This association of SNPs on 15q25 with lung cancer risk has been suggested to be mediated through changes in propensity to smoke tobacco(13,18). In the present study the estimated risk effect of 15q25 SNPs was attenuated to varying degrees when controlling for various smoking variables, including CPD, duration of smoking and cotinine levels. In this study adjustment for regression-dilution bias corrected cotinine in current smokers led to attenuation of the rs16969968 OR, thus supporting the hypothesis of the 15q25 association with lung cancer risk being mediated by smoking behavior. However, the regression-dilution method is not perfect and relies on several assumptions that may not hold. Firstly, the correction is estimated using measurements taken 3 years apart and assuming a constant mean rate. The issue is further complicated by our incomplete understanding of the relation between life-course smoking and lung cancer risk: a life-time mean may not be the ideal predictor even if we were able to estimate it with precision. In addition, our estimates of regression dilution were obtained from a distinct population, geographically unrepresentative of the EPIC Lung study,
