the risk associated to smoking. The French study is included in a European project, which aims a joint analysis of all case-control studies in Europe. This very large-scale analysis will provide the statistical power required to evaluate the effect of potential modifying factors or the interaction between radon and smoking.